diff options
Diffstat (limited to 'fs')
503 files changed, 23401 insertions, 16030 deletions
diff --git a/fs/9p/Kconfig b/fs/9p/Kconfig index 814ac4e213a8..0a93dc1cb4ac 100644 --- a/fs/9p/Kconfig +++ b/fs/9p/Kconfig @@ -1,6 +1,6 @@ config 9P_FS - tristate "Plan 9 Resource Sharing Support (9P2000) (Experimental)" - depends on INET && NET_9P && EXPERIMENTAL + tristate "Plan 9 Resource Sharing Support (9P2000)" + depends on INET && NET_9P help If you say Y here, you will get experimental support for Plan 9 resource sharing via the 9P2000 protocol. @@ -10,7 +10,6 @@ config 9P_FS If unsure, say N. if 9P_FS - config 9P_FSCACHE bool "Enable 9P client caching support (EXPERIMENTAL)" depends on EXPERIMENTAL diff --git a/fs/9p/fid.c b/fs/9p/fid.c index 0ee594569dcc..85b67ffa2a43 100644 --- a/fs/9p/fid.c +++ b/fs/9p/fid.c @@ -286,11 +286,9 @@ static struct p9_fid *v9fs_fid_clone_with_uid(struct dentry *dentry, uid_t uid) struct p9_fid *v9fs_writeback_fid(struct dentry *dentry) { - int err, flags; + int err; struct p9_fid *fid; - struct v9fs_session_info *v9ses; - v9ses = v9fs_dentry2v9ses(dentry); fid = v9fs_fid_clone_with_uid(dentry, 0); if (IS_ERR(fid)) goto error_out; @@ -299,17 +297,8 @@ struct p9_fid *v9fs_writeback_fid(struct dentry *dentry) * dirty pages. We always request for the open fid in read-write * mode so that a partial page write which result in page * read can work. - * - * we don't have a tsyncfs operation for older version - * of protocol. So make sure the write back fid is - * opened in O_SYNC mode. */ - if (!v9fs_proto_dotl(v9ses)) - flags = O_RDWR | O_SYNC; - else - flags = O_RDWR; - - err = p9_client_open(fid, flags); + err = p9_client_open(fid, O_RDWR); if (err < 0) { p9_client_clunk(fid); fid = ERR_PTR(err); diff --git a/fs/9p/v9fs.h b/fs/9p/v9fs.h index 9665c2b840e6..e5ebedfc5ed8 100644 --- a/fs/9p/v9fs.h +++ b/fs/9p/v9fs.h @@ -116,7 +116,6 @@ struct v9fs_session_info { struct list_head slist; /* list of sessions registered with v9fs */ struct backing_dev_info bdi; struct rw_semaphore rename_sem; - struct p9_fid *root_fid; /* Used for file system sync */ }; /* cache_validity flags */ diff --git a/fs/9p/vfs_dentry.c b/fs/9p/vfs_dentry.c index b6a3b9f7fe4d..e022890c6f40 100644 --- a/fs/9p/vfs_dentry.c +++ b/fs/9p/vfs_dentry.c @@ -126,7 +126,9 @@ static int v9fs_lookup_revalidate(struct dentry *dentry, struct nameidata *nd) retval = v9fs_refresh_inode_dotl(fid, inode); else retval = v9fs_refresh_inode(fid, inode); - if (retval <= 0) + if (retval == -ENOENT) + return 0; + if (retval < 0) return retval; } out_valid: diff --git a/fs/9p/vfs_inode_dotl.c b/fs/9p/vfs_inode_dotl.c index ffbb113d5f33..691c78f58bef 100644 --- a/fs/9p/vfs_inode_dotl.c +++ b/fs/9p/vfs_inode_dotl.c @@ -259,7 +259,7 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, if (IS_ERR(inode_fid)) { err = PTR_ERR(inode_fid); mutex_unlock(&v9inode->v_mutex); - goto error; + goto err_clunk_old_fid; } v9inode->writeback_fid = (void *) inode_fid; } @@ -267,8 +267,8 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, /* Since we are opening a file, assign the open fid to the file */ filp = lookup_instantiate_filp(nd, dentry, generic_file_open); if (IS_ERR(filp)) { - p9_client_clunk(ofid); - return PTR_ERR(filp); + err = PTR_ERR(filp); + goto err_clunk_old_fid; } filp->private_data = ofid; #ifdef CONFIG_9P_FSCACHE @@ -278,10 +278,11 @@ v9fs_vfs_create_dotl(struct inode *dir, struct dentry *dentry, int omode, return 0; error: - if (ofid) - p9_client_clunk(ofid); if (fid) p9_client_clunk(fid); +err_clunk_old_fid: + if (ofid) + p9_client_clunk(ofid); return err; } @@ -811,7 +812,7 @@ v9fs_vfs_follow_link_dotl(struct dentry *dentry, struct nameidata *nd) fid = v9fs_fid_lookup(dentry); if (IS_ERR(fid)) { __putname(link); - link = ERR_PTR(PTR_ERR(fid)); + link = ERR_CAST(fid); goto ndset; } retval = p9_client_readlink(fid, &target); diff --git a/fs/9p/vfs_super.c b/fs/9p/vfs_super.c index f3eed3383e4f..feef6cdc1fd2 100644 --- a/fs/9p/vfs_super.c +++ b/fs/9p/vfs_super.c @@ -154,6 +154,7 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, retval = PTR_ERR(inode); goto release_sb; } + root = d_alloc_root(inode); if (!root) { iput(inode); @@ -185,21 +186,10 @@ static struct dentry *v9fs_mount(struct file_system_type *fs_type, int flags, p9stat_free(st); kfree(st); } - v9fs_fid_add(root, fid); retval = v9fs_get_acl(inode, fid); if (retval) goto release_sb; - /* - * Add the root fid to session info. This is used - * for file system sync. We want a cloned fid here - * so that we can do a sync_filesystem after a - * shrink_dcache_for_umount - */ - v9ses->root_fid = v9fs_fid_clone(root); - if (IS_ERR(v9ses->root_fid)) { - retval = PTR_ERR(v9ses->root_fid); - goto release_sb; - } + v9fs_fid_add(root, fid); P9_DPRINTK(P9_DEBUG_VFS, " simple set mount, return 0\n"); return dget(sb->s_root); @@ -210,11 +200,15 @@ close_session: v9fs_session_close(v9ses); kfree(v9ses); return ERR_PTR(retval); + release_sb: /* - * we will do the session_close and root dentry - * release in the below call. + * we will do the session_close and root dentry release + * in the below call. But we need to clunk fid, because we haven't + * attached the fid to dentry so it won't get clunked + * automatically. */ + p9_client_clunk(fid); deactivate_locked_super(sb); return ERR_PTR(retval); } @@ -232,7 +226,7 @@ static void v9fs_kill_super(struct super_block *s) P9_DPRINTK(P9_DEBUG_VFS, " %p\n", s); kill_anon_super(s); - p9_client_clunk(v9ses->root_fid); + v9fs_session_cancel(v9ses); v9fs_session_close(v9ses); kfree(v9ses); @@ -285,14 +279,6 @@ done: return res; } -static int v9fs_sync_fs(struct super_block *sb, int wait) -{ - struct v9fs_session_info *v9ses = sb->s_fs_info; - - P9_DPRINTK(P9_DEBUG_VFS, "v9fs_sync_fs: super_block %p\n", sb); - return p9_client_sync_fs(v9ses->root_fid); -} - static int v9fs_drop_inode(struct inode *inode) { struct v9fs_session_info *v9ses; @@ -307,6 +293,51 @@ static int v9fs_drop_inode(struct inode *inode) return 1; } +static int v9fs_write_inode(struct inode *inode, + struct writeback_control *wbc) +{ + int ret; + struct p9_wstat wstat; + struct v9fs_inode *v9inode; + /* + * send an fsync request to server irrespective of + * wbc->sync_mode. + */ + P9_DPRINTK(P9_DEBUG_VFS, "%s: inode %p\n", __func__, inode); + v9inode = V9FS_I(inode); + if (!v9inode->writeback_fid) + return 0; + v9fs_blank_wstat(&wstat); + + ret = p9_client_wstat(v9inode->writeback_fid, &wstat); + if (ret < 0) { + __mark_inode_dirty(inode, I_DIRTY_DATASYNC); + return ret; + } + return 0; +} + +static int v9fs_write_inode_dotl(struct inode *inode, + struct writeback_control *wbc) +{ + int ret; + struct v9fs_inode *v9inode; + /* + * send an fsync request to server irrespective of + * wbc->sync_mode. + */ + P9_DPRINTK(P9_DEBUG_VFS, "%s: inode %p\n", __func__, inode); + v9inode = V9FS_I(inode); + if (!v9inode->writeback_fid) + return 0; + ret = p9_client_fsync(v9inode->writeback_fid, 0); + if (ret < 0) { + __mark_inode_dirty(inode, I_DIRTY_DATASYNC); + return ret; + } + return 0; +} + static const struct super_operations v9fs_super_ops = { .alloc_inode = v9fs_alloc_inode, .destroy_inode = v9fs_destroy_inode, @@ -314,17 +345,18 @@ static const struct super_operations v9fs_super_ops = { .evict_inode = v9fs_evict_inode, .show_options = generic_show_options, .umount_begin = v9fs_umount_begin, + .write_inode = v9fs_write_inode, }; static const struct super_operations v9fs_super_ops_dotl = { .alloc_inode = v9fs_alloc_inode, .destroy_inode = v9fs_destroy_inode, - .sync_fs = v9fs_sync_fs, .statfs = v9fs_statfs, .drop_inode = v9fs_drop_inode, .evict_inode = v9fs_evict_inode, .show_options = generic_show_options, .umount_begin = v9fs_umount_begin, + .write_inode = v9fs_write_inode_dotl, }; struct file_system_type v9fs_fs_type = { diff --git a/fs/Kconfig b/fs/Kconfig index f3aa9b08b228..19891aab9c6e 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -47,7 +47,7 @@ config FS_POSIX_ACL def_bool n config EXPORTFS - bool + tristate config FILE_LOCKING bool "Enable POSIX file locking API" if EXPERT @@ -124,6 +124,7 @@ config TMPFS config TMPFS_POSIX_ACL bool "Tmpfs POSIX Access Control Lists" depends on TMPFS + select TMPFS_XATTR select GENERIC_ACL help POSIX Access Control Lists (ACLs) support permissions for users and @@ -134,6 +135,22 @@ config TMPFS_POSIX_ACL If you don't know what Access Control Lists are, say N. +config TMPFS_XATTR + bool "Tmpfs extended attributes" + depends on TMPFS + default n + help + Extended attributes are name:value pairs associated with inodes by + the kernel or by users (see the attr(5) manual page, or visit + <http://acl.bestbits.at/> for details). + + Currently this enables support for the trusted.* and + security.* namespaces. + + You need this for POSIX ACL support on tmpfs. + + If unsure, say N. + config HUGETLBFS bool "HugeTLB file system support" depends on X86 || IA64 || SPARC64 || (S390 && 64BIT) || \ diff --git a/fs/adfs/map.c b/fs/adfs/map.c index d1a5932bb0f1..6935f05202ac 100644 --- a/fs/adfs/map.c +++ b/fs/adfs/map.c @@ -51,7 +51,7 @@ static DEFINE_RWLOCK(adfs_map_lock); /* * This is fun. We need to load up to 19 bits from the map at an - * arbitary bit alignment. (We're limited to 19 bits by F+ version 2). + * arbitrary bit alignment. (We're limited to 19 bits by F+ version 2). */ #define GET_FRAG_ID(_map,_start,_idmask) \ ({ \ diff --git a/fs/afs/cache.c b/fs/afs/cache.c index 0fb315dd4d2a..577763c3d88b 100644 --- a/fs/afs/cache.c +++ b/fs/afs/cache.c @@ -98,7 +98,7 @@ static uint16_t afs_cell_cache_get_key(const void *cookie_netfs_data, } /* - * provide new auxilliary cache data + * provide new auxiliary cache data */ static uint16_t afs_cell_cache_get_aux(const void *cookie_netfs_data, void *buffer, uint16_t bufmax) @@ -117,7 +117,7 @@ static uint16_t afs_cell_cache_get_aux(const void *cookie_netfs_data, } /* - * check that the auxilliary data indicates that the entry is still valid + * check that the auxiliary data indicates that the entry is still valid */ static enum fscache_checkaux afs_cell_cache_check_aux(void *cookie_netfs_data, const void *buffer, @@ -150,7 +150,7 @@ static uint16_t afs_vlocation_cache_get_key(const void *cookie_netfs_data, } /* - * provide new auxilliary cache data + * provide new auxiliary cache data */ static uint16_t afs_vlocation_cache_get_aux(const void *cookie_netfs_data, void *buffer, uint16_t bufmax) @@ -172,7 +172,7 @@ static uint16_t afs_vlocation_cache_get_aux(const void *cookie_netfs_data, } /* - * check that the auxilliary data indicates that the entry is still valid + * check that the auxiliary data indicates that the entry is still valid */ static enum fscache_checkaux afs_vlocation_cache_check_aux(void *cookie_netfs_data, @@ -283,7 +283,7 @@ static void afs_vnode_cache_get_attr(const void *cookie_netfs_data, } /* - * provide new auxilliary cache data + * provide new auxiliary cache data */ static uint16_t afs_vnode_cache_get_aux(const void *cookie_netfs_data, void *buffer, uint16_t bufmax) @@ -309,7 +309,7 @@ static uint16_t afs_vnode_cache_get_aux(const void *cookie_netfs_data, } /* - * check that the auxilliary data indicates that the entry is still valid + * check that the auxiliary data indicates that the entry is still valid */ static enum fscache_checkaux afs_vnode_cache_check_aux(void *cookie_netfs_data, const void *buffer, diff --git a/fs/afs/cell.c b/fs/afs/cell.c index 0d5eeadf6121..3c090b7555ea 100644 --- a/fs/afs/cell.c +++ b/fs/afs/cell.c @@ -293,7 +293,7 @@ struct afs_cell *afs_cell_lookup(const char *name, unsigned namesz, if (!cell) { /* this should not happen unless user tries to mount * when root cell is not set. Return an impossibly - * bizzare errno to alert the user. Things like + * bizarre errno to alert the user. Things like * ENOENT might be "more appropriate" but they happen * for other reasons. */ diff --git a/fs/attr.c b/fs/attr.c index 1007ed616314..caf2aa521e2b 100644 --- a/fs/attr.c +++ b/fs/attr.c @@ -128,7 +128,7 @@ EXPORT_SYMBOL(inode_newsize_ok); * setattr_copy must be called with i_mutex held. * * setattr_copy updates the inode's metadata with that specified - * in attr. Noticably missing is inode size update, which is more complex + * in attr. Noticeably missing is inode size update, which is more complex * as it requires pagecache updates. * * The inode is not marked as dirty after this operation. The rationale is @@ -175,6 +175,13 @@ int notify_change(struct dentry * dentry, struct iattr * attr) return -EPERM; } + if ((ia_valid & ATTR_MODE)) { + mode_t amode = attr->ia_mode; + /* Flag setting protected by i_mutex */ + if (is_sxid(amode)) + inode->i_flags &= ~S_NOSEC; + } + now = current_fs_time(inode->i_sb); attr->ia_ctime = now; diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index 96804a17bbd0..87d95a8cddbc 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -583,6 +583,8 @@ static int autofs4_dir_unlink(struct inode *dir, struct dentry *dentry) if (!autofs4_oz_mode(sbi) && !capable(CAP_SYS_ADMIN)) return -EACCES; + dentry_unhash(dentry); + if (atomic_dec_and_test(&ino->count)) { p_ino = autofs4_dentry_ino(dentry->d_parent); if (p_ino && dentry->d_parent != dentry) @@ -612,7 +614,7 @@ static int autofs4_dir_unlink(struct inode *dir, struct dentry *dentry) * set the DMANAGED_AUTOMOUNT and DMANAGED_TRANSIT flags on the leaves * of the directory tree. There is no need to clear the automount flag * following a mount or restore it after an expire because these mounts - * are always covered. However, it is neccessary to ensure that these + * are always covered. However, it is necessary to ensure that these * flags are clear on non-empty directories to avoid unnecessary calls * during path walks. */ diff --git a/fs/befs/ChangeLog b/fs/befs/ChangeLog index ce8c787916be..75a461cfaca6 100644 --- a/fs/befs/ChangeLog +++ b/fs/befs/ChangeLog @@ -24,7 +24,7 @@ Version 0.9 (2002-03-14) Version 0.64 (2002-02-07) ========== -* Did the string comparision really right this time (btree.c) [WD] +* Did the string comparison really right this time (btree.c) [WD] * Fixed up some places where I assumed that a long int could hold a pointer value. (btree.c) [WD] @@ -114,7 +114,7 @@ Version 0.6 (2001-12-15) More flexible. Will soon be controllable at mount time (see TODO). [WD] -* Rewrote datastream positon lookups. +* Rewrote datastream position lookups. (datastream.c) [WD] * Moved the TODO list to its own file. @@ -150,7 +150,7 @@ Version 0.50 (2001-11-13) * Anton also told me that the blocksize is not allowed to be larger than the page size in linux, which is 4k i386. Oops. Added a test for (blocksize > PAGE_SIZE), and refuse to mount in that case. What this - practicaly means is that 8k blocksize volumes won't work without a major + practically means is that 8k blocksize volumes won't work without a major restructuring of the driver (or an alpha or other 64bit hardware). [WD] * Cleaned up the befs_count_blocks() function. Much smarter now. @@ -183,7 +183,7 @@ Version 0.45 (2001-10-29) structures into the generic pointer fields of the public structures with kmalloc(). put_super and put_inode free them. This allows us not to have to touch the definitions of the public structures in - include/linux/fs.h. Also, befs_inode_info is huge (becuase of the + include/linux/fs.h. Also, befs_inode_info is huge (because of the symlink string). (super.c, inode.c, befs_fs.h) [WD] * Fixed a thinko that was corrupting file reads after the first block_run @@ -404,7 +404,7 @@ Version 0.4 (2001-10-28) * Fixed compile errors on 2.4.1 kernel (WD) Resolve rejected patches - Accomodate changed NLS interface (util.h) + Accommodate changed NLS interface (util.h) Needed to include <linux/slab.h> in most files Makefile changes fs/Config.in changes diff --git a/fs/befs/befs_fs_types.h b/fs/befs/befs_fs_types.h index 7893eaa1e58c..eb557d9dc8be 100644 --- a/fs/befs/befs_fs_types.h +++ b/fs/befs/befs_fs_types.h @@ -234,7 +234,7 @@ typedef struct { } PACKED befs_btree_super; /* - * Header stucture of each btree node + * Header structure of each btree node */ typedef struct { fs64 left; diff --git a/fs/befs/btree.c b/fs/befs/btree.c index 4202db7496cb..a66c9b1136e0 100644 --- a/fs/befs/btree.c +++ b/fs/befs/btree.c @@ -5,7 +5,7 @@ * * Licensed under the GNU GPL. See the file COPYING for details. * - * 2002-02-05: Sergey S. Kostyliov added binary search withing + * 2002-02-05: Sergey S. Kostyliov added binary search within * btree nodes. * * Many thanks to: diff --git a/fs/befs/linuxvfs.c b/fs/befs/linuxvfs.c index 06457ed8f3e7..54b8c28bebc8 100644 --- a/fs/befs/linuxvfs.c +++ b/fs/befs/linuxvfs.c @@ -734,7 +734,7 @@ parse_options(char *options, befs_mount_options * opts) /* This function has the responsibiltiy of getting the * filesystem ready for unmounting. - * Basicly, we free everything that we allocated in + * Basically, we free everything that we allocated in * befs_read_inode */ static void diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index f34078d702d3..303983fabfd6 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -941,9 +941,13 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) current->mm->start_stack = bprm->p; #ifdef arch_randomize_brk - if ((current->flags & PF_RANDOMIZE) && (randomize_va_space > 1)) + if ((current->flags & PF_RANDOMIZE) && (randomize_va_space > 1)) { current->mm->brk = current->mm->start_brk = arch_randomize_brk(current->mm); +#ifdef CONFIG_COMPAT_BRK + current->brk_randomized = 1; +#endif + } #endif if (current->personality & MMAP_PAGE_ZERO) { diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index 811384bec8de..1bffbe0ed778 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -717,7 +717,7 @@ static int load_flat_file(struct linux_binprm * bprm, * help simplify all this mumbo jumbo * * We've got two different sections of relocation entries. - * The first is the GOT which resides at the begining of the data segment + * The first is the GOT which resides at the beginning of the data segment * and is terminated with a -1. This one can be relocated in place. * The second is the extra relocation entries tacked after the image's * data segment. These require a little more processing as the entry is @@ -820,6 +820,8 @@ static int load_flat_shared_library(int id, struct lib_info *libs) int res; char buf[16]; + memset(&bprm, 0, sizeof(bprm)); + /* Create the file name */ sprintf(buf, "/lib/lib%d.so", id); @@ -835,6 +837,12 @@ static int load_flat_shared_library(int id, struct lib_info *libs) if (!bprm.cred) goto out; + /* We don't really care about recalculating credentials at this point + * as we're past the point of no return and are dealing with shared + * libraries. + */ + bprm.cred_prepared = 1; + res = prepare_binprm(&bprm); if (!IS_ERR_VALUE(res)) @@ -638,10 +638,11 @@ static int __bio_add_page(struct request_queue *q, struct bio *bio, struct page * @offset: vec entry offset * * Attempt to add a page to the bio_vec maplist. This can fail for a - * number of reasons, such as the bio being full or target block - * device limitations. The target block device must allow bio's - * smaller than PAGE_SIZE, so it is always possible to add a single - * page to an empty bio. This should only be used by REQ_PC bios. + * number of reasons, such as the bio being full or target block device + * limitations. The target block device must allow bio's up to PAGE_SIZE, + * so it is always possible to add a single page to an empty bio. + * + * This should only be used by REQ_PC bios. */ int bio_add_pc_page(struct request_queue *q, struct bio *bio, struct page *page, unsigned int len, unsigned int offset) @@ -659,10 +660,9 @@ EXPORT_SYMBOL(bio_add_pc_page); * @offset: vec entry offset * * Attempt to add a page to the bio_vec maplist. This can fail for a - * number of reasons, such as the bio being full or target block - * device limitations. The target block device must allow bio's - * smaller than PAGE_SIZE, so it is always possible to add a single - * page to an empty bio. + * number of reasons, such as the bio being full or target block device + * limitations. The target block device must allow bio's up to PAGE_SIZE, + * so it is always possible to add a single page to an empty bio. */ int bio_add_page(struct bio *bio, struct page *page, unsigned int len, unsigned int offset) @@ -1436,7 +1436,7 @@ EXPORT_SYMBOL(bio_flush_dcache_pages); * preferred way to end I/O on a bio, it takes care of clearing * BIO_UPTODATE on error. @error is 0 on success, and and one of the * established -Exxxx (-EIO, for instance) error values in case - * something went wrong. Noone should call bi_end_io() directly on a + * something went wrong. No one should call bi_end_io() directly on a * bio unless they own it and thus know that it has an end_io * function. **/ diff --git a/fs/block_dev.c b/fs/block_dev.c index c1511c674f53..1f2b19978333 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -653,7 +653,7 @@ void bd_forget(struct inode *inode) * @whole: whole block device containing @bdev, may equal @bdev * @holder: holder trying to claim @bdev * - * Test whther @bdev can be claimed by @holder. + * Test whether @bdev can be claimed by @holder. * * CONTEXT: * spin_lock(&bdev_lock). @@ -1102,6 +1102,7 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) if (!bdev->bd_part) goto out_clear; + ret = 0; if (disk->fops->open) { ret = disk->fops->open(bdev, mode); if (ret == -ERESTARTSYS) { @@ -1118,18 +1119,26 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) put_disk(disk); goto restart; } - if (ret) - goto out_clear; } - if (!bdev->bd_openers) { + + if (!ret && !bdev->bd_openers) { bd_set_size(bdev,(loff_t)get_capacity(disk)<<9); bdi = blk_get_backing_dev_info(bdev); if (bdi == NULL) bdi = &default_backing_dev_info; bdev_inode_switch_bdi(bdev->bd_inode, bdi); } - if (bdev->bd_invalidated) + + /* + * If the device is invalidated, rescan partition + * if open succeeded or failed with -ENOMEDIUM. + * The latter is necessary to prevent ghost + * partitions on a removed medium. + */ + if (bdev->bd_invalidated && (!ret || ret == -ENOMEDIUM)) rescan_partitions(disk, bdev); + if (ret) + goto out_clear; } else { struct block_device *whole; whole = bdget_disk(disk, 0); @@ -1153,13 +1162,14 @@ static int __blkdev_get(struct block_device *bdev, fmode_t mode, int for_part) } } else { if (bdev->bd_contains == bdev) { - if (bdev->bd_disk->fops->open) { + ret = 0; + if (bdev->bd_disk->fops->open) ret = bdev->bd_disk->fops->open(bdev, mode); - if (ret) - goto out_unlock_bdev; - } - if (bdev->bd_invalidated) + /* the same as first opener case, read comment there */ + if (bdev->bd_invalidated && (!ret || ret == -ENOMEDIUM)) rescan_partitions(bdev->bd_disk, bdev); + if (ret) + goto out_unlock_bdev; } /* only one opener holds refs to the module and disk */ module_put(disk->fops->owner); @@ -1228,6 +1238,8 @@ int blkdev_get(struct block_device *bdev, fmode_t mode, void *holder) res = __blkdev_get(bdev, mode, 0); if (whole) { + struct gendisk *disk = whole->bd_disk; + /* finish claiming */ mutex_lock(&bdev->bd_mutex); spin_lock(&bdev_lock); @@ -1254,15 +1266,16 @@ int blkdev_get(struct block_device *bdev, fmode_t mode, void *holder) spin_unlock(&bdev_lock); /* - * Block event polling for write claims. Any write - * holder makes the write_holder state stick until all - * are released. This is good enough and tracking - * individual writeable reference is too fragile given - * the way @mode is used in blkdev_get/put(). + * Block event polling for write claims if requested. Any + * write holder makes the write_holder state stick until + * all are released. This is good enough and tracking + * individual writeable reference is too fragile given the + * way @mode is used in blkdev_get/put(). */ - if (!res && (mode & FMODE_WRITE) && !bdev->bd_write_holder) { + if ((disk->flags & GENHD_FL_BLOCK_EVENTS_ON_EXCL_WRITE) && + !res && (mode & FMODE_WRITE) && !bdev->bd_write_holder) { bdev->bd_write_holder = true; - disk_block_events(bdev->bd_disk); + disk_block_events(disk); } mutex_unlock(&bdev->bd_mutex); diff --git a/fs/btrfs/Makefile b/fs/btrfs/Makefile index 31610ea73aec..9b72dcf1cd25 100644 --- a/fs/btrfs/Makefile +++ b/fs/btrfs/Makefile @@ -7,4 +7,4 @@ btrfs-y += super.o ctree.o extent-tree.o print-tree.o root-tree.o dir-item.o \ extent_map.o sysfs.o struct-funcs.o xattr.o ordered-data.o \ extent_io.o volumes.o async-thread.o ioctl.o locking.o orphan.o \ export.o tree-log.o acl.o free-space-cache.o zlib.o lzo.o \ - compression.o delayed-ref.o relocation.o + compression.o delayed-ref.o relocation.o delayed-inode.o scrub.o diff --git a/fs/btrfs/acl.c b/fs/btrfs/acl.c index de34bfad9ec3..f66fc9959733 100644 --- a/fs/btrfs/acl.c +++ b/fs/btrfs/acl.c @@ -178,16 +178,18 @@ static int btrfs_xattr_acl_set(struct dentry *dentry, const char *name, if (value) { acl = posix_acl_from_xattr(value, size); - if (acl == NULL) { - value = NULL; - size = 0; - } else if (IS_ERR(acl)) { + if (IS_ERR(acl)) return PTR_ERR(acl); + + if (acl) { + ret = posix_acl_valid(acl); + if (ret) + goto out; } } ret = btrfs_set_acl(NULL, dentry->d_inode, acl, type); - +out: posix_acl_release(acl); return ret; @@ -286,7 +288,7 @@ int btrfs_acl_chmod(struct inode *inode) return 0; acl = btrfs_get_acl(inode, ACL_TYPE_ACCESS); - if (IS_ERR(acl) || !acl) + if (IS_ERR_OR_NULL(acl)) return PTR_ERR(acl); clone = posix_acl_clone(acl, GFP_KERNEL); diff --git a/fs/btrfs/btrfs_inode.h b/fs/btrfs/btrfs_inode.h index 57c3bb2884ce..93b1aa932014 100644 --- a/fs/btrfs/btrfs_inode.h +++ b/fs/btrfs/btrfs_inode.h @@ -22,6 +22,7 @@ #include "extent_map.h" #include "extent_io.h" #include "ordered-data.h" +#include "delayed-inode.h" /* in memory btrfs inode */ struct btrfs_inode { @@ -152,20 +153,34 @@ struct btrfs_inode { unsigned ordered_data_close:1; unsigned orphan_meta_reserved:1; unsigned dummy_inode:1; + unsigned in_defrag:1; /* * always compress this one file */ unsigned force_compress:4; + struct btrfs_delayed_node *delayed_node; + struct inode vfs_inode; }; +extern unsigned char btrfs_filetype_table[]; + static inline struct btrfs_inode *BTRFS_I(struct inode *inode) { return container_of(inode, struct btrfs_inode, vfs_inode); } +static inline u64 btrfs_ino(struct inode *inode) +{ + u64 ino = BTRFS_I(inode)->location.objectid; + + if (ino <= BTRFS_FIRST_FREE_OBJECTID) + ino = inode->i_ino; + return ino; +} + static inline void btrfs_i_size_write(struct inode *inode, u64 size) { i_size_write(inode, size); diff --git a/fs/btrfs/compression.c b/fs/btrfs/compression.c index 41d1d7c70e29..bfe42b03eaf9 100644 --- a/fs/btrfs/compression.c +++ b/fs/btrfs/compression.c @@ -125,9 +125,10 @@ static int check_compressed_csum(struct inode *inode, kunmap_atomic(kaddr, KM_USER0); if (csum != *cb_sum) { - printk(KERN_INFO "btrfs csum failed ino %lu " + printk(KERN_INFO "btrfs csum failed ino %llu " "extent %llu csum %u " - "wanted %u mirror %d\n", inode->i_ino, + "wanted %u mirror %d\n", + (unsigned long long)btrfs_ino(inode), (unsigned long long)disk_start, csum, *cb_sum, cb->mirror_num); ret = -EIO; @@ -332,7 +333,7 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start, struct compressed_bio *cb; unsigned long bytes_left; struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; - int page_index = 0; + int pg_index = 0; struct page *page; u64 first_byte = disk_start; struct block_device *bdev; @@ -366,8 +367,8 @@ int btrfs_submit_compressed_write(struct inode *inode, u64 start, /* create and submit bios for the compressed pages */ bytes_left = compressed_len; - for (page_index = 0; page_index < cb->nr_pages; page_index++) { - page = compressed_pages[page_index]; + for (pg_index = 0; pg_index < cb->nr_pages; pg_index++) { + page = compressed_pages[pg_index]; page->mapping = inode->i_mapping; if (bio->bi_size) ret = io_tree->ops->merge_bio_hook(page, 0, @@ -432,7 +433,7 @@ static noinline int add_ra_bio_pages(struct inode *inode, struct compressed_bio *cb) { unsigned long end_index; - unsigned long page_index; + unsigned long pg_index; u64 last_offset; u64 isize = i_size_read(inode); int ret; @@ -456,13 +457,13 @@ static noinline int add_ra_bio_pages(struct inode *inode, end_index = (i_size_read(inode) - 1) >> PAGE_CACHE_SHIFT; while (last_offset < compressed_end) { - page_index = last_offset >> PAGE_CACHE_SHIFT; + pg_index = last_offset >> PAGE_CACHE_SHIFT; - if (page_index > end_index) + if (pg_index > end_index) break; rcu_read_lock(); - page = radix_tree_lookup(&mapping->page_tree, page_index); + page = radix_tree_lookup(&mapping->page_tree, pg_index); rcu_read_unlock(); if (page) { misses++; @@ -476,7 +477,7 @@ static noinline int add_ra_bio_pages(struct inode *inode, if (!page) break; - if (add_to_page_cache_lru(page, mapping, page_index, + if (add_to_page_cache_lru(page, mapping, pg_index, GFP_NOFS)) { page_cache_release(page); goto next; @@ -560,7 +561,7 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, unsigned long uncompressed_len = bio->bi_vcnt * PAGE_CACHE_SIZE; unsigned long compressed_len; unsigned long nr_pages; - unsigned long page_index; + unsigned long pg_index; struct page *page; struct block_device *bdev; struct bio *comp_bio; @@ -613,10 +614,10 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, bdev = BTRFS_I(inode)->root->fs_info->fs_devices->latest_bdev; - for (page_index = 0; page_index < nr_pages; page_index++) { - cb->compressed_pages[page_index] = alloc_page(GFP_NOFS | + for (pg_index = 0; pg_index < nr_pages; pg_index++) { + cb->compressed_pages[pg_index] = alloc_page(GFP_NOFS | __GFP_HIGHMEM); - if (!cb->compressed_pages[page_index]) + if (!cb->compressed_pages[pg_index]) goto fail2; } cb->nr_pages = nr_pages; @@ -634,8 +635,8 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, comp_bio->bi_end_io = end_compressed_bio_read; atomic_inc(&cb->pending_bios); - for (page_index = 0; page_index < nr_pages; page_index++) { - page = cb->compressed_pages[page_index]; + for (pg_index = 0; pg_index < nr_pages; pg_index++) { + page = cb->compressed_pages[pg_index]; page->mapping = inode->i_mapping; page->index = em_start >> PAGE_CACHE_SHIFT; @@ -702,8 +703,8 @@ int btrfs_submit_compressed_read(struct inode *inode, struct bio *bio, return 0; fail2: - for (page_index = 0; page_index < nr_pages; page_index++) - free_page((unsigned long)cb->compressed_pages[page_index]); + for (pg_index = 0; pg_index < nr_pages; pg_index++) + free_page((unsigned long)cb->compressed_pages[pg_index]); kfree(cb->compressed_pages); fail1: @@ -945,7 +946,7 @@ void btrfs_exit_compress(void) int btrfs_decompress_buf2page(char *buf, unsigned long buf_start, unsigned long total_out, u64 disk_start, struct bio_vec *bvec, int vcnt, - unsigned long *page_index, + unsigned long *pg_index, unsigned long *pg_offset) { unsigned long buf_offset; @@ -954,7 +955,7 @@ int btrfs_decompress_buf2page(char *buf, unsigned long buf_start, unsigned long working_bytes = total_out - buf_start; unsigned long bytes; char *kaddr; - struct page *page_out = bvec[*page_index].bv_page; + struct page *page_out = bvec[*pg_index].bv_page; /* * start byte is the first byte of the page we're currently @@ -995,11 +996,11 @@ int btrfs_decompress_buf2page(char *buf, unsigned long buf_start, /* check if we need to pick another page */ if (*pg_offset == PAGE_CACHE_SIZE) { - (*page_index)++; - if (*page_index >= vcnt) + (*pg_index)++; + if (*pg_index >= vcnt) return 0; - page_out = bvec[*page_index].bv_page; + page_out = bvec[*pg_index].bv_page; *pg_offset = 0; start_byte = page_offset(page_out) - disk_start; diff --git a/fs/btrfs/compression.h b/fs/btrfs/compression.h index 51000174b9d7..a12059f4f0fd 100644 --- a/fs/btrfs/compression.h +++ b/fs/btrfs/compression.h @@ -37,7 +37,7 @@ int btrfs_decompress(int type, unsigned char *data_in, struct page *dest_page, int btrfs_decompress_buf2page(char *buf, unsigned long buf_start, unsigned long total_out, u64 disk_start, struct bio_vec *bvec, int vcnt, - unsigned long *page_index, + unsigned long *pg_index, unsigned long *pg_offset); int btrfs_submit_compressed_write(struct inode *inode, u64 start, diff --git a/fs/btrfs/ctree.c b/fs/btrfs/ctree.c index 84d7ca1fe0ba..b0e18d986e0a 100644 --- a/fs/btrfs/ctree.c +++ b/fs/btrfs/ctree.c @@ -38,11 +38,6 @@ static int balance_node_right(struct btrfs_trans_handle *trans, struct extent_buffer *src_buf); static int del_ptr(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, int level, int slot); -static int setup_items_for_insert(struct btrfs_trans_handle *trans, - struct btrfs_root *root, struct btrfs_path *path, - struct btrfs_key *cpu_key, u32 *data_size, - u32 total_data, u32 total_size, int nr); - struct btrfs_path *btrfs_alloc_path(void) { @@ -107,7 +102,7 @@ void btrfs_free_path(struct btrfs_path *p) { if (!p) return; - btrfs_release_path(NULL, p); + btrfs_release_path(p); kmem_cache_free(btrfs_path_cachep, p); } @@ -117,7 +112,7 @@ void btrfs_free_path(struct btrfs_path *p) * * It is safe to call this on paths that no locks or extent buffers held. */ -noinline void btrfs_release_path(struct btrfs_root *root, struct btrfs_path *p) +noinline void btrfs_release_path(struct btrfs_path *p) { int i; @@ -1328,7 +1323,7 @@ static noinline int reada_for_balance(struct btrfs_root *root, ret = -EAGAIN; /* release the whole path */ - btrfs_release_path(root, path); + btrfs_release_path(path); /* read the blocks */ if (block1) @@ -1475,7 +1470,7 @@ read_block_for_search(struct btrfs_trans_handle *trans, return 0; } free_extent_buffer(tmp); - btrfs_release_path(NULL, p); + btrfs_release_path(p); return -EIO; } } @@ -1494,7 +1489,7 @@ read_block_for_search(struct btrfs_trans_handle *trans, if (p->reada) reada_for_search(root, p, level, slot, key->objectid); - btrfs_release_path(NULL, p); + btrfs_release_path(p); ret = -EAGAIN; tmp = read_tree_block(root, blocknr, blocksize, 0); @@ -1563,7 +1558,7 @@ setup_nodes_for_search(struct btrfs_trans_handle *trans, } b = p->nodes[level]; if (!b) { - btrfs_release_path(NULL, p); + btrfs_release_path(p); goto again; } BUG_ON(btrfs_header_nritems(b) == 1); @@ -1753,7 +1748,7 @@ done: if (!p->leave_spinning) btrfs_set_path_blocking(p); if (ret < 0) - btrfs_release_path(root, p); + btrfs_release_path(p); return ret; } @@ -3026,7 +3021,7 @@ static noinline int setup_leaf_for_split(struct btrfs_trans_handle *trans, struct btrfs_file_extent_item); extent_len = btrfs_file_extent_num_bytes(leaf, fi); } - btrfs_release_path(root, path); + btrfs_release_path(path); path->keep_locks = 1; path->search_for_split = 1; @@ -3216,7 +3211,6 @@ int btrfs_truncate_item(struct btrfs_trans_handle *trans, struct btrfs_path *path, u32 new_size, int from_end) { - int ret = 0; int slot; struct extent_buffer *leaf; struct btrfs_item *item; @@ -3314,12 +3308,11 @@ int btrfs_truncate_item(struct btrfs_trans_handle *trans, btrfs_set_item_size(leaf, item, new_size); btrfs_mark_buffer_dirty(leaf); - ret = 0; if (btrfs_leaf_free_space(root, leaf) < 0) { btrfs_print_leaf(root, leaf); BUG(); } - return ret; + return 0; } /* @@ -3329,7 +3322,6 @@ int btrfs_extend_item(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, u32 data_size) { - int ret = 0; int slot; struct extent_buffer *leaf; struct btrfs_item *item; @@ -3394,12 +3386,11 @@ int btrfs_extend_item(struct btrfs_trans_handle *trans, btrfs_set_item_size(leaf, item, old_size + data_size); btrfs_mark_buffer_dirty(leaf); - ret = 0; if (btrfs_leaf_free_space(root, leaf) < 0) { btrfs_print_leaf(root, leaf); BUG(); } - return ret; + return 0; } /* @@ -3559,11 +3550,10 @@ out: * to save stack depth by doing the bulk of the work in a function * that doesn't call btrfs_search_slot */ -static noinline_for_stack int -setup_items_for_insert(struct btrfs_trans_handle *trans, - struct btrfs_root *root, struct btrfs_path *path, - struct btrfs_key *cpu_key, u32 *data_size, - u32 total_data, u32 total_size, int nr) +int setup_items_for_insert(struct btrfs_trans_handle *trans, + struct btrfs_root *root, struct btrfs_path *path, + struct btrfs_key *cpu_key, u32 *data_size, + u32 total_data, u32 total_size, int nr) { struct btrfs_item *item; int i; @@ -3647,7 +3637,6 @@ setup_items_for_insert(struct btrfs_trans_handle *trans, ret = 0; if (slot == 0) { - struct btrfs_disk_key disk_key; btrfs_cpu_key_to_disk(&disk_key, cpu_key); ret = fixup_low_keys(trans, root, path, &disk_key, 1); } @@ -3949,7 +3938,7 @@ int btrfs_prev_leaf(struct btrfs_root *root, struct btrfs_path *path) else return 1; - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) return ret; @@ -4073,7 +4062,7 @@ find_next_key: sret = btrfs_find_next_key(root, path, min_key, level, cache_only, min_trans); if (sret == 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto again; } else { goto out; @@ -4152,7 +4141,7 @@ next: btrfs_node_key_to_cpu(c, &cur_key, slot); orig_lowest = path->lowest_level; - btrfs_release_path(root, path); + btrfs_release_path(path); path->lowest_level = level; ret = btrfs_search_slot(NULL, root, &cur_key, path, 0, 0); @@ -4229,7 +4218,7 @@ int btrfs_next_leaf(struct btrfs_root *root, struct btrfs_path *path) again: level = 1; next = NULL; - btrfs_release_path(root, path); + btrfs_release_path(path); path->keep_locks = 1; @@ -4285,7 +4274,7 @@ again: goto again; if (ret < 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto done; } @@ -4324,7 +4313,7 @@ again: goto again; if (ret < 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto done; } diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index 3458b5725540..6c093fa98f61 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -23,6 +23,7 @@ #include <linux/mm.h> #include <linux/highmem.h> #include <linux/fs.h> +#include <linux/rwsem.h> #include <linux/completion.h> #include <linux/backing-dev.h> #include <linux/wait.h> @@ -33,6 +34,7 @@ #include "extent_io.h" #include "extent_map.h" #include "async-thread.h" +#include "ioctl.h" struct btrfs_trans_handle; struct btrfs_transaction; @@ -105,6 +107,12 @@ struct btrfs_ordered_sum; /* For storing free space cache */ #define BTRFS_FREE_SPACE_OBJECTID -11ULL +/* + * The inode number assigned to the special inode for sotring + * free ino cache + */ +#define BTRFS_FREE_INO_OBJECTID -12ULL + /* dummy objectid represents multiple objectids */ #define BTRFS_MULTIPLE_OBJECTIDS -255ULL @@ -187,7 +195,6 @@ struct btrfs_mapping_tree { struct extent_map_tree map_tree; }; -#define BTRFS_UUID_SIZE 16 struct btrfs_dev_item { /* the internal btrfs device id */ __le64 devid; @@ -294,7 +301,6 @@ static inline unsigned long btrfs_chunk_item_size(int num_stripes) sizeof(struct btrfs_stripe) * (num_stripes - 1); } -#define BTRFS_FSID_SIZE 16 #define BTRFS_HEADER_FLAG_WRITTEN (1ULL << 0) #define BTRFS_HEADER_FLAG_RELOC (1ULL << 1) @@ -510,6 +516,12 @@ struct btrfs_extent_item_v0 { /* use full backrefs for extent pointers in the block */ #define BTRFS_BLOCK_FLAG_FULL_BACKREF (1ULL << 8) +/* + * this flag is only used internally by scrub and may be changed at any time + * it is only declared here to avoid collisions + */ +#define BTRFS_EXTENT_FLAG_SUPER (1ULL << 48) + struct btrfs_tree_block_info { struct btrfs_disk_key key; u8 level; @@ -718,7 +730,7 @@ struct btrfs_space_info { u64 total_bytes; /* total bytes in the space, this doesn't take mirrors into account */ u64 bytes_used; /* total bytes used, - this does't take mirrors into account */ + this doesn't take mirrors into account */ u64 bytes_pinned; /* total bytes pinned, will be freed when the transaction finishes */ u64 bytes_reserved; /* total bytes the allocator has reserved for @@ -740,10 +752,12 @@ struct btrfs_space_info { */ unsigned long reservation_progress; - int full; /* indicates that we cannot allocate any more + unsigned int full:1; /* indicates that we cannot allocate any more chunks for this space */ - int force_alloc; /* set if we need to force a chunk alloc for - this space */ + unsigned int chunk_alloc:1; /* set if we are allocating a chunk */ + + unsigned int force_alloc; /* set if we need to force a chunk + alloc for this space */ struct list_head list; @@ -828,9 +842,6 @@ struct btrfs_block_group_cache { u64 bytes_super; u64 flags; u64 sectorsize; - int extents_thresh; - int free_extents; - int total_bitmaps; unsigned int ro:1; unsigned int dirty:1; unsigned int iref:1; @@ -845,9 +856,7 @@ struct btrfs_block_group_cache { struct btrfs_space_info *space_info; /* free space cache stuff */ - spinlock_t tree_lock; - struct rb_root free_space_offset; - u64 free_space; + struct btrfs_free_space_ctl *free_space_ctl; /* block group cache stuff */ struct rb_node cache_node; @@ -867,6 +876,7 @@ struct btrfs_block_group_cache { struct reloc_control; struct btrfs_device; struct btrfs_fs_devices; +struct btrfs_delayed_root; struct btrfs_fs_info { u8 fsid[BTRFS_FSID_SIZE]; u8 chunk_tree_uuid[BTRFS_UUID_SIZE]; @@ -893,7 +903,10 @@ struct btrfs_fs_info { /* logical->physical extent mapping */ struct btrfs_mapping_tree mapping_tree; - /* block reservation for extent, checksum and root tree */ + /* + * block reservation for extent, checksum, root tree and + * delayed dir index item + */ struct btrfs_block_rsv global_block_rsv; /* block reservation for delay allocation */ struct btrfs_block_rsv delalloc_block_rsv; @@ -1020,6 +1033,7 @@ struct btrfs_fs_info { * for the sys_munmap function call path */ struct btrfs_workers fixup_workers; + struct btrfs_workers delayed_workers; struct task_struct *transaction_kthread; struct task_struct *cleaner_kthread; int thread_pool_size; @@ -1060,6 +1074,11 @@ struct btrfs_fs_info { /* all metadata allocations go through this cluster */ struct btrfs_free_cluster meta_alloc_cluster; + /* auto defrag inodes go here */ + spinlock_t defrag_inodes_lock; + struct rb_root defrag_inodes; + atomic_t defrag_running; + spinlock_t ref_cache_lock; u64 total_ref_cache_size; @@ -1075,8 +1094,21 @@ struct btrfs_fs_info { void *bdev_holder; + /* private scrub information */ + struct mutex scrub_lock; + atomic_t scrubs_running; + atomic_t scrub_pause_req; + atomic_t scrubs_paused; + atomic_t scrub_cancel_req; + wait_queue_head_t scrub_pause_wait; + struct rw_semaphore scrub_super_lock; + int scrub_workers_refcnt; + struct btrfs_workers scrub_workers; + /* filesystem state */ u64 fs_state; + + struct btrfs_delayed_root *delayed_root; }; /* @@ -1086,9 +1118,6 @@ struct btrfs_fs_info { struct btrfs_root { struct extent_buffer *node; - /* the node lock is held while changing the node pointer */ - spinlock_t node_lock; - struct extent_buffer *commit_root; struct btrfs_root *log_root; struct btrfs_root *reloc_root; @@ -1105,6 +1134,16 @@ struct btrfs_root { spinlock_t accounting_lock; struct btrfs_block_rsv *block_rsv; + /* free ino cache stuff */ + struct mutex fs_commit_mutex; + struct btrfs_free_space_ctl *free_ino_ctl; + enum btrfs_caching_type cached; + spinlock_t cache_lock; + wait_queue_head_t cache_wait; + struct btrfs_free_space_ctl *free_ino_pinned; + u64 cache_progress; + struct inode *cache_inode; + struct mutex log_mutex; wait_queue_head_t log_writer_wait; wait_queue_head_t log_commit_wait[2]; @@ -1160,12 +1199,49 @@ struct btrfs_root { struct rb_root inode_tree; /* + * radix tree that keeps track of delayed nodes of every inode, + * protected by inode_lock + */ + struct radix_tree_root delayed_nodes_tree; + /* * right now this just gets used so that a root has its own devid * for stat. It may be used for more later */ struct super_block anon_super; }; +struct btrfs_ioctl_defrag_range_args { + /* start of the defrag operation */ + __u64 start; + + /* number of bytes to defrag, use (u64)-1 to say all */ + __u64 len; + + /* + * flags for the operation, which can include turning + * on compression for this one defrag + */ + __u64 flags; + + /* + * any extent bigger than this will be considered + * already defragged. Use 0 to take the kernel default + * Use 1 to say every single extent must be rewritten + */ + __u32 extent_thresh; + + /* + * which compression method to use if turning on compression + * for this defrag operation. If unspecified, zlib will + * be used + */ + __u32 compress_type; + + /* spare for later */ + __u32 unused[4]; +}; + + /* * inode items have the data typically returned from stat and store other * info about object characteristics. There is one for every file and dir in @@ -1263,6 +1339,7 @@ struct btrfs_root { #define BTRFS_MOUNT_CLEAR_CACHE (1 << 13) #define BTRFS_MOUNT_USER_SUBVOL_RM_ALLOWED (1 << 14) #define BTRFS_MOUNT_ENOSPC_DEBUG (1 << 15) +#define BTRFS_MOUNT_AUTO_DEFRAG (1 << 16) #define btrfs_clear_opt(o, opt) ((o) &= ~BTRFS_MOUNT_##opt) #define btrfs_set_opt(o, opt) ((o) |= BTRFS_MOUNT_##opt) @@ -1438,26 +1515,12 @@ static inline u64 btrfs_stripe_offset_nr(struct extent_buffer *eb, return btrfs_stripe_offset(eb, btrfs_stripe_nr(c, nr)); } -static inline void btrfs_set_stripe_offset_nr(struct extent_buffer *eb, - struct btrfs_chunk *c, int nr, - u64 val) -{ - btrfs_set_stripe_offset(eb, btrfs_stripe_nr(c, nr), val); -} - static inline u64 btrfs_stripe_devid_nr(struct extent_buffer *eb, struct btrfs_chunk *c, int nr) { return btrfs_stripe_devid(eb, btrfs_stripe_nr(c, nr)); } -static inline void btrfs_set_stripe_devid_nr(struct extent_buffer *eb, - struct btrfs_chunk *c, int nr, - u64 val) -{ - btrfs_set_stripe_devid(eb, btrfs_stripe_nr(c, nr), val); -} - /* struct btrfs_block_group_item */ BTRFS_SETGET_STACK_FUNCS(block_group_used, struct btrfs_block_group_item, used, 64); @@ -1515,14 +1578,6 @@ btrfs_inode_ctime(struct btrfs_inode_item *inode_item) return (struct btrfs_timespec *)ptr; } -static inline struct btrfs_timespec * -btrfs_inode_otime(struct btrfs_inode_item *inode_item) -{ - unsigned long ptr = (unsigned long)inode_item; - ptr += offsetof(struct btrfs_inode_item, otime); - return (struct btrfs_timespec *)ptr; -} - BTRFS_SETGET_FUNCS(timespec_sec, struct btrfs_timespec, sec, 64); BTRFS_SETGET_FUNCS(timespec_nsec, struct btrfs_timespec, nsec, 32); @@ -1873,33 +1928,6 @@ static inline u8 *btrfs_header_chunk_tree_uuid(struct extent_buffer *eb) return (u8 *)ptr; } -static inline u8 *btrfs_super_fsid(struct extent_buffer *eb) -{ - unsigned long ptr = offsetof(struct btrfs_super_block, fsid); - return (u8 *)ptr; -} - -static inline u8 *btrfs_header_csum(struct extent_buffer *eb) -{ - unsigned long ptr = offsetof(struct btrfs_header, csum); - return (u8 *)ptr; -} - -static inline struct btrfs_node *btrfs_buffer_node(struct extent_buffer *eb) -{ - return NULL; -} - -static inline struct btrfs_leaf *btrfs_buffer_leaf(struct extent_buffer *eb) -{ - return NULL; -} - -static inline struct btrfs_header *btrfs_buffer_header(struct extent_buffer *eb) -{ - return NULL; -} - static inline int btrfs_is_leaf(struct extent_buffer *eb) { return btrfs_header_level(eb) == 0; @@ -2053,22 +2081,6 @@ static inline struct btrfs_root *btrfs_sb(struct super_block *sb) return sb->s_fs_info; } -static inline int btrfs_set_root_name(struct btrfs_root *root, - const char *name, int len) -{ - /* if we already have a name just free it */ - kfree(root->name); - - root->name = kmalloc(len+1, GFP_KERNEL); - if (!root->name) - return -ENOMEM; - - memcpy(root->name, name, len); - root->name[len] = '\0'; - - return 0; -} - static inline u32 btrfs_level_size(struct btrfs_root *root, int level) { if (level == 0) @@ -2097,6 +2109,13 @@ static inline bool btrfs_mixed_space_info(struct btrfs_space_info *space_info) } /* extent-tree.c */ +static inline u64 btrfs_calc_trans_metadata_size(struct btrfs_root *root, + int num_items) +{ + return (root->leafsize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) * + 3 * num_items; +} + void btrfs_put_block_group(struct btrfs_block_group_cache *cache); int btrfs_run_delayed_refs(struct btrfs_trans_handle *trans, struct btrfs_root *root, unsigned long count); @@ -2106,12 +2125,9 @@ int btrfs_lookup_extent_info(struct btrfs_trans_handle *trans, u64 num_bytes, u64 *refs, u64 *flags); int btrfs_pin_extent(struct btrfs_root *root, u64 bytenr, u64 num, int reserved); -int btrfs_drop_leaf_ref(struct btrfs_trans_handle *trans, - struct btrfs_root *root, struct extent_buffer *leaf); int btrfs_cross_ref_exist(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 objectid, u64 offset, u64 bytenr); -int btrfs_copy_pinned(struct btrfs_root *root, struct extent_io_tree *copy); struct btrfs_block_group_cache *btrfs_lookup_block_group( struct btrfs_fs_info *info, u64 bytenr); @@ -2288,10 +2304,12 @@ int btrfs_realloc_node(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *parent, int start_slot, int cache_only, u64 *last_ret, struct btrfs_key *progress); -void btrfs_release_path(struct btrfs_root *root, struct btrfs_path *p); +void btrfs_release_path(struct btrfs_path *p); struct btrfs_path *btrfs_alloc_path(void); void btrfs_free_path(struct btrfs_path *p); void btrfs_set_path_blocking(struct btrfs_path *p); +void btrfs_clear_path_blocking(struct btrfs_path *p, + struct extent_buffer *held); void btrfs_unlock_up_safe(struct btrfs_path *p, int level); int btrfs_del_items(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -2303,13 +2321,12 @@ static inline int btrfs_del_item(struct btrfs_trans_handle *trans, return btrfs_del_items(trans, root, path, path->slots[0], 1); } +int setup_items_for_insert(struct btrfs_trans_handle *trans, + struct btrfs_root *root, struct btrfs_path *path, + struct btrfs_key *cpu_key, u32 *data_size, + u32 total_data, u32 total_size, int nr); int btrfs_insert_item(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_key *key, void *data, u32 data_size); -int btrfs_insert_some_items(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_path *path, - struct btrfs_key *cpu_key, u32 *data_size, - int nr); int btrfs_insert_empty_items(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, @@ -2355,8 +2372,6 @@ int btrfs_update_root(struct btrfs_trans_handle *trans, struct btrfs_root *item); int btrfs_find_last_root(struct btrfs_root *root, u64 objectid, struct btrfs_root_item *item, struct btrfs_key *key); -int btrfs_search_root(struct btrfs_root *root, u64 search_start, - u64 *found_objectid); int btrfs_find_dead_roots(struct btrfs_root *root, u64 objectid); int btrfs_find_orphan_roots(struct btrfs_root *tree_root); int btrfs_set_root_node(struct btrfs_root_item *item, @@ -2366,7 +2381,7 @@ void btrfs_check_and_init_root_item(struct btrfs_root_item *item); /* dir-item.c */ int btrfs_insert_dir_item(struct btrfs_trans_handle *trans, struct btrfs_root *root, const char *name, - int name_len, u64 dir, + int name_len, struct inode *dir, struct btrfs_key *location, u8 type, u64 index); struct btrfs_dir_item *btrfs_lookup_dir_item(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -2411,12 +2426,6 @@ int btrfs_del_orphan_item(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 offset); int btrfs_find_orphan_item(struct btrfs_root *root, u64 offset); -/* inode-map.c */ -int btrfs_find_free_objectid(struct btrfs_trans_handle *trans, - struct btrfs_root *fs_root, - u64 dirid, u64 *objectid); -int btrfs_find_highest_inode(struct btrfs_root *fs_root, u64 *objectid); - /* inode-item.c */ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -2461,8 +2470,6 @@ int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans, struct btrfs_ordered_sum *sums); int btrfs_csum_one_bio(struct btrfs_root *root, struct inode *inode, struct bio *bio, u64 file_start, int contig); -int btrfs_csum_file_bytes(struct btrfs_root *root, struct inode *inode, - u64 start, unsigned long len); struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, @@ -2470,8 +2477,8 @@ struct btrfs_csum_item *btrfs_lookup_csum(struct btrfs_trans_handle *trans, int btrfs_csum_truncate(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_path *path, u64 isize); -int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, - u64 end, struct list_head *list); +int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end, + struct list_head *list, int search_commit); /* inode.c */ /* RHEL and EL kernels have a patch that renames PG_checked to FsMisc */ @@ -2500,8 +2507,6 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans, u32 min_type); int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput); -int btrfs_start_one_delalloc_inode(struct btrfs_root *root, int delay_iput, - int sync); int btrfs_set_extent_delalloc(struct inode *inode, u64 start, u64 end, struct extent_state **cached_state); int btrfs_writepages(struct address_space *mapping, @@ -2518,9 +2523,8 @@ unsigned long btrfs_force_ra(struct address_space *mapping, int btrfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf); int btrfs_readpage(struct file *file, struct page *page); void btrfs_evict_inode(struct inode *inode); -void btrfs_put_inode(struct inode *inode); int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc); -void btrfs_dirty_inode(struct inode *inode); +void btrfs_dirty_inode(struct inode *inode, int flags); struct inode *btrfs_alloc_inode(struct super_block *sb); void btrfs_destroy_inode(struct inode *inode); int btrfs_drop_inode(struct inode *inode); @@ -2529,10 +2533,8 @@ void btrfs_destroy_cachep(void); long btrfs_ioctl_trans_end(struct file *file); struct inode *btrfs_iget(struct super_block *s, struct btrfs_key *location, struct btrfs_root *root, int *was_new); -int btrfs_commit_write(struct file *file, struct page *page, - unsigned from, unsigned to); struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page, - size_t page_offset, u64 start, u64 end, + size_t pg_offset, u64 start, u64 end, int create); int btrfs_update_inode(struct btrfs_trans_handle *trans, struct btrfs_root *root, @@ -2564,18 +2566,27 @@ extern const struct dentry_operations btrfs_dentry_operations; long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg); void btrfs_update_iflags(struct inode *inode); void btrfs_inherit_iflags(struct inode *inode, struct inode *dir); - +int btrfs_defrag_file(struct inode *inode, struct file *file, + struct btrfs_ioctl_defrag_range_args *range, + u64 newer_than, unsigned long max_pages); /* file.c */ +int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans, + struct inode *inode); +int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info); int btrfs_sync_file(struct file *file, int datasync); int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end, int skip_pinned); -int btrfs_check_file(struct btrfs_root *root, struct inode *inode); extern const struct file_operations btrfs_file_operations; int btrfs_drop_extents(struct btrfs_trans_handle *trans, struct inode *inode, u64 start, u64 end, u64 *hint_byte, int drop_cache); int btrfs_mark_extent_written(struct btrfs_trans_handle *trans, struct inode *inode, u64 start, u64 end); int btrfs_release_file(struct inode *inode, struct file *file); +void btrfs_drop_pages(struct page **pages, size_t num_pages); +int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode, + struct page **pages, size_t num_pages, + loff_t pos, size_t write_bytes, + struct extent_state **cached); /* tree-defrag.c */ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, @@ -2584,10 +2595,6 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, /* sysfs.c */ int btrfs_init_sysfs(void); void btrfs_exit_sysfs(void); -int btrfs_sysfs_add_super(struct btrfs_fs_info *fs); -int btrfs_sysfs_add_root(struct btrfs_root *root); -void btrfs_sysfs_del_root(struct btrfs_root *root); -void btrfs_sysfs_del_super(struct btrfs_fs_info *root); /* xattr.c */ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size); @@ -2630,4 +2637,18 @@ void btrfs_reloc_pre_snapshot(struct btrfs_trans_handle *trans, u64 *bytes_to_reserve); void btrfs_reloc_post_snapshot(struct btrfs_trans_handle *trans, struct btrfs_pending_snapshot *pending); + +/* scrub.c */ +int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end, + struct btrfs_scrub_progress *progress, int readonly); +int btrfs_scrub_pause(struct btrfs_root *root); +int btrfs_scrub_pause_super(struct btrfs_root *root); +int btrfs_scrub_continue(struct btrfs_root *root); +int btrfs_scrub_continue_super(struct btrfs_root *root); +int btrfs_scrub_cancel(struct btrfs_root *root); +int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev); +int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid); +int btrfs_scrub_progress(struct btrfs_root *root, u64 devid, + struct btrfs_scrub_progress *progress); + #endif diff --git a/fs/btrfs/delayed-inode.c b/fs/btrfs/delayed-inode.c new file mode 100644 index 000000000000..01e29503a54b --- /dev/null +++ b/fs/btrfs/delayed-inode.c @@ -0,0 +1,1695 @@ +/* + * Copyright (C) 2011 Fujitsu. All rights reserved. + * Written by Miao Xie <miaox@cn.fujitsu.com> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License v2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public + * License along with this program; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 021110-1307, USA. + */ + +#include <linux/slab.h> +#include "delayed-inode.h" +#include "disk-io.h" +#include "transaction.h" + +#define BTRFS_DELAYED_WRITEBACK 400 +#define BTRFS_DELAYED_BACKGROUND 100 + +static struct kmem_cache *delayed_node_cache; + +int __init btrfs_delayed_inode_init(void) +{ + delayed_node_cache = kmem_cache_create("delayed_node", + sizeof(struct btrfs_delayed_node), + 0, + SLAB_RECLAIM_ACCOUNT | SLAB_MEM_SPREAD, + NULL); + if (!delayed_node_cache) + return -ENOMEM; + return 0; +} + +void btrfs_delayed_inode_exit(void) +{ + if (delayed_node_cache) + kmem_cache_destroy(delayed_node_cache); +} + +static inline void btrfs_init_delayed_node( + struct btrfs_delayed_node *delayed_node, + struct btrfs_root *root, u64 inode_id) +{ + delayed_node->root = root; + delayed_node->inode_id = inode_id; + atomic_set(&delayed_node->refs, 0); + delayed_node->count = 0; + delayed_node->in_list = 0; + delayed_node->inode_dirty = 0; + delayed_node->ins_root = RB_ROOT; + delayed_node->del_root = RB_ROOT; + mutex_init(&delayed_node->mutex); + delayed_node->index_cnt = 0; + INIT_LIST_HEAD(&delayed_node->n_list); + INIT_LIST_HEAD(&delayed_node->p_list); + delayed_node->bytes_reserved = 0; +} + +static inline int btrfs_is_continuous_delayed_item( + struct btrfs_delayed_item *item1, + struct btrfs_delayed_item *item2) +{ + if (item1->key.type == BTRFS_DIR_INDEX_KEY && + item1->key.objectid == item2->key.objectid && + item1->key.type == item2->key.type && + item1->key.offset + 1 == item2->key.offset) + return 1; + return 0; +} + +static inline struct btrfs_delayed_root *btrfs_get_delayed_root( + struct btrfs_root *root) +{ + return root->fs_info->delayed_root; +} + +static struct btrfs_delayed_node *btrfs_get_or_create_delayed_node( + struct inode *inode) +{ + struct btrfs_delayed_node *node; + struct btrfs_inode *btrfs_inode = BTRFS_I(inode); + struct btrfs_root *root = btrfs_inode->root; + u64 ino = btrfs_ino(inode); + int ret; + +again: + node = ACCESS_ONCE(btrfs_inode->delayed_node); + if (node) { + atomic_inc(&node->refs); /* can be accessed */ + return node; + } + + spin_lock(&root->inode_lock); + node = radix_tree_lookup(&root->delayed_nodes_tree, ino); + if (node) { + if (btrfs_inode->delayed_node) { + spin_unlock(&root->inode_lock); + goto again; + } + btrfs_inode->delayed_node = node; + atomic_inc(&node->refs); /* can be accessed */ + atomic_inc(&node->refs); /* cached in the inode */ + spin_unlock(&root->inode_lock); + return node; + } + spin_unlock(&root->inode_lock); + + node = kmem_cache_alloc(delayed_node_cache, GFP_NOFS); + if (!node) + return ERR_PTR(-ENOMEM); + btrfs_init_delayed_node(node, root, ino); + + atomic_inc(&node->refs); /* cached in the btrfs inode */ + atomic_inc(&node->refs); /* can be accessed */ + + ret = radix_tree_preload(GFP_NOFS & ~__GFP_HIGHMEM); + if (ret) { + kmem_cache_free(delayed_node_cache, node); + return ERR_PTR(ret); + } + + spin_lock(&root->inode_lock); + ret = radix_tree_insert(&root->delayed_nodes_tree, ino, node); + if (ret == -EEXIST) { + kmem_cache_free(delayed_node_cache, node); + spin_unlock(&root->inode_lock); + radix_tree_preload_end(); + goto again; + } + btrfs_inode->delayed_node = node; + spin_unlock(&root->inode_lock); + radix_tree_preload_end(); + + return node; +} + +/* + * Call it when holding delayed_node->mutex + * + * If mod = 1, add this node into the prepared list. + */ +static void btrfs_queue_delayed_node(struct btrfs_delayed_root *root, + struct btrfs_delayed_node *node, + int mod) +{ + spin_lock(&root->lock); + if (node->in_list) { + if (!list_empty(&node->p_list)) + list_move_tail(&node->p_list, &root->prepare_list); + else if (mod) + list_add_tail(&node->p_list, &root->prepare_list); + } else { + list_add_tail(&node->n_list, &root->node_list); + list_add_tail(&node->p_list, &root->prepare_list); + atomic_inc(&node->refs); /* inserted into list */ + root->nodes++; + node->in_list = 1; + } + spin_unlock(&root->lock); +} + +/* Call it when holding delayed_node->mutex */ +static void btrfs_dequeue_delayed_node(struct btrfs_delayed_root *root, + struct btrfs_delayed_node *node) +{ + spin_lock(&root->lock); + if (node->in_list) { + root->nodes--; + atomic_dec(&node->refs); /* not in the list */ + list_del_init(&node->n_list); + if (!list_empty(&node->p_list)) + list_del_init(&node->p_list); + node->in_list = 0; + } + spin_unlock(&root->lock); +} + +struct btrfs_delayed_node *btrfs_first_delayed_node( + struct btrfs_delayed_root *delayed_root) +{ + struct list_head *p; + struct btrfs_delayed_node *node = NULL; + + spin_lock(&delayed_root->lock); + if (list_empty(&delayed_root->node_list)) + goto out; + + p = delayed_root->node_list.next; + node = list_entry(p, struct btrfs_delayed_node, n_list); + atomic_inc(&node->refs); +out: + spin_unlock(&delayed_root->lock); + + return node; +} + +struct btrfs_delayed_node *btrfs_next_delayed_node( + struct btrfs_delayed_node *node) +{ + struct btrfs_delayed_root *delayed_root; + struct list_head *p; + struct btrfs_delayed_node *next = NULL; + + delayed_root = node->root->fs_info->delayed_root; + spin_lock(&delayed_root->lock); + if (!node->in_list) { /* not in the list */ + if (list_empty(&delayed_root->node_list)) + goto out; + p = delayed_root->node_list.next; + } else if (list_is_last(&node->n_list, &delayed_root->node_list)) + goto out; + else + p = node->n_list.next; + + next = list_entry(p, struct btrfs_delayed_node, n_list); + atomic_inc(&next->refs); +out: + spin_unlock(&delayed_root->lock); + + return next; +} + +static void __btrfs_release_delayed_node( + struct btrfs_delayed_node *delayed_node, + int mod) +{ + struct btrfs_delayed_root *delayed_root; + + if (!delayed_node) + return; + + delayed_root = delayed_node->root->fs_info->delayed_root; + + mutex_lock(&delayed_node->mutex); + if (delayed_node->count) + btrfs_queue_delayed_node(delayed_root, delayed_node, mod); + else + btrfs_dequeue_delayed_node(delayed_root, delayed_node); + mutex_unlock(&delayed_node->mutex); + + if (atomic_dec_and_test(&delayed_node->refs)) { + struct btrfs_root *root = delayed_node->root; + spin_lock(&root->inode_lock); + if (atomic_read(&delayed_node->refs) == 0) { + radix_tree_delete(&root->delayed_nodes_tree, + delayed_node->inode_id); + kmem_cache_free(delayed_node_cache, delayed_node); + } + spin_unlock(&root->inode_lock); + } +} + +static inline void btrfs_release_delayed_node(struct btrfs_delayed_node *node) +{ + __btrfs_release_delayed_node(node, 0); +} + +struct btrfs_delayed_node *btrfs_first_prepared_delayed_node( + struct btrfs_delayed_root *delayed_root) +{ + struct list_head *p; + struct btrfs_delayed_node *node = NULL; + + spin_lock(&delayed_root->lock); + if (list_empty(&delayed_root->prepare_list)) + goto out; + + p = delayed_root->prepare_list.next; + list_del_init(p); + node = list_entry(p, struct btrfs_delayed_node, p_list); + atomic_inc(&node->refs); +out: + spin_unlock(&delayed_root->lock); + + return node; +} + +static inline void btrfs_release_prepared_delayed_node( + struct btrfs_delayed_node *node) +{ + __btrfs_release_delayed_node(node, 1); +} + +struct btrfs_delayed_item *btrfs_alloc_delayed_item(u32 data_len) +{ + struct btrfs_delayed_item *item; + item = kmalloc(sizeof(*item) + data_len, GFP_NOFS); + if (item) { + item->data_len = data_len; + item->ins_or_del = 0; + item->bytes_reserved = 0; + item->block_rsv = NULL; + item->delayed_node = NULL; + atomic_set(&item->refs, 1); + } + return item; +} + +/* + * __btrfs_lookup_delayed_item - look up the delayed item by key + * @delayed_node: pointer to the delayed node + * @key: the key to look up + * @prev: used to store the prev item if the right item isn't found + * @next: used to store the next item if the right item isn't found + * + * Note: if we don't find the right item, we will return the prev item and + * the next item. + */ +static struct btrfs_delayed_item *__btrfs_lookup_delayed_item( + struct rb_root *root, + struct btrfs_key *key, + struct btrfs_delayed_item **prev, + struct btrfs_delayed_item **next) +{ + struct rb_node *node, *prev_node = NULL; + struct btrfs_delayed_item *delayed_item = NULL; + int ret = 0; + + node = root->rb_node; + + while (node) { + delayed_item = rb_entry(node, struct btrfs_delayed_item, + rb_node); + prev_node = node; + ret = btrfs_comp_cpu_keys(&delayed_item->key, key); + if (ret < 0) + node = node->rb_right; + else if (ret > 0) + node = node->rb_left; + else + return delayed_item; + } + + if (prev) { + if (!prev_node) + *prev = NULL; + else if (ret < 0) + *prev = delayed_item; + else if ((node = rb_prev(prev_node)) != NULL) { + *prev = rb_entry(node, struct btrfs_delayed_item, + rb_node); + } else + *prev = NULL; + } + + if (next) { + if (!prev_node) + *next = NULL; + else if (ret > 0) + *next = delayed_item; + else if ((node = rb_next(prev_node)) != NULL) { + *next = rb_entry(node, struct btrfs_delayed_item, + rb_node); + } else + *next = NULL; + } + return NULL; +} + +struct btrfs_delayed_item *__btrfs_lookup_delayed_insertion_item( + struct btrfs_delayed_node *delayed_node, + struct btrfs_key *key) +{ + struct btrfs_delayed_item *item; + + item = __btrfs_lookup_delayed_item(&delayed_node->ins_root, key, + NULL, NULL); + return item; +} + +struct btrfs_delayed_item *__btrfs_lookup_delayed_deletion_item( + struct btrfs_delayed_node *delayed_node, + struct btrfs_key *key) +{ + struct btrfs_delayed_item *item; + + item = __btrfs_lookup_delayed_item(&delayed_node->del_root, key, + NULL, NULL); + return item; +} + +struct btrfs_delayed_item *__btrfs_search_delayed_insertion_item( + struct btrfs_delayed_node *delayed_node, + struct btrfs_key *key) +{ + struct btrfs_delayed_item *item, *next; + + item = __btrfs_lookup_delayed_item(&delayed_node->ins_root, key, + NULL, &next); + if (!item) + item = next; + + return item; +} + +struct btrfs_delayed_item *__btrfs_search_delayed_deletion_item( + struct btrfs_delayed_node *delayed_node, + struct btrfs_key *key) +{ + struct btrfs_delayed_item *item, *next; + + item = __btrfs_lookup_delayed_item(&delayed_node->del_root, key, + NULL, &next); + if (!item) + item = next; + + return item; +} + +static int __btrfs_add_delayed_item(struct btrfs_delayed_node *delayed_node, + struct btrfs_delayed_item *ins, + int action) +{ + struct rb_node **p, *node; + struct rb_node *parent_node = NULL; + struct rb_root *root; + struct btrfs_delayed_item *item; + int cmp; + + if (action == BTRFS_DELAYED_INSERTION_ITEM) + root = &delayed_node->ins_root; + else if (action == BTRFS_DELAYED_DELETION_ITEM) + root = &delayed_node->del_root; + else + BUG(); + p = &root->rb_node; + node = &ins->rb_node; + + while (*p) { + parent_node = *p; + item = rb_entry(parent_node, struct btrfs_delayed_item, + rb_node); + + cmp = btrfs_comp_cpu_keys(&item->key, &ins->key); + if (cmp < 0) + p = &(*p)->rb_right; + else if (cmp > 0) + p = &(*p)->rb_left; + else + return -EEXIST; + } + + rb_link_node(node, parent_node, p); + rb_insert_color(node, root); + ins->delayed_node = delayed_node; + ins->ins_or_del = action; + + if (ins->key.type == BTRFS_DIR_INDEX_KEY && + action == BTRFS_DELAYED_INSERTION_ITEM && + ins->key.offset >= delayed_node->index_cnt) + delayed_node->index_cnt = ins->key.offset + 1; + + delayed_node->count++; + atomic_inc(&delayed_node->root->fs_info->delayed_root->items); + return 0; +} + +static int __btrfs_add_delayed_insertion_item(struct btrfs_delayed_node *node, + struct btrfs_delayed_item *item) +{ + return __btrfs_add_delayed_item(node, item, + BTRFS_DELAYED_INSERTION_ITEM); +} + +static int __btrfs_add_delayed_deletion_item(struct btrfs_delayed_node *node, + struct btrfs_delayed_item *item) +{ + return __btrfs_add_delayed_item(node, item, + BTRFS_DELAYED_DELETION_ITEM); +} + +static void __btrfs_remove_delayed_item(struct btrfs_delayed_item *delayed_item) +{ + struct rb_root *root; + struct btrfs_delayed_root *delayed_root; + + delayed_root = delayed_item->delayed_node->root->fs_info->delayed_root; + + BUG_ON(!delayed_root); + BUG_ON(delayed_item->ins_or_del != BTRFS_DELAYED_DELETION_ITEM && + delayed_item->ins_or_del != BTRFS_DELAYED_INSERTION_ITEM); + + if (delayed_item->ins_or_del == BTRFS_DELAYED_INSERTION_ITEM) + root = &delayed_item->delayed_node->ins_root; + else + root = &delayed_item->delayed_node->del_root; + + rb_erase(&delayed_item->rb_node, root); + delayed_item->delayed_node->count--; + atomic_dec(&delayed_root->items); + if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND && + waitqueue_active(&delayed_root->wait)) + wake_up(&delayed_root->wait); +} + +static void btrfs_release_delayed_item(struct btrfs_delayed_item *item) +{ + if (item) { + __btrfs_remove_delayed_item(item); + if (atomic_dec_and_test(&item->refs)) + kfree(item); + } +} + +struct btrfs_delayed_item *__btrfs_first_delayed_insertion_item( + struct btrfs_delayed_node *delayed_node) +{ + struct rb_node *p; + struct btrfs_delayed_item *item = NULL; + + p = rb_first(&delayed_node->ins_root); + if (p) + item = rb_entry(p, struct btrfs_delayed_item, rb_node); + + return item; +} + +struct btrfs_delayed_item *__btrfs_first_delayed_deletion_item( + struct btrfs_delayed_node *delayed_node) +{ + struct rb_node *p; + struct btrfs_delayed_item *item = NULL; + + p = rb_first(&delayed_node->del_root); + if (p) + item = rb_entry(p, struct btrfs_delayed_item, rb_node); + + return item; +} + +struct btrfs_delayed_item *__btrfs_next_delayed_item( + struct btrfs_delayed_item *item) +{ + struct rb_node *p; + struct btrfs_delayed_item *next = NULL; + + p = rb_next(&item->rb_node); + if (p) + next = rb_entry(p, struct btrfs_delayed_item, rb_node); + + return next; +} + +static inline struct btrfs_delayed_node *btrfs_get_delayed_node( + struct inode *inode) +{ + struct btrfs_inode *btrfs_inode = BTRFS_I(inode); + struct btrfs_delayed_node *delayed_node; + + delayed_node = btrfs_inode->delayed_node; + if (delayed_node) + atomic_inc(&delayed_node->refs); + + return delayed_node; +} + +static inline struct btrfs_root *btrfs_get_fs_root(struct btrfs_root *root, + u64 root_id) +{ + struct btrfs_key root_key; + + if (root->objectid == root_id) + return root; + + root_key.objectid = root_id; + root_key.type = BTRFS_ROOT_ITEM_KEY; + root_key.offset = (u64)-1; + return btrfs_read_fs_root_no_name(root->fs_info, &root_key); +} + +static int btrfs_delayed_item_reserve_metadata(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_delayed_item *item) +{ + struct btrfs_block_rsv *src_rsv; + struct btrfs_block_rsv *dst_rsv; + u64 num_bytes; + int ret; + + if (!trans->bytes_reserved) + return 0; + + src_rsv = trans->block_rsv; + dst_rsv = &root->fs_info->global_block_rsv; + + num_bytes = btrfs_calc_trans_metadata_size(root, 1); + ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes); + if (!ret) { + item->bytes_reserved = num_bytes; + item->block_rsv = dst_rsv; + } + + return ret; +} + +static void btrfs_delayed_item_release_metadata(struct btrfs_root *root, + struct btrfs_delayed_item *item) +{ + if (!item->bytes_reserved) + return; + + btrfs_block_rsv_release(root, item->block_rsv, + item->bytes_reserved); +} + +static int btrfs_delayed_inode_reserve_metadata( + struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_delayed_node *node) +{ + struct btrfs_block_rsv *src_rsv; + struct btrfs_block_rsv *dst_rsv; + u64 num_bytes; + int ret; + + if (!trans->bytes_reserved) + return 0; + + src_rsv = trans->block_rsv; + dst_rsv = &root->fs_info->global_block_rsv; + + num_bytes = btrfs_calc_trans_metadata_size(root, 1); + ret = btrfs_block_rsv_migrate(src_rsv, dst_rsv, num_bytes); + if (!ret) + node->bytes_reserved = num_bytes; + + return ret; +} + +static void btrfs_delayed_inode_release_metadata(struct btrfs_root *root, + struct btrfs_delayed_node *node) +{ + struct btrfs_block_rsv *rsv; + + if (!node->bytes_reserved) + return; + + rsv = &root->fs_info->global_block_rsv; + btrfs_block_rsv_release(root, rsv, + node->bytes_reserved); + node->bytes_reserved = 0; +} + +/* + * This helper will insert some continuous items into the same leaf according + * to the free space of the leaf. + */ +static int btrfs_batch_insert_items(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct btrfs_delayed_item *item) +{ + struct btrfs_delayed_item *curr, *next; + int free_space; + int total_data_size = 0, total_size = 0; + struct extent_buffer *leaf; + char *data_ptr; + struct btrfs_key *keys; + u32 *data_size; + struct list_head head; + int slot; + int nitems; + int i; + int ret = 0; + + BUG_ON(!path->nodes[0]); + + leaf = path->nodes[0]; + free_space = btrfs_leaf_free_space(root, leaf); + INIT_LIST_HEAD(&head); + + next = item; + + /* + * count the number of the continuous items that we can insert in batch + */ + while (total_size + next->data_len + sizeof(struct btrfs_item) <= + free_space) { + total_data_size += next->data_len; + total_size += next->data_len + sizeof(struct btrfs_item); + list_add_tail(&next->tree_list, &head); + nitems++; + + curr = next; + next = __btrfs_next_delayed_item(curr); + if (!next) + break; + + if (!btrfs_is_continuous_delayed_item(curr, next)) + break; + } + + if (!nitems) { + ret = 0; + goto out; + } + + /* + * we need allocate some memory space, but it might cause the task + * to sleep, so we set all locked nodes in the path to blocking locks + * first. + */ + btrfs_set_path_blocking(path); + + keys = kmalloc(sizeof(struct btrfs_key) * nitems, GFP_NOFS); + if (!keys) { + ret = -ENOMEM; + goto out; + } + + data_size = kmalloc(sizeof(u32) * nitems, GFP_NOFS); + if (!data_size) { + ret = -ENOMEM; + goto error; + } + + /* get keys of all the delayed items */ + i = 0; + list_for_each_entry(next, &head, tree_list) { + keys[i] = next->key; + data_size[i] = next->data_len; + i++; + } + + /* reset all the locked nodes in the patch to spinning locks. */ + btrfs_clear_path_blocking(path, NULL); + + /* insert the keys of the items */ + ret = setup_items_for_insert(trans, root, path, keys, data_size, + total_data_size, total_size, nitems); + if (ret) + goto error; + + /* insert the dir index items */ + slot = path->slots[0]; + list_for_each_entry_safe(curr, next, &head, tree_list) { + data_ptr = btrfs_item_ptr(leaf, slot, char); + write_extent_buffer(leaf, &curr->data, + (unsigned long)data_ptr, + curr->data_len); + slot++; + + btrfs_delayed_item_release_metadata(root, curr); + + list_del(&curr->tree_list); + btrfs_release_delayed_item(curr); + } + +error: + kfree(data_size); + kfree(keys); +out: + return ret; +} + +/* + * This helper can just do simple insertion that needn't extend item for new + * data, such as directory name index insertion, inode insertion. + */ +static int btrfs_insert_delayed_item(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct btrfs_delayed_item *delayed_item) +{ + struct extent_buffer *leaf; + struct btrfs_item *item; + char *ptr; + int ret; + + ret = btrfs_insert_empty_item(trans, root, path, &delayed_item->key, + delayed_item->data_len); + if (ret < 0 && ret != -EEXIST) + return ret; + + leaf = path->nodes[0]; + + item = btrfs_item_nr(leaf, path->slots[0]); + ptr = btrfs_item_ptr(leaf, path->slots[0], char); + + write_extent_buffer(leaf, delayed_item->data, (unsigned long)ptr, + delayed_item->data_len); + btrfs_mark_buffer_dirty(leaf); + + btrfs_delayed_item_release_metadata(root, delayed_item); + return 0; +} + +/* + * we insert an item first, then if there are some continuous items, we try + * to insert those items into the same leaf. + */ +static int btrfs_insert_delayed_items(struct btrfs_trans_handle *trans, + struct btrfs_path *path, + struct btrfs_root *root, + struct btrfs_delayed_node *node) +{ + struct btrfs_delayed_item *curr, *prev; + int ret = 0; + +do_again: + mutex_lock(&node->mutex); + curr = __btrfs_first_delayed_insertion_item(node); + if (!curr) + goto insert_end; + + ret = btrfs_insert_delayed_item(trans, root, path, curr); + if (ret < 0) { + btrfs_release_path(path); + goto insert_end; + } + + prev = curr; + curr = __btrfs_next_delayed_item(prev); + if (curr && btrfs_is_continuous_delayed_item(prev, curr)) { + /* insert the continuous items into the same leaf */ + path->slots[0]++; + btrfs_batch_insert_items(trans, root, path, curr); + } + btrfs_release_delayed_item(prev); + btrfs_mark_buffer_dirty(path->nodes[0]); + + btrfs_release_path(path); + mutex_unlock(&node->mutex); + goto do_again; + +insert_end: + mutex_unlock(&node->mutex); + return ret; +} + +static int btrfs_batch_delete_items(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct btrfs_delayed_item *item) +{ + struct btrfs_delayed_item *curr, *next; + struct extent_buffer *leaf; + struct btrfs_key key; + struct list_head head; + int nitems, i, last_item; + int ret = 0; + + BUG_ON(!path->nodes[0]); + + leaf = path->nodes[0]; + + i = path->slots[0]; + last_item = btrfs_header_nritems(leaf) - 1; + if (i > last_item) + return -ENOENT; /* FIXME: Is errno suitable? */ + + next = item; + INIT_LIST_HEAD(&head); + btrfs_item_key_to_cpu(leaf, &key, i); + nitems = 0; + /* + * count the number of the dir index items that we can delete in batch + */ + while (btrfs_comp_cpu_keys(&next->key, &key) == 0) { + list_add_tail(&next->tree_list, &head); + nitems++; + + curr = next; + next = __btrfs_next_delayed_item(curr); + if (!next) + break; + + if (!btrfs_is_continuous_delayed_item(curr, next)) + break; + + i++; + if (i > last_item) + break; + btrfs_item_key_to_cpu(leaf, &key, i); + } + + if (!nitems) + return 0; + + ret = btrfs_del_items(trans, root, path, path->slots[0], nitems); + if (ret) + goto out; + + list_for_each_entry_safe(curr, next, &head, tree_list) { + btrfs_delayed_item_release_metadata(root, curr); + list_del(&curr->tree_list); + btrfs_release_delayed_item(curr); + } + +out: + return ret; +} + +static int btrfs_delete_delayed_items(struct btrfs_trans_handle *trans, + struct btrfs_path *path, + struct btrfs_root *root, + struct btrfs_delayed_node *node) +{ + struct btrfs_delayed_item *curr, *prev; + int ret = 0; + +do_again: + mutex_lock(&node->mutex); + curr = __btrfs_first_delayed_deletion_item(node); + if (!curr) + goto delete_fail; + + ret = btrfs_search_slot(trans, root, &curr->key, path, -1, 1); + if (ret < 0) + goto delete_fail; + else if (ret > 0) { + /* + * can't find the item which the node points to, so this node + * is invalid, just drop it. + */ + prev = curr; + curr = __btrfs_next_delayed_item(prev); + btrfs_release_delayed_item(prev); + ret = 0; + btrfs_release_path(path); + if (curr) + goto do_again; + else + goto delete_fail; + } + + btrfs_batch_delete_items(trans, root, path, curr); + btrfs_release_path(path); + mutex_unlock(&node->mutex); + goto do_again; + +delete_fail: + btrfs_release_path(path); + mutex_unlock(&node->mutex); + return ret; +} + +static void btrfs_release_delayed_inode(struct btrfs_delayed_node *delayed_node) +{ + struct btrfs_delayed_root *delayed_root; + + if (delayed_node && delayed_node->inode_dirty) { + BUG_ON(!delayed_node->root); + delayed_node->inode_dirty = 0; + delayed_node->count--; + + delayed_root = delayed_node->root->fs_info->delayed_root; + atomic_dec(&delayed_root->items); + if (atomic_read(&delayed_root->items) < + BTRFS_DELAYED_BACKGROUND && + waitqueue_active(&delayed_root->wait)) + wake_up(&delayed_root->wait); + } +} + +static int btrfs_update_delayed_inode(struct btrfs_trans_handle *trans, + struct btrfs_root *root, + struct btrfs_path *path, + struct btrfs_delayed_node *node) +{ + struct btrfs_key key; + struct btrfs_inode_item *inode_item; + struct extent_buffer *leaf; + int ret; + + mutex_lock(&node->mutex); + if (!node->inode_dirty) { + mutex_unlock(&node->mutex); + return 0; + } + + key.objectid = node->inode_id; + btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY); + key.offset = 0; + ret = btrfs_lookup_inode(trans, root, path, &key, 1); + if (ret > 0) { + btrfs_release_path(path); + mutex_unlock(&node->mutex); + return -ENOENT; + } else if (ret < 0) { + mutex_unlock(&node->mutex); + return ret; + } + + btrfs_unlock_up_safe(path, 1); + leaf = path->nodes[0]; + inode_item = btrfs_item_ptr(leaf, path->slots[0], + struct btrfs_inode_item); + write_extent_buffer(leaf, &node->inode_item, (unsigned long)inode_item, + sizeof(struct btrfs_inode_item)); + btrfs_mark_buffer_dirty(leaf); + btrfs_release_path(path); + + btrfs_delayed_inode_release_metadata(root, node); + btrfs_release_delayed_inode(node); + mutex_unlock(&node->mutex); + + return 0; +} + +/* Called when committing the transaction. */ +int btrfs_run_delayed_items(struct btrfs_trans_handle *trans, + struct btrfs_root *root) +{ + struct btrfs_delayed_root *delayed_root; + struct btrfs_delayed_node *curr_node, *prev_node; + struct btrfs_path *path; + int ret = 0; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + path->leave_spinning = 1; + + delayed_root = btrfs_get_delayed_root(root); + + curr_node = btrfs_first_delayed_node(delayed_root); + while (curr_node) { + root = curr_node->root; + ret = btrfs_insert_delayed_items(trans, path, root, + curr_node); + if (!ret) + ret = btrfs_delete_delayed_items(trans, path, root, + curr_node); + if (!ret) + ret = btrfs_update_delayed_inode(trans, root, path, + curr_node); + if (ret) { + btrfs_release_delayed_node(curr_node); + break; + } + + prev_node = curr_node; + curr_node = btrfs_next_delayed_node(curr_node); + btrfs_release_delayed_node(prev_node); + } + + btrfs_free_path(path); + return ret; +} + +static int __btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, + struct btrfs_delayed_node *node) +{ + struct btrfs_path *path; + int ret; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + path->leave_spinning = 1; + + ret = btrfs_insert_delayed_items(trans, path, node->root, node); + if (!ret) + ret = btrfs_delete_delayed_items(trans, path, node->root, node); + if (!ret) + ret = btrfs_update_delayed_inode(trans, node->root, path, node); + btrfs_free_path(path); + + return ret; +} + +int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, + struct inode *inode) +{ + struct btrfs_delayed_node *delayed_node = btrfs_get_delayed_node(inode); + int ret; + + if (!delayed_node) + return 0; + + mutex_lock(&delayed_node->mutex); + if (!delayed_node->count) { + mutex_unlock(&delayed_node->mutex); + btrfs_release_delayed_node(delayed_node); + return 0; + } + mutex_unlock(&delayed_node->mutex); + + ret = __btrfs_commit_inode_delayed_items(trans, delayed_node); + btrfs_release_delayed_node(delayed_node); + return ret; +} + +void btrfs_remove_delayed_node(struct inode *inode) +{ + struct btrfs_delayed_node *delayed_node; + + delayed_node = ACCESS_ONCE(BTRFS_I(inode)->delayed_node); + if (!delayed_node) + return; + + BTRFS_I(inode)->delayed_node = NULL; + btrfs_release_delayed_node(delayed_node); +} + +struct btrfs_async_delayed_node { + struct btrfs_root *root; + struct btrfs_delayed_node *delayed_node; + struct btrfs_work work; +}; + +static void btrfs_async_run_delayed_node_done(struct btrfs_work *work) +{ + struct btrfs_async_delayed_node *async_node; + struct btrfs_trans_handle *trans; + struct btrfs_path *path; + struct btrfs_delayed_node *delayed_node = NULL; + struct btrfs_root *root; + unsigned long nr = 0; + int need_requeue = 0; + int ret; + + async_node = container_of(work, struct btrfs_async_delayed_node, work); + + path = btrfs_alloc_path(); + if (!path) + goto out; + path->leave_spinning = 1; + + delayed_node = async_node->delayed_node; + root = delayed_node->root; + + trans = btrfs_join_transaction(root, 0); + if (IS_ERR(trans)) + goto free_path; + + ret = btrfs_insert_delayed_items(trans, path, root, delayed_node); + if (!ret) + ret = btrfs_delete_delayed_items(trans, path, root, + delayed_node); + + if (!ret) + btrfs_update_delayed_inode(trans, root, path, delayed_node); + + /* + * Maybe new delayed items have been inserted, so we need requeue + * the work. Besides that, we must dequeue the empty delayed nodes + * to avoid the race between delayed items balance and the worker. + * The race like this: + * Task1 Worker thread + * count == 0, needn't requeue + * also needn't insert the + * delayed node into prepare + * list again. + * add lots of delayed items + * queue the delayed node + * already in the list, + * and not in the prepare + * list, it means the delayed + * node is being dealt with + * by the worker. + * do delayed items balance + * the delayed node is being + * dealt with by the worker + * now, just wait. + * the worker goto idle. + * Task1 will sleep until the transaction is commited. + */ + mutex_lock(&delayed_node->mutex); + if (delayed_node->count) + need_requeue = 1; + else + btrfs_dequeue_delayed_node(root->fs_info->delayed_root, + delayed_node); + mutex_unlock(&delayed_node->mutex); + + nr = trans->blocks_used; + + btrfs_end_transaction_dmeta(trans, root); + __btrfs_btree_balance_dirty(root, nr); +free_path: + btrfs_free_path(path); +out: + if (need_requeue) + btrfs_requeue_work(&async_node->work); + else { + btrfs_release_prepared_delayed_node(delayed_node); + kfree(async_node); + } +} + +static int btrfs_wq_run_delayed_node(struct btrfs_delayed_root *delayed_root, + struct btrfs_root *root, int all) +{ + struct btrfs_async_delayed_node *async_node; + struct btrfs_delayed_node *curr; + int count = 0; + +again: + curr = btrfs_first_prepared_delayed_node(delayed_root); + if (!curr) + return 0; + + async_node = kmalloc(sizeof(*async_node), GFP_NOFS); + if (!async_node) { + btrfs_release_prepared_delayed_node(curr); + return -ENOMEM; + } + + async_node->root = root; + async_node->delayed_node = curr; + + async_node->work.func = btrfs_async_run_delayed_node_done; + async_node->work.flags = 0; + + btrfs_queue_worker(&root->fs_info->delayed_workers, &async_node->work); + count++; + + if (all || count < 4) + goto again; + + return 0; +} + +void btrfs_balance_delayed_items(struct btrfs_root *root) +{ + struct btrfs_delayed_root *delayed_root; + + delayed_root = btrfs_get_delayed_root(root); + + if (atomic_read(&delayed_root->items) < BTRFS_DELAYED_BACKGROUND) + return; + + if (atomic_read(&delayed_root->items) >= BTRFS_DELAYED_WRITEBACK) { + int ret; + ret = btrfs_wq_run_delayed_node(delayed_root, root, 1); + if (ret) + return; + + wait_event_interruptible_timeout( + delayed_root->wait, + (atomic_read(&delayed_root->items) < + BTRFS_DELAYED_BACKGROUND), + HZ); + return; + } + + btrfs_wq_run_delayed_node(delayed_root, root, 0); +} + +int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans, + struct btrfs_root *root, const char *name, + int name_len, struct inode *dir, + struct btrfs_disk_key *disk_key, u8 type, + u64 index) +{ + struct btrfs_delayed_node *delayed_node; + struct btrfs_delayed_item *delayed_item; + struct btrfs_dir_item *dir_item; + int ret; + + delayed_node = btrfs_get_or_create_delayed_node(dir); + if (IS_ERR(delayed_node)) + return PTR_ERR(delayed_node); + + delayed_item = btrfs_alloc_delayed_item(sizeof(*dir_item) + name_len); + if (!delayed_item) { + ret = -ENOMEM; + goto release_node; + } + + ret = btrfs_delayed_item_reserve_metadata(trans, root, delayed_item); + /* + * we have reserved enough space when we start a new transaction, + * so reserving metadata failure is impossible + */ + BUG_ON(ret); + + delayed_item->key.objectid = btrfs_ino(dir); + btrfs_set_key_type(&delayed_item->key, BTRFS_DIR_INDEX_KEY); + delayed_item->key.offset = index; + + dir_item = (struct btrfs_dir_item *)delayed_item->data; + dir_item->location = *disk_key; + dir_item->transid = cpu_to_le64(trans->transid); + dir_item->data_len = 0; + dir_item->name_len = cpu_to_le16(name_len); + dir_item->type = type; + memcpy((char *)(dir_item + 1), name, name_len); + + mutex_lock(&delayed_node->mutex); + ret = __btrfs_add_delayed_insertion_item(delayed_node, delayed_item); + if (unlikely(ret)) { + printk(KERN_ERR "err add delayed dir index item(name: %s) into " + "the insertion tree of the delayed node" + "(root id: %llu, inode id: %llu, errno: %d)\n", + name, + (unsigned long long)delayed_node->root->objectid, + (unsigned long long)delayed_node->inode_id, + ret); + BUG(); + } + mutex_unlock(&delayed_node->mutex); + +release_node: + btrfs_release_delayed_node(delayed_node); + return ret; +} + +static int btrfs_delete_delayed_insertion_item(struct btrfs_root *root, + struct btrfs_delayed_node *node, + struct btrfs_key *key) +{ + struct btrfs_delayed_item *item; + + mutex_lock(&node->mutex); + item = __btrfs_lookup_delayed_insertion_item(node, key); + if (!item) { + mutex_unlock(&node->mutex); + return 1; + } + + btrfs_delayed_item_release_metadata(root, item); + btrfs_release_delayed_item(item); + mutex_unlock(&node->mutex); + return 0; +} + +int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans, + struct btrfs_root *root, struct inode *dir, + u64 index) +{ + struct btrfs_delayed_node *node; + struct btrfs_delayed_item *item; + struct btrfs_key item_key; + int ret; + + node = btrfs_get_or_create_delayed_node(dir); + if (IS_ERR(node)) + return PTR_ERR(node); + + item_key.objectid = btrfs_ino(dir); + btrfs_set_key_type(&item_key, BTRFS_DIR_INDEX_KEY); + item_key.offset = index; + + ret = btrfs_delete_delayed_insertion_item(root, node, &item_key); + if (!ret) + goto end; + + item = btrfs_alloc_delayed_item(0); + if (!item) { + ret = -ENOMEM; + goto end; + } + + item->key = item_key; + + ret = btrfs_delayed_item_reserve_metadata(trans, root, item); + /* + * we have reserved enough space when we start a new transaction, + * so reserving metadata failure is impossible. + */ + BUG_ON(ret); + + mutex_lock(&node->mutex); + ret = __btrfs_add_delayed_deletion_item(node, item); + if (unlikely(ret)) { + printk(KERN_ERR "err add delayed dir index item(index: %llu) " + "into the deletion tree of the delayed node" + "(root id: %llu, inode id: %llu, errno: %d)\n", + (unsigned long long)index, + (unsigned long long)node->root->objectid, + (unsigned long long)node->inode_id, + ret); + BUG(); + } + mutex_unlock(&node->mutex); +end: + btrfs_release_delayed_node(node); + return ret; +} + +int btrfs_inode_delayed_dir_index_count(struct inode *inode) +{ + struct btrfs_delayed_node *delayed_node = BTRFS_I(inode)->delayed_node; + int ret = 0; + + if (!delayed_node) + return -ENOENT; + + /* + * Since we have held i_mutex of this directory, it is impossible that + * a new directory index is added into the delayed node and index_cnt + * is updated now. So we needn't lock the delayed node. + */ + if (!delayed_node->index_cnt) + return -EINVAL; + + BTRFS_I(inode)->index_cnt = delayed_node->index_cnt; + return ret; +} + +void btrfs_get_delayed_items(struct inode *inode, struct list_head *ins_list, + struct list_head *del_list) +{ + struct btrfs_delayed_node *delayed_node; + struct btrfs_delayed_item *item; + + delayed_node = btrfs_get_delayed_node(inode); + if (!delayed_node) + return; + + mutex_lock(&delayed_node->mutex); + item = __btrfs_first_delayed_insertion_item(delayed_node); + while (item) { + atomic_inc(&item->refs); + list_add_tail(&item->readdir_list, ins_list); + item = __btrfs_next_delayed_item(item); + } + + item = __btrfs_first_delayed_deletion_item(delayed_node); + while (item) { + atomic_inc(&item->refs); + list_add_tail(&item->readdir_list, del_list); + item = __btrfs_next_delayed_item(item); + } + mutex_unlock(&delayed_node->mutex); + /* + * This delayed node is still cached in the btrfs inode, so refs + * must be > 1 now, and we needn't check it is going to be freed + * or not. + * + * Besides that, this function is used to read dir, we do not + * insert/delete delayed items in this period. So we also needn't + * requeue or dequeue this delayed node. + */ + atomic_dec(&delayed_node->refs); +} + +void btrfs_put_delayed_items(struct list_head *ins_list, + struct list_head *del_list) +{ + struct btrfs_delayed_item *curr, *next; + + list_for_each_entry_safe(curr, next, ins_list, readdir_list) { + list_del(&curr->readdir_list); + if (atomic_dec_and_test(&curr->refs)) + kfree(curr); + } + + list_for_each_entry_safe(curr, next, del_list, readdir_list) { + list_del(&curr->readdir_list); + if (atomic_dec_and_test(&curr->refs)) + kfree(curr); + } +} + +int btrfs_should_delete_dir_index(struct list_head *del_list, + u64 index) +{ + struct btrfs_delayed_item *curr, *next; + int ret; + + if (list_empty(del_list)) + return 0; + + list_for_each_entry_safe(curr, next, del_list, readdir_list) { + if (curr->key.offset > index) + break; + + list_del(&curr->readdir_list); + ret = (curr->key.offset == index); + + if (atomic_dec_and_test(&curr->refs)) + kfree(curr); + + if (ret) + return 1; + else + continue; + } + return 0; +} + +/* + * btrfs_readdir_delayed_dir_index - read dir info stored in the delayed tree + * + */ +int btrfs_readdir_delayed_dir_index(struct file *filp, void *dirent, + filldir_t filldir, + struct list_head *ins_list) +{ + struct btrfs_dir_item *di; + struct btrfs_delayed_item *curr, *next; + struct btrfs_key location; + char *name; + int name_len; + int over = 0; + unsigned char d_type; + + if (list_empty(ins_list)) + return 0; + + /* + * Changing the data of the delayed item is impossible. So + * we needn't lock them. And we have held i_mutex of the + * directory, nobody can delete any directory indexes now. + */ + list_for_each_entry_safe(curr, next, ins_list, readdir_list) { + list_del(&curr->readdir_list); + + if (curr->key.offset < filp->f_pos) { + if (atomic_dec_and_test(&curr->refs)) + kfree(curr); + continue; + } + + filp->f_pos = curr->key.offset; + + di = (struct btrfs_dir_item *)curr->data; + name = (char *)(di + 1); + name_len = le16_to_cpu(di->name_len); + + d_type = btrfs_filetype_table[di->type]; + btrfs_disk_key_to_cpu(&location, &di->location); + + over = filldir(dirent, name, name_len, curr->key.offset, + location.objectid, d_type); + + if (atomic_dec_and_test(&curr->refs)) + kfree(curr); + + if (over) + return 1; + } + return 0; +} + +BTRFS_SETGET_STACK_FUNCS(stack_inode_generation, struct btrfs_inode_item, + generation, 64); +BTRFS_SETGET_STACK_FUNCS(stack_inode_sequence, struct btrfs_inode_item, + sequence, 64); +BTRFS_SETGET_STACK_FUNCS(stack_inode_transid, struct btrfs_inode_item, + transid, 64); +BTRFS_SETGET_STACK_FUNCS(stack_inode_size, struct btrfs_inode_item, size, 64); +BTRFS_SETGET_STACK_FUNCS(stack_inode_nbytes, struct btrfs_inode_item, + nbytes, 64); +BTRFS_SETGET_STACK_FUNCS(stack_inode_block_group, struct btrfs_inode_item, + block_group, 64); +BTRFS_SETGET_STACK_FUNCS(stack_inode_nlink, struct btrfs_inode_item, nlink, 32); +BTRFS_SETGET_STACK_FUNCS(stack_inode_uid, struct btrfs_inode_item, uid, 32); +BTRFS_SETGET_STACK_FUNCS(stack_inode_gid, struct btrfs_inode_item, gid, 32); +BTRFS_SETGET_STACK_FUNCS(stack_inode_mode, struct btrfs_inode_item, mode, 32); +BTRFS_SETGET_STACK_FUNCS(stack_inode_rdev, struct btrfs_inode_item, rdev, 64); +BTRFS_SETGET_STACK_FUNCS(stack_inode_flags, struct btrfs_inode_item, flags, 64); + +BTRFS_SETGET_STACK_FUNCS(stack_timespec_sec, struct btrfs_timespec, sec, 64); +BTRFS_SETGET_STACK_FUNCS(stack_timespec_nsec, struct btrfs_timespec, nsec, 32); + +static void fill_stack_inode_item(struct btrfs_trans_handle *trans, + struct btrfs_inode_item *inode_item, + struct inode *inode) +{ + btrfs_set_stack_inode_uid(inode_item, inode->i_uid); + btrfs_set_stack_inode_gid(inode_item, inode->i_gid); + btrfs_set_stack_inode_size(inode_item, BTRFS_I(inode)->disk_i_size); + btrfs_set_stack_inode_mode(inode_item, inode->i_mode); + btrfs_set_stack_inode_nlink(inode_item, inode->i_nlink); + btrfs_set_stack_inode_nbytes(inode_item, inode_get_bytes(inode)); + btrfs_set_stack_inode_generation(inode_item, + BTRFS_I(inode)->generation); + btrfs_set_stack_inode_sequence(inode_item, BTRFS_I(inode)->sequence); + btrfs_set_stack_inode_transid(inode_item, trans->transid); + btrfs_set_stack_inode_rdev(inode_item, inode->i_rdev); + btrfs_set_stack_inode_flags(inode_item, BTRFS_I(inode)->flags); + btrfs_set_stack_inode_block_group(inode_item, + BTRFS_I(inode)->block_group); + + btrfs_set_stack_timespec_sec(btrfs_inode_atime(inode_item), + inode->i_atime.tv_sec); + btrfs_set_stack_timespec_nsec(btrfs_inode_atime(inode_item), + inode->i_atime.tv_nsec); + + btrfs_set_stack_timespec_sec(btrfs_inode_mtime(inode_item), + inode->i_mtime.tv_sec); + btrfs_set_stack_timespec_nsec(btrfs_inode_mtime(inode_item), + inode->i_mtime.tv_nsec); + + btrfs_set_stack_timespec_sec(btrfs_inode_ctime(inode_item), + inode->i_ctime.tv_sec); + btrfs_set_stack_timespec_nsec(btrfs_inode_ctime(inode_item), + inode->i_ctime.tv_nsec); +} + +int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans, + struct btrfs_root *root, struct inode *inode) +{ + struct btrfs_delayed_node *delayed_node; + int ret; + + delayed_node = btrfs_get_or_create_delayed_node(inode); + if (IS_ERR(delayed_node)) + return PTR_ERR(delayed_node); + + mutex_lock(&delayed_node->mutex); + if (delayed_node->inode_dirty) { + fill_stack_inode_item(trans, &delayed_node->inode_item, inode); + goto release_node; + } + + ret = btrfs_delayed_inode_reserve_metadata(trans, root, delayed_node); + /* + * we must reserve enough space when we start a new transaction, + * so reserving metadata failure is impossible + */ + BUG_ON(ret); + + fill_stack_inode_item(trans, &delayed_node->inode_item, inode); + delayed_node->inode_dirty = 1; + delayed_node->count++; + atomic_inc(&root->fs_info->delayed_root->items); +release_node: + mutex_unlock(&delayed_node->mutex); + btrfs_release_delayed_node(delayed_node); + return ret; +} + +static void __btrfs_kill_delayed_node(struct btrfs_delayed_node *delayed_node) +{ + struct btrfs_root *root = delayed_node->root; + struct btrfs_delayed_item *curr_item, *prev_item; + + mutex_lock(&delayed_node->mutex); + curr_item = __btrfs_first_delayed_insertion_item(delayed_node); + while (curr_item) { + btrfs_delayed_item_release_metadata(root, curr_item); + prev_item = curr_item; + curr_item = __btrfs_next_delayed_item(prev_item); + btrfs_release_delayed_item(prev_item); + } + + curr_item = __btrfs_first_delayed_deletion_item(delayed_node); + while (curr_item) { + btrfs_delayed_item_release_metadata(root, curr_item); + prev_item = curr_item; + curr_item = __btrfs_next_delayed_item(prev_item); + btrfs_release_delayed_item(prev_item); + } + + if (delayed_node->inode_dirty) { + btrfs_delayed_inode_release_metadata(root, delayed_node); + btrfs_release_delayed_inode(delayed_node); + } + mutex_unlock(&delayed_node->mutex); +} + +void btrfs_kill_delayed_inode_items(struct inode *inode) +{ + struct btrfs_delayed_node *delayed_node; + + delayed_node = btrfs_get_delayed_node(inode); + if (!delayed_node) + return; + + __btrfs_kill_delayed_node(delayed_node); + btrfs_release_delayed_node(delayed_node); +} + +void btrfs_kill_all_delayed_nodes(struct btrfs_root *root) +{ + u64 inode_id = 0; + struct btrfs_delayed_node *delayed_nodes[8]; + int i, n; + + while (1) { + spin_lock(&root->inode_lock); + n = radix_tree_gang_lookup(&root->delayed_nodes_tree, + (void **)delayed_nodes, inode_id, + ARRAY_SIZE(delayed_nodes)); + if (!n) { + spin_unlock(&root->inode_lock); + break; + } + + inode_id = delayed_nodes[n - 1]->inode_id + 1; + + for (i = 0; i < n; i++) + atomic_inc(&delayed_nodes[i]->refs); + spin_unlock(&root->inode_lock); + + for (i = 0; i < n; i++) { + __btrfs_kill_delayed_node(delayed_nodes[i]); + btrfs_release_delayed_node(delayed_nodes[i]); + } + } +} diff --git a/fs/btrfs/delayed-inode.h b/fs/btrfs/delayed-inode.h new file mode 100644 index 000000000000..eb7d240aa648 --- /dev/null +++ b/fs/btrfs/delayed-inode.h @@ -0,0 +1,141 @@ +/* + * Copyright (C) 2011 Fujitsu. All rights reserved. + * Written by Miao Xie <miaox@cn.fujitsu.com> + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License v2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public + * License along with this program; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 021110-1307, USA. + */ + +#ifndef __DELAYED_TREE_OPERATION_H +#define __DELAYED_TREE_OPERATION_H + +#include <linux/rbtree.h> +#include <linux/spinlock.h> +#include <linux/mutex.h> +#include <linux/list.h> +#include <linux/wait.h> +#include <asm/atomic.h> + +#include "ctree.h" + +/* types of the delayed item */ +#define BTRFS_DELAYED_INSERTION_ITEM 1 +#define BTRFS_DELAYED_DELETION_ITEM 2 + +struct btrfs_delayed_root { + spinlock_t lock; + struct list_head node_list; + /* + * Used for delayed nodes which is waiting to be dealt with by the + * worker. If the delayed node is inserted into the work queue, we + * drop it from this list. + */ + struct list_head prepare_list; + atomic_t items; /* for delayed items */ + int nodes; /* for delayed nodes */ + wait_queue_head_t wait; +}; + +struct btrfs_delayed_node { + u64 inode_id; + u64 bytes_reserved; + struct btrfs_root *root; + /* Used to add the node into the delayed root's node list. */ + struct list_head n_list; + /* + * Used to add the node into the prepare list, the nodes in this list + * is waiting to be dealt with by the async worker. + */ + struct list_head p_list; + struct rb_root ins_root; + struct rb_root del_root; + struct mutex mutex; + struct btrfs_inode_item inode_item; + atomic_t refs; + u64 index_cnt; + bool in_list; + bool inode_dirty; + int count; +}; + +struct btrfs_delayed_item { + struct rb_node rb_node; + struct btrfs_key key; + struct list_head tree_list; /* used for batch insert/delete items */ + struct list_head readdir_list; /* used for readdir items */ + u64 bytes_reserved; + struct btrfs_block_rsv *block_rsv; + struct btrfs_delayed_node *delayed_node; + atomic_t refs; + int ins_or_del; + u32 data_len; + char data[0]; +}; + +static inline void btrfs_init_delayed_root( + struct btrfs_delayed_root *delayed_root) +{ + atomic_set(&delayed_root->items, 0); + delayed_root->nodes = 0; + spin_lock_init(&delayed_root->lock); + init_waitqueue_head(&delayed_root->wait); + INIT_LIST_HEAD(&delayed_root->node_list); + INIT_LIST_HEAD(&delayed_root->prepare_list); +} + +int btrfs_insert_delayed_dir_index(struct btrfs_trans_handle *trans, + struct btrfs_root *root, const char *name, + int name_len, struct inode *dir, + struct btrfs_disk_key *disk_key, u8 type, + u64 index); + +int btrfs_delete_delayed_dir_index(struct btrfs_trans_handle *trans, + struct btrfs_root *root, struct inode *dir, + u64 index); + +int btrfs_inode_delayed_dir_index_count(struct inode *inode); + +int btrfs_run_delayed_items(struct btrfs_trans_handle *trans, + struct btrfs_root *root); + +void btrfs_balance_delayed_items(struct btrfs_root *root); + +int btrfs_commit_inode_delayed_items(struct btrfs_trans_handle *trans, + struct inode *inode); +/* Used for evicting the inode. */ +void btrfs_remove_delayed_node(struct inode *inode); +void btrfs_kill_delayed_inode_items(struct inode *inode); + + +int btrfs_delayed_update_inode(struct btrfs_trans_handle *trans, + struct btrfs_root *root, struct inode *inode); + +/* Used for drop dead root */ +void btrfs_kill_all_delayed_nodes(struct btrfs_root *root); + +/* Used for readdir() */ +void btrfs_get_delayed_items(struct inode *inode, struct list_head *ins_list, + struct list_head *del_list); +void btrfs_put_delayed_items(struct list_head *ins_list, + struct list_head *del_list); +int btrfs_should_delete_dir_index(struct list_head *del_list, + u64 index); +int btrfs_readdir_delayed_dir_index(struct file *filp, void *dirent, + filldir_t filldir, + struct list_head *ins_list); + +/* for init */ +int __init btrfs_delayed_inode_init(void); +void btrfs_delayed_inode_exit(void); +#endif diff --git a/fs/btrfs/delayed-ref.c b/fs/btrfs/delayed-ref.c index bce28f653899..125cf76fcd08 100644 --- a/fs/btrfs/delayed-ref.c +++ b/fs/btrfs/delayed-ref.c @@ -281,44 +281,6 @@ again: } /* - * This checks to see if there are any delayed refs in the - * btree for a given bytenr. It returns one if it finds any - * and zero otherwise. - * - * If it only finds a head node, it returns 0. - * - * The idea is to use this when deciding if you can safely delete an - * extent from the extent allocation tree. There may be a pending - * ref in the rbtree that adds or removes references, so as long as this - * returns one you need to leave the BTRFS_EXTENT_ITEM in the extent - * allocation tree. - */ -int btrfs_delayed_ref_pending(struct btrfs_trans_handle *trans, u64 bytenr) -{ - struct btrfs_delayed_ref_node *ref; - struct btrfs_delayed_ref_root *delayed_refs; - struct rb_node *prev_node; - int ret = 0; - - delayed_refs = &trans->transaction->delayed_refs; - spin_lock(&delayed_refs->lock); - - ref = find_ref_head(&delayed_refs->root, bytenr, NULL); - if (ref) { - prev_node = rb_prev(&ref->rb_node); - if (!prev_node) - goto out; - ref = rb_entry(prev_node, struct btrfs_delayed_ref_node, - rb_node); - if (ref->bytenr == bytenr) - ret = 1; - } -out: - spin_unlock(&delayed_refs->lock); - return ret; -} - -/* * helper function to update an extent delayed ref in the * rbtree. existing and update must both have the same * bytenr and parent @@ -747,79 +709,3 @@ btrfs_find_delayed_ref_head(struct btrfs_trans_handle *trans, u64 bytenr) return btrfs_delayed_node_to_head(ref); return NULL; } - -/* - * add a delayed ref to the tree. This does all of the accounting required - * to make sure the delayed ref is eventually processed before this - * transaction commits. - * - * The main point of this call is to add and remove a backreference in a single - * shot, taking the lock only once, and only searching for the head node once. - * - * It is the same as doing a ref add and delete in two separate calls. - */ -#if 0 -int btrfs_update_delayed_ref(struct btrfs_trans_handle *trans, - u64 bytenr, u64 num_bytes, u64 orig_parent, - u64 parent, u64 orig_ref_root, u64 ref_root, - u64 orig_ref_generation, u64 ref_generation, - u64 owner_objectid, int pin) -{ - struct btrfs_delayed_ref *ref; - struct btrfs_delayed_ref *old_ref; - struct btrfs_delayed_ref_head *head_ref; - struct btrfs_delayed_ref_root *delayed_refs; - int ret; - - ref = kmalloc(sizeof(*ref), GFP_NOFS); - if (!ref) - return -ENOMEM; - - old_ref = kmalloc(sizeof(*old_ref), GFP_NOFS); - if (!old_ref) { - kfree(ref); - return -ENOMEM; - } - - /* - * the parent = 0 case comes from cases where we don't actually - * know the parent yet. It will get updated later via a add/drop - * pair. - */ - if (parent == 0) - parent = bytenr; - if (orig_parent == 0) - orig_parent = bytenr; - - head_ref = kmalloc(sizeof(*head_ref), GFP_NOFS); - if (!head_ref) { - kfree(ref); - kfree(old_ref); - return -ENOMEM; - } - delayed_refs = &trans->transaction->delayed_refs; - spin_lock(&delayed_refs->lock); - - /* - * insert both the head node and the new ref without dropping - * the spin lock - */ - ret = __btrfs_add_delayed_ref(trans, &head_ref->node, bytenr, num_bytes, - (u64)-1, 0, 0, 0, - BTRFS_UPDATE_DELAYED_HEAD, 0); - BUG_ON(ret); - - ret = __btrfs_add_delayed_ref(trans, &ref->node, bytenr, num_bytes, - parent, ref_root, ref_generation, - owner_objectid, BTRFS_ADD_DELAYED_REF, 0); - BUG_ON(ret); - - ret = __btrfs_add_delayed_ref(trans, &old_ref->node, bytenr, num_bytes, - orig_parent, orig_ref_root, - orig_ref_generation, owner_objectid, - BTRFS_DROP_DELAYED_REF, pin); - BUG_ON(ret); - spin_unlock(&delayed_refs->lock); - return 0; -} -#endif diff --git a/fs/btrfs/delayed-ref.h b/fs/btrfs/delayed-ref.h index 50e3cf92fbda..e287e3b0eab0 100644 --- a/fs/btrfs/delayed-ref.h +++ b/fs/btrfs/delayed-ref.h @@ -166,12 +166,6 @@ int btrfs_add_delayed_extent_op(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head * btrfs_find_delayed_ref_head(struct btrfs_trans_handle *trans, u64 bytenr); -int btrfs_delayed_ref_pending(struct btrfs_trans_handle *trans, u64 bytenr); -int btrfs_update_delayed_ref(struct btrfs_trans_handle *trans, - u64 bytenr, u64 num_bytes, u64 orig_parent, - u64 parent, u64 orig_ref_root, u64 ref_root, - u64 orig_ref_generation, u64 ref_generation, - u64 owner_objectid, int pin); int btrfs_delayed_ref_lock(struct btrfs_trans_handle *trans, struct btrfs_delayed_ref_head *head); int btrfs_find_ref_cluster(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/dir-item.c b/fs/btrfs/dir-item.c index c62f02f6ae69..685f2593c4f0 100644 --- a/fs/btrfs/dir-item.c +++ b/fs/btrfs/dir-item.c @@ -50,7 +50,6 @@ static struct btrfs_dir_item *insert_with_overflow(struct btrfs_trans_handle if (di) return ERR_PTR(-EEXIST); ret = btrfs_extend_item(trans, root, path, data_size); - WARN_ON(ret > 0); } if (ret < 0) return ERR_PTR(ret); @@ -124,8 +123,9 @@ int btrfs_insert_xattr_item(struct btrfs_trans_handle *trans, * to use for the second index (if one is created). */ int btrfs_insert_dir_item(struct btrfs_trans_handle *trans, struct btrfs_root - *root, const char *name, int name_len, u64 dir, - struct btrfs_key *location, u8 type, u64 index) + *root, const char *name, int name_len, + struct inode *dir, struct btrfs_key *location, + u8 type, u64 index) { int ret = 0; int ret2 = 0; @@ -137,13 +137,17 @@ int btrfs_insert_dir_item(struct btrfs_trans_handle *trans, struct btrfs_root struct btrfs_disk_key disk_key; u32 data_size; - key.objectid = dir; + key.objectid = btrfs_ino(dir); btrfs_set_key_type(&key, BTRFS_DIR_ITEM_KEY); key.offset = btrfs_name_hash(name, name_len); path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; path->leave_spinning = 1; + btrfs_cpu_key_to_disk(&disk_key, location); + data_size = sizeof(*dir_item) + name_len; dir_item = insert_with_overflow(trans, root, path, &key, data_size, name, name_len); @@ -155,7 +159,6 @@ int btrfs_insert_dir_item(struct btrfs_trans_handle *trans, struct btrfs_root } leaf = path->nodes[0]; - btrfs_cpu_key_to_disk(&disk_key, location); btrfs_set_dir_item_key(leaf, dir_item, &disk_key); btrfs_set_dir_type(leaf, dir_item, type); btrfs_set_dir_data_len(leaf, dir_item, 0); @@ -172,29 +175,11 @@ second_insert: ret = 0; goto out_free; } - btrfs_release_path(root, path); - - btrfs_set_key_type(&key, BTRFS_DIR_INDEX_KEY); - key.offset = index; - dir_item = insert_with_overflow(trans, root, path, &key, data_size, - name, name_len); - if (IS_ERR(dir_item)) { - ret2 = PTR_ERR(dir_item); - goto out_free; - } - leaf = path->nodes[0]; - btrfs_cpu_key_to_disk(&disk_key, location); - btrfs_set_dir_item_key(leaf, dir_item, &disk_key); - btrfs_set_dir_type(leaf, dir_item, type); - btrfs_set_dir_data_len(leaf, dir_item, 0); - btrfs_set_dir_name_len(leaf, dir_item, name_len); - btrfs_set_dir_transid(leaf, dir_item, trans->transid); - name_ptr = (unsigned long)(dir_item + 1); - write_extent_buffer(leaf, name, name_ptr, name_len); - btrfs_mark_buffer_dirty(leaf); + btrfs_release_path(path); + ret2 = btrfs_insert_delayed_dir_index(trans, root, name, name_len, dir, + &disk_key, type, index); out_free: - btrfs_free_path(path); if (ret) return ret; @@ -452,7 +437,7 @@ int verify_dir_item(struct btrfs_root *root, namelen = XATTR_NAME_MAX; if (btrfs_dir_name_len(leaf, dir_item) > namelen) { - printk(KERN_CRIT "btrfS: invalid dir item name len: %u\n", + printk(KERN_CRIT "btrfs: invalid dir item name len: %u\n", (unsigned)btrfs_dir_data_len(leaf, dir_item)); return 1; } diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 8f1d44ba332f..98b6a71decba 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -29,6 +29,7 @@ #include <linux/crc32c.h> #include <linux/slab.h> #include <linux/migrate.h> +#include <linux/ratelimit.h> #include <asm/unaligned.h> #include "compat.h" #include "ctree.h" @@ -41,6 +42,7 @@ #include "locking.h" #include "tree-log.h" #include "free-space-cache.h" +#include "inode-map.h" static struct extent_io_ops btree_extent_io_ops; static void end_workqueue_fn(struct btrfs_work *work); @@ -137,7 +139,7 @@ static const char *btrfs_eb_name[BTRFS_MAX_LEVEL + 1] = { * that covers the entire device */ static struct extent_map *btree_get_extent(struct inode *inode, - struct page *page, size_t page_offset, u64 start, u64 len, + struct page *page, size_t pg_offset, u64 start, u64 len, int create) { struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; @@ -154,7 +156,7 @@ static struct extent_map *btree_get_extent(struct inode *inode, } read_unlock(&em_tree->lock); - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); if (!em) { em = ERR_PTR(-ENOMEM); goto out; @@ -254,14 +256,12 @@ static int csum_tree_block(struct btrfs_root *root, struct extent_buffer *buf, memcpy(&found, result, csum_size); read_extent_buffer(buf, &val, 0, csum_size); - if (printk_ratelimit()) { - printk(KERN_INFO "btrfs: %s checksum verify " + printk_ratelimited(KERN_INFO "btrfs: %s checksum verify " "failed on %llu wanted %X found %X " "level %d\n", root->fs_info->sb->s_id, (unsigned long long)buf->start, val, found, btrfs_header_level(buf)); - } if (result != (char *)&inline_result) kfree(result); return 1; @@ -296,13 +296,11 @@ static int verify_parent_transid(struct extent_io_tree *io_tree, ret = 0; goto out; } - if (printk_ratelimit()) { - printk("parent transid verify failed on %llu wanted %llu " + printk_ratelimited("parent transid verify failed on %llu wanted %llu " "found %llu\n", (unsigned long long)eb->start, (unsigned long long)parent_transid, (unsigned long long)btrfs_header_generation(eb)); - } ret = 1; clear_extent_buffer_uptodate(io_tree, eb, &cached_state); out: @@ -380,7 +378,7 @@ static int csum_dirty_buffer(struct btrfs_root *root, struct page *page) len = page->private >> 2; WARN_ON(len == 0); - eb = alloc_extent_buffer(tree, start, len, page, GFP_NOFS); + eb = alloc_extent_buffer(tree, start, len, page); if (eb == NULL) { WARN_ON(1); goto out; @@ -525,7 +523,7 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end, len = page->private >> 2; WARN_ON(len == 0); - eb = alloc_extent_buffer(tree, start, len, page, GFP_NOFS); + eb = alloc_extent_buffer(tree, start, len, page); if (eb == NULL) { ret = -EIO; goto out; @@ -533,12 +531,10 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end, found_start = btrfs_header_bytenr(eb); if (found_start != start) { - if (printk_ratelimit()) { - printk(KERN_INFO "btrfs bad tree block start " + printk_ratelimited(KERN_INFO "btrfs bad tree block start " "%llu %llu\n", (unsigned long long)found_start, (unsigned long long)eb->start); - } ret = -EIO; goto err; } @@ -550,10 +546,8 @@ static int btree_readpage_end_io_hook(struct page *page, u64 start, u64 end, goto err; } if (check_tree_block_fsid(root, eb)) { - if (printk_ratelimit()) { - printk(KERN_INFO "btrfs bad fsid on block %llu\n", + printk_ratelimited(KERN_INFO "btrfs bad fsid on block %llu\n", (unsigned long long)eb->start); - } ret = -EIO; goto err; } @@ -650,12 +644,6 @@ unsigned long btrfs_async_submit_limit(struct btrfs_fs_info *info) return 256 * limit; } -int btrfs_congested_async(struct btrfs_fs_info *info, int iodone) -{ - return atomic_read(&info->nr_async_bios) > - btrfs_async_submit_limit(info); -} - static void run_one_async_start(struct btrfs_work *work) { struct async_submit_bio *async; @@ -963,7 +951,7 @@ struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root, struct inode *btree_inode = root->fs_info->btree_inode; struct extent_buffer *eb; eb = find_extent_buffer(&BTRFS_I(btree_inode)->io_tree, - bytenr, blocksize, GFP_NOFS); + bytenr, blocksize); return eb; } @@ -974,7 +962,7 @@ struct extent_buffer *btrfs_find_create_tree_block(struct btrfs_root *root, struct extent_buffer *eb; eb = alloc_extent_buffer(&BTRFS_I(btree_inode)->io_tree, - bytenr, blocksize, NULL, GFP_NOFS); + bytenr, blocksize, NULL); return eb; } @@ -1058,13 +1046,13 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, root->name = NULL; root->in_sysfs = 0; root->inode_tree = RB_ROOT; + INIT_RADIX_TREE(&root->delayed_nodes_tree, GFP_ATOMIC); root->block_rsv = NULL; root->orphan_block_rsv = NULL; INIT_LIST_HEAD(&root->dirty_list); INIT_LIST_HEAD(&root->orphan_list); INIT_LIST_HEAD(&root->root_list); - spin_lock_init(&root->node_lock); spin_lock_init(&root->orphan_lock); spin_lock_init(&root->inode_lock); spin_lock_init(&root->accounting_lock); @@ -1080,7 +1068,7 @@ static int __setup_root(u32 nodesize, u32 leafsize, u32 sectorsize, root->log_transid = 0; root->last_log_commit = 0; extent_io_tree_init(&root->dirty_log_pages, - fs_info->btree_inode->i_mapping, GFP_NOFS); + fs_info->btree_inode->i_mapping); memset(&root->root_key, 0, sizeof(root->root_key)); memset(&root->root_item, 0, sizeof(root->root_item)); @@ -1283,21 +1271,6 @@ out: return root; } -struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info, - u64 root_objectid) -{ - struct btrfs_root *root; - - if (root_objectid == BTRFS_ROOT_TREE_OBJECTID) - return fs_info->tree_root; - if (root_objectid == BTRFS_EXTENT_TREE_OBJECTID) - return fs_info->extent_root; - - root = radix_tree_lookup(&fs_info->fs_roots_radix, - (unsigned long)root_objectid); - return root; -} - struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info, struct btrfs_key *location) { @@ -1326,6 +1299,19 @@ again: if (IS_ERR(root)) return root; + root->free_ino_ctl = kzalloc(sizeof(*root->free_ino_ctl), GFP_NOFS); + if (!root->free_ino_ctl) + goto fail; + root->free_ino_pinned = kzalloc(sizeof(*root->free_ino_pinned), + GFP_NOFS); + if (!root->free_ino_pinned) + goto fail; + + btrfs_init_free_ino_ctl(root); + mutex_init(&root->fs_commit_mutex); + spin_lock_init(&root->cache_lock); + init_waitqueue_head(&root->cache_wait); + set_anon_super(&root->anon_super, NULL); if (btrfs_root_refs(&root->root_item) == 0) { @@ -1369,41 +1355,6 @@ fail: return ERR_PTR(ret); } -struct btrfs_root *btrfs_read_fs_root(struct btrfs_fs_info *fs_info, - struct btrfs_key *location, - const char *name, int namelen) -{ - return btrfs_read_fs_root_no_name(fs_info, location); -#if 0 - struct btrfs_root *root; - int ret; - - root = btrfs_read_fs_root_no_name(fs_info, location); - if (!root) - return NULL; - - if (root->in_sysfs) - return root; - - ret = btrfs_set_root_name(root, name, namelen); - if (ret) { - free_extent_buffer(root->node); - kfree(root); - return ERR_PTR(ret); - } - - ret = btrfs_sysfs_add_root(root); - if (ret) { - free_extent_buffer(root->node); - kfree(root->name); - kfree(root); - return ERR_PTR(ret); - } - root->in_sysfs = 1; - return root; -#endif -} - static int btrfs_congested_fn(void *congested_data, int bdi_bits) { struct btrfs_fs_info *info = (struct btrfs_fs_info *)congested_data; @@ -1411,7 +1362,8 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits) struct btrfs_device *device; struct backing_dev_info *bdi; - list_for_each_entry(device, &info->fs_devices->devices, dev_list) { + rcu_read_lock(); + list_for_each_entry_rcu(device, &info->fs_devices->devices, dev_list) { if (!device->bdev) continue; bdi = blk_get_backing_dev_info(device->bdev); @@ -1420,6 +1372,7 @@ static int btrfs_congested_fn(void *congested_data, int bdi_bits) break; } } + rcu_read_unlock(); return ret; } @@ -1522,6 +1475,7 @@ static int cleaner_kthread(void *arg) btrfs_run_delayed_iputs(root); btrfs_clean_old_snapshots(root); mutex_unlock(&root->fs_info->cleaner_mutex); + btrfs_run_defrag_inodes(root->fs_info); } if (freezing(current)) { @@ -1611,7 +1565,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, struct btrfs_root *csum_root = kzalloc(sizeof(struct btrfs_root), GFP_NOFS); struct btrfs_root *tree_root = btrfs_sb(sb); - struct btrfs_fs_info *fs_info = tree_root->fs_info; + struct btrfs_fs_info *fs_info = NULL; struct btrfs_root *chunk_root = kzalloc(sizeof(struct btrfs_root), GFP_NOFS); struct btrfs_root *dev_root = kzalloc(sizeof(struct btrfs_root), @@ -1623,11 +1577,12 @@ struct btrfs_root *open_ctree(struct super_block *sb, struct btrfs_super_block *disk_super; - if (!extent_root || !tree_root || !fs_info || + if (!extent_root || !tree_root || !tree_root->fs_info || !chunk_root || !dev_root || !csum_root) { err = -ENOMEM; goto fail; } + fs_info = tree_root->fs_info; ret = init_srcu_struct(&fs_info->subvol_srcu); if (ret) { @@ -1662,6 +1617,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, spin_lock_init(&fs_info->ref_cache_lock); spin_lock_init(&fs_info->fs_roots_radix_lock); spin_lock_init(&fs_info->delayed_iput_lock); + spin_lock_init(&fs_info->defrag_inodes_lock); init_completion(&fs_info->kobj_unregister); fs_info->tree_root = tree_root; @@ -1684,15 +1640,35 @@ struct btrfs_root *open_ctree(struct super_block *sb, atomic_set(&fs_info->async_delalloc_pages, 0); atomic_set(&fs_info->async_submit_draining, 0); atomic_set(&fs_info->nr_async_bios, 0); + atomic_set(&fs_info->defrag_running, 0); fs_info->sb = sb; fs_info->max_inline = 8192 * 1024; fs_info->metadata_ratio = 0; + fs_info->defrag_inodes = RB_ROOT; fs_info->thread_pool_size = min_t(unsigned long, num_online_cpus() + 2, 8); INIT_LIST_HEAD(&fs_info->ordered_extents); spin_lock_init(&fs_info->ordered_extent_lock); + fs_info->delayed_root = kmalloc(sizeof(struct btrfs_delayed_root), + GFP_NOFS); + if (!fs_info->delayed_root) { + err = -ENOMEM; + goto fail_iput; + } + btrfs_init_delayed_root(fs_info->delayed_root); + + mutex_init(&fs_info->scrub_lock); + atomic_set(&fs_info->scrubs_running, 0); + atomic_set(&fs_info->scrub_pause_req, 0); + atomic_set(&fs_info->scrubs_paused, 0); + atomic_set(&fs_info->scrub_cancel_req, 0); + init_waitqueue_head(&fs_info->scrub_pause_wait); + init_rwsem(&fs_info->scrub_super_lock); + fs_info->scrub_workers_refcnt = 0; + btrfs_init_workers(&fs_info->scrub_workers, "scrub", + fs_info->thread_pool_size, &fs_info->generic_worker); sb->s_blocksize = 4096; sb->s_blocksize_bits = blksize_bits(4096); @@ -1711,10 +1687,8 @@ struct btrfs_root *open_ctree(struct super_block *sb, RB_CLEAR_NODE(&BTRFS_I(fs_info->btree_inode)->rb_node); extent_io_tree_init(&BTRFS_I(fs_info->btree_inode)->io_tree, - fs_info->btree_inode->i_mapping, - GFP_NOFS); - extent_map_tree_init(&BTRFS_I(fs_info->btree_inode)->extent_tree, - GFP_NOFS); + fs_info->btree_inode->i_mapping); + extent_map_tree_init(&BTRFS_I(fs_info->btree_inode)->extent_tree); BTRFS_I(fs_info->btree_inode)->io_tree.ops = &btree_extent_io_ops; @@ -1728,9 +1702,9 @@ struct btrfs_root *open_ctree(struct super_block *sb, fs_info->block_group_cache_tree = RB_ROOT; extent_io_tree_init(&fs_info->freed_extents[0], - fs_info->btree_inode->i_mapping, GFP_NOFS); + fs_info->btree_inode->i_mapping); extent_io_tree_init(&fs_info->freed_extents[1], - fs_info->btree_inode->i_mapping, GFP_NOFS); + fs_info->btree_inode->i_mapping); fs_info->pinned_extents = &fs_info->freed_extents[0]; fs_info->do_barriers = 1; @@ -1760,7 +1734,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, bh = btrfs_read_dev_super(fs_devices->latest_bdev); if (!bh) { err = -EINVAL; - goto fail_iput; + goto fail_alloc; } memcpy(&fs_info->super_copy, bh->b_data, sizeof(fs_info->super_copy)); @@ -1772,7 +1746,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, disk_super = &fs_info->super_copy; if (!btrfs_super_root(disk_super)) - goto fail_iput; + goto fail_alloc; /* check FS state, whether FS is broken. */ fs_info->fs_state |= btrfs_super_flags(disk_super); @@ -1788,7 +1762,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, ret = btrfs_parse_options(tree_root, options); if (ret) { err = ret; - goto fail_iput; + goto fail_alloc; } features = btrfs_super_incompat_flags(disk_super) & @@ -1798,7 +1772,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, "unsupported optional features (%Lx).\n", (unsigned long long)features); err = -EINVAL; - goto fail_iput; + goto fail_alloc; } features = btrfs_super_incompat_flags(disk_super); @@ -1814,7 +1788,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, "unsupported option features (%Lx).\n", (unsigned long long)features); err = -EINVAL; - goto fail_iput; + goto fail_alloc; } btrfs_init_workers(&fs_info->generic_worker, @@ -1861,6 +1835,9 @@ struct btrfs_root *open_ctree(struct super_block *sb, &fs_info->generic_worker); btrfs_init_workers(&fs_info->endio_freespace_worker, "freespace-write", 1, &fs_info->generic_worker); + btrfs_init_workers(&fs_info->delayed_workers, "delayed-meta", + fs_info->thread_pool_size, + &fs_info->generic_worker); /* * endios are largely parallel and should have a very @@ -1882,6 +1859,7 @@ struct btrfs_root *open_ctree(struct super_block *sb, btrfs_start_workers(&fs_info->endio_meta_write_workers, 1); btrfs_start_workers(&fs_info->endio_write_workers, 1); btrfs_start_workers(&fs_info->endio_freespace_worker, 1); + btrfs_start_workers(&fs_info->delayed_workers, 1); fs_info->bdi.ra_pages *= btrfs_super_num_devices(disk_super); fs_info->bdi.ra_pages = max(fs_info->bdi.ra_pages, @@ -2138,6 +2116,9 @@ fail_sb_buffer: btrfs_stop_workers(&fs_info->endio_write_workers); btrfs_stop_workers(&fs_info->endio_freespace_worker); btrfs_stop_workers(&fs_info->submit_workers); + btrfs_stop_workers(&fs_info->delayed_workers); +fail_alloc: + kfree(fs_info->delayed_root); fail_iput: invalidate_inode_pages2(fs_info->btree_inode->i_mapping); iput(fs_info->btree_inode); @@ -2165,11 +2146,9 @@ static void btrfs_end_buffer_write_sync(struct buffer_head *bh, int uptodate) if (uptodate) { set_buffer_uptodate(bh); } else { - if (printk_ratelimit()) { - printk(KERN_WARNING "lost page write due to " + printk_ratelimited(KERN_WARNING "lost page write due to " "I/O error on %s\n", bdevname(bh->b_bdev, b)); - } /* note, we dont' set_buffer_write_io_error because we have * our own ways of dealing with the IO errors */ @@ -2333,7 +2312,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors) mutex_lock(&root->fs_info->fs_devices->device_list_mutex); head = &root->fs_info->fs_devices->devices; - list_for_each_entry(dev, head, dev_list) { + list_for_each_entry_rcu(dev, head, dev_list) { if (!dev->bdev) { total_errors++; continue; @@ -2366,7 +2345,7 @@ int write_all_supers(struct btrfs_root *root, int max_mirrors) } total_errors = 0; - list_for_each_entry(dev, head, dev_list) { + list_for_each_entry_rcu(dev, head, dev_list) { if (!dev->bdev) continue; if (!dev->in_fs_metadata || !dev->writeable) @@ -2404,12 +2383,15 @@ int btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root) if (btrfs_root_refs(&root->root_item) == 0) synchronize_srcu(&fs_info->subvol_srcu); + __btrfs_remove_free_space_cache(root->free_ino_pinned); + __btrfs_remove_free_space_cache(root->free_ino_ctl); free_fs_root(root); return 0; } static void free_fs_root(struct btrfs_root *root) { + iput(root->cache_inode); WARN_ON(!RB_EMPTY_ROOT(&root->inode_tree)); if (root->anon_super.s_dev) { down_write(&root->anon_super.s_umount); @@ -2417,6 +2399,8 @@ static void free_fs_root(struct btrfs_root *root) } free_extent_buffer(root->node); free_extent_buffer(root->commit_root); + kfree(root->free_ino_ctl); + kfree(root->free_ino_pinned); kfree(root->name); kfree(root); } @@ -2520,6 +2504,15 @@ int close_ctree(struct btrfs_root *root) fs_info->closing = 1; smp_mb(); + btrfs_scrub_cancel(root); + + /* wait for any defraggers to finish */ + wait_event(fs_info->transaction_wait, + (atomic_read(&fs_info->defrag_running) == 0)); + + /* clear out the rbtree of defraggable inodes */ + btrfs_run_defrag_inodes(root->fs_info); + btrfs_put_block_group_cache(fs_info); /* @@ -2578,6 +2571,7 @@ int close_ctree(struct btrfs_root *root) del_fs_roots(fs_info); iput(fs_info->btree_inode); + kfree(fs_info->delayed_root); btrfs_stop_workers(&fs_info->generic_worker); btrfs_stop_workers(&fs_info->fixup_workers); @@ -2589,6 +2583,7 @@ int close_ctree(struct btrfs_root *root) btrfs_stop_workers(&fs_info->endio_write_workers); btrfs_stop_workers(&fs_info->endio_freespace_worker); btrfs_stop_workers(&fs_info->submit_workers); + btrfs_stop_workers(&fs_info->delayed_workers); btrfs_close_devices(fs_info->fs_devices); btrfs_mapping_tree_free(&fs_info->mapping_tree); @@ -2665,6 +2660,29 @@ void btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr) if (current->flags & PF_MEMALLOC) return; + btrfs_balance_delayed_items(root); + + num_dirty = root->fs_info->dirty_metadata_bytes; + + if (num_dirty > thresh) { + balance_dirty_pages_ratelimited_nr( + root->fs_info->btree_inode->i_mapping, 1); + } + return; +} + +void __btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr) +{ + /* + * looks as though older kernels can get into trouble with + * this code, they end up stuck in balance_dirty_pages forever + */ + u64 num_dirty; + unsigned long thresh = 32 * 1024 * 1024; + + if (current->flags & PF_MEMALLOC) + return; + num_dirty = root->fs_info->dirty_metadata_bytes; if (num_dirty > thresh) { @@ -2697,7 +2715,7 @@ int btree_lock_page_hook(struct page *page) goto out; len = page->private >> 2; - eb = find_extent_buffer(io_tree, bytenr, len, GFP_NOFS); + eb = find_extent_buffer(io_tree, bytenr, len); if (!eb) goto out; @@ -2824,6 +2842,7 @@ static int btrfs_destroy_delayed_refs(struct btrfs_transaction *trans, spin_lock(&delayed_refs->lock); if (delayed_refs->num_entries == 0) { + spin_unlock(&delayed_refs->lock); printk(KERN_INFO "delayed_refs has NO entry\n"); return ret; } @@ -3057,7 +3076,7 @@ static int btrfs_cleanup_transaction(struct btrfs_root *root) btrfs_destroy_pinned_extent(root, root->fs_info->pinned_extents); - t->use_count = 0; + atomic_set(&t->use_count, 0); list_del_init(&t->list); memset(t, 0, sizeof(*t)); kmem_cache_free(btrfs_transaction_cachep, t); diff --git a/fs/btrfs/disk-io.h b/fs/btrfs/disk-io.h index 07b20dc2fd95..a0b610a67aae 100644 --- a/fs/btrfs/disk-io.h +++ b/fs/btrfs/disk-io.h @@ -55,35 +55,20 @@ int btrfs_commit_super(struct btrfs_root *root); int btrfs_error_commit_super(struct btrfs_root *root); struct extent_buffer *btrfs_find_tree_block(struct btrfs_root *root, u64 bytenr, u32 blocksize); -struct btrfs_root *btrfs_lookup_fs_root(struct btrfs_fs_info *fs_info, - u64 root_objectid); -struct btrfs_root *btrfs_read_fs_root(struct btrfs_fs_info *fs_info, - struct btrfs_key *location, - const char *name, int namelen); struct btrfs_root *btrfs_read_fs_root_no_radix(struct btrfs_root *tree_root, struct btrfs_key *location); struct btrfs_root *btrfs_read_fs_root_no_name(struct btrfs_fs_info *fs_info, struct btrfs_key *location); int btrfs_cleanup_fs_roots(struct btrfs_fs_info *fs_info); -int btrfs_insert_dev_radix(struct btrfs_root *root, - struct block_device *bdev, - u64 device_id, - u64 block_start, - u64 num_blocks); void btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr); +void __btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr); int btrfs_free_fs_root(struct btrfs_fs_info *fs_info, struct btrfs_root *root); void btrfs_mark_buffer_dirty(struct extent_buffer *buf); -void btrfs_mark_buffer_dirty_nonblocking(struct extent_buffer *buf); int btrfs_buffer_uptodate(struct extent_buffer *buf, u64 parent_transid); int btrfs_set_buffer_uptodate(struct extent_buffer *buf); -int wait_on_tree_block_writeback(struct btrfs_root *root, - struct extent_buffer *buf); int btrfs_read_buffer(struct extent_buffer *buf, u64 parent_transid); u32 btrfs_csum_data(struct btrfs_root *root, char *data, u32 seed, size_t len); void btrfs_csum_final(u32 crc, char *result); -int btrfs_open_device(struct btrfs_device *dev); -int btrfs_verify_block_csum(struct btrfs_root *root, - struct extent_buffer *buf); int btrfs_bio_wq_end_io(struct btrfs_fs_info *info, struct bio *bio, int metadata); int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode, @@ -91,8 +76,6 @@ int btrfs_wq_submit_bio(struct btrfs_fs_info *fs_info, struct inode *inode, unsigned long bio_flags, u64 bio_offset, extent_submit_bio_hook_t *submit_bio_start, extent_submit_bio_hook_t *submit_bio_done); - -int btrfs_congested_async(struct btrfs_fs_info *info, int iodone); unsigned long btrfs_async_submit_limit(struct btrfs_fs_info *info); int btrfs_write_tree_block(struct extent_buffer *buf); int btrfs_wait_tree_block_writeback(struct extent_buffer *buf); diff --git a/fs/btrfs/export.c b/fs/btrfs/export.c index b4ffad859adb..1b8dc33778f9 100644 --- a/fs/btrfs/export.c +++ b/fs/btrfs/export.c @@ -32,7 +32,7 @@ static int btrfs_encode_fh(struct dentry *dentry, u32 *fh, int *max_len, len = BTRFS_FID_SIZE_NON_CONNECTABLE; type = FILEID_BTRFS_WITHOUT_PARENT; - fid->objectid = inode->i_ino; + fid->objectid = btrfs_ino(inode); fid->root_objectid = BTRFS_I(inode)->root->objectid; fid->gen = inode->i_generation; @@ -178,13 +178,13 @@ static struct dentry *btrfs_get_parent(struct dentry *child) if (!path) return ERR_PTR(-ENOMEM); - if (dir->i_ino == BTRFS_FIRST_FREE_OBJECTID) { + if (btrfs_ino(dir) == BTRFS_FIRST_FREE_OBJECTID) { key.objectid = root->root_key.objectid; key.type = BTRFS_ROOT_BACKREF_KEY; key.offset = (u64)-1; root = root->fs_info->tree_root; } else { - key.objectid = dir->i_ino; + key.objectid = btrfs_ino(dir); key.type = BTRFS_INODE_REF_KEY; key.offset = (u64)-1; } @@ -244,6 +244,7 @@ static int btrfs_get_name(struct dentry *parent, char *name, struct btrfs_key key; int name_len; int ret; + u64 ino; if (!dir || !inode) return -EINVAL; @@ -251,19 +252,21 @@ static int btrfs_get_name(struct dentry *parent, char *name, if (!S_ISDIR(dir->i_mode)) return -EINVAL; + ino = btrfs_ino(inode); + path = btrfs_alloc_path(); if (!path) return -ENOMEM; path->leave_spinning = 1; - if (inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) { + if (ino == BTRFS_FIRST_FREE_OBJECTID) { key.objectid = BTRFS_I(inode)->root->root_key.objectid; key.type = BTRFS_ROOT_BACKREF_KEY; key.offset = (u64)-1; root = root->fs_info->tree_root; } else { - key.objectid = inode->i_ino; - key.offset = dir->i_ino; + key.objectid = ino; + key.offset = btrfs_ino(dir); key.type = BTRFS_INODE_REF_KEY; } @@ -272,7 +275,7 @@ static int btrfs_get_name(struct dentry *parent, char *name, btrfs_free_path(path); return ret; } else if (ret > 0) { - if (inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) { + if (ino == BTRFS_FIRST_FREE_OBJECTID) { path->slots[0]--; } else { btrfs_free_path(path); @@ -281,11 +284,11 @@ static int btrfs_get_name(struct dentry *parent, char *name, } leaf = path->nodes[0]; - if (inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) { - rref = btrfs_item_ptr(leaf, path->slots[0], + if (ino == BTRFS_FIRST_FREE_OBJECTID) { + rref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref); - name_ptr = (unsigned long)(rref + 1); - name_len = btrfs_root_ref_name_len(leaf, rref); + name_ptr = (unsigned long)(rref + 1); + name_len = btrfs_root_ref_name_len(leaf, rref); } else { iref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_inode_ref); diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index f619c3cb13b7..169bd62ce776 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -33,6 +33,25 @@ #include "locking.h" #include "free-space-cache.h" +/* control flags for do_chunk_alloc's force field + * CHUNK_ALLOC_NO_FORCE means to only allocate a chunk + * if we really need one. + * + * CHUNK_ALLOC_FORCE means it must try to allocate one + * + * CHUNK_ALLOC_LIMITED means to only try and allocate one + * if we have very few chunks already allocated. This is + * used as part of the clustering code to help make sure + * we have a good pool of storage to cluster in, without + * filling the FS with empty chunks + * + */ +enum { + CHUNK_ALLOC_NO_FORCE = 0, + CHUNK_ALLOC_FORCE = 1, + CHUNK_ALLOC_LIMITED = 2, +}; + static int update_block_group(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 bytenr, u64 num_bytes, int alloc); @@ -75,7 +94,7 @@ static int block_group_bits(struct btrfs_block_group_cache *cache, u64 bits) return (cache->flags & bits) == bits; } -void btrfs_get_block_group(struct btrfs_block_group_cache *cache) +static void btrfs_get_block_group(struct btrfs_block_group_cache *cache) { atomic_inc(&cache->count); } @@ -86,6 +105,7 @@ void btrfs_put_block_group(struct btrfs_block_group_cache *cache) WARN_ON(cache->pinned > 0); WARN_ON(cache->reserved > 0); WARN_ON(cache->reserved_pinned > 0); + kfree(cache->free_space_ctl); kfree(cache); } } @@ -360,7 +380,7 @@ again: break; caching_ctl->progress = last; - btrfs_release_path(extent_root, path); + btrfs_release_path(path); up_read(&fs_info->extent_commit_sem); mutex_unlock(&caching_ctl->mutex); if (btrfs_transaction_in_commit(fs_info)) @@ -735,8 +755,12 @@ again: atomic_inc(&head->node.refs); spin_unlock(&delayed_refs->lock); - btrfs_release_path(root->fs_info->extent_root, path); + btrfs_release_path(path); + /* + * Mutex was contended, block until it's released and try + * again + */ mutex_lock(&head->mutex); mutex_unlock(&head->mutex); btrfs_put_delayed_ref(&head->node); @@ -915,7 +939,7 @@ static int convert_extent_item_v0(struct btrfs_trans_handle *trans, break; } } - btrfs_release_path(root, path); + btrfs_release_path(path); if (owner < BTRFS_FIRST_FREE_OBJECTID) new_size += sizeof(*bi); @@ -928,7 +952,6 @@ static int convert_extent_item_v0(struct btrfs_trans_handle *trans, BUG_ON(ret); ret = btrfs_extend_item(trans, root, path, new_size); - BUG_ON(ret); leaf = path->nodes[0]; item = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item); @@ -1023,7 +1046,7 @@ again: return 0; #ifdef BTRFS_COMPAT_EXTENT_TREE_V0 key.type = BTRFS_EXTENT_REF_V0_KEY; - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_search_slot(trans, root, &key, path, -1, 1); if (ret < 0) { err = ret; @@ -1061,7 +1084,7 @@ again: if (match_extent_data_ref(leaf, ref, root_objectid, owner, offset)) { if (recow) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto again; } err = 0; @@ -1122,7 +1145,7 @@ static noinline int insert_extent_data_ref(struct btrfs_trans_handle *trans, if (match_extent_data_ref(leaf, ref, root_objectid, owner, offset)) break; - btrfs_release_path(root, path); + btrfs_release_path(path); key.offset++; ret = btrfs_insert_empty_item(trans, root, path, &key, size); @@ -1148,7 +1171,7 @@ static noinline int insert_extent_data_ref(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(leaf); ret = 0; fail: - btrfs_release_path(root, path); + btrfs_release_path(path); return ret; } @@ -1274,7 +1297,7 @@ static noinline int lookup_tree_block_ref(struct btrfs_trans_handle *trans, ret = -ENOENT; #ifdef BTRFS_COMPAT_EXTENT_TREE_V0 if (ret == -ENOENT && parent) { - btrfs_release_path(root, path); + btrfs_release_path(path); key.type = BTRFS_EXTENT_REF_V0_KEY; ret = btrfs_search_slot(trans, root, &key, path, -1, 1); if (ret > 0) @@ -1303,7 +1326,7 @@ static noinline int insert_tree_block_ref(struct btrfs_trans_handle *trans, } ret = btrfs_insert_empty_item(trans, root, path, &key, 0); - btrfs_release_path(root, path); + btrfs_release_path(path); return ret; } @@ -1536,7 +1559,6 @@ int setup_inline_extent_backref(struct btrfs_trans_handle *trans, size = btrfs_extent_inline_ref_size(type); ret = btrfs_extend_item(trans, root, path, size); - BUG_ON(ret); ei = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_extent_item); refs = btrfs_extent_refs(leaf, ei); @@ -1589,7 +1611,7 @@ static int lookup_extent_backref(struct btrfs_trans_handle *trans, if (ret != -ENOENT) return ret; - btrfs_release_path(root, path); + btrfs_release_path(path); *ref_ret = NULL; if (owner < BTRFS_FIRST_FREE_OBJECTID) { @@ -1665,7 +1687,6 @@ int update_inline_extent_backref(struct btrfs_trans_handle *trans, end - ptr - size); item_size -= size; ret = btrfs_truncate_item(trans, root, path, item_size, 1); - BUG_ON(ret); } btrfs_mark_buffer_dirty(leaf); return 0; @@ -1843,7 +1864,7 @@ static int __btrfs_inc_extent_ref(struct btrfs_trans_handle *trans, __run_delayed_extent_op(extent_op, leaf, item); btrfs_mark_buffer_dirty(leaf); - btrfs_release_path(root->fs_info->extent_root, path); + btrfs_release_path(path); path->reada = 1; path->leave_spinning = 1; @@ -2278,6 +2299,10 @@ again: atomic_inc(&ref->refs); spin_unlock(&delayed_refs->lock); + /* + * Mutex was contended, block until it's + * released and try again + */ mutex_lock(&head->mutex); mutex_unlock(&head->mutex); @@ -2342,8 +2367,12 @@ static noinline int check_delayed_ref(struct btrfs_trans_handle *trans, atomic_inc(&head->node.refs); spin_unlock(&delayed_refs->lock); - btrfs_release_path(root->fs_info->extent_root, path); + btrfs_release_path(path); + /* + * Mutex was contended, block until it's released and let + * caller try again + */ mutex_lock(&head->mutex); mutex_unlock(&head->mutex); btrfs_put_delayed_ref(&head->node); @@ -2491,126 +2520,6 @@ out: return ret; } -#if 0 -int btrfs_cache_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, - struct extent_buffer *buf, u32 nr_extents) -{ - struct btrfs_key key; - struct btrfs_file_extent_item *fi; - u64 root_gen; - u32 nritems; - int i; - int level; - int ret = 0; - int shared = 0; - - if (!root->ref_cows) - return 0; - - if (root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID) { - shared = 0; - root_gen = root->root_key.offset; - } else { - shared = 1; - root_gen = trans->transid - 1; - } - - level = btrfs_header_level(buf); - nritems = btrfs_header_nritems(buf); - - if (level == 0) { - struct btrfs_leaf_ref *ref; - struct btrfs_extent_info *info; - - ref = btrfs_alloc_leaf_ref(root, nr_extents); - if (!ref) { - ret = -ENOMEM; - goto out; - } - - ref->root_gen = root_gen; - ref->bytenr = buf->start; - ref->owner = btrfs_header_owner(buf); - ref->generation = btrfs_header_generation(buf); - ref->nritems = nr_extents; - info = ref->extents; - - for (i = 0; nr_extents > 0 && i < nritems; i++) { - u64 disk_bytenr; - btrfs_item_key_to_cpu(buf, &key, i); - if (btrfs_key_type(&key) != BTRFS_EXTENT_DATA_KEY) - continue; - fi = btrfs_item_ptr(buf, i, - struct btrfs_file_extent_item); - if (btrfs_file_extent_type(buf, fi) == - BTRFS_FILE_EXTENT_INLINE) - continue; - disk_bytenr = btrfs_file_extent_disk_bytenr(buf, fi); - if (disk_bytenr == 0) - continue; - - info->bytenr = disk_bytenr; - info->num_bytes = - btrfs_file_extent_disk_num_bytes(buf, fi); - info->objectid = key.objectid; - info->offset = key.offset; - info++; - } - - ret = btrfs_add_leaf_ref(root, ref, shared); - if (ret == -EEXIST && shared) { - struct btrfs_leaf_ref *old; - old = btrfs_lookup_leaf_ref(root, ref->bytenr); - BUG_ON(!old); - btrfs_remove_leaf_ref(root, old); - btrfs_free_leaf_ref(root, old); - ret = btrfs_add_leaf_ref(root, ref, shared); - } - WARN_ON(ret); - btrfs_free_leaf_ref(root, ref); - } -out: - return ret; -} - -/* when a block goes through cow, we update the reference counts of - * everything that block points to. The internal pointers of the block - * can be in just about any order, and it is likely to have clusters of - * things that are close together and clusters of things that are not. - * - * To help reduce the seeks that come with updating all of these reference - * counts, sort them by byte number before actual updates are done. - * - * struct refsort is used to match byte number to slot in the btree block. - * we sort based on the byte number and then use the slot to actually - * find the item. - * - * struct refsort is smaller than strcut btrfs_item and smaller than - * struct btrfs_key_ptr. Since we're currently limited to the page size - * for a btree block, there's no way for a kmalloc of refsorts for a - * single node to be bigger than a page. - */ -struct refsort { - u64 bytenr; - u32 slot; -}; - -/* - * for passing into sort() - */ -static int refsort_cmp(const void *a_void, const void *b_void) -{ - const struct refsort *a = a_void; - const struct refsort *b = b_void; - - if (a->bytenr < b->bytenr) - return -1; - if (a->bytenr > b->bytenr) - return 1; - return 0; -} -#endif - static int __btrfs_mod_ref(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct extent_buffer *buf, @@ -2713,7 +2622,7 @@ static int write_one_cache_group(struct btrfs_trans_handle *trans, bi = btrfs_item_ptr_offset(leaf, path->slots[0]); write_extent_buffer(leaf, &cache->item, bi, sizeof(cache->item)); btrfs_mark_buffer_dirty(leaf); - btrfs_release_path(extent_root, path); + btrfs_release_path(path); fail: if (ret) return ret; @@ -2766,7 +2675,7 @@ again: inode = lookup_free_space_inode(root, block_group, path); if (IS_ERR(inode) && PTR_ERR(inode) != -ENOENT) { ret = PTR_ERR(inode); - btrfs_release_path(root, path); + btrfs_release_path(path); goto out; } @@ -2835,7 +2744,7 @@ again: out_put: iput(inode); out_free: - btrfs_release_path(root, path); + btrfs_release_path(path); out: spin_lock(&block_group->lock); block_group->disk_cache_state = dcs; @@ -3019,7 +2928,8 @@ static int update_space_info(struct btrfs_fs_info *info, u64 flags, found->bytes_readonly = 0; found->bytes_may_use = 0; found->full = 0; - found->force_alloc = 0; + found->force_alloc = CHUNK_ALLOC_NO_FORCE; + found->chunk_alloc = 0; *space_info = found; list_add_rcu(&found->list, &info->space_info); atomic_set(&found->caching_threads, 0); @@ -3124,7 +3034,8 @@ int btrfs_check_data_free_space(struct inode *inode, u64 bytes) /* make sure bytes are sectorsize aligned */ bytes = (bytes + root->sectorsize - 1) & ~((u64)root->sectorsize - 1); - if (root == root->fs_info->tree_root) { + if (root == root->fs_info->tree_root || + BTRFS_I(inode)->location.objectid == BTRFS_FREE_INO_OBJECTID) { alloc_chunk = 0; committed = 1; } @@ -3150,7 +3061,7 @@ again: if (!data_sinfo->full && alloc_chunk) { u64 alloc_target; - data_sinfo->force_alloc = 1; + data_sinfo->force_alloc = CHUNK_ALLOC_FORCE; spin_unlock(&data_sinfo->lock); alloc: alloc_target = btrfs_get_alloc_profile(root, 1); @@ -3160,7 +3071,8 @@ alloc: ret = do_chunk_alloc(trans, root->fs_info->extent_root, bytes + 2 * 1024 * 1024, - alloc_target, 0); + alloc_target, + CHUNK_ALLOC_NO_FORCE); btrfs_end_transaction(trans, root); if (ret < 0) { if (ret != -ENOSPC) @@ -3190,18 +3102,6 @@ commit_trans: goto again; } -#if 0 /* I hope we never need this code again, just in case */ - printk(KERN_ERR "no space left, need %llu, %llu bytes_used, " - "%llu bytes_reserved, " "%llu bytes_pinned, " - "%llu bytes_readonly, %llu may use %llu total\n", - (unsigned long long)bytes, - (unsigned long long)data_sinfo->bytes_used, - (unsigned long long)data_sinfo->bytes_reserved, - (unsigned long long)data_sinfo->bytes_pinned, - (unsigned long long)data_sinfo->bytes_readonly, - (unsigned long long)data_sinfo->bytes_may_use, - (unsigned long long)data_sinfo->total_bytes); -#endif return -ENOSPC; } data_sinfo->bytes_may_use += bytes; @@ -3239,31 +3139,56 @@ static void force_metadata_allocation(struct btrfs_fs_info *info) rcu_read_lock(); list_for_each_entry_rcu(found, head, list) { if (found->flags & BTRFS_BLOCK_GROUP_METADATA) - found->force_alloc = 1; + found->force_alloc = CHUNK_ALLOC_FORCE; } rcu_read_unlock(); } static int should_alloc_chunk(struct btrfs_root *root, - struct btrfs_space_info *sinfo, u64 alloc_bytes) + struct btrfs_space_info *sinfo, u64 alloc_bytes, + int force) { u64 num_bytes = sinfo->total_bytes - sinfo->bytes_readonly; + u64 num_allocated = sinfo->bytes_used + sinfo->bytes_reserved; u64 thresh; - if (sinfo->bytes_used + sinfo->bytes_reserved + - alloc_bytes + 256 * 1024 * 1024 < num_bytes) + if (force == CHUNK_ALLOC_FORCE) + return 1; + + /* + * in limited mode, we want to have some free space up to + * about 1% of the FS size. + */ + if (force == CHUNK_ALLOC_LIMITED) { + thresh = btrfs_super_total_bytes(&root->fs_info->super_copy); + thresh = max_t(u64, 64 * 1024 * 1024, + div_factor_fine(thresh, 1)); + + if (num_bytes - num_allocated < thresh) + return 1; + } + + /* + * we have two similar checks here, one based on percentage + * and once based on a hard number of 256MB. The idea + * is that if we have a good amount of free + * room, don't allocate a chunk. A good mount is + * less than 80% utilized of the chunks we have allocated, + * or more than 256MB free + */ + if (num_allocated + alloc_bytes + 256 * 1024 * 1024 < num_bytes) return 0; - if (sinfo->bytes_used + sinfo->bytes_reserved + - alloc_bytes < div_factor(num_bytes, 8)) + if (num_allocated + alloc_bytes < div_factor(num_bytes, 8)) return 0; thresh = btrfs_super_total_bytes(&root->fs_info->super_copy); + + /* 256MB or 5% of the FS */ thresh = max_t(u64, 256 * 1024 * 1024, div_factor_fine(thresh, 5)); if (num_bytes > thresh && sinfo->bytes_used < div_factor(num_bytes, 3)) return 0; - return 1; } @@ -3273,10 +3198,9 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, { struct btrfs_space_info *space_info; struct btrfs_fs_info *fs_info = extent_root->fs_info; + int wait_for_alloc = 0; int ret = 0; - mutex_lock(&fs_info->chunk_mutex); - flags = btrfs_reduce_alloc_profile(extent_root, flags); space_info = __find_space_info(extent_root->fs_info, flags); @@ -3287,21 +3211,40 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, } BUG_ON(!space_info); +again: spin_lock(&space_info->lock); if (space_info->force_alloc) - force = 1; + force = space_info->force_alloc; if (space_info->full) { spin_unlock(&space_info->lock); - goto out; + return 0; } - if (!force && !should_alloc_chunk(extent_root, space_info, - alloc_bytes)) { + if (!should_alloc_chunk(extent_root, space_info, alloc_bytes, force)) { spin_unlock(&space_info->lock); - goto out; + return 0; + } else if (space_info->chunk_alloc) { + wait_for_alloc = 1; + } else { + space_info->chunk_alloc = 1; } + spin_unlock(&space_info->lock); + mutex_lock(&fs_info->chunk_mutex); + + /* + * The chunk_mutex is held throughout the entirety of a chunk + * allocation, so once we've acquired the chunk_mutex we know that the + * other guy is done and we need to recheck and see if we should + * allocate. + */ + if (wait_for_alloc) { + mutex_unlock(&fs_info->chunk_mutex); + wait_for_alloc = 0; + goto again; + } + /* * If we have mixed data/metadata chunks we want to make sure we keep * allocating mixed chunks instead of individual chunks. @@ -3327,9 +3270,10 @@ static int do_chunk_alloc(struct btrfs_trans_handle *trans, space_info->full = 1; else ret = 1; - space_info->force_alloc = 0; + + space_info->force_alloc = CHUNK_ALLOC_NO_FORCE; + space_info->chunk_alloc = 0; spin_unlock(&space_info->lock); -out: mutex_unlock(&extent_root->fs_info->chunk_mutex); return ret; } @@ -3360,6 +3304,10 @@ static int shrink_delalloc(struct btrfs_trans_handle *trans, if (reserved == 0) return 0; + /* nothing to shrink - nothing to reclaim */ + if (root->fs_info->delalloc_bytes == 0) + return 0; + max_reclaim = min(reserved, to_reclaim); while (loops < 1024) { @@ -3586,8 +3534,8 @@ static void block_rsv_add_bytes(struct btrfs_block_rsv *block_rsv, spin_unlock(&block_rsv->lock); } -void block_rsv_release_bytes(struct btrfs_block_rsv *block_rsv, - struct btrfs_block_rsv *dest, u64 num_bytes) +static void block_rsv_release_bytes(struct btrfs_block_rsv *block_rsv, + struct btrfs_block_rsv *dest, u64 num_bytes) { struct btrfs_space_info *space_info = block_rsv->space_info; @@ -3790,23 +3738,7 @@ static u64 calc_global_metadata_size(struct btrfs_fs_info *fs_info) u64 meta_used; u64 data_used; int csum_size = btrfs_super_csum_size(&fs_info->super_copy); -#if 0 - /* - * per tree used space accounting can be inaccuracy, so we - * can't rely on it. - */ - spin_lock(&fs_info->extent_root->accounting_lock); - num_bytes = btrfs_root_used(&fs_info->extent_root->root_item); - spin_unlock(&fs_info->extent_root->accounting_lock); - - spin_lock(&fs_info->csum_root->accounting_lock); - num_bytes += btrfs_root_used(&fs_info->csum_root->root_item); - spin_unlock(&fs_info->csum_root->accounting_lock); - spin_lock(&fs_info->tree_root->accounting_lock); - num_bytes += btrfs_root_used(&fs_info->tree_root->root_item); - spin_unlock(&fs_info->tree_root->accounting_lock); -#endif sinfo = __find_space_info(fs_info, BTRFS_BLOCK_GROUP_DATA); spin_lock(&sinfo->lock); data_used = sinfo->bytes_used; @@ -3859,10 +3791,7 @@ static void update_global_block_rsv(struct btrfs_fs_info *fs_info) block_rsv->reserved = block_rsv->size; block_rsv->full = 1; } -#if 0 - printk(KERN_INFO"global block rsv size %llu reserved %llu\n", - block_rsv->size, block_rsv->reserved); -#endif + spin_unlock(&sinfo->lock); spin_unlock(&block_rsv->lock); } @@ -3908,12 +3837,6 @@ static void release_global_block_rsv(struct btrfs_fs_info *fs_info) WARN_ON(fs_info->chunk_block_rsv.reserved > 0); } -static u64 calc_trans_metadata_size(struct btrfs_root *root, int num_items) -{ - return (root->leafsize + root->nodesize * (BTRFS_MAX_LEVEL - 1)) * - 3 * num_items; -} - int btrfs_trans_reserve_metadata(struct btrfs_trans_handle *trans, struct btrfs_root *root, int num_items) @@ -3924,7 +3847,7 @@ int btrfs_trans_reserve_metadata(struct btrfs_trans_handle *trans, if (num_items == 0 || root->fs_info->chunk_root == root) return 0; - num_bytes = calc_trans_metadata_size(root, num_items); + num_bytes = btrfs_calc_trans_metadata_size(root, num_items); ret = btrfs_block_rsv_add(trans, root, &root->fs_info->trans_block_rsv, num_bytes); if (!ret) { @@ -3963,14 +3886,14 @@ int btrfs_orphan_reserve_metadata(struct btrfs_trans_handle *trans, * If all of the metadata space is used, we can commit * transaction and use space it freed. */ - u64 num_bytes = calc_trans_metadata_size(root, 4); + u64 num_bytes = btrfs_calc_trans_metadata_size(root, 4); return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes); } void btrfs_orphan_release_metadata(struct inode *inode) { struct btrfs_root *root = BTRFS_I(inode)->root; - u64 num_bytes = calc_trans_metadata_size(root, 4); + u64 num_bytes = btrfs_calc_trans_metadata_size(root, 4); btrfs_block_rsv_release(root, root->orphan_block_rsv, num_bytes); } @@ -3984,7 +3907,7 @@ int btrfs_snap_reserve_metadata(struct btrfs_trans_handle *trans, * two for root back/forward refs, two for directory entries * and one for root of the snapshot. */ - u64 num_bytes = calc_trans_metadata_size(root, 5); + u64 num_bytes = btrfs_calc_trans_metadata_size(root, 5); dst_rsv->space_info = src_rsv->space_info; return block_rsv_migrate_bytes(src_rsv, dst_rsv, num_bytes); } @@ -4013,7 +3936,7 @@ int btrfs_delalloc_reserve_metadata(struct inode *inode, u64 num_bytes) if (nr_extents > reserved_extents) { nr_extents -= reserved_extents; - to_reserve = calc_trans_metadata_size(root, nr_extents); + to_reserve = btrfs_calc_trans_metadata_size(root, nr_extents); } else { nr_extents = 0; to_reserve = 0; @@ -4067,7 +3990,7 @@ void btrfs_delalloc_release_metadata(struct inode *inode, u64 num_bytes) to_free = calc_csum_metadata_size(inode, num_bytes); if (nr_extents > 0) - to_free += calc_trans_metadata_size(root, nr_extents); + to_free += btrfs_calc_trans_metadata_size(root, nr_extents); btrfs_block_rsv_release(root, &root->fs_info->delalloc_block_rsv, to_free); @@ -4476,7 +4399,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, NULL, refs_to_drop, is_data); BUG_ON(ret); - btrfs_release_path(extent_root, path); + btrfs_release_path(path); path->leave_spinning = 1; key.objectid = bytenr; @@ -4515,7 +4438,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, owner_objectid, 0); BUG_ON(ret < 0); - btrfs_release_path(extent_root, path); + btrfs_release_path(path); path->leave_spinning = 1; key.objectid = bytenr; @@ -4585,7 +4508,7 @@ static int __btrfs_free_extent(struct btrfs_trans_handle *trans, ret = btrfs_del_items(trans, extent_root, path, path->slots[0], num_to_del); BUG_ON(ret); - btrfs_release_path(extent_root, path); + btrfs_release_path(path); if (is_data) { ret = btrfs_del_csums(trans, root, bytenr, num_bytes); @@ -4828,7 +4751,7 @@ wait_block_group_cache_progress(struct btrfs_block_group_cache *cache, return 0; wait_event(caching_ctl->wait, block_group_cache_done(cache) || - (cache->free_space >= num_bytes)); + (cache->free_space_ctl->free_space >= num_bytes)); put_caching_control(caching_ctl); return 0; @@ -5303,11 +5226,13 @@ loop: if (allowed_chunk_alloc) { ret = do_chunk_alloc(trans, root, num_bytes + - 2 * 1024 * 1024, data, 1); + 2 * 1024 * 1024, data, + CHUNK_ALLOC_LIMITED); allowed_chunk_alloc = 0; done_chunk_alloc = 1; - } else if (!done_chunk_alloc) { - space_info->force_alloc = 1; + } else if (!done_chunk_alloc && + space_info->force_alloc == CHUNK_ALLOC_NO_FORCE) { + space_info->force_alloc = CHUNK_ALLOC_LIMITED; } if (loop < LOOP_NO_EMPTY_SIZE) { @@ -5393,7 +5318,8 @@ again: */ if (empty_size || root->ref_cows) ret = do_chunk_alloc(trans, root->fs_info->extent_root, - num_bytes + 2 * 1024 * 1024, data, 0); + num_bytes + 2 * 1024 * 1024, data, + CHUNK_ALLOC_NO_FORCE); WARN_ON(num_bytes < root->sectorsize); ret = find_free_extent(trans, root, num_bytes, empty_size, @@ -5405,7 +5331,7 @@ again: num_bytes = num_bytes & ~(root->sectorsize - 1); num_bytes = max(num_bytes, min_alloc_size); do_chunk_alloc(trans, root->fs_info->extent_root, - num_bytes, data, 1); + num_bytes, data, CHUNK_ALLOC_FORCE); goto again; } if (ret == -ENOSPC && btrfs_test_opt(root, ENOSPC_DEBUG)) { @@ -6412,7 +6338,7 @@ int btrfs_drop_snapshot(struct btrfs_root *root, trans->block_rsv = block_rsv; } } - btrfs_release_path(root, path); + btrfs_release_path(path); BUG_ON(err); ret = btrfs_del_root(trans, tree_root, &root->root_key); @@ -6516,1510 +6442,6 @@ int btrfs_drop_subtree(struct btrfs_trans_handle *trans, return ret; } -#if 0 -static unsigned long calc_ra(unsigned long start, unsigned long last, - unsigned long nr) -{ - return min(last, start + nr - 1); -} - -static noinline int relocate_inode_pages(struct inode *inode, u64 start, - u64 len) -{ - u64 page_start; - u64 page_end; - unsigned long first_index; - unsigned long last_index; - unsigned long i; - struct page *page; - struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; - struct file_ra_state *ra; - struct btrfs_ordered_extent *ordered; - unsigned int total_read = 0; - unsigned int total_dirty = 0; - int ret = 0; - - ra = kzalloc(sizeof(*ra), GFP_NOFS); - if (!ra) - return -ENOMEM; - - mutex_lock(&inode->i_mutex); - first_index = start >> PAGE_CACHE_SHIFT; - last_index = (start + len - 1) >> PAGE_CACHE_SHIFT; - - /* make sure the dirty trick played by the caller work */ - ret = invalidate_inode_pages2_range(inode->i_mapping, - first_index, last_index); - if (ret) - goto out_unlock; - - file_ra_state_init(ra, inode->i_mapping); - - for (i = first_index ; i <= last_index; i++) { - if (total_read % ra->ra_pages == 0) { - btrfs_force_ra(inode->i_mapping, ra, NULL, i, - calc_ra(i, last_index, ra->ra_pages)); - } - total_read++; -again: - if (((u64)i << PAGE_CACHE_SHIFT) > i_size_read(inode)) - BUG_ON(1); - page = grab_cache_page(inode->i_mapping, i); - if (!page) { - ret = -ENOMEM; - goto out_unlock; - } - if (!PageUptodate(page)) { - btrfs_readpage(NULL, page); - lock_page(page); - if (!PageUptodate(page)) { - unlock_page(page); - page_cache_release(page); - ret = -EIO; - goto out_unlock; - } - } - wait_on_page_writeback(page); - - page_start = (u64)page->index << PAGE_CACHE_SHIFT; - page_end = page_start + PAGE_CACHE_SIZE - 1; - lock_extent(io_tree, page_start, page_end, GFP_NOFS); - - ordered = btrfs_lookup_ordered_extent(inode, page_start); - if (ordered) { - unlock_extent(io_tree, page_start, page_end, GFP_NOFS); - unlock_page(page); - page_cache_release(page); - btrfs_start_ordered_extent(inode, ordered, 1); - btrfs_put_ordered_extent(ordered); - goto again; - } - set_page_extent_mapped(page); - - if (i == first_index) - set_extent_bits(io_tree, page_start, page_end, - EXTENT_BOUNDARY, GFP_NOFS); - btrfs_set_extent_delalloc(inode, page_start, page_end); - - set_page_dirty(page); - total_dirty++; - - unlock_extent(io_tree, page_start, page_end, GFP_NOFS); - unlock_page(page); - page_cache_release(page); - } - -out_unlock: - kfree(ra); - mutex_unlock(&inode->i_mutex); - balance_dirty_pages_ratelimited_nr(inode->i_mapping, total_dirty); - return ret; -} - -static noinline int relocate_data_extent(struct inode *reloc_inode, - struct btrfs_key *extent_key, - u64 offset) -{ - struct btrfs_root *root = BTRFS_I(reloc_inode)->root; - struct extent_map_tree *em_tree = &BTRFS_I(reloc_inode)->extent_tree; - struct extent_map *em; - u64 start = extent_key->objectid - offset; - u64 end = start + extent_key->offset - 1; - - em = alloc_extent_map(GFP_NOFS); - BUG_ON(!em); - - em->start = start; - em->len = extent_key->offset; - em->block_len = extent_key->offset; - em->block_start = extent_key->objectid; - em->bdev = root->fs_info->fs_devices->latest_bdev; - set_bit(EXTENT_FLAG_PINNED, &em->flags); - - /* setup extent map to cheat btrfs_readpage */ - lock_extent(&BTRFS_I(reloc_inode)->io_tree, start, end, GFP_NOFS); - while (1) { - int ret; - write_lock(&em_tree->lock); - ret = add_extent_mapping(em_tree, em); - write_unlock(&em_tree->lock); - if (ret != -EEXIST) { - free_extent_map(em); - break; - } - btrfs_drop_extent_cache(reloc_inode, start, end, 0); - } - unlock_extent(&BTRFS_I(reloc_inode)->io_tree, start, end, GFP_NOFS); - - return relocate_inode_pages(reloc_inode, start, extent_key->offset); -} - -struct btrfs_ref_path { - u64 extent_start; - u64 nodes[BTRFS_MAX_LEVEL]; - u64 root_objectid; - u64 root_generation; - u64 owner_objectid; - u32 num_refs; - int lowest_level; - int current_level; - int shared_level; - - struct btrfs_key node_keys[BTRFS_MAX_LEVEL]; - u64 new_nodes[BTRFS_MAX_LEVEL]; -}; - -struct disk_extent { - u64 ram_bytes; - u64 disk_bytenr; - u64 disk_num_bytes; - u64 offset; - u64 num_bytes; - u8 compression; - u8 encryption; - u16 other_encoding; -}; - -static int is_cowonly_root(u64 root_objectid) -{ - if (root_objectid == BTRFS_ROOT_TREE_OBJECTID || - root_objectid == BTRFS_EXTENT_TREE_OBJECTID || - root_objectid == BTRFS_CHUNK_TREE_OBJECTID || - root_objectid == BTRFS_DEV_TREE_OBJECTID || - root_objectid == BTRFS_TREE_LOG_OBJECTID || - root_objectid == BTRFS_CSUM_TREE_OBJECTID) - return 1; - return 0; -} - -static noinline int __next_ref_path(struct btrfs_trans_handle *trans, - struct btrfs_root *extent_root, - struct btrfs_ref_path *ref_path, - int first_time) -{ - struct extent_buffer *leaf; - struct btrfs_path *path; - struct btrfs_extent_ref *ref; - struct btrfs_key key; - struct btrfs_key found_key; - u64 bytenr; - u32 nritems; - int level; - int ret = 1; - - path = btrfs_alloc_path(); - if (!path) - return -ENOMEM; - - if (first_time) { - ref_path->lowest_level = -1; - ref_path->current_level = -1; - ref_path->shared_level = -1; - goto walk_up; - } -walk_down: - level = ref_path->current_level - 1; - while (level >= -1) { - u64 parent; - if (level < ref_path->lowest_level) - break; - - if (level >= 0) - bytenr = ref_path->nodes[level]; - else - bytenr = ref_path->extent_start; - BUG_ON(bytenr == 0); - - parent = ref_path->nodes[level + 1]; - ref_path->nodes[level + 1] = 0; - ref_path->current_level = level; - BUG_ON(parent == 0); - - key.objectid = bytenr; - key.offset = parent + 1; - key.type = BTRFS_EXTENT_REF_KEY; - - ret = btrfs_search_slot(trans, extent_root, &key, path, 0, 0); - if (ret < 0) - goto out; - BUG_ON(ret == 0); - - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - if (path->slots[0] >= nritems) { - ret = btrfs_next_leaf(extent_root, path); - if (ret < 0) - goto out; - if (ret > 0) - goto next; - leaf = path->nodes[0]; - } - - btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); - if (found_key.objectid == bytenr && - found_key.type == BTRFS_EXTENT_REF_KEY) { - if (level < ref_path->shared_level) - ref_path->shared_level = level; - goto found; - } -next: - level--; - btrfs_release_path(extent_root, path); - cond_resched(); - } - /* reached lowest level */ - ret = 1; - goto out; -walk_up: - level = ref_path->current_level; - while (level < BTRFS_MAX_LEVEL - 1) { - u64 ref_objectid; - - if (level >= 0) - bytenr = ref_path->nodes[level]; - else - bytenr = ref_path->extent_start; - - BUG_ON(bytenr == 0); - - key.objectid = bytenr; - key.offset = 0; - key.type = BTRFS_EXTENT_REF_KEY; - - ret = btrfs_search_slot(trans, extent_root, &key, path, 0, 0); - if (ret < 0) - goto out; - - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - if (path->slots[0] >= nritems) { - ret = btrfs_next_leaf(extent_root, path); - if (ret < 0) - goto out; - if (ret > 0) { - /* the extent was freed by someone */ - if (ref_path->lowest_level == level) - goto out; - btrfs_release_path(extent_root, path); - goto walk_down; - } - leaf = path->nodes[0]; - } - - btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); - if (found_key.objectid != bytenr || - found_key.type != BTRFS_EXTENT_REF_KEY) { - /* the extent was freed by someone */ - if (ref_path->lowest_level == level) { - ret = 1; - goto out; - } - btrfs_release_path(extent_root, path); - goto walk_down; - } -found: - ref = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_extent_ref); - ref_objectid = btrfs_ref_objectid(leaf, ref); - if (ref_objectid < BTRFS_FIRST_FREE_OBJECTID) { - if (first_time) { - level = (int)ref_objectid; - BUG_ON(level >= BTRFS_MAX_LEVEL); - ref_path->lowest_level = level; - ref_path->current_level = level; - ref_path->nodes[level] = bytenr; - } else { - WARN_ON(ref_objectid != level); - } - } else { - WARN_ON(level != -1); - } - first_time = 0; - - if (ref_path->lowest_level == level) { - ref_path->owner_objectid = ref_objectid; - ref_path->num_refs = btrfs_ref_num_refs(leaf, ref); - } - - /* - * the block is tree root or the block isn't in reference - * counted tree. - */ - if (found_key.objectid == found_key.offset || - is_cowonly_root(btrfs_ref_root(leaf, ref))) { - ref_path->root_objectid = btrfs_ref_root(leaf, ref); - ref_path->root_generation = - btrfs_ref_generation(leaf, ref); - if (level < 0) { - /* special reference from the tree log */ - ref_path->nodes[0] = found_key.offset; - ref_path->current_level = 0; - } - ret = 0; - goto out; - } - - level++; - BUG_ON(ref_path->nodes[level] != 0); - ref_path->nodes[level] = found_key.offset; - ref_path->current_level = level; - - /* - * the reference was created in the running transaction, - * no need to continue walking up. - */ - if (btrfs_ref_generation(leaf, ref) == trans->transid) { - ref_path->root_objectid = btrfs_ref_root(leaf, ref); - ref_path->root_generation = - btrfs_ref_generation(leaf, ref); - ret = 0; - goto out; - } - - btrfs_release_path(extent_root, path); - cond_resched(); - } - /* reached max tree level, but no tree root found. */ - BUG(); -out: - btrfs_free_path(path); - return ret; -} - -static int btrfs_first_ref_path(struct btrfs_trans_handle *trans, - struct btrfs_root *extent_root, - struct btrfs_ref_path *ref_path, - u64 extent_start) -{ - memset(ref_path, 0, sizeof(*ref_path)); - ref_path->extent_start = extent_start; - - return __next_ref_path(trans, extent_root, ref_path, 1); -} - -static int btrfs_next_ref_path(struct btrfs_trans_handle *trans, - struct btrfs_root *extent_root, - struct btrfs_ref_path *ref_path) -{ - return __next_ref_path(trans, extent_root, ref_path, 0); -} - -static noinline int get_new_locations(struct inode *reloc_inode, - struct btrfs_key *extent_key, - u64 offset, int no_fragment, - struct disk_extent **extents, - int *nr_extents) -{ - struct btrfs_root *root = BTRFS_I(reloc_inode)->root; - struct btrfs_path *path; - struct btrfs_file_extent_item *fi; - struct extent_buffer *leaf; - struct disk_extent *exts = *extents; - struct btrfs_key found_key; - u64 cur_pos; - u64 last_byte; - u32 nritems; - int nr = 0; - int max = *nr_extents; - int ret; - - WARN_ON(!no_fragment && *extents); - if (!exts) { - max = 1; - exts = kmalloc(sizeof(*exts) * max, GFP_NOFS); - if (!exts) - return -ENOMEM; - } - - path = btrfs_alloc_path(); - if (!path) { - if (exts != *extents) - kfree(exts); - return -ENOMEM; - } - - cur_pos = extent_key->objectid - offset; - last_byte = extent_key->objectid + extent_key->offset; - ret = btrfs_lookup_file_extent(NULL, root, path, reloc_inode->i_ino, - cur_pos, 0); - if (ret < 0) - goto out; - if (ret > 0) { - ret = -ENOENT; - goto out; - } - - while (1) { - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - if (path->slots[0] >= nritems) { - ret = btrfs_next_leaf(root, path); - if (ret < 0) - goto out; - if (ret > 0) - break; - leaf = path->nodes[0]; - } - - btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); - if (found_key.offset != cur_pos || - found_key.type != BTRFS_EXTENT_DATA_KEY || - found_key.objectid != reloc_inode->i_ino) - break; - - fi = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_file_extent_item); - if (btrfs_file_extent_type(leaf, fi) != - BTRFS_FILE_EXTENT_REG || - btrfs_file_extent_disk_bytenr(leaf, fi) == 0) - break; - - if (nr == max) { - struct disk_extent *old = exts; - max *= 2; - exts = kzalloc(sizeof(*exts) * max, GFP_NOFS); - if (!exts) { - ret = -ENOMEM; - goto out; - } - memcpy(exts, old, sizeof(*exts) * nr); - if (old != *extents) - kfree(old); - } - - exts[nr].disk_bytenr = - btrfs_file_extent_disk_bytenr(leaf, fi); - exts[nr].disk_num_bytes = - btrfs_file_extent_disk_num_bytes(leaf, fi); - exts[nr].offset = btrfs_file_extent_offset(leaf, fi); - exts[nr].num_bytes = btrfs_file_extent_num_bytes(leaf, fi); - exts[nr].ram_bytes = btrfs_file_extent_ram_bytes(leaf, fi); - exts[nr].compression = btrfs_file_extent_compression(leaf, fi); - exts[nr].encryption = btrfs_file_extent_encryption(leaf, fi); - exts[nr].other_encoding = btrfs_file_extent_other_encoding(leaf, - fi); - BUG_ON(exts[nr].offset > 0); - BUG_ON(exts[nr].compression || exts[nr].encryption); - BUG_ON(exts[nr].num_bytes != exts[nr].disk_num_bytes); - - cur_pos += exts[nr].num_bytes; - nr++; - - if (cur_pos + offset >= last_byte) - break; - - if (no_fragment) { - ret = 1; - goto out; - } - path->slots[0]++; - } - - BUG_ON(cur_pos + offset > last_byte); - if (cur_pos + offset < last_byte) { - ret = -ENOENT; - goto out; - } - ret = 0; -out: - btrfs_free_path(path); - if (ret) { - if (exts != *extents) - kfree(exts); - } else { - *extents = exts; - *nr_extents = nr; - } - return ret; -} - -static noinline int replace_one_extent(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_path *path, - struct btrfs_key *extent_key, - struct btrfs_key *leaf_key, - struct btrfs_ref_path *ref_path, - struct disk_extent *new_extents, - int nr_extents) -{ - struct extent_buffer *leaf; - struct btrfs_file_extent_item *fi; - struct inode *inode = NULL; - struct btrfs_key key; - u64 lock_start = 0; - u64 lock_end = 0; - u64 num_bytes; - u64 ext_offset; - u64 search_end = (u64)-1; - u32 nritems; - int nr_scaned = 0; - int extent_locked = 0; - int extent_type; - int ret; - - memcpy(&key, leaf_key, sizeof(key)); - if (ref_path->owner_objectid != BTRFS_MULTIPLE_OBJECTIDS) { - if (key.objectid < ref_path->owner_objectid || - (key.objectid == ref_path->owner_objectid && - key.type < BTRFS_EXTENT_DATA_KEY)) { - key.objectid = ref_path->owner_objectid; - key.type = BTRFS_EXTENT_DATA_KEY; - key.offset = 0; - } - } - - while (1) { - ret = btrfs_search_slot(trans, root, &key, path, 0, 1); - if (ret < 0) - goto out; - - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); -next: - if (extent_locked && ret > 0) { - /* - * the file extent item was modified by someone - * before the extent got locked. - */ - unlock_extent(&BTRFS_I(inode)->io_tree, lock_start, - lock_end, GFP_NOFS); - extent_locked = 0; - } - - if (path->slots[0] >= nritems) { - if (++nr_scaned > 2) - break; - - BUG_ON(extent_locked); - ret = btrfs_next_leaf(root, path); - if (ret < 0) - goto out; - if (ret > 0) - break; - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - } - - btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - - if (ref_path->owner_objectid != BTRFS_MULTIPLE_OBJECTIDS) { - if ((key.objectid > ref_path->owner_objectid) || - (key.objectid == ref_path->owner_objectid && - key.type > BTRFS_EXTENT_DATA_KEY) || - key.offset >= search_end) - break; - } - - if (inode && key.objectid != inode->i_ino) { - BUG_ON(extent_locked); - btrfs_release_path(root, path); - mutex_unlock(&inode->i_mutex); - iput(inode); - inode = NULL; - continue; - } - - if (key.type != BTRFS_EXTENT_DATA_KEY) { - path->slots[0]++; - ret = 1; - goto next; - } - fi = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_file_extent_item); - extent_type = btrfs_file_extent_type(leaf, fi); - if ((extent_type != BTRFS_FILE_EXTENT_REG && - extent_type != BTRFS_FILE_EXTENT_PREALLOC) || - (btrfs_file_extent_disk_bytenr(leaf, fi) != - extent_key->objectid)) { - path->slots[0]++; - ret = 1; - goto next; - } - - num_bytes = btrfs_file_extent_num_bytes(leaf, fi); - ext_offset = btrfs_file_extent_offset(leaf, fi); - - if (search_end == (u64)-1) { - search_end = key.offset - ext_offset + - btrfs_file_extent_ram_bytes(leaf, fi); - } - - if (!extent_locked) { - lock_start = key.offset; - lock_end = lock_start + num_bytes - 1; - } else { - if (lock_start > key.offset || - lock_end + 1 < key.offset + num_bytes) { - unlock_extent(&BTRFS_I(inode)->io_tree, - lock_start, lock_end, GFP_NOFS); - extent_locked = 0; - } - } - - if (!inode) { - btrfs_release_path(root, path); - - inode = btrfs_iget_locked(root->fs_info->sb, - key.objectid, root); - if (inode->i_state & I_NEW) { - BTRFS_I(inode)->root = root; - BTRFS_I(inode)->location.objectid = - key.objectid; - BTRFS_I(inode)->location.type = - BTRFS_INODE_ITEM_KEY; - BTRFS_I(inode)->location.offset = 0; - btrfs_read_locked_inode(inode); - unlock_new_inode(inode); - } - /* - * some code call btrfs_commit_transaction while - * holding the i_mutex, so we can't use mutex_lock - * here. - */ - if (is_bad_inode(inode) || - !mutex_trylock(&inode->i_mutex)) { - iput(inode); - inode = NULL; - key.offset = (u64)-1; - goto skip; - } - } - - if (!extent_locked) { - struct btrfs_ordered_extent *ordered; - - btrfs_release_path(root, path); - - lock_extent(&BTRFS_I(inode)->io_tree, lock_start, - lock_end, GFP_NOFS); - ordered = btrfs_lookup_first_ordered_extent(inode, - lock_end); - if (ordered && - ordered->file_offset <= lock_end && - ordered->file_offset + ordered->len > lock_start) { - unlock_extent(&BTRFS_I(inode)->io_tree, - lock_start, lock_end, GFP_NOFS); - btrfs_start_ordered_extent(inode, ordered, 1); - btrfs_put_ordered_extent(ordered); - key.offset += num_bytes; - goto skip; - } - if (ordered) - btrfs_put_ordered_extent(ordered); - - extent_locked = 1; - continue; - } - - if (nr_extents == 1) { - /* update extent pointer in place */ - btrfs_set_file_extent_disk_bytenr(leaf, fi, - new_extents[0].disk_bytenr); - btrfs_set_file_extent_disk_num_bytes(leaf, fi, - new_extents[0].disk_num_bytes); - btrfs_mark_buffer_dirty(leaf); - - btrfs_drop_extent_cache(inode, key.offset, - key.offset + num_bytes - 1, 0); - - ret = btrfs_inc_extent_ref(trans, root, - new_extents[0].disk_bytenr, - new_extents[0].disk_num_bytes, - leaf->start, - root->root_key.objectid, - trans->transid, - key.objectid); - BUG_ON(ret); - - ret = btrfs_free_extent(trans, root, - extent_key->objectid, - extent_key->offset, - leaf->start, - btrfs_header_owner(leaf), - btrfs_header_generation(leaf), - key.objectid, 0); - BUG_ON(ret); - - btrfs_release_path(root, path); - key.offset += num_bytes; - } else { - BUG_ON(1); -#if 0 - u64 alloc_hint; - u64 extent_len; - int i; - /* - * drop old extent pointer at first, then insert the - * new pointers one bye one - */ - btrfs_release_path(root, path); - ret = btrfs_drop_extents(trans, root, inode, key.offset, - key.offset + num_bytes, - key.offset, &alloc_hint); - BUG_ON(ret); - - for (i = 0; i < nr_extents; i++) { - if (ext_offset >= new_extents[i].num_bytes) { - ext_offset -= new_extents[i].num_bytes; - continue; - } - extent_len = min(new_extents[i].num_bytes - - ext_offset, num_bytes); - - ret = btrfs_insert_empty_item(trans, root, - path, &key, - sizeof(*fi)); - BUG_ON(ret); - - leaf = path->nodes[0]; - fi = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_file_extent_item); - btrfs_set_file_extent_generation(leaf, fi, - trans->transid); - btrfs_set_file_extent_type(leaf, fi, - BTRFS_FILE_EXTENT_REG); - btrfs_set_file_extent_disk_bytenr(leaf, fi, - new_extents[i].disk_bytenr); - btrfs_set_file_extent_disk_num_bytes(leaf, fi, - new_extents[i].disk_num_bytes); - btrfs_set_file_extent_ram_bytes(leaf, fi, - new_extents[i].ram_bytes); - - btrfs_set_file_extent_compression(leaf, fi, - new_extents[i].compression); - btrfs_set_file_extent_encryption(leaf, fi, - new_extents[i].encryption); - btrfs_set_file_extent_other_encoding(leaf, fi, - new_extents[i].other_encoding); - - btrfs_set_file_extent_num_bytes(leaf, fi, - extent_len); - ext_offset += new_extents[i].offset; - btrfs_set_file_extent_offset(leaf, fi, - ext_offset); - btrfs_mark_buffer_dirty(leaf); - - btrfs_drop_extent_cache(inode, key.offset, - key.offset + extent_len - 1, 0); - - ret = btrfs_inc_extent_ref(trans, root, - new_extents[i].disk_bytenr, - new_extents[i].disk_num_bytes, - leaf->start, - root->root_key.objectid, - trans->transid, key.objectid); - BUG_ON(ret); - btrfs_release_path(root, path); - - inode_add_bytes(inode, extent_len); - - ext_offset = 0; - num_bytes -= extent_len; - key.offset += extent_len; - - if (num_bytes == 0) - break; - } - BUG_ON(i >= nr_extents); -#endif - } - - if (extent_locked) { - unlock_extent(&BTRFS_I(inode)->io_tree, lock_start, - lock_end, GFP_NOFS); - extent_locked = 0; - } -skip: - if (ref_path->owner_objectid != BTRFS_MULTIPLE_OBJECTIDS && - key.offset >= search_end) - break; - - cond_resched(); - } - ret = 0; -out: - btrfs_release_path(root, path); - if (inode) { - mutex_unlock(&inode->i_mutex); - if (extent_locked) { - unlock_extent(&BTRFS_I(inode)->io_tree, lock_start, - lock_end, GFP_NOFS); - } - iput(inode); - } - return ret; -} - -int btrfs_reloc_tree_cache_ref(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct extent_buffer *buf, u64 orig_start) -{ - int level; - int ret; - - BUG_ON(btrfs_header_generation(buf) != trans->transid); - BUG_ON(root->root_key.objectid != BTRFS_TREE_RELOC_OBJECTID); - - level = btrfs_header_level(buf); - if (level == 0) { - struct btrfs_leaf_ref *ref; - struct btrfs_leaf_ref *orig_ref; - - orig_ref = btrfs_lookup_leaf_ref(root, orig_start); - if (!orig_ref) - return -ENOENT; - - ref = btrfs_alloc_leaf_ref(root, orig_ref->nritems); - if (!ref) { - btrfs_free_leaf_ref(root, orig_ref); - return -ENOMEM; - } - - ref->nritems = orig_ref->nritems; - memcpy(ref->extents, orig_ref->extents, - sizeof(ref->extents[0]) * ref->nritems); - - btrfs_free_leaf_ref(root, orig_ref); - - ref->root_gen = trans->transid; - ref->bytenr = buf->start; - ref->owner = btrfs_header_owner(buf); - ref->generation = btrfs_header_generation(buf); - - ret = btrfs_add_leaf_ref(root, ref, 0); - WARN_ON(ret); - btrfs_free_leaf_ref(root, ref); - } - return 0; -} - -static noinline int invalidate_extent_cache(struct btrfs_root *root, - struct extent_buffer *leaf, - struct btrfs_block_group_cache *group, - struct btrfs_root *target_root) -{ - struct btrfs_key key; - struct inode *inode = NULL; - struct btrfs_file_extent_item *fi; - struct extent_state *cached_state = NULL; - u64 num_bytes; - u64 skip_objectid = 0; - u32 nritems; - u32 i; - - nritems = btrfs_header_nritems(leaf); - for (i = 0; i < nritems; i++) { - btrfs_item_key_to_cpu(leaf, &key, i); - if (key.objectid == skip_objectid || - key.type != BTRFS_EXTENT_DATA_KEY) - continue; - fi = btrfs_item_ptr(leaf, i, struct btrfs_file_extent_item); - if (btrfs_file_extent_type(leaf, fi) == - BTRFS_FILE_EXTENT_INLINE) - continue; - if (btrfs_file_extent_disk_bytenr(leaf, fi) == 0) - continue; - if (!inode || inode->i_ino != key.objectid) { - iput(inode); - inode = btrfs_ilookup(target_root->fs_info->sb, - key.objectid, target_root, 1); - } - if (!inode) { - skip_objectid = key.objectid; - continue; - } - num_bytes = btrfs_file_extent_num_bytes(leaf, fi); - - lock_extent_bits(&BTRFS_I(inode)->io_tree, key.offset, - key.offset + num_bytes - 1, 0, &cached_state, - GFP_NOFS); - btrfs_drop_extent_cache(inode, key.offset, - key.offset + num_bytes - 1, 1); - unlock_extent_cached(&BTRFS_I(inode)->io_tree, key.offset, - key.offset + num_bytes - 1, &cached_state, - GFP_NOFS); - cond_resched(); - } - iput(inode); - return 0; -} - -static noinline int replace_extents_in_leaf(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct extent_buffer *leaf, - struct btrfs_block_group_cache *group, - struct inode *reloc_inode) -{ - struct btrfs_key key; - struct btrfs_key extent_key; - struct btrfs_file_extent_item *fi; - struct btrfs_leaf_ref *ref; - struct disk_extent *new_extent; - u64 bytenr; - u64 num_bytes; - u32 nritems; - u32 i; - int ext_index; - int nr_extent; - int ret; - - new_extent = kmalloc(sizeof(*new_extent), GFP_NOFS); - if (!new_extent) - return -ENOMEM; - - ref = btrfs_lookup_leaf_ref(root, leaf->start); - BUG_ON(!ref); - - ext_index = -1; - nritems = btrfs_header_nritems(leaf); - for (i = 0; i < nritems; i++) { - btrfs_item_key_to_cpu(leaf, &key, i); - if (btrfs_key_type(&key) != BTRFS_EXTENT_DATA_KEY) - continue; - fi = btrfs_item_ptr(leaf, i, struct btrfs_file_extent_item); - if (btrfs_file_extent_type(leaf, fi) == - BTRFS_FILE_EXTENT_INLINE) - continue; - bytenr = btrfs_file_extent_disk_bytenr(leaf, fi); - num_bytes = btrfs_file_extent_disk_num_bytes(leaf, fi); - if (bytenr == 0) - continue; - - ext_index++; - if (bytenr >= group->key.objectid + group->key.offset || - bytenr + num_bytes <= group->key.objectid) - continue; - - extent_key.objectid = bytenr; - extent_key.offset = num_bytes; - extent_key.type = BTRFS_EXTENT_ITEM_KEY; - nr_extent = 1; - ret = get_new_locations(reloc_inode, &extent_key, - group->key.objectid, 1, - &new_extent, &nr_extent); - if (ret > 0) - continue; - BUG_ON(ret < 0); - - BUG_ON(ref->extents[ext_index].bytenr != bytenr); - BUG_ON(ref->extents[ext_index].num_bytes != num_bytes); - ref->extents[ext_index].bytenr = new_extent->disk_bytenr; - ref->extents[ext_index].num_bytes = new_extent->disk_num_bytes; - - btrfs_set_file_extent_disk_bytenr(leaf, fi, - new_extent->disk_bytenr); - btrfs_set_file_extent_disk_num_bytes(leaf, fi, - new_extent->disk_num_bytes); - btrfs_mark_buffer_dirty(leaf); - - ret = btrfs_inc_extent_ref(trans, root, - new_extent->disk_bytenr, - new_extent->disk_num_bytes, - leaf->start, - root->root_key.objectid, - trans->transid, key.objectid); - BUG_ON(ret); - - ret = btrfs_free_extent(trans, root, - bytenr, num_bytes, leaf->start, - btrfs_header_owner(leaf), - btrfs_header_generation(leaf), - key.objectid, 0); - BUG_ON(ret); - cond_resched(); - } - kfree(new_extent); - BUG_ON(ext_index + 1 != ref->nritems); - btrfs_free_leaf_ref(root, ref); - return 0; -} - -int btrfs_free_reloc_root(struct btrfs_trans_handle *trans, - struct btrfs_root *root) -{ - struct btrfs_root *reloc_root; - int ret; - - if (root->reloc_root) { - reloc_root = root->reloc_root; - root->reloc_root = NULL; - list_add(&reloc_root->dead_list, - &root->fs_info->dead_reloc_roots); - - btrfs_set_root_bytenr(&reloc_root->root_item, - reloc_root->node->start); - btrfs_set_root_level(&root->root_item, - btrfs_header_level(reloc_root->node)); - memset(&reloc_root->root_item.drop_progress, 0, - sizeof(struct btrfs_disk_key)); - reloc_root->root_item.drop_level = 0; - - ret = btrfs_update_root(trans, root->fs_info->tree_root, - &reloc_root->root_key, - &reloc_root->root_item); - BUG_ON(ret); - } - return 0; -} - -int btrfs_drop_dead_reloc_roots(struct btrfs_root *root) -{ - struct btrfs_trans_handle *trans; - struct btrfs_root *reloc_root; - struct btrfs_root *prev_root = NULL; - struct list_head dead_roots; - int ret; - unsigned long nr; - - INIT_LIST_HEAD(&dead_roots); - list_splice_init(&root->fs_info->dead_reloc_roots, &dead_roots); - - while (!list_empty(&dead_roots)) { - reloc_root = list_entry(dead_roots.prev, - struct btrfs_root, dead_list); - list_del_init(&reloc_root->dead_list); - - BUG_ON(reloc_root->commit_root != NULL); - while (1) { - trans = btrfs_join_transaction(root, 1); - BUG_ON(IS_ERR(trans)); - - mutex_lock(&root->fs_info->drop_mutex); - ret = btrfs_drop_snapshot(trans, reloc_root); - if (ret != -EAGAIN) - break; - mutex_unlock(&root->fs_info->drop_mutex); - - nr = trans->blocks_used; - ret = btrfs_end_transaction(trans, root); - BUG_ON(ret); - btrfs_btree_balance_dirty(root, nr); - } - - free_extent_buffer(reloc_root->node); - - ret = btrfs_del_root(trans, root->fs_info->tree_root, - &reloc_root->root_key); - BUG_ON(ret); - mutex_unlock(&root->fs_info->drop_mutex); - - nr = trans->blocks_used; - ret = btrfs_end_transaction(trans, root); - BUG_ON(ret); - btrfs_btree_balance_dirty(root, nr); - - kfree(prev_root); - prev_root = reloc_root; - } - if (prev_root) { - btrfs_remove_leaf_refs(prev_root, (u64)-1, 0); - kfree(prev_root); - } - return 0; -} - -int btrfs_add_dead_reloc_root(struct btrfs_root *root) -{ - list_add(&root->dead_list, &root->fs_info->dead_reloc_roots); - return 0; -} - -int btrfs_cleanup_reloc_trees(struct btrfs_root *root) -{ - struct btrfs_root *reloc_root; - struct btrfs_trans_handle *trans; - struct btrfs_key location; - int found; - int ret; - - mutex_lock(&root->fs_info->tree_reloc_mutex); - ret = btrfs_find_dead_roots(root, BTRFS_TREE_RELOC_OBJECTID, NULL); - BUG_ON(ret); - found = !list_empty(&root->fs_info->dead_reloc_roots); - mutex_unlock(&root->fs_info->tree_reloc_mutex); - - if (found) { - trans = btrfs_start_transaction(root, 1); - BUG_ON(IS_ERR(trans)); - ret = btrfs_commit_transaction(trans, root); - BUG_ON(ret); - } - - location.objectid = BTRFS_DATA_RELOC_TREE_OBJECTID; - location.offset = (u64)-1; - location.type = BTRFS_ROOT_ITEM_KEY; - - reloc_root = btrfs_read_fs_root_no_name(root->fs_info, &location); - BUG_ON(!reloc_root); - ret = btrfs_orphan_cleanup(reloc_root); - BUG_ON(ret); - return 0; -} - -static noinline int init_reloc_tree(struct btrfs_trans_handle *trans, - struct btrfs_root *root) -{ - struct btrfs_root *reloc_root; - struct extent_buffer *eb; - struct btrfs_root_item *root_item; - struct btrfs_key root_key; - int ret; - - BUG_ON(!root->ref_cows); - if (root->reloc_root) - return 0; - - root_item = kmalloc(sizeof(*root_item), GFP_NOFS); - if (!root_item) - return -ENOMEM; - - ret = btrfs_copy_root(trans, root, root->commit_root, - &eb, BTRFS_TREE_RELOC_OBJECTID); - BUG_ON(ret); - - root_key.objectid = BTRFS_TREE_RELOC_OBJECTID; - root_key.offset = root->root_key.objectid; - root_key.type = BTRFS_ROOT_ITEM_KEY; - - memcpy(root_item, &root->root_item, sizeof(root_item)); - btrfs_set_root_refs(root_item, 0); - btrfs_set_root_bytenr(root_item, eb->start); - btrfs_set_root_level(root_item, btrfs_header_level(eb)); - btrfs_set_root_generation(root_item, trans->transid); - - btrfs_tree_unlock(eb); - free_extent_buffer(eb); - - ret = btrfs_insert_root(trans, root->fs_info->tree_root, - &root_key, root_item); - BUG_ON(ret); - kfree(root_item); - - reloc_root = btrfs_read_fs_root_no_radix(root->fs_info->tree_root, - &root_key); - BUG_ON(IS_ERR(reloc_root)); - reloc_root->last_trans = trans->transid; - reloc_root->commit_root = NULL; - reloc_root->ref_tree = &root->fs_info->reloc_ref_tree; - - root->reloc_root = reloc_root; - return 0; -} - -/* - * Core function of space balance. - * - * The idea is using reloc trees to relocate tree blocks in reference - * counted roots. There is one reloc tree for each subvol, and all - * reloc trees share same root key objectid. Reloc trees are snapshots - * of the latest committed roots of subvols (root->commit_root). - * - * To relocate a tree block referenced by a subvol, there are two steps. - * COW the block through subvol's reloc tree, then update block pointer - * in the subvol to point to the new block. Since all reloc trees share - * same root key objectid, doing special handing for tree blocks owned - * by them is easy. Once a tree block has been COWed in one reloc tree, - * we can use the resulting new block directly when the same block is - * required to COW again through other reloc trees. By this way, relocated - * tree blocks are shared between reloc trees, so they are also shared - * between subvols. - */ -static noinline int relocate_one_path(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_path *path, - struct btrfs_key *first_key, - struct btrfs_ref_path *ref_path, - struct btrfs_block_group_cache *group, - struct inode *reloc_inode) -{ - struct btrfs_root *reloc_root; - struct extent_buffer *eb = NULL; - struct btrfs_key *keys; - u64 *nodes; - int level; - int shared_level; - int lowest_level = 0; - int ret; - - if (ref_path->owner_objectid < BTRFS_FIRST_FREE_OBJECTID) - lowest_level = ref_path->owner_objectid; - - if (!root->ref_cows) { - path->lowest_level = lowest_level; - ret = btrfs_search_slot(trans, root, first_key, path, 0, 1); - BUG_ON(ret < 0); - path->lowest_level = 0; - btrfs_release_path(root, path); - return 0; - } - - mutex_lock(&root->fs_info->tree_reloc_mutex); - ret = init_reloc_tree(trans, root); - BUG_ON(ret); - reloc_root = root->reloc_root; - - shared_level = ref_path->shared_level; - ref_path->shared_level = BTRFS_MAX_LEVEL - 1; - - keys = ref_path->node_keys; - nodes = ref_path->new_nodes; - memset(&keys[shared_level + 1], 0, - sizeof(*keys) * (BTRFS_MAX_LEVEL - shared_level - 1)); - memset(&nodes[shared_level + 1], 0, - sizeof(*nodes) * (BTRFS_MAX_LEVEL - shared_level - 1)); - - if (nodes[lowest_level] == 0) { - path->lowest_level = lowest_level; - ret = btrfs_search_slot(trans, reloc_root, first_key, path, - 0, 1); - BUG_ON(ret); - for (level = lowest_level; level < BTRFS_MAX_LEVEL; level++) { - eb = path->nodes[level]; - if (!eb || eb == reloc_root->node) - break; - nodes[level] = eb->start; - if (level == 0) - btrfs_item_key_to_cpu(eb, &keys[level], 0); - else - btrfs_node_key_to_cpu(eb, &keys[level], 0); - } - if (nodes[0] && - ref_path->owner_objectid >= BTRFS_FIRST_FREE_OBJECTID) { - eb = path->nodes[0]; - ret = replace_extents_in_leaf(trans, reloc_root, eb, - group, reloc_inode); - BUG_ON(ret); - } - btrfs_release_path(reloc_root, path); - } else { - ret = btrfs_merge_path(trans, reloc_root, keys, nodes, - lowest_level); - BUG_ON(ret); - } - - /* - * replace tree blocks in the fs tree with tree blocks in - * the reloc tree. - */ - ret = btrfs_merge_path(trans, root, keys, nodes, lowest_level); - BUG_ON(ret < 0); - - if (ref_path->owner_objectid >= BTRFS_FIRST_FREE_OBJECTID) { - ret = btrfs_search_slot(trans, reloc_root, first_key, path, - 0, 0); - BUG_ON(ret); - extent_buffer_get(path->nodes[0]); - eb = path->nodes[0]; - btrfs_release_path(reloc_root, path); - ret = invalidate_extent_cache(reloc_root, eb, group, root); - BUG_ON(ret); - free_extent_buffer(eb); - } - - mutex_unlock(&root->fs_info->tree_reloc_mutex); - path->lowest_level = 0; - return 0; -} - -static noinline int relocate_tree_block(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_path *path, - struct btrfs_key *first_key, - struct btrfs_ref_path *ref_path) -{ - int ret; - - ret = relocate_one_path(trans, root, path, first_key, - ref_path, NULL, NULL); - BUG_ON(ret); - - return 0; -} - -static noinline int del_extent_zero(struct btrfs_trans_handle *trans, - struct btrfs_root *extent_root, - struct btrfs_path *path, - struct btrfs_key *extent_key) -{ - int ret; - - ret = btrfs_search_slot(trans, extent_root, extent_key, path, -1, 1); - if (ret) - goto out; - ret = btrfs_del_item(trans, extent_root, path); -out: - btrfs_release_path(extent_root, path); - return ret; -} - -static noinline struct btrfs_root *read_ref_root(struct btrfs_fs_info *fs_info, - struct btrfs_ref_path *ref_path) -{ - struct btrfs_key root_key; - - root_key.objectid = ref_path->root_objectid; - root_key.type = BTRFS_ROOT_ITEM_KEY; - if (is_cowonly_root(ref_path->root_objectid)) - root_key.offset = 0; - else - root_key.offset = (u64)-1; - - return btrfs_read_fs_root_no_name(fs_info, &root_key); -} - -static noinline int relocate_one_extent(struct btrfs_root *extent_root, - struct btrfs_path *path, - struct btrfs_key *extent_key, - struct btrfs_block_group_cache *group, - struct inode *reloc_inode, int pass) -{ - struct btrfs_trans_handle *trans; - struct btrfs_root *found_root; - struct btrfs_ref_path *ref_path = NULL; - struct disk_extent *new_extents = NULL; - int nr_extents = 0; - int loops; - int ret; - int level; - struct btrfs_key first_key; - u64 prev_block = 0; - - - trans = btrfs_start_transaction(extent_root, 1); - BUG_ON(IS_ERR(trans)); - - if (extent_key->objectid == 0) { - ret = del_extent_zero(trans, extent_root, path, extent_key); - goto out; - } - - ref_path = kmalloc(sizeof(*ref_path), GFP_NOFS); - if (!ref_path) { - ret = -ENOMEM; - goto out; - } - - for (loops = 0; ; loops++) { - if (loops == 0) { - ret = btrfs_first_ref_path(trans, extent_root, ref_path, - extent_key->objectid); - } else { - ret = btrfs_next_ref_path(trans, extent_root, ref_path); - } - if (ret < 0) - goto out; - if (ret > 0) - break; - - if (ref_path->root_objectid == BTRFS_TREE_LOG_OBJECTID || - ref_path->root_objectid == BTRFS_TREE_RELOC_OBJECTID) - continue; - - found_root = read_ref_root(extent_root->fs_info, ref_path); - BUG_ON(!found_root); - /* - * for reference counted tree, only process reference paths - * rooted at the latest committed root. - */ - if (found_root->ref_cows && - ref_path->root_generation != found_root->root_key.offset) - continue; - - if (ref_path->owner_objectid >= BTRFS_FIRST_FREE_OBJECTID) { - if (pass == 0) { - /* - * copy data extents to new locations - */ - u64 group_start = group->key.objectid; - ret = relocate_data_extent(reloc_inode, - extent_key, - group_start); - if (ret < 0) - goto out; - break; - } - level = 0; - } else { - level = ref_path->owner_objectid; - } - - if (prev_block != ref_path->nodes[level]) { - struct extent_buffer *eb; - u64 block_start = ref_path->nodes[level]; - u64 block_size = btrfs_level_size(found_root, level); - - eb = read_tree_block(found_root, block_start, - block_size, 0); - if (!eb) { - ret = -EIO; - goto out; - } - btrfs_tree_lock(eb); - BUG_ON(level != btrfs_header_level(eb)); - - if (level == 0) - btrfs_item_key_to_cpu(eb, &first_key, 0); - else - btrfs_node_key_to_cpu(eb, &first_key, 0); - - btrfs_tree_unlock(eb); - free_extent_buffer(eb); - prev_block = block_start; - } - - mutex_lock(&extent_root->fs_info->trans_mutex); - btrfs_record_root_in_trans(found_root); - mutex_unlock(&extent_root->fs_info->trans_mutex); - if (ref_path->owner_objectid >= BTRFS_FIRST_FREE_OBJECTID) { - /* - * try to update data extent references while - * keeping metadata shared between snapshots. - */ - if (pass == 1) { - ret = relocate_one_path(trans, found_root, - path, &first_key, ref_path, - group, reloc_inode); - if (ret < 0) - goto out; - continue; - } - /* - * use fallback method to process the remaining - * references. - */ - if (!new_extents) { - u64 group_start = group->key.objectid; - new_extents = kmalloc(sizeof(*new_extents), - GFP_NOFS); - nr_extents = 1; - ret = get_new_locations(reloc_inode, - extent_key, - group_start, 1, - &new_extents, - &nr_extents); - if (ret) - goto out; - } - ret = replace_one_extent(trans, found_root, - path, extent_key, - &first_key, ref_path, - new_extents, nr_extents); - } else { - ret = relocate_tree_block(trans, found_root, path, - &first_key, ref_path); - } - if (ret < 0) - goto out; - } - ret = 0; -out: - btrfs_end_transaction(trans, extent_root); - kfree(new_extents); - kfree(ref_path); - return ret; -} -#endif - static u64 update_block_group_flags(struct btrfs_root *root, u64 flags) { u64 num_devices; @@ -8109,13 +6531,15 @@ int btrfs_set_block_group_ro(struct btrfs_root *root, alloc_flags = update_block_group_flags(root, cache->flags); if (alloc_flags != cache->flags) - do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1); + do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, + CHUNK_ALLOC_FORCE); ret = set_block_group_ro(cache); if (!ret) goto out; alloc_flags = get_alloc_profile(root, cache->space_info->flags); - ret = do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1); + ret = do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, + CHUNK_ALLOC_FORCE); if (ret < 0) goto out; ret = set_block_group_ro(cache); @@ -8128,7 +6552,8 @@ int btrfs_force_chunk_alloc(struct btrfs_trans_handle *trans, struct btrfs_root *root, u64 type) { u64 alloc_flags = get_alloc_profile(root, type); - return do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, 1); + return do_chunk_alloc(trans, root, 2 * 1024 * 1024, alloc_flags, + CHUNK_ALLOC_FORCE); } /* @@ -8480,10 +6905,16 @@ int btrfs_read_block_groups(struct btrfs_root *root) ret = -ENOMEM; goto error; } + cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl), + GFP_NOFS); + if (!cache->free_space_ctl) { + kfree(cache); + ret = -ENOMEM; + goto error; + } atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); - spin_lock_init(&cache->tree_lock); cache->fs_info = info; INIT_LIST_HEAD(&cache->list); INIT_LIST_HEAD(&cache->cluster_list); @@ -8491,24 +6922,18 @@ int btrfs_read_block_groups(struct btrfs_root *root) if (need_clear) cache->disk_cache_state = BTRFS_DC_CLEAR; - /* - * we only want to have 32k of ram per block group for keeping - * track of free space, and if we pass 1/2 of that we want to - * start converting things over to using bitmaps - */ - cache->extents_thresh = ((1024 * 32) / 2) / - sizeof(struct btrfs_free_space); - read_extent_buffer(leaf, &cache->item, btrfs_item_ptr_offset(leaf, path->slots[0]), sizeof(cache->item)); memcpy(&cache->key, &found_key, sizeof(found_key)); key.objectid = found_key.objectid + found_key.offset; - btrfs_release_path(root, path); + btrfs_release_path(path); cache->flags = btrfs_block_group_flags(&cache->item); cache->sectorsize = root->sectorsize; + btrfs_init_free_space_ctl(cache); + /* * We need to exclude the super stripes now so that the space * info has super bytes accounted for, otherwise we'll think @@ -8595,6 +7020,12 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, cache = kzalloc(sizeof(*cache), GFP_NOFS); if (!cache) return -ENOMEM; + cache->free_space_ctl = kzalloc(sizeof(*cache->free_space_ctl), + GFP_NOFS); + if (!cache->free_space_ctl) { + kfree(cache); + return -ENOMEM; + } cache->key.objectid = chunk_offset; cache->key.offset = size; @@ -8602,19 +7033,13 @@ int btrfs_make_block_group(struct btrfs_trans_handle *trans, cache->sectorsize = root->sectorsize; cache->fs_info = root->fs_info; - /* - * we only want to have 32k of ram per block group for keeping track - * of free space, and if we pass 1/2 of that we want to start - * converting things over to using bitmaps - */ - cache->extents_thresh = ((1024 * 32) / 2) / - sizeof(struct btrfs_free_space); atomic_set(&cache->count, 1); spin_lock_init(&cache->lock); - spin_lock_init(&cache->tree_lock); INIT_LIST_HEAD(&cache->list); INIT_LIST_HEAD(&cache->cluster_list); + btrfs_init_free_space_ctl(cache); + btrfs_set_block_group_used(&cache->item, bytes_used); btrfs_set_block_group_chunk_objectid(&cache->item, chunk_objectid); cache->flags = type; @@ -8727,12 +7152,12 @@ int btrfs_remove_block_group(struct btrfs_trans_handle *trans, if (ret < 0) goto out; if (ret > 0) - btrfs_release_path(tree_root, path); + btrfs_release_path(path); if (ret == 0) { ret = btrfs_del_item(trans, tree_root, path); if (ret) goto out; - btrfs_release_path(tree_root, path); + btrfs_release_path(path); } spin_lock(&root->fs_info->block_group_cache_lock); @@ -8781,23 +7206,38 @@ out: int btrfs_init_space_info(struct btrfs_fs_info *fs_info) { struct btrfs_space_info *space_info; + struct btrfs_super_block *disk_super; + u64 features; + u64 flags; + int mixed = 0; int ret; - ret = update_space_info(fs_info, BTRFS_BLOCK_GROUP_SYSTEM, 0, 0, - &space_info); - if (ret) - return ret; + disk_super = &fs_info->super_copy; + if (!btrfs_super_root(disk_super)) + return 1; - ret = update_space_info(fs_info, BTRFS_BLOCK_GROUP_METADATA, 0, 0, - &space_info); - if (ret) - return ret; + features = btrfs_super_incompat_flags(disk_super); + if (features & BTRFS_FEATURE_INCOMPAT_MIXED_GROUPS) + mixed = 1; - ret = update_space_info(fs_info, BTRFS_BLOCK_GROUP_DATA, 0, 0, - &space_info); + flags = BTRFS_BLOCK_GROUP_SYSTEM; + ret = update_space_info(fs_info, flags, 0, 0, &space_info); if (ret) - return ret; + goto out; + if (mixed) { + flags = BTRFS_BLOCK_GROUP_METADATA | BTRFS_BLOCK_GROUP_DATA; + ret = update_space_info(fs_info, flags, 0, 0, &space_info); + } else { + flags = BTRFS_BLOCK_GROUP_METADATA; + ret = update_space_info(fs_info, flags, 0, 0, &space_info); + if (ret) + goto out; + + flags = BTRFS_BLOCK_GROUP_DATA; + ret = update_space_info(fs_info, flags, 0, 0, &space_info); + } +out: return ret; } diff --git a/fs/btrfs/extent_io.c b/fs/btrfs/extent_io.c index 20ddb28602a8..c5d9fbb92bc3 100644 --- a/fs/btrfs/extent_io.c +++ b/fs/btrfs/extent_io.c @@ -10,6 +10,8 @@ #include <linux/swap.h> #include <linux/writeback.h> #include <linux/pagevec.h> +#include <linux/prefetch.h> +#include <linux/cleancache.h> #include "extent_io.h" #include "extent_map.h" #include "compat.h" @@ -101,7 +103,7 @@ void extent_io_exit(void) } void extent_io_tree_init(struct extent_io_tree *tree, - struct address_space *mapping, gfp_t mask) + struct address_space *mapping) { tree->state = RB_ROOT; INIT_RADIX_TREE(&tree->buffer, GFP_ATOMIC); @@ -439,6 +441,15 @@ static int clear_state_bit(struct extent_io_tree *tree, return ret; } +static struct extent_state * +alloc_extent_state_atomic(struct extent_state *prealloc) +{ + if (!prealloc) + prealloc = alloc_extent_state(GFP_ATOMIC); + + return prealloc; +} + /* * clear some bits on a range in the tree. This may require splitting * or inserting elements in the tree, so the gfp mask is used to @@ -529,8 +540,8 @@ hit_next: */ if (state->start < start) { - if (!prealloc) - prealloc = alloc_extent_state(GFP_ATOMIC); + prealloc = alloc_extent_state_atomic(prealloc); + BUG_ON(!prealloc); err = split_state(tree, state, prealloc, start); BUG_ON(err == -EEXIST); prealloc = NULL; @@ -551,8 +562,8 @@ hit_next: * on the first half */ if (state->start <= end && state->end > end) { - if (!prealloc) - prealloc = alloc_extent_state(GFP_ATOMIC); + prealloc = alloc_extent_state_atomic(prealloc); + BUG_ON(!prealloc); err = split_state(tree, state, prealloc, end + 1); BUG_ON(err == -EEXIST); if (wake) @@ -690,6 +701,15 @@ static void cache_state(struct extent_state *state, } } +static void uncache_state(struct extent_state **cached_ptr) +{ + if (cached_ptr && (*cached_ptr)) { + struct extent_state *state = *cached_ptr; + *cached_ptr = NULL; + free_extent_state(state); + } +} + /* * set some bits on a range in the tree. This may require allocations or * sleeping, so the gfp mask is used to indicate what is allowed. @@ -716,8 +736,7 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, again: if (!prealloc && (mask & __GFP_WAIT)) { prealloc = alloc_extent_state(mask); - if (!prealloc) - return -ENOMEM; + BUG_ON(!prealloc); } spin_lock(&tree->lock); @@ -734,6 +753,8 @@ again: */ node = tree_search(tree, start); if (!node) { + prealloc = alloc_extent_state_atomic(prealloc); + BUG_ON(!prealloc); err = insert_state(tree, prealloc, start, end, &bits); prealloc = NULL; BUG_ON(err == -EEXIST); @@ -762,20 +783,18 @@ hit_next: if (err) goto out; + next_node = rb_next(node); cache_state(state, cached_state); merge_state(tree, state); if (last_end == (u64)-1) goto out; start = last_end + 1; - if (start < end && prealloc && !need_resched()) { - next_node = rb_next(node); - if (next_node) { - state = rb_entry(next_node, struct extent_state, - rb_node); - if (state->start == start) - goto hit_next; - } + if (next_node && start < end && prealloc && !need_resched()) { + state = rb_entry(next_node, struct extent_state, + rb_node); + if (state->start == start) + goto hit_next; } goto search_again; } @@ -802,6 +821,9 @@ hit_next: err = -EEXIST; goto out; } + + prealloc = alloc_extent_state_atomic(prealloc); + BUG_ON(!prealloc); err = split_state(tree, state, prealloc, start); BUG_ON(err == -EEXIST); prealloc = NULL; @@ -832,14 +854,25 @@ hit_next: this_end = end; else this_end = last_start - 1; + + prealloc = alloc_extent_state_atomic(prealloc); + BUG_ON(!prealloc); + + /* + * Avoid to free 'prealloc' if it can be merged with + * the later extent. + */ + atomic_inc(&prealloc->refs); err = insert_state(tree, prealloc, start, this_end, &bits); BUG_ON(err == -EEXIST); if (err) { + free_extent_state(prealloc); prealloc = NULL; goto out; } cache_state(prealloc, cached_state); + free_extent_state(prealloc); prealloc = NULL; start = this_end + 1; goto search_again; @@ -856,6 +889,9 @@ hit_next: err = -EEXIST; goto out; } + + prealloc = alloc_extent_state_atomic(prealloc); + BUG_ON(!prealloc); err = split_state(tree, state, prealloc, end + 1); BUG_ON(err == -EEXIST); @@ -932,18 +968,11 @@ int set_extent_new(struct extent_io_tree *tree, u64 start, u64 end, NULL, mask); } -static int clear_extent_new(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask) -{ - return clear_extent_bit(tree, start, end, EXTENT_NEW, 0, 0, - NULL, mask); -} - int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask) + struct extent_state **cached_state, gfp_t mask) { - return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, NULL, - NULL, mask); + return set_extent_bit(tree, start, end, EXTENT_UPTODATE, 0, + NULL, cached_state, mask); } static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start, @@ -954,11 +983,6 @@ static int clear_extent_uptodate(struct extent_io_tree *tree, u64 start, cached_state, mask); } -int wait_on_extent_writeback(struct extent_io_tree *tree, u64 start, u64 end) -{ - return wait_extent_bit(tree, start, end, EXTENT_WRITEBACK); -} - /* * either insert or lock state struct between start and end use mask to tell * us if waiting is desired. @@ -1012,33 +1036,13 @@ int unlock_extent_cached(struct extent_io_tree *tree, u64 start, u64 end, mask); } -int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask) +int unlock_extent(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask) { return clear_extent_bit(tree, start, end, EXTENT_LOCKED, 1, 0, NULL, mask); } /* - * helper function to set pages and extents in the tree dirty - */ -int set_range_dirty(struct extent_io_tree *tree, u64 start, u64 end) -{ - unsigned long index = start >> PAGE_CACHE_SHIFT; - unsigned long end_index = end >> PAGE_CACHE_SHIFT; - struct page *page; - - while (index <= end_index) { - page = find_get_page(tree->mapping, index); - BUG_ON(!page); - __set_page_dirty_nobuffers(page); - page_cache_release(page); - index++; - } - return 0; -} - -/* * helper function to set both pages and extents in the tree writeback */ static int set_range_writeback(struct extent_io_tree *tree, u64 start, u64 end) @@ -1735,6 +1739,9 @@ static void end_bio_extent_readpage(struct bio *bio, int err) do { struct page *page = bvec->bv_page; + struct extent_state *cached = NULL; + struct extent_state *state; + tree = &BTRFS_I(page->mapping->host)->io_tree; start = ((u64)page->index << PAGE_CACHE_SHIFT) + @@ -1749,9 +1756,20 @@ static void end_bio_extent_readpage(struct bio *bio, int err) if (++bvec <= bvec_end) prefetchw(&bvec->bv_page->flags); + spin_lock(&tree->lock); + state = find_first_extent_bit_state(tree, start, EXTENT_LOCKED); + if (state && state->start == start) { + /* + * take a reference on the state, unlock will drop + * the ref + */ + cache_state(state, &cached); + } + spin_unlock(&tree->lock); + if (uptodate && tree->ops && tree->ops->readpage_end_io_hook) { ret = tree->ops->readpage_end_io_hook(page, start, end, - NULL); + state); if (ret) uptodate = 0; } @@ -1764,15 +1782,16 @@ static void end_bio_extent_readpage(struct bio *bio, int err) test_bit(BIO_UPTODATE, &bio->bi_flags); if (err) uptodate = 0; + uncache_state(&cached); continue; } } if (uptodate) { - set_extent_uptodate(tree, start, end, + set_extent_uptodate(tree, start, end, &cached, GFP_ATOMIC); } - unlock_extent(tree, start, end, GFP_ATOMIC); + unlock_extent_cached(tree, start, end, &cached, GFP_ATOMIC); if (whole_page) { if (uptodate) { @@ -1796,44 +1815,6 @@ static void end_bio_extent_readpage(struct bio *bio, int err) bio_put(bio); } -/* - * IO done from prepare_write is pretty simple, we just unlock - * the structs in the extent tree when done, and set the uptodate bits - * as appropriate. - */ -static void end_bio_extent_preparewrite(struct bio *bio, int err) -{ - const int uptodate = test_bit(BIO_UPTODATE, &bio->bi_flags); - struct bio_vec *bvec = bio->bi_io_vec + bio->bi_vcnt - 1; - struct extent_io_tree *tree; - u64 start; - u64 end; - - do { - struct page *page = bvec->bv_page; - tree = &BTRFS_I(page->mapping->host)->io_tree; - - start = ((u64)page->index << PAGE_CACHE_SHIFT) + - bvec->bv_offset; - end = start + bvec->bv_len - 1; - - if (--bvec >= bio->bi_io_vec) - prefetchw(&bvec->bv_page->flags); - - if (uptodate) { - set_extent_uptodate(tree, start, end, GFP_ATOMIC); - } else { - ClearPageUptodate(page); - SetPageError(page); - } - - unlock_extent(tree, start, end, GFP_ATOMIC); - - } while (bvec >= bio->bi_io_vec); - - bio_put(bio); -} - struct bio * btrfs_bio_alloc(struct block_device *bdev, u64 first_sector, int nr_vecs, gfp_t gfp_flags) @@ -1982,7 +1963,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree, struct btrfs_ordered_extent *ordered; int ret; int nr = 0; - size_t page_offset = 0; + size_t pg_offset = 0; size_t iosize; size_t disk_io_size; size_t blocksize = inode->i_sb->s_blocksize; @@ -1990,6 +1971,13 @@ static int __extent_read_full_page(struct extent_io_tree *tree, set_page_extent_mapped(page); + if (!PageUptodate(page)) { + if (cleancache_get_page(page) == 0) { + BUG_ON(blocksize != PAGE_SIZE); + goto out; + } + } + end = page_end; while (1) { lock_extent(tree, start, end, GFP_NOFS); @@ -2016,19 +2004,22 @@ static int __extent_read_full_page(struct extent_io_tree *tree, while (cur <= end) { if (cur >= last_byte) { char *userpage; - iosize = PAGE_CACHE_SIZE - page_offset; + struct extent_state *cached = NULL; + + iosize = PAGE_CACHE_SIZE - pg_offset; userpage = kmap_atomic(page, KM_USER0); - memset(userpage + page_offset, 0, iosize); + memset(userpage + pg_offset, 0, iosize); flush_dcache_page(page); kunmap_atomic(userpage, KM_USER0); set_extent_uptodate(tree, cur, cur + iosize - 1, - GFP_NOFS); - unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); + &cached, GFP_NOFS); + unlock_extent_cached(tree, cur, cur + iosize - 1, + &cached, GFP_NOFS); break; } - em = get_extent(inode, page, page_offset, cur, + em = get_extent(inode, page, pg_offset, cur, end - cur + 1, 0); - if (IS_ERR(em) || !em) { + if (IS_ERR_OR_NULL(em)) { SetPageError(page); unlock_extent(tree, cur, end, GFP_NOFS); break; @@ -2063,16 +2054,19 @@ static int __extent_read_full_page(struct extent_io_tree *tree, /* we've found a hole, just zero and go on */ if (block_start == EXTENT_MAP_HOLE) { char *userpage; + struct extent_state *cached = NULL; + userpage = kmap_atomic(page, KM_USER0); - memset(userpage + page_offset, 0, iosize); + memset(userpage + pg_offset, 0, iosize); flush_dcache_page(page); kunmap_atomic(userpage, KM_USER0); set_extent_uptodate(tree, cur, cur + iosize - 1, - GFP_NOFS); - unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); + &cached, GFP_NOFS); + unlock_extent_cached(tree, cur, cur + iosize - 1, + &cached, GFP_NOFS); cur = cur + iosize; - page_offset += iosize; + pg_offset += iosize; continue; } /* the get_extent function already copied into the page */ @@ -2081,7 +2075,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree, check_page_uptodate(tree, page); unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); cur = cur + iosize; - page_offset += iosize; + pg_offset += iosize; continue; } /* we have an inline extent but it didn't get marked up @@ -2091,7 +2085,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree, SetPageError(page); unlock_extent(tree, cur, cur + iosize - 1, GFP_NOFS); cur = cur + iosize; - page_offset += iosize; + pg_offset += iosize; continue; } @@ -2104,7 +2098,7 @@ static int __extent_read_full_page(struct extent_io_tree *tree, unsigned long pnr = (last_byte >> PAGE_CACHE_SHIFT) + 1; pnr -= page->index; ret = submit_extent_page(READ, tree, page, - sector, disk_io_size, page_offset, + sector, disk_io_size, pg_offset, bdev, bio, pnr, end_bio_extent_readpage, mirror_num, *bio_flags, @@ -2115,8 +2109,9 @@ static int __extent_read_full_page(struct extent_io_tree *tree, if (ret) SetPageError(page); cur = cur + iosize; - page_offset += iosize; + pg_offset += iosize; } +out: if (!nr) { if (!PageError(page)) SetPageUptodate(page); @@ -2310,7 +2305,7 @@ static int __extent_writepage(struct page *page, struct writeback_control *wbc, } em = epd->get_extent(inode, page, pg_offset, cur, end - cur + 1, 1); - if (IS_ERR(em) || !em) { + if (IS_ERR_OR_NULL(em)) { SetPageError(page); break; } @@ -2650,7 +2645,7 @@ int extent_readpages(struct extent_io_tree *tree, prefetchw(&page->flags); list_del(&page->lru); if (!add_to_page_cache_lru(page, mapping, - page->index, GFP_KERNEL)) { + page->index, GFP_NOFS)) { __extent_read_full_page(tree, page, get_extent, &bio, 0, &bio_flags); } @@ -2689,125 +2684,6 @@ int extent_invalidatepage(struct extent_io_tree *tree, } /* - * simple commit_write call, set_range_dirty is used to mark both - * the pages and the extent records as dirty - */ -int extent_commit_write(struct extent_io_tree *tree, - struct inode *inode, struct page *page, - unsigned from, unsigned to) -{ - loff_t pos = ((loff_t)page->index << PAGE_CACHE_SHIFT) + to; - - set_page_extent_mapped(page); - set_page_dirty(page); - - if (pos > inode->i_size) { - i_size_write(inode, pos); - mark_inode_dirty(inode); - } - return 0; -} - -int extent_prepare_write(struct extent_io_tree *tree, - struct inode *inode, struct page *page, - unsigned from, unsigned to, get_extent_t *get_extent) -{ - u64 page_start = (u64)page->index << PAGE_CACHE_SHIFT; - u64 page_end = page_start + PAGE_CACHE_SIZE - 1; - u64 block_start; - u64 orig_block_start; - u64 block_end; - u64 cur_end; - struct extent_map *em; - unsigned blocksize = 1 << inode->i_blkbits; - size_t page_offset = 0; - size_t block_off_start; - size_t block_off_end; - int err = 0; - int iocount = 0; - int ret = 0; - int isnew; - - set_page_extent_mapped(page); - - block_start = (page_start + from) & ~((u64)blocksize - 1); - block_end = (page_start + to - 1) | (blocksize - 1); - orig_block_start = block_start; - - lock_extent(tree, page_start, page_end, GFP_NOFS); - while (block_start <= block_end) { - em = get_extent(inode, page, page_offset, block_start, - block_end - block_start + 1, 1); - if (IS_ERR(em) || !em) - goto err; - - cur_end = min(block_end, extent_map_end(em) - 1); - block_off_start = block_start & (PAGE_CACHE_SIZE - 1); - block_off_end = block_off_start + blocksize; - isnew = clear_extent_new(tree, block_start, cur_end, GFP_NOFS); - - if (!PageUptodate(page) && isnew && - (block_off_end > to || block_off_start < from)) { - void *kaddr; - - kaddr = kmap_atomic(page, KM_USER0); - if (block_off_end > to) - memset(kaddr + to, 0, block_off_end - to); - if (block_off_start < from) - memset(kaddr + block_off_start, 0, - from - block_off_start); - flush_dcache_page(page); - kunmap_atomic(kaddr, KM_USER0); - } - if ((em->block_start != EXTENT_MAP_HOLE && - em->block_start != EXTENT_MAP_INLINE) && - !isnew && !PageUptodate(page) && - (block_off_end > to || block_off_start < from) && - !test_range_bit(tree, block_start, cur_end, - EXTENT_UPTODATE, 1, NULL)) { - u64 sector; - u64 extent_offset = block_start - em->start; - size_t iosize; - sector = (em->block_start + extent_offset) >> 9; - iosize = (cur_end - block_start + blocksize) & - ~((u64)blocksize - 1); - /* - * we've already got the extent locked, but we - * need to split the state such that our end_bio - * handler can clear the lock. - */ - set_extent_bit(tree, block_start, - block_start + iosize - 1, - EXTENT_LOCKED, 0, NULL, NULL, GFP_NOFS); - ret = submit_extent_page(READ, tree, page, - sector, iosize, page_offset, em->bdev, - NULL, 1, - end_bio_extent_preparewrite, 0, - 0, 0); - if (ret && !err) - err = ret; - iocount++; - block_start = block_start + iosize; - } else { - set_extent_uptodate(tree, block_start, cur_end, - GFP_NOFS); - unlock_extent(tree, block_start, cur_end, GFP_NOFS); - block_start = cur_end + 1; - } - page_offset = block_start & (PAGE_CACHE_SIZE - 1); - free_extent_map(em); - } - if (iocount) { - wait_extent_bit(tree, orig_block_start, - block_end, EXTENT_LOCKED); - } - check_page_uptodate(tree, page); -err: - /* FIXME, zero out newly allocated blocks on error */ - return err; -} - -/* * a helper for releasepage, this tests for areas of the page that * are locked or under IO and drops the related state bits if it is safe * to drop the page. @@ -2865,7 +2741,7 @@ int try_release_extent_mapping(struct extent_map_tree *map, len = end - start + 1; write_lock(&map->lock); em = lookup_extent_mapping(map, start, len); - if (!em || IS_ERR(em)) { + if (IS_ERR_OR_NULL(em)) { write_unlock(&map->lock); break; } @@ -2893,33 +2769,6 @@ int try_release_extent_mapping(struct extent_map_tree *map, return try_release_extent_state(map, tree, page, mask); } -sector_t extent_bmap(struct address_space *mapping, sector_t iblock, - get_extent_t *get_extent) -{ - struct inode *inode = mapping->host; - struct extent_state *cached_state = NULL; - u64 start = iblock << inode->i_blkbits; - sector_t sector = 0; - size_t blksize = (1 << inode->i_blkbits); - struct extent_map *em; - - lock_extent_bits(&BTRFS_I(inode)->io_tree, start, start + blksize - 1, - 0, &cached_state, GFP_NOFS); - em = get_extent(inode, NULL, 0, start, blksize, 0); - unlock_extent_cached(&BTRFS_I(inode)->io_tree, start, - start + blksize - 1, &cached_state, GFP_NOFS); - if (!em || IS_ERR(em)) - return 0; - - if (em->block_start > EXTENT_MAP_LAST_BYTE) - goto out; - - sector = (em->block_start + start - em->start) >> inode->i_blkbits; -out: - free_extent_map(em); - return sector; -} - /* * helper function for fiemap, which doesn't want to see any holes. * This maps until we find something past 'last' @@ -2942,7 +2791,7 @@ static struct extent_map *get_extent_skip_holes(struct inode *inode, break; len = (len + sectorsize - 1) & ~(sectorsize - 1); em = get_extent(inode, NULL, 0, offset, len, 0); - if (!em || IS_ERR(em)) + if (IS_ERR_OR_NULL(em)) return em; /* if this isn't a hole return it */ @@ -2996,7 +2845,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, * because there might be preallocation past i_size */ ret = btrfs_lookup_file_extent(NULL, BTRFS_I(inode)->root, - path, inode->i_ino, -1, 0); + path, btrfs_ino(inode), -1, 0); if (ret < 0) { btrfs_free_path(path); return ret; @@ -3009,7 +2858,7 @@ int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, found_type = btrfs_key_type(&found_key); /* No extents, but there might be delalloc bits */ - if (found_key.objectid != inode->i_ino || + if (found_key.objectid != btrfs_ino(inode) || found_type != BTRFS_EXTENT_DATA_KEY) { /* have to trust i_size as the end */ last = (u64)-1; @@ -3232,8 +3081,7 @@ static inline void btrfs_release_extent_buffer(struct extent_buffer *eb) struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree, u64 start, unsigned long len, - struct page *page0, - gfp_t mask) + struct page *page0) { unsigned long num_pages = num_extent_pages(start, len); unsigned long i; @@ -3254,7 +3102,7 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree, } rcu_read_unlock(); - eb = __alloc_extent_buffer(tree, start, len, mask); + eb = __alloc_extent_buffer(tree, start, len, GFP_NOFS); if (!eb) return NULL; @@ -3271,7 +3119,7 @@ struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree, i = 0; } for (; i < num_pages; i++, index++) { - p = find_or_create_page(mapping, index, mask | __GFP_HIGHMEM); + p = find_or_create_page(mapping, index, GFP_NOFS | __GFP_HIGHMEM); if (!p) { WARN_ON(1); goto free_eb; @@ -3343,8 +3191,7 @@ free_eb: } struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree, - u64 start, unsigned long len, - gfp_t mask) + u64 start, unsigned long len) { struct extent_buffer *eb; @@ -3405,13 +3252,6 @@ int clear_extent_buffer_dirty(struct extent_io_tree *tree, return 0; } -int wait_on_extent_buffer_writeback(struct extent_io_tree *tree, - struct extent_buffer *eb) -{ - return wait_on_extent_writeback(tree, eb->start, - eb->start + eb->len - 1); -} - int set_extent_buffer_dirty(struct extent_io_tree *tree, struct extent_buffer *eb) { @@ -3457,7 +3297,7 @@ int set_extent_buffer_uptodate(struct extent_io_tree *tree, num_pages = num_extent_pages(eb->start, eb->len); set_extent_uptodate(tree, eb->start, eb->start + eb->len - 1, - GFP_NOFS); + NULL, GFP_NOFS); for (i = 0; i < num_pages; i++) { page = extent_buffer_page(eb, i); if ((i == 0 && (eb->start & (PAGE_CACHE_SIZE - 1))) || @@ -3885,6 +3725,12 @@ static void move_pages(struct page *dst_page, struct page *src_page, kunmap_atomic(dst_kaddr, KM_USER0); } +static inline bool areas_overlap(unsigned long src, unsigned long dst, unsigned long len) +{ + unsigned long distance = (src > dst) ? src - dst : dst - src; + return distance < len; +} + static void copy_pages(struct page *dst_page, struct page *src_page, unsigned long dst_off, unsigned long src_off, unsigned long len) @@ -3892,10 +3738,12 @@ static void copy_pages(struct page *dst_page, struct page *src_page, char *dst_kaddr = kmap_atomic(dst_page, KM_USER0); char *src_kaddr; - if (dst_page != src_page) + if (dst_page != src_page) { src_kaddr = kmap_atomic(src_page, KM_USER1); - else + } else { src_kaddr = dst_kaddr; + BUG_ON(areas_overlap(src_off, dst_off, len)); + } memcpy(dst_kaddr + dst_off, src_kaddr + src_off, len); kunmap_atomic(dst_kaddr, KM_USER0); @@ -3970,7 +3818,7 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset, "len %lu len %lu\n", dst_offset, len, dst->len); BUG_ON(1); } - if (dst_offset < src_offset) { + if (!areas_overlap(src_offset, dst_offset, len)) { memcpy_extent_buffer(dst, dst_offset, src_offset, len); return; } diff --git a/fs/btrfs/extent_io.h b/fs/btrfs/extent_io.h index f62c5442835d..4e8445a4757c 100644 --- a/fs/btrfs/extent_io.h +++ b/fs/btrfs/extent_io.h @@ -153,23 +153,14 @@ static inline int extent_compress_type(unsigned long bio_flags) struct extent_map_tree; -static inline struct extent_state *extent_state_next(struct extent_state *state) -{ - struct rb_node *node; - node = rb_next(&state->rb_node); - if (!node) - return NULL; - return rb_entry(node, struct extent_state, rb_node); -} - typedef struct extent_map *(get_extent_t)(struct inode *inode, struct page *page, - size_t page_offset, + size_t pg_offset, u64 start, u64 len, int create); void extent_io_tree_init(struct extent_io_tree *tree, - struct address_space *mapping, gfp_t mask); + struct address_space *mapping); int try_release_extent_mapping(struct extent_map_tree *map, struct extent_io_tree *tree, struct page *page, gfp_t mask); @@ -208,21 +199,15 @@ int set_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits, int exclusive_bits, u64 *failed_start, struct extent_state **cached_state, gfp_t mask); int set_extent_uptodate(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask); + struct extent_state **cached_state, gfp_t mask); int set_extent_new(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask); int set_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask); int clear_extent_dirty(struct extent_io_tree *tree, u64 start, u64 end, gfp_t mask); -int clear_extent_ordered(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask); -int clear_extent_ordered_metadata(struct extent_io_tree *tree, u64 start, - u64 end, gfp_t mask); int set_extent_delalloc(struct extent_io_tree *tree, u64 start, u64 end, struct extent_state **cached_state, gfp_t mask); -int set_extent_ordered(struct extent_io_tree *tree, u64 start, u64 end, - gfp_t mask); int find_first_extent_bit(struct extent_io_tree *tree, u64 start, u64 *start_ret, u64 *end_ret, int bits); struct extent_state *find_first_extent_bit_state(struct extent_io_tree *tree, @@ -243,28 +228,17 @@ int extent_readpages(struct extent_io_tree *tree, struct address_space *mapping, struct list_head *pages, unsigned nr_pages, get_extent_t get_extent); -int extent_prepare_write(struct extent_io_tree *tree, - struct inode *inode, struct page *page, - unsigned from, unsigned to, get_extent_t *get_extent); -int extent_commit_write(struct extent_io_tree *tree, - struct inode *inode, struct page *page, - unsigned from, unsigned to); -sector_t extent_bmap(struct address_space *mapping, sector_t iblock, - get_extent_t *get_extent); int extent_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len, get_extent_t *get_extent); -int set_range_dirty(struct extent_io_tree *tree, u64 start, u64 end); int set_state_private(struct extent_io_tree *tree, u64 start, u64 private); int get_state_private(struct extent_io_tree *tree, u64 start, u64 *private); void set_page_extent_mapped(struct page *page); struct extent_buffer *alloc_extent_buffer(struct extent_io_tree *tree, u64 start, unsigned long len, - struct page *page0, - gfp_t mask); + struct page *page0); struct extent_buffer *find_extent_buffer(struct extent_io_tree *tree, - u64 start, unsigned long len, - gfp_t mask); + u64 start, unsigned long len); void free_extent_buffer(struct extent_buffer *eb); int read_extent_buffer_pages(struct extent_io_tree *tree, struct extent_buffer *eb, u64 start, int wait, @@ -292,16 +266,11 @@ void memmove_extent_buffer(struct extent_buffer *dst, unsigned long dst_offset, unsigned long src_offset, unsigned long len); void memset_extent_buffer(struct extent_buffer *eb, char c, unsigned long start, unsigned long len); -int wait_on_extent_buffer_writeback(struct extent_io_tree *tree, - struct extent_buffer *eb); -int wait_on_extent_writeback(struct extent_io_tree *tree, u64 start, u64 end); int wait_extent_bit(struct extent_io_tree *tree, u64 start, u64 end, int bits); int clear_extent_buffer_dirty(struct extent_io_tree *tree, struct extent_buffer *eb); int set_extent_buffer_dirty(struct extent_io_tree *tree, struct extent_buffer *eb); -int test_extent_buffer_dirty(struct extent_io_tree *tree, - struct extent_buffer *eb); int set_extent_buffer_uptodate(struct extent_io_tree *tree, struct extent_buffer *eb); int clear_extent_buffer_uptodate(struct extent_io_tree *tree, @@ -319,7 +288,6 @@ int map_private_extent_buffer(struct extent_buffer *eb, unsigned long offset, unsigned long *map_start, unsigned long *map_len, int km); void unmap_extent_buffer(struct extent_buffer *eb, char *token, int km); -int release_extent_buffer_tail_pages(struct extent_buffer *eb); int extent_range_uptodate(struct extent_io_tree *tree, u64 start, u64 end); int extent_clear_unlock_delalloc(struct inode *inode, diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index 2b6c12e983b3..2d0410344ea3 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -28,12 +28,11 @@ void extent_map_exit(void) /** * extent_map_tree_init - initialize extent map tree * @tree: tree to initialize - * @mask: flags for memory allocations during tree operations * * Initialize the extent tree @tree. Should be called for each new inode * or other user of the extent_map interface. */ -void extent_map_tree_init(struct extent_map_tree *tree, gfp_t mask) +void extent_map_tree_init(struct extent_map_tree *tree) { tree->map = RB_ROOT; rwlock_init(&tree->lock); @@ -41,16 +40,15 @@ void extent_map_tree_init(struct extent_map_tree *tree, gfp_t mask) /** * alloc_extent_map - allocate new extent map structure - * @mask: memory allocation flags * * Allocate a new extent_map structure. The new structure is * returned with a reference count of one and needs to be * freed using free_extent_map() */ -struct extent_map *alloc_extent_map(gfp_t mask) +struct extent_map *alloc_extent_map(void) { struct extent_map *em; - em = kmem_cache_alloc(extent_map_cache, mask); + em = kmem_cache_alloc(extent_map_cache, GFP_NOFS); if (!em) return NULL; em->in_tree = 0; @@ -243,7 +241,7 @@ out: * Insert @em into @tree or perform a simple forward/backward merge with * existing mappings. The extent_map struct passed in will be inserted * into the tree directly, with an additional reference taken, or a - * reference dropped if the merge attempt was successfull. + * reference dropped if the merge attempt was successful. */ int add_extent_mapping(struct extent_map_tree *tree, struct extent_map *em) diff --git a/fs/btrfs/extent_map.h b/fs/btrfs/extent_map.h index 28b44dbd1e35..33a7890b1f40 100644 --- a/fs/btrfs/extent_map.h +++ b/fs/btrfs/extent_map.h @@ -49,14 +49,14 @@ static inline u64 extent_map_block_end(struct extent_map *em) return em->block_start + em->block_len; } -void extent_map_tree_init(struct extent_map_tree *tree, gfp_t mask); +void extent_map_tree_init(struct extent_map_tree *tree); struct extent_map *lookup_extent_mapping(struct extent_map_tree *tree, u64 start, u64 len); int add_extent_mapping(struct extent_map_tree *tree, struct extent_map *em); int remove_extent_mapping(struct extent_map_tree *tree, struct extent_map *em); -struct extent_map *alloc_extent_map(gfp_t mask); +struct extent_map *alloc_extent_map(void); void free_extent_map(struct extent_map *em); int __init extent_map_init(void); void extent_map_exit(void); diff --git a/fs/btrfs/file-item.c b/fs/btrfs/file-item.c index a6a9d4e8b491..90d4ee52cd45 100644 --- a/fs/btrfs/file-item.c +++ b/fs/btrfs/file-item.c @@ -193,7 +193,7 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root, u32 item_size; if (item) - btrfs_release_path(root, path); + btrfs_release_path(path); item = btrfs_lookup_csum(NULL, root->fs_info->csum_root, path, disk_bytenr, 0); if (IS_ERR(item)) { @@ -208,12 +208,13 @@ static int __btrfs_lookup_bio_sums(struct btrfs_root *root, EXTENT_NODATASUM, GFP_NOFS); } else { printk(KERN_INFO "btrfs no csum found " - "for inode %lu start %llu\n", - inode->i_ino, + "for inode %llu start %llu\n", + (unsigned long long) + btrfs_ino(inode), (unsigned long long)offset); } item = NULL; - btrfs_release_path(root, path); + btrfs_release_path(path); goto found; } btrfs_item_key_to_cpu(path->nodes[0], &found_key, @@ -266,7 +267,7 @@ int btrfs_lookup_bio_sums_dio(struct btrfs_root *root, struct inode *inode, } int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end, - struct list_head *list) + struct list_head *list, int search_commit) { struct btrfs_key key; struct btrfs_path *path; @@ -283,6 +284,12 @@ int btrfs_lookup_csums_range(struct btrfs_root *root, u64 start, u64 end, path = btrfs_alloc_path(); BUG_ON(!path); + if (search_commit) { + path->skip_locking = 1; + path->reada = 2; + path->search_commit_root = 1; + } + key.objectid = BTRFS_EXTENT_CSUM_OBJECTID; key.offset = start; key.type = BTRFS_EXTENT_CSUM_KEY; @@ -495,7 +502,6 @@ static noinline int truncate_one_csum(struct btrfs_trans_handle *trans, u32 new_size = (bytenr - key->offset) >> blocksize_bits; new_size *= csum_size; ret = btrfs_truncate_item(trans, root, path, new_size, 1); - BUG_ON(ret); } else if (key->offset >= bytenr && csum_end > end_byte && end_byte > key->offset) { /* @@ -508,7 +514,6 @@ static noinline int truncate_one_csum(struct btrfs_trans_handle *trans, new_size *= csum_size; ret = btrfs_truncate_item(trans, root, path, new_size, 0); - BUG_ON(ret); key->offset = end_byte; ret = btrfs_set_item_key_safe(trans, root, path, key); @@ -551,10 +556,10 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans, ret = btrfs_search_slot(trans, root, &key, path, -1, 1); if (ret > 0) { if (path->slots[0] == 0) - goto out; + break; path->slots[0]--; } else if (ret < 0) { - goto out; + break; } leaf = path->nodes[0]; @@ -579,7 +584,8 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans, /* delete the entire item, it is inside our range */ if (key.offset >= bytenr && csum_end <= end_byte) { ret = btrfs_del_item(trans, root, path); - BUG_ON(ret); + if (ret) + goto out; if (key.offset == bytenr) break; } else if (key.offset < bytenr && csum_end > end_byte) { @@ -631,11 +637,12 @@ int btrfs_del_csums(struct btrfs_trans_handle *trans, if (key.offset < bytenr) break; } - btrfs_release_path(root, path); + btrfs_release_path(path); } + ret = 0; out: btrfs_free_path(path); - return 0; + return ret; } int btrfs_csum_file_blocks(struct btrfs_trans_handle *trans, @@ -722,7 +729,7 @@ again: * at this point, we know the tree has an item, but it isn't big * enough yet to put our csum in. Grow it */ - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_search_slot(trans, root, &file_key, path, csum_size, 1); if (ret < 0) @@ -761,12 +768,11 @@ again: goto insert; ret = btrfs_extend_item(trans, root, path, diff); - BUG_ON(ret); goto csum; } insert: - btrfs_release_path(root, path); + btrfs_release_path(path); csum_offset = 0; if (found_next) { u64 tmp = total_bytes + root->sectorsize; @@ -850,7 +856,7 @@ next_sector: } btrfs_mark_buffer_dirty(path->nodes[0]); if (total_bytes < sums->len) { - btrfs_release_path(root, path); + btrfs_release_path(path); cond_resched(); goto again; } diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index e621ea54a3fd..c6a22d783c35 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -40,6 +40,263 @@ #include "locking.h" #include "compat.h" +/* + * when auto defrag is enabled we + * queue up these defrag structs to remember which + * inodes need defragging passes + */ +struct inode_defrag { + struct rb_node rb_node; + /* objectid */ + u64 ino; + /* + * transid where the defrag was added, we search for + * extents newer than this + */ + u64 transid; + + /* root objectid */ + u64 root; + + /* last offset we were able to defrag */ + u64 last_offset; + + /* if we've wrapped around back to zero once already */ + int cycled; +}; + +/* pop a record for an inode into the defrag tree. The lock + * must be held already + * + * If you're inserting a record for an older transid than an + * existing record, the transid already in the tree is lowered + * + * If an existing record is found the defrag item you + * pass in is freed + */ +static int __btrfs_add_inode_defrag(struct inode *inode, + struct inode_defrag *defrag) +{ + struct btrfs_root *root = BTRFS_I(inode)->root; + struct inode_defrag *entry; + struct rb_node **p; + struct rb_node *parent = NULL; + + p = &root->fs_info->defrag_inodes.rb_node; + while (*p) { + parent = *p; + entry = rb_entry(parent, struct inode_defrag, rb_node); + + if (defrag->ino < entry->ino) + p = &parent->rb_left; + else if (defrag->ino > entry->ino) + p = &parent->rb_right; + else { + /* if we're reinserting an entry for + * an old defrag run, make sure to + * lower the transid of our existing record + */ + if (defrag->transid < entry->transid) + entry->transid = defrag->transid; + if (defrag->last_offset > entry->last_offset) + entry->last_offset = defrag->last_offset; + goto exists; + } + } + BTRFS_I(inode)->in_defrag = 1; + rb_link_node(&defrag->rb_node, parent, p); + rb_insert_color(&defrag->rb_node, &root->fs_info->defrag_inodes); + return 0; + +exists: + kfree(defrag); + return 0; + +} + +/* + * insert a defrag record for this inode if auto defrag is + * enabled + */ +int btrfs_add_inode_defrag(struct btrfs_trans_handle *trans, + struct inode *inode) +{ + struct btrfs_root *root = BTRFS_I(inode)->root; + struct inode_defrag *defrag; + int ret = 0; + u64 transid; + + if (!btrfs_test_opt(root, AUTO_DEFRAG)) + return 0; + + if (root->fs_info->closing) + return 0; + + if (BTRFS_I(inode)->in_defrag) + return 0; + + if (trans) + transid = trans->transid; + else + transid = BTRFS_I(inode)->root->last_trans; + + defrag = kzalloc(sizeof(*defrag), GFP_NOFS); + if (!defrag) + return -ENOMEM; + + defrag->ino = inode->i_ino; + defrag->transid = transid; + defrag->root = root->root_key.objectid; + + spin_lock(&root->fs_info->defrag_inodes_lock); + if (!BTRFS_I(inode)->in_defrag) + ret = __btrfs_add_inode_defrag(inode, defrag); + spin_unlock(&root->fs_info->defrag_inodes_lock); + return ret; +} + +/* + * must be called with the defrag_inodes lock held + */ +struct inode_defrag *btrfs_find_defrag_inode(struct btrfs_fs_info *info, u64 ino, + struct rb_node **next) +{ + struct inode_defrag *entry = NULL; + struct rb_node *p; + struct rb_node *parent = NULL; + + p = info->defrag_inodes.rb_node; + while (p) { + parent = p; + entry = rb_entry(parent, struct inode_defrag, rb_node); + + if (ino < entry->ino) + p = parent->rb_left; + else if (ino > entry->ino) + p = parent->rb_right; + else + return entry; + } + + if (next) { + while (parent && ino > entry->ino) { + parent = rb_next(parent); + entry = rb_entry(parent, struct inode_defrag, rb_node); + } + *next = parent; + } + return NULL; +} + +/* + * run through the list of inodes in the FS that need + * defragging + */ +int btrfs_run_defrag_inodes(struct btrfs_fs_info *fs_info) +{ + struct inode_defrag *defrag; + struct btrfs_root *inode_root; + struct inode *inode; + struct rb_node *n; + struct btrfs_key key; + struct btrfs_ioctl_defrag_range_args range; + u64 first_ino = 0; + int num_defrag; + int defrag_batch = 1024; + + memset(&range, 0, sizeof(range)); + range.len = (u64)-1; + + atomic_inc(&fs_info->defrag_running); + spin_lock(&fs_info->defrag_inodes_lock); + while(1) { + n = NULL; + + /* find an inode to defrag */ + defrag = btrfs_find_defrag_inode(fs_info, first_ino, &n); + if (!defrag) { + if (n) + defrag = rb_entry(n, struct inode_defrag, rb_node); + else if (first_ino) { + first_ino = 0; + continue; + } else { + break; + } + } + + /* remove it from the rbtree */ + first_ino = defrag->ino + 1; + rb_erase(&defrag->rb_node, &fs_info->defrag_inodes); + + if (fs_info->closing) + goto next_free; + + spin_unlock(&fs_info->defrag_inodes_lock); + + /* get the inode */ + key.objectid = defrag->root; + btrfs_set_key_type(&key, BTRFS_ROOT_ITEM_KEY); + key.offset = (u64)-1; + inode_root = btrfs_read_fs_root_no_name(fs_info, &key); + if (IS_ERR(inode_root)) + goto next; + + key.objectid = defrag->ino; + btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY); + key.offset = 0; + + inode = btrfs_iget(fs_info->sb, &key, inode_root, NULL); + if (IS_ERR(inode)) + goto next; + + /* do a chunk of defrag */ + BTRFS_I(inode)->in_defrag = 0; + range.start = defrag->last_offset; + num_defrag = btrfs_defrag_file(inode, NULL, &range, defrag->transid, + defrag_batch); + /* + * if we filled the whole defrag batch, there + * must be more work to do. Queue this defrag + * again + */ + if (num_defrag == defrag_batch) { + defrag->last_offset = range.start; + __btrfs_add_inode_defrag(inode, defrag); + /* + * we don't want to kfree defrag, we added it back to + * the rbtree + */ + defrag = NULL; + } else if (defrag->last_offset && !defrag->cycled) { + /* + * we didn't fill our defrag batch, but + * we didn't start at zero. Make sure we loop + * around to the start of the file. + */ + defrag->last_offset = 0; + defrag->cycled = 1; + __btrfs_add_inode_defrag(inode, defrag); + defrag = NULL; + } + + iput(inode); +next: + spin_lock(&fs_info->defrag_inodes_lock); +next_free: + kfree(defrag); + } + spin_unlock(&fs_info->defrag_inodes_lock); + + atomic_dec(&fs_info->defrag_running); + + /* + * during unmount, we use the transaction_wait queue to + * wait for the defragger to stop + */ + wake_up(&fs_info->transaction_wait); + return 0; +} /* simple helper to fault in pages and copy. This should go away * and be replaced with calls into generic code. @@ -104,7 +361,7 @@ static noinline int btrfs_copy_from_user(loff_t pos, int num_pages, /* * unlocks pages after btrfs_file_write is done with them */ -static noinline void btrfs_drop_pages(struct page **pages, size_t num_pages) +void btrfs_drop_pages(struct page **pages, size_t num_pages) { size_t i; for (i = 0; i < num_pages; i++) { @@ -127,16 +384,13 @@ static noinline void btrfs_drop_pages(struct page **pages, size_t num_pages) * this also makes the decision about creating an inline extent vs * doing real data extents, marking pages dirty and delalloc as required. */ -static noinline int dirty_and_release_pages(struct btrfs_root *root, - struct file *file, - struct page **pages, - size_t num_pages, - loff_t pos, - size_t write_bytes) +int btrfs_dirty_pages(struct btrfs_root *root, struct inode *inode, + struct page **pages, size_t num_pages, + loff_t pos, size_t write_bytes, + struct extent_state **cached) { int err = 0; int i; - struct inode *inode = fdentry(file)->d_inode; u64 num_bytes; u64 start_pos; u64 end_of_last_block; @@ -149,7 +403,7 @@ static noinline int dirty_and_release_pages(struct btrfs_root *root, end_of_last_block = start_pos + num_bytes - 1; err = btrfs_set_extent_delalloc(inode, start_pos, end_of_last_block, - NULL); + cached); if (err) return err; @@ -194,9 +448,9 @@ int btrfs_drop_extent_cache(struct inode *inode, u64 start, u64 end, } while (1) { if (!split) - split = alloc_extent_map(GFP_NOFS); + split = alloc_extent_map(); if (!split2) - split2 = alloc_extent_map(GFP_NOFS); + split2 = alloc_extent_map(); BUG_ON(!split || !split2); write_lock(&em_tree->lock); @@ -301,6 +555,7 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans, struct inode *inode, struct btrfs_path *path; struct btrfs_key key; struct btrfs_key new_key; + u64 ino = btrfs_ino(inode); u64 search_start = start; u64 disk_bytenr = 0; u64 num_bytes = 0; @@ -321,14 +576,14 @@ int btrfs_drop_extents(struct btrfs_trans_handle *trans, struct inode *inode, while (1) { recow = 0; - ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino, + ret = btrfs_lookup_file_extent(trans, root, path, ino, search_start, -1); if (ret < 0) break; if (ret > 0 && path->slots[0] > 0 && search_start == start) { leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, path->slots[0] - 1); - if (key.objectid == inode->i_ino && + if (key.objectid == ino && key.type == BTRFS_EXTENT_DATA_KEY) path->slots[0]--; } @@ -349,7 +604,7 @@ next_slot: } btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - if (key.objectid > inode->i_ino || + if (key.objectid > ino || key.type > BTRFS_EXTENT_DATA_KEY || key.offset >= end) break; @@ -379,7 +634,7 @@ next_slot: search_start = max(key.offset, start); if (recow) { - btrfs_release_path(root, path); + btrfs_release_path(path); continue; } @@ -396,7 +651,7 @@ next_slot: ret = btrfs_duplicate_item(trans, root, path, &new_key); if (ret == -EAGAIN) { - btrfs_release_path(root, path); + btrfs_release_path(path); continue; } if (ret < 0) @@ -519,7 +774,7 @@ next_slot: del_nr = 0; del_slot = 0; - btrfs_release_path(root, path); + btrfs_release_path(path); continue; } @@ -595,6 +850,7 @@ int btrfs_mark_extent_written(struct btrfs_trans_handle *trans, int del_slot = 0; int recow; int ret; + u64 ino = btrfs_ino(inode); btrfs_drop_extent_cache(inode, start, end - 1, 0); @@ -603,7 +859,7 @@ int btrfs_mark_extent_written(struct btrfs_trans_handle *trans, again: recow = 0; split = start; - key.objectid = inode->i_ino; + key.objectid = ino; key.type = BTRFS_EXTENT_DATA_KEY; key.offset = split; @@ -615,8 +871,7 @@ again: leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - BUG_ON(key.objectid != inode->i_ino || - key.type != BTRFS_EXTENT_DATA_KEY); + BUG_ON(key.objectid != ino || key.type != BTRFS_EXTENT_DATA_KEY); fi = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item); BUG_ON(btrfs_file_extent_type(leaf, fi) != @@ -633,7 +888,7 @@ again: other_start = 0; other_end = start; if (extent_mergeable(leaf, path->slots[0] - 1, - inode->i_ino, bytenr, orig_offset, + ino, bytenr, orig_offset, &other_start, &other_end)) { new_key.offset = end; btrfs_set_item_key_safe(trans, root, path, &new_key); @@ -656,7 +911,7 @@ again: other_start = end; other_end = 0; if (extent_mergeable(leaf, path->slots[0] + 1, - inode->i_ino, bytenr, orig_offset, + ino, bytenr, orig_offset, &other_start, &other_end)) { fi = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_file_extent_item); @@ -684,7 +939,7 @@ again: new_key.offset = split; ret = btrfs_duplicate_item(trans, root, path, &new_key); if (ret == -EAGAIN) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto again; } BUG_ON(ret < 0); @@ -705,7 +960,7 @@ again: ret = btrfs_inc_extent_ref(trans, root, bytenr, num_bytes, 0, root->root_key.objectid, - inode->i_ino, orig_offset); + ino, orig_offset); BUG_ON(ret); if (split == start) { @@ -721,10 +976,10 @@ again: other_start = end; other_end = 0; if (extent_mergeable(leaf, path->slots[0] + 1, - inode->i_ino, bytenr, orig_offset, + ino, bytenr, orig_offset, &other_start, &other_end)) { if (recow) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto again; } extent_end = other_end; @@ -732,16 +987,16 @@ again: del_nr++; ret = btrfs_free_extent(trans, root, bytenr, num_bytes, 0, root->root_key.objectid, - inode->i_ino, orig_offset); + ino, orig_offset); BUG_ON(ret); } other_start = 0; other_end = start; if (extent_mergeable(leaf, path->slots[0] - 1, - inode->i_ino, bytenr, orig_offset, + ino, bytenr, orig_offset, &other_start, &other_end)) { if (recow) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto again; } key.offset = other_start; @@ -749,7 +1004,7 @@ again: del_nr++; ret = btrfs_free_extent(trans, root, bytenr, num_bytes, 0, root->root_key.objectid, - inode->i_ino, orig_offset); + ino, orig_offset); BUG_ON(ret); } if (del_nr == 0) { @@ -992,9 +1247,9 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, } if (copied > 0) { - ret = dirty_and_release_pages(root, file, pages, - dirty_pages, pos, - copied); + ret = btrfs_dirty_pages(root, inode, pages, + dirty_pages, pos, copied, + NULL); if (ret) { btrfs_delalloc_release_space(inode, dirty_pages << PAGE_CACHE_SHIFT); @@ -1378,7 +1633,7 @@ static long btrfs_fallocate(struct file *file, int mode, while (1) { em = btrfs_get_extent(inode, NULL, 0, cur_offset, alloc_end - cur_offset, 0); - BUG_ON(IS_ERR(em) || !em); + BUG_ON(IS_ERR_OR_NULL(em)); last_byte = min(extent_map_end(em), alloc_end); last_byte = (last_byte + mask) & ~mask; if (em->block_start == EXTENT_MAP_HOLE || diff --git a/fs/btrfs/free-space-cache.c b/fs/btrfs/free-space-cache.c index f561c953205b..70d45795d758 100644 --- a/fs/btrfs/free-space-cache.c +++ b/fs/btrfs/free-space-cache.c @@ -25,18 +25,17 @@ #include "transaction.h" #include "disk-io.h" #include "extent_io.h" +#include "inode-map.h" #define BITS_PER_BITMAP (PAGE_CACHE_SIZE * 8) #define MAX_CACHE_BYTES_PER_GIG (32 * 1024) -static void recalculate_thresholds(struct btrfs_block_group_cache - *block_group); -static int link_free_space(struct btrfs_block_group_cache *block_group, +static int link_free_space(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info); -struct inode *lookup_free_space_inode(struct btrfs_root *root, - struct btrfs_block_group_cache - *block_group, struct btrfs_path *path) +static struct inode *__lookup_free_space_inode(struct btrfs_root *root, + struct btrfs_path *path, + u64 offset) { struct btrfs_key key; struct btrfs_key location; @@ -46,22 +45,15 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root, struct inode *inode = NULL; int ret; - spin_lock(&block_group->lock); - if (block_group->inode) - inode = igrab(block_group->inode); - spin_unlock(&block_group->lock); - if (inode) - return inode; - key.objectid = BTRFS_FREE_SPACE_OBJECTID; - key.offset = block_group->key.objectid; + key.offset = offset; key.type = 0; ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) return ERR_PTR(ret); if (ret > 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); return ERR_PTR(-ENOENT); } @@ -70,7 +62,7 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root, struct btrfs_free_space_header); btrfs_free_space_key(leaf, header, &disk_key); btrfs_disk_key_to_cpu(&location, &disk_key); - btrfs_release_path(root, path); + btrfs_release_path(path); inode = btrfs_iget(root->fs_info->sb, &location, root, NULL); if (!inode) @@ -84,6 +76,27 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root, inode->i_mapping->flags &= ~__GFP_FS; + return inode; +} + +struct inode *lookup_free_space_inode(struct btrfs_root *root, + struct btrfs_block_group_cache + *block_group, struct btrfs_path *path) +{ + struct inode *inode = NULL; + + spin_lock(&block_group->lock); + if (block_group->inode) + inode = igrab(block_group->inode); + spin_unlock(&block_group->lock); + if (inode) + return inode; + + inode = __lookup_free_space_inode(root, path, + block_group->key.objectid); + if (IS_ERR(inode)) + return inode; + spin_lock(&block_group->lock); if (!root->fs_info->closing) { block_group->inode = igrab(inode); @@ -94,24 +107,18 @@ struct inode *lookup_free_space_inode(struct btrfs_root *root, return inode; } -int create_free_space_inode(struct btrfs_root *root, - struct btrfs_trans_handle *trans, - struct btrfs_block_group_cache *block_group, - struct btrfs_path *path) +int __create_free_space_inode(struct btrfs_root *root, + struct btrfs_trans_handle *trans, + struct btrfs_path *path, u64 ino, u64 offset) { struct btrfs_key key; struct btrfs_disk_key disk_key; struct btrfs_free_space_header *header; struct btrfs_inode_item *inode_item; struct extent_buffer *leaf; - u64 objectid; int ret; - ret = btrfs_find_free_objectid(trans, root, 0, &objectid); - if (ret < 0) - return ret; - - ret = btrfs_insert_empty_inode(trans, root, path, objectid); + ret = btrfs_insert_empty_inode(trans, root, path, ino); if (ret) return ret; @@ -131,19 +138,18 @@ int create_free_space_inode(struct btrfs_root *root, BTRFS_INODE_PREALLOC | BTRFS_INODE_NODATASUM); btrfs_set_inode_nlink(leaf, inode_item, 1); btrfs_set_inode_transid(leaf, inode_item, trans->transid); - btrfs_set_inode_block_group(leaf, inode_item, - block_group->key.objectid); + btrfs_set_inode_block_group(leaf, inode_item, offset); btrfs_mark_buffer_dirty(leaf); - btrfs_release_path(root, path); + btrfs_release_path(path); key.objectid = BTRFS_FREE_SPACE_OBJECTID; - key.offset = block_group->key.objectid; + key.offset = offset; key.type = 0; ret = btrfs_insert_empty_item(trans, root, path, &key, sizeof(struct btrfs_free_space_header)); if (ret < 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); return ret; } leaf = path->nodes[0]; @@ -152,11 +158,27 @@ int create_free_space_inode(struct btrfs_root *root, memset_extent_buffer(leaf, 0, (unsigned long)header, sizeof(*header)); btrfs_set_free_space_key(leaf, header, &disk_key); btrfs_mark_buffer_dirty(leaf); - btrfs_release_path(root, path); + btrfs_release_path(path); return 0; } +int create_free_space_inode(struct btrfs_root *root, + struct btrfs_trans_handle *trans, + struct btrfs_block_group_cache *block_group, + struct btrfs_path *path) +{ + int ret; + u64 ino; + + ret = btrfs_find_free_objectid(root, &ino); + if (ret < 0) + return ret; + + return __create_free_space_inode(root, trans, path, ino, + block_group->key.objectid); +} + int btrfs_truncate_free_space_cache(struct btrfs_root *root, struct btrfs_trans_handle *trans, struct btrfs_path *path, @@ -187,7 +209,8 @@ int btrfs_truncate_free_space_cache(struct btrfs_root *root, return ret; } - return btrfs_update_inode(trans, root, inode); + ret = btrfs_update_inode(trans, root, inode); + return ret; } static int readahead_cache(struct inode *inode) @@ -209,15 +232,13 @@ static int readahead_cache(struct inode *inode) return 0; } -int load_free_space_cache(struct btrfs_fs_info *fs_info, - struct btrfs_block_group_cache *block_group) +int __load_free_space_cache(struct btrfs_root *root, struct inode *inode, + struct btrfs_free_space_ctl *ctl, + struct btrfs_path *path, u64 offset) { - struct btrfs_root *root = fs_info->tree_root; - struct inode *inode; struct btrfs_free_space_header *header; struct extent_buffer *leaf; struct page *page; - struct btrfs_path *path; u32 *checksums = NULL, *crc; char *disk_crcs = NULL; struct btrfs_key key; @@ -225,76 +246,47 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, u64 num_entries; u64 num_bitmaps; u64 generation; - u64 used = btrfs_block_group_used(&block_group->item); u32 cur_crc = ~(u32)0; pgoff_t index = 0; unsigned long first_page_offset; int num_checksums; - int ret = 0; - - /* - * If we're unmounting then just return, since this does a search on the - * normal root and not the commit root and we could deadlock. - */ - smp_mb(); - if (fs_info->closing) - return 0; - - /* - * If this block group has been marked to be cleared for one reason or - * another then we can't trust the on disk cache, so just return. - */ - spin_lock(&block_group->lock); - if (block_group->disk_cache_state != BTRFS_DC_WRITTEN) { - spin_unlock(&block_group->lock); - return 0; - } - spin_unlock(&block_group->lock); + int ret = 0, ret2; INIT_LIST_HEAD(&bitmaps); - path = btrfs_alloc_path(); - if (!path) - return 0; - - inode = lookup_free_space_inode(root, block_group, path); - if (IS_ERR(inode)) { - btrfs_free_path(path); - return 0; - } - /* Nothing in the space cache, goodbye */ - if (!i_size_read(inode)) { - btrfs_free_path(path); + if (!i_size_read(inode)) goto out; - } key.objectid = BTRFS_FREE_SPACE_OBJECTID; - key.offset = block_group->key.objectid; + key.offset = offset; key.type = 0; ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); - if (ret) { - btrfs_free_path(path); + if (ret < 0) + goto out; + else if (ret > 0) { + btrfs_release_path(path); + ret = 0; goto out; } + ret = -1; + leaf = path->nodes[0]; header = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_free_space_header); num_entries = btrfs_free_space_entries(leaf, header); num_bitmaps = btrfs_free_space_bitmaps(leaf, header); generation = btrfs_free_space_generation(leaf, header); - btrfs_free_path(path); + btrfs_release_path(path); if (BTRFS_I(inode)->generation != generation) { printk(KERN_ERR "btrfs: free space inode generation (%llu) did" - " not match free space cache generation (%llu) for " - "block group %llu\n", + " not match free space cache generation (%llu)\n", (unsigned long long)BTRFS_I(inode)->generation, - (unsigned long long)generation, - (unsigned long long)block_group->key.objectid); - goto free_cache; + (unsigned long long)generation); + goto out; } if (!num_entries) @@ -311,10 +303,8 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, goto out; ret = readahead_cache(inode); - if (ret) { - ret = 0; + if (ret) goto out; - } while (1) { struct btrfs_free_space_entry *entry; @@ -333,10 +323,8 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, } page = grab_cache_page(inode->i_mapping, index); - if (!page) { - ret = 0; + if (!page) goto free_cache; - } if (!PageUptodate(page)) { btrfs_readpage(NULL, page); @@ -345,9 +333,7 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, unlock_page(page); page_cache_release(page); printk(KERN_ERR "btrfs: error reading free " - "space cache: %llu\n", - (unsigned long long) - block_group->key.objectid); + "space cache\n"); goto free_cache; } } @@ -360,13 +346,10 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, gen = addr + (sizeof(u32) * num_checksums); if (*gen != BTRFS_I(inode)->generation) { printk(KERN_ERR "btrfs: space cache generation" - " (%llu) does not match inode (%llu) " - "for block group %llu\n", + " (%llu) does not match inode (%llu)\n", (unsigned long long)*gen, (unsigned long long) - BTRFS_I(inode)->generation, - (unsigned long long) - block_group->key.objectid); + BTRFS_I(inode)->generation); kunmap(page); unlock_page(page); page_cache_release(page); @@ -382,9 +365,8 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, PAGE_CACHE_SIZE - start_offset); btrfs_csum_final(cur_crc, (char *)&cur_crc); if (cur_crc != *crc) { - printk(KERN_ERR "btrfs: crc mismatch for page %lu in " - "block group %llu\n", index, - (unsigned long long)block_group->key.objectid); + printk(KERN_ERR "btrfs: crc mismatch for page %lu\n", + index); kunmap(page); unlock_page(page); page_cache_release(page); @@ -417,9 +399,9 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, } if (entry->type == BTRFS_FREE_SPACE_EXTENT) { - spin_lock(&block_group->tree_lock); - ret = link_free_space(block_group, e); - spin_unlock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); + ret = link_free_space(ctl, e); + spin_unlock(&ctl->tree_lock); BUG_ON(ret); } else { e->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); @@ -431,11 +413,11 @@ int load_free_space_cache(struct btrfs_fs_info *fs_info, page_cache_release(page); goto free_cache; } - spin_lock(&block_group->tree_lock); - ret = link_free_space(block_group, e); - block_group->total_bitmaps++; - recalculate_thresholds(block_group); - spin_unlock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); + ret2 = link_free_space(ctl, e); + ctl->total_bitmaps++; + ctl->op->recalc_thresholds(ctl); + spin_unlock(&ctl->tree_lock); list_add_tail(&e->list, &bitmaps); } @@ -471,43 +453,100 @@ next: index++; } - spin_lock(&block_group->tree_lock); - if (block_group->free_space != (block_group->key.offset - used - - block_group->bytes_super)) { - spin_unlock(&block_group->tree_lock); - printk(KERN_ERR "block group %llu has an wrong amount of free " - "space\n", block_group->key.objectid); - ret = 0; - goto free_cache; - } - spin_unlock(&block_group->tree_lock); - ret = 1; out: kfree(checksums); kfree(disk_crcs); - iput(inode); return ret; - free_cache: - /* This cache is bogus, make sure it gets cleared */ + __btrfs_remove_free_space_cache(ctl); + goto out; +} + +int load_free_space_cache(struct btrfs_fs_info *fs_info, + struct btrfs_block_group_cache *block_group) +{ + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; + struct btrfs_root *root = fs_info->tree_root; + struct inode *inode; + struct btrfs_path *path; + int ret; + bool matched; + u64 used = btrfs_block_group_used(&block_group->item); + + /* + * If we're unmounting then just return, since this does a search on the + * normal root and not the commit root and we could deadlock. + */ + smp_mb(); + if (fs_info->closing) + return 0; + + /* + * If this block group has been marked to be cleared for one reason or + * another then we can't trust the on disk cache, so just return. + */ spin_lock(&block_group->lock); - block_group->disk_cache_state = BTRFS_DC_CLEAR; + if (block_group->disk_cache_state != BTRFS_DC_WRITTEN) { + spin_unlock(&block_group->lock); + return 0; + } spin_unlock(&block_group->lock); - btrfs_remove_free_space_cache(block_group); - goto out; + + path = btrfs_alloc_path(); + if (!path) + return 0; + + inode = lookup_free_space_inode(root, block_group, path); + if (IS_ERR(inode)) { + btrfs_free_path(path); + return 0; + } + + ret = __load_free_space_cache(fs_info->tree_root, inode, ctl, + path, block_group->key.objectid); + btrfs_free_path(path); + if (ret <= 0) + goto out; + + spin_lock(&ctl->tree_lock); + matched = (ctl->free_space == (block_group->key.offset - used - + block_group->bytes_super)); + spin_unlock(&ctl->tree_lock); + + if (!matched) { + __btrfs_remove_free_space_cache(ctl); + printk(KERN_ERR "block group %llu has an wrong amount of free " + "space\n", block_group->key.objectid); + ret = -1; + } +out: + if (ret < 0) { + /* This cache is bogus, make sure it gets cleared */ + spin_lock(&block_group->lock); + block_group->disk_cache_state = BTRFS_DC_CLEAR; + spin_unlock(&block_group->lock); + ret = 0; + + printk(KERN_ERR "btrfs: failed to load free space cache " + "for block group %llu\n", block_group->key.objectid); + } + + iput(inode); + return ret; } -int btrfs_write_out_cache(struct btrfs_root *root, - struct btrfs_trans_handle *trans, - struct btrfs_block_group_cache *block_group, - struct btrfs_path *path) +int __btrfs_write_out_cache(struct btrfs_root *root, struct inode *inode, + struct btrfs_free_space_ctl *ctl, + struct btrfs_block_group_cache *block_group, + struct btrfs_trans_handle *trans, + struct btrfs_path *path, u64 offset) { struct btrfs_free_space_header *header; struct extent_buffer *leaf; - struct inode *inode; struct rb_node *node; struct list_head *pos, *n; + struct page **pages; struct page *page; struct extent_state *cached_state = NULL; struct btrfs_free_cluster *cluster = NULL; @@ -517,61 +556,48 @@ int btrfs_write_out_cache(struct btrfs_root *root, u64 start, end, len; u64 bytes = 0; u32 *crc, *checksums; - pgoff_t index = 0, last_index = 0; unsigned long first_page_offset; - int num_checksums; + int index = 0, num_pages = 0; int entries = 0; int bitmaps = 0; - int ret = 0; + int ret = -1; bool next_page = false; - - root = root->fs_info->tree_root; + bool out_of_space = false; INIT_LIST_HEAD(&bitmap_list); - spin_lock(&block_group->lock); - if (block_group->disk_cache_state < BTRFS_DC_SETUP) { - spin_unlock(&block_group->lock); - return 0; - } - spin_unlock(&block_group->lock); - - inode = lookup_free_space_inode(root, block_group, path); - if (IS_ERR(inode)) + node = rb_first(&ctl->free_space_offset); + if (!node) return 0; - if (!i_size_read(inode)) { - iput(inode); - return 0; - } + if (!i_size_read(inode)) + return -1; - node = rb_first(&block_group->free_space_offset); - if (!node) { - iput(inode); - return 0; - } - - last_index = (i_size_read(inode) - 1) >> PAGE_CACHE_SHIFT; + num_pages = (i_size_read(inode) + PAGE_CACHE_SIZE - 1) >> + PAGE_CACHE_SHIFT; filemap_write_and_wait(inode->i_mapping); btrfs_wait_ordered_range(inode, inode->i_size & ~(root->sectorsize - 1), (u64)-1); /* We need a checksum per page. */ - num_checksums = i_size_read(inode) / PAGE_CACHE_SIZE; - crc = checksums = kzalloc(sizeof(u32) * num_checksums, GFP_NOFS); - if (!crc) { - iput(inode); - return 0; + crc = checksums = kzalloc(sizeof(u32) * num_pages, GFP_NOFS); + if (!crc) + return -1; + + pages = kzalloc(sizeof(struct page *) * num_pages, GFP_NOFS); + if (!pages) { + kfree(crc); + return -1; } /* Since the first page has all of our checksums and our generation we * need to calculate the offset into the page that we can start writing * our entries. */ - first_page_offset = (sizeof(u32) * num_checksums) + sizeof(u64); + first_page_offset = (sizeof(u32) * num_pages) + sizeof(u64); /* Get the cluster for this block_group if it exists */ - if (!list_empty(&block_group->cluster_list)) + if (block_group && !list_empty(&block_group->cluster_list)) cluster = list_entry(block_group->cluster_list.next, struct btrfs_free_cluster, block_group_list); @@ -590,20 +616,18 @@ int btrfs_write_out_cache(struct btrfs_root *root, * after find_get_page at this point. Just putting this here so people * know and don't freak out. */ - while (index <= last_index) { + while (index < num_pages) { page = grab_cache_page(inode->i_mapping, index); if (!page) { - pgoff_t i = 0; + int i; - while (i < index) { - page = find_get_page(inode->i_mapping, i); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); - i++; + for (i = 0; i < num_pages; i++) { + unlock_page(pages[i]); + page_cache_release(pages[i]); } goto out_free; } + pages[index] = page; index++; } @@ -615,7 +639,8 @@ int btrfs_write_out_cache(struct btrfs_root *root, * When searching for pinned extents, we need to start at our start * offset. */ - start = block_group->key.objectid; + if (block_group) + start = block_group->key.objectid; /* Write out the extent entries */ do { @@ -631,7 +656,12 @@ int btrfs_write_out_cache(struct btrfs_root *root, offset = start_offset; } - page = find_get_page(inode->i_mapping, index); + if (index >= num_pages) { + out_of_space = true; + break; + } + + page = pages[index]; addr = kmap(page); entry = addr + start_offset; @@ -668,8 +698,9 @@ int btrfs_write_out_cache(struct btrfs_root *root, * We want to add any pinned extents to our free space cache * so we don't leak the space */ - while (!next_page && (start < block_group->key.objectid + - block_group->key.offset)) { + while (block_group && !next_page && + (start < block_group->key.objectid + + block_group->key.offset)) { ret = find_first_extent_bit(unpin, start, &start, &end, EXTENT_DIRTY); if (ret) { @@ -708,23 +739,6 @@ int btrfs_write_out_cache(struct btrfs_root *root, bytes += PAGE_CACHE_SIZE; - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - - /* - * We need to release our reference we got for grab_cache_page, - * except for the first page which will hold our checksums, we - * do that below. - */ - if (index != 0) { - unlock_page(page); - page_cache_release(page); - } - - page_cache_release(page); - index++; } while (node || next_page); @@ -734,7 +748,11 @@ int btrfs_write_out_cache(struct btrfs_root *root, struct btrfs_free_space *entry = list_entry(pos, struct btrfs_free_space, list); - page = find_get_page(inode->i_mapping, index); + if (index >= num_pages) { + out_of_space = true; + break; + } + page = pages[index]; addr = kmap(page); memcpy(addr, entry->bitmap, PAGE_CACHE_SIZE); @@ -745,73 +763,67 @@ int btrfs_write_out_cache(struct btrfs_root *root, crc++; bytes += PAGE_CACHE_SIZE; - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); list_del_init(&entry->list); index++; } + if (out_of_space) { + btrfs_drop_pages(pages, num_pages); + unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, + i_size_read(inode) - 1, &cached_state, + GFP_NOFS); + ret = 0; + goto out_free; + } + /* Zero out the rest of the pages just to make sure */ - while (index <= last_index) { + while (index < num_pages) { void *addr; - page = find_get_page(inode->i_mapping, index); - + page = pages[index]; addr = kmap(page); memset(addr, 0, PAGE_CACHE_SIZE); kunmap(page); - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); bytes += PAGE_CACHE_SIZE; index++; } - btrfs_set_extent_delalloc(inode, 0, bytes - 1, &cached_state); - /* Write the checksums and trans id to the first page */ { void *addr; u64 *gen; - page = find_get_page(inode->i_mapping, 0); + page = pages[0]; addr = kmap(page); - memcpy(addr, checksums, sizeof(u32) * num_checksums); - gen = addr + (sizeof(u32) * num_checksums); + memcpy(addr, checksums, sizeof(u32) * num_pages); + gen = addr + (sizeof(u32) * num_pages); *gen = trans->transid; kunmap(page); - ClearPageChecked(page); - set_page_extent_mapped(page); - SetPageUptodate(page); - set_page_dirty(page); - unlock_page(page); - page_cache_release(page); - page_cache_release(page); } - BTRFS_I(inode)->generation = trans->transid; + ret = btrfs_dirty_pages(root, inode, pages, num_pages, 0, + bytes, &cached_state); + btrfs_drop_pages(pages, num_pages); unlock_extent_cached(&BTRFS_I(inode)->io_tree, 0, i_size_read(inode) - 1, &cached_state, GFP_NOFS); + if (ret) { + ret = 0; + goto out_free; + } + + BTRFS_I(inode)->generation = trans->transid; + filemap_write_and_wait(inode->i_mapping); key.objectid = BTRFS_FREE_SPACE_OBJECTID; - key.offset = block_group->key.objectid; + key.offset = offset; key.type = 0; ret = btrfs_search_slot(trans, root, &key, path, 1, 1); if (ret < 0) { - ret = 0; + ret = -1; clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, bytes - 1, EXTENT_DIRTY | EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING, 0, 0, NULL, GFP_NOFS); @@ -824,13 +836,13 @@ int btrfs_write_out_cache(struct btrfs_root *root, path->slots[0]--; btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); if (found_key.objectid != BTRFS_FREE_SPACE_OBJECTID || - found_key.offset != block_group->key.objectid) { - ret = 0; + found_key.offset != offset) { + ret = -1; clear_extent_bit(&BTRFS_I(inode)->io_tree, 0, bytes - 1, EXTENT_DIRTY | EXTENT_DELALLOC | EXTENT_DO_ACCOUNTING, 0, 0, NULL, GFP_NOFS); - btrfs_release_path(root, path); + btrfs_release_path(path); goto out_free; } } @@ -840,48 +852,83 @@ int btrfs_write_out_cache(struct btrfs_root *root, btrfs_set_free_space_bitmaps(leaf, header, bitmaps); btrfs_set_free_space_generation(leaf, header, trans->transid); btrfs_mark_buffer_dirty(leaf); - btrfs_release_path(root, path); + btrfs_release_path(path); ret = 1; out_free: - if (ret == 0) { + if (ret != 1) { invalidate_inode_pages2_range(inode->i_mapping, 0, index); - spin_lock(&block_group->lock); - block_group->disk_cache_state = BTRFS_DC_ERROR; - spin_unlock(&block_group->lock); BTRFS_I(inode)->generation = 0; } kfree(checksums); + kfree(pages); btrfs_update_inode(trans, root, inode); + return ret; +} + +int btrfs_write_out_cache(struct btrfs_root *root, + struct btrfs_trans_handle *trans, + struct btrfs_block_group_cache *block_group, + struct btrfs_path *path) +{ + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; + struct inode *inode; + int ret = 0; + + root = root->fs_info->tree_root; + + spin_lock(&block_group->lock); + if (block_group->disk_cache_state < BTRFS_DC_SETUP) { + spin_unlock(&block_group->lock); + return 0; + } + spin_unlock(&block_group->lock); + + inode = lookup_free_space_inode(root, block_group, path); + if (IS_ERR(inode)) + return 0; + + ret = __btrfs_write_out_cache(root, inode, ctl, block_group, trans, + path, block_group->key.objectid); + if (ret < 0) { + spin_lock(&block_group->lock); + block_group->disk_cache_state = BTRFS_DC_ERROR; + spin_unlock(&block_group->lock); + ret = 0; + + printk(KERN_ERR "btrfs: failed to write free space cace " + "for block group %llu\n", block_group->key.objectid); + } + iput(inode); return ret; } -static inline unsigned long offset_to_bit(u64 bitmap_start, u64 sectorsize, +static inline unsigned long offset_to_bit(u64 bitmap_start, u32 unit, u64 offset) { BUG_ON(offset < bitmap_start); offset -= bitmap_start; - return (unsigned long)(div64_u64(offset, sectorsize)); + return (unsigned long)(div_u64(offset, unit)); } -static inline unsigned long bytes_to_bits(u64 bytes, u64 sectorsize) +static inline unsigned long bytes_to_bits(u64 bytes, u32 unit) { - return (unsigned long)(div64_u64(bytes, sectorsize)); + return (unsigned long)(div_u64(bytes, unit)); } -static inline u64 offset_to_bitmap(struct btrfs_block_group_cache *block_group, +static inline u64 offset_to_bitmap(struct btrfs_free_space_ctl *ctl, u64 offset) { u64 bitmap_start; u64 bytes_per_bitmap; - bytes_per_bitmap = BITS_PER_BITMAP * block_group->sectorsize; - bitmap_start = offset - block_group->key.objectid; + bytes_per_bitmap = BITS_PER_BITMAP * ctl->unit; + bitmap_start = offset - ctl->start; bitmap_start = div64_u64(bitmap_start, bytes_per_bitmap); bitmap_start *= bytes_per_bitmap; - bitmap_start += block_group->key.objectid; + bitmap_start += ctl->start; return bitmap_start; } @@ -939,10 +986,10 @@ static int tree_insert_offset(struct rb_root *root, u64 offset, * offset. */ static struct btrfs_free_space * -tree_search_offset(struct btrfs_block_group_cache *block_group, +tree_search_offset(struct btrfs_free_space_ctl *ctl, u64 offset, int bitmap_only, int fuzzy) { - struct rb_node *n = block_group->free_space_offset.rb_node; + struct rb_node *n = ctl->free_space_offset.rb_node; struct btrfs_free_space *entry, *prev = NULL; /* find entry that is closest to the 'offset' */ @@ -1038,8 +1085,7 @@ tree_search_offset(struct btrfs_block_group_cache *block_group, break; } } - if (entry->offset + BITS_PER_BITMAP * - block_group->sectorsize > offset) + if (entry->offset + BITS_PER_BITMAP * ctl->unit > offset) return entry; } else if (entry->offset + entry->bytes > offset) return entry; @@ -1050,7 +1096,7 @@ tree_search_offset(struct btrfs_block_group_cache *block_group, while (1) { if (entry->bitmap) { if (entry->offset + BITS_PER_BITMAP * - block_group->sectorsize > offset) + ctl->unit > offset) break; } else { if (entry->offset + entry->bytes > offset) @@ -1066,42 +1112,47 @@ tree_search_offset(struct btrfs_block_group_cache *block_group, } static inline void -__unlink_free_space(struct btrfs_block_group_cache *block_group, +__unlink_free_space(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info) { - rb_erase(&info->offset_index, &block_group->free_space_offset); - block_group->free_extents--; + rb_erase(&info->offset_index, &ctl->free_space_offset); + ctl->free_extents--; } -static void unlink_free_space(struct btrfs_block_group_cache *block_group, +static void unlink_free_space(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info) { - __unlink_free_space(block_group, info); - block_group->free_space -= info->bytes; + __unlink_free_space(ctl, info); + ctl->free_space -= info->bytes; } -static int link_free_space(struct btrfs_block_group_cache *block_group, +static int link_free_space(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info) { int ret = 0; BUG_ON(!info->bitmap && !info->bytes); - ret = tree_insert_offset(&block_group->free_space_offset, info->offset, + ret = tree_insert_offset(&ctl->free_space_offset, info->offset, &info->offset_index, (info->bitmap != NULL)); if (ret) return ret; - block_group->free_space += info->bytes; - block_group->free_extents++; + ctl->free_space += info->bytes; + ctl->free_extents++; return ret; } -static void recalculate_thresholds(struct btrfs_block_group_cache *block_group) +static void recalculate_thresholds(struct btrfs_free_space_ctl *ctl) { + struct btrfs_block_group_cache *block_group = ctl->private; u64 max_bytes; u64 bitmap_bytes; u64 extent_bytes; u64 size = block_group->key.offset; + u64 bytes_per_bg = BITS_PER_BITMAP * block_group->sectorsize; + int max_bitmaps = div64_u64(size + bytes_per_bg - 1, bytes_per_bg); + + BUG_ON(ctl->total_bitmaps > max_bitmaps); /* * The goal is to keep the total amount of memory used per 1gb of space @@ -1119,10 +1170,10 @@ static void recalculate_thresholds(struct btrfs_block_group_cache *block_group) * sure we don't go over our overall goal of MAX_CACHE_BYTES_PER_GIG as * we add more bitmaps. */ - bitmap_bytes = (block_group->total_bitmaps + 1) * PAGE_CACHE_SIZE; + bitmap_bytes = (ctl->total_bitmaps + 1) * PAGE_CACHE_SIZE; if (bitmap_bytes >= max_bytes) { - block_group->extents_thresh = 0; + ctl->extents_thresh = 0; return; } @@ -1133,47 +1184,43 @@ static void recalculate_thresholds(struct btrfs_block_group_cache *block_group) extent_bytes = max_bytes - bitmap_bytes; extent_bytes = min_t(u64, extent_bytes, div64_u64(max_bytes, 2)); - block_group->extents_thresh = + ctl->extents_thresh = div64_u64(extent_bytes, (sizeof(struct btrfs_free_space))); } -static void bitmap_clear_bits(struct btrfs_block_group_cache *block_group, +static void bitmap_clear_bits(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, u64 offset, u64 bytes) { - unsigned long start, end; - unsigned long i; + unsigned long start, count; - start = offset_to_bit(info->offset, block_group->sectorsize, offset); - end = start + bytes_to_bits(bytes, block_group->sectorsize); - BUG_ON(end > BITS_PER_BITMAP); + start = offset_to_bit(info->offset, ctl->unit, offset); + count = bytes_to_bits(bytes, ctl->unit); + BUG_ON(start + count > BITS_PER_BITMAP); - for (i = start; i < end; i++) - clear_bit(i, info->bitmap); + bitmap_clear(info->bitmap, start, count); info->bytes -= bytes; - block_group->free_space -= bytes; + ctl->free_space -= bytes; } -static void bitmap_set_bits(struct btrfs_block_group_cache *block_group, +static void bitmap_set_bits(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, u64 offset, u64 bytes) { - unsigned long start, end; - unsigned long i; + unsigned long start, count; - start = offset_to_bit(info->offset, block_group->sectorsize, offset); - end = start + bytes_to_bits(bytes, block_group->sectorsize); - BUG_ON(end > BITS_PER_BITMAP); + start = offset_to_bit(info->offset, ctl->unit, offset); + count = bytes_to_bits(bytes, ctl->unit); + BUG_ON(start + count > BITS_PER_BITMAP); - for (i = start; i < end; i++) - set_bit(i, info->bitmap); + bitmap_set(info->bitmap, start, count); info->bytes += bytes; - block_group->free_space += bytes; + ctl->free_space += bytes; } -static int search_bitmap(struct btrfs_block_group_cache *block_group, +static int search_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *bitmap_info, u64 *offset, u64 *bytes) { @@ -1181,9 +1228,9 @@ static int search_bitmap(struct btrfs_block_group_cache *block_group, unsigned long bits, i; unsigned long next_zero; - i = offset_to_bit(bitmap_info->offset, block_group->sectorsize, + i = offset_to_bit(bitmap_info->offset, ctl->unit, max_t(u64, *offset, bitmap_info->offset)); - bits = bytes_to_bits(*bytes, block_group->sectorsize); + bits = bytes_to_bits(*bytes, ctl->unit); for (i = find_next_bit(bitmap_info->bitmap, BITS_PER_BITMAP, i); i < BITS_PER_BITMAP; @@ -1198,29 +1245,25 @@ static int search_bitmap(struct btrfs_block_group_cache *block_group, } if (found_bits) { - *offset = (u64)(i * block_group->sectorsize) + - bitmap_info->offset; - *bytes = (u64)(found_bits) * block_group->sectorsize; + *offset = (u64)(i * ctl->unit) + bitmap_info->offset; + *bytes = (u64)(found_bits) * ctl->unit; return 0; } return -1; } -static struct btrfs_free_space *find_free_space(struct btrfs_block_group_cache - *block_group, u64 *offset, - u64 *bytes, int debug) +static struct btrfs_free_space * +find_free_space(struct btrfs_free_space_ctl *ctl, u64 *offset, u64 *bytes) { struct btrfs_free_space *entry; struct rb_node *node; int ret; - if (!block_group->free_space_offset.rb_node) + if (!ctl->free_space_offset.rb_node) return NULL; - entry = tree_search_offset(block_group, - offset_to_bitmap(block_group, *offset), - 0, 1); + entry = tree_search_offset(ctl, offset_to_bitmap(ctl, *offset), 0, 1); if (!entry) return NULL; @@ -1230,7 +1273,7 @@ static struct btrfs_free_space *find_free_space(struct btrfs_block_group_cache continue; if (entry->bitmap) { - ret = search_bitmap(block_group, entry, offset, bytes); + ret = search_bitmap(ctl, entry, offset, bytes); if (!ret) return entry; continue; @@ -1244,33 +1287,28 @@ static struct btrfs_free_space *find_free_space(struct btrfs_block_group_cache return NULL; } -static void add_new_bitmap(struct btrfs_block_group_cache *block_group, +static void add_new_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, u64 offset) { - u64 bytes_per_bg = BITS_PER_BITMAP * block_group->sectorsize; - int max_bitmaps = (int)div64_u64(block_group->key.offset + - bytes_per_bg - 1, bytes_per_bg); - BUG_ON(block_group->total_bitmaps >= max_bitmaps); - - info->offset = offset_to_bitmap(block_group, offset); + info->offset = offset_to_bitmap(ctl, offset); info->bytes = 0; - link_free_space(block_group, info); - block_group->total_bitmaps++; + link_free_space(ctl, info); + ctl->total_bitmaps++; - recalculate_thresholds(block_group); + ctl->op->recalc_thresholds(ctl); } -static void free_bitmap(struct btrfs_block_group_cache *block_group, +static void free_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *bitmap_info) { - unlink_free_space(block_group, bitmap_info); + unlink_free_space(ctl, bitmap_info); kfree(bitmap_info->bitmap); kmem_cache_free(btrfs_free_space_cachep, bitmap_info); - block_group->total_bitmaps--; - recalculate_thresholds(block_group); + ctl->total_bitmaps--; + ctl->op->recalc_thresholds(ctl); } -static noinline int remove_from_bitmap(struct btrfs_block_group_cache *block_group, +static noinline int remove_from_bitmap(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *bitmap_info, u64 *offset, u64 *bytes) { @@ -1279,8 +1317,7 @@ static noinline int remove_from_bitmap(struct btrfs_block_group_cache *block_gro int ret; again: - end = bitmap_info->offset + - (u64)(BITS_PER_BITMAP * block_group->sectorsize) - 1; + end = bitmap_info->offset + (u64)(BITS_PER_BITMAP * ctl->unit) - 1; /* * XXX - this can go away after a few releases. @@ -1295,24 +1332,22 @@ again: search_start = *offset; search_bytes = *bytes; search_bytes = min(search_bytes, end - search_start + 1); - ret = search_bitmap(block_group, bitmap_info, &search_start, - &search_bytes); + ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes); BUG_ON(ret < 0 || search_start != *offset); if (*offset > bitmap_info->offset && *offset + *bytes > end) { - bitmap_clear_bits(block_group, bitmap_info, *offset, - end - *offset + 1); + bitmap_clear_bits(ctl, bitmap_info, *offset, end - *offset + 1); *bytes -= end - *offset + 1; *offset = end + 1; } else if (*offset >= bitmap_info->offset && *offset + *bytes <= end) { - bitmap_clear_bits(block_group, bitmap_info, *offset, *bytes); + bitmap_clear_bits(ctl, bitmap_info, *offset, *bytes); *bytes = 0; } if (*bytes) { struct rb_node *next = rb_next(&bitmap_info->offset_index); if (!bitmap_info->bytes) - free_bitmap(block_group, bitmap_info); + free_bitmap(ctl, bitmap_info); /* * no entry after this bitmap, but we still have bytes to @@ -1339,31 +1374,28 @@ again: */ search_start = *offset; search_bytes = *bytes; - ret = search_bitmap(block_group, bitmap_info, &search_start, + ret = search_bitmap(ctl, bitmap_info, &search_start, &search_bytes); if (ret < 0 || search_start != *offset) return -EAGAIN; goto again; } else if (!bitmap_info->bytes) - free_bitmap(block_group, bitmap_info); + free_bitmap(ctl, bitmap_info); return 0; } -static int insert_into_bitmap(struct btrfs_block_group_cache *block_group, - struct btrfs_free_space *info) +static bool use_bitmap(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *info) { - struct btrfs_free_space *bitmap_info; - int added = 0; - u64 bytes, offset, end; - int ret; + struct btrfs_block_group_cache *block_group = ctl->private; /* * If we are below the extents threshold then we can add this as an * extent, and don't have to deal with the bitmap */ - if (block_group->free_extents < block_group->extents_thresh) { + if (ctl->free_extents < ctl->extents_thresh) { /* * If this block group has some small extents we don't want to * use up all of our free slots in the cache with them, we want @@ -1372,11 +1404,10 @@ static int insert_into_bitmap(struct btrfs_block_group_cache *block_group, * the overhead of a bitmap if we don't have to. */ if (info->bytes <= block_group->sectorsize * 4) { - if (block_group->free_extents * 2 <= - block_group->extents_thresh) - return 0; + if (ctl->free_extents * 2 <= ctl->extents_thresh) + return false; } else { - return 0; + return false; } } @@ -1386,31 +1417,42 @@ static int insert_into_bitmap(struct btrfs_block_group_cache *block_group, */ if (BITS_PER_BITMAP * block_group->sectorsize > block_group->key.offset) - return 0; + return false; + + return true; +} + +static int insert_into_bitmap(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *info) +{ + struct btrfs_free_space *bitmap_info; + int added = 0; + u64 bytes, offset, end; + int ret; bytes = info->bytes; offset = info->offset; + if (!ctl->op->use_bitmap(ctl, info)) + return 0; + again: - bitmap_info = tree_search_offset(block_group, - offset_to_bitmap(block_group, offset), + bitmap_info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), 1, 0); if (!bitmap_info) { BUG_ON(added); goto new_bitmap; } - end = bitmap_info->offset + - (u64)(BITS_PER_BITMAP * block_group->sectorsize); + end = bitmap_info->offset + (u64)(BITS_PER_BITMAP * ctl->unit); if (offset >= bitmap_info->offset && offset + bytes > end) { - bitmap_set_bits(block_group, bitmap_info, offset, - end - offset); + bitmap_set_bits(ctl, bitmap_info, offset, end - offset); bytes -= end - offset; offset = end; added = 0; } else if (offset >= bitmap_info->offset && offset + bytes <= end) { - bitmap_set_bits(block_group, bitmap_info, offset, bytes); + bitmap_set_bits(ctl, bitmap_info, offset, bytes); bytes = 0; } else { BUG(); @@ -1424,19 +1466,19 @@ again: new_bitmap: if (info && info->bitmap) { - add_new_bitmap(block_group, info, offset); + add_new_bitmap(ctl, info, offset); added = 1; info = NULL; goto again; } else { - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); /* no pre-allocated info, allocate a new one */ if (!info) { info = kmem_cache_zalloc(btrfs_free_space_cachep, GFP_NOFS); if (!info) { - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); ret = -ENOMEM; goto out; } @@ -1444,7 +1486,7 @@ new_bitmap: /* allocate the bitmap */ info->bitmap = kzalloc(PAGE_CACHE_SIZE, GFP_NOFS); - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); if (!info->bitmap) { ret = -ENOMEM; goto out; @@ -1462,7 +1504,7 @@ out: return ret; } -bool try_merge_free_space(struct btrfs_block_group_cache *block_group, +static bool try_merge_free_space(struct btrfs_free_space_ctl *ctl, struct btrfs_free_space *info, bool update_stat) { struct btrfs_free_space *left_info; @@ -1476,18 +1518,18 @@ bool try_merge_free_space(struct btrfs_block_group_cache *block_group, * are adding, if there is remove that struct and add a new one to * cover the entire range */ - right_info = tree_search_offset(block_group, offset + bytes, 0, 0); + right_info = tree_search_offset(ctl, offset + bytes, 0, 0); if (right_info && rb_prev(&right_info->offset_index)) left_info = rb_entry(rb_prev(&right_info->offset_index), struct btrfs_free_space, offset_index); else - left_info = tree_search_offset(block_group, offset - 1, 0, 0); + left_info = tree_search_offset(ctl, offset - 1, 0, 0); if (right_info && !right_info->bitmap) { if (update_stat) - unlink_free_space(block_group, right_info); + unlink_free_space(ctl, right_info); else - __unlink_free_space(block_group, right_info); + __unlink_free_space(ctl, right_info); info->bytes += right_info->bytes; kmem_cache_free(btrfs_free_space_cachep, right_info); merged = true; @@ -1496,9 +1538,9 @@ bool try_merge_free_space(struct btrfs_block_group_cache *block_group, if (left_info && !left_info->bitmap && left_info->offset + left_info->bytes == offset) { if (update_stat) - unlink_free_space(block_group, left_info); + unlink_free_space(ctl, left_info); else - __unlink_free_space(block_group, left_info); + __unlink_free_space(ctl, left_info); info->offset = left_info->offset; info->bytes += left_info->bytes; kmem_cache_free(btrfs_free_space_cachep, left_info); @@ -1508,8 +1550,8 @@ bool try_merge_free_space(struct btrfs_block_group_cache *block_group, return merged; } -int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, - u64 offset, u64 bytes) +int __btrfs_add_free_space(struct btrfs_free_space_ctl *ctl, + u64 offset, u64 bytes) { struct btrfs_free_space *info; int ret = 0; @@ -1521,9 +1563,9 @@ int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, info->offset = offset; info->bytes = bytes; - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); - if (try_merge_free_space(block_group, info, true)) + if (try_merge_free_space(ctl, info, true)) goto link; /* @@ -1531,7 +1573,7 @@ int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, * extent then we know we're going to have to allocate a new extent, so * before we do that see if we need to drop this into a bitmap */ - ret = insert_into_bitmap(block_group, info); + ret = insert_into_bitmap(ctl, info); if (ret < 0) { goto out; } else if (ret) { @@ -1539,11 +1581,11 @@ int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, goto out; } link: - ret = link_free_space(block_group, info); + ret = link_free_space(ctl, info); if (ret) kmem_cache_free(btrfs_free_space_cachep, info); out: - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); if (ret) { printk(KERN_CRIT "btrfs: unable to add free space :%d\n", ret); @@ -1556,21 +1598,21 @@ out: int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, u64 offset, u64 bytes) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *info; struct btrfs_free_space *next_info = NULL; int ret = 0; - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); again: - info = tree_search_offset(block_group, offset, 0, 0); + info = tree_search_offset(ctl, offset, 0, 0); if (!info) { /* * oops didn't find an extent that matched the space we wanted * to remove, look for a bitmap instead */ - info = tree_search_offset(block_group, - offset_to_bitmap(block_group, offset), + info = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), 1, 0); if (!info) { WARN_ON(1); @@ -1585,8 +1627,8 @@ again: offset_index); if (next_info->bitmap) - end = next_info->offset + BITS_PER_BITMAP * - block_group->sectorsize - 1; + end = next_info->offset + + BITS_PER_BITMAP * ctl->unit - 1; else end = next_info->offset + next_info->bytes; @@ -1606,20 +1648,20 @@ again: } if (info->bytes == bytes) { - unlink_free_space(block_group, info); + unlink_free_space(ctl, info); if (info->bitmap) { kfree(info->bitmap); - block_group->total_bitmaps--; + ctl->total_bitmaps--; } kmem_cache_free(btrfs_free_space_cachep, info); goto out_lock; } if (!info->bitmap && info->offset == offset) { - unlink_free_space(block_group, info); + unlink_free_space(ctl, info); info->offset += bytes; info->bytes -= bytes; - link_free_space(block_group, info); + link_free_space(ctl, info); goto out_lock; } @@ -1633,13 +1675,13 @@ again: * first unlink the old info and then * insert it again after the hole we're creating */ - unlink_free_space(block_group, info); + unlink_free_space(ctl, info); if (offset + bytes < info->offset + info->bytes) { u64 old_end = info->offset + info->bytes; info->offset = offset + bytes; info->bytes = old_end - info->offset; - ret = link_free_space(block_group, info); + ret = link_free_space(ctl, info); WARN_ON(ret); if (ret) goto out_lock; @@ -1649,7 +1691,7 @@ again: */ kmem_cache_free(btrfs_free_space_cachep, info); } - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); /* step two, insert a new info struct to cover * anything before the hole @@ -1660,12 +1702,12 @@ again: goto out; } - ret = remove_from_bitmap(block_group, info, &offset, &bytes); + ret = remove_from_bitmap(ctl, info, &offset, &bytes); if (ret == -EAGAIN) goto again; BUG_ON(ret); out_lock: - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); out: return ret; } @@ -1673,11 +1715,12 @@ out: void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group, u64 bytes) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *info; struct rb_node *n; int count = 0; - for (n = rb_first(&block_group->free_space_offset); n; n = rb_next(n)) { + for (n = rb_first(&ctl->free_space_offset); n; n = rb_next(n)) { info = rb_entry(n, struct btrfs_free_space, offset_index); if (info->bytes >= bytes) count++; @@ -1692,19 +1735,28 @@ void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group, "\n", count); } -u64 btrfs_block_group_free_space(struct btrfs_block_group_cache *block_group) +static struct btrfs_free_space_op free_space_op = { + .recalc_thresholds = recalculate_thresholds, + .use_bitmap = use_bitmap, +}; + +void btrfs_init_free_space_ctl(struct btrfs_block_group_cache *block_group) { - struct btrfs_free_space *info; - struct rb_node *n; - u64 ret = 0; + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; - for (n = rb_first(&block_group->free_space_offset); n; - n = rb_next(n)) { - info = rb_entry(n, struct btrfs_free_space, offset_index); - ret += info->bytes; - } + spin_lock_init(&ctl->tree_lock); + ctl->unit = block_group->sectorsize; + ctl->start = block_group->key.objectid; + ctl->private = block_group; + ctl->op = &free_space_op; - return ret; + /* + * we only want to have 32k of ram per block group for keeping + * track of free space, and if we pass 1/2 of that we want to + * start converting things over to using bitmaps + */ + ctl->extents_thresh = ((1024 * 32) / 2) / + sizeof(struct btrfs_free_space); } /* @@ -1718,6 +1770,7 @@ __btrfs_return_cluster_to_free_space( struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry; struct rb_node *node; @@ -1739,8 +1792,8 @@ __btrfs_return_cluster_to_free_space( bitmap = (entry->bitmap != NULL); if (!bitmap) - try_merge_free_space(block_group, entry, false); - tree_insert_offset(&block_group->free_space_offset, + try_merge_free_space(ctl, entry, false); + tree_insert_offset(&ctl->free_space_offset, entry->offset, &entry->offset_index, bitmap); } cluster->root = RB_ROOT; @@ -1751,14 +1804,38 @@ out: return 0; } -void btrfs_remove_free_space_cache(struct btrfs_block_group_cache *block_group) +void __btrfs_remove_free_space_cache_locked(struct btrfs_free_space_ctl *ctl) { struct btrfs_free_space *info; struct rb_node *node; + + while ((node = rb_last(&ctl->free_space_offset)) != NULL) { + info = rb_entry(node, struct btrfs_free_space, offset_index); + unlink_free_space(ctl, info); + kfree(info->bitmap); + kmem_cache_free(btrfs_free_space_cachep, info); + if (need_resched()) { + spin_unlock(&ctl->tree_lock); + cond_resched(); + spin_lock(&ctl->tree_lock); + } + } +} + +void __btrfs_remove_free_space_cache(struct btrfs_free_space_ctl *ctl) +{ + spin_lock(&ctl->tree_lock); + __btrfs_remove_free_space_cache_locked(ctl); + spin_unlock(&ctl->tree_lock); +} + +void btrfs_remove_free_space_cache(struct btrfs_block_group_cache *block_group) +{ + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_cluster *cluster; struct list_head *head; - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); while ((head = block_group->cluster_list.next) != &block_group->cluster_list) { cluster = list_entry(head, struct btrfs_free_cluster, @@ -1767,57 +1844,46 @@ void btrfs_remove_free_space_cache(struct btrfs_block_group_cache *block_group) WARN_ON(cluster->block_group != block_group); __btrfs_return_cluster_to_free_space(block_group, cluster); if (need_resched()) { - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); cond_resched(); - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); } } + __btrfs_remove_free_space_cache_locked(ctl); + spin_unlock(&ctl->tree_lock); - while ((node = rb_last(&block_group->free_space_offset)) != NULL) { - info = rb_entry(node, struct btrfs_free_space, offset_index); - unlink_free_space(block_group, info); - if (info->bitmap) - kfree(info->bitmap); - kmem_cache_free(btrfs_free_space_cachep, info); - if (need_resched()) { - spin_unlock(&block_group->tree_lock); - cond_resched(); - spin_lock(&block_group->tree_lock); - } - } - - spin_unlock(&block_group->tree_lock); } u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, u64 offset, u64 bytes, u64 empty_size) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry = NULL; u64 bytes_search = bytes + empty_size; u64 ret = 0; - spin_lock(&block_group->tree_lock); - entry = find_free_space(block_group, &offset, &bytes_search, 0); + spin_lock(&ctl->tree_lock); + entry = find_free_space(ctl, &offset, &bytes_search); if (!entry) goto out; ret = offset; if (entry->bitmap) { - bitmap_clear_bits(block_group, entry, offset, bytes); + bitmap_clear_bits(ctl, entry, offset, bytes); if (!entry->bytes) - free_bitmap(block_group, entry); + free_bitmap(ctl, entry); } else { - unlink_free_space(block_group, entry); + unlink_free_space(ctl, entry); entry->offset += bytes; entry->bytes -= bytes; if (!entry->bytes) kmem_cache_free(btrfs_free_space_cachep, entry); else - link_free_space(block_group, entry); + link_free_space(ctl, entry); } out: - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); return ret; } @@ -1834,6 +1900,7 @@ int btrfs_return_cluster_to_free_space( struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster) { + struct btrfs_free_space_ctl *ctl; int ret; /* first, get a safe pointer to the block group */ @@ -1852,10 +1919,12 @@ int btrfs_return_cluster_to_free_space( atomic_inc(&block_group->count); spin_unlock(&cluster->lock); + ctl = block_group->free_space_ctl; + /* now return any extents the cluster had on it */ - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); ret = __btrfs_return_cluster_to_free_space(block_group, cluster); - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); /* finally drop our ref */ btrfs_put_block_group(block_group); @@ -1867,6 +1936,7 @@ static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group, struct btrfs_free_space *entry, u64 bytes, u64 min_start) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; int err; u64 search_start = cluster->window_start; u64 search_bytes = bytes; @@ -1875,13 +1945,12 @@ static u64 btrfs_alloc_from_bitmap(struct btrfs_block_group_cache *block_group, search_start = min_start; search_bytes = bytes; - err = search_bitmap(block_group, entry, &search_start, - &search_bytes); + err = search_bitmap(ctl, entry, &search_start, &search_bytes); if (err) return 0; ret = search_start; - bitmap_clear_bits(block_group, entry, ret, bytes); + bitmap_clear_bits(ctl, entry, ret, bytes); return ret; } @@ -1895,6 +1964,7 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster, u64 bytes, u64 min_start) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry = NULL; struct rb_node *node; u64 ret = 0; @@ -1914,8 +1984,6 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group, while(1) { if (entry->bytes < bytes || (!entry->bitmap && entry->offset < min_start)) { - struct rb_node *node; - node = rb_next(&entry->offset_index); if (!node) break; @@ -1929,7 +1997,6 @@ u64 btrfs_alloc_from_cluster(struct btrfs_block_group_cache *block_group, cluster, entry, bytes, min_start); if (ret == 0) { - struct rb_node *node; node = rb_next(&entry->offset_index); if (!node) break; @@ -1955,20 +2022,20 @@ out: if (!ret) return 0; - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); - block_group->free_space -= bytes; + ctl->free_space -= bytes; if (entry->bytes == 0) { - block_group->free_extents--; + ctl->free_extents--; if (entry->bitmap) { kfree(entry->bitmap); - block_group->total_bitmaps--; - recalculate_thresholds(block_group); + ctl->total_bitmaps--; + ctl->op->recalc_thresholds(ctl); } kmem_cache_free(btrfs_free_space_cachep, entry); } - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); return ret; } @@ -1978,6 +2045,7 @@ static int btrfs_bitmap_cluster(struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster, u64 offset, u64 bytes, u64 min_bytes) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; unsigned long next_zero; unsigned long i; unsigned long search_bits; @@ -2032,7 +2100,7 @@ again: cluster->window_start = start * block_group->sectorsize + entry->offset; - rb_erase(&entry->offset_index, &block_group->free_space_offset); + rb_erase(&entry->offset_index, &ctl->free_space_offset); ret = tree_insert_offset(&cluster->root, entry->offset, &entry->offset_index, 1); BUG_ON(ret); @@ -2047,6 +2115,7 @@ static int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster, u64 offset, u64 bytes, u64 min_bytes) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *first = NULL; struct btrfs_free_space *entry = NULL; struct btrfs_free_space *prev = NULL; @@ -2057,7 +2126,7 @@ static int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, u64 max_extent; u64 max_gap = 128 * 1024; - entry = tree_search_offset(block_group, offset, 0, 1); + entry = tree_search_offset(ctl, offset, 0, 1); if (!entry) return -ENOSPC; @@ -2123,7 +2192,7 @@ static int setup_cluster_no_bitmap(struct btrfs_block_group_cache *block_group, if (entry->bitmap) continue; - rb_erase(&entry->offset_index, &block_group->free_space_offset); + rb_erase(&entry->offset_index, &ctl->free_space_offset); ret = tree_insert_offset(&cluster->root, entry->offset, &entry->offset_index, 0); BUG_ON(ret); @@ -2142,16 +2211,15 @@ static int setup_cluster_bitmap(struct btrfs_block_group_cache *block_group, struct btrfs_free_cluster *cluster, u64 offset, u64 bytes, u64 min_bytes) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry; struct rb_node *node; int ret = -ENOSPC; - if (block_group->total_bitmaps == 0) + if (ctl->total_bitmaps == 0) return -ENOSPC; - entry = tree_search_offset(block_group, - offset_to_bitmap(block_group, offset), - 0, 1); + entry = tree_search_offset(ctl, offset_to_bitmap(ctl, offset), 0, 1); if (!entry) return -ENOSPC; @@ -2184,6 +2252,7 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, struct btrfs_free_cluster *cluster, u64 offset, u64 bytes, u64 empty_size) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; u64 min_bytes; int ret; @@ -2203,14 +2272,14 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, } else min_bytes = max(bytes, (bytes + empty_size) >> 2); - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); /* * If we know we don't have enough space to make a cluster don't even * bother doing all the work to try and find one. */ - if (block_group->free_space < min_bytes) { - spin_unlock(&block_group->tree_lock); + if (ctl->free_space < min_bytes) { + spin_unlock(&ctl->tree_lock); return -ENOSPC; } @@ -2236,7 +2305,7 @@ int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, } out: spin_unlock(&cluster->lock); - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); return ret; } @@ -2257,6 +2326,7 @@ void btrfs_init_free_cluster(struct btrfs_free_cluster *cluster) int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, u64 *trimmed, u64 start, u64 end, u64 minlen) { + struct btrfs_free_space_ctl *ctl = block_group->free_space_ctl; struct btrfs_free_space *entry = NULL; struct btrfs_fs_info *fs_info = block_group->fs_info; u64 bytes = 0; @@ -2266,52 +2336,50 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, *trimmed = 0; while (start < end) { - spin_lock(&block_group->tree_lock); + spin_lock(&ctl->tree_lock); - if (block_group->free_space < minlen) { - spin_unlock(&block_group->tree_lock); + if (ctl->free_space < minlen) { + spin_unlock(&ctl->tree_lock); break; } - entry = tree_search_offset(block_group, start, 0, 1); + entry = tree_search_offset(ctl, start, 0, 1); if (!entry) - entry = tree_search_offset(block_group, - offset_to_bitmap(block_group, - start), + entry = tree_search_offset(ctl, + offset_to_bitmap(ctl, start), 1, 1); if (!entry || entry->offset >= end) { - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); break; } if (entry->bitmap) { - ret = search_bitmap(block_group, entry, &start, &bytes); + ret = search_bitmap(ctl, entry, &start, &bytes); if (!ret) { if (start >= end) { - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); break; } bytes = min(bytes, end - start); - bitmap_clear_bits(block_group, entry, - start, bytes); + bitmap_clear_bits(ctl, entry, start, bytes); if (entry->bytes == 0) - free_bitmap(block_group, entry); + free_bitmap(ctl, entry); } else { start = entry->offset + BITS_PER_BITMAP * block_group->sectorsize; - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); ret = 0; continue; } } else { start = entry->offset; bytes = min(entry->bytes, end - start); - unlink_free_space(block_group, entry); - kfree(entry); + unlink_free_space(ctl, entry); + kmem_cache_free(btrfs_free_space_cachep, entry); } - spin_unlock(&block_group->tree_lock); + spin_unlock(&ctl->tree_lock); if (bytes >= minlen) { int update_ret; @@ -2323,8 +2391,7 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, bytes, &actually_trimmed); - btrfs_add_free_space(block_group, - start, bytes); + btrfs_add_free_space(block_group, start, bytes); if (!update_ret) btrfs_update_reserved_bytes(block_group, bytes, 0, 1); @@ -2346,3 +2413,145 @@ int btrfs_trim_block_group(struct btrfs_block_group_cache *block_group, return ret; } + +/* + * Find the left-most item in the cache tree, and then return the + * smallest inode number in the item. + * + * Note: the returned inode number may not be the smallest one in + * the tree, if the left-most item is a bitmap. + */ +u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root) +{ + struct btrfs_free_space_ctl *ctl = fs_root->free_ino_ctl; + struct btrfs_free_space *entry = NULL; + u64 ino = 0; + + spin_lock(&ctl->tree_lock); + + if (RB_EMPTY_ROOT(&ctl->free_space_offset)) + goto out; + + entry = rb_entry(rb_first(&ctl->free_space_offset), + struct btrfs_free_space, offset_index); + + if (!entry->bitmap) { + ino = entry->offset; + + unlink_free_space(ctl, entry); + entry->offset++; + entry->bytes--; + if (!entry->bytes) + kmem_cache_free(btrfs_free_space_cachep, entry); + else + link_free_space(ctl, entry); + } else { + u64 offset = 0; + u64 count = 1; + int ret; + + ret = search_bitmap(ctl, entry, &offset, &count); + BUG_ON(ret); + + ino = offset; + bitmap_clear_bits(ctl, entry, offset, 1); + if (entry->bytes == 0) + free_bitmap(ctl, entry); + } +out: + spin_unlock(&ctl->tree_lock); + + return ino; +} + +struct inode *lookup_free_ino_inode(struct btrfs_root *root, + struct btrfs_path *path) +{ + struct inode *inode = NULL; + + spin_lock(&root->cache_lock); + if (root->cache_inode) + inode = igrab(root->cache_inode); + spin_unlock(&root->cache_lock); + if (inode) + return inode; + + inode = __lookup_free_space_inode(root, path, 0); + if (IS_ERR(inode)) + return inode; + + spin_lock(&root->cache_lock); + if (!root->fs_info->closing) + root->cache_inode = igrab(inode); + spin_unlock(&root->cache_lock); + + return inode; +} + +int create_free_ino_inode(struct btrfs_root *root, + struct btrfs_trans_handle *trans, + struct btrfs_path *path) +{ + return __create_free_space_inode(root, trans, path, + BTRFS_FREE_INO_OBJECTID, 0); +} + +int load_free_ino_cache(struct btrfs_fs_info *fs_info, struct btrfs_root *root) +{ + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct btrfs_path *path; + struct inode *inode; + int ret = 0; + u64 root_gen = btrfs_root_generation(&root->root_item); + + /* + * If we're unmounting then just return, since this does a search on the + * normal root and not the commit root and we could deadlock. + */ + smp_mb(); + if (fs_info->closing) + return 0; + + path = btrfs_alloc_path(); + if (!path) + return 0; + + inode = lookup_free_ino_inode(root, path); + if (IS_ERR(inode)) + goto out; + + if (root_gen != BTRFS_I(inode)->generation) + goto out_put; + + ret = __load_free_space_cache(root, inode, ctl, path, 0); + + if (ret < 0) + printk(KERN_ERR "btrfs: failed to load free ino cache for " + "root %llu\n", root->root_key.objectid); +out_put: + iput(inode); +out: + btrfs_free_path(path); + return ret; +} + +int btrfs_write_out_ino_cache(struct btrfs_root *root, + struct btrfs_trans_handle *trans, + struct btrfs_path *path) +{ + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct inode *inode; + int ret; + + inode = lookup_free_ino_inode(root, path); + if (IS_ERR(inode)) + return 0; + + ret = __btrfs_write_out_cache(root, inode, ctl, NULL, trans, path, 0); + if (ret < 0) + printk(KERN_ERR "btrfs: failed to write free ino cache " + "for root %llu\n", root->root_key.objectid); + + iput(inode); + return ret; +} diff --git a/fs/btrfs/free-space-cache.h b/fs/btrfs/free-space-cache.h index 65c3b935289f..8f2613f779ed 100644 --- a/fs/btrfs/free-space-cache.h +++ b/fs/btrfs/free-space-cache.h @@ -27,6 +27,25 @@ struct btrfs_free_space { struct list_head list; }; +struct btrfs_free_space_ctl { + spinlock_t tree_lock; + struct rb_root free_space_offset; + u64 free_space; + int extents_thresh; + int free_extents; + int total_bitmaps; + int unit; + u64 start; + struct btrfs_free_space_op *op; + void *private; +}; + +struct btrfs_free_space_op { + void (*recalc_thresholds)(struct btrfs_free_space_ctl *ctl); + bool (*use_bitmap)(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *info); +}; + struct inode *lookup_free_space_inode(struct btrfs_root *root, struct btrfs_block_group_cache *block_group, struct btrfs_path *path); @@ -45,17 +64,38 @@ int btrfs_write_out_cache(struct btrfs_root *root, struct btrfs_trans_handle *trans, struct btrfs_block_group_cache *block_group, struct btrfs_path *path); -int btrfs_add_free_space(struct btrfs_block_group_cache *block_group, - u64 bytenr, u64 size); + +struct inode *lookup_free_ino_inode(struct btrfs_root *root, + struct btrfs_path *path); +int create_free_ino_inode(struct btrfs_root *root, + struct btrfs_trans_handle *trans, + struct btrfs_path *path); +int load_free_ino_cache(struct btrfs_fs_info *fs_info, + struct btrfs_root *root); +int btrfs_write_out_ino_cache(struct btrfs_root *root, + struct btrfs_trans_handle *trans, + struct btrfs_path *path); + +void btrfs_init_free_space_ctl(struct btrfs_block_group_cache *block_group); +int __btrfs_add_free_space(struct btrfs_free_space_ctl *ctl, + u64 bytenr, u64 size); +static inline int +btrfs_add_free_space(struct btrfs_block_group_cache *block_group, + u64 bytenr, u64 size) +{ + return __btrfs_add_free_space(block_group->free_space_ctl, + bytenr, size); +} int btrfs_remove_free_space(struct btrfs_block_group_cache *block_group, u64 bytenr, u64 size); +void __btrfs_remove_free_space_cache(struct btrfs_free_space_ctl *ctl); void btrfs_remove_free_space_cache(struct btrfs_block_group_cache - *block_group); + *block_group); u64 btrfs_find_space_for_alloc(struct btrfs_block_group_cache *block_group, u64 offset, u64 bytes, u64 empty_size); +u64 btrfs_find_ino_for_alloc(struct btrfs_root *fs_root); void btrfs_dump_free_space(struct btrfs_block_group_cache *block_group, u64 bytes); -u64 btrfs_block_group_free_space(struct btrfs_block_group_cache *block_group); int btrfs_find_space_cluster(struct btrfs_trans_handle *trans, struct btrfs_root *root, struct btrfs_block_group_cache *block_group, diff --git a/fs/btrfs/inode-item.c b/fs/btrfs/inode-item.c index 64f1150bb48d..baa74f3db691 100644 --- a/fs/btrfs/inode-item.c +++ b/fs/btrfs/inode-item.c @@ -130,7 +130,6 @@ int btrfs_del_inode_ref(struct btrfs_trans_handle *trans, item_size - (ptr + sub_item_len - item_start)); ret = btrfs_truncate_item(trans, root, path, item_size - sub_item_len, 1); - BUG_ON(ret); out: btrfs_free_path(path); return ret; @@ -167,7 +166,6 @@ int btrfs_insert_inode_ref(struct btrfs_trans_handle *trans, old_size = btrfs_item_size_nr(path->nodes[0], path->slots[0]); ret = btrfs_extend_item(trans, root, path, ins_len); - BUG_ON(ret); ref = btrfs_item_ptr(path->nodes[0], path->slots[0], struct btrfs_inode_ref); ref = (struct btrfs_inode_ref *)((unsigned long)ref + old_size); diff --git a/fs/btrfs/inode-map.c b/fs/btrfs/inode-map.c index c05a08f4c411..3262cd17a12f 100644 --- a/fs/btrfs/inode-map.c +++ b/fs/btrfs/inode-map.c @@ -16,11 +16,446 @@ * Boston, MA 021110-1307, USA. */ +#include <linux/delay.h> +#include <linux/kthread.h> +#include <linux/pagemap.h> + #include "ctree.h" #include "disk-io.h" +#include "free-space-cache.h" +#include "inode-map.h" #include "transaction.h" -int btrfs_find_highest_inode(struct btrfs_root *root, u64 *objectid) +static int caching_kthread(void *data) +{ + struct btrfs_root *root = data; + struct btrfs_fs_info *fs_info = root->fs_info; + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct btrfs_key key; + struct btrfs_path *path; + struct extent_buffer *leaf; + u64 last = (u64)-1; + int slot; + int ret; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + /* Since the commit root is read-only, we can safely skip locking. */ + path->skip_locking = 1; + path->search_commit_root = 1; + path->reada = 2; + + key.objectid = BTRFS_FIRST_FREE_OBJECTID; + key.offset = 0; + key.type = BTRFS_INODE_ITEM_KEY; +again: + /* need to make sure the commit_root doesn't disappear */ + mutex_lock(&root->fs_commit_mutex); + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + goto out; + + while (1) { + smp_mb(); + if (fs_info->closing) + goto out; + + leaf = path->nodes[0]; + slot = path->slots[0]; + if (slot >= btrfs_header_nritems(leaf)) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) + goto out; + else if (ret > 0) + break; + + if (need_resched() || + btrfs_transaction_in_commit(fs_info)) { + leaf = path->nodes[0]; + + if (btrfs_header_nritems(leaf) == 0) { + WARN_ON(1); + break; + } + + /* + * Save the key so we can advances forward + * in the next search. + */ + btrfs_item_key_to_cpu(leaf, &key, 0); + btrfs_release_path(path); + root->cache_progress = last; + mutex_unlock(&root->fs_commit_mutex); + schedule_timeout(1); + goto again; + } else + continue; + } + + btrfs_item_key_to_cpu(leaf, &key, slot); + + if (key.type != BTRFS_INODE_ITEM_KEY) + goto next; + + if (key.objectid >= root->highest_objectid) + break; + + if (last != (u64)-1 && last + 1 != key.objectid) { + __btrfs_add_free_space(ctl, last + 1, + key.objectid - last - 1); + wake_up(&root->cache_wait); + } + + last = key.objectid; +next: + path->slots[0]++; + } + + if (last < root->highest_objectid - 1) { + __btrfs_add_free_space(ctl, last + 1, + root->highest_objectid - last - 1); + } + + spin_lock(&root->cache_lock); + root->cached = BTRFS_CACHE_FINISHED; + spin_unlock(&root->cache_lock); + + root->cache_progress = (u64)-1; + btrfs_unpin_free_ino(root); +out: + wake_up(&root->cache_wait); + mutex_unlock(&root->fs_commit_mutex); + + btrfs_free_path(path); + + return ret; +} + +static void start_caching(struct btrfs_root *root) +{ + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct task_struct *tsk; + int ret; + u64 objectid; + + spin_lock(&root->cache_lock); + if (root->cached != BTRFS_CACHE_NO) { + spin_unlock(&root->cache_lock); + return; + } + + root->cached = BTRFS_CACHE_STARTED; + spin_unlock(&root->cache_lock); + + ret = load_free_ino_cache(root->fs_info, root); + if (ret == 1) { + spin_lock(&root->cache_lock); + root->cached = BTRFS_CACHE_FINISHED; + spin_unlock(&root->cache_lock); + return; + } + + /* + * It can be quite time-consuming to fill the cache by searching + * through the extent tree, and this can keep ino allocation path + * waiting. Therefore at start we quickly find out the highest + * inode number and we know we can use inode numbers which fall in + * [highest_ino + 1, BTRFS_LAST_FREE_OBJECTID]. + */ + ret = btrfs_find_free_objectid(root, &objectid); + if (!ret && objectid <= BTRFS_LAST_FREE_OBJECTID) { + __btrfs_add_free_space(ctl, objectid, + BTRFS_LAST_FREE_OBJECTID - objectid + 1); + } + + tsk = kthread_run(caching_kthread, root, "btrfs-ino-cache-%llu\n", + root->root_key.objectid); + BUG_ON(IS_ERR(tsk)); +} + +int btrfs_find_free_ino(struct btrfs_root *root, u64 *objectid) +{ +again: + *objectid = btrfs_find_ino_for_alloc(root); + + if (*objectid != 0) + return 0; + + start_caching(root); + + wait_event(root->cache_wait, + root->cached == BTRFS_CACHE_FINISHED || + root->free_ino_ctl->free_space > 0); + + if (root->cached == BTRFS_CACHE_FINISHED && + root->free_ino_ctl->free_space == 0) + return -ENOSPC; + else + goto again; +} + +void btrfs_return_ino(struct btrfs_root *root, u64 objectid) +{ + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct btrfs_free_space_ctl *pinned = root->free_ino_pinned; +again: + if (root->cached == BTRFS_CACHE_FINISHED) { + __btrfs_add_free_space(ctl, objectid, 1); + } else { + /* + * If we are in the process of caching free ino chunks, + * to avoid adding the same inode number to the free_ino + * tree twice due to cross transaction, we'll leave it + * in the pinned tree until a transaction is committed + * or the caching work is done. + */ + + mutex_lock(&root->fs_commit_mutex); + spin_lock(&root->cache_lock); + if (root->cached == BTRFS_CACHE_FINISHED) { + spin_unlock(&root->cache_lock); + mutex_unlock(&root->fs_commit_mutex); + goto again; + } + spin_unlock(&root->cache_lock); + + start_caching(root); + + if (objectid <= root->cache_progress || + objectid > root->highest_objectid) + __btrfs_add_free_space(ctl, objectid, 1); + else + __btrfs_add_free_space(pinned, objectid, 1); + + mutex_unlock(&root->fs_commit_mutex); + } +} + +/* + * When a transaction is committed, we'll move those inode numbers which + * are smaller than root->cache_progress from pinned tree to free_ino tree, + * and others will just be dropped, because the commit root we were + * searching has changed. + * + * Must be called with root->fs_commit_mutex held + */ +void btrfs_unpin_free_ino(struct btrfs_root *root) +{ + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct rb_root *rbroot = &root->free_ino_pinned->free_space_offset; + struct btrfs_free_space *info; + struct rb_node *n; + u64 count; + + while (1) { + n = rb_first(rbroot); + if (!n) + break; + + info = rb_entry(n, struct btrfs_free_space, offset_index); + BUG_ON(info->bitmap); + + if (info->offset > root->cache_progress) + goto free; + else if (info->offset + info->bytes > root->cache_progress) + count = root->cache_progress - info->offset + 1; + else + count = info->bytes; + + __btrfs_add_free_space(ctl, info->offset, count); +free: + rb_erase(&info->offset_index, rbroot); + kfree(info); + } +} + +#define INIT_THRESHOLD (((1024 * 32) / 2) / sizeof(struct btrfs_free_space)) +#define INODES_PER_BITMAP (PAGE_CACHE_SIZE * 8) + +/* + * The goal is to keep the memory used by the free_ino tree won't + * exceed the memory if we use bitmaps only. + */ +static void recalculate_thresholds(struct btrfs_free_space_ctl *ctl) +{ + struct btrfs_free_space *info; + struct rb_node *n; + int max_ino; + int max_bitmaps; + + n = rb_last(&ctl->free_space_offset); + if (!n) { + ctl->extents_thresh = INIT_THRESHOLD; + return; + } + info = rb_entry(n, struct btrfs_free_space, offset_index); + + /* + * Find the maximum inode number in the filesystem. Note we + * ignore the fact that this can be a bitmap, because we are + * not doing precise calculation. + */ + max_ino = info->bytes - 1; + + max_bitmaps = ALIGN(max_ino, INODES_PER_BITMAP) / INODES_PER_BITMAP; + if (max_bitmaps <= ctl->total_bitmaps) { + ctl->extents_thresh = 0; + return; + } + + ctl->extents_thresh = (max_bitmaps - ctl->total_bitmaps) * + PAGE_CACHE_SIZE / sizeof(*info); +} + +/* + * We don't fall back to bitmap, if we are below the extents threshold + * or this chunk of inode numbers is a big one. + */ +static bool use_bitmap(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *info) +{ + if (ctl->free_extents < ctl->extents_thresh || + info->bytes > INODES_PER_BITMAP / 10) + return false; + + return true; +} + +static struct btrfs_free_space_op free_ino_op = { + .recalc_thresholds = recalculate_thresholds, + .use_bitmap = use_bitmap, +}; + +static void pinned_recalc_thresholds(struct btrfs_free_space_ctl *ctl) +{ +} + +static bool pinned_use_bitmap(struct btrfs_free_space_ctl *ctl, + struct btrfs_free_space *info) +{ + /* + * We always use extents for two reasons: + * + * - The pinned tree is only used during the process of caching + * work. + * - Make code simpler. See btrfs_unpin_free_ino(). + */ + return false; +} + +static struct btrfs_free_space_op pinned_free_ino_op = { + .recalc_thresholds = pinned_recalc_thresholds, + .use_bitmap = pinned_use_bitmap, +}; + +void btrfs_init_free_ino_ctl(struct btrfs_root *root) +{ + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct btrfs_free_space_ctl *pinned = root->free_ino_pinned; + + spin_lock_init(&ctl->tree_lock); + ctl->unit = 1; + ctl->start = 0; + ctl->private = NULL; + ctl->op = &free_ino_op; + + /* + * Initially we allow to use 16K of ram to cache chunks of + * inode numbers before we resort to bitmaps. This is somewhat + * arbitrary, but it will be adjusted in runtime. + */ + ctl->extents_thresh = INIT_THRESHOLD; + + spin_lock_init(&pinned->tree_lock); + pinned->unit = 1; + pinned->start = 0; + pinned->private = NULL; + pinned->extents_thresh = 0; + pinned->op = &pinned_free_ino_op; +} + +int btrfs_save_ino_cache(struct btrfs_root *root, + struct btrfs_trans_handle *trans) +{ + struct btrfs_free_space_ctl *ctl = root->free_ino_ctl; + struct btrfs_path *path; + struct inode *inode; + u64 alloc_hint = 0; + int ret; + int prealloc; + bool retry = false; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; +again: + inode = lookup_free_ino_inode(root, path); + if (IS_ERR(inode) && PTR_ERR(inode) != -ENOENT) { + ret = PTR_ERR(inode); + goto out; + } + + if (IS_ERR(inode)) { + BUG_ON(retry); + retry = true; + + ret = create_free_ino_inode(root, trans, path); + if (ret) + goto out; + goto again; + } + + BTRFS_I(inode)->generation = 0; + ret = btrfs_update_inode(trans, root, inode); + WARN_ON(ret); + + if (i_size_read(inode) > 0) { + ret = btrfs_truncate_free_space_cache(root, trans, path, inode); + if (ret) + goto out_put; + } + + spin_lock(&root->cache_lock); + if (root->cached != BTRFS_CACHE_FINISHED) { + ret = -1; + spin_unlock(&root->cache_lock); + goto out_put; + } + spin_unlock(&root->cache_lock); + + spin_lock(&ctl->tree_lock); + prealloc = sizeof(struct btrfs_free_space) * ctl->free_extents; + prealloc = ALIGN(prealloc, PAGE_CACHE_SIZE); + prealloc += ctl->total_bitmaps * PAGE_CACHE_SIZE; + spin_unlock(&ctl->tree_lock); + + /* Just to make sure we have enough space */ + prealloc += 8 * PAGE_CACHE_SIZE; + + ret = btrfs_check_data_free_space(inode, prealloc); + if (ret) + goto out_put; + + ret = btrfs_prealloc_file_range_trans(inode, trans, 0, 0, prealloc, + prealloc, prealloc, &alloc_hint); + if (ret) + goto out_put; + btrfs_free_reserved_data_space(inode, prealloc); + +out_put: + iput(inode); +out: + if (ret == 0) + ret = btrfs_write_out_ino_cache(root, trans, path); + + btrfs_free_path(path); + return ret; +} + +static int btrfs_find_highest_objectid(struct btrfs_root *root, u64 *objectid) { struct btrfs_path *path; int ret; @@ -55,15 +490,14 @@ error: return ret; } -int btrfs_find_free_objectid(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - u64 dirid, u64 *objectid) +int btrfs_find_free_objectid(struct btrfs_root *root, u64 *objectid) { int ret; mutex_lock(&root->objectid_mutex); if (unlikely(root->highest_objectid < BTRFS_FIRST_FREE_OBJECTID)) { - ret = btrfs_find_highest_inode(root, &root->highest_objectid); + ret = btrfs_find_highest_objectid(root, + &root->highest_objectid); if (ret) goto out; } diff --git a/fs/btrfs/inode-map.h b/fs/btrfs/inode-map.h new file mode 100644 index 000000000000..ddb347bfee23 --- /dev/null +++ b/fs/btrfs/inode-map.h @@ -0,0 +1,13 @@ +#ifndef __BTRFS_INODE_MAP +#define __BTRFS_INODE_MAP + +void btrfs_init_free_ino_ctl(struct btrfs_root *root); +void btrfs_unpin_free_ino(struct btrfs_root *root); +void btrfs_return_ino(struct btrfs_root *root, u64 objectid); +int btrfs_find_free_ino(struct btrfs_root *root, u64 *objectid); +int btrfs_save_ino_cache(struct btrfs_root *root, + struct btrfs_trans_handle *trans); + +int btrfs_find_free_objectid(struct btrfs_root *root, u64 *objectid); + +#endif diff --git a/fs/btrfs/inode.c b/fs/btrfs/inode.c index 65413394daef..39a9d5750efd 100644 --- a/fs/btrfs/inode.c +++ b/fs/btrfs/inode.c @@ -37,6 +37,7 @@ #include <linux/posix_acl.h> #include <linux/falloc.h> #include <linux/slab.h> +#include <linux/ratelimit.h> #include "compat.h" #include "ctree.h" #include "disk-io.h" @@ -51,6 +52,7 @@ #include "compression.h" #include "locking.h" #include "free-space-cache.h" +#include "inode-map.h" struct btrfs_iget_args { u64 ino; @@ -138,7 +140,7 @@ static noinline int insert_inline_extent(struct btrfs_trans_handle *trans, path->leave_spinning = 1; btrfs_set_trans_block_group(trans, inode); - key.objectid = inode->i_ino; + key.objectid = btrfs_ino(inode); key.offset = start; btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY); datasize = btrfs_file_extent_calc_inline_size(cur_size); @@ -340,6 +342,10 @@ static noinline int compress_file_range(struct inode *inode, int will_compress; int compress_type = root->fs_info->compress_type; + /* if this is a small write inside eof, kick off a defragbot */ + if (end <= BTRFS_I(inode)->disk_i_size && (end - start + 1) < 16 * 1024) + btrfs_add_inode_defrag(NULL, inode); + actual_end = min_t(u64, isize, end + 1); again: will_compress = 0; @@ -649,7 +655,7 @@ retry: async_extent->start + async_extent->ram_size - 1, 0); - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); BUG_ON(!em); em->start = async_extent->start; em->len = async_extent->ram_size; @@ -745,6 +751,15 @@ static u64 get_extent_allocation_hint(struct inode *inode, u64 start, return alloc_hint; } +static inline bool is_free_space_inode(struct btrfs_root *root, + struct inode *inode) +{ + if (root == root->fs_info->tree_root || + BTRFS_I(inode)->location.objectid == BTRFS_FREE_INO_OBJECTID) + return true; + return false; +} + /* * when extent_io.c finds a delayed allocation range in the file, * the call backs end up in this code. The basic idea is to @@ -777,7 +792,7 @@ static noinline int cow_file_range(struct inode *inode, struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; int ret = 0; - BUG_ON(root == root->fs_info->tree_root); + BUG_ON(is_free_space_inode(root, inode)); trans = btrfs_join_transaction(root, 1); BUG_ON(IS_ERR(trans)); btrfs_set_trans_block_group(trans, inode); @@ -788,6 +803,10 @@ static noinline int cow_file_range(struct inode *inode, disk_num_bytes = num_bytes; ret = 0; + /* if this is a small write inside eof, kick off defrag */ + if (end <= BTRFS_I(inode)->disk_i_size && num_bytes < 64 * 1024) + btrfs_add_inode_defrag(trans, inode); + if (start == 0) { /* lets try to make an inline extent */ ret = cow_file_range_inline(trans, root, inode, @@ -826,7 +845,7 @@ static noinline int cow_file_range(struct inode *inode, (u64)-1, &ins, 1); BUG_ON(ret); - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); BUG_ON(!em); em->start = start; em->orig_start = em->start; @@ -954,6 +973,7 @@ static int cow_file_range_async(struct inode *inode, struct page *locked_page, 1, 0, NULL, GFP_NOFS); while (start < end) { async_cow = kmalloc(sizeof(*async_cow), GFP_NOFS); + BUG_ON(!async_cow); async_cow->inode = inode; async_cow->root = root; async_cow->locked_page = locked_page; @@ -1007,7 +1027,7 @@ static noinline int csum_exist_in_range(struct btrfs_root *root, LIST_HEAD(list); ret = btrfs_lookup_csums_range(root->fs_info->csum_root, bytenr, - bytenr + num_bytes - 1, &list); + bytenr + num_bytes - 1, &list, 0); if (ret == 0 && list_empty(&list)) return 0; @@ -1048,29 +1068,31 @@ static noinline int run_delalloc_nocow(struct inode *inode, int type; int nocow; int check_prev = 1; - bool nolock = false; + bool nolock; + u64 ino = btrfs_ino(inode); path = btrfs_alloc_path(); BUG_ON(!path); - if (root == root->fs_info->tree_root) { - nolock = true; + + nolock = is_free_space_inode(root, inode); + + if (nolock) trans = btrfs_join_transaction_nolock(root, 1); - } else { + else trans = btrfs_join_transaction(root, 1); - } BUG_ON(IS_ERR(trans)); cow_start = (u64)-1; cur_offset = start; while (1) { - ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino, + ret = btrfs_lookup_file_extent(trans, root, path, ino, cur_offset, 0); BUG_ON(ret < 0); if (ret > 0 && path->slots[0] > 0 && check_prev) { leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0] - 1); - if (found_key.objectid == inode->i_ino && + if (found_key.objectid == ino && found_key.type == BTRFS_EXTENT_DATA_KEY) path->slots[0]--; } @@ -1091,7 +1113,7 @@ next_slot: num_bytes = 0; btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); - if (found_key.objectid > inode->i_ino || + if (found_key.objectid > ino || found_key.type > BTRFS_EXTENT_DATA_KEY || found_key.offset > end) break; @@ -1126,7 +1148,7 @@ next_slot: goto out_check; if (btrfs_extent_readonly(root, disk_bytenr)) goto out_check; - if (btrfs_cross_ref_exist(trans, root, inode->i_ino, + if (btrfs_cross_ref_exist(trans, root, ino, found_key.offset - extent_offset, disk_bytenr)) goto out_check; @@ -1163,7 +1185,7 @@ out_check: goto next_slot; } - btrfs_release_path(root, path); + btrfs_release_path(path); if (cow_start != (u64)-1) { ret = cow_file_range(inode, locked_page, cow_start, found_key.offset - 1, page_started, @@ -1176,7 +1198,7 @@ out_check: struct extent_map *em; struct extent_map_tree *em_tree; em_tree = &BTRFS_I(inode)->extent_tree; - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); BUG_ON(!em); em->start = cur_offset; em->orig_start = em->start; @@ -1221,7 +1243,7 @@ out_check: if (cur_offset > end) break; } - btrfs_release_path(root, path); + btrfs_release_path(path); if (cur_offset <= end && cow_start == (u64)-1) cow_start = cur_offset; @@ -1309,14 +1331,13 @@ static int btrfs_set_bit_hook(struct inode *inode, /* * set_bit and clear bit hooks normally require _irqsave/restore - * but in this case, we are only testeing for the DELALLOC + * but in this case, we are only testing for the DELALLOC * bit, which is only set or cleared with irqs on */ if (!(state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) { struct btrfs_root *root = BTRFS_I(inode)->root; u64 len = state->end + 1 - state->start; - int do_list = (root->root_key.objectid != - BTRFS_ROOT_TREE_OBJECTID); + bool do_list = !is_free_space_inode(root, inode); if (*bits & EXTENT_FIRST_DELALLOC) *bits &= ~EXTENT_FIRST_DELALLOC; @@ -1343,14 +1364,13 @@ static int btrfs_clear_bit_hook(struct inode *inode, { /* * set_bit and clear bit hooks normally require _irqsave/restore - * but in this case, we are only testeing for the DELALLOC + * but in this case, we are only testing for the DELALLOC * bit, which is only set or cleared with irqs on */ if ((state->state & EXTENT_DELALLOC) && (*bits & EXTENT_DELALLOC)) { struct btrfs_root *root = BTRFS_I(inode)->root; u64 len = state->end + 1 - state->start; - int do_list = (root->root_key.objectid != - BTRFS_ROOT_TREE_OBJECTID); + bool do_list = !is_free_space_inode(root, inode); if (*bits & EXTENT_FIRST_DELALLOC) *bits &= ~EXTENT_FIRST_DELALLOC; @@ -1457,7 +1477,7 @@ static int btrfs_submit_bio_hook(struct inode *inode, int rw, struct bio *bio, skip_sum = BTRFS_I(inode)->flags & BTRFS_INODE_NODATASUM; - if (root == root->fs_info->tree_root) + if (is_free_space_inode(root, inode)) ret = btrfs_bio_wq_end_io(root->fs_info, bio, 2); else ret = btrfs_bio_wq_end_io(root->fs_info, bio, 0); @@ -1643,7 +1663,7 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans, &hint, 0); BUG_ON(ret); - ins.objectid = inode->i_ino; + ins.objectid = btrfs_ino(inode); ins.offset = file_pos; ins.type = BTRFS_EXTENT_DATA_KEY; ret = btrfs_insert_empty_item(trans, root, path, &ins, sizeof(*fi)); @@ -1674,7 +1694,7 @@ static int insert_reserved_file_extent(struct btrfs_trans_handle *trans, ins.type = BTRFS_EXTENT_ITEM_KEY; ret = btrfs_alloc_reserved_file_extent(trans, root, root->root_key.objectid, - inode->i_ino, file_pos, &ins); + btrfs_ino(inode), file_pos, &ins); BUG_ON(ret); btrfs_free_path(path); @@ -1700,7 +1720,7 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end) struct extent_state *cached_state = NULL; int compress_type = 0; int ret; - bool nolock = false; + bool nolock; ret = btrfs_dec_test_ordered_pending(inode, &ordered_extent, start, end - start + 1); @@ -1708,7 +1728,7 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end) return 0; BUG_ON(!ordered_extent); - nolock = (root == root->fs_info->tree_root); + nolock = is_free_space_inode(root, inode); if (test_bit(BTRFS_ORDERED_NOCOW, &ordered_extent->flags)) { BUG_ON(!list_empty(&ordered_extent->list)); @@ -1770,9 +1790,12 @@ static int btrfs_finish_ordered_io(struct inode *inode, u64 start, u64 end) add_pending_csums(trans, inode, ordered_extent->file_offset, &ordered_extent->list); - btrfs_ordered_update_i_size(inode, 0, ordered_extent); - ret = btrfs_update_inode(trans, root, inode); - BUG_ON(ret); + ret = btrfs_ordered_update_i_size(inode, 0, ordered_extent); + if (!ret) { + ret = btrfs_update_inode(trans, root, inode); + BUG_ON(ret); + } + ret = 0; out: if (nolock) { if (trans) @@ -1851,7 +1874,7 @@ static int btrfs_io_failed_hook(struct bio *failed_bio, } read_unlock(&em_tree->lock); - if (!em || IS_ERR(em)) { + if (IS_ERR_OR_NULL(em)) { kfree(failrec); return -EIO; } @@ -2000,12 +2023,11 @@ good: return 0; zeroit: - if (printk_ratelimit()) { - printk(KERN_INFO "btrfs csum failed ino %lu off %llu csum %u " - "private %llu\n", page->mapping->host->i_ino, + printk_ratelimited(KERN_INFO "btrfs csum failed ino %llu off %llu csum %u " + "private %llu\n", + (unsigned long long)btrfs_ino(page->mapping->host), (unsigned long long)start, csum, (unsigned long long)private); - } memset(kaddr + offset, 1, end - start + 1); flush_dcache_page(page); kunmap_atomic(kaddr, KM_USER0); @@ -2240,7 +2262,7 @@ int btrfs_orphan_add(struct btrfs_trans_handle *trans, struct inode *inode) /* insert an orphan item to track this unlinked/truncated file */ if (insert >= 1) { - ret = btrfs_insert_orphan_item(trans, root, inode->i_ino); + ret = btrfs_insert_orphan_item(trans, root, btrfs_ino(inode)); BUG_ON(ret); } @@ -2277,7 +2299,7 @@ int btrfs_orphan_del(struct btrfs_trans_handle *trans, struct inode *inode) spin_unlock(&root->orphan_lock); if (trans && delete_item) { - ret = btrfs_del_orphan_item(trans, root, inode->i_ino); + ret = btrfs_del_orphan_item(trans, root, btrfs_ino(inode)); BUG_ON(ret); } @@ -2321,7 +2343,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) /* * if ret == 0 means we found what we were searching for, which - * is weird, but possible, so only screw with path if we didnt + * is weird, but possible, so only screw with path if we didn't * find the key and see if we have stuff that matches */ if (ret > 0) { @@ -2342,7 +2364,7 @@ int btrfs_orphan_cleanup(struct btrfs_root *root) break; /* release the path since we're done with it */ - btrfs_release_path(root, path); + btrfs_release_path(path); /* * this is where we are basically btrfs_lookup, without the @@ -2539,7 +2561,8 @@ static void btrfs_read_locked_inode(struct inode *inode) * try to precache a NULL acl entry for files that don't have * any xattrs or acls */ - maybe_acls = acls_after_inode_item(leaf, path->slots[0], inode->i_ino); + maybe_acls = acls_after_inode_item(leaf, path->slots[0], + btrfs_ino(inode)); if (!maybe_acls) cache_no_acl(inode); @@ -2590,6 +2613,13 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, struct btrfs_inode_item *item, struct inode *inode) { + if (!leaf->map_token) + map_private_extent_buffer(leaf, (unsigned long)item, + sizeof(struct btrfs_inode_item), + &leaf->map_token, &leaf->kaddr, + &leaf->map_start, &leaf->map_len, + KM_USER1); + btrfs_set_inode_uid(leaf, item, inode->i_uid); btrfs_set_inode_gid(leaf, item, inode->i_gid); btrfs_set_inode_size(leaf, item, BTRFS_I(inode)->disk_i_size); @@ -2618,6 +2648,11 @@ static void fill_inode_item(struct btrfs_trans_handle *trans, btrfs_set_inode_rdev(leaf, item, inode->i_rdev); btrfs_set_inode_flags(leaf, item, BTRFS_I(inode)->flags); btrfs_set_inode_block_group(leaf, item, BTRFS_I(inode)->block_group); + + if (leaf->map_token) { + unmap_extent_buffer(leaf, leaf->map_token, KM_USER1); + leaf->map_token = NULL; + } } /* @@ -2631,11 +2666,26 @@ noinline int btrfs_update_inode(struct btrfs_trans_handle *trans, struct extent_buffer *leaf; int ret; + /* + * If root is tree root, it means this inode is used to + * store free space information. And these inodes are updated + * when committing the transaction, so they needn't delaye to + * be updated, or deadlock will occured. + */ + if (!is_free_space_inode(root, inode)) { + ret = btrfs_delayed_update_inode(trans, root, inode); + if (!ret) + btrfs_set_inode_last_trans(trans, inode); + return ret; + } + path = btrfs_alloc_path(); - BUG_ON(!path); + if (!path) + return -ENOMEM; + path->leave_spinning = 1; - ret = btrfs_lookup_inode(trans, root, path, - &BTRFS_I(inode)->location, 1); + ret = btrfs_lookup_inode(trans, root, path, &BTRFS_I(inode)->location, + 1); if (ret) { if (ret > 0) ret = -ENOENT; @@ -2645,7 +2695,7 @@ noinline int btrfs_update_inode(struct btrfs_trans_handle *trans, btrfs_unlock_up_safe(path, 1); leaf = path->nodes[0]; inode_item = btrfs_item_ptr(leaf, path->slots[0], - struct btrfs_inode_item); + struct btrfs_inode_item); fill_inode_item(trans, leaf, inode_item, inode); btrfs_mark_buffer_dirty(leaf); @@ -2656,7 +2706,6 @@ failed: return ret; } - /* * unlink helper that gets used here in inode.c and in the tree logging * recovery code. It remove a link in a directory with a given name, and @@ -2673,6 +2722,8 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans, struct btrfs_dir_item *di; struct btrfs_key key; u64 index; + u64 ino = btrfs_ino(inode); + u64 dir_ino = btrfs_ino(dir); path = btrfs_alloc_path(); if (!path) { @@ -2681,7 +2732,7 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans, } path->leave_spinning = 1; - di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino, + di = btrfs_lookup_dir_item(trans, root, path, dir_ino, name, name_len, -1); if (IS_ERR(di)) { ret = PTR_ERR(di); @@ -2696,33 +2747,23 @@ static int __btrfs_unlink_inode(struct btrfs_trans_handle *trans, ret = btrfs_delete_one_dir_name(trans, root, path, di); if (ret) goto err; - btrfs_release_path(root, path); + btrfs_release_path(path); - ret = btrfs_del_inode_ref(trans, root, name, name_len, - inode->i_ino, - dir->i_ino, &index); + ret = btrfs_del_inode_ref(trans, root, name, name_len, ino, + dir_ino, &index); if (ret) { printk(KERN_INFO "btrfs failed to delete reference to %.*s, " - "inode %lu parent %lu\n", name_len, name, - inode->i_ino, dir->i_ino); + "inode %llu parent %llu\n", name_len, name, + (unsigned long long)ino, (unsigned long long)dir_ino); goto err; } - di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino, - index, name, name_len, -1); - if (IS_ERR(di)) { - ret = PTR_ERR(di); - goto err; - } - if (!di) { - ret = -ENOENT; + ret = btrfs_delete_delayed_dir_index(trans, root, dir, index); + if (ret) goto err; - } - ret = btrfs_delete_one_dir_name(trans, root, path, di); - btrfs_release_path(root, path); ret = btrfs_del_inode_ref_in_log(trans, root, name, name_len, - inode, dir->i_ino); + inode, dir_ino); BUG_ON(ret != 0 && ret != -ENOENT); ret = btrfs_del_dir_entries_in_log(trans, root, name, name_len, @@ -2800,12 +2841,14 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, int check_link = 1; int err = -ENOSPC; int ret; + u64 ino = btrfs_ino(inode); + u64 dir_ino = btrfs_ino(dir); trans = btrfs_start_transaction(root, 10); if (!IS_ERR(trans) || PTR_ERR(trans) != -ENOSPC) return trans; - if (inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID) + if (ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID) return ERR_PTR(-ENOSPC); /* check if there is someone else holds reference */ @@ -2846,7 +2889,7 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, } else { check_link = 0; } - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_lookup_inode(trans, root, path, &BTRFS_I(inode)->location, 0); @@ -2860,11 +2903,11 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, } else { check_link = 0; } - btrfs_release_path(root, path); + btrfs_release_path(path); if (ret == 0 && S_ISREG(inode->i_mode)) { ret = btrfs_lookup_file_extent(trans, root, path, - inode->i_ino, (u64)-1, 0); + ino, (u64)-1, 0); if (ret < 0) { err = ret; goto out; @@ -2872,7 +2915,7 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, BUG_ON(ret == 0); if (check_path_shared(root, path)) goto out; - btrfs_release_path(root, path); + btrfs_release_path(path); } if (!check_link) { @@ -2880,7 +2923,7 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, goto out; } - di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino, + di = btrfs_lookup_dir_item(trans, root, path, dir_ino, dentry->d_name.name, dentry->d_name.len, 0); if (IS_ERR(di)) { err = PTR_ERR(di); @@ -2893,11 +2936,11 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, err = 0; goto out; } - btrfs_release_path(root, path); + btrfs_release_path(path); ref = btrfs_lookup_inode_ref(trans, root, path, dentry->d_name.name, dentry->d_name.len, - inode->i_ino, dir->i_ino, 0); + ino, dir_ino, 0); if (IS_ERR(ref)) { err = PTR_ERR(ref); goto out; @@ -2906,9 +2949,17 @@ static struct btrfs_trans_handle *__unlink_start_trans(struct inode *dir, if (check_path_shared(root, path)) goto out; index = btrfs_inode_ref_index(path->nodes[0], ref); - btrfs_release_path(root, path); + btrfs_release_path(path); - di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino, index, + /* + * This is a commit root search, if we can lookup inode item and other + * relative items in the commit root, it means the transaction of + * dir/file creation has been committed, and the dir index item that we + * delay to insert has also been inserted into the commit root. So + * we needn't worry about the delayed insertion of the dir index item + * here. + */ + di = btrfs_lookup_dir_index_item(trans, root, path, dir_ino, index, dentry->d_name.name, dentry->d_name.len, 0); if (IS_ERR(di)) { err = PTR_ERR(di); @@ -2983,54 +3034,47 @@ int btrfs_unlink_subvol(struct btrfs_trans_handle *trans, struct btrfs_key key; u64 index; int ret; + u64 dir_ino = btrfs_ino(dir); path = btrfs_alloc_path(); if (!path) return -ENOMEM; - di = btrfs_lookup_dir_item(trans, root, path, dir->i_ino, + di = btrfs_lookup_dir_item(trans, root, path, dir_ino, name, name_len, -1); - BUG_ON(!di || IS_ERR(di)); + BUG_ON(IS_ERR_OR_NULL(di)); leaf = path->nodes[0]; btrfs_dir_item_key_to_cpu(leaf, di, &key); WARN_ON(key.type != BTRFS_ROOT_ITEM_KEY || key.objectid != objectid); ret = btrfs_delete_one_dir_name(trans, root, path, di); BUG_ON(ret); - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_del_root_ref(trans, root->fs_info->tree_root, objectid, root->root_key.objectid, - dir->i_ino, &index, name, name_len); + dir_ino, &index, name, name_len); if (ret < 0) { BUG_ON(ret != -ENOENT); - di = btrfs_search_dir_index_item(root, path, dir->i_ino, + di = btrfs_search_dir_index_item(root, path, dir_ino, name, name_len); - BUG_ON(!di || IS_ERR(di)); + BUG_ON(IS_ERR_OR_NULL(di)); leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - btrfs_release_path(root, path); + btrfs_release_path(path); index = key.offset; } + btrfs_release_path(path); - di = btrfs_lookup_dir_index_item(trans, root, path, dir->i_ino, - index, name, name_len, -1); - BUG_ON(!di || IS_ERR(di)); - - leaf = path->nodes[0]; - btrfs_dir_item_key_to_cpu(leaf, di, &key); - WARN_ON(key.type != BTRFS_ROOT_ITEM_KEY || key.objectid != objectid); - ret = btrfs_delete_one_dir_name(trans, root, path, di); + ret = btrfs_delete_delayed_dir_index(trans, root, dir, index); BUG_ON(ret); - btrfs_release_path(root, path); btrfs_i_size_write(dir, dir->i_size - name_len * 2); dir->i_mtime = dir->i_ctime = CURRENT_TIME; ret = btrfs_update_inode(trans, root, dir); BUG_ON(ret); - btrfs_free_path(path); return 0; } @@ -3043,7 +3087,7 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) unsigned long nr = 0; if (inode->i_size > BTRFS_EMPTY_DIR_SIZE || - inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) + btrfs_ino(inode) == BTRFS_FIRST_FREE_OBJECTID) return -ENOTEMPTY; trans = __unlink_start_trans(dir, dentry); @@ -3052,7 +3096,7 @@ static int btrfs_rmdir(struct inode *dir, struct dentry *dentry) btrfs_set_trans_block_group(trans, dir); - if (unlikely(inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) { + if (unlikely(btrfs_ino(inode) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) { err = btrfs_unlink_subvol(trans, root, dir, BTRFS_I(inode)->location.objectid, dentry->d_name.name, @@ -3077,178 +3121,6 @@ out: return err; } -#if 0 -/* - * when truncating bytes in a file, it is possible to avoid reading - * the leaves that contain only checksum items. This can be the - * majority of the IO required to delete a large file, but it must - * be done carefully. - * - * The keys in the level just above the leaves are checked to make sure - * the lowest key in a given leaf is a csum key, and starts at an offset - * after the new size. - * - * Then the key for the next leaf is checked to make sure it also has - * a checksum item for the same file. If it does, we know our target leaf - * contains only checksum items, and it can be safely freed without reading - * it. - * - * This is just an optimization targeted at large files. It may do - * nothing. It will return 0 unless things went badly. - */ -static noinline int drop_csum_leaves(struct btrfs_trans_handle *trans, - struct btrfs_root *root, - struct btrfs_path *path, - struct inode *inode, u64 new_size) -{ - struct btrfs_key key; - int ret; - int nritems; - struct btrfs_key found_key; - struct btrfs_key other_key; - struct btrfs_leaf_ref *ref; - u64 leaf_gen; - u64 leaf_start; - - path->lowest_level = 1; - key.objectid = inode->i_ino; - key.type = BTRFS_CSUM_ITEM_KEY; - key.offset = new_size; -again: - ret = btrfs_search_slot(trans, root, &key, path, -1, 1); - if (ret < 0) - goto out; - - if (path->nodes[1] == NULL) { - ret = 0; - goto out; - } - ret = 0; - btrfs_node_key_to_cpu(path->nodes[1], &found_key, path->slots[1]); - nritems = btrfs_header_nritems(path->nodes[1]); - - if (!nritems) - goto out; - - if (path->slots[1] >= nritems) - goto next_node; - - /* did we find a key greater than anything we want to delete? */ - if (found_key.objectid > inode->i_ino || - (found_key.objectid == inode->i_ino && found_key.type > key.type)) - goto out; - - /* we check the next key in the node to make sure the leave contains - * only checksum items. This comparison doesn't work if our - * leaf is the last one in the node - */ - if (path->slots[1] + 1 >= nritems) { -next_node: - /* search forward from the last key in the node, this - * will bring us into the next node in the tree - */ - btrfs_node_key_to_cpu(path->nodes[1], &found_key, nritems - 1); - - /* unlikely, but we inc below, so check to be safe */ - if (found_key.offset == (u64)-1) - goto out; - - /* search_forward needs a path with locks held, do the - * search again for the original key. It is possible - * this will race with a balance and return a path that - * we could modify, but this drop is just an optimization - * and is allowed to miss some leaves. - */ - btrfs_release_path(root, path); - found_key.offset++; - - /* setup a max key for search_forward */ - other_key.offset = (u64)-1; - other_key.type = key.type; - other_key.objectid = key.objectid; - - path->keep_locks = 1; - ret = btrfs_search_forward(root, &found_key, &other_key, - path, 0, 0); - path->keep_locks = 0; - if (ret || found_key.objectid != key.objectid || - found_key.type != key.type) { - ret = 0; - goto out; - } - - key.offset = found_key.offset; - btrfs_release_path(root, path); - cond_resched(); - goto again; - } - - /* we know there's one more slot after us in the tree, - * read that key so we can verify it is also a checksum item - */ - btrfs_node_key_to_cpu(path->nodes[1], &other_key, path->slots[1] + 1); - - if (found_key.objectid < inode->i_ino) - goto next_key; - - if (found_key.type != key.type || found_key.offset < new_size) - goto next_key; - - /* - * if the key for the next leaf isn't a csum key from this objectid, - * we can't be sure there aren't good items inside this leaf. - * Bail out - */ - if (other_key.objectid != inode->i_ino || other_key.type != key.type) - goto out; - - leaf_start = btrfs_node_blockptr(path->nodes[1], path->slots[1]); - leaf_gen = btrfs_node_ptr_generation(path->nodes[1], path->slots[1]); - /* - * it is safe to delete this leaf, it contains only - * csum items from this inode at an offset >= new_size - */ - ret = btrfs_del_leaf(trans, root, path, leaf_start); - BUG_ON(ret); - - if (root->ref_cows && leaf_gen < trans->transid) { - ref = btrfs_alloc_leaf_ref(root, 0); - if (ref) { - ref->root_gen = root->root_key.offset; - ref->bytenr = leaf_start; - ref->owner = 0; - ref->generation = leaf_gen; - ref->nritems = 0; - - btrfs_sort_leaf_ref(ref); - - ret = btrfs_add_leaf_ref(root, ref, 0); - WARN_ON(ret); - btrfs_free_leaf_ref(root, ref); - } else { - WARN_ON(1); - } - } -next_key: - btrfs_release_path(root, path); - - if (other_key.objectid == inode->i_ino && - other_key.type == key.type && other_key.offset > key.offset) { - key.offset = other_key.offset; - cond_resched(); - goto again; - } - ret = 0; -out: - /* fixup any changes we've made to the path */ - path->lowest_level = 0; - path->keep_locks = 0; - btrfs_release_path(root, path); - return ret; -} - -#endif - /* * this can truncate away extent items, csum items and directory items. * It starts at a high offset and removes keys until it can't find @@ -3284,17 +3156,27 @@ int btrfs_truncate_inode_items(struct btrfs_trans_handle *trans, int encoding; int ret; int err = 0; + u64 ino = btrfs_ino(inode); BUG_ON(new_size > 0 && min_type != BTRFS_EXTENT_DATA_KEY); if (root->ref_cows || root == root->fs_info->tree_root) btrfs_drop_extent_cache(inode, new_size & (~mask), (u64)-1, 0); + /* + * This function is also used to drop the items in the log tree before + * we relog the inode, so if root != BTRFS_I(inode)->root, it means + * it is used to drop the loged items. So we shouldn't kill the delayed + * items. + */ + if (min_type == 0 && root == BTRFS_I(inode)->root) + btrfs_kill_delayed_inode_items(inode); + path = btrfs_alloc_path(); BUG_ON(!path); path->reada = -1; - key.objectid = inode->i_ino; + key.objectid = ino; key.offset = (u64)-1; key.type = (u8)-1; @@ -3322,7 +3204,7 @@ search_again: found_type = btrfs_key_type(&found_key); encoding = 0; - if (found_key.objectid != inode->i_ino) + if (found_key.objectid != ino) break; if (found_type < min_type) @@ -3412,7 +3294,6 @@ search_again: btrfs_file_extent_calc_inline_size(size); ret = btrfs_truncate_item(trans, root, path, size, 1); - BUG_ON(ret); } else if (root->ref_cows) { inode_sub_bytes(inode, item_end + 1 - found_key.offset); @@ -3441,7 +3322,7 @@ delete: ret = btrfs_free_extent(trans, root, extent_start, extent_num_bytes, 0, btrfs_header_owner(leaf), - inode->i_ino, extent_offset); + ino, extent_offset); BUG_ON(ret); } @@ -3450,7 +3331,9 @@ delete: if (path->slots[0] == 0 || path->slots[0] != pending_del_slot) { - if (root->ref_cows) { + if (root->ref_cows && + BTRFS_I(inode)->location.objectid != + BTRFS_FREE_INO_OBJECTID) { err = -EAGAIN; goto out; } @@ -3461,7 +3344,7 @@ delete: BUG_ON(ret); pending_del_nr = 0; } - btrfs_release_path(root, path); + btrfs_release_path(path); goto search_again; } else { path->slots[0]--; @@ -3619,7 +3502,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size) while (1) { em = btrfs_get_extent(inode, NULL, 0, cur_offset, block_end - cur_offset, 0); - BUG_ON(IS_ERR(em) || !em); + BUG_ON(IS_ERR_OR_NULL(em)); last_byte = min(extent_map_end(em), block_end); last_byte = (last_byte + mask) & ~mask; if (!test_bit(EXTENT_FLAG_PREALLOC, &em->flags)) { @@ -3640,7 +3523,7 @@ int btrfs_cont_expand(struct inode *inode, loff_t oldsize, loff_t size) break; err = btrfs_insert_file_extent(trans, root, - inode->i_ino, cur_offset, 0, + btrfs_ino(inode), cur_offset, 0, 0, hole_size, 0, hole_size, 0, 0, 0); if (err) @@ -3742,7 +3625,7 @@ void btrfs_evict_inode(struct inode *inode) truncate_inode_pages(&inode->i_data, 0); if (inode->i_nlink && (btrfs_root_refs(&root->root_item) != 0 || - root == root->fs_info->tree_root)) + is_free_space_inode(root, inode))) goto no_delete; if (is_bad_inode(inode)) { @@ -3795,6 +3678,10 @@ void btrfs_evict_inode(struct inode *inode) BUG_ON(ret); } + if (!(root == root->fs_info->tree_root || + root->root_key.objectid == BTRFS_TREE_RELOC_OBJECTID)) + btrfs_return_ino(root, btrfs_ino(inode)); + nr = trans->blocks_used; btrfs_end_transaction(trans, root); btrfs_btree_balance_dirty(root, nr); @@ -3820,12 +3707,12 @@ static int btrfs_inode_by_name(struct inode *dir, struct dentry *dentry, path = btrfs_alloc_path(); BUG_ON(!path); - di = btrfs_lookup_dir_item(NULL, root, path, dir->i_ino, name, + di = btrfs_lookup_dir_item(NULL, root, path, btrfs_ino(dir), name, namelen, 0); if (IS_ERR(di)) ret = PTR_ERR(di); - if (!di || IS_ERR(di)) + if (IS_ERR_OR_NULL(di)) goto out_err; btrfs_dir_item_key_to_cpu(path->nodes[0], di, location); @@ -3873,7 +3760,7 @@ static int fixup_tree_root_location(struct btrfs_root *root, leaf = path->nodes[0]; ref = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_root_ref); - if (btrfs_root_ref_dirid(leaf, ref) != dir->i_ino || + if (btrfs_root_ref_dirid(leaf, ref) != btrfs_ino(dir) || btrfs_root_ref_name_len(leaf, ref) != dentry->d_name.len) goto out; @@ -3883,7 +3770,7 @@ static int fixup_tree_root_location(struct btrfs_root *root, if (ret) goto out; - btrfs_release_path(root->fs_info->tree_root, path); + btrfs_release_path(path); new_root = btrfs_read_fs_root_no_name(root->fs_info, location); if (IS_ERR(new_root)) { @@ -3912,6 +3799,7 @@ static void inode_tree_add(struct inode *inode) struct btrfs_inode *entry; struct rb_node **p; struct rb_node *parent; + u64 ino = btrfs_ino(inode); again: p = &root->inode_tree.rb_node; parent = NULL; @@ -3924,9 +3812,9 @@ again: parent = *p; entry = rb_entry(parent, struct btrfs_inode, rb_node); - if (inode->i_ino < entry->vfs_inode.i_ino) + if (ino < btrfs_ino(&entry->vfs_inode)) p = &parent->rb_left; - else if (inode->i_ino > entry->vfs_inode.i_ino) + else if (ino > btrfs_ino(&entry->vfs_inode)) p = &parent->rb_right; else { WARN_ON(!(entry->vfs_inode.i_state & @@ -3990,9 +3878,9 @@ again: prev = node; entry = rb_entry(node, struct btrfs_inode, rb_node); - if (objectid < entry->vfs_inode.i_ino) + if (objectid < btrfs_ino(&entry->vfs_inode)) node = node->rb_left; - else if (objectid > entry->vfs_inode.i_ino) + else if (objectid > btrfs_ino(&entry->vfs_inode)) node = node->rb_right; else break; @@ -4000,7 +3888,7 @@ again: if (!node) { while (prev) { entry = rb_entry(prev, struct btrfs_inode, rb_node); - if (objectid <= entry->vfs_inode.i_ino) { + if (objectid <= btrfs_ino(&entry->vfs_inode)) { node = prev; break; } @@ -4009,7 +3897,7 @@ again: } while (node) { entry = rb_entry(node, struct btrfs_inode, rb_node); - objectid = entry->vfs_inode.i_ino + 1; + objectid = btrfs_ino(&entry->vfs_inode) + 1; inode = igrab(&entry->vfs_inode); if (inode) { spin_unlock(&root->inode_lock); @@ -4047,7 +3935,7 @@ static int btrfs_init_locked_inode(struct inode *inode, void *p) static int btrfs_find_actor(struct inode *inode, void *opaque) { struct btrfs_iget_args *args = opaque; - return args->ino == inode->i_ino && + return args->ino == btrfs_ino(inode) && args->root == BTRFS_I(inode)->root; } @@ -4192,7 +4080,7 @@ static struct dentry *btrfs_lookup(struct inode *dir, struct dentry *dentry, return d_splice_alias(inode, dentry); } -static unsigned char btrfs_filetype_table[] = { +unsigned char btrfs_filetype_table[] = { DT_UNKNOWN, DT_REG, DT_DIR, DT_CHR, DT_BLK, DT_FIFO, DT_SOCK, DT_LNK }; @@ -4206,11 +4094,11 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, struct btrfs_key key; struct btrfs_key found_key; struct btrfs_path *path; + struct list_head ins_list; + struct list_head del_list; int ret; - u32 nritems; struct extent_buffer *leaf; int slot; - int advance; unsigned char d_type; int over = 0; u32 di_cur; @@ -4220,6 +4108,7 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, char tmp_name[32]; char *name_ptr; int name_len; + int is_curr = 0; /* filp->f_pos points to the current index? */ /* FIXME, use a real flag for deciding about the key type */ if (root->fs_info->tree_root == root) @@ -4227,9 +4116,7 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, /* special case for "." */ if (filp->f_pos == 0) { - over = filldir(dirent, ".", 1, - 1, inode->i_ino, - DT_DIR); + over = filldir(dirent, ".", 1, 1, btrfs_ino(inode), DT_DIR); if (over) return 0; filp->f_pos = 1; @@ -4244,36 +4131,36 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, filp->f_pos = 2; } path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; path->reada = 2; + if (key_type == BTRFS_DIR_INDEX_KEY) { + INIT_LIST_HEAD(&ins_list); + INIT_LIST_HEAD(&del_list); + btrfs_get_delayed_items(inode, &ins_list, &del_list); + } + btrfs_set_key_type(&key, key_type); key.offset = filp->f_pos; - key.objectid = inode->i_ino; + key.objectid = btrfs_ino(inode); ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto err; - advance = 0; while (1) { leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); slot = path->slots[0]; - if (advance || slot >= nritems) { - if (slot >= nritems - 1) { - ret = btrfs_next_leaf(root, path); - if (ret) - break; - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - slot = path->slots[0]; - } else { - slot++; - path->slots[0]++; - } + if (slot >= btrfs_header_nritems(leaf)) { + ret = btrfs_next_leaf(root, path); + if (ret < 0) + goto err; + else if (ret > 0) + break; + continue; } - advance = 1; item = btrfs_item_nr(leaf, slot); btrfs_item_key_to_cpu(leaf, &found_key, slot); @@ -4282,9 +4169,14 @@ static int btrfs_real_readdir(struct file *filp, void *dirent, if (btrfs_key_type(&found_key) != key_type) break; if (found_key.offset < filp->f_pos) - continue; + goto next; + if (key_type == BTRFS_DIR_INDEX_KEY && + btrfs_should_delete_dir_index(&del_list, + found_key.offset)) + goto next; filp->f_pos = found_key.offset; + is_curr = 1; di = btrfs_item_ptr(leaf, slot, struct btrfs_dir_item); di_cur = 0; @@ -4335,6 +4227,17 @@ skip: di_cur += di_len; di = (struct btrfs_dir_item *)((char *)di + di_len); } +next: + path->slots[0]++; + } + + if (key_type == BTRFS_DIR_INDEX_KEY) { + if (is_curr) + filp->f_pos++; + ret = btrfs_readdir_delayed_dir_index(filp, dirent, filldir, + &ins_list); + if (ret) + goto nopos; } /* Reached end of directory/root. Bump pos past the last item. */ @@ -4349,6 +4252,8 @@ skip: nopos: ret = 0; err: + if (key_type == BTRFS_DIR_INDEX_KEY) + btrfs_put_delayed_items(&ins_list, &del_list); btrfs_free_path(path); return ret; } @@ -4364,7 +4269,8 @@ int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc) return 0; smp_mb(); - nolock = (root->fs_info->closing && root == root->fs_info->tree_root); + if (root->fs_info->closing && is_free_space_inode(root, inode)) + nolock = true; if (wbc->sync_mode == WB_SYNC_ALL) { if (nolock) @@ -4388,7 +4294,7 @@ int btrfs_write_inode(struct inode *inode, struct writeback_control *wbc) * FIXME, needs more benchmarking...there are no reasons other than performance * to keep or drop this code. */ -void btrfs_dirty_inode(struct inode *inode) +void btrfs_dirty_inode(struct inode *inode, int flags) { struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_trans_handle *trans; @@ -4407,25 +4313,25 @@ void btrfs_dirty_inode(struct inode *inode) btrfs_end_transaction(trans, root); trans = btrfs_start_transaction(root, 1); if (IS_ERR(trans)) { - if (printk_ratelimit()) { - printk(KERN_ERR "btrfs: fail to " - "dirty inode %lu error %ld\n", - inode->i_ino, PTR_ERR(trans)); - } + printk_ratelimited(KERN_ERR "btrfs: fail to " + "dirty inode %llu error %ld\n", + (unsigned long long)btrfs_ino(inode), + PTR_ERR(trans)); return; } btrfs_set_trans_block_group(trans, inode); ret = btrfs_update_inode(trans, root, inode); if (ret) { - if (printk_ratelimit()) { - printk(KERN_ERR "btrfs: fail to " - "dirty inode %lu error %d\n", - inode->i_ino, ret); - } + printk_ratelimited(KERN_ERR "btrfs: fail to " + "dirty inode %llu error %d\n", + (unsigned long long)btrfs_ino(inode), + ret); } } btrfs_end_transaction(trans, root); + if (BTRFS_I(inode)->delayed_node) + btrfs_balance_delayed_items(root); } /* @@ -4441,7 +4347,7 @@ static int btrfs_set_inode_index_count(struct inode *inode) struct extent_buffer *leaf; int ret; - key.objectid = inode->i_ino; + key.objectid = btrfs_ino(inode); btrfs_set_key_type(&key, BTRFS_DIR_INDEX_KEY); key.offset = (u64)-1; @@ -4473,7 +4379,7 @@ static int btrfs_set_inode_index_count(struct inode *inode) leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); - if (found_key.objectid != inode->i_ino || + if (found_key.objectid != btrfs_ino(inode) || btrfs_key_type(&found_key) != BTRFS_DIR_INDEX_KEY) { BTRFS_I(inode)->index_cnt = 2; goto out; @@ -4494,9 +4400,12 @@ int btrfs_set_inode_index(struct inode *dir, u64 *index) int ret = 0; if (BTRFS_I(dir)->index_cnt == (u64)-1) { - ret = btrfs_set_inode_index_count(dir); - if (ret) - return ret; + ret = btrfs_inode_delayed_dir_index_count(dir); + if (ret) { + ret = btrfs_set_inode_index_count(dir); + if (ret) + return ret; + } } *index = BTRFS_I(dir)->index_cnt; @@ -4527,14 +4436,23 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, BUG_ON(!path); inode = new_inode(root->fs_info->sb); - if (!inode) + if (!inode) { + btrfs_free_path(path); return ERR_PTR(-ENOMEM); + } + + /* + * we have to initialize this early, so we can reclaim the inode + * number if we fail afterwards in this function. + */ + inode->i_ino = objectid; if (dir) { trace_btrfs_inode_request(dir); ret = btrfs_set_inode_index(dir, index); if (ret) { + btrfs_free_path(path); iput(inode); return ERR_PTR(ret); } @@ -4574,7 +4492,6 @@ static struct inode *btrfs_new_inode(struct btrfs_trans_handle *trans, goto fail; inode_init_owner(inode, dir, mode); - inode->i_ino = objectid; inode_set_bytes(inode, 0); inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; inode_item = btrfs_item_ptr(path->nodes[0], path->slots[0], @@ -4638,29 +4555,29 @@ int btrfs_add_link(struct btrfs_trans_handle *trans, int ret = 0; struct btrfs_key key; struct btrfs_root *root = BTRFS_I(parent_inode)->root; + u64 ino = btrfs_ino(inode); + u64 parent_ino = btrfs_ino(parent_inode); - if (unlikely(inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) { + if (unlikely(ino == BTRFS_FIRST_FREE_OBJECTID)) { memcpy(&key, &BTRFS_I(inode)->root->root_key, sizeof(key)); } else { - key.objectid = inode->i_ino; + key.objectid = ino; btrfs_set_key_type(&key, BTRFS_INODE_ITEM_KEY); key.offset = 0; } - if (unlikely(inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) { + if (unlikely(ino == BTRFS_FIRST_FREE_OBJECTID)) { ret = btrfs_add_root_ref(trans, root->fs_info->tree_root, key.objectid, root->root_key.objectid, - parent_inode->i_ino, - index, name, name_len); + parent_ino, index, name, name_len); } else if (add_backref) { - ret = btrfs_insert_inode_ref(trans, root, - name, name_len, inode->i_ino, - parent_inode->i_ino, index); + ret = btrfs_insert_inode_ref(trans, root, name, name_len, ino, + parent_ino, index); } if (ret == 0) { ret = btrfs_insert_dir_item(trans, root, name, name_len, - parent_inode->i_ino, &key, + parent_inode, &key, btrfs_inode_type(inode), index); BUG_ON(ret); @@ -4703,10 +4620,6 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry, if (!new_valid_dev(rdev)) return -EINVAL; - err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid); - if (err) - return err; - /* * 2 for inode item and ref * 2 for dir items @@ -4718,12 +4631,17 @@ static int btrfs_mknod(struct inode *dir, struct dentry *dentry, btrfs_set_trans_block_group(trans, dir); + err = btrfs_find_free_ino(root, &objectid); + if (err) + goto out_unlock; + inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name, - dentry->d_name.len, dir->i_ino, objectid, + dentry->d_name.len, btrfs_ino(dir), objectid, BTRFS_I(dir)->block_group, mode, &index); - err = PTR_ERR(inode); - if (IS_ERR(inode)) + if (IS_ERR(inode)) { + err = PTR_ERR(inode); goto out_unlock; + } err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name); if (err) { @@ -4765,9 +4683,6 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry, u64 objectid; u64 index = 0; - err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid); - if (err) - return err; /* * 2 for inode item and ref * 2 for dir items @@ -4779,12 +4694,17 @@ static int btrfs_create(struct inode *dir, struct dentry *dentry, btrfs_set_trans_block_group(trans, dir); + err = btrfs_find_free_ino(root, &objectid); + if (err) + goto out_unlock; + inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name, - dentry->d_name.len, dir->i_ino, objectid, + dentry->d_name.len, btrfs_ino(dir), objectid, BTRFS_I(dir)->block_group, mode, &index); - err = PTR_ERR(inode); - if (IS_ERR(inode)) + if (IS_ERR(inode)) { + err = PTR_ERR(inode); goto out_unlock; + } err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name); if (err) { @@ -4834,9 +4754,6 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, if (inode->i_nlink == ~0U) return -EMLINK; - btrfs_inc_nlink(inode); - inode->i_ctime = CURRENT_TIME; - err = btrfs_set_inode_index(dir, &index); if (err) goto fail; @@ -4852,6 +4769,9 @@ static int btrfs_link(struct dentry *old_dentry, struct inode *dir, goto fail; } + btrfs_inc_nlink(inode); + inode->i_ctime = CURRENT_TIME; + btrfs_set_trans_block_group(trans, dir); ihold(inode); @@ -4890,10 +4810,6 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) u64 index = 0; unsigned long nr = 1; - err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid); - if (err) - return err; - /* * 2 items for inode and ref * 2 items for dir items @@ -4904,8 +4820,12 @@ static int btrfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) return PTR_ERR(trans); btrfs_set_trans_block_group(trans, dir); + err = btrfs_find_free_ino(root, &objectid); + if (err) + goto out_fail; + inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name, - dentry->d_name.len, dir->i_ino, objectid, + dentry->d_name.len, btrfs_ino(dir), objectid, BTRFS_I(dir)->block_group, S_IFDIR | mode, &index); if (IS_ERR(inode)) { @@ -4989,6 +4909,8 @@ static noinline int uncompress_inline(struct btrfs_path *path, inline_size = btrfs_file_extent_inline_item_len(leaf, btrfs_item_nr(leaf, path->slots[0])); tmp = kmalloc(inline_size, GFP_NOFS); + if (!tmp) + return -ENOMEM; ptr = btrfs_file_extent_inline_start(item); read_extent_buffer(leaf, tmp, ptr, inline_size); @@ -5026,7 +4948,7 @@ struct extent_map *btrfs_get_extent(struct inode *inode, struct page *page, u64 bytenr; u64 extent_start = 0; u64 extent_end = 0; - u64 objectid = inode->i_ino; + u64 objectid = btrfs_ino(inode); u32 found_type; struct btrfs_path *path = NULL; struct btrfs_root *root = BTRFS_I(inode)->root; @@ -5054,7 +4976,7 @@ again: else goto out; } - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); if (!em) { err = -ENOMEM; goto out; @@ -5208,7 +5130,7 @@ again: kunmap(page); free_extent_map(em); em = NULL; - btrfs_release_path(root, path); + btrfs_release_path(path); trans = btrfs_join_transaction(root, 1); if (IS_ERR(trans)) return ERR_CAST(trans); @@ -5221,7 +5143,7 @@ again: btrfs_mark_buffer_dirty(leaf); } set_extent_uptodate(io_tree, em->start, - extent_map_end(em) - 1, GFP_NOFS); + extent_map_end(em) - 1, NULL, GFP_NOFS); goto insert; } else { printk(KERN_ERR "btrfs unknown found_type %d\n", found_type); @@ -5234,7 +5156,7 @@ not_found_em: em->block_start = EXTENT_MAP_HOLE; set_bit(EXTENT_FLAG_VACANCY, &em->flags); insert: - btrfs_release_path(root, path); + btrfs_release_path(path); if (em->start > start || extent_map_end(em) <= start) { printk(KERN_ERR "Btrfs: bad extent! em: [%llu %llu] passed " "[%llu %llu]\n", (unsigned long long)em->start, @@ -5367,7 +5289,7 @@ struct extent_map *btrfs_get_extent_fiemap(struct inode *inode, struct page *pag u64 hole_start = start; u64 hole_len = len; - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); if (!em) { err = -ENOMEM; goto out; @@ -5428,22 +5350,38 @@ out: } static struct extent_map *btrfs_new_extent_direct(struct inode *inode, + struct extent_map *em, u64 start, u64 len) { struct btrfs_root *root = BTRFS_I(inode)->root; struct btrfs_trans_handle *trans; - struct extent_map *em; struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; struct btrfs_key ins; u64 alloc_hint; int ret; + bool insert = false; - btrfs_drop_extent_cache(inode, start, start + len - 1, 0); + /* + * Ok if the extent map we looked up is a hole and is for the exact + * range we want, there is no reason to allocate a new one, however if + * it is not right then we need to free this one and drop the cache for + * our range. + */ + if (em->block_start != EXTENT_MAP_HOLE || em->start != start || + em->len != len) { + free_extent_map(em); + em = NULL; + insert = true; + btrfs_drop_extent_cache(inode, start, start + len - 1, 0); + } trans = btrfs_join_transaction(root, 0); if (IS_ERR(trans)) return ERR_CAST(trans); + if (start <= BTRFS_I(inode)->disk_i_size && len < 64 * 1024) + btrfs_add_inode_defrag(trans, inode); + trans->block_rsv = &root->fs_info->delalloc_block_rsv; alloc_hint = get_extent_allocation_hint(inode, start, len); @@ -5454,10 +5392,12 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode, goto out; } - em = alloc_extent_map(GFP_NOFS); if (!em) { - em = ERR_PTR(-ENOMEM); - goto out; + em = alloc_extent_map(); + if (!em) { + em = ERR_PTR(-ENOMEM); + goto out; + } } em->start = start; @@ -5467,9 +5407,15 @@ static struct extent_map *btrfs_new_extent_direct(struct inode *inode, em->block_start = ins.objectid; em->block_len = ins.offset; em->bdev = root->fs_info->fs_devices->latest_bdev; + + /* + * We need to do this because if we're using the original em we searched + * for, we could have EXTENT_FLAG_VACANCY set, and we don't want that. + */ + em->flags = 0; set_bit(EXTENT_FLAG_PINNED, &em->flags); - while (1) { + while (insert) { write_lock(&em_tree->lock); ret = add_extent_mapping(em_tree, em); write_unlock(&em_tree->lock); @@ -5513,7 +5459,7 @@ static noinline int can_nocow_odirect(struct btrfs_trans_handle *trans, if (!path) return -ENOMEM; - ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino, + ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode), offset, 0); if (ret < 0) goto out; @@ -5530,7 +5476,7 @@ static noinline int can_nocow_odirect(struct btrfs_trans_handle *trans, ret = 0; leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, slot); - if (key.objectid != inode->i_ino || + if (key.objectid != btrfs_ino(inode) || key.type != BTRFS_EXTENT_DATA_KEY) { /* not our file or wrong item type, must cow */ goto out; @@ -5564,7 +5510,7 @@ static noinline int can_nocow_odirect(struct btrfs_trans_handle *trans, * look for other files referencing this extent, if we * find any we must cow */ - if (btrfs_cross_ref_exist(trans, root, inode->i_ino, + if (btrfs_cross_ref_exist(trans, root, btrfs_ino(inode), key.offset - backref_offset, disk_bytenr)) goto out; @@ -5687,8 +5633,7 @@ must_cow: * it above */ len = bh_result->b_size; - free_extent_map(em); - em = btrfs_new_extent_direct(inode, start, len); + em = btrfs_new_extent_direct(inode, em, start, len); if (IS_ERR(em)) return PTR_ERR(em); len = min(len, em->len - (start - em->start)); @@ -5755,9 +5700,10 @@ static void btrfs_endio_direct_read(struct bio *bio, int err) flush_dcache_page(bvec->bv_page); if (csum != *private) { - printk(KERN_ERR "btrfs csum failed ino %lu off" + printk(KERN_ERR "btrfs csum failed ino %llu off" " %llu csum %u private %u\n", - inode->i_ino, (unsigned long long)start, + (unsigned long long)btrfs_ino(inode), + (unsigned long long)start, csum, *private); err = -EIO; } @@ -5851,8 +5797,10 @@ again: } add_pending_csums(trans, inode, ordered->file_offset, &ordered->list); - btrfs_ordered_update_i_size(inode, 0, ordered); - btrfs_update_inode(trans, root, inode); + ret = btrfs_ordered_update_i_size(inode, 0, ordered); + if (!ret) + btrfs_update_inode(trans, root, inode); + ret = 0; out_unlock: unlock_extent_cached(&BTRFS_I(inode)->io_tree, ordered->file_offset, ordered->file_offset + ordered->len - 1, @@ -5902,9 +5850,9 @@ static void btrfs_end_dio_bio(struct bio *bio, int err) struct btrfs_dio_private *dip = bio->bi_private; if (err) { - printk(KERN_ERR "btrfs direct IO failed ino %lu rw %lu " + printk(KERN_ERR "btrfs direct IO failed ino %llu rw %lu " "sector %#Lx len %u err no %d\n", - dip->inode->i_ino, bio->bi_rw, + (unsigned long long)btrfs_ino(dip->inode), bio->bi_rw, (unsigned long long)bio->bi_sector, bio->bi_size, err); dip->errors = 1; @@ -5938,7 +5886,7 @@ static struct bio *btrfs_dio_bio_alloc(struct block_device *bdev, static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode, int rw, u64 file_offset, int skip_sum, - u32 *csums) + u32 *csums, int async_submit) { int write = rw & REQ_WRITE; struct btrfs_root *root = BTRFS_I(inode)->root; @@ -5949,13 +5897,24 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode, if (ret) goto err; - if (write && !skip_sum) { + if (skip_sum) + goto map; + + if (write && async_submit) { ret = btrfs_wq_submit_bio(root->fs_info, inode, rw, bio, 0, 0, file_offset, __btrfs_submit_bio_start_direct_io, __btrfs_submit_bio_done); goto err; + } else if (write) { + /* + * If we aren't doing async submit, calculate the csum of the + * bio now. + */ + ret = btrfs_csum_one_bio(root, inode, bio, file_offset, 1); + if (ret) + goto err; } else if (!skip_sum) { ret = btrfs_lookup_bio_sums_dio(root, inode, bio, file_offset, csums); @@ -5963,7 +5922,8 @@ static inline int __btrfs_submit_dio_bio(struct bio *bio, struct inode *inode, goto err; } - ret = btrfs_map_bio(root, rw, bio, 0, 1); +map: + ret = btrfs_map_bio(root, rw, bio, 0, async_submit); err: bio_put(bio); return ret; @@ -5985,23 +5945,30 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, int nr_pages = 0; u32 *csums = dip->csums; int ret = 0; + int async_submit = 0; int write = rw & REQ_WRITE; - bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS); - if (!bio) - return -ENOMEM; - bio->bi_private = dip; - bio->bi_end_io = btrfs_end_dio_bio; - atomic_inc(&dip->pending_bios); - map_length = orig_bio->bi_size; ret = btrfs_map_block(map_tree, READ, start_sector << 9, &map_length, NULL, 0); if (ret) { - bio_put(bio); + bio_put(orig_bio); return -EIO; } + if (map_length >= orig_bio->bi_size) { + bio = orig_bio; + goto submit; + } + + async_submit = 1; + bio = btrfs_dio_bio_alloc(orig_bio->bi_bdev, start_sector, GFP_NOFS); + if (!bio) + return -ENOMEM; + bio->bi_private = dip; + bio->bi_end_io = btrfs_end_dio_bio; + atomic_inc(&dip->pending_bios); + while (bvec <= (orig_bio->bi_io_vec + orig_bio->bi_vcnt - 1)) { if (unlikely(map_length < submit_len + bvec->bv_len || bio_add_page(bio, bvec->bv_page, bvec->bv_len, @@ -6015,7 +5982,7 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, atomic_inc(&dip->pending_bios); ret = __btrfs_submit_dio_bio(bio, inode, rw, file_offset, skip_sum, - csums); + csums, async_submit); if (ret) { bio_put(bio); atomic_dec(&dip->pending_bios); @@ -6052,8 +6019,9 @@ static int btrfs_submit_direct_hook(int rw, struct btrfs_dio_private *dip, } } +submit: ret = __btrfs_submit_dio_bio(bio, inode, rw, file_offset, skip_sum, - csums); + csums, async_submit); if (!ret) return 0; @@ -6148,6 +6116,7 @@ static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *io unsigned long nr_segs) { int seg; + int i; size_t size; unsigned long addr; unsigned blocksize_mask = root->sectorsize - 1; @@ -6162,8 +6131,22 @@ static ssize_t check_direct_IO(struct btrfs_root *root, int rw, struct kiocb *io addr = (unsigned long)iov[seg].iov_base; size = iov[seg].iov_len; end += size; - if ((addr & blocksize_mask) || (size & blocksize_mask)) + if ((addr & blocksize_mask) || (size & blocksize_mask)) goto out; + + /* If this is a write we don't need to check anymore */ + if (rw & WRITE) + continue; + + /* + * Check to make sure we don't have duplicate iov_base's in this + * iovec, if so return EINVAL, otherwise we'll get csum errors + * when reading back. + */ + for (i = seg + 1; i < nr_segs; i++) { + if (iov[seg].iov_base == iov[i].iov_base) + goto out; + } } retval = 0; out: @@ -6710,12 +6693,15 @@ struct inode *btrfs_alloc_inode(struct super_block *sb) ei->ordered_data_close = 0; ei->orphan_meta_reserved = 0; ei->dummy_inode = 0; + ei->in_defrag = 0; ei->force_compress = BTRFS_COMPRESS_NONE; + ei->delayed_node = NULL; + inode = &ei->vfs_inode; - extent_map_tree_init(&ei->extent_tree, GFP_NOFS); - extent_io_tree_init(&ei->io_tree, &inode->i_data, GFP_NOFS); - extent_io_tree_init(&ei->io_failure_tree, &inode->i_data, GFP_NOFS); + extent_map_tree_init(&ei->extent_tree); + extent_io_tree_init(&ei->io_tree, &inode->i_data); + extent_io_tree_init(&ei->io_failure_tree, &inode->i_data); mutex_init(&ei->log_mutex); btrfs_ordered_inode_tree_init(&ei->ordered_tree); INIT_LIST_HEAD(&ei->i_orphan); @@ -6779,8 +6765,8 @@ void btrfs_destroy_inode(struct inode *inode) spin_lock(&root->orphan_lock); if (!list_empty(&BTRFS_I(inode)->i_orphan)) { - printk(KERN_INFO "BTRFS: inode %lu still on the orphan list\n", - inode->i_ino); + printk(KERN_INFO "BTRFS: inode %llu still on the orphan list\n", + (unsigned long long)btrfs_ino(inode)); list_del_init(&BTRFS_I(inode)->i_orphan); } spin_unlock(&root->orphan_lock); @@ -6802,6 +6788,7 @@ void btrfs_destroy_inode(struct inode *inode) inode_tree_del(inode); btrfs_drop_extent_cache(inode, 0, (u64)-1, 0); free: + btrfs_remove_delayed_node(inode); call_rcu(&inode->i_rcu, btrfs_i_callback); } @@ -6810,7 +6797,7 @@ int btrfs_drop_inode(struct inode *inode) struct btrfs_root *root = BTRFS_I(inode)->root; if (btrfs_root_refs(&root->root_item) == 0 && - root != root->fs_info->tree_root) + !is_free_space_inode(root, inode)) return 1; else return generic_drop_inode(inode); @@ -6919,16 +6906,17 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, u64 index = 0; u64 root_objectid; int ret; + u64 old_ino = btrfs_ino(old_inode); - if (new_dir->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID) + if (btrfs_ino(new_dir) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID) return -EPERM; /* we only allow rename subvolume link between subvolumes */ - if (old_inode->i_ino != BTRFS_FIRST_FREE_OBJECTID && root != dest) + if (old_ino != BTRFS_FIRST_FREE_OBJECTID && root != dest) return -EXDEV; - if (old_inode->i_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID || - (new_inode && new_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) + if (old_ino == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID || + (new_inode && btrfs_ino(new_inode) == BTRFS_FIRST_FREE_OBJECTID)) return -ENOTEMPTY; if (S_ISDIR(old_inode->i_mode) && new_inode && @@ -6944,7 +6932,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, filemap_flush(old_inode->i_mapping); /* close the racy window with snapshot create/destroy ioctl */ - if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) + if (old_ino == BTRFS_FIRST_FREE_OBJECTID) down_read(&root->fs_info->subvol_sem); /* * We want to reserve the absolute worst case amount of items. So if @@ -6969,15 +6957,15 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (ret) goto out_fail; - if (unlikely(old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) { + if (unlikely(old_ino == BTRFS_FIRST_FREE_OBJECTID)) { /* force full log commit if subvolume involved. */ root->fs_info->last_trans_log_full_commit = trans->transid; } else { ret = btrfs_insert_inode_ref(trans, dest, new_dentry->d_name.name, new_dentry->d_name.len, - old_inode->i_ino, - new_dir->i_ino, index); + old_ino, + btrfs_ino(new_dir), index); if (ret) goto out_fail; /* @@ -6993,10 +6981,8 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, * make sure the inode gets flushed if it is replacing * something. */ - if (new_inode && new_inode->i_size && - old_inode && S_ISREG(old_inode->i_mode)) { + if (new_inode && new_inode->i_size && S_ISREG(old_inode->i_mode)) btrfs_add_ordered_operation(trans, root, old_inode); - } old_dir->i_ctime = old_dir->i_mtime = ctime; new_dir->i_ctime = new_dir->i_mtime = ctime; @@ -7005,7 +6991,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (old_dentry->d_parent != new_dentry->d_parent) btrfs_record_unlink_dir(trans, old_dir, old_inode, 1); - if (unlikely(old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID)) { + if (unlikely(old_ino == BTRFS_FIRST_FREE_OBJECTID)) { root_objectid = BTRFS_I(old_inode)->root->root_key.objectid; ret = btrfs_unlink_subvol(trans, root, old_dir, root_objectid, old_dentry->d_name.name, @@ -7022,7 +7008,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (new_inode) { new_inode->i_ctime = CURRENT_TIME; - if (unlikely(new_inode->i_ino == + if (unlikely(btrfs_ino(new_inode) == BTRFS_EMPTY_SUBVOL_DIR_OBJECTID)) { root_objectid = BTRFS_I(new_inode)->location.objectid; ret = btrfs_unlink_subvol(trans, dest, new_dir, @@ -7050,7 +7036,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, new_dentry->d_name.len, 0, index); BUG_ON(ret); - if (old_inode->i_ino != BTRFS_FIRST_FREE_OBJECTID) { + if (old_ino != BTRFS_FIRST_FREE_OBJECTID) { struct dentry *parent = dget_parent(new_dentry); btrfs_log_new_name(trans, old_inode, old_dir, parent); dput(parent); @@ -7059,7 +7045,7 @@ static int btrfs_rename(struct inode *old_dir, struct dentry *old_dentry, out_fail: btrfs_end_transaction_throttle(trans, root); out_notrans: - if (old_inode->i_ino == BTRFS_FIRST_FREE_OBJECTID) + if (old_ino == BTRFS_FIRST_FREE_OBJECTID) up_read(&root->fs_info->subvol_sem); return ret; @@ -7113,58 +7099,6 @@ int btrfs_start_delalloc_inodes(struct btrfs_root *root, int delay_iput) return 0; } -int btrfs_start_one_delalloc_inode(struct btrfs_root *root, int delay_iput, - int sync) -{ - struct btrfs_inode *binode; - struct inode *inode = NULL; - - spin_lock(&root->fs_info->delalloc_lock); - while (!list_empty(&root->fs_info->delalloc_inodes)) { - binode = list_entry(root->fs_info->delalloc_inodes.next, - struct btrfs_inode, delalloc_inodes); - inode = igrab(&binode->vfs_inode); - if (inode) { - list_move_tail(&binode->delalloc_inodes, - &root->fs_info->delalloc_inodes); - break; - } - - list_del_init(&binode->delalloc_inodes); - cond_resched_lock(&root->fs_info->delalloc_lock); - } - spin_unlock(&root->fs_info->delalloc_lock); - - if (inode) { - if (sync) { - filemap_write_and_wait(inode->i_mapping); - /* - * We have to do this because compression doesn't - * actually set PG_writeback until it submits the pages - * for IO, which happens in an async thread, so we could - * race and not actually wait for any writeback pages - * because they've not been submitted yet. Technically - * this could still be the case for the ordered stuff - * since the async thread may not have started to do its - * work yet. If this becomes the case then we need to - * figure out a way to make sure that in writepage we - * wait for any async pages to be submitted before - * returning so that fdatawait does what its supposed to - * do. - */ - btrfs_wait_ordered_range(inode, 0, (u64)-1); - } else { - filemap_flush(inode->i_mapping); - } - if (delay_iput) - btrfs_add_delayed_iput(inode); - else - iput(inode); - return 1; - } - return 0; -} - static int btrfs_symlink(struct inode *dir, struct dentry *dentry, const char *symname) { @@ -7188,9 +7122,6 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry, if (name_len > BTRFS_MAX_INLINE_DATA_SIZE(root)) return -ENAMETOOLONG; - err = btrfs_find_free_objectid(NULL, root, dir->i_ino, &objectid); - if (err) - return err; /* * 2 items for inode item and ref * 2 items for dir items @@ -7202,13 +7133,18 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry, btrfs_set_trans_block_group(trans, dir); + err = btrfs_find_free_ino(root, &objectid); + if (err) + goto out_unlock; + inode = btrfs_new_inode(trans, root, dir, dentry->d_name.name, - dentry->d_name.len, dir->i_ino, objectid, + dentry->d_name.len, btrfs_ino(dir), objectid, BTRFS_I(dir)->block_group, S_IFLNK|S_IRWXUGO, &index); - err = PTR_ERR(inode); - if (IS_ERR(inode)) + if (IS_ERR(inode)) { + err = PTR_ERR(inode); goto out_unlock; + } err = btrfs_init_inode_security(trans, inode, dir, &dentry->d_name); if (err) { @@ -7234,7 +7170,7 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry, path = btrfs_alloc_path(); BUG_ON(!path); - key.objectid = inode->i_ino; + key.objectid = btrfs_ino(inode); key.offset = 0; btrfs_set_key_type(&key, BTRFS_EXTENT_DATA_KEY); datasize = btrfs_file_extent_calc_inline_size(name_len); @@ -7242,6 +7178,7 @@ static int btrfs_symlink(struct inode *dir, struct dentry *dentry, datasize); if (err) { drop_inode = 1; + btrfs_free_path(path); goto out_unlock; } leaf = path->nodes[0]; diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index cfc264fefdb0..85e818ce00c5 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -50,6 +50,7 @@ #include "print-tree.h" #include "volumes.h" #include "locking.h" +#include "inode-map.h" /* Mask out flags that are inappropriate for the given type of inode. */ static inline __u32 btrfs_mask_flags(umode_t mode, __u32 flags) @@ -81,6 +82,13 @@ static unsigned int btrfs_flags_to_ioctl(unsigned int flags) iflags |= FS_NOATIME_FL; if (flags & BTRFS_INODE_DIRSYNC) iflags |= FS_DIRSYNC_FL; + if (flags & BTRFS_INODE_NODATACOW) + iflags |= FS_NOCOW_FL; + + if ((flags & BTRFS_INODE_COMPRESS) && !(flags & BTRFS_INODE_NOCOMPRESS)) + iflags |= FS_COMPR_FL; + else if (flags & BTRFS_INODE_NOCOMPRESS) + iflags |= FS_NOCOMP_FL; return iflags; } @@ -144,16 +152,13 @@ static int check_flags(unsigned int flags) if (flags & ~(FS_IMMUTABLE_FL | FS_APPEND_FL | \ FS_NOATIME_FL | FS_NODUMP_FL | \ FS_SYNC_FL | FS_DIRSYNC_FL | \ - FS_NOCOMP_FL | FS_COMPR_FL | \ - FS_NOCOW_FL | FS_COW_FL)) + FS_NOCOMP_FL | FS_COMPR_FL | + FS_NOCOW_FL)) return -EOPNOTSUPP; if ((flags & FS_NOCOMP_FL) && (flags & FS_COMPR_FL)) return -EINVAL; - if ((flags & FS_NOCOW_FL) && (flags & FS_COW_FL)) - return -EINVAL; - return 0; } @@ -218,6 +223,10 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) ip->flags |= BTRFS_INODE_DIRSYNC; else ip->flags &= ~BTRFS_INODE_DIRSYNC; + if (flags & FS_NOCOW_FL) + ip->flags |= BTRFS_INODE_NODATACOW; + else + ip->flags &= ~BTRFS_INODE_NODATACOW; /* * The COMPRESS flag can only be changed by users, while the NOCOMPRESS @@ -230,11 +239,9 @@ static int btrfs_ioctl_setflags(struct file *file, void __user *arg) } else if (flags & FS_COMPR_FL) { ip->flags |= BTRFS_INODE_COMPRESS; ip->flags &= ~BTRFS_INODE_NOCOMPRESS; + } else { + ip->flags &= ~(BTRFS_INODE_COMPRESS | BTRFS_INODE_NOCOMPRESS); } - if (flags & FS_NOCOW_FL) - ip->flags |= BTRFS_INODE_NODATACOW; - else if (flags & FS_COW_FL) - ip->flags &= ~BTRFS_INODE_NODATACOW; trans = btrfs_join_transaction(root, 1); BUG_ON(IS_ERR(trans)); @@ -275,8 +282,9 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg) if (!capable(CAP_SYS_ADMIN)) return -EPERM; - mutex_lock(&fs_info->fs_devices->device_list_mutex); - list_for_each_entry(device, &fs_info->fs_devices->devices, dev_list) { + rcu_read_lock(); + list_for_each_entry_rcu(device, &fs_info->fs_devices->devices, + dev_list) { if (!device->bdev) continue; q = bdev_get_queue(device->bdev); @@ -286,7 +294,7 @@ static noinline int btrfs_ioctl_fitrim(struct file *file, void __user *arg) minlen); } } - mutex_unlock(&fs_info->fs_devices->device_list_mutex); + rcu_read_unlock(); if (!num_devices) return -EOPNOTSUPP; @@ -323,8 +331,7 @@ static noinline int create_subvol(struct btrfs_root *root, u64 new_dirid = BTRFS_FIRST_FREE_OBJECTID; u64 index = 0; - ret = btrfs_find_free_objectid(NULL, root->fs_info->tree_root, - 0, &objectid); + ret = btrfs_find_free_objectid(root->fs_info->tree_root, &objectid); if (ret) { dput(parent); return ret; @@ -416,7 +423,7 @@ static noinline int create_subvol(struct btrfs_root *root, BUG_ON(ret); ret = btrfs_insert_dir_item(trans, root, - name, namelen, dir->i_ino, &key, + name, namelen, dir, &key, BTRFS_FT_DIR, index); if (ret) goto fail; @@ -427,7 +434,7 @@ static noinline int create_subvol(struct btrfs_root *root, ret = btrfs_add_root_ref(trans, root->fs_info->tree_root, objectid, root->root_key.objectid, - dir->i_ino, index, name, namelen); + btrfs_ino(dir), index, name, namelen); BUG_ON(ret); @@ -649,6 +656,106 @@ out_unlock: return error; } +/* + * When we're defragging a range, we don't want to kick it off again + * if it is really just waiting for delalloc to send it down. + * If we find a nice big extent or delalloc range for the bytes in the + * file you want to defrag, we return 0 to let you know to skip this + * part of the file + */ +static int check_defrag_in_cache(struct inode *inode, u64 offset, int thresh) +{ + struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; + struct extent_map *em = NULL; + struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; + u64 end; + + read_lock(&em_tree->lock); + em = lookup_extent_mapping(em_tree, offset, PAGE_CACHE_SIZE); + read_unlock(&em_tree->lock); + + if (em) { + end = extent_map_end(em); + free_extent_map(em); + if (end - offset > thresh) + return 0; + } + /* if we already have a nice delalloc here, just stop */ + thresh /= 2; + end = count_range_bits(io_tree, &offset, offset + thresh, + thresh, EXTENT_DELALLOC, 1); + if (end >= thresh) + return 0; + return 1; +} + +/* + * helper function to walk through a file and find extents + * newer than a specific transid, and smaller than thresh. + * + * This is used by the defragging code to find new and small + * extents + */ +static int find_new_extents(struct btrfs_root *root, + struct inode *inode, u64 newer_than, + u64 *off, int thresh) +{ + struct btrfs_path *path; + struct btrfs_key min_key; + struct btrfs_key max_key; + struct extent_buffer *leaf; + struct btrfs_file_extent_item *extent; + int type; + int ret; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + min_key.objectid = inode->i_ino; + min_key.type = BTRFS_EXTENT_DATA_KEY; + min_key.offset = *off; + + max_key.objectid = inode->i_ino; + max_key.type = (u8)-1; + max_key.offset = (u64)-1; + + path->keep_locks = 1; + + while(1) { + ret = btrfs_search_forward(root, &min_key, &max_key, + path, 0, newer_than); + if (ret != 0) + goto none; + if (min_key.objectid != inode->i_ino) + goto none; + if (min_key.type != BTRFS_EXTENT_DATA_KEY) + goto none; + + leaf = path->nodes[0]; + extent = btrfs_item_ptr(leaf, path->slots[0], + struct btrfs_file_extent_item); + + type = btrfs_file_extent_type(leaf, extent); + if (type == BTRFS_FILE_EXTENT_REG && + btrfs_file_extent_num_bytes(leaf, extent) < thresh && + check_defrag_in_cache(inode, min_key.offset, thresh)) { + *off = min_key.offset; + btrfs_free_path(path); + return 0; + } + + if (min_key.offset == (u64)-1) + goto none; + + min_key.offset++; + btrfs_release_path(path); + } +none: + btrfs_free_path(path); + return -ENOENT; +} + static int should_defrag_range(struct inode *inode, u64 start, u64 len, int thresh, u64 *last_len, u64 *skip, u64 *defrag_end) @@ -658,10 +765,6 @@ static int should_defrag_range(struct inode *inode, u64 start, u64 len, struct extent_map_tree *em_tree = &BTRFS_I(inode)->extent_tree; int ret = 1; - - if (thresh == 0) - thresh = 256 * 1024; - /* * make sure that once we start defragging and extent, we keep on * defragging it @@ -720,27 +823,176 @@ static int should_defrag_range(struct inode *inode, u64 start, u64 len, return ret; } -static int btrfs_defrag_file(struct file *file, - struct btrfs_ioctl_defrag_range_args *range) +/* + * it doesn't do much good to defrag one or two pages + * at a time. This pulls in a nice chunk of pages + * to COW and defrag. + * + * It also makes sure the delalloc code has enough + * dirty data to avoid making new small extents as part + * of the defrag + * + * It's a good idea to start RA on this range + * before calling this. + */ +static int cluster_pages_for_defrag(struct inode *inode, + struct page **pages, + unsigned long start_index, + int num_pages) +{ + unsigned long file_end; + u64 isize = i_size_read(inode); + u64 page_start; + u64 page_end; + int ret; + int i; + int i_done; + struct btrfs_ordered_extent *ordered; + struct extent_state *cached_state = NULL; + + if (isize == 0) + return 0; + file_end = (isize - 1) >> PAGE_CACHE_SHIFT; + + ret = btrfs_delalloc_reserve_space(inode, + num_pages << PAGE_CACHE_SHIFT); + if (ret) + return ret; +again: + ret = 0; + i_done = 0; + + /* step one, lock all the pages */ + for (i = 0; i < num_pages; i++) { + struct page *page; + page = grab_cache_page(inode->i_mapping, + start_index + i); + if (!page) + break; + + if (!PageUptodate(page)) { + btrfs_readpage(NULL, page); + lock_page(page); + if (!PageUptodate(page)) { + unlock_page(page); + page_cache_release(page); + ret = -EIO; + break; + } + } + isize = i_size_read(inode); + file_end = (isize - 1) >> PAGE_CACHE_SHIFT; + if (!isize || page->index > file_end || + page->mapping != inode->i_mapping) { + /* whoops, we blew past eof, skip this page */ + unlock_page(page); + page_cache_release(page); + break; + } + pages[i] = page; + i_done++; + } + if (!i_done || ret) + goto out; + + if (!(inode->i_sb->s_flags & MS_ACTIVE)) + goto out; + + /* + * so now we have a nice long stream of locked + * and up to date pages, lets wait on them + */ + for (i = 0; i < i_done; i++) + wait_on_page_writeback(pages[i]); + + page_start = page_offset(pages[0]); + page_end = page_offset(pages[i_done - 1]) + PAGE_CACHE_SIZE; + + lock_extent_bits(&BTRFS_I(inode)->io_tree, + page_start, page_end - 1, 0, &cached_state, + GFP_NOFS); + ordered = btrfs_lookup_first_ordered_extent(inode, page_end - 1); + if (ordered && + ordered->file_offset + ordered->len > page_start && + ordered->file_offset < page_end) { + btrfs_put_ordered_extent(ordered); + unlock_extent_cached(&BTRFS_I(inode)->io_tree, + page_start, page_end - 1, + &cached_state, GFP_NOFS); + for (i = 0; i < i_done; i++) { + unlock_page(pages[i]); + page_cache_release(pages[i]); + } + btrfs_wait_ordered_range(inode, page_start, + page_end - page_start); + goto again; + } + if (ordered) + btrfs_put_ordered_extent(ordered); + + clear_extent_bit(&BTRFS_I(inode)->io_tree, page_start, + page_end - 1, EXTENT_DIRTY | EXTENT_DELALLOC | + EXTENT_DO_ACCOUNTING, 0, 0, &cached_state, + GFP_NOFS); + + if (i_done != num_pages) { + atomic_inc(&BTRFS_I(inode)->outstanding_extents); + btrfs_delalloc_release_space(inode, + (num_pages - i_done) << PAGE_CACHE_SHIFT); + } + + + btrfs_set_extent_delalloc(inode, page_start, page_end - 1, + &cached_state); + + unlock_extent_cached(&BTRFS_I(inode)->io_tree, + page_start, page_end - 1, &cached_state, + GFP_NOFS); + + for (i = 0; i < i_done; i++) { + clear_page_dirty_for_io(pages[i]); + ClearPageChecked(pages[i]); + set_page_extent_mapped(pages[i]); + set_page_dirty(pages[i]); + unlock_page(pages[i]); + page_cache_release(pages[i]); + } + return i_done; +out: + for (i = 0; i < i_done; i++) { + unlock_page(pages[i]); + page_cache_release(pages[i]); + } + btrfs_delalloc_release_space(inode, num_pages << PAGE_CACHE_SHIFT); + return ret; + +} + +int btrfs_defrag_file(struct inode *inode, struct file *file, + struct btrfs_ioctl_defrag_range_args *range, + u64 newer_than, unsigned long max_to_defrag) { - struct inode *inode = fdentry(file)->d_inode; struct btrfs_root *root = BTRFS_I(inode)->root; - struct extent_io_tree *io_tree = &BTRFS_I(inode)->io_tree; - struct btrfs_ordered_extent *ordered; - struct page *page; struct btrfs_super_block *disk_super; + struct file_ra_state *ra = NULL; unsigned long last_index; - unsigned long ra_pages = root->fs_info->bdi.ra_pages; - unsigned long total_read = 0; u64 features; - u64 page_start; - u64 page_end; u64 last_len = 0; u64 skip = 0; u64 defrag_end = 0; + u64 newer_off = range->start; + int newer_left = 0; unsigned long i; int ret; + int defrag_count = 0; int compress_type = BTRFS_COMPRESS_ZLIB; + int extent_thresh = range->extent_thresh; + int newer_cluster = (256 * 1024) >> PAGE_CACHE_SHIFT; + u64 new_align = ~((u64)128 * 1024 - 1); + struct page **pages = NULL; + + if (extent_thresh == 0) + extent_thresh = 256 * 1024; if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) { if (range->compress_type > BTRFS_COMPRESS_TYPES) @@ -752,6 +1004,27 @@ static int btrfs_defrag_file(struct file *file, if (inode->i_size == 0) return 0; + /* + * if we were not given a file, allocate a readahead + * context + */ + if (!file) { + ra = kzalloc(sizeof(*ra), GFP_NOFS); + if (!ra) + return -ENOMEM; + file_ra_state_init(ra, inode->i_mapping); + } else { + ra = &file->f_ra; + } + + pages = kmalloc(sizeof(struct page *) * newer_cluster, + GFP_NOFS); + if (!pages) { + ret = -ENOMEM; + goto out_ra; + } + + /* find the last page to defrag */ if (range->start + range->len > range->start) { last_index = min_t(u64, inode->i_size - 1, range->start + range->len - 1) >> PAGE_CACHE_SHIFT; @@ -759,11 +1032,37 @@ static int btrfs_defrag_file(struct file *file, last_index = (inode->i_size - 1) >> PAGE_CACHE_SHIFT; } - i = range->start >> PAGE_CACHE_SHIFT; - while (i <= last_index) { - if (!should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT, + if (newer_than) { + ret = find_new_extents(root, inode, newer_than, + &newer_off, 64 * 1024); + if (!ret) { + range->start = newer_off; + /* + * we always align our defrag to help keep + * the extents in the file evenly spaced + */ + i = (newer_off & new_align) >> PAGE_CACHE_SHIFT; + newer_left = newer_cluster; + } else + goto out_ra; + } else { + i = range->start >> PAGE_CACHE_SHIFT; + } + if (!max_to_defrag) + max_to_defrag = last_index - 1; + + while (i <= last_index && defrag_count < max_to_defrag) { + /* + * make sure we stop running if someone unmounts + * the FS + */ + if (!(inode->i_sb->s_flags & MS_ACTIVE)) + break; + + if (!newer_than && + !should_defrag_range(inode, (u64)i << PAGE_CACHE_SHIFT, PAGE_CACHE_SIZE, - range->extent_thresh, + extent_thresh, &last_len, &skip, &defrag_end)) { unsigned long next; @@ -775,92 +1074,39 @@ static int btrfs_defrag_file(struct file *file, i = max(i + 1, next); continue; } - - if (total_read % ra_pages == 0) { - btrfs_force_ra(inode->i_mapping, &file->f_ra, file, i, - min(last_index, i + ra_pages - 1)); - } - total_read++; - mutex_lock(&inode->i_mutex); if (range->flags & BTRFS_DEFRAG_RANGE_COMPRESS) BTRFS_I(inode)->force_compress = compress_type; - ret = btrfs_delalloc_reserve_space(inode, PAGE_CACHE_SIZE); - if (ret) - goto err_unlock; -again: - if (inode->i_size == 0 || - i > ((inode->i_size - 1) >> PAGE_CACHE_SHIFT)) { - ret = 0; - goto err_reservations; - } - - page = grab_cache_page(inode->i_mapping, i); - if (!page) { - ret = -ENOMEM; - goto err_reservations; - } - - if (!PageUptodate(page)) { - btrfs_readpage(NULL, page); - lock_page(page); - if (!PageUptodate(page)) { - unlock_page(page); - page_cache_release(page); - ret = -EIO; - goto err_reservations; - } - } - - if (page->mapping != inode->i_mapping) { - unlock_page(page); - page_cache_release(page); - goto again; - } + btrfs_force_ra(inode->i_mapping, ra, file, i, newer_cluster); - wait_on_page_writeback(page); + ret = cluster_pages_for_defrag(inode, pages, i, newer_cluster); + if (ret < 0) + goto out_ra; - if (PageDirty(page)) { - btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE); - goto loop_unlock; - } + defrag_count += ret; + balance_dirty_pages_ratelimited_nr(inode->i_mapping, ret); + i += ret; - page_start = (u64)page->index << PAGE_CACHE_SHIFT; - page_end = page_start + PAGE_CACHE_SIZE - 1; - lock_extent(io_tree, page_start, page_end, GFP_NOFS); + if (newer_than) { + if (newer_off == (u64)-1) + break; - ordered = btrfs_lookup_ordered_extent(inode, page_start); - if (ordered) { - unlock_extent(io_tree, page_start, page_end, GFP_NOFS); - unlock_page(page); - page_cache_release(page); - btrfs_start_ordered_extent(inode, ordered, 1); - btrfs_put_ordered_extent(ordered); - goto again; + newer_off = max(newer_off + 1, + (u64)i << PAGE_CACHE_SHIFT); + + ret = find_new_extents(root, inode, + newer_than, &newer_off, + 64 * 1024); + if (!ret) { + range->start = newer_off; + i = (newer_off & new_align) >> PAGE_CACHE_SHIFT; + newer_left = newer_cluster; + } else { + break; + } + } else { + i++; } - set_page_extent_mapped(page); - - /* - * this makes sure page_mkwrite is called on the - * page if it is dirtied again later - */ - clear_page_dirty_for_io(page); - clear_extent_bits(&BTRFS_I(inode)->io_tree, page_start, - page_end, EXTENT_DIRTY | EXTENT_DELALLOC | - EXTENT_DO_ACCOUNTING, GFP_NOFS); - - btrfs_set_extent_delalloc(inode, page_start, page_end, NULL); - ClearPageChecked(page); - set_page_dirty(page); - unlock_extent(io_tree, page_start, page_end, GFP_NOFS); - -loop_unlock: - unlock_page(page); - page_cache_release(page); - mutex_unlock(&inode->i_mutex); - - balance_dirty_pages_ratelimited_nr(inode->i_mapping, 1); - i++; } if ((range->flags & BTRFS_DEFRAG_RANGE_START_IO)) @@ -892,12 +1138,14 @@ loop_unlock: btrfs_set_super_incompat_flags(disk_super, features); } - return 0; + if (!file) + kfree(ra); + return defrag_count; -err_reservations: - btrfs_delalloc_release_space(inode, PAGE_CACHE_SIZE); -err_unlock: - mutex_unlock(&inode->i_mutex); +out_ra: + if (!file) + kfree(ra); + kfree(pages); return ret; } @@ -1123,7 +1371,7 @@ static noinline int btrfs_ioctl_subvol_getflags(struct file *file, int ret = 0; u64 flags = 0; - if (inode->i_ino != BTRFS_FIRST_FREE_OBJECTID) + if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) return -EINVAL; down_read(&root->fs_info->subvol_sem); @@ -1150,7 +1398,7 @@ static noinline int btrfs_ioctl_subvol_setflags(struct file *file, if (root->fs_info->sb->s_flags & MS_RDONLY) return -EROFS; - if (inode->i_ino != BTRFS_FIRST_FREE_OBJECTID) + if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) return -EINVAL; if (copy_from_user(&flags, arg, sizeof(flags))) @@ -1273,7 +1521,6 @@ static noinline int copy_to_sk(struct btrfs_root *root, int nritems; int i; int slot; - int found = 0; int ret = 0; leaf = path->nodes[0]; @@ -1320,7 +1567,7 @@ static noinline int copy_to_sk(struct btrfs_root *root, item_off, item_len); *sk_offset += item_len; } - found++; + (*num_found)++; if (*num_found >= sk->nr_items) break; @@ -1339,7 +1586,6 @@ advance_key: } else ret = 1; overflow: - *num_found += found; return ret; } @@ -1396,7 +1642,7 @@ static noinline int search_ioctl(struct inode *inode, } ret = copy_to_sk(root, path, &key, sk, args->buf, &sk_offset, &num_found); - btrfs_release_path(root, path); + btrfs_release_path(path); if (ret || num_found >= sk->nr_items) break; @@ -1503,7 +1749,7 @@ static noinline int btrfs_search_path_in_tree(struct btrfs_fs_info *info, if (key.offset == BTRFS_FIRST_FREE_OBJECTID) break; - btrfs_release_path(root, path); + btrfs_release_path(path); key.objectid = key.offset; key.offset = (u64)-1; dirid = key.objectid; @@ -1633,7 +1879,7 @@ static noinline int btrfs_ioctl_snap_destroy(struct file *file, goto out_dput; } - if (inode->i_ino != BTRFS_FIRST_FREE_OBJECTID) { + if (btrfs_ino(inode) != BTRFS_FIRST_FREE_OBJECTID) { err = -EINVAL; goto out_dput; } @@ -1751,7 +1997,10 @@ static int btrfs_ioctl_defrag(struct file *file, void __user *argp) /* the rest are all set to zero by kzalloc */ range->len = (u64)-1; } - ret = btrfs_defrag_file(file, range); + ret = btrfs_defrag_file(fdentry(file)->d_inode, file, + range, 0, 0); + if (ret > 0) + ret = 0; kfree(range); break; default: @@ -1803,6 +2052,75 @@ static long btrfs_ioctl_rm_dev(struct btrfs_root *root, void __user *arg) return ret; } +static long btrfs_ioctl_fs_info(struct btrfs_root *root, void __user *arg) +{ + struct btrfs_ioctl_fs_info_args fi_args; + struct btrfs_device *device; + struct btrfs_device *next; + struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + fi_args.num_devices = fs_devices->num_devices; + fi_args.max_id = 0; + memcpy(&fi_args.fsid, root->fs_info->fsid, sizeof(fi_args.fsid)); + + mutex_lock(&fs_devices->device_list_mutex); + list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) { + if (device->devid > fi_args.max_id) + fi_args.max_id = device->devid; + } + mutex_unlock(&fs_devices->device_list_mutex); + + if (copy_to_user(arg, &fi_args, sizeof(fi_args))) + return -EFAULT; + + return 0; +} + +static long btrfs_ioctl_dev_info(struct btrfs_root *root, void __user *arg) +{ + struct btrfs_ioctl_dev_info_args *di_args; + struct btrfs_device *dev; + struct btrfs_fs_devices *fs_devices = root->fs_info->fs_devices; + int ret = 0; + char *s_uuid = NULL; + char empty_uuid[BTRFS_UUID_SIZE] = {0}; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + di_args = memdup_user(arg, sizeof(*di_args)); + if (IS_ERR(di_args)) + return PTR_ERR(di_args); + + if (memcmp(empty_uuid, di_args->uuid, BTRFS_UUID_SIZE) != 0) + s_uuid = di_args->uuid; + + mutex_lock(&fs_devices->device_list_mutex); + dev = btrfs_find_device(root, di_args->devid, s_uuid, NULL); + mutex_unlock(&fs_devices->device_list_mutex); + + if (!dev) { + ret = -ENODEV; + goto out; + } + + di_args->devid = dev->devid; + di_args->bytes_used = dev->bytes_used; + di_args->total_bytes = dev->total_bytes; + memcpy(di_args->uuid, dev->uuid, sizeof(di_args->uuid)); + strncpy(di_args->path, dev->name, sizeof(di_args->path)); + +out: + if (ret == 0 && copy_to_user(arg, di_args, sizeof(*di_args))) + ret = -EFAULT; + + kfree(di_args); + return ret; +} + static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, u64 off, u64 olen, u64 destoff) { @@ -1919,7 +2237,7 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, } /* clone data */ - key.objectid = src->i_ino; + key.objectid = btrfs_ino(src); key.type = BTRFS_EXTENT_DATA_KEY; key.offset = 0; @@ -1946,7 +2264,7 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, btrfs_item_key_to_cpu(leaf, &key, slot); if (btrfs_key_type(&key) > BTRFS_EXTENT_DATA_KEY || - key.objectid != src->i_ino) + key.objectid != btrfs_ino(src)) break; if (btrfs_key_type(&key) == BTRFS_EXTENT_DATA_KEY) { @@ -1982,14 +2300,14 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, datal = btrfs_file_extent_ram_bytes(leaf, extent); } - btrfs_release_path(root, path); + btrfs_release_path(path); if (key.offset + datal <= off || key.offset >= off+len) goto next; memcpy(&new_key, &key, sizeof(new_key)); - new_key.objectid = inode->i_ino; + new_key.objectid = btrfs_ino(inode); if (off <= key.offset) new_key.offset = key.offset + destoff - off; else @@ -2043,7 +2361,7 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, ret = btrfs_inc_extent_ref(trans, root, disko, diskl, 0, root->root_key.objectid, - inode->i_ino, + btrfs_ino(inode), new_key.offset - datao); BUG_ON(ret); } @@ -2092,7 +2410,7 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, } btrfs_mark_buffer_dirty(leaf); - btrfs_release_path(root, path); + btrfs_release_path(path); inode->i_mtime = inode->i_ctime = CURRENT_TIME; @@ -2113,12 +2431,12 @@ static noinline long btrfs_ioctl_clone(struct file *file, unsigned long srcfd, btrfs_end_transaction(trans, root); } next: - btrfs_release_path(root, path); + btrfs_release_path(path); key.offset++; } ret = 0; out: - btrfs_release_path(root, path); + btrfs_release_path(path); unlock_extent(&BTRFS_I(src)->io_tree, off, off+len, GFP_NOFS); out_unlock: mutex_unlock(&src->i_mutex); @@ -2287,7 +2605,7 @@ long btrfs_ioctl_space_info(struct btrfs_root *root, void __user *arg) struct btrfs_ioctl_space_info space; struct btrfs_ioctl_space_info *dest; struct btrfs_ioctl_space_info *dest_orig; - struct btrfs_ioctl_space_info *user_dest; + struct btrfs_ioctl_space_info __user *user_dest; struct btrfs_space_info *info; u64 types[] = {BTRFS_BLOCK_GROUP_DATA, BTRFS_BLOCK_GROUP_SYSTEM, @@ -2465,6 +2783,58 @@ static noinline long btrfs_ioctl_wait_sync(struct file *file, void __user *argp) return btrfs_wait_for_commit(root, transid); } +static long btrfs_ioctl_scrub(struct btrfs_root *root, void __user *arg) +{ + int ret; + struct btrfs_ioctl_scrub_args *sa; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + sa = memdup_user(arg, sizeof(*sa)); + if (IS_ERR(sa)) + return PTR_ERR(sa); + + ret = btrfs_scrub_dev(root, sa->devid, sa->start, sa->end, + &sa->progress, sa->flags & BTRFS_SCRUB_READONLY); + + if (copy_to_user(arg, sa, sizeof(*sa))) + ret = -EFAULT; + + kfree(sa); + return ret; +} + +static long btrfs_ioctl_scrub_cancel(struct btrfs_root *root, void __user *arg) +{ + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + return btrfs_scrub_cancel(root); +} + +static long btrfs_ioctl_scrub_progress(struct btrfs_root *root, + void __user *arg) +{ + struct btrfs_ioctl_scrub_args *sa; + int ret; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + sa = memdup_user(arg, sizeof(*sa)); + if (IS_ERR(sa)) + return PTR_ERR(sa); + + ret = btrfs_scrub_progress(root, sa->devid, &sa->progress); + + if (copy_to_user(arg, sa, sizeof(*sa))) + ret = -EFAULT; + + kfree(sa); + return ret; +} + long btrfs_ioctl(struct file *file, unsigned int cmd, unsigned long arg) { @@ -2504,6 +2874,10 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_add_dev(root, argp); case BTRFS_IOC_RM_DEV: return btrfs_ioctl_rm_dev(root, argp); + case BTRFS_IOC_FS_INFO: + return btrfs_ioctl_fs_info(root, argp); + case BTRFS_IOC_DEV_INFO: + return btrfs_ioctl_dev_info(root, argp); case BTRFS_IOC_BALANCE: return btrfs_balance(root->fs_info->dev_root); case BTRFS_IOC_CLONE: @@ -2527,6 +2901,12 @@ long btrfs_ioctl(struct file *file, unsigned int return btrfs_ioctl_start_sync(file, argp); case BTRFS_IOC_WAIT_SYNC: return btrfs_ioctl_wait_sync(file, argp); + case BTRFS_IOC_SCRUB: + return btrfs_ioctl_scrub(root, argp); + case BTRFS_IOC_SCRUB_CANCEL: + return btrfs_ioctl_scrub_cancel(root, argp); + case BTRFS_IOC_SCRUB_PROGRESS: + return btrfs_ioctl_scrub_progress(root, argp); } return -ENOTTY; diff --git a/fs/btrfs/ioctl.h b/fs/btrfs/ioctl.h index 8fb382167b13..ad1ea789fcb4 100644 --- a/fs/btrfs/ioctl.h +++ b/fs/btrfs/ioctl.h @@ -32,6 +32,8 @@ struct btrfs_ioctl_vol_args { #define BTRFS_SUBVOL_CREATE_ASYNC (1ULL << 0) #define BTRFS_SUBVOL_RDONLY (1ULL << 1) +#define BTRFS_FSID_SIZE 16 +#define BTRFS_UUID_SIZE 16 #define BTRFS_SUBVOL_NAME_MAX 4039 struct btrfs_ioctl_vol_args_v2 { @@ -42,6 +44,71 @@ struct btrfs_ioctl_vol_args_v2 { char name[BTRFS_SUBVOL_NAME_MAX + 1]; }; +/* + * structure to report errors and progress to userspace, either as a + * result of a finished scrub, a canceled scrub or a progress inquiry + */ +struct btrfs_scrub_progress { + __u64 data_extents_scrubbed; /* # of data extents scrubbed */ + __u64 tree_extents_scrubbed; /* # of tree extents scrubbed */ + __u64 data_bytes_scrubbed; /* # of data bytes scrubbed */ + __u64 tree_bytes_scrubbed; /* # of tree bytes scrubbed */ + __u64 read_errors; /* # of read errors encountered (EIO) */ + __u64 csum_errors; /* # of failed csum checks */ + __u64 verify_errors; /* # of occurences, where the metadata + * of a tree block did not match the + * expected values, like generation or + * logical */ + __u64 no_csum; /* # of 4k data block for which no csum + * is present, probably the result of + * data written with nodatasum */ + __u64 csum_discards; /* # of csum for which no data was found + * in the extent tree. */ + __u64 super_errors; /* # of bad super blocks encountered */ + __u64 malloc_errors; /* # of internal kmalloc errors. These + * will likely cause an incomplete + * scrub */ + __u64 uncorrectable_errors; /* # of errors where either no intact + * copy was found or the writeback + * failed */ + __u64 corrected_errors; /* # of errors corrected */ + __u64 last_physical; /* last physical address scrubbed. In + * case a scrub was aborted, this can + * be used to restart the scrub */ + __u64 unverified_errors; /* # of occurences where a read for a + * full (64k) bio failed, but the re- + * check succeeded for each 4k piece. + * Intermittent error. */ +}; + +#define BTRFS_SCRUB_READONLY 1 +struct btrfs_ioctl_scrub_args { + __u64 devid; /* in */ + __u64 start; /* in */ + __u64 end; /* in */ + __u64 flags; /* in */ + struct btrfs_scrub_progress progress; /* out */ + /* pad to 1k */ + __u64 unused[(1024-32-sizeof(struct btrfs_scrub_progress))/8]; +}; + +#define BTRFS_DEVICE_PATH_NAME_MAX 1024 +struct btrfs_ioctl_dev_info_args { + __u64 devid; /* in/out */ + __u8 uuid[BTRFS_UUID_SIZE]; /* in/out */ + __u64 bytes_used; /* out */ + __u64 total_bytes; /* out */ + __u64 unused[379]; /* pad to 4k */ + __u8 path[BTRFS_DEVICE_PATH_NAME_MAX]; /* out */ +}; + +struct btrfs_ioctl_fs_info_args { + __u64 max_id; /* out */ + __u64 num_devices; /* out */ + __u8 fsid[BTRFS_FSID_SIZE]; /* out */ + __u64 reserved[124]; /* pad to 1k */ +}; + #define BTRFS_INO_LOOKUP_PATH_MAX 4080 struct btrfs_ioctl_ino_lookup_args { __u64 treeid; @@ -114,37 +181,6 @@ struct btrfs_ioctl_clone_range_args { #define BTRFS_DEFRAG_RANGE_COMPRESS 1 #define BTRFS_DEFRAG_RANGE_START_IO 2 -struct btrfs_ioctl_defrag_range_args { - /* start of the defrag operation */ - __u64 start; - - /* number of bytes to defrag, use (u64)-1 to say all */ - __u64 len; - - /* - * flags for the operation, which can include turning - * on compression for this one defrag - */ - __u64 flags; - - /* - * any extent bigger than this will be considered - * already defragged. Use 0 to take the kernel default - * Use 1 to say every single extent must be rewritten - */ - __u32 extent_thresh; - - /* - * which compression method to use if turning on compression - * for this defrag operation. If unspecified, zlib will - * be used - */ - __u32 compress_type; - - /* spare for later */ - __u32 unused[4]; -}; - struct btrfs_ioctl_space_info { __u64 flags; __u64 total_bytes; @@ -203,4 +239,13 @@ struct btrfs_ioctl_space_args { struct btrfs_ioctl_vol_args_v2) #define BTRFS_IOC_SUBVOL_GETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 25, __u64) #define BTRFS_IOC_SUBVOL_SETFLAGS _IOW(BTRFS_IOCTL_MAGIC, 26, __u64) +#define BTRFS_IOC_SCRUB _IOWR(BTRFS_IOCTL_MAGIC, 27, \ + struct btrfs_ioctl_scrub_args) +#define BTRFS_IOC_SCRUB_CANCEL _IO(BTRFS_IOCTL_MAGIC, 28) +#define BTRFS_IOC_SCRUB_PROGRESS _IOWR(BTRFS_IOCTL_MAGIC, 29, \ + struct btrfs_ioctl_scrub_args) +#define BTRFS_IOC_DEV_INFO _IOWR(BTRFS_IOCTL_MAGIC, 30, \ + struct btrfs_ioctl_dev_info_args) +#define BTRFS_IOC_FS_INFO _IOR(BTRFS_IOCTL_MAGIC, 31, \ + struct btrfs_ioctl_fs_info_args) #endif diff --git a/fs/btrfs/locking.c b/fs/btrfs/locking.c index 6151f2ea38bb..66fa43dc3f0f 100644 --- a/fs/btrfs/locking.c +++ b/fs/btrfs/locking.c @@ -185,31 +185,6 @@ sleep: return 0; } -/* - * Very quick trylock, this does not spin or schedule. It returns - * 1 with the spinlock held if it was able to take the lock, or it - * returns zero if it was unable to take the lock. - * - * After this call, scheduling is not safe without first calling - * btrfs_set_lock_blocking() - */ -int btrfs_try_tree_lock(struct extent_buffer *eb) -{ - if (spin_trylock(&eb->lock)) { - if (test_bit(EXTENT_BUFFER_BLOCKING, &eb->bflags)) { - /* - * we've got the spinlock, but the real owner is - * blocking. Drop the spinlock and return failure - */ - spin_unlock(&eb->lock); - return 0; - } - return 1; - } - /* someone else has the spinlock giveup */ - return 0; -} - int btrfs_tree_unlock(struct extent_buffer *eb) { /* diff --git a/fs/btrfs/locking.h b/fs/btrfs/locking.h index 6c4ce457168c..5c33a560a2f1 100644 --- a/fs/btrfs/locking.h +++ b/fs/btrfs/locking.h @@ -21,8 +21,6 @@ int btrfs_tree_lock(struct extent_buffer *eb); int btrfs_tree_unlock(struct extent_buffer *eb); - -int btrfs_try_tree_lock(struct extent_buffer *eb); int btrfs_try_spin_lock(struct extent_buffer *eb); void btrfs_set_lock_blocking(struct extent_buffer *eb); diff --git a/fs/btrfs/ref-cache.c b/fs/btrfs/ref-cache.c index a97314cf6bd6..82d569cb6267 100644 --- a/fs/btrfs/ref-cache.c +++ b/fs/btrfs/ref-cache.c @@ -23,56 +23,6 @@ #include "ref-cache.h" #include "transaction.h" -/* - * leaf refs are used to cache the information about which extents - * a given leaf has references on. This allows us to process that leaf - * in btrfs_drop_snapshot without needing to read it back from disk. - */ - -/* - * kmalloc a leaf reference struct and update the counters for the - * total ref cache size - */ -struct btrfs_leaf_ref *btrfs_alloc_leaf_ref(struct btrfs_root *root, - int nr_extents) -{ - struct btrfs_leaf_ref *ref; - size_t size = btrfs_leaf_ref_size(nr_extents); - - ref = kmalloc(size, GFP_NOFS); - if (ref) { - spin_lock(&root->fs_info->ref_cache_lock); - root->fs_info->total_ref_cache_size += size; - spin_unlock(&root->fs_info->ref_cache_lock); - - memset(ref, 0, sizeof(*ref)); - atomic_set(&ref->usage, 1); - INIT_LIST_HEAD(&ref->list); - } - return ref; -} - -/* - * free a leaf reference struct and update the counters for the - * total ref cache size - */ -void btrfs_free_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref) -{ - if (!ref) - return; - WARN_ON(atomic_read(&ref->usage) == 0); - if (atomic_dec_and_test(&ref->usage)) { - size_t size = btrfs_leaf_ref_size(ref->nritems); - - BUG_ON(ref->in_tree); - kfree(ref); - - spin_lock(&root->fs_info->ref_cache_lock); - root->fs_info->total_ref_cache_size -= size; - spin_unlock(&root->fs_info->ref_cache_lock); - } -} - static struct rb_node *tree_insert(struct rb_root *root, u64 bytenr, struct rb_node *node) { @@ -116,117 +66,3 @@ static struct rb_node *tree_search(struct rb_root *root, u64 bytenr) } return NULL; } - -int btrfs_remove_leaf_refs(struct btrfs_root *root, u64 max_root_gen, - int shared) -{ - struct btrfs_leaf_ref *ref = NULL; - struct btrfs_leaf_ref_tree *tree = root->ref_tree; - - if (shared) - tree = &root->fs_info->shared_ref_tree; - if (!tree) - return 0; - - spin_lock(&tree->lock); - while (!list_empty(&tree->list)) { - ref = list_entry(tree->list.next, struct btrfs_leaf_ref, list); - BUG_ON(ref->tree != tree); - if (ref->root_gen > max_root_gen) - break; - if (!xchg(&ref->in_tree, 0)) { - cond_resched_lock(&tree->lock); - continue; - } - - rb_erase(&ref->rb_node, &tree->root); - list_del_init(&ref->list); - - spin_unlock(&tree->lock); - btrfs_free_leaf_ref(root, ref); - cond_resched(); - spin_lock(&tree->lock); - } - spin_unlock(&tree->lock); - return 0; -} - -/* - * find the leaf ref for a given extent. This returns the ref struct with - * a usage reference incremented - */ -struct btrfs_leaf_ref *btrfs_lookup_leaf_ref(struct btrfs_root *root, - u64 bytenr) -{ - struct rb_node *rb; - struct btrfs_leaf_ref *ref = NULL; - struct btrfs_leaf_ref_tree *tree = root->ref_tree; -again: - if (tree) { - spin_lock(&tree->lock); - rb = tree_search(&tree->root, bytenr); - if (rb) - ref = rb_entry(rb, struct btrfs_leaf_ref, rb_node); - if (ref) - atomic_inc(&ref->usage); - spin_unlock(&tree->lock); - if (ref) - return ref; - } - if (tree != &root->fs_info->shared_ref_tree) { - tree = &root->fs_info->shared_ref_tree; - goto again; - } - return NULL; -} - -/* - * add a fully filled in leaf ref struct - * remove all the refs older than a given root generation - */ -int btrfs_add_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref, - int shared) -{ - int ret = 0; - struct rb_node *rb; - struct btrfs_leaf_ref_tree *tree = root->ref_tree; - - if (shared) - tree = &root->fs_info->shared_ref_tree; - - spin_lock(&tree->lock); - rb = tree_insert(&tree->root, ref->bytenr, &ref->rb_node); - if (rb) { - ret = -EEXIST; - } else { - atomic_inc(&ref->usage); - ref->tree = tree; - ref->in_tree = 1; - list_add_tail(&ref->list, &tree->list); - } - spin_unlock(&tree->lock); - return ret; -} - -/* - * remove a single leaf ref from the tree. This drops the ref held by the tree - * only - */ -int btrfs_remove_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref) -{ - struct btrfs_leaf_ref_tree *tree; - - if (!xchg(&ref->in_tree, 0)) - return 0; - - tree = ref->tree; - spin_lock(&tree->lock); - - rb_erase(&ref->rb_node, &tree->root); - list_del_init(&ref->list); - - spin_unlock(&tree->lock); - - btrfs_free_leaf_ref(root, ref); - return 0; -} diff --git a/fs/btrfs/ref-cache.h b/fs/btrfs/ref-cache.h index e2a55cb2072b..24f7001f6387 100644 --- a/fs/btrfs/ref-cache.h +++ b/fs/btrfs/ref-cache.h @@ -49,28 +49,4 @@ static inline size_t btrfs_leaf_ref_size(int nr_extents) return sizeof(struct btrfs_leaf_ref) + sizeof(struct btrfs_extent_info) * nr_extents; } - -static inline void btrfs_leaf_ref_tree_init(struct btrfs_leaf_ref_tree *tree) -{ - tree->root = RB_ROOT; - INIT_LIST_HEAD(&tree->list); - spin_lock_init(&tree->lock); -} - -static inline int btrfs_leaf_ref_tree_empty(struct btrfs_leaf_ref_tree *tree) -{ - return RB_EMPTY_ROOT(&tree->root); -} - -void btrfs_leaf_ref_tree_init(struct btrfs_leaf_ref_tree *tree); -struct btrfs_leaf_ref *btrfs_alloc_leaf_ref(struct btrfs_root *root, - int nr_extents); -void btrfs_free_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref); -struct btrfs_leaf_ref *btrfs_lookup_leaf_ref(struct btrfs_root *root, - u64 bytenr); -int btrfs_add_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref, - int shared); -int btrfs_remove_leaf_refs(struct btrfs_root *root, u64 max_root_gen, - int shared); -int btrfs_remove_leaf_ref(struct btrfs_root *root, struct btrfs_leaf_ref *ref); #endif diff --git a/fs/btrfs/relocation.c b/fs/btrfs/relocation.c index 58250e09eb05..ca38eca70af0 100644 --- a/fs/btrfs/relocation.c +++ b/fs/btrfs/relocation.c @@ -30,6 +30,7 @@ #include "btrfs_inode.h" #include "async-thread.h" #include "free-space-cache.h" +#include "inode-map.h" /* * backref_node, mapping_node and tree_block start with this @@ -507,6 +508,7 @@ static int update_backref_cache(struct btrfs_trans_handle *trans, return 1; } + static int should_ignore_root(struct btrfs_root *root) { struct btrfs_root *reloc_root; @@ -529,7 +531,6 @@ static int should_ignore_root(struct btrfs_root *root) */ return 1; } - /* * find reloc tree by address of tree root */ @@ -709,7 +710,7 @@ again: WARN_ON(cur->checked); if (!list_empty(&cur->upper)) { /* - * the backref was added previously when processsing + * the backref was added previously when processing * backref of type BTRFS_TREE_BLOCK_REF_KEY */ BUG_ON(!list_is_singular(&cur->upper)); @@ -961,7 +962,7 @@ again: lower = upper; upper = NULL; } - btrfs_release_path(root, path2); + btrfs_release_path(path2); next: if (ptr < end) { ptr += btrfs_extent_inline_ref_size(key.type); @@ -974,7 +975,7 @@ next: if (ptr >= end) path1->slots[0]++; } - btrfs_release_path(rc->extent_root, path1); + btrfs_release_path(path1); cur->checked = 1; WARN_ON(exist); @@ -1409,9 +1410,9 @@ again: prev = node; entry = rb_entry(node, struct btrfs_inode, rb_node); - if (objectid < entry->vfs_inode.i_ino) + if (objectid < btrfs_ino(&entry->vfs_inode)) node = node->rb_left; - else if (objectid > entry->vfs_inode.i_ino) + else if (objectid > btrfs_ino(&entry->vfs_inode)) node = node->rb_right; else break; @@ -1419,7 +1420,7 @@ again: if (!node) { while (prev) { entry = rb_entry(prev, struct btrfs_inode, rb_node); - if (objectid <= entry->vfs_inode.i_ino) { + if (objectid <= btrfs_ino(&entry->vfs_inode)) { node = prev; break; } @@ -1434,7 +1435,7 @@ again: return inode; } - objectid = entry->vfs_inode.i_ino + 1; + objectid = btrfs_ino(&entry->vfs_inode) + 1; if (cond_resched_lock(&root->inode_lock)) goto again; @@ -1470,7 +1471,7 @@ static int get_new_location(struct inode *reloc_inode, u64 *new_bytenr, return -ENOMEM; bytenr -= BTRFS_I(reloc_inode)->index_cnt; - ret = btrfs_lookup_file_extent(NULL, root, path, reloc_inode->i_ino, + ret = btrfs_lookup_file_extent(NULL, root, path, btrfs_ino(reloc_inode), bytenr, 0); if (ret < 0) goto out; @@ -1558,11 +1559,11 @@ int replace_file_extents(struct btrfs_trans_handle *trans, if (first) { inode = find_next_inode(root, key.objectid); first = 0; - } else if (inode && inode->i_ino < key.objectid) { + } else if (inode && btrfs_ino(inode) < key.objectid) { btrfs_add_delayed_iput(inode); inode = find_next_inode(root, key.objectid); } - if (inode && inode->i_ino == key.objectid) { + if (inode && btrfs_ino(inode) == key.objectid) { end = key.offset + btrfs_file_extent_num_bytes(leaf, fi); WARN_ON(!IS_ALIGNED(key.offset, @@ -1749,7 +1750,7 @@ again: btrfs_node_key_to_cpu(path->nodes[level], &key, path->slots[level]); - btrfs_release_path(src, path); + btrfs_release_path(path); path->lowest_level = level; ret = btrfs_search_slot(trans, src, &key, path, 0, 1); @@ -1893,6 +1894,7 @@ static int invalidate_extent_cache(struct btrfs_root *root, struct inode *inode = NULL; u64 objectid; u64 start, end; + u64 ino; objectid = min_key->objectid; while (1) { @@ -1905,17 +1907,18 @@ static int invalidate_extent_cache(struct btrfs_root *root, inode = find_next_inode(root, objectid); if (!inode) break; + ino = btrfs_ino(inode); - if (inode->i_ino > max_key->objectid) { + if (ino > max_key->objectid) { iput(inode); break; } - objectid = inode->i_ino + 1; + objectid = ino + 1; if (!S_ISREG(inode->i_mode)) continue; - if (unlikely(min_key->objectid == inode->i_ino)) { + if (unlikely(min_key->objectid == ino)) { if (min_key->type > BTRFS_EXTENT_DATA_KEY) continue; if (min_key->type < BTRFS_EXTENT_DATA_KEY) @@ -1928,7 +1931,7 @@ static int invalidate_extent_cache(struct btrfs_root *root, start = 0; } - if (unlikely(max_key->objectid == inode->i_ino)) { + if (unlikely(max_key->objectid == ino)) { if (max_key->type < BTRFS_EXTENT_DATA_KEY) continue; if (max_key->type > BTRFS_EXTENT_DATA_KEY) { @@ -2346,7 +2349,7 @@ struct btrfs_root *select_one_root(struct btrfs_trans_handle *trans, root = next->root; BUG_ON(!root); - /* no other choice for non-refernce counted tree */ + /* no other choice for non-references counted tree */ if (!root->ref_cows) return root; @@ -2496,7 +2499,7 @@ static int do_relocation(struct btrfs_trans_handle *trans, path->locks[upper->level] = 0; slot = path->slots[upper->level]; - btrfs_release_path(NULL, path); + btrfs_release_path(path); } else { ret = btrfs_bin_search(upper->eb, key, upper->level, &slot); @@ -2737,7 +2740,7 @@ static int relocate_tree_block(struct btrfs_trans_handle *trans, } else { path->lowest_level = node->level; ret = btrfs_search_slot(trans, root, key, path, 0, 1); - btrfs_release_path(root, path); + btrfs_release_path(path); if (ret > 0) ret = 0; } @@ -2870,7 +2873,7 @@ int setup_extent_mapping(struct inode *inode, u64 start, u64 end, struct extent_map *em; int ret = 0; - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); if (!em) return -ENOMEM; @@ -3119,7 +3122,7 @@ static int add_tree_block(struct reloc_control *rc, #endif } - btrfs_release_path(rc->extent_root, path); + btrfs_release_path(path); BUG_ON(level == -1); @@ -3220,7 +3223,7 @@ static int delete_block_group_cache(struct btrfs_fs_info *fs_info, key.offset = 0; inode = btrfs_iget(fs_info->sb, &key, root, NULL); - if (!inode || IS_ERR(inode) || is_bad_inode(inode)) { + if (IS_ERR_OR_NULL(inode) || is_bad_inode(inode)) { if (inode && !IS_ERR(inode)) iput(inode); return -ENOENT; @@ -3505,7 +3508,7 @@ int add_data_references(struct reloc_control *rc, } path->slots[0]++; } - btrfs_release_path(rc->extent_root, path); + btrfs_release_path(path); if (err) free_block_list(blocks); return err; @@ -3568,7 +3571,7 @@ next: EXTENT_DIRTY); if (ret == 0 && start <= key.objectid) { - btrfs_release_path(rc->extent_root, path); + btrfs_release_path(path); rc->search_start = end + 1; } else { rc->search_start = key.objectid + key.offset; @@ -3576,7 +3579,7 @@ next: return 0; } } - btrfs_release_path(rc->extent_root, path); + btrfs_release_path(path); return ret; } @@ -3713,7 +3716,7 @@ restart: flags = BTRFS_EXTENT_FLAG_DATA; if (path_change) { - btrfs_release_path(rc->extent_root, path); + btrfs_release_path(path); path->search_commit_root = 1; path->skip_locking = 1; @@ -3736,7 +3739,7 @@ restart: (flags & BTRFS_EXTENT_FLAG_DATA)) { ret = add_data_references(rc, &key, path, &blocks); } else { - btrfs_release_path(rc->extent_root, path); + btrfs_release_path(path); ret = 0; } if (ret < 0) { @@ -3799,7 +3802,7 @@ restart: } } - btrfs_release_path(rc->extent_root, path); + btrfs_release_path(path); clear_extent_bits(&rc->processed_blocks, 0, (u64)-1, EXTENT_DIRTY, GFP_NOFS); @@ -3867,7 +3870,7 @@ static int __insert_orphan_inode(struct btrfs_trans_handle *trans, btrfs_set_inode_flags(leaf, item, BTRFS_INODE_NOCOMPRESS | BTRFS_INODE_PREALLOC); btrfs_mark_buffer_dirty(leaf); - btrfs_release_path(root, path); + btrfs_release_path(path); out: btrfs_free_path(path); return ret; @@ -3897,7 +3900,7 @@ struct inode *create_reloc_inode(struct btrfs_fs_info *fs_info, if (IS_ERR(trans)) return ERR_CAST(trans); - err = btrfs_find_free_objectid(trans, root, objectid, &objectid); + err = btrfs_find_free_objectid(root, &objectid); if (err) goto out; @@ -3935,7 +3938,7 @@ static struct reloc_control *alloc_reloc_control(void) INIT_LIST_HEAD(&rc->reloc_roots); backref_cache_init(&rc->backref_cache); mapping_tree_init(&rc->reloc_root_tree); - extent_io_tree_init(&rc->processed_blocks, NULL, GFP_NOFS); + extent_io_tree_init(&rc->processed_blocks, NULL); return rc; } @@ -4109,7 +4112,7 @@ int btrfs_recover_relocation(struct btrfs_root *root) } leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - btrfs_release_path(root->fs_info->tree_root, path); + btrfs_release_path(path); if (key.objectid != BTRFS_TREE_RELOC_OBJECTID || key.type != BTRFS_ROOT_ITEM_KEY) @@ -4141,7 +4144,7 @@ int btrfs_recover_relocation(struct btrfs_root *root) key.offset--; } - btrfs_release_path(root->fs_info->tree_root, path); + btrfs_release_path(path); if (list_empty(&reloc_roots)) goto out; @@ -4242,7 +4245,7 @@ int btrfs_reloc_clone_csums(struct inode *inode, u64 file_pos, u64 len) disk_bytenr = file_pos + BTRFS_I(inode)->index_cnt; ret = btrfs_lookup_csums_range(root->fs_info->csum_root, disk_bytenr, - disk_bytenr + len - 1, &list); + disk_bytenr + len - 1, &list, 0); while (!list_empty(&list)) { sums = list_entry(list.next, struct btrfs_ordered_sum, list); diff --git a/fs/btrfs/root-tree.c b/fs/btrfs/root-tree.c index 6928bff62daa..ebe45443de06 100644 --- a/fs/btrfs/root-tree.c +++ b/fs/btrfs/root-tree.c @@ -22,53 +22,6 @@ #include "print-tree.h" /* - * search forward for a root, starting with objectid 'search_start' - * if a root key is found, the objectid we find is filled into 'found_objectid' - * and 0 is returned. < 0 is returned on error, 1 if there is nothing - * left in the tree. - */ -int btrfs_search_root(struct btrfs_root *root, u64 search_start, - u64 *found_objectid) -{ - struct btrfs_path *path; - struct btrfs_key search_key; - int ret; - - root = root->fs_info->tree_root; - search_key.objectid = search_start; - search_key.type = (u8)-1; - search_key.offset = (u64)-1; - - path = btrfs_alloc_path(); - BUG_ON(!path); -again: - ret = btrfs_search_slot(NULL, root, &search_key, path, 0, 0); - if (ret < 0) - goto out; - if (ret == 0) { - ret = 1; - goto out; - } - if (path->slots[0] >= btrfs_header_nritems(path->nodes[0])) { - ret = btrfs_next_leaf(root, path); - if (ret) - goto out; - } - btrfs_item_key_to_cpu(path->nodes[0], &search_key, path->slots[0]); - if (search_key.type != BTRFS_ROOT_ITEM_KEY) { - search_key.offset++; - btrfs_release_path(root, path); - goto again; - } - ret = 0; - *found_objectid = search_key.objectid; - -out: - btrfs_free_path(path); - return ret; -} - -/* * lookup the root with the highest offset for a given objectid. The key we do * find is copied into 'key'. If we find something return 0, otherwise 1, < 0 * on error. @@ -230,7 +183,7 @@ again: memcpy(&found_key, &key, sizeof(key)); key.offset++; - btrfs_release_path(root, path); + btrfs_release_path(path); dead_root = btrfs_read_fs_root_no_radix(root->fs_info->tree_root, &found_key); @@ -292,7 +245,7 @@ int btrfs_find_orphan_roots(struct btrfs_root *tree_root) } btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - btrfs_release_path(tree_root, path); + btrfs_release_path(path); if (key.objectid != BTRFS_ORPHAN_OBJECTID || key.type != BTRFS_ORPHAN_ITEM_KEY) @@ -385,18 +338,22 @@ again: *sequence = btrfs_root_ref_sequence(leaf, ref); ret = btrfs_del_item(trans, tree_root, path); - BUG_ON(ret); + if (ret) { + err = ret; + goto out; + } } else err = -ENOENT; if (key.type == BTRFS_ROOT_BACKREF_KEY) { - btrfs_release_path(tree_root, path); + btrfs_release_path(path); key.objectid = ref_id; key.type = BTRFS_ROOT_REF_KEY; key.offset = root_id; goto again; } +out: btrfs_free_path(path); return err; } @@ -463,7 +420,7 @@ again: btrfs_mark_buffer_dirty(leaf); if (key.type == BTRFS_ROOT_BACKREF_KEY) { - btrfs_release_path(tree_root, path); + btrfs_release_path(path); key.objectid = ref_id; key.type = BTRFS_ROOT_REF_KEY; key.offset = root_id; diff --git a/fs/btrfs/scrub.c b/fs/btrfs/scrub.c new file mode 100644 index 000000000000..6dfed0c27ac3 --- /dev/null +++ b/fs/btrfs/scrub.c @@ -0,0 +1,1369 @@ +/* + * Copyright (C) 2011 STRATO. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License v2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + * + * You should have received a copy of the GNU General Public + * License along with this program; if not, write to the + * Free Software Foundation, Inc., 59 Temple Place - Suite 330, + * Boston, MA 021110-1307, USA. + */ + +#include <linux/sched.h> +#include <linux/pagemap.h> +#include <linux/writeback.h> +#include <linux/blkdev.h> +#include <linux/rbtree.h> +#include <linux/slab.h> +#include <linux/workqueue.h> +#include "ctree.h" +#include "volumes.h" +#include "disk-io.h" +#include "ordered-data.h" + +/* + * This is only the first step towards a full-features scrub. It reads all + * extent and super block and verifies the checksums. In case a bad checksum + * is found or the extent cannot be read, good data will be written back if + * any can be found. + * + * Future enhancements: + * - To enhance the performance, better read-ahead strategies for the + * extent-tree can be employed. + * - In case an unrepairable extent is encountered, track which files are + * affected and report them + * - In case of a read error on files with nodatasum, map the file and read + * the extent to trigger a writeback of the good copy + * - track and record media errors, throw out bad devices + * - add a mode to also read unallocated space + * - make the prefetch cancellable + */ + +struct scrub_bio; +struct scrub_page; +struct scrub_dev; +static void scrub_bio_end_io(struct bio *bio, int err); +static void scrub_checksum(struct btrfs_work *work); +static int scrub_checksum_data(struct scrub_dev *sdev, + struct scrub_page *spag, void *buffer); +static int scrub_checksum_tree_block(struct scrub_dev *sdev, + struct scrub_page *spag, u64 logical, + void *buffer); +static int scrub_checksum_super(struct scrub_bio *sbio, void *buffer); +static int scrub_fixup_check(struct scrub_bio *sbio, int ix); +static void scrub_fixup_end_io(struct bio *bio, int err); +static int scrub_fixup_io(int rw, struct block_device *bdev, sector_t sector, + struct page *page); +static void scrub_fixup(struct scrub_bio *sbio, int ix); + +#define SCRUB_PAGES_PER_BIO 16 /* 64k per bio */ +#define SCRUB_BIOS_PER_DEV 16 /* 1 MB per device in flight */ + +struct scrub_page { + u64 flags; /* extent flags */ + u64 generation; + u64 mirror_num; + int have_csum; + u8 csum[BTRFS_CSUM_SIZE]; +}; + +struct scrub_bio { + int index; + struct scrub_dev *sdev; + struct bio *bio; + int err; + u64 logical; + u64 physical; + struct scrub_page spag[SCRUB_PAGES_PER_BIO]; + u64 count; + int next_free; + struct btrfs_work work; +}; + +struct scrub_dev { + struct scrub_bio *bios[SCRUB_BIOS_PER_DEV]; + struct btrfs_device *dev; + int first_free; + int curr; + atomic_t in_flight; + spinlock_t list_lock; + wait_queue_head_t list_wait; + u16 csum_size; + struct list_head csum_list; + atomic_t cancel_req; + int readonly; + /* + * statistics + */ + struct btrfs_scrub_progress stat; + spinlock_t stat_lock; +}; + +static void scrub_free_csums(struct scrub_dev *sdev) +{ + while (!list_empty(&sdev->csum_list)) { + struct btrfs_ordered_sum *sum; + sum = list_first_entry(&sdev->csum_list, + struct btrfs_ordered_sum, list); + list_del(&sum->list); + kfree(sum); + } +} + +static noinline_for_stack void scrub_free_dev(struct scrub_dev *sdev) +{ + int i; + int j; + struct page *last_page; + + if (!sdev) + return; + + for (i = 0; i < SCRUB_BIOS_PER_DEV; ++i) { + struct scrub_bio *sbio = sdev->bios[i]; + struct bio *bio; + + if (!sbio) + break; + + bio = sbio->bio; + if (bio) { + last_page = NULL; + for (j = 0; j < bio->bi_vcnt; ++j) { + if (bio->bi_io_vec[j].bv_page == last_page) + continue; + last_page = bio->bi_io_vec[j].bv_page; + __free_page(last_page); + } + bio_put(bio); + } + kfree(sbio); + } + + scrub_free_csums(sdev); + kfree(sdev); +} + +static noinline_for_stack +struct scrub_dev *scrub_setup_dev(struct btrfs_device *dev) +{ + struct scrub_dev *sdev; + int i; + int j; + int ret; + struct btrfs_fs_info *fs_info = dev->dev_root->fs_info; + + sdev = kzalloc(sizeof(*sdev), GFP_NOFS); + if (!sdev) + goto nomem; + sdev->dev = dev; + for (i = 0; i < SCRUB_BIOS_PER_DEV; ++i) { + struct bio *bio; + struct scrub_bio *sbio; + + sbio = kzalloc(sizeof(*sbio), GFP_NOFS); + if (!sbio) + goto nomem; + sdev->bios[i] = sbio; + + bio = bio_kmalloc(GFP_NOFS, SCRUB_PAGES_PER_BIO); + if (!bio) + goto nomem; + + sbio->index = i; + sbio->sdev = sdev; + sbio->bio = bio; + sbio->count = 0; + sbio->work.func = scrub_checksum; + bio->bi_private = sdev->bios[i]; + bio->bi_end_io = scrub_bio_end_io; + bio->bi_sector = 0; + bio->bi_bdev = dev->bdev; + bio->bi_size = 0; + + for (j = 0; j < SCRUB_PAGES_PER_BIO; ++j) { + struct page *page; + page = alloc_page(GFP_NOFS); + if (!page) + goto nomem; + + ret = bio_add_page(bio, page, PAGE_SIZE, 0); + if (!ret) + goto nomem; + } + WARN_ON(bio->bi_vcnt != SCRUB_PAGES_PER_BIO); + + if (i != SCRUB_BIOS_PER_DEV-1) + sdev->bios[i]->next_free = i + 1; + else + sdev->bios[i]->next_free = -1; + } + sdev->first_free = 0; + sdev->curr = -1; + atomic_set(&sdev->in_flight, 0); + atomic_set(&sdev->cancel_req, 0); + sdev->csum_size = btrfs_super_csum_size(&fs_info->super_copy); + INIT_LIST_HEAD(&sdev->csum_list); + + spin_lock_init(&sdev->list_lock); + spin_lock_init(&sdev->stat_lock); + init_waitqueue_head(&sdev->list_wait); + return sdev; + +nomem: + scrub_free_dev(sdev); + return ERR_PTR(-ENOMEM); +} + +/* + * scrub_recheck_error gets called when either verification of the page + * failed or the bio failed to read, e.g. with EIO. In the latter case, + * recheck_error gets called for every page in the bio, even though only + * one may be bad + */ +static void scrub_recheck_error(struct scrub_bio *sbio, int ix) +{ + if (sbio->err) { + if (scrub_fixup_io(READ, sbio->sdev->dev->bdev, + (sbio->physical + ix * PAGE_SIZE) >> 9, + sbio->bio->bi_io_vec[ix].bv_page) == 0) { + if (scrub_fixup_check(sbio, ix) == 0) + return; + } + } + + scrub_fixup(sbio, ix); +} + +static int scrub_fixup_check(struct scrub_bio *sbio, int ix) +{ + int ret = 1; + struct page *page; + void *buffer; + u64 flags = sbio->spag[ix].flags; + + page = sbio->bio->bi_io_vec[ix].bv_page; + buffer = kmap_atomic(page, KM_USER0); + if (flags & BTRFS_EXTENT_FLAG_DATA) { + ret = scrub_checksum_data(sbio->sdev, + sbio->spag + ix, buffer); + } else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { + ret = scrub_checksum_tree_block(sbio->sdev, + sbio->spag + ix, + sbio->logical + ix * PAGE_SIZE, + buffer); + } else { + WARN_ON(1); + } + kunmap_atomic(buffer, KM_USER0); + + return ret; +} + +static void scrub_fixup_end_io(struct bio *bio, int err) +{ + complete((struct completion *)bio->bi_private); +} + +static void scrub_fixup(struct scrub_bio *sbio, int ix) +{ + struct scrub_dev *sdev = sbio->sdev; + struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info; + struct btrfs_mapping_tree *map_tree = &fs_info->mapping_tree; + struct btrfs_multi_bio *multi = NULL; + u64 logical = sbio->logical + ix * PAGE_SIZE; + u64 length; + int i; + int ret; + DECLARE_COMPLETION_ONSTACK(complete); + + if ((sbio->spag[ix].flags & BTRFS_EXTENT_FLAG_DATA) && + (sbio->spag[ix].have_csum == 0)) { + /* + * nodatasum, don't try to fix anything + * FIXME: we can do better, open the inode and trigger a + * writeback + */ + goto uncorrectable; + } + + length = PAGE_SIZE; + ret = btrfs_map_block(map_tree, REQ_WRITE, logical, &length, + &multi, 0); + if (ret || !multi || length < PAGE_SIZE) { + printk(KERN_ERR + "scrub_fixup: btrfs_map_block failed us for %llu\n", + (unsigned long long)logical); + WARN_ON(1); + return; + } + + if (multi->num_stripes == 1) + /* there aren't any replicas */ + goto uncorrectable; + + /* + * first find a good copy + */ + for (i = 0; i < multi->num_stripes; ++i) { + if (i == sbio->spag[ix].mirror_num) + continue; + + if (scrub_fixup_io(READ, multi->stripes[i].dev->bdev, + multi->stripes[i].physical >> 9, + sbio->bio->bi_io_vec[ix].bv_page)) { + /* I/O-error, this is not a good copy */ + continue; + } + + if (scrub_fixup_check(sbio, ix) == 0) + break; + } + if (i == multi->num_stripes) + goto uncorrectable; + + if (!sdev->readonly) { + /* + * bi_io_vec[ix].bv_page now contains good data, write it back + */ + if (scrub_fixup_io(WRITE, sdev->dev->bdev, + (sbio->physical + ix * PAGE_SIZE) >> 9, + sbio->bio->bi_io_vec[ix].bv_page)) { + /* I/O-error, writeback failed, give up */ + goto uncorrectable; + } + } + + kfree(multi); + spin_lock(&sdev->stat_lock); + ++sdev->stat.corrected_errors; + spin_unlock(&sdev->stat_lock); + + if (printk_ratelimit()) + printk(KERN_ERR "btrfs: fixed up at %llu\n", + (unsigned long long)logical); + return; + +uncorrectable: + kfree(multi); + spin_lock(&sdev->stat_lock); + ++sdev->stat.uncorrectable_errors; + spin_unlock(&sdev->stat_lock); + + if (printk_ratelimit()) + printk(KERN_ERR "btrfs: unable to fixup at %llu\n", + (unsigned long long)logical); +} + +static int scrub_fixup_io(int rw, struct block_device *bdev, sector_t sector, + struct page *page) +{ + struct bio *bio = NULL; + int ret; + DECLARE_COMPLETION_ONSTACK(complete); + + /* we are going to wait on this IO */ + rw |= REQ_SYNC; + + bio = bio_alloc(GFP_NOFS, 1); + bio->bi_bdev = bdev; + bio->bi_sector = sector; + bio_add_page(bio, page, PAGE_SIZE, 0); + bio->bi_end_io = scrub_fixup_end_io; + bio->bi_private = &complete; + submit_bio(rw, bio); + + wait_for_completion(&complete); + + ret = !test_bit(BIO_UPTODATE, &bio->bi_flags); + bio_put(bio); + return ret; +} + +static void scrub_bio_end_io(struct bio *bio, int err) +{ + struct scrub_bio *sbio = bio->bi_private; + struct scrub_dev *sdev = sbio->sdev; + struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info; + + sbio->err = err; + + btrfs_queue_worker(&fs_info->scrub_workers, &sbio->work); +} + +static void scrub_checksum(struct btrfs_work *work) +{ + struct scrub_bio *sbio = container_of(work, struct scrub_bio, work); + struct scrub_dev *sdev = sbio->sdev; + struct page *page; + void *buffer; + int i; + u64 flags; + u64 logical; + int ret; + + if (sbio->err) { + for (i = 0; i < sbio->count; ++i) + scrub_recheck_error(sbio, i); + + sbio->bio->bi_flags &= ~(BIO_POOL_MASK - 1); + sbio->bio->bi_flags |= 1 << BIO_UPTODATE; + sbio->bio->bi_phys_segments = 0; + sbio->bio->bi_idx = 0; + + for (i = 0; i < sbio->count; i++) { + struct bio_vec *bi; + bi = &sbio->bio->bi_io_vec[i]; + bi->bv_offset = 0; + bi->bv_len = PAGE_SIZE; + } + + spin_lock(&sdev->stat_lock); + ++sdev->stat.read_errors; + spin_unlock(&sdev->stat_lock); + goto out; + } + for (i = 0; i < sbio->count; ++i) { + page = sbio->bio->bi_io_vec[i].bv_page; + buffer = kmap_atomic(page, KM_USER0); + flags = sbio->spag[i].flags; + logical = sbio->logical + i * PAGE_SIZE; + ret = 0; + if (flags & BTRFS_EXTENT_FLAG_DATA) { + ret = scrub_checksum_data(sdev, sbio->spag + i, buffer); + } else if (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK) { + ret = scrub_checksum_tree_block(sdev, sbio->spag + i, + logical, buffer); + } else if (flags & BTRFS_EXTENT_FLAG_SUPER) { + BUG_ON(i); + (void)scrub_checksum_super(sbio, buffer); + } else { + WARN_ON(1); + } + kunmap_atomic(buffer, KM_USER0); + if (ret) + scrub_recheck_error(sbio, i); + } + +out: + spin_lock(&sdev->list_lock); + sbio->next_free = sdev->first_free; + sdev->first_free = sbio->index; + spin_unlock(&sdev->list_lock); + atomic_dec(&sdev->in_flight); + wake_up(&sdev->list_wait); +} + +static int scrub_checksum_data(struct scrub_dev *sdev, + struct scrub_page *spag, void *buffer) +{ + u8 csum[BTRFS_CSUM_SIZE]; + u32 crc = ~(u32)0; + int fail = 0; + struct btrfs_root *root = sdev->dev->dev_root; + + if (!spag->have_csum) + return 0; + + crc = btrfs_csum_data(root, buffer, crc, PAGE_SIZE); + btrfs_csum_final(crc, csum); + if (memcmp(csum, spag->csum, sdev->csum_size)) + fail = 1; + + spin_lock(&sdev->stat_lock); + ++sdev->stat.data_extents_scrubbed; + sdev->stat.data_bytes_scrubbed += PAGE_SIZE; + if (fail) + ++sdev->stat.csum_errors; + spin_unlock(&sdev->stat_lock); + + return fail; +} + +static int scrub_checksum_tree_block(struct scrub_dev *sdev, + struct scrub_page *spag, u64 logical, + void *buffer) +{ + struct btrfs_header *h; + struct btrfs_root *root = sdev->dev->dev_root; + struct btrfs_fs_info *fs_info = root->fs_info; + u8 csum[BTRFS_CSUM_SIZE]; + u32 crc = ~(u32)0; + int fail = 0; + int crc_fail = 0; + + /* + * we don't use the getter functions here, as we + * a) don't have an extent buffer and + * b) the page is already kmapped + */ + h = (struct btrfs_header *)buffer; + + if (logical != le64_to_cpu(h->bytenr)) + ++fail; + + if (spag->generation != le64_to_cpu(h->generation)) + ++fail; + + if (memcmp(h->fsid, fs_info->fsid, BTRFS_UUID_SIZE)) + ++fail; + + if (memcmp(h->chunk_tree_uuid, fs_info->chunk_tree_uuid, + BTRFS_UUID_SIZE)) + ++fail; + + crc = btrfs_csum_data(root, buffer + BTRFS_CSUM_SIZE, crc, + PAGE_SIZE - BTRFS_CSUM_SIZE); + btrfs_csum_final(crc, csum); + if (memcmp(csum, h->csum, sdev->csum_size)) + ++crc_fail; + + spin_lock(&sdev->stat_lock); + ++sdev->stat.tree_extents_scrubbed; + sdev->stat.tree_bytes_scrubbed += PAGE_SIZE; + if (crc_fail) + ++sdev->stat.csum_errors; + if (fail) + ++sdev->stat.verify_errors; + spin_unlock(&sdev->stat_lock); + + return fail || crc_fail; +} + +static int scrub_checksum_super(struct scrub_bio *sbio, void *buffer) +{ + struct btrfs_super_block *s; + u64 logical; + struct scrub_dev *sdev = sbio->sdev; + struct btrfs_root *root = sdev->dev->dev_root; + struct btrfs_fs_info *fs_info = root->fs_info; + u8 csum[BTRFS_CSUM_SIZE]; + u32 crc = ~(u32)0; + int fail = 0; + + s = (struct btrfs_super_block *)buffer; + logical = sbio->logical; + + if (logical != le64_to_cpu(s->bytenr)) + ++fail; + + if (sbio->spag[0].generation != le64_to_cpu(s->generation)) + ++fail; + + if (memcmp(s->fsid, fs_info->fsid, BTRFS_UUID_SIZE)) + ++fail; + + crc = btrfs_csum_data(root, buffer + BTRFS_CSUM_SIZE, crc, + PAGE_SIZE - BTRFS_CSUM_SIZE); + btrfs_csum_final(crc, csum); + if (memcmp(csum, s->csum, sbio->sdev->csum_size)) + ++fail; + + if (fail) { + /* + * if we find an error in a super block, we just report it. + * They will get written with the next transaction commit + * anyway + */ + spin_lock(&sdev->stat_lock); + ++sdev->stat.super_errors; + spin_unlock(&sdev->stat_lock); + } + + return fail; +} + +static int scrub_submit(struct scrub_dev *sdev) +{ + struct scrub_bio *sbio; + + if (sdev->curr == -1) + return 0; + + sbio = sdev->bios[sdev->curr]; + + sbio->bio->bi_sector = sbio->physical >> 9; + sbio->bio->bi_size = sbio->count * PAGE_SIZE; + sbio->bio->bi_next = NULL; + sbio->bio->bi_flags |= 1 << BIO_UPTODATE; + sbio->bio->bi_comp_cpu = -1; + sbio->bio->bi_bdev = sdev->dev->bdev; + sbio->err = 0; + sdev->curr = -1; + atomic_inc(&sdev->in_flight); + + submit_bio(0, sbio->bio); + + return 0; +} + +static int scrub_page(struct scrub_dev *sdev, u64 logical, u64 len, + u64 physical, u64 flags, u64 gen, u64 mirror_num, + u8 *csum, int force) +{ + struct scrub_bio *sbio; + +again: + /* + * grab a fresh bio or wait for one to become available + */ + while (sdev->curr == -1) { + spin_lock(&sdev->list_lock); + sdev->curr = sdev->first_free; + if (sdev->curr != -1) { + sdev->first_free = sdev->bios[sdev->curr]->next_free; + sdev->bios[sdev->curr]->next_free = -1; + sdev->bios[sdev->curr]->count = 0; + spin_unlock(&sdev->list_lock); + } else { + spin_unlock(&sdev->list_lock); + wait_event(sdev->list_wait, sdev->first_free != -1); + } + } + sbio = sdev->bios[sdev->curr]; + if (sbio->count == 0) { + sbio->physical = physical; + sbio->logical = logical; + } else if (sbio->physical + sbio->count * PAGE_SIZE != physical || + sbio->logical + sbio->count * PAGE_SIZE != logical) { + scrub_submit(sdev); + goto again; + } + sbio->spag[sbio->count].flags = flags; + sbio->spag[sbio->count].generation = gen; + sbio->spag[sbio->count].have_csum = 0; + sbio->spag[sbio->count].mirror_num = mirror_num; + if (csum) { + sbio->spag[sbio->count].have_csum = 1; + memcpy(sbio->spag[sbio->count].csum, csum, sdev->csum_size); + } + ++sbio->count; + if (sbio->count == SCRUB_PAGES_PER_BIO || force) + scrub_submit(sdev); + + return 0; +} + +static int scrub_find_csum(struct scrub_dev *sdev, u64 logical, u64 len, + u8 *csum) +{ + struct btrfs_ordered_sum *sum = NULL; + int ret = 0; + unsigned long i; + unsigned long num_sectors; + u32 sectorsize = sdev->dev->dev_root->sectorsize; + + while (!list_empty(&sdev->csum_list)) { + sum = list_first_entry(&sdev->csum_list, + struct btrfs_ordered_sum, list); + if (sum->bytenr > logical) + return 0; + if (sum->bytenr + sum->len > logical) + break; + + ++sdev->stat.csum_discards; + list_del(&sum->list); + kfree(sum); + sum = NULL; + } + if (!sum) + return 0; + + num_sectors = sum->len / sectorsize; + for (i = 0; i < num_sectors; ++i) { + if (sum->sums[i].bytenr == logical) { + memcpy(csum, &sum->sums[i].sum, sdev->csum_size); + ret = 1; + break; + } + } + if (ret && i == num_sectors - 1) { + list_del(&sum->list); + kfree(sum); + } + return ret; +} + +/* scrub extent tries to collect up to 64 kB for each bio */ +static int scrub_extent(struct scrub_dev *sdev, u64 logical, u64 len, + u64 physical, u64 flags, u64 gen, u64 mirror_num) +{ + int ret; + u8 csum[BTRFS_CSUM_SIZE]; + + while (len) { + u64 l = min_t(u64, len, PAGE_SIZE); + int have_csum = 0; + + if (flags & BTRFS_EXTENT_FLAG_DATA) { + /* push csums to sbio */ + have_csum = scrub_find_csum(sdev, logical, l, csum); + if (have_csum == 0) + ++sdev->stat.no_csum; + } + ret = scrub_page(sdev, logical, l, physical, flags, gen, + mirror_num, have_csum ? csum : NULL, 0); + if (ret) + return ret; + len -= l; + logical += l; + physical += l; + } + return 0; +} + +static noinline_for_stack int scrub_stripe(struct scrub_dev *sdev, + struct map_lookup *map, int num, u64 base, u64 length) +{ + struct btrfs_path *path; + struct btrfs_fs_info *fs_info = sdev->dev->dev_root->fs_info; + struct btrfs_root *root = fs_info->extent_root; + struct btrfs_root *csum_root = fs_info->csum_root; + struct btrfs_extent_item *extent; + u64 flags; + int ret; + int slot; + int i; + u64 nstripes; + int start_stripe; + struct extent_buffer *l; + struct btrfs_key key; + u64 physical; + u64 logical; + u64 generation; + u64 mirror_num; + + u64 increment = map->stripe_len; + u64 offset; + + nstripes = length; + offset = 0; + do_div(nstripes, map->stripe_len); + if (map->type & BTRFS_BLOCK_GROUP_RAID0) { + offset = map->stripe_len * num; + increment = map->stripe_len * map->num_stripes; + mirror_num = 0; + } else if (map->type & BTRFS_BLOCK_GROUP_RAID10) { + int factor = map->num_stripes / map->sub_stripes; + offset = map->stripe_len * (num / map->sub_stripes); + increment = map->stripe_len * factor; + mirror_num = num % map->sub_stripes; + } else if (map->type & BTRFS_BLOCK_GROUP_RAID1) { + increment = map->stripe_len; + mirror_num = num % map->num_stripes; + } else if (map->type & BTRFS_BLOCK_GROUP_DUP) { + increment = map->stripe_len; + mirror_num = num % map->num_stripes; + } else { + increment = map->stripe_len; + mirror_num = 0; + } + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + path->reada = 2; + path->search_commit_root = 1; + path->skip_locking = 1; + + /* + * find all extents for each stripe and just read them to get + * them into the page cache + * FIXME: we can do better. build a more intelligent prefetching + */ + logical = base + offset; + physical = map->stripes[num].physical; + ret = 0; + for (i = 0; i < nstripes; ++i) { + key.objectid = logical; + key.type = BTRFS_EXTENT_ITEM_KEY; + key.offset = (u64)0; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + goto out; + + l = path->nodes[0]; + slot = path->slots[0]; + btrfs_item_key_to_cpu(l, &key, slot); + if (key.objectid != logical) { + ret = btrfs_previous_item(root, path, 0, + BTRFS_EXTENT_ITEM_KEY); + if (ret < 0) + goto out; + } + + while (1) { + l = path->nodes[0]; + slot = path->slots[0]; + if (slot >= btrfs_header_nritems(l)) { + ret = btrfs_next_leaf(root, path); + if (ret == 0) + continue; + if (ret < 0) + goto out; + + break; + } + btrfs_item_key_to_cpu(l, &key, slot); + + if (key.objectid >= logical + map->stripe_len) + break; + + path->slots[0]++; + } + btrfs_release_path(path); + logical += increment; + physical += map->stripe_len; + cond_resched(); + } + + /* + * collect all data csums for the stripe to avoid seeking during + * the scrub. This might currently (crc32) end up to be about 1MB + */ + start_stripe = 0; +again: + logical = base + offset + start_stripe * increment; + for (i = start_stripe; i < nstripes; ++i) { + ret = btrfs_lookup_csums_range(csum_root, logical, + logical + map->stripe_len - 1, + &sdev->csum_list, 1); + if (ret) + goto out; + + logical += increment; + cond_resched(); + } + /* + * now find all extents for each stripe and scrub them + */ + logical = base + offset + start_stripe * increment; + physical = map->stripes[num].physical + start_stripe * map->stripe_len; + ret = 0; + for (i = start_stripe; i < nstripes; ++i) { + /* + * canceled? + */ + if (atomic_read(&fs_info->scrub_cancel_req) || + atomic_read(&sdev->cancel_req)) { + ret = -ECANCELED; + goto out; + } + /* + * check to see if we have to pause + */ + if (atomic_read(&fs_info->scrub_pause_req)) { + /* push queued extents */ + scrub_submit(sdev); + wait_event(sdev->list_wait, + atomic_read(&sdev->in_flight) == 0); + atomic_inc(&fs_info->scrubs_paused); + wake_up(&fs_info->scrub_pause_wait); + mutex_lock(&fs_info->scrub_lock); + while (atomic_read(&fs_info->scrub_pause_req)) { + mutex_unlock(&fs_info->scrub_lock); + wait_event(fs_info->scrub_pause_wait, + atomic_read(&fs_info->scrub_pause_req) == 0); + mutex_lock(&fs_info->scrub_lock); + } + atomic_dec(&fs_info->scrubs_paused); + mutex_unlock(&fs_info->scrub_lock); + wake_up(&fs_info->scrub_pause_wait); + scrub_free_csums(sdev); + start_stripe = i; + goto again; + } + + key.objectid = logical; + key.type = BTRFS_EXTENT_ITEM_KEY; + key.offset = (u64)0; + + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + goto out; + + l = path->nodes[0]; + slot = path->slots[0]; + btrfs_item_key_to_cpu(l, &key, slot); + if (key.objectid != logical) { + ret = btrfs_previous_item(root, path, 0, + BTRFS_EXTENT_ITEM_KEY); + if (ret < 0) + goto out; + } + + while (1) { + l = path->nodes[0]; + slot = path->slots[0]; + if (slot >= btrfs_header_nritems(l)) { + ret = btrfs_next_leaf(root, path); + if (ret == 0) + continue; + if (ret < 0) + goto out; + + break; + } + btrfs_item_key_to_cpu(l, &key, slot); + + if (key.objectid + key.offset <= logical) + goto next; + + if (key.objectid >= logical + map->stripe_len) + break; + + if (btrfs_key_type(&key) != BTRFS_EXTENT_ITEM_KEY) + goto next; + + extent = btrfs_item_ptr(l, slot, + struct btrfs_extent_item); + flags = btrfs_extent_flags(l, extent); + generation = btrfs_extent_generation(l, extent); + + if (key.objectid < logical && + (flags & BTRFS_EXTENT_FLAG_TREE_BLOCK)) { + printk(KERN_ERR + "btrfs scrub: tree block %llu spanning " + "stripes, ignored. logical=%llu\n", + (unsigned long long)key.objectid, + (unsigned long long)logical); + goto next; + } + + /* + * trim extent to this stripe + */ + if (key.objectid < logical) { + key.offset -= logical - key.objectid; + key.objectid = logical; + } + if (key.objectid + key.offset > + logical + map->stripe_len) { + key.offset = logical + map->stripe_len - + key.objectid; + } + + ret = scrub_extent(sdev, key.objectid, key.offset, + key.objectid - logical + physical, + flags, generation, mirror_num); + if (ret) + goto out; + +next: + path->slots[0]++; + } + btrfs_release_path(path); + logical += increment; + physical += map->stripe_len; + spin_lock(&sdev->stat_lock); + sdev->stat.last_physical = physical; + spin_unlock(&sdev->stat_lock); + } + /* push queued extents */ + scrub_submit(sdev); + +out: + btrfs_free_path(path); + return ret < 0 ? ret : 0; +} + +static noinline_for_stack int scrub_chunk(struct scrub_dev *sdev, + u64 chunk_tree, u64 chunk_objectid, u64 chunk_offset, u64 length) +{ + struct btrfs_mapping_tree *map_tree = + &sdev->dev->dev_root->fs_info->mapping_tree; + struct map_lookup *map; + struct extent_map *em; + int i; + int ret = -EINVAL; + + read_lock(&map_tree->map_tree.lock); + em = lookup_extent_mapping(&map_tree->map_tree, chunk_offset, 1); + read_unlock(&map_tree->map_tree.lock); + + if (!em) + return -EINVAL; + + map = (struct map_lookup *)em->bdev; + if (em->start != chunk_offset) + goto out; + + if (em->len < length) + goto out; + + for (i = 0; i < map->num_stripes; ++i) { + if (map->stripes[i].dev == sdev->dev) { + ret = scrub_stripe(sdev, map, i, chunk_offset, length); + if (ret) + goto out; + } + } +out: + free_extent_map(em); + + return ret; +} + +static noinline_for_stack +int scrub_enumerate_chunks(struct scrub_dev *sdev, u64 start, u64 end) +{ + struct btrfs_dev_extent *dev_extent = NULL; + struct btrfs_path *path; + struct btrfs_root *root = sdev->dev->dev_root; + struct btrfs_fs_info *fs_info = root->fs_info; + u64 length; + u64 chunk_tree; + u64 chunk_objectid; + u64 chunk_offset; + int ret; + int slot; + struct extent_buffer *l; + struct btrfs_key key; + struct btrfs_key found_key; + struct btrfs_block_group_cache *cache; + + path = btrfs_alloc_path(); + if (!path) + return -ENOMEM; + + path->reada = 2; + path->search_commit_root = 1; + path->skip_locking = 1; + + key.objectid = sdev->dev->devid; + key.offset = 0ull; + key.type = BTRFS_DEV_EXTENT_KEY; + + + while (1) { + ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); + if (ret < 0) + goto out; + ret = 0; + + l = path->nodes[0]; + slot = path->slots[0]; + + btrfs_item_key_to_cpu(l, &found_key, slot); + + if (found_key.objectid != sdev->dev->devid) + break; + + if (btrfs_key_type(&key) != BTRFS_DEV_EXTENT_KEY) + break; + + if (found_key.offset >= end) + break; + + if (found_key.offset < key.offset) + break; + + dev_extent = btrfs_item_ptr(l, slot, struct btrfs_dev_extent); + length = btrfs_dev_extent_length(l, dev_extent); + + if (found_key.offset + length <= start) { + key.offset = found_key.offset + length; + btrfs_release_path(path); + continue; + } + + chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent); + chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent); + chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent); + + /* + * get a reference on the corresponding block group to prevent + * the chunk from going away while we scrub it + */ + cache = btrfs_lookup_block_group(fs_info, chunk_offset); + if (!cache) { + ret = -ENOENT; + goto out; + } + ret = scrub_chunk(sdev, chunk_tree, chunk_objectid, + chunk_offset, length); + btrfs_put_block_group(cache); + if (ret) + break; + + key.offset = found_key.offset + length; + btrfs_release_path(path); + } + +out: + btrfs_free_path(path); + return ret; +} + +static noinline_for_stack int scrub_supers(struct scrub_dev *sdev) +{ + int i; + u64 bytenr; + u64 gen; + int ret; + struct btrfs_device *device = sdev->dev; + struct btrfs_root *root = device->dev_root; + + gen = root->fs_info->last_trans_committed; + + for (i = 0; i < BTRFS_SUPER_MIRROR_MAX; i++) { + bytenr = btrfs_sb_offset(i); + if (bytenr + BTRFS_SUPER_INFO_SIZE >= device->total_bytes) + break; + + ret = scrub_page(sdev, bytenr, PAGE_SIZE, bytenr, + BTRFS_EXTENT_FLAG_SUPER, gen, i, NULL, 1); + if (ret) + return ret; + } + wait_event(sdev->list_wait, atomic_read(&sdev->in_flight) == 0); + + return 0; +} + +/* + * get a reference count on fs_info->scrub_workers. start worker if necessary + */ +static noinline_for_stack int scrub_workers_get(struct btrfs_root *root) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + + mutex_lock(&fs_info->scrub_lock); + if (fs_info->scrub_workers_refcnt == 0) + btrfs_start_workers(&fs_info->scrub_workers, 1); + ++fs_info->scrub_workers_refcnt; + mutex_unlock(&fs_info->scrub_lock); + + return 0; +} + +static noinline_for_stack void scrub_workers_put(struct btrfs_root *root) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + + mutex_lock(&fs_info->scrub_lock); + if (--fs_info->scrub_workers_refcnt == 0) + btrfs_stop_workers(&fs_info->scrub_workers); + WARN_ON(fs_info->scrub_workers_refcnt < 0); + mutex_unlock(&fs_info->scrub_lock); +} + + +int btrfs_scrub_dev(struct btrfs_root *root, u64 devid, u64 start, u64 end, + struct btrfs_scrub_progress *progress, int readonly) +{ + struct scrub_dev *sdev; + struct btrfs_fs_info *fs_info = root->fs_info; + int ret; + struct btrfs_device *dev; + + if (root->fs_info->closing) + return -EINVAL; + + /* + * check some assumptions + */ + if (root->sectorsize != PAGE_SIZE || + root->sectorsize != root->leafsize || + root->sectorsize != root->nodesize) { + printk(KERN_ERR "btrfs_scrub: size assumptions fail\n"); + return -EINVAL; + } + + ret = scrub_workers_get(root); + if (ret) + return ret; + + mutex_lock(&root->fs_info->fs_devices->device_list_mutex); + dev = btrfs_find_device(root, devid, NULL, NULL); + if (!dev || dev->missing) { + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + scrub_workers_put(root); + return -ENODEV; + } + mutex_lock(&fs_info->scrub_lock); + + if (!dev->in_fs_metadata) { + mutex_unlock(&fs_info->scrub_lock); + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + scrub_workers_put(root); + return -ENODEV; + } + + if (dev->scrub_device) { + mutex_unlock(&fs_info->scrub_lock); + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + scrub_workers_put(root); + return -EINPROGRESS; + } + sdev = scrub_setup_dev(dev); + if (IS_ERR(sdev)) { + mutex_unlock(&fs_info->scrub_lock); + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + scrub_workers_put(root); + return PTR_ERR(sdev); + } + sdev->readonly = readonly; + dev->scrub_device = sdev; + + atomic_inc(&fs_info->scrubs_running); + mutex_unlock(&fs_info->scrub_lock); + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + + down_read(&fs_info->scrub_super_lock); + ret = scrub_supers(sdev); + up_read(&fs_info->scrub_super_lock); + + if (!ret) + ret = scrub_enumerate_chunks(sdev, start, end); + + wait_event(sdev->list_wait, atomic_read(&sdev->in_flight) == 0); + + atomic_dec(&fs_info->scrubs_running); + wake_up(&fs_info->scrub_pause_wait); + + if (progress) + memcpy(progress, &sdev->stat, sizeof(*progress)); + + mutex_lock(&fs_info->scrub_lock); + dev->scrub_device = NULL; + mutex_unlock(&fs_info->scrub_lock); + + scrub_free_dev(sdev); + scrub_workers_put(root); + + return ret; +} + +int btrfs_scrub_pause(struct btrfs_root *root) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + + mutex_lock(&fs_info->scrub_lock); + atomic_inc(&fs_info->scrub_pause_req); + while (atomic_read(&fs_info->scrubs_paused) != + atomic_read(&fs_info->scrubs_running)) { + mutex_unlock(&fs_info->scrub_lock); + wait_event(fs_info->scrub_pause_wait, + atomic_read(&fs_info->scrubs_paused) == + atomic_read(&fs_info->scrubs_running)); + mutex_lock(&fs_info->scrub_lock); + } + mutex_unlock(&fs_info->scrub_lock); + + return 0; +} + +int btrfs_scrub_continue(struct btrfs_root *root) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + + atomic_dec(&fs_info->scrub_pause_req); + wake_up(&fs_info->scrub_pause_wait); + return 0; +} + +int btrfs_scrub_pause_super(struct btrfs_root *root) +{ + down_write(&root->fs_info->scrub_super_lock); + return 0; +} + +int btrfs_scrub_continue_super(struct btrfs_root *root) +{ + up_write(&root->fs_info->scrub_super_lock); + return 0; +} + +int btrfs_scrub_cancel(struct btrfs_root *root) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + + mutex_lock(&fs_info->scrub_lock); + if (!atomic_read(&fs_info->scrubs_running)) { + mutex_unlock(&fs_info->scrub_lock); + return -ENOTCONN; + } + + atomic_inc(&fs_info->scrub_cancel_req); + while (atomic_read(&fs_info->scrubs_running)) { + mutex_unlock(&fs_info->scrub_lock); + wait_event(fs_info->scrub_pause_wait, + atomic_read(&fs_info->scrubs_running) == 0); + mutex_lock(&fs_info->scrub_lock); + } + atomic_dec(&fs_info->scrub_cancel_req); + mutex_unlock(&fs_info->scrub_lock); + + return 0; +} + +int btrfs_scrub_cancel_dev(struct btrfs_root *root, struct btrfs_device *dev) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + struct scrub_dev *sdev; + + mutex_lock(&fs_info->scrub_lock); + sdev = dev->scrub_device; + if (!sdev) { + mutex_unlock(&fs_info->scrub_lock); + return -ENOTCONN; + } + atomic_inc(&sdev->cancel_req); + while (dev->scrub_device) { + mutex_unlock(&fs_info->scrub_lock); + wait_event(fs_info->scrub_pause_wait, + dev->scrub_device == NULL); + mutex_lock(&fs_info->scrub_lock); + } + mutex_unlock(&fs_info->scrub_lock); + + return 0; +} +int btrfs_scrub_cancel_devid(struct btrfs_root *root, u64 devid) +{ + struct btrfs_fs_info *fs_info = root->fs_info; + struct btrfs_device *dev; + int ret; + + /* + * we have to hold the device_list_mutex here so the device + * does not go away in cancel_dev. FIXME: find a better solution + */ + mutex_lock(&fs_info->fs_devices->device_list_mutex); + dev = btrfs_find_device(root, devid, NULL, NULL); + if (!dev) { + mutex_unlock(&fs_info->fs_devices->device_list_mutex); + return -ENODEV; + } + ret = btrfs_scrub_cancel_dev(root, dev); + mutex_unlock(&fs_info->fs_devices->device_list_mutex); + + return ret; +} + +int btrfs_scrub_progress(struct btrfs_root *root, u64 devid, + struct btrfs_scrub_progress *progress) +{ + struct btrfs_device *dev; + struct scrub_dev *sdev = NULL; + + mutex_lock(&root->fs_info->fs_devices->device_list_mutex); + dev = btrfs_find_device(root, devid, NULL, NULL); + if (dev) + sdev = dev->scrub_device; + if (sdev) + memcpy(progress, &sdev->stat, sizeof(*progress)); + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + + return dev ? (sdev ? 0 : -ENOTCONN) : -ENODEV; +} diff --git a/fs/btrfs/super.c b/fs/btrfs/super.c index 58e7de9cc90c..9b2e7e5bc3ef 100644 --- a/fs/btrfs/super.c +++ b/fs/btrfs/super.c @@ -39,7 +39,9 @@ #include <linux/miscdevice.h> #include <linux/magic.h> #include <linux/slab.h> +#include <linux/cleancache.h> #include "compat.h" +#include "delayed-inode.h" #include "ctree.h" #include "disk-io.h" #include "transaction.h" @@ -159,7 +161,7 @@ enum { Opt_compress_type, Opt_compress_force, Opt_compress_force_type, Opt_notreelog, Opt_ratio, Opt_flushoncommit, Opt_discard, Opt_space_cache, Opt_clear_cache, Opt_user_subvol_rm_allowed, - Opt_enospc_debug, Opt_err, + Opt_enospc_debug, Opt_subvolrootid, Opt_defrag, Opt_err, }; static match_table_t tokens = { @@ -189,6 +191,8 @@ static match_table_t tokens = { {Opt_clear_cache, "clear_cache"}, {Opt_user_subvol_rm_allowed, "user_subvol_rm_allowed"}, {Opt_enospc_debug, "enospc_debug"}, + {Opt_subvolrootid, "subvolrootid=%d"}, + {Opt_defrag, "autodefrag"}, {Opt_err, NULL}, }; @@ -232,6 +236,7 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) break; case Opt_subvol: case Opt_subvolid: + case Opt_subvolrootid: case Opt_device: /* * These are parsed by btrfs_parse_early_options @@ -366,6 +371,10 @@ int btrfs_parse_options(struct btrfs_root *root, char *options) case Opt_enospc_debug: btrfs_set_opt(info->mount_opt, ENOSPC_DEBUG); break; + case Opt_defrag: + printk(KERN_INFO "btrfs: enabling auto defrag"); + btrfs_set_opt(info->mount_opt, AUTO_DEFRAG); + break; case Opt_err: printk(KERN_INFO "btrfs: unrecognized mount option " "'%s'\n", p); @@ -388,7 +397,7 @@ out: */ static int btrfs_parse_early_options(const char *options, fmode_t flags, void *holder, char **subvol_name, u64 *subvol_objectid, - struct btrfs_fs_devices **fs_devices) + u64 *subvol_rootid, struct btrfs_fs_devices **fs_devices) { substring_t args[MAX_OPT_ARGS]; char *opts, *orig, *p; @@ -429,6 +438,18 @@ static int btrfs_parse_early_options(const char *options, fmode_t flags, *subvol_objectid = intarg; } break; + case Opt_subvolrootid: + intarg = 0; + error = match_int(&args[0], &intarg); + if (!error) { + /* we want the original fs_tree */ + if (!intarg) + *subvol_rootid = + BTRFS_FS_TREE_OBJECTID; + else + *subvol_rootid = intarg; + } + break; case Opt_device: error = btrfs_scan_one_device(match_strdup(&args[0]), flags, holder, fs_devices); @@ -492,8 +513,10 @@ static struct dentry *get_default_root(struct super_block *sb, */ dir_id = btrfs_super_root_dir(&root->fs_info->super_copy); di = btrfs_lookup_dir_item(NULL, root, path, dir_id, "default", 7, 0); - if (IS_ERR(di)) + if (IS_ERR(di)) { + btrfs_free_path(path); return ERR_CAST(di); + } if (!di) { /* * Ok the default dir item isn't there. This is weird since @@ -610,6 +633,7 @@ static int btrfs_fill_super(struct super_block *sb, sb->s_root = root_dentry; save_mount_options(sb, data); + cleancache_init_fs(sb); return 0; fail_close: @@ -725,7 +749,7 @@ static int btrfs_set_super(struct super_block *s, void *data) * for multiple device setup. Make sure to keep it in sync. */ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, - const char *dev_name, void *data) + const char *device_name, void *data) { struct block_device *bdev = NULL; struct super_block *s; @@ -736,6 +760,7 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, fmode_t mode = FMODE_READ; char *subvol_name = NULL; u64 subvol_objectid = 0; + u64 subvol_rootid = 0; int error = 0; if (!(flags & MS_RDONLY)) @@ -743,11 +768,11 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, error = btrfs_parse_early_options(data, mode, fs_type, &subvol_name, &subvol_objectid, - &fs_devices); + &subvol_rootid, &fs_devices); if (error) return ERR_PTR(error); - error = btrfs_scan_one_device(dev_name, mode, fs_type, &fs_devices); + error = btrfs_scan_one_device(device_name, mode, fs_type, &fs_devices); if (error) goto error_free_subvol_name; @@ -807,15 +832,17 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, s->s_flags |= MS_ACTIVE; } - root = get_default_root(s, subvol_objectid); - if (IS_ERR(root)) { - error = PTR_ERR(root); - deactivate_locked_super(s); - goto error_free_subvol_name; - } /* if they gave us a subvolume name bind mount into that */ if (strcmp(subvol_name, ".")) { struct dentry *new_root; + + root = get_default_root(s, subvol_rootid); + if (IS_ERR(root)) { + error = PTR_ERR(root); + deactivate_locked_super(s); + goto error_free_subvol_name; + } + mutex_lock(&root->d_inode->i_mutex); new_root = lookup_one_len(subvol_name, root, strlen(subvol_name)); @@ -836,6 +863,13 @@ static struct dentry *btrfs_mount(struct file_system_type *fs_type, int flags, } dput(root); root = new_root; + } else { + root = get_default_root(s, subvol_objectid); + if (IS_ERR(root)) { + error = PTR_ERR(root); + deactivate_locked_super(s); + goto error_free_subvol_name; + } } kfree(subvol_name); @@ -889,6 +923,32 @@ static int btrfs_remount(struct super_block *sb, int *flags, char *data) return 0; } +/* Used to sort the devices by max_avail(descending sort) */ +static int btrfs_cmp_device_free_bytes(const void *dev_info1, + const void *dev_info2) +{ + if (((struct btrfs_device_info *)dev_info1)->max_avail > + ((struct btrfs_device_info *)dev_info2)->max_avail) + return -1; + else if (((struct btrfs_device_info *)dev_info1)->max_avail < + ((struct btrfs_device_info *)dev_info2)->max_avail) + return 1; + else + return 0; +} + +/* + * sort the devices by max_avail, in which max free extent size of each device + * is stored.(Descending Sort) + */ +static inline void btrfs_descending_sort_devices( + struct btrfs_device_info *devices, + size_t nr_devices) +{ + sort(devices, nr_devices, sizeof(struct btrfs_device_info), + btrfs_cmp_device_free_bytes, NULL); +} + /* * The helper to calc the free space on the devices that can be used to store * file data. @@ -1182,10 +1242,14 @@ static int __init init_btrfs_fs(void) if (err) goto free_extent_io; - err = btrfs_interface_init(); + err = btrfs_delayed_inode_init(); if (err) goto free_extent_map; + err = btrfs_interface_init(); + if (err) + goto free_delayed_inode; + err = register_filesystem(&btrfs_fs_type); if (err) goto unregister_ioctl; @@ -1195,6 +1259,8 @@ static int __init init_btrfs_fs(void) unregister_ioctl: btrfs_interface_exit(); +free_delayed_inode: + btrfs_delayed_inode_exit(); free_extent_map: extent_map_exit(); free_extent_io: @@ -1211,6 +1277,7 @@ free_sysfs: static void __exit exit_btrfs_fs(void) { btrfs_destroy_cachep(); + btrfs_delayed_inode_exit(); extent_map_exit(); extent_io_exit(); btrfs_interface_exit(); diff --git a/fs/btrfs/sysfs.c b/fs/btrfs/sysfs.c index 4ce16ef702a3..c3c223ae6691 100644 --- a/fs/btrfs/sysfs.c +++ b/fs/btrfs/sysfs.c @@ -174,86 +174,9 @@ static const struct sysfs_ops btrfs_root_attr_ops = { .store = btrfs_root_attr_store, }; -static struct kobj_type btrfs_root_ktype = { - .default_attrs = btrfs_root_attrs, - .sysfs_ops = &btrfs_root_attr_ops, - .release = btrfs_root_release, -}; - -static struct kobj_type btrfs_super_ktype = { - .default_attrs = btrfs_super_attrs, - .sysfs_ops = &btrfs_super_attr_ops, - .release = btrfs_super_release, -}; - /* /sys/fs/btrfs/ entry */ static struct kset *btrfs_kset; -int btrfs_sysfs_add_super(struct btrfs_fs_info *fs) -{ - int error; - char *name; - char c; - int len = strlen(fs->sb->s_id) + 1; - int i; - - name = kmalloc(len, GFP_NOFS); - if (!name) { - error = -ENOMEM; - goto fail; - } - - for (i = 0; i < len; i++) { - c = fs->sb->s_id[i]; - if (c == '/' || c == '\\') - c = '!'; - name[i] = c; - } - name[len] = '\0'; - - fs->super_kobj.kset = btrfs_kset; - error = kobject_init_and_add(&fs->super_kobj, &btrfs_super_ktype, - NULL, "%s", name); - kfree(name); - if (error) - goto fail; - - return 0; - -fail: - printk(KERN_ERR "btrfs: sysfs creation for super failed\n"); - return error; -} - -int btrfs_sysfs_add_root(struct btrfs_root *root) -{ - int error; - - error = kobject_init_and_add(&root->root_kobj, &btrfs_root_ktype, - &root->fs_info->super_kobj, - "%s", root->name); - if (error) - goto fail; - - return 0; - -fail: - printk(KERN_ERR "btrfs: sysfs creation for root failed\n"); - return error; -} - -void btrfs_sysfs_del_root(struct btrfs_root *root) -{ - kobject_put(&root->root_kobj); - wait_for_completion(&root->kobj_unregister); -} - -void btrfs_sysfs_del_super(struct btrfs_fs_info *fs) -{ - kobject_put(&fs->super_kobj); - wait_for_completion(&fs->kobj_unregister); -} - int btrfs_init_sysfs(void) { btrfs_kset = kset_create_and_add("btrfs", NULL, fs_kobj); diff --git a/fs/btrfs/transaction.c b/fs/btrfs/transaction.c index 5b158da7e0bb..dc80f7156923 100644 --- a/fs/btrfs/transaction.c +++ b/fs/btrfs/transaction.c @@ -27,15 +27,14 @@ #include "transaction.h" #include "locking.h" #include "tree-log.h" +#include "inode-map.h" #define BTRFS_ROOT_TRANS_TAG 0 static noinline void put_transaction(struct btrfs_transaction *transaction) { - WARN_ON(transaction->use_count == 0); - transaction->use_count--; - if (transaction->use_count == 0) { - list_del_init(&transaction->list); + WARN_ON(atomic_read(&transaction->use_count) == 0); + if (atomic_dec_and_test(&transaction->use_count)) { memset(transaction, 0, sizeof(*transaction)); kmem_cache_free(btrfs_transaction_cachep, transaction); } @@ -60,14 +59,14 @@ static noinline int join_transaction(struct btrfs_root *root) if (!cur_trans) return -ENOMEM; root->fs_info->generation++; - cur_trans->num_writers = 1; + atomic_set(&cur_trans->num_writers, 1); cur_trans->num_joined = 0; cur_trans->transid = root->fs_info->generation; init_waitqueue_head(&cur_trans->writer_wait); init_waitqueue_head(&cur_trans->commit_wait); cur_trans->in_commit = 0; cur_trans->blocked = 0; - cur_trans->use_count = 1; + atomic_set(&cur_trans->use_count, 1); cur_trans->commit_done = 0; cur_trans->start_time = get_seconds(); @@ -82,13 +81,12 @@ static noinline int join_transaction(struct btrfs_root *root) INIT_LIST_HEAD(&cur_trans->pending_snapshots); list_add_tail(&cur_trans->list, &root->fs_info->trans_list); extent_io_tree_init(&cur_trans->dirty_pages, - root->fs_info->btree_inode->i_mapping, - GFP_NOFS); + root->fs_info->btree_inode->i_mapping); spin_lock(&root->fs_info->new_trans_lock); root->fs_info->running_transaction = cur_trans; spin_unlock(&root->fs_info->new_trans_lock); } else { - cur_trans->num_writers++; + atomic_inc(&cur_trans->num_writers); cur_trans->num_joined++; } @@ -145,7 +143,7 @@ static void wait_current_trans(struct btrfs_root *root) cur_trans = root->fs_info->running_transaction; if (cur_trans && cur_trans->blocked) { DEFINE_WAIT(wait); - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); while (1) { prepare_to_wait(&root->fs_info->transaction_wait, &wait, TASK_UNINTERRUPTIBLE); @@ -181,6 +179,7 @@ static struct btrfs_trans_handle *start_transaction(struct btrfs_root *root, { struct btrfs_trans_handle *h; struct btrfs_transaction *cur_trans; + int retries = 0; int ret; if (root->fs_info->fs_state & BTRFS_SUPER_FLAG_ERROR) @@ -204,7 +203,7 @@ again: } cur_trans = root->fs_info->running_transaction; - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); if (type != TRANS_JOIN_NOLOCK) mutex_unlock(&root->fs_info->trans_mutex); @@ -224,10 +223,18 @@ again: if (num_items > 0) { ret = btrfs_trans_reserve_metadata(h, root, num_items); - if (ret == -EAGAIN) { + if (ret == -EAGAIN && !retries) { + retries++; btrfs_commit_transaction(h, root); goto again; + } else if (ret == -EAGAIN) { + /* + * We have already retried and got EAGAIN, so really we + * don't have space, so set ret to -ENOSPC. + */ + ret = -ENOSPC; } + if (ret < 0) { btrfs_end_transaction(h, root); return ERR_PTR(ret); @@ -327,7 +334,7 @@ int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid) goto out_unlock; /* nothing committing|committed */ } - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); wait_for_commit(root, cur_trans); @@ -340,49 +347,6 @@ out_unlock: return ret; } -#if 0 -/* - * rate limit against the drop_snapshot code. This helps to slow down new - * operations if the drop_snapshot code isn't able to keep up. - */ -static void throttle_on_drops(struct btrfs_root *root) -{ - struct btrfs_fs_info *info = root->fs_info; - int harder_count = 0; - -harder: - if (atomic_read(&info->throttles)) { - DEFINE_WAIT(wait); - int thr; - thr = atomic_read(&info->throttle_gen); - - do { - prepare_to_wait(&info->transaction_throttle, - &wait, TASK_UNINTERRUPTIBLE); - if (!atomic_read(&info->throttles)) { - finish_wait(&info->transaction_throttle, &wait); - break; - } - schedule(); - finish_wait(&info->transaction_throttle, &wait); - } while (thr == atomic_read(&info->throttle_gen)); - harder_count++; - - if (root->fs_info->total_ref_cache_size > 1 * 1024 * 1024 && - harder_count < 2) - goto harder; - - if (root->fs_info->total_ref_cache_size > 5 * 1024 * 1024 && - harder_count < 10) - goto harder; - - if (root->fs_info->total_ref_cache_size > 10 * 1024 * 1024 && - harder_count < 20) - goto harder; - } -} -#endif - void btrfs_throttle(struct btrfs_root *root) { mutex_lock(&root->fs_info->trans_mutex); @@ -457,18 +421,14 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, wake_up_process(info->transaction_kthread); } - if (lock) - mutex_lock(&info->trans_mutex); WARN_ON(cur_trans != info->running_transaction); - WARN_ON(cur_trans->num_writers < 1); - cur_trans->num_writers--; + WARN_ON(atomic_read(&cur_trans->num_writers) < 1); + atomic_dec(&cur_trans->num_writers); smp_mb(); if (waitqueue_active(&cur_trans->writer_wait)) wake_up(&cur_trans->writer_wait); put_transaction(cur_trans); - if (lock) - mutex_unlock(&info->trans_mutex); if (current->journal_info == trans) current->journal_info = NULL; @@ -484,19 +444,40 @@ static int __btrfs_end_transaction(struct btrfs_trans_handle *trans, int btrfs_end_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root) { - return __btrfs_end_transaction(trans, root, 0, 1); + int ret; + + ret = __btrfs_end_transaction(trans, root, 0, 1); + if (ret) + return ret; + return 0; } int btrfs_end_transaction_throttle(struct btrfs_trans_handle *trans, struct btrfs_root *root) { - return __btrfs_end_transaction(trans, root, 1, 1); + int ret; + + ret = __btrfs_end_transaction(trans, root, 1, 1); + if (ret) + return ret; + return 0; } int btrfs_end_transaction_nolock(struct btrfs_trans_handle *trans, struct btrfs_root *root) { - return __btrfs_end_transaction(trans, root, 0, 0); + int ret; + + ret = __btrfs_end_transaction(trans, root, 0, 0); + if (ret) + return ret; + return 0; +} + +int btrfs_end_transaction_dmeta(struct btrfs_trans_handle *trans, + struct btrfs_root *root) +{ + return __btrfs_end_transaction(trans, root, 1, 1); } /* @@ -757,8 +738,14 @@ static noinline int commit_fs_roots(struct btrfs_trans_handle *trans, btrfs_update_reloc_root(trans, root); btrfs_orphan_commit_root(trans, root); + btrfs_save_ino_cache(root, trans); + if (root->commit_root != root->node) { + mutex_lock(&root->fs_commit_mutex); switch_commit_root(root); + btrfs_unpin_free_ino(root); + mutex_unlock(&root->fs_commit_mutex); + btrfs_set_root_node(&root->root_item, root->node); } @@ -806,97 +793,6 @@ int btrfs_defrag_root(struct btrfs_root *root, int cacheonly) return ret; } -#if 0 -/* - * when dropping snapshots, we generate a ton of delayed refs, and it makes - * sense not to join the transaction while it is trying to flush the current - * queue of delayed refs out. - * - * This is used by the drop snapshot code only - */ -static noinline int wait_transaction_pre_flush(struct btrfs_fs_info *info) -{ - DEFINE_WAIT(wait); - - mutex_lock(&info->trans_mutex); - while (info->running_transaction && - info->running_transaction->delayed_refs.flushing) { - prepare_to_wait(&info->transaction_wait, &wait, - TASK_UNINTERRUPTIBLE); - mutex_unlock(&info->trans_mutex); - - schedule(); - - mutex_lock(&info->trans_mutex); - finish_wait(&info->transaction_wait, &wait); - } - mutex_unlock(&info->trans_mutex); - return 0; -} - -/* - * Given a list of roots that need to be deleted, call btrfs_drop_snapshot on - * all of them - */ -int btrfs_drop_dead_root(struct btrfs_root *root) -{ - struct btrfs_trans_handle *trans; - struct btrfs_root *tree_root = root->fs_info->tree_root; - unsigned long nr; - int ret; - - while (1) { - /* - * we don't want to jump in and create a bunch of - * delayed refs if the transaction is starting to close - */ - wait_transaction_pre_flush(tree_root->fs_info); - trans = btrfs_start_transaction(tree_root, 1); - - /* - * we've joined a transaction, make sure it isn't - * closing right now - */ - if (trans->transaction->delayed_refs.flushing) { - btrfs_end_transaction(trans, tree_root); - continue; - } - - ret = btrfs_drop_snapshot(trans, root); - if (ret != -EAGAIN) - break; - - ret = btrfs_update_root(trans, tree_root, - &root->root_key, - &root->root_item); - if (ret) - break; - - nr = trans->blocks_used; - ret = btrfs_end_transaction(trans, tree_root); - BUG_ON(ret); - - btrfs_btree_balance_dirty(tree_root, nr); - cond_resched(); - } - BUG_ON(ret); - - ret = btrfs_del_root(trans, tree_root, &root->root_key); - BUG_ON(ret); - - nr = trans->blocks_used; - ret = btrfs_end_transaction(trans, tree_root); - BUG_ON(ret); - - free_extent_buffer(root->node); - free_extent_buffer(root->commit_root); - kfree(root); - - btrfs_btree_balance_dirty(tree_root, nr); - return ret; -} -#endif - /* * new snapshots need to be created at a very specific time in the * transaction commit. This does the actual creation @@ -927,7 +823,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, goto fail; } - ret = btrfs_find_free_objectid(trans, tree_root, 0, &objectid); + ret = btrfs_find_free_objectid(tree_root, &objectid); if (ret) { pending->error = ret; goto fail; @@ -964,7 +860,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, BUG_ON(ret); ret = btrfs_insert_dir_item(trans, parent_root, dentry->d_name.name, dentry->d_name.len, - parent_inode->i_ino, &key, + parent_inode, &key, BTRFS_FT_DIR, index); BUG_ON(ret); @@ -1006,7 +902,7 @@ static noinline int create_pending_snapshot(struct btrfs_trans_handle *trans, */ ret = btrfs_add_root_ref(trans, tree_root, objectid, parent_root->root_key.objectid, - parent_inode->i_ino, index, + btrfs_ino(parent_inode), index, dentry->d_name.name, dentry->d_name.len); BUG_ON(ret); dput(parent); @@ -1034,6 +930,14 @@ static noinline int create_pending_snapshots(struct btrfs_trans_handle *trans, int ret; list_for_each_entry(pending, head, list) { + /* + * We must deal with the delayed items before creating + * snapshots, or we will create a snapthot with inconsistent + * information. + */ + ret = btrfs_run_delayed_items(trans, fs_info->fs_root); + BUG_ON(ret); + ret = create_pending_snapshot(trans, fs_info, pending); BUG_ON(ret); } @@ -1178,7 +1082,7 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans, /* take transaction reference */ mutex_lock(&root->fs_info->trans_mutex); cur_trans = trans->transaction; - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); btrfs_end_transaction(trans, root); @@ -1237,7 +1141,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, mutex_lock(&root->fs_info->trans_mutex); if (cur_trans->in_commit) { - cur_trans->use_count++; + atomic_inc(&cur_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); btrfs_end_transaction(trans, root); @@ -1259,7 +1163,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, prev_trans = list_entry(cur_trans->list.prev, struct btrfs_transaction, list); if (!prev_trans->commit_done) { - prev_trans->use_count++; + atomic_inc(&prev_trans->use_count); mutex_unlock(&root->fs_info->trans_mutex); wait_for_commit(root, prev_trans); @@ -1287,6 +1191,9 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, BUG_ON(ret); } + ret = btrfs_run_delayed_items(trans, root); + BUG_ON(ret); + /* * rename don't use btrfs_join_transaction, so, once we * set the transaction to blocked above, we aren't going @@ -1300,24 +1207,28 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, TASK_UNINTERRUPTIBLE); smp_mb(); - if (cur_trans->num_writers > 1) + if (atomic_read(&cur_trans->num_writers) > 1) schedule_timeout(MAX_SCHEDULE_TIMEOUT); else if (should_grow) schedule_timeout(1); mutex_lock(&root->fs_info->trans_mutex); finish_wait(&cur_trans->writer_wait, &wait); - } while (cur_trans->num_writers > 1 || + } while (atomic_read(&cur_trans->num_writers) > 1 || (should_grow && cur_trans->num_joined != joined)); ret = create_pending_snapshots(trans, root->fs_info); BUG_ON(ret); + ret = btrfs_run_delayed_items(trans, root); + BUG_ON(ret); + ret = btrfs_run_delayed_refs(trans, root, (unsigned long)-1); BUG_ON(ret); WARN_ON(cur_trans != trans->transaction); + btrfs_scrub_pause(root); /* btrfs_commit_tree_roots is responsible for getting the * various roots consistent with each other. Every pointer * in the tree of tree roots has to point to the most up to date @@ -1394,6 +1305,7 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, wake_up(&cur_trans->commit_wait); + list_del_init(&cur_trans->list); put_transaction(cur_trans); put_transaction(cur_trans); @@ -1401,6 +1313,8 @@ int btrfs_commit_transaction(struct btrfs_trans_handle *trans, mutex_unlock(&root->fs_info->trans_mutex); + btrfs_scrub_continue(root); + if (current->journal_info == trans) current->journal_info = NULL; @@ -1428,6 +1342,8 @@ int btrfs_clean_old_snapshots(struct btrfs_root *root) root = list_entry(list.next, struct btrfs_root, root_list); list_del(&root->root_list); + btrfs_kill_all_delayed_nodes(root); + if (btrfs_header_backref_rev(root->node) < BTRFS_MIXED_BACKREF_REV) btrfs_drop_snapshot(root, NULL, 0); diff --git a/fs/btrfs/transaction.h b/fs/btrfs/transaction.h index 229a594cacd5..804c88639e5d 100644 --- a/fs/btrfs/transaction.h +++ b/fs/btrfs/transaction.h @@ -27,11 +27,11 @@ struct btrfs_transaction { * total writers in this transaction, it must be zero before the * transaction can end */ - unsigned long num_writers; + atomic_t num_writers; unsigned long num_joined; int in_commit; - int use_count; + atomic_t use_count; int commit_done; int blocked; struct list_head list; @@ -101,11 +101,8 @@ struct btrfs_trans_handle *btrfs_start_ioctl_transaction(struct btrfs_root *r, int btrfs_wait_for_commit(struct btrfs_root *root, u64 transid); int btrfs_write_and_wait_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root); -int btrfs_commit_tree_roots(struct btrfs_trans_handle *trans, - struct btrfs_root *root); int btrfs_add_dead_root(struct btrfs_root *root); -int btrfs_drop_dead_root(struct btrfs_root *root); int btrfs_defrag_root(struct btrfs_root *root, int cacheonly); int btrfs_clean_old_snapshots(struct btrfs_root *root); int btrfs_commit_transaction(struct btrfs_trans_handle *trans, @@ -115,6 +112,8 @@ int btrfs_commit_transaction_async(struct btrfs_trans_handle *trans, int wait_for_unblock); int btrfs_end_transaction_throttle(struct btrfs_trans_handle *trans, struct btrfs_root *root); +int btrfs_end_transaction_dmeta(struct btrfs_trans_handle *trans, + struct btrfs_root *root); int btrfs_should_end_transaction(struct btrfs_trans_handle *trans, struct btrfs_root *root); void btrfs_throttle(struct btrfs_root *root); diff --git a/fs/btrfs/tree-defrag.c b/fs/btrfs/tree-defrag.c index 992ab425599d..3b580ee8ab1d 100644 --- a/fs/btrfs/tree-defrag.c +++ b/fs/btrfs/tree-defrag.c @@ -97,7 +97,7 @@ int btrfs_defrag_leaves(struct btrfs_trans_handle *trans, ret = 0; goto out; } - btrfs_release_path(root, path); + btrfs_release_path(path); wret = btrfs_search_slot(trans, root, &key, path, 0, 1); if (wret < 0) { diff --git a/fs/btrfs/tree-log.c b/fs/btrfs/tree-log.c index c50271ad3157..592396c6dc47 100644 --- a/fs/btrfs/tree-log.c +++ b/fs/btrfs/tree-log.c @@ -333,13 +333,13 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans, goto insert; if (item_size == 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); return 0; } dst_copy = kmalloc(item_size, GFP_NOFS); src_copy = kmalloc(item_size, GFP_NOFS); if (!dst_copy || !src_copy) { - btrfs_release_path(root, path); + btrfs_release_path(path); kfree(dst_copy); kfree(src_copy); return -ENOMEM; @@ -361,13 +361,13 @@ static noinline int overwrite_item(struct btrfs_trans_handle *trans, * sync */ if (ret == 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); return 0; } } insert: - btrfs_release_path(root, path); + btrfs_release_path(path); /* try to insert the key into the destination tree */ ret = btrfs_insert_empty_item(trans, root, path, key, item_size); @@ -382,7 +382,6 @@ insert: } else if (found_size < item_size) { ret = btrfs_extend_item(trans, root, path, item_size - found_size); - BUG_ON(ret); } } else if (ret) { return ret; @@ -438,7 +437,7 @@ insert: } no_copy: btrfs_mark_buffer_dirty(path->nodes[0]); - btrfs_release_path(root, path); + btrfs_release_path(path); return 0; } @@ -519,7 +518,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, * file. This must be done before the btrfs_drop_extents run * so we don't try to drop this extent. */ - ret = btrfs_lookup_file_extent(trans, root, path, inode->i_ino, + ret = btrfs_lookup_file_extent(trans, root, path, btrfs_ino(inode), start, 0); if (ret == 0 && @@ -544,11 +543,11 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, * we don't have to do anything */ if (memcmp(&cmp1, &cmp2, sizeof(cmp1)) == 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); goto out; } } - btrfs_release_path(root, path); + btrfs_release_path(path); saved_nbytes = inode_get_bytes(inode); /* drop any overlapping extents */ @@ -590,6 +589,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, ins.objectid, ins.offset, 0, root->root_key.objectid, key->objectid, offset); + BUG_ON(ret); } else { /* * insert the extent pointer in the extent @@ -600,7 +600,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, key->objectid, offset, &ins); BUG_ON(ret); } - btrfs_release_path(root, path); + btrfs_release_path(path); if (btrfs_file_extent_compression(eb, item)) { csum_start = ins.objectid; @@ -614,7 +614,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, ret = btrfs_lookup_csums_range(root->log_root, csum_start, csum_end - 1, - &ordered_sums); + &ordered_sums, 0); BUG_ON(ret); while (!list_empty(&ordered_sums)) { struct btrfs_ordered_sum *sums; @@ -629,7 +629,7 @@ static noinline int replay_one_extent(struct btrfs_trans_handle *trans, kfree(sums); } } else { - btrfs_release_path(root, path); + btrfs_release_path(path); } } else if (found_type == BTRFS_FILE_EXTENT_INLINE) { /* inline extents are easy, we just overwrite them */ @@ -675,10 +675,13 @@ static noinline int drop_one_dir_item(struct btrfs_trans_handle *trans, return -ENOMEM; read_extent_buffer(leaf, name, (unsigned long)(di + 1), name_len); - btrfs_release_path(root, path); + btrfs_release_path(path); inode = read_one_inode(root, location.objectid); - BUG_ON(!inode); + if (!inode) { + kfree(name); + return -EIO; + } ret = link_to_fixup_dir(trans, root, path, location.objectid); BUG_ON(ret); @@ -713,7 +716,7 @@ static noinline int inode_in_dir(struct btrfs_root *root, goto out; } else goto out; - btrfs_release_path(root, path); + btrfs_release_path(path); di = btrfs_lookup_dir_item(NULL, root, path, dirid, name, name_len, 0); if (di && !IS_ERR(di)) { @@ -724,7 +727,7 @@ static noinline int inode_in_dir(struct btrfs_root *root, goto out; match = 1; out: - btrfs_release_path(root, path); + btrfs_release_path(path); return match; } @@ -817,7 +820,10 @@ static noinline int add_inode_ref(struct btrfs_trans_handle *trans, return -ENOENT; inode = read_one_inode(root, key->objectid); - BUG_ON(!inode); + if (!inode) { + iput(dir); + return -EIO; + } ref_ptr = btrfs_item_ptr_offset(eb, slot); ref_end = ref_ptr + btrfs_item_size_nr(eb, slot); @@ -832,7 +838,7 @@ again: read_extent_buffer(eb, name, (unsigned long)(ref + 1), namelen); /* if we already have a perfect match, we're done */ - if (inode_in_dir(root, path, dir->i_ino, inode->i_ino, + if (inode_in_dir(root, path, btrfs_ino(dir), btrfs_ino(inode), btrfs_inode_ref_index(eb, ref), name, namelen)) { goto out; @@ -884,7 +890,7 @@ again: if (!backref_in_log(log, key, victim_name, victim_name_len)) { btrfs_inc_nlink(inode); - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_unlink_inode(trans, root, dir, inode, victim_name, @@ -901,7 +907,7 @@ again: */ search_done = 1; } - btrfs_release_path(root, path); + btrfs_release_path(path); insert: /* insert our name */ @@ -922,7 +928,7 @@ out: BUG_ON(ret); out_nowrite: - btrfs_release_path(root, path); + btrfs_release_path(path); iput(dir); iput(inode); return 0; @@ -960,8 +966,9 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, unsigned long ptr; unsigned long ptr_end; int name_len; + u64 ino = btrfs_ino(inode); - key.objectid = inode->i_ino; + key.objectid = ino; key.type = BTRFS_INODE_REF_KEY; key.offset = (u64)-1; @@ -980,7 +987,7 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, } btrfs_item_key_to_cpu(path->nodes[0], &key, path->slots[0]); - if (key.objectid != inode->i_ino || + if (key.objectid != ino || key.type != BTRFS_INODE_REF_KEY) break; ptr = btrfs_item_ptr_offset(path->nodes[0], path->slots[0]); @@ -999,9 +1006,9 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, if (key.offset == 0) break; key.offset--; - btrfs_release_path(root, path); + btrfs_release_path(path); } - btrfs_release_path(root, path); + btrfs_release_path(path); if (nlink != inode->i_nlink) { inode->i_nlink = nlink; btrfs_update_inode(trans, root, inode); @@ -1011,10 +1018,10 @@ static noinline int fixup_inode_link_count(struct btrfs_trans_handle *trans, if (inode->i_nlink == 0) { if (S_ISDIR(inode->i_mode)) { ret = replay_dir_deletes(trans, root, NULL, path, - inode->i_ino, 1); + ino, 1); BUG_ON(ret); } - ret = insert_orphan_item(trans, root, inode->i_ino); + ret = insert_orphan_item(trans, root, ino); BUG_ON(ret); } btrfs_free_path(path); @@ -1050,11 +1057,13 @@ static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans, break; ret = btrfs_del_item(trans, root, path); - BUG_ON(ret); + if (ret) + goto out; - btrfs_release_path(root, path); + btrfs_release_path(path); inode = read_one_inode(root, key.offset); - BUG_ON(!inode); + if (!inode) + return -EIO; ret = fixup_inode_link_count(trans, root, inode); BUG_ON(ret); @@ -1068,8 +1077,10 @@ static noinline int fixup_inode_link_counts(struct btrfs_trans_handle *trans, */ key.offset = (u64)-1; } - btrfs_release_path(root, path); - return 0; + ret = 0; +out: + btrfs_release_path(path); + return ret; } @@ -1088,7 +1099,8 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans, struct inode *inode; inode = read_one_inode(root, objectid); - BUG_ON(!inode); + if (!inode) + return -EIO; key.objectid = BTRFS_TREE_LOG_FIXUP_OBJECTID; btrfs_set_key_type(&key, BTRFS_ORPHAN_ITEM_KEY); @@ -1096,7 +1108,7 @@ static noinline int link_to_fixup_dir(struct btrfs_trans_handle *trans, ret = btrfs_insert_empty_item(trans, root, path, &key, 0); - btrfs_release_path(root, path); + btrfs_release_path(path); if (ret == 0) { btrfs_inc_nlink(inode); btrfs_update_inode(trans, root, inode); @@ -1175,7 +1187,8 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, int ret; dir = read_one_inode(root, key->objectid); - BUG_ON(!dir); + if (!dir) + return -EIO; name_len = btrfs_dir_name_len(eb, di); name = kmalloc(name_len, GFP_NOFS); @@ -1192,7 +1205,7 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, exists = 1; else exists = 0; - btrfs_release_path(root, path); + btrfs_release_path(path); if (key->type == BTRFS_DIR_ITEM_KEY) { dst_di = btrfs_lookup_dir_item(trans, root, path, key->objectid, @@ -1205,7 +1218,7 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, } else { BUG(); } - if (!dst_di || IS_ERR(dst_di)) { + if (IS_ERR_OR_NULL(dst_di)) { /* we need a sequence number to insert, so we only * do inserts for the BTRFS_DIR_INDEX_KEY types */ @@ -1236,13 +1249,13 @@ static noinline int replay_one_name(struct btrfs_trans_handle *trans, if (key->type == BTRFS_DIR_INDEX_KEY) goto insert; out: - btrfs_release_path(root, path); + btrfs_release_path(path); kfree(name); iput(dir); return 0; insert: - btrfs_release_path(root, path); + btrfs_release_path(path); ret = insert_one_name(trans, root, path, key->objectid, key->offset, name, name_len, log_type, &log_key); @@ -1363,7 +1376,7 @@ next: *end_ret = found_end; ret = 0; out: - btrfs_release_path(root, path); + btrfs_release_path(path); return ret; } @@ -1426,12 +1439,15 @@ again: dir_key->offset, name, name_len, 0); } - if (!log_di || IS_ERR(log_di)) { + if (IS_ERR_OR_NULL(log_di)) { btrfs_dir_item_key_to_cpu(eb, di, &location); - btrfs_release_path(root, path); - btrfs_release_path(log, log_path); + btrfs_release_path(path); + btrfs_release_path(log_path); inode = read_one_inode(root, location.objectid); - BUG_ON(!inode); + if (!inode) { + kfree(name); + return -EIO; + } ret = link_to_fixup_dir(trans, root, path, location.objectid); @@ -1453,7 +1469,7 @@ again: ret = 0; goto out; } - btrfs_release_path(log, log_path); + btrfs_release_path(log_path); kfree(name); ptr = (unsigned long)(di + 1); @@ -1461,8 +1477,8 @@ again: } ret = 0; out: - btrfs_release_path(root, path); - btrfs_release_path(log, log_path); + btrfs_release_path(path); + btrfs_release_path(log_path); return ret; } @@ -1550,7 +1566,7 @@ again: break; dir_key.offset = found_key.offset + 1; } - btrfs_release_path(root, path); + btrfs_release_path(path); if (range_end == (u64)-1) break; range_start = range_end + 1; @@ -1561,11 +1577,11 @@ next_type: if (key_type == BTRFS_DIR_LOG_ITEM_KEY) { key_type = BTRFS_DIR_LOG_INDEX_KEY; dir_key.type = BTRFS_DIR_INDEX_KEY; - btrfs_release_path(root, path); + btrfs_release_path(path); goto again; } out: - btrfs_release_path(root, path); + btrfs_release_path(path); btrfs_free_path(log_path); iput(dir); return ret; @@ -2093,7 +2109,9 @@ int btrfs_sync_log(struct btrfs_trans_handle *trans, * the running transaction open, so a full commit can't hop * in and cause problems either. */ + btrfs_scrub_pause_super(root); write_ctree_super(trans, root->fs_info->tree_root, 1); + btrfs_scrub_continue_super(root); ret = 0; mutex_lock(&root->log_mutex); @@ -2197,6 +2215,7 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, int ret; int err = 0; int bytes_del = 0; + u64 dir_ino = btrfs_ino(dir); if (BTRFS_I(dir)->logged_trans < trans->transid) return 0; @@ -2209,10 +2228,12 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, log = root->log_root; path = btrfs_alloc_path(); - if (!path) - return -ENOMEM; + if (!path) { + err = -ENOMEM; + goto out_unlock; + } - di = btrfs_lookup_dir_item(trans, log, path, dir->i_ino, + di = btrfs_lookup_dir_item(trans, log, path, dir_ino, name, name_len, -1); if (IS_ERR(di)) { err = PTR_ERR(di); @@ -2223,8 +2244,8 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, bytes_del += name_len; BUG_ON(ret); } - btrfs_release_path(log, path); - di = btrfs_lookup_dir_index_item(trans, log, path, dir->i_ino, + btrfs_release_path(path); + di = btrfs_lookup_dir_index_item(trans, log, path, dir_ino, index, name, name_len, -1); if (IS_ERR(di)) { err = PTR_ERR(di); @@ -2242,10 +2263,10 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, if (bytes_del) { struct btrfs_key key; - key.objectid = dir->i_ino; + key.objectid = dir_ino; key.offset = 0; key.type = BTRFS_INODE_ITEM_KEY; - btrfs_release_path(log, path); + btrfs_release_path(path); ret = btrfs_search_slot(trans, log, &key, path, 0, 1); if (ret < 0) { @@ -2267,10 +2288,11 @@ int btrfs_del_dir_entries_in_log(struct btrfs_trans_handle *trans, btrfs_mark_buffer_dirty(path->nodes[0]); } else ret = 0; - btrfs_release_path(log, path); + btrfs_release_path(path); } fail: btrfs_free_path(path); +out_unlock: mutex_unlock(&BTRFS_I(dir)->log_mutex); if (ret == -ENOSPC) { root->fs_info->last_trans_log_full_commit = trans->transid; @@ -2300,7 +2322,7 @@ int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans, log = root->log_root; mutex_lock(&BTRFS_I(inode)->log_mutex); - ret = btrfs_del_inode_ref(trans, log, name, name_len, inode->i_ino, + ret = btrfs_del_inode_ref(trans, log, name, name_len, btrfs_ino(inode), dirid, &index); mutex_unlock(&BTRFS_I(inode)->log_mutex); if (ret == -ENOSPC) { @@ -2341,7 +2363,7 @@ static noinline int insert_dir_log_key(struct btrfs_trans_handle *trans, struct btrfs_dir_log_item); btrfs_set_dir_log_end(path->nodes[0], item, last_offset); btrfs_mark_buffer_dirty(path->nodes[0]); - btrfs_release_path(log, path); + btrfs_release_path(path); return 0; } @@ -2366,13 +2388,14 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, int nritems; u64 first_offset = min_offset; u64 last_offset = (u64)-1; + u64 ino = btrfs_ino(inode); log = root->log_root; - max_key.objectid = inode->i_ino; + max_key.objectid = ino; max_key.offset = (u64)-1; max_key.type = key_type; - min_key.objectid = inode->i_ino; + min_key.objectid = ino; min_key.type = key_type; min_key.offset = min_offset; @@ -2385,18 +2408,17 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, * we didn't find anything from this transaction, see if there * is anything at all */ - if (ret != 0 || min_key.objectid != inode->i_ino || - min_key.type != key_type) { - min_key.objectid = inode->i_ino; + if (ret != 0 || min_key.objectid != ino || min_key.type != key_type) { + min_key.objectid = ino; min_key.type = key_type; min_key.offset = (u64)-1; - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); if (ret < 0) { - btrfs_release_path(root, path); + btrfs_release_path(path); return ret; } - ret = btrfs_previous_item(root, path, inode->i_ino, key_type); + ret = btrfs_previous_item(root, path, ino, key_type); /* if ret == 0 there are items for this type, * create a range to tell us the last key of this type. @@ -2414,7 +2436,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, } /* go backward to find any previous key */ - ret = btrfs_previous_item(root, path, inode->i_ino, key_type); + ret = btrfs_previous_item(root, path, ino, key_type); if (ret == 0) { struct btrfs_key tmp; btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]); @@ -2429,7 +2451,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, } } } - btrfs_release_path(root, path); + btrfs_release_path(path); /* find the first key from this transaction again */ ret = btrfs_search_slot(NULL, root, &min_key, path, 0, 0); @@ -2449,8 +2471,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, for (i = path->slots[0]; i < nritems; i++) { btrfs_item_key_to_cpu(src, &min_key, i); - if (min_key.objectid != inode->i_ino || - min_key.type != key_type) + if (min_key.objectid != ino || min_key.type != key_type) goto done; ret = overwrite_item(trans, log, dst_path, src, i, &min_key); @@ -2471,7 +2492,7 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, goto done; } btrfs_item_key_to_cpu(path->nodes[0], &tmp, path->slots[0]); - if (tmp.objectid != inode->i_ino || tmp.type != key_type) { + if (tmp.objectid != ino || tmp.type != key_type) { last_offset = (u64)-1; goto done; } @@ -2487,8 +2508,8 @@ static noinline int log_dir_items(struct btrfs_trans_handle *trans, } } done: - btrfs_release_path(root, path); - btrfs_release_path(log, dst_path); + btrfs_release_path(path); + btrfs_release_path(dst_path); if (err == 0) { *last_offset_ret = last_offset; @@ -2497,8 +2518,7 @@ done: * is valid */ ret = insert_dir_log_key(trans, log, path, key_type, - inode->i_ino, first_offset, - last_offset); + ino, first_offset, last_offset); if (ret) err = ret; } @@ -2584,10 +2604,11 @@ static int drop_objectid_items(struct btrfs_trans_handle *trans, break; ret = btrfs_del_item(trans, log, path); - BUG_ON(ret); - btrfs_release_path(log, path); + if (ret) + break; + btrfs_release_path(path); } - btrfs_release_path(log, path); + btrfs_release_path(path); return ret; } @@ -2662,6 +2683,9 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, extent = btrfs_item_ptr(src, start_slot + i, struct btrfs_file_extent_item); + if (btrfs_file_extent_generation(src, extent) < trans->transid) + continue; + found_type = btrfs_file_extent_type(src, extent); if (found_type == BTRFS_FILE_EXTENT_REG || found_type == BTRFS_FILE_EXTENT_PREALLOC) { @@ -2686,14 +2710,14 @@ static noinline int copy_items(struct btrfs_trans_handle *trans, ret = btrfs_lookup_csums_range( log->fs_info->csum_root, ds + cs, ds + cs + cl - 1, - &ordered_sums); + &ordered_sums, 0); BUG_ON(ret); } } } btrfs_mark_buffer_dirty(dst_path->nodes[0]); - btrfs_release_path(log, dst_path); + btrfs_release_path(dst_path); kfree(ins_data); /* @@ -2742,6 +2766,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, int nritems; int ins_start_slot = 0; int ins_nr; + u64 ino = btrfs_ino(inode); log = root->log_root; @@ -2754,11 +2779,11 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, return -ENOMEM; } - min_key.objectid = inode->i_ino; + min_key.objectid = ino; min_key.type = BTRFS_INODE_ITEM_KEY; min_key.offset = 0; - max_key.objectid = inode->i_ino; + max_key.objectid = ino; /* today the code can only do partial logging of directories */ if (!S_ISDIR(inode->i_mode)) @@ -2770,6 +2795,13 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, max_key.type = (u8)-1; max_key.offset = (u64)-1; + ret = btrfs_commit_inode_delayed_items(trans, inode); + if (ret) { + btrfs_free_path(path); + btrfs_free_path(dst_path); + return ret; + } + mutex_lock(&BTRFS_I(inode)->log_mutex); /* @@ -2781,8 +2813,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, if (inode_only == LOG_INODE_EXISTS) max_key_type = BTRFS_XATTR_ITEM_KEY; - ret = drop_objectid_items(trans, log, path, - inode->i_ino, max_key_type); + ret = drop_objectid_items(trans, log, path, ino, max_key_type); } else { ret = btrfs_truncate_inode_items(trans, log, inode, 0, 0); } @@ -2800,7 +2831,7 @@ static int btrfs_log_inode(struct btrfs_trans_handle *trans, break; again: /* note, ins_nr might be > 0 here, cleanup outside the loop */ - if (min_key.objectid != inode->i_ino) + if (min_key.objectid != ino) break; if (min_key.type > max_key.type) break; @@ -2842,7 +2873,7 @@ next_slot: } ins_nr = 0; } - btrfs_release_path(root, path); + btrfs_release_path(path); if (min_key.offset < (u64)-1) min_key.offset++; @@ -2865,8 +2896,8 @@ next_slot: } WARN_ON(ins_nr); if (inode_only == LOG_INODE_ALL && S_ISDIR(inode->i_mode)) { - btrfs_release_path(root, path); - btrfs_release_path(log, dst_path); + btrfs_release_path(path); + btrfs_release_path(dst_path); ret = log_directory_changes(trans, root, inode, path, dst_path); if (ret) { err = ret; @@ -3133,7 +3164,7 @@ again: } btrfs_item_key_to_cpu(path->nodes[0], &found_key, path->slots[0]); - btrfs_release_path(log_root_tree, path); + btrfs_release_path(path); if (found_key.objectid != BTRFS_TREE_LOG_OBJECTID) break; @@ -3168,7 +3199,7 @@ again: if (found_key.offset == 0) break; } - btrfs_release_path(log_root_tree, path); + btrfs_release_path(path); /* step one is to pin it all, step two is to replay just inodes */ if (wc.pin) { diff --git a/fs/btrfs/tree-log.h b/fs/btrfs/tree-log.h index 3dfae84c8cc8..2270ac58d746 100644 --- a/fs/btrfs/tree-log.h +++ b/fs/btrfs/tree-log.h @@ -38,7 +38,6 @@ int btrfs_del_inode_ref_in_log(struct btrfs_trans_handle *trans, struct btrfs_root *root, const char *name, int name_len, struct inode *inode, u64 dirid); -int btrfs_join_running_log_trans(struct btrfs_root *root); int btrfs_end_log_trans(struct btrfs_root *root); int btrfs_pin_log_trans(struct btrfs_root *root); int btrfs_log_inode_parent(struct btrfs_trans_handle *trans, diff --git a/fs/btrfs/version.sh b/fs/btrfs/version.sh deleted file mode 100644 index 1ca1952fd917..000000000000 --- a/fs/btrfs/version.sh +++ /dev/null @@ -1,43 +0,0 @@ -#!/bin/bash -# -# determine-version -- report a useful version for releases -# -# Copyright 2008, Aron Griffis <agriffis@n01se.net> -# Copyright 2008, Oracle -# Released under the GNU GPLv2 - -v="v0.16" - -which git &> /dev/null -if [ $? == 0 ]; then - git branch >& /dev/null - if [ $? == 0 ]; then - if head=`git rev-parse --verify HEAD 2>/dev/null`; then - if tag=`git describe --tags 2>/dev/null`; then - v="$tag" - fi - - # Are there uncommitted changes? - git update-index --refresh --unmerged > /dev/null - if git diff-index --name-only HEAD | \ - grep -v "^scripts/package" \ - | read dummy; then - v="$v"-dirty - fi - fi - fi -fi - -echo "#ifndef __BUILD_VERSION" > .build-version.h -echo "#define __BUILD_VERSION" >> .build-version.h -echo "#define BTRFS_BUILD_VERSION \"Btrfs $v\"" >> .build-version.h -echo "#endif" >> .build-version.h - -diff -q version.h .build-version.h >& /dev/null - -if [ $? == 0 ]; then - rm .build-version.h - exit 0 -fi - -mv .build-version.h version.h diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 309a57b9fc85..c48214ef5c09 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -38,22 +38,9 @@ static int init_first_rw_device(struct btrfs_trans_handle *trans, struct btrfs_device *device); static int btrfs_relocate_sys_chunks(struct btrfs_root *root); -#define map_lookup_size(n) (sizeof(struct map_lookup) + \ - (sizeof(struct btrfs_bio_stripe) * (n))) - static DEFINE_MUTEX(uuid_mutex); static LIST_HEAD(fs_uuids); -void btrfs_lock_volumes(void) -{ - mutex_lock(&uuid_mutex); -} - -void btrfs_unlock_volumes(void) -{ - mutex_unlock(&uuid_mutex); -} - static void lock_chunks(struct btrfs_root *root) { mutex_lock(&root->fs_info->chunk_mutex); @@ -155,6 +142,15 @@ static noinline int run_scheduled_bios(struct btrfs_device *device) unsigned long limit; unsigned long last_waited = 0; int force_reg = 0; + struct blk_plug plug; + + /* + * this function runs all the bios we've collected for + * a particular device. We don't want to wander off to + * another device without first sending all of these down. + * So, setup a plug here and finish it off before we return + */ + blk_start_plug(&plug); bdi = blk_get_backing_dev_info(device->bdev); fs_info = device->dev_root->fs_info; @@ -294,6 +290,7 @@ loop_lock: spin_unlock(&device->io_lock); done: + blk_finish_plug(&plug); return 0; } @@ -353,7 +350,7 @@ static noinline int device_list_add(const char *path, INIT_LIST_HEAD(&device->dev_alloc_list); mutex_lock(&fs_devices->device_list_mutex); - list_add(&device->dev_list, &fs_devices->devices); + list_add_rcu(&device->dev_list, &fs_devices->devices); mutex_unlock(&fs_devices->device_list_mutex); device->fs_devices = fs_devices; @@ -396,7 +393,7 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig) fs_devices->latest_trans = orig->latest_trans; memcpy(fs_devices->fsid, orig->fsid, sizeof(fs_devices->fsid)); - mutex_lock(&orig->device_list_mutex); + /* We have held the volume lock, it is safe to get the devices. */ list_for_each_entry(orig_dev, &orig->devices, dev_list) { device = kzalloc(sizeof(*device), GFP_NOFS); if (!device) @@ -419,10 +416,8 @@ static struct btrfs_fs_devices *clone_fs_devices(struct btrfs_fs_devices *orig) device->fs_devices = fs_devices; fs_devices->num_devices++; } - mutex_unlock(&orig->device_list_mutex); return fs_devices; error: - mutex_unlock(&orig->device_list_mutex); free_fs_devices(fs_devices); return ERR_PTR(-ENOMEM); } @@ -433,7 +428,7 @@ int btrfs_close_extra_devices(struct btrfs_fs_devices *fs_devices) mutex_lock(&uuid_mutex); again: - mutex_lock(&fs_devices->device_list_mutex); + /* This is the initialized path, it is safe to release the devices. */ list_for_each_entry_safe(device, next, &fs_devices->devices, dev_list) { if (device->in_fs_metadata) continue; @@ -453,7 +448,6 @@ again: kfree(device->name); kfree(device); } - mutex_unlock(&fs_devices->device_list_mutex); if (fs_devices->seed) { fs_devices = fs_devices->seed; @@ -464,6 +458,29 @@ again: return 0; } +static void __free_device(struct work_struct *work) +{ + struct btrfs_device *device; + + device = container_of(work, struct btrfs_device, rcu_work); + + if (device->bdev) + blkdev_put(device->bdev, device->mode); + + kfree(device->name); + kfree(device); +} + +static void free_device(struct rcu_head *head) +{ + struct btrfs_device *device; + + device = container_of(head, struct btrfs_device, rcu); + + INIT_WORK(&device->rcu_work, __free_device); + schedule_work(&device->rcu_work); +} + static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices) { struct btrfs_device *device; @@ -471,20 +488,32 @@ static int __btrfs_close_devices(struct btrfs_fs_devices *fs_devices) if (--fs_devices->opened > 0) return 0; + mutex_lock(&fs_devices->device_list_mutex); list_for_each_entry(device, &fs_devices->devices, dev_list) { - if (device->bdev) { - blkdev_put(device->bdev, device->mode); + struct btrfs_device *new_device; + + if (device->bdev) fs_devices->open_devices--; - } + if (device->writeable) { list_del_init(&device->dev_alloc_list); fs_devices->rw_devices--; } - device->bdev = NULL; - device->writeable = 0; - device->in_fs_metadata = 0; + new_device = kmalloc(sizeof(*new_device), GFP_NOFS); + BUG_ON(!new_device); + memcpy(new_device, device, sizeof(*new_device)); + new_device->name = kstrdup(device->name, GFP_NOFS); + BUG_ON(!new_device->name); + new_device->bdev = NULL; + new_device->writeable = 0; + new_device->in_fs_metadata = 0; + list_replace_rcu(&device->dev_list, &new_device->dev_list); + + call_rcu(&device->rcu, free_device); } + mutex_unlock(&fs_devices->device_list_mutex); + WARN_ON(fs_devices->open_devices); WARN_ON(fs_devices->rw_devices); fs_devices->opened = 0; @@ -587,6 +616,7 @@ static int __btrfs_open_devices(struct btrfs_fs_devices *fs_devices, list_add(&device->dev_alloc_list, &fs_devices->alloc_list); } + brelse(bh); continue; error_brelse: @@ -805,10 +835,7 @@ int find_free_dev_extent(struct btrfs_trans_handle *trans, /* we don't want to overwrite the superblock on the drive, * so we make sure to start at an offset of at least 1MB */ - search_start = 1024 * 1024; - - if (root->fs_info->alloc_start + num_bytes <= search_end) - search_start = max(root->fs_info->alloc_start, search_start); + search_start = max(root->fs_info->alloc_start, 1024ull * 1024); max_hole_start = search_start; max_hole_size = 0; @@ -939,14 +966,14 @@ static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans, if (ret > 0) { ret = btrfs_previous_item(root, path, key.objectid, BTRFS_DEV_EXTENT_KEY); - BUG_ON(ret); + if (ret) + goto out; leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &found_key, path->slots[0]); extent = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_dev_extent); BUG_ON(found_key.offset > start || found_key.offset + btrfs_dev_extent_length(leaf, extent) < start); - ret = 0; } else if (ret == 0) { leaf = path->nodes[0]; extent = btrfs_item_ptr(leaf, path->slots[0], @@ -957,8 +984,8 @@ static int btrfs_free_dev_extent(struct btrfs_trans_handle *trans, if (device->bytes_used > 0) device->bytes_used -= btrfs_dev_extent_length(leaf, extent); ret = btrfs_del_item(trans, root, path); - BUG_ON(ret); +out: btrfs_free_path(path); return ret; } @@ -1193,11 +1220,13 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) struct block_device *bdev; struct buffer_head *bh = NULL; struct btrfs_super_block *disk_super; + struct btrfs_fs_devices *cur_devices; u64 all_avail; u64 devid; u64 num_devices; u8 *dev_uuid; int ret = 0; + bool clear_super = false; mutex_lock(&uuid_mutex); mutex_lock(&root->fs_info->volume_mutex); @@ -1228,14 +1257,16 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) device = NULL; devices = &root->fs_info->fs_devices->devices; - mutex_lock(&root->fs_info->fs_devices->device_list_mutex); + /* + * It is safe to read the devices since the volume_mutex + * is held. + */ list_for_each_entry(tmp, devices, dev_list) { if (tmp->in_fs_metadata && !tmp->bdev) { device = tmp; break; } } - mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); bdev = NULL; bh = NULL; disk_super = NULL; @@ -1277,8 +1308,11 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) } if (device->writeable) { + lock_chunks(root); list_del_init(&device->dev_alloc_list); + unlock_chunks(root); root->fs_info->fs_devices->rw_devices--; + clear_super = true; } ret = btrfs_shrink_device(device, 0); @@ -1290,15 +1324,17 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) goto error_undo; device->in_fs_metadata = 0; + btrfs_scrub_cancel_dev(root, device); /* * the device list mutex makes sure that we don't change * the device list while someone else is writing out all * the device supers. */ + + cur_devices = device->fs_devices; mutex_lock(&root->fs_info->fs_devices->device_list_mutex); - list_del_init(&device->dev_list); - mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + list_del_rcu(&device->dev_list); device->fs_devices->num_devices--; @@ -1312,34 +1348,36 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) if (device->bdev == root->fs_info->fs_devices->latest_bdev) root->fs_info->fs_devices->latest_bdev = next_device->bdev; - if (device->bdev) { - blkdev_put(device->bdev, device->mode); - device->bdev = NULL; + if (device->bdev) device->fs_devices->open_devices--; - } + + call_rcu(&device->rcu, free_device); + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); num_devices = btrfs_super_num_devices(&root->fs_info->super_copy) - 1; btrfs_set_super_num_devices(&root->fs_info->super_copy, num_devices); - if (device->fs_devices->open_devices == 0) { + if (cur_devices->open_devices == 0) { struct btrfs_fs_devices *fs_devices; fs_devices = root->fs_info->fs_devices; while (fs_devices) { - if (fs_devices->seed == device->fs_devices) + if (fs_devices->seed == cur_devices) break; fs_devices = fs_devices->seed; } - fs_devices->seed = device->fs_devices->seed; - device->fs_devices->seed = NULL; - __btrfs_close_devices(device->fs_devices); - free_fs_devices(device->fs_devices); + fs_devices->seed = cur_devices->seed; + cur_devices->seed = NULL; + lock_chunks(root); + __btrfs_close_devices(cur_devices); + unlock_chunks(root); + free_fs_devices(cur_devices); } /* * at this point, the device is zero sized. We want to * remove it from the devices list and zero out the old super */ - if (device->writeable) { + if (clear_super) { /* make sure this device isn't detected as part of * the FS anymore */ @@ -1348,8 +1386,6 @@ int btrfs_rm_device(struct btrfs_root *root, char *device_path) sync_dirty_buffer(bh); } - kfree(device->name); - kfree(device); ret = 0; error_brelse: @@ -1363,8 +1399,10 @@ out: return ret; error_undo: if (device->writeable) { + lock_chunks(root); list_add(&device->dev_alloc_list, &root->fs_info->fs_devices->alloc_list); + unlock_chunks(root); root->fs_info->fs_devices->rw_devices++; } goto error_brelse; @@ -1404,7 +1442,12 @@ static int btrfs_prepare_sprout(struct btrfs_trans_handle *trans, INIT_LIST_HEAD(&seed_devices->devices); INIT_LIST_HEAD(&seed_devices->alloc_list); mutex_init(&seed_devices->device_list_mutex); - list_splice_init(&fs_devices->devices, &seed_devices->devices); + + mutex_lock(&root->fs_info->fs_devices->device_list_mutex); + list_splice_init_rcu(&fs_devices->devices, &seed_devices->devices, + synchronize_rcu); + mutex_unlock(&root->fs_info->fs_devices->device_list_mutex); + list_splice_init(&fs_devices->alloc_list, &seed_devices->alloc_list); list_for_each_entry(device, &seed_devices->devices, dev_list) { device->fs_devices = seed_devices; @@ -1465,7 +1508,7 @@ next_slot: goto error; leaf = path->nodes[0]; btrfs_item_key_to_cpu(leaf, &key, path->slots[0]); - btrfs_release_path(root, path); + btrfs_release_path(path); continue; } @@ -1601,7 +1644,7 @@ int btrfs_init_new_device(struct btrfs_root *root, char *device_path) * half setup */ mutex_lock(&root->fs_info->fs_devices->device_list_mutex); - list_add(&device->dev_list, &root->fs_info->fs_devices->devices); + list_add_rcu(&device->dev_list, &root->fs_info->fs_devices->devices); list_add(&device->dev_alloc_list, &root->fs_info->fs_devices->alloc_list); root->fs_info->fs_devices->num_devices++; @@ -1759,10 +1802,9 @@ static int btrfs_free_chunk(struct btrfs_trans_handle *trans, BUG_ON(ret); ret = btrfs_del_item(trans, root, path); - BUG_ON(ret); btrfs_free_path(path); - return 0; + return ret; } static int btrfs_del_sys_chunk(struct btrfs_root *root, u64 chunk_objectid, u64 @@ -1937,7 +1979,7 @@ again: chunk = btrfs_item_ptr(leaf, path->slots[0], struct btrfs_chunk); chunk_type = btrfs_chunk_type(leaf, chunk); - btrfs_release_path(chunk_root, path); + btrfs_release_path(path); if (chunk_type & BTRFS_BLOCK_GROUP_SYSTEM) { ret = btrfs_relocate_chunk(chunk_root, chunk_tree, @@ -2055,7 +2097,7 @@ int btrfs_balance(struct btrfs_root *dev_root) if (found_key.offset == 0) break; - btrfs_release_path(chunk_root, path); + btrfs_release_path(path); ret = btrfs_relocate_chunk(chunk_root, chunk_root->root_key.objectid, found_key.objectid, @@ -2127,7 +2169,7 @@ again: goto done; if (ret) { ret = 0; - btrfs_release_path(root, path); + btrfs_release_path(path); break; } @@ -2136,7 +2178,7 @@ again: btrfs_item_key_to_cpu(l, &key, path->slots[0]); if (key.objectid != device->devid) { - btrfs_release_path(root, path); + btrfs_release_path(path); break; } @@ -2144,14 +2186,14 @@ again: length = btrfs_dev_extent_length(l, dev_extent); if (key.offset + length <= new_size) { - btrfs_release_path(root, path); + btrfs_release_path(path); break; } chunk_tree = btrfs_dev_extent_chunk_tree(l, dev_extent); chunk_objectid = btrfs_dev_extent_chunk_objectid(l, dev_extent); chunk_offset = btrfs_dev_extent_chunk_offset(l, dev_extent); - btrfs_release_path(root, path); + btrfs_release_path(path); ret = btrfs_relocate_chunk(root, chunk_tree, chunk_objectid, chunk_offset); @@ -2227,275 +2269,204 @@ static int btrfs_add_system_chunk(struct btrfs_trans_handle *trans, return 0; } -static noinline u64 chunk_bytes_by_type(u64 type, u64 calc_size, - int num_stripes, int sub_stripes) +/* + * sort the devices in descending order by max_avail, total_avail + */ +static int btrfs_cmp_device_info(const void *a, const void *b) { - if (type & (BTRFS_BLOCK_GROUP_RAID1 | BTRFS_BLOCK_GROUP_DUP)) - return calc_size; - else if (type & BTRFS_BLOCK_GROUP_RAID10) - return calc_size * (num_stripes / sub_stripes); - else - return calc_size * num_stripes; -} + const struct btrfs_device_info *di_a = a; + const struct btrfs_device_info *di_b = b; -/* Used to sort the devices by max_avail(descending sort) */ -int btrfs_cmp_device_free_bytes(const void *dev_info1, const void *dev_info2) -{ - if (((struct btrfs_device_info *)dev_info1)->max_avail > - ((struct btrfs_device_info *)dev_info2)->max_avail) + if (di_a->max_avail > di_b->max_avail) return -1; - else if (((struct btrfs_device_info *)dev_info1)->max_avail < - ((struct btrfs_device_info *)dev_info2)->max_avail) + if (di_a->max_avail < di_b->max_avail) return 1; - else - return 0; + if (di_a->total_avail > di_b->total_avail) + return -1; + if (di_a->total_avail < di_b->total_avail) + return 1; + return 0; } -static int __btrfs_calc_nstripes(struct btrfs_fs_devices *fs_devices, u64 type, - int *num_stripes, int *min_stripes, - int *sub_stripes) +static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, + struct btrfs_root *extent_root, + struct map_lookup **map_ret, + u64 *num_bytes_out, u64 *stripe_size_out, + u64 start, u64 type) { - *num_stripes = 1; - *min_stripes = 1; - *sub_stripes = 0; + struct btrfs_fs_info *info = extent_root->fs_info; + struct btrfs_fs_devices *fs_devices = info->fs_devices; + struct list_head *cur; + struct map_lookup *map = NULL; + struct extent_map_tree *em_tree; + struct extent_map *em; + struct btrfs_device_info *devices_info = NULL; + u64 total_avail; + int num_stripes; /* total number of stripes to allocate */ + int sub_stripes; /* sub_stripes info for map */ + int dev_stripes; /* stripes per dev */ + int devs_max; /* max devs to use */ + int devs_min; /* min devs needed */ + int devs_increment; /* ndevs has to be a multiple of this */ + int ncopies; /* how many copies to data has */ + int ret; + u64 max_stripe_size; + u64 max_chunk_size; + u64 stripe_size; + u64 num_bytes; + int ndevs; + int i; + int j; - if (type & (BTRFS_BLOCK_GROUP_RAID0)) { - *num_stripes = fs_devices->rw_devices; - *min_stripes = 2; - } - if (type & (BTRFS_BLOCK_GROUP_DUP)) { - *num_stripes = 2; - *min_stripes = 2; - } - if (type & (BTRFS_BLOCK_GROUP_RAID1)) { - if (fs_devices->rw_devices < 2) - return -ENOSPC; - *num_stripes = 2; - *min_stripes = 2; - } - if (type & (BTRFS_BLOCK_GROUP_RAID10)) { - *num_stripes = fs_devices->rw_devices; - if (*num_stripes < 4) - return -ENOSPC; - *num_stripes &= ~(u32)1; - *sub_stripes = 2; - *min_stripes = 4; + if ((type & BTRFS_BLOCK_GROUP_RAID1) && + (type & BTRFS_BLOCK_GROUP_DUP)) { + WARN_ON(1); + type &= ~BTRFS_BLOCK_GROUP_DUP; } - return 0; -} + if (list_empty(&fs_devices->alloc_list)) + return -ENOSPC; -static u64 __btrfs_calc_stripe_size(struct btrfs_fs_devices *fs_devices, - u64 proposed_size, u64 type, - int num_stripes, int small_stripe) -{ - int min_stripe_size = 1 * 1024 * 1024; - u64 calc_size = proposed_size; - u64 max_chunk_size = calc_size; - int ncopies = 1; + sub_stripes = 1; + dev_stripes = 1; + devs_increment = 1; + ncopies = 1; + devs_max = 0; /* 0 == as many as possible */ + devs_min = 1; - if (type & (BTRFS_BLOCK_GROUP_RAID1 | - BTRFS_BLOCK_GROUP_DUP | - BTRFS_BLOCK_GROUP_RAID10)) + /* + * define the properties of each RAID type. + * FIXME: move this to a global table and use it in all RAID + * calculation code + */ + if (type & (BTRFS_BLOCK_GROUP_DUP)) { + dev_stripes = 2; ncopies = 2; + devs_max = 1; + } else if (type & (BTRFS_BLOCK_GROUP_RAID0)) { + devs_min = 2; + } else if (type & (BTRFS_BLOCK_GROUP_RAID1)) { + devs_increment = 2; + ncopies = 2; + devs_max = 2; + devs_min = 2; + } else if (type & (BTRFS_BLOCK_GROUP_RAID10)) { + sub_stripes = 2; + devs_increment = 2; + ncopies = 2; + devs_min = 4; + } else { + devs_max = 1; + } if (type & BTRFS_BLOCK_GROUP_DATA) { - max_chunk_size = 10 * calc_size; - min_stripe_size = 64 * 1024 * 1024; + max_stripe_size = 1024 * 1024 * 1024; + max_chunk_size = 10 * max_stripe_size; } else if (type & BTRFS_BLOCK_GROUP_METADATA) { - max_chunk_size = 256 * 1024 * 1024; - min_stripe_size = 32 * 1024 * 1024; + max_stripe_size = 256 * 1024 * 1024; + max_chunk_size = max_stripe_size; } else if (type & BTRFS_BLOCK_GROUP_SYSTEM) { - calc_size = 8 * 1024 * 1024; - max_chunk_size = calc_size * 2; - min_stripe_size = 1 * 1024 * 1024; + max_stripe_size = 8 * 1024 * 1024; + max_chunk_size = 2 * max_stripe_size; + } else { + printk(KERN_ERR "btrfs: invalid chunk type 0x%llx requested\n", + type); + BUG_ON(1); } /* we don't want a chunk larger than 10% of writeable space */ max_chunk_size = min(div_factor(fs_devices->total_rw_bytes, 1), max_chunk_size); - if (calc_size * num_stripes > max_chunk_size * ncopies) { - calc_size = max_chunk_size * ncopies; - do_div(calc_size, num_stripes); - do_div(calc_size, BTRFS_STRIPE_LEN); - calc_size *= BTRFS_STRIPE_LEN; - } + devices_info = kzalloc(sizeof(*devices_info) * fs_devices->rw_devices, + GFP_NOFS); + if (!devices_info) + return -ENOMEM; - /* we don't want tiny stripes */ - if (!small_stripe) - calc_size = max_t(u64, min_stripe_size, calc_size); + cur = fs_devices->alloc_list.next; /* - * we're about to do_div by the BTRFS_STRIPE_LEN so lets make sure - * we end up with something bigger than a stripe + * in the first pass through the devices list, we gather information + * about the available holes on each device. */ - calc_size = max_t(u64, calc_size, BTRFS_STRIPE_LEN); - - do_div(calc_size, BTRFS_STRIPE_LEN); - calc_size *= BTRFS_STRIPE_LEN; - - return calc_size; -} - -static struct map_lookup *__shrink_map_lookup_stripes(struct map_lookup *map, - int num_stripes) -{ - struct map_lookup *new; - size_t len = map_lookup_size(num_stripes); - - BUG_ON(map->num_stripes < num_stripes); - - if (map->num_stripes == num_stripes) - return map; - - new = kmalloc(len, GFP_NOFS); - if (!new) { - /* just change map->num_stripes */ - map->num_stripes = num_stripes; - return map; - } - - memcpy(new, map, len); - new->num_stripes = num_stripes; - kfree(map); - return new; -} + ndevs = 0; + while (cur != &fs_devices->alloc_list) { + struct btrfs_device *device; + u64 max_avail; + u64 dev_offset; -/* - * helper to allocate device space from btrfs_device_info, in which we stored - * max free space information of every device. It is used when we can not - * allocate chunks by default size. - * - * By this helper, we can allocate a new chunk as larger as possible. - */ -static int __btrfs_alloc_tiny_space(struct btrfs_trans_handle *trans, - struct btrfs_fs_devices *fs_devices, - struct btrfs_device_info *devices, - int nr_device, u64 type, - struct map_lookup **map_lookup, - int min_stripes, u64 *stripe_size) -{ - int i, index, sort_again = 0; - int min_devices = min_stripes; - u64 max_avail, min_free; - struct map_lookup *map = *map_lookup; - int ret; + device = list_entry(cur, struct btrfs_device, dev_alloc_list); - if (nr_device < min_stripes) - return -ENOSPC; + cur = cur->next; - btrfs_descending_sort_devices(devices, nr_device); + if (!device->writeable) { + printk(KERN_ERR + "btrfs: read-only device in alloc_list\n"); + WARN_ON(1); + continue; + } - max_avail = devices[0].max_avail; - if (!max_avail) - return -ENOSPC; + if (!device->in_fs_metadata) + continue; - for (i = 0; i < nr_device; i++) { - /* - * if dev_offset = 0, it means the free space of this device - * is less than what we need, and we didn't search max avail - * extent on this device, so do it now. + if (device->total_bytes > device->bytes_used) + total_avail = device->total_bytes - device->bytes_used; + else + total_avail = 0; + /* avail is off by max(alloc_start, 1MB), but that is the same + * for all devices, so it doesn't hurt the sorting later on */ - if (!devices[i].dev_offset) { - ret = find_free_dev_extent(trans, devices[i].dev, - max_avail, - &devices[i].dev_offset, - &devices[i].max_avail); - if (ret != 0 && ret != -ENOSPC) - return ret; - sort_again = 1; - } - } - /* we update the max avail free extent of each devices, sort again */ - if (sort_again) - btrfs_descending_sort_devices(devices, nr_device); - - if (type & BTRFS_BLOCK_GROUP_DUP) - min_devices = 1; + ret = find_free_dev_extent(trans, device, + max_stripe_size * dev_stripes, + &dev_offset, &max_avail); + if (ret && ret != -ENOSPC) + goto error; - if (!devices[min_devices - 1].max_avail) - return -ENOSPC; + if (ret == 0) + max_avail = max_stripe_size * dev_stripes; - max_avail = devices[min_devices - 1].max_avail; - if (type & BTRFS_BLOCK_GROUP_DUP) - do_div(max_avail, 2); + if (max_avail < BTRFS_STRIPE_LEN * dev_stripes) + continue; - max_avail = __btrfs_calc_stripe_size(fs_devices, max_avail, type, - min_stripes, 1); - if (type & BTRFS_BLOCK_GROUP_DUP) - min_free = max_avail * 2; - else - min_free = max_avail; + devices_info[ndevs].dev_offset = dev_offset; + devices_info[ndevs].max_avail = max_avail; + devices_info[ndevs].total_avail = total_avail; + devices_info[ndevs].dev = device; + ++ndevs; + } - if (min_free > devices[min_devices - 1].max_avail) - return -ENOSPC; + /* + * now sort the devices by hole size / available space + */ + sort(devices_info, ndevs, sizeof(struct btrfs_device_info), + btrfs_cmp_device_info, NULL); - map = __shrink_map_lookup_stripes(map, min_stripes); - *stripe_size = max_avail; + /* round down to number of usable stripes */ + ndevs -= ndevs % devs_increment; - index = 0; - for (i = 0; i < min_stripes; i++) { - map->stripes[i].dev = devices[index].dev; - map->stripes[i].physical = devices[index].dev_offset; - if (type & BTRFS_BLOCK_GROUP_DUP) { - i++; - map->stripes[i].dev = devices[index].dev; - map->stripes[i].physical = devices[index].dev_offset + - max_avail; - } - index++; + if (ndevs < devs_increment * sub_stripes || ndevs < devs_min) { + ret = -ENOSPC; + goto error; } - *map_lookup = map; - - return 0; -} -static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, - struct btrfs_root *extent_root, - struct map_lookup **map_ret, - u64 *num_bytes, u64 *stripe_size, - u64 start, u64 type) -{ - struct btrfs_fs_info *info = extent_root->fs_info; - struct btrfs_device *device = NULL; - struct btrfs_fs_devices *fs_devices = info->fs_devices; - struct list_head *cur; - struct map_lookup *map; - struct extent_map_tree *em_tree; - struct extent_map *em; - struct btrfs_device_info *devices_info; - struct list_head private_devs; - u64 calc_size = 1024 * 1024 * 1024; - u64 min_free; - u64 avail; - u64 dev_offset; - int num_stripes; - int min_stripes; - int sub_stripes; - int min_devices; /* the min number of devices we need */ - int i; - int ret; - int index; + if (devs_max && ndevs > devs_max) + ndevs = devs_max; + /* + * the primary goal is to maximize the number of stripes, so use as many + * devices as possible, even if the stripes are not maximum sized. + */ + stripe_size = devices_info[ndevs-1].max_avail; + num_stripes = ndevs * dev_stripes; - if ((type & BTRFS_BLOCK_GROUP_RAID1) && - (type & BTRFS_BLOCK_GROUP_DUP)) { - WARN_ON(1); - type &= ~BTRFS_BLOCK_GROUP_DUP; + if (stripe_size * num_stripes > max_chunk_size * ncopies) { + stripe_size = max_chunk_size * ncopies; + do_div(stripe_size, num_stripes); } - if (list_empty(&fs_devices->alloc_list)) - return -ENOSPC; - - ret = __btrfs_calc_nstripes(fs_devices, type, &num_stripes, - &min_stripes, &sub_stripes); - if (ret) - return ret; - devices_info = kzalloc(sizeof(*devices_info) * fs_devices->rw_devices, - GFP_NOFS); - if (!devices_info) - return -ENOMEM; + do_div(stripe_size, dev_stripes); + do_div(stripe_size, BTRFS_STRIPE_LEN); + stripe_size *= BTRFS_STRIPE_LEN; map = kmalloc(map_lookup_size(num_stripes), GFP_NOFS); if (!map) { @@ -2504,85 +2475,12 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, } map->num_stripes = num_stripes; - cur = fs_devices->alloc_list.next; - index = 0; - i = 0; - - calc_size = __btrfs_calc_stripe_size(fs_devices, calc_size, type, - num_stripes, 0); - - if (type & BTRFS_BLOCK_GROUP_DUP) { - min_free = calc_size * 2; - min_devices = 1; - } else { - min_free = calc_size; - min_devices = min_stripes; - } - - INIT_LIST_HEAD(&private_devs); - while (index < num_stripes) { - device = list_entry(cur, struct btrfs_device, dev_alloc_list); - BUG_ON(!device->writeable); - if (device->total_bytes > device->bytes_used) - avail = device->total_bytes - device->bytes_used; - else - avail = 0; - cur = cur->next; - - if (device->in_fs_metadata && avail >= min_free) { - ret = find_free_dev_extent(trans, device, min_free, - &devices_info[i].dev_offset, - &devices_info[i].max_avail); - if (ret == 0) { - list_move_tail(&device->dev_alloc_list, - &private_devs); - map->stripes[index].dev = device; - map->stripes[index].physical = - devices_info[i].dev_offset; - index++; - if (type & BTRFS_BLOCK_GROUP_DUP) { - map->stripes[index].dev = device; - map->stripes[index].physical = - devices_info[i].dev_offset + - calc_size; - index++; - } - } else if (ret != -ENOSPC) - goto error; - - devices_info[i].dev = device; - i++; - } else if (device->in_fs_metadata && - avail >= BTRFS_STRIPE_LEN) { - devices_info[i].dev = device; - devices_info[i].max_avail = avail; - i++; - } - - if (cur == &fs_devices->alloc_list) - break; - } - - list_splice(&private_devs, &fs_devices->alloc_list); - if (index < num_stripes) { - if (index >= min_stripes) { - num_stripes = index; - if (type & (BTRFS_BLOCK_GROUP_RAID10)) { - num_stripes /= sub_stripes; - num_stripes *= sub_stripes; - } - - map = __shrink_map_lookup_stripes(map, num_stripes); - } else if (i >= min_devices) { - ret = __btrfs_alloc_tiny_space(trans, fs_devices, - devices_info, i, type, - &map, min_stripes, - &calc_size); - if (ret) - goto error; - } else { - ret = -ENOSPC; - goto error; + for (i = 0; i < ndevs; ++i) { + for (j = 0; j < dev_stripes; ++j) { + int s = i * dev_stripes + j; + map->stripes[s].dev = devices_info[i].dev; + map->stripes[s].physical = devices_info[i].dev_offset + + j * stripe_size; } } map->sector_size = extent_root->sectorsize; @@ -2593,20 +2491,21 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, map->sub_stripes = sub_stripes; *map_ret = map; - *stripe_size = calc_size; - *num_bytes = chunk_bytes_by_type(type, calc_size, - map->num_stripes, sub_stripes); + num_bytes = stripe_size * (num_stripes / ncopies); - trace_btrfs_chunk_alloc(info->chunk_root, map, start, *num_bytes); + *stripe_size_out = stripe_size; + *num_bytes_out = num_bytes; - em = alloc_extent_map(GFP_NOFS); + trace_btrfs_chunk_alloc(info->chunk_root, map, start, num_bytes); + + em = alloc_extent_map(); if (!em) { ret = -ENOMEM; goto error; } em->bdev = (struct block_device *)map; em->start = start; - em->len = *num_bytes; + em->len = num_bytes; em->block_start = 0; em->block_len = em->len; @@ -2619,20 +2518,21 @@ static int __btrfs_alloc_chunk(struct btrfs_trans_handle *trans, ret = btrfs_make_block_group(trans, extent_root, 0, type, BTRFS_FIRST_CHUNK_TREE_OBJECTID, - start, *num_bytes); + start, num_bytes); BUG_ON(ret); - index = 0; - while (index < map->num_stripes) { - device = map->stripes[index].dev; - dev_offset = map->stripes[index].physical; + for (i = 0; i < map->num_stripes; ++i) { + struct btrfs_device *device; + u64 dev_offset; + + device = map->stripes[i].dev; + dev_offset = map->stripes[i].physical; ret = btrfs_alloc_dev_extent(trans, device, info->chunk_root->root_key.objectid, BTRFS_FIRST_CHUNK_TREE_OBJECTID, - start, dev_offset, calc_size); + start, dev_offset, stripe_size); BUG_ON(ret); - index++; } kfree(devices_info); @@ -2839,7 +2739,7 @@ int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset) void btrfs_mapping_init(struct btrfs_mapping_tree *tree) { - extent_map_tree_init(&tree->map_tree, GFP_NOFS); + extent_map_tree_init(&tree->map_tree); } void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree) @@ -3489,7 +3389,7 @@ static int read_one_chunk(struct btrfs_root *root, struct btrfs_key *key, free_extent_map(em); } - em = alloc_extent_map(GFP_NOFS); + em = alloc_extent_map(); if (!em) return -ENOMEM; num_stripes = btrfs_chunk_num_stripes(leaf, chunk); @@ -3678,15 +3578,6 @@ static int read_one_dev(struct btrfs_root *root, return ret; } -int btrfs_read_super_device(struct btrfs_root *root, struct extent_buffer *buf) -{ - struct btrfs_dev_item *dev_item; - - dev_item = (struct btrfs_dev_item *)offsetof(struct btrfs_super_block, - dev_item); - return read_one_dev(root, buf, dev_item); -} - int btrfs_read_sys_array(struct btrfs_root *root) { struct btrfs_super_block *super_copy = &root->fs_info->super_copy; @@ -3803,7 +3694,7 @@ again: } if (key.objectid == BTRFS_DEV_ITEMS_OBJECTID) { key.objectid = 0; - btrfs_release_path(root, path); + btrfs_release_path(path); goto again; } ret = 0; diff --git a/fs/btrfs/volumes.h b/fs/btrfs/volumes.h index cc2eadaf7a27..7c12d61ae7ae 100644 --- a/fs/btrfs/volumes.h +++ b/fs/btrfs/volumes.h @@ -85,7 +85,12 @@ struct btrfs_device { /* physical drive uuid (or lvm uuid) */ u8 uuid[BTRFS_UUID_SIZE]; + /* per-device scrub information */ + struct scrub_dev *scrub_device; + struct btrfs_work work; + struct rcu_head rcu; + struct work_struct rcu_work; }; struct btrfs_fs_devices { @@ -144,6 +149,7 @@ struct btrfs_device_info { struct btrfs_device *dev; u64 dev_offset; u64 max_avail; + u64 total_avail; }; struct map_lookup { @@ -157,20 +163,8 @@ struct map_lookup { struct btrfs_bio_stripe stripes[]; }; -/* Used to sort the devices by max_avail(descending sort) */ -int btrfs_cmp_device_free_bytes(const void *dev_info1, const void *dev_info2); - -/* - * sort the devices by max_avail, in which max free extent size of each device - * is stored.(Descending Sort) - */ -static inline void btrfs_descending_sort_devices( - struct btrfs_device_info *devices, - size_t nr_devices) -{ - sort(devices, nr_devices, sizeof(struct btrfs_device_info), - btrfs_cmp_device_free_bytes, NULL); -} +#define map_lookup_size(n) (sizeof(struct map_lookup) + \ + (sizeof(struct btrfs_bio_stripe) * (n))) int btrfs_account_dev_extents_size(struct btrfs_device *device, u64 start, u64 end, u64 *length); @@ -196,7 +190,6 @@ void btrfs_mapping_init(struct btrfs_mapping_tree *tree); void btrfs_mapping_tree_free(struct btrfs_mapping_tree *tree); int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, int mirror_num, int async_submit); -int btrfs_read_super_device(struct btrfs_root *root, struct extent_buffer *buf); int btrfs_open_devices(struct btrfs_fs_devices *fs_devices, fmode_t flags, void *holder); int btrfs_scan_one_device(const char *path, fmode_t flags, void *holder, @@ -209,8 +202,6 @@ int btrfs_add_device(struct btrfs_trans_handle *trans, int btrfs_rm_device(struct btrfs_root *root, char *device_path); int btrfs_cleanup_fs_uuids(void); int btrfs_num_copies(struct btrfs_mapping_tree *map_tree, u64 logical, u64 len); -int btrfs_unplug_page(struct btrfs_mapping_tree *map_tree, - u64 logical, struct page *page); int btrfs_grow_device(struct btrfs_trans_handle *trans, struct btrfs_device *device, u64 new_size); struct btrfs_device *btrfs_find_device(struct btrfs_root *root, u64 devid, @@ -218,8 +209,6 @@ struct btrfs_device *btrfs_find_device(struct btrfs_root *root, u64 devid, int btrfs_shrink_device(struct btrfs_device *device, u64 new_size); int btrfs_init_new_device(struct btrfs_root *root, char *path); int btrfs_balance(struct btrfs_root *dev_root); -void btrfs_unlock_volumes(void); -void btrfs_lock_volumes(void); int btrfs_chunk_readonly(struct btrfs_root *root, u64 chunk_offset); int find_free_dev_extent(struct btrfs_trans_handle *trans, struct btrfs_device *device, u64 num_bytes, diff --git a/fs/btrfs/xattr.c b/fs/btrfs/xattr.c index a5303b871b13..f3107e4b4d56 100644 --- a/fs/btrfs/xattr.c +++ b/fs/btrfs/xattr.c @@ -44,7 +44,7 @@ ssize_t __btrfs_getxattr(struct inode *inode, const char *name, return -ENOMEM; /* lookup the xattr by name */ - di = btrfs_lookup_xattr(NULL, root, path, inode->i_ino, name, + di = btrfs_lookup_xattr(NULL, root, path, btrfs_ino(inode), name, strlen(name), 0); if (!di) { ret = -ENODATA; @@ -103,7 +103,7 @@ static int do_setxattr(struct btrfs_trans_handle *trans, return -ENOMEM; /* first lets see if we already have this xattr */ - di = btrfs_lookup_xattr(trans, root, path, inode->i_ino, name, + di = btrfs_lookup_xattr(trans, root, path, btrfs_ino(inode), name, strlen(name), -1); if (IS_ERR(di)) { ret = PTR_ERR(di); @@ -120,13 +120,13 @@ static int do_setxattr(struct btrfs_trans_handle *trans, ret = btrfs_delete_one_dir_name(trans, root, path, di); BUG_ON(ret); - btrfs_release_path(root, path); + btrfs_release_path(path); /* if we don't have a value then we are removing the xattr */ if (!value) goto out; } else { - btrfs_release_path(root, path); + btrfs_release_path(path); if (flags & XATTR_REPLACE) { /* we couldn't find the attr to replace */ @@ -136,7 +136,7 @@ static int do_setxattr(struct btrfs_trans_handle *trans, } /* ok we have to create a completely new xattr */ - ret = btrfs_insert_xattr_item(trans, root, path, inode->i_ino, + ret = btrfs_insert_xattr_item(trans, root, path, btrfs_ino(inode), name, name_len, value, size); BUG_ON(ret); out: @@ -180,18 +180,17 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) struct btrfs_path *path; struct extent_buffer *leaf; struct btrfs_dir_item *di; - int ret = 0, slot, advance; + int ret = 0, slot; size_t total_size = 0, size_left = size; unsigned long name_ptr; size_t name_len; - u32 nritems; /* * ok we want all objects associated with this id. * NOTE: we set key.offset = 0; because we want to start with the * first xattr that we find and walk forward */ - key.objectid = inode->i_ino; + key.objectid = btrfs_ino(inode); btrfs_set_key_type(&key, BTRFS_XATTR_ITEM_KEY); key.offset = 0; @@ -204,34 +203,24 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) ret = btrfs_search_slot(NULL, root, &key, path, 0, 0); if (ret < 0) goto err; - advance = 0; + while (1) { leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); slot = path->slots[0]; /* this is where we start walking through the path */ - if (advance || slot >= nritems) { + if (slot >= btrfs_header_nritems(leaf)) { /* * if we've reached the last slot in this leaf we need * to go to the next leaf and reset everything */ - if (slot >= nritems-1) { - ret = btrfs_next_leaf(root, path); - if (ret) - break; - leaf = path->nodes[0]; - nritems = btrfs_header_nritems(leaf); - slot = path->slots[0]; - } else { - /* - * just walking through the slots on this leaf - */ - slot++; - path->slots[0]++; - } + ret = btrfs_next_leaf(root, path); + if (ret < 0) + goto err; + else if (ret > 0) + break; + continue; } - advance = 1; btrfs_item_key_to_cpu(leaf, &found_key, slot); @@ -250,7 +239,7 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) /* we are just looking for how big our buffer needs to be */ if (!size) - continue; + goto next; if (!buffer || (name_len + 1) > size_left) { ret = -ERANGE; @@ -263,6 +252,8 @@ ssize_t btrfs_listxattr(struct dentry *dentry, char *buffer, size_t size) size_left -= name_len + 1; buffer += name_len + 1; +next: + path->slots[0]++; } ret = total_size; diff --git a/fs/buffer.c b/fs/buffer.c index a08bb8e61c6f..49c9aada0374 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -41,6 +41,7 @@ #include <linux/bitops.h> #include <linux/mpage.h> #include <linux/bit_spinlock.h> +#include <linux/cleancache.h> static int fsync_buffers_list(spinlock_t *lock, struct list_head *list); @@ -269,6 +270,10 @@ void invalidate_bdev(struct block_device *bdev) invalidate_bh_lrus(); lru_add_drain_all(); /* make sure all lru add caches are flushed */ invalidate_mapping_pages(mapping, 0, -1); + /* 99% of the time, we don't need to flush the cleancache on the bdev. + * But, for the strange corners, lets be cautious + */ + cleancache_flush_inode(mapping); } EXPORT_SYMBOL(invalidate_bdev); @@ -2331,24 +2336,26 @@ EXPORT_SYMBOL(block_commit_write); * page lock we can determine safely if the page is beyond EOF. If it is not * beyond EOF, then the page is guaranteed safe against truncation until we * unlock the page. + * + * Direct callers of this function should call vfs_check_frozen() so that page + * fault does not busyloop until the fs is thawed. */ -int -block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, - get_block_t get_block) +int __block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, + get_block_t get_block) { struct page *page = vmf->page; struct inode *inode = vma->vm_file->f_path.dentry->d_inode; unsigned long end; loff_t size; - int ret = VM_FAULT_NOPAGE; /* make the VM retry the fault */ + int ret; lock_page(page); size = i_size_read(inode); if ((page->mapping != inode->i_mapping) || (page_offset(page) > size)) { - /* page got truncated out from underneath us */ - unlock_page(page); - goto out; + /* We overload EFAULT to mean page got truncated */ + ret = -EFAULT; + goto out_unlock; } /* page is wholly or partially inside EOF */ @@ -2361,18 +2368,42 @@ block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, if (!ret) ret = block_commit_write(page, 0, end); - if (unlikely(ret)) { - unlock_page(page); - if (ret == -ENOMEM) - ret = VM_FAULT_OOM; - else /* -ENOSPC, -EIO, etc */ - ret = VM_FAULT_SIGBUS; - } else - ret = VM_FAULT_LOCKED; - -out: + if (unlikely(ret < 0)) + goto out_unlock; + /* + * Freezing in progress? We check after the page is marked dirty and + * with page lock held so if the test here fails, we are sure freezing + * code will wait during syncing until the page fault is done - at that + * point page will be dirty and unlocked so freezing code will write it + * and writeprotect it again. + */ + set_page_dirty(page); + if (inode->i_sb->s_frozen != SB_UNFROZEN) { + ret = -EAGAIN; + goto out_unlock; + } + wait_on_page_writeback(page); + return 0; +out_unlock: + unlock_page(page); return ret; } +EXPORT_SYMBOL(__block_page_mkwrite); + +int block_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf, + get_block_t get_block) +{ + int ret; + struct super_block *sb = vma->vm_file->f_path.dentry->d_inode->i_sb; + + /* + * This check is racy but catches the common case. The check in + * __block_page_mkwrite() is reliable. + */ + vfs_check_frozen(sb, SB_FREEZE_WRITE); + ret = __block_page_mkwrite(vma, vmf, get_block); + return block_page_mkwrite_return(ret); +} EXPORT_SYMBOL(block_page_mkwrite); /* diff --git a/fs/cachefiles/interface.c b/fs/cachefiles/interface.c index 37fe101a4e0d..1064805e653b 100644 --- a/fs/cachefiles/interface.c +++ b/fs/cachefiles/interface.c @@ -197,7 +197,7 @@ struct fscache_object *cachefiles_grab_object(struct fscache_object *_object) } /* - * update the auxilliary data for an object object on disk + * update the auxiliary data for an object object on disk */ static void cachefiles_update_object(struct fscache_object *_object) { diff --git a/fs/ceph/addr.c b/fs/ceph/addr.c index 37368ba2e67c..33da49dc3cc6 100644 --- a/fs/ceph/addr.c +++ b/fs/ceph/addr.c @@ -24,7 +24,7 @@ * context needs to be associated with the osd write during writeback. * * Similarly, struct ceph_inode_info maintains a set of counters to - * count dirty pages on the inode. In the absense of snapshots, + * count dirty pages on the inode. In the absence of snapshots, * i_wrbuffer_ref == i_wrbuffer_ref_head == the dirty page count. * * When a snapshot is taken (that is, when the client receives @@ -775,6 +775,13 @@ get_more_pages: ci->i_truncate_seq, ci->i_truncate_size, &inode->i_mtime, true, 1, 0); + + if (!req) { + rc = -ENOMEM; + unlock_page(page); + break; + } + max_pages = req->r_num_pages; alloc_page_vec(fsc, req); @@ -841,7 +848,8 @@ get_more_pages: op->payload_len = cpu_to_le32(len); req->r_request->hdr.data_len = cpu_to_le32(len); - ceph_osdc_start_request(&fsc->client->osdc, req, true); + rc = ceph_osdc_start_request(&fsc->client->osdc, req, true); + BUG_ON(rc); req = NULL; /* continue? */ @@ -873,8 +881,6 @@ release_pvec_pages: out: if (req) ceph_osdc_put_request(req); - if (rc > 0) - rc = 0; /* vfs expects us to return 0 */ ceph_put_snap_context(snapc); dout("writepages done, rc = %d\n", rc); return rc; diff --git a/fs/ceph/caps.c b/fs/ceph/caps.c index 6b61ded701e1..1f72b00447c4 100644 --- a/fs/ceph/caps.c +++ b/fs/ceph/caps.c @@ -569,7 +569,8 @@ retry: list_add_tail(&cap->session_caps, &session->s_caps); session->s_nr_caps++; spin_unlock(&session->s_cap_lock); - } + } else if (new_cap) + ceph_put_cap(mdsc, new_cap); if (!ci->i_snap_realm) { /* @@ -765,7 +766,7 @@ int __ceph_caps_issued_mask(struct ceph_inode_info *ci, int mask, int touch) if (touch) { struct rb_node *q; - /* touch this + preceeding caps */ + /* touch this + preceding caps */ __touch_cap(cap); for (q = rb_first(&ci->i_caps); q != p; q = rb_next(q)) { @@ -819,7 +820,7 @@ int __ceph_caps_used(struct ceph_inode_info *ci) used |= CEPH_CAP_FILE_CACHE; if (ci->i_wr_ref) used |= CEPH_CAP_FILE_WR; - if (ci->i_wrbuffer_ref) + if (ci->i_wb_ref || ci->i_wrbuffer_ref) used |= CEPH_CAP_FILE_BUFFER; return used; } @@ -1331,10 +1332,11 @@ static void ceph_flush_snaps(struct ceph_inode_info *ci) } /* - * Mark caps dirty. If inode is newly dirty, add to the global dirty - * list. + * Mark caps dirty. If inode is newly dirty, return the dirty flags. + * Caller is then responsible for calling __mark_inode_dirty with the + * returned flags value. */ -void __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) +int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) { struct ceph_mds_client *mdsc = ceph_sb_to_client(ci->vfs_inode.i_sb)->mdsc; @@ -1357,7 +1359,7 @@ void __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) list_add(&ci->i_dirty_item, &mdsc->cap_dirty); spin_unlock(&mdsc->cap_dirty_lock); if (ci->i_flushing_caps == 0) { - igrab(inode); + ihold(inode); dirty |= I_DIRTY_SYNC; } } @@ -1365,9 +1367,8 @@ void __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask) if (((was | ci->i_flushing_caps) & CEPH_CAP_FILE_BUFFER) && (mask & CEPH_CAP_FILE_BUFFER)) dirty |= I_DIRTY_DATASYNC; - if (dirty) - __mark_inode_dirty(inode, dirty); __cap_delay_requeue(mdsc, ci); + return dirty; } /* @@ -1990,11 +1991,11 @@ static void __take_cap_refs(struct ceph_inode_info *ci, int got) if (got & CEPH_CAP_FILE_WR) ci->i_wr_ref++; if (got & CEPH_CAP_FILE_BUFFER) { - if (ci->i_wrbuffer_ref == 0) - igrab(&ci->vfs_inode); - ci->i_wrbuffer_ref++; - dout("__take_cap_refs %p wrbuffer %d -> %d (?)\n", - &ci->vfs_inode, ci->i_wrbuffer_ref-1, ci->i_wrbuffer_ref); + if (ci->i_wb_ref == 0) + ihold(&ci->vfs_inode); + ci->i_wb_ref++; + dout("__take_cap_refs %p wb %d -> %d (?)\n", + &ci->vfs_inode, ci->i_wb_ref-1, ci->i_wb_ref); } } @@ -2169,12 +2170,12 @@ void ceph_put_cap_refs(struct ceph_inode_info *ci, int had) if (--ci->i_rdcache_ref == 0) last++; if (had & CEPH_CAP_FILE_BUFFER) { - if (--ci->i_wrbuffer_ref == 0) { + if (--ci->i_wb_ref == 0) { last++; put++; } - dout("put_cap_refs %p wrbuffer %d -> %d (?)\n", - inode, ci->i_wrbuffer_ref+1, ci->i_wrbuffer_ref); + dout("put_cap_refs %p wb %d -> %d (?)\n", + inode, ci->i_wb_ref+1, ci->i_wb_ref); } if (had & CEPH_CAP_FILE_WR) if (--ci->i_wr_ref == 0) { @@ -2634,6 +2635,7 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex, struct ceph_mds_session *session, int *open_target_sessions) { + struct ceph_mds_client *mdsc = ceph_inode_to_client(inode)->mdsc; struct ceph_inode_info *ci = ceph_inode(inode); int mds = session->s_mds; unsigned mseq = le32_to_cpu(ex->migrate_seq); @@ -2670,6 +2672,19 @@ static void handle_cap_export(struct inode *inode, struct ceph_mds_caps *ex, * export targets, so that we get the matching IMPORT */ *open_target_sessions = 1; + + /* + * we can't flush dirty caps that we've seen the + * EXPORT but no IMPORT for + */ + spin_lock(&mdsc->cap_dirty_lock); + if (!list_empty(&ci->i_dirty_item)) { + dout(" moving %p to cap_dirty_migrating\n", + inode); + list_move(&ci->i_dirty_item, + &mdsc->cap_dirty_migrating); + } + spin_unlock(&mdsc->cap_dirty_lock); } __ceph_remove_cap(cap); } @@ -2707,6 +2722,13 @@ static void handle_cap_import(struct ceph_mds_client *mdsc, ci->i_cap_exporting_issued = 0; ci->i_cap_exporting_mseq = 0; ci->i_cap_exporting_mds = -1; + + spin_lock(&mdsc->cap_dirty_lock); + if (!list_empty(&ci->i_dirty_item)) { + dout(" moving %p back to cap_dirty\n", inode); + list_move(&ci->i_dirty_item, &mdsc->cap_dirty); + } + spin_unlock(&mdsc->cap_dirty_lock); } else { dout("handle_cap_import inode %p ci %p mds%d mseq %d\n", inode, ci, mds, mseq); @@ -2910,38 +2932,16 @@ void ceph_check_delayed_caps(struct ceph_mds_client *mdsc) */ void ceph_flush_dirty_caps(struct ceph_mds_client *mdsc) { - struct ceph_inode_info *ci, *nci = NULL; - struct inode *inode, *ninode = NULL; - struct list_head *p, *n; + struct ceph_inode_info *ci; + struct inode *inode; dout("flush_dirty_caps\n"); spin_lock(&mdsc->cap_dirty_lock); - list_for_each_safe(p, n, &mdsc->cap_dirty) { - if (nci) { - ci = nci; - inode = ninode; - ci->i_ceph_flags &= ~CEPH_I_NOFLUSH; - dout("flush_dirty_caps inode %p (was next inode)\n", - inode); - } else { - ci = list_entry(p, struct ceph_inode_info, - i_dirty_item); - inode = igrab(&ci->vfs_inode); - BUG_ON(!inode); - dout("flush_dirty_caps inode %p\n", inode); - } - if (n != &mdsc->cap_dirty) { - nci = list_entry(n, struct ceph_inode_info, - i_dirty_item); - ninode = igrab(&nci->vfs_inode); - BUG_ON(!ninode); - nci->i_ceph_flags |= CEPH_I_NOFLUSH; - dout("flush_dirty_caps next inode %p, noflush\n", - ninode); - } else { - nci = NULL; - ninode = NULL; - } + while (!list_empty(&mdsc->cap_dirty)) { + ci = list_first_entry(&mdsc->cap_dirty, struct ceph_inode_info, + i_dirty_item); + inode = igrab(&ci->vfs_inode); + dout("flush_dirty_caps %p\n", inode); spin_unlock(&mdsc->cap_dirty_lock); if (inode) { ceph_check_caps(ci, CHECK_CAPS_NODELAY|CHECK_CAPS_FLUSH, @@ -2951,6 +2951,7 @@ void ceph_flush_dirty_caps(struct ceph_mds_client *mdsc) spin_lock(&mdsc->cap_dirty_lock); } spin_unlock(&mdsc->cap_dirty_lock); + dout("flush_dirty_caps done\n"); } /* diff --git a/fs/ceph/dir.c b/fs/ceph/dir.c index 1a867a3601ae..33729e822bb9 100644 --- a/fs/ceph/dir.c +++ b/fs/ceph/dir.c @@ -360,7 +360,7 @@ more: rinfo = &fi->last_readdir->r_reply_info; dout("readdir frag %x num %d off %d chunkoff %d\n", frag, rinfo->dir_nr, off, fi->offset); - while (off - fi->offset >= 0 && off - fi->offset < rinfo->dir_nr) { + while (off >= fi->offset && off - fi->offset < rinfo->dir_nr) { u64 pos = ceph_make_fpos(frag, off); struct ceph_mds_reply_inode *in = rinfo->dir_in[off - fi->offset].in; @@ -1066,16 +1066,17 @@ static ssize_t ceph_read_dir(struct file *file, char __user *buf, size_t size, struct inode *inode = file->f_dentry->d_inode; struct ceph_inode_info *ci = ceph_inode(inode); int left; + const int bufsize = 1024; if (!ceph_test_mount_opt(ceph_sb_to_client(inode->i_sb), DIRSTAT)) return -EISDIR; if (!cf->dir_info) { - cf->dir_info = kmalloc(1024, GFP_NOFS); + cf->dir_info = kmalloc(bufsize, GFP_NOFS); if (!cf->dir_info) return -ENOMEM; cf->dir_info_len = - sprintf(cf->dir_info, + snprintf(cf->dir_info, bufsize, "entries: %20lld\n" " files: %20lld\n" " subdirs: %20lld\n" diff --git a/fs/ceph/export.c b/fs/ceph/export.c index e41056174bf8..a610d3d67488 100644 --- a/fs/ceph/export.c +++ b/fs/ceph/export.c @@ -86,6 +86,7 @@ static int ceph_encode_fh(struct dentry *dentry, u32 *rawfh, int *max_len, static struct dentry *__fh_to_dentry(struct super_block *sb, struct ceph_nfs_fh *fh) { + struct ceph_mds_client *mdsc = ceph_sb_to_client(sb)->mdsc; struct inode *inode; struct dentry *dentry; struct ceph_vino vino; @@ -95,8 +96,24 @@ static struct dentry *__fh_to_dentry(struct super_block *sb, vino.ino = fh->ino; vino.snap = CEPH_NOSNAP; inode = ceph_find_inode(sb, vino); - if (!inode) - return ERR_PTR(-ESTALE); + if (!inode) { + struct ceph_mds_request *req; + + req = ceph_mdsc_create_request(mdsc, CEPH_MDS_OP_LOOKUPINO, + USE_ANY_MDS); + if (IS_ERR(req)) + return ERR_CAST(req); + + req->r_ino1 = vino; + req->r_num_caps = 1; + err = ceph_mdsc_do_request(mdsc, NULL, req); + inode = req->r_target_inode; + if (inode) + igrab(inode); + ceph_mdsc_put_request(req); + if (!inode) + return ERR_PTR(-ESTALE); + } dentry = d_obtain_alias(inode); if (IS_ERR(dentry)) { @@ -148,8 +165,10 @@ static struct dentry *__cfh_to_dentry(struct super_block *sb, snprintf(req->r_path2, 16, "%d", cfh->parent_name_hash); req->r_num_caps = 1; err = ceph_mdsc_do_request(mdsc, NULL, req); + inode = req->r_target_inode; + if (inode) + igrab(inode); ceph_mdsc_put_request(req); - inode = ceph_find_inode(sb, vino); if (!inode) return ERR_PTR(err ? err : -ESTALE); } diff --git a/fs/ceph/file.c b/fs/ceph/file.c index 159b512d5a27..203252d88d9f 100644 --- a/fs/ceph/file.c +++ b/fs/ceph/file.c @@ -734,9 +734,12 @@ retry_snap: } } if (ret >= 0) { + int dirty; spin_lock(&inode->i_lock); - __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_FILE_WR); spin_unlock(&inode->i_lock); + if (dirty) + __mark_inode_dirty(inode, dirty); } out: diff --git a/fs/ceph/inode.c b/fs/ceph/inode.c index b54c97da1c43..70b6a4839c38 100644 --- a/fs/ceph/inode.c +++ b/fs/ceph/inode.c @@ -355,6 +355,7 @@ struct inode *ceph_alloc_inode(struct super_block *sb) ci->i_rd_ref = 0; ci->i_rdcache_ref = 0; ci->i_wr_ref = 0; + ci->i_wb_ref = 0; ci->i_wrbuffer_ref = 0; ci->i_wrbuffer_ref_head = 0; ci->i_shared_gen = 0; @@ -1567,6 +1568,7 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) int release = 0, dirtied = 0; int mask = 0; int err = 0; + int inode_dirty_flags = 0; if (ceph_snap(inode) != CEPH_NOSNAP) return -EROFS; @@ -1725,13 +1727,16 @@ int ceph_setattr(struct dentry *dentry, struct iattr *attr) dout("setattr %p ATTR_FILE ... hrm!\n", inode); if (dirtied) { - __ceph_mark_dirty_caps(ci, dirtied); + inode_dirty_flags = __ceph_mark_dirty_caps(ci, dirtied); inode->i_ctime = CURRENT_TIME; } release &= issued; spin_unlock(&inode->i_lock); + if (inode_dirty_flags) + __mark_inode_dirty(inode, inode_dirty_flags); + if (mask) { req->r_inode = igrab(inode); req->r_inode_drop = release; diff --git a/fs/ceph/mds_client.c b/fs/ceph/mds_client.c index f60b07b0feb0..79743d146be6 100644 --- a/fs/ceph/mds_client.c +++ b/fs/ceph/mds_client.c @@ -578,6 +578,7 @@ static void __register_request(struct ceph_mds_client *mdsc, if (dir) { struct ceph_inode_info *ci = ceph_inode(dir); + ihold(dir); spin_lock(&ci->i_unsafe_lock); req->r_unsafe_dir = dir; list_add_tail(&req->r_unsafe_dir_item, &ci->i_unsafe_dirops); @@ -598,6 +599,9 @@ static void __unregister_request(struct ceph_mds_client *mdsc, spin_lock(&ci->i_unsafe_lock); list_del_init(&req->r_unsafe_dir_item); spin_unlock(&ci->i_unsafe_lock); + + iput(req->r_unsafe_dir); + req->r_unsafe_dir = NULL; } ceph_mdsc_put_request(req); @@ -2691,7 +2695,6 @@ static void handle_lease(struct ceph_mds_client *mdsc, { struct super_block *sb = mdsc->fsc->sb; struct inode *inode; - struct ceph_inode_info *ci; struct dentry *parent, *dentry; struct ceph_dentry_info *di; int mds = session->s_mds; @@ -2728,7 +2731,6 @@ static void handle_lease(struct ceph_mds_client *mdsc, dout("handle_lease no inode %llx\n", vino.ino); goto release; } - ci = ceph_inode(inode); /* dentry */ parent = d_find_alias(inode); @@ -3002,6 +3004,7 @@ int ceph_mdsc_init(struct ceph_fs_client *fsc) spin_lock_init(&mdsc->snap_flush_lock); mdsc->cap_flush_seq = 0; INIT_LIST_HEAD(&mdsc->cap_dirty); + INIT_LIST_HEAD(&mdsc->cap_dirty_migrating); mdsc->num_cap_flushing = 0; spin_lock_init(&mdsc->cap_dirty_lock); init_waitqueue_head(&mdsc->cap_flushing_wq); @@ -3304,8 +3307,8 @@ static void con_put(struct ceph_connection *con) { struct ceph_mds_session *s = con->private; + dout("mdsc con_put %p (%d)\n", s, atomic_read(&s->s_ref) - 1); ceph_put_mds_session(s); - dout("mdsc con_put %p (%d)\n", s, atomic_read(&s->s_ref)); } /* diff --git a/fs/ceph/mds_client.h b/fs/ceph/mds_client.h index 4e3a9cc0bba6..7d8a0d662d56 100644 --- a/fs/ceph/mds_client.h +++ b/fs/ceph/mds_client.h @@ -278,6 +278,7 @@ struct ceph_mds_client { u64 cap_flush_seq; struct list_head cap_dirty; /* inodes with dirty caps */ + struct list_head cap_dirty_migrating; /* ...that are migration... */ int num_cap_flushing; /* # caps we are flushing */ spinlock_t cap_dirty_lock; /* protects above items */ wait_queue_head_t cap_flushing_wq; diff --git a/fs/ceph/snap.c b/fs/ceph/snap.c index 0aee66b92af3..24067d68a554 100644 --- a/fs/ceph/snap.c +++ b/fs/ceph/snap.c @@ -206,7 +206,7 @@ void ceph_put_snap_realm(struct ceph_mds_client *mdsc, up_write(&mdsc->snap_rwsem); } else { spin_lock(&mdsc->snap_empty_lock); - list_add(&mdsc->snap_empty, &realm->empty_item); + list_add(&realm->empty_item, &mdsc->snap_empty); spin_unlock(&mdsc->snap_empty_lock); } } @@ -342,7 +342,7 @@ static int build_snap_context(struct ceph_snap_realm *realm) num = 0; snapc->seq = realm->seq; if (parent) { - /* include any of parent's snaps occuring _after_ my + /* include any of parent's snaps occurring _after_ my parent became my parent */ for (i = 0; i < parent->cached_context->num_snaps; i++) if (parent->cached_context->snaps[i] >= diff --git a/fs/ceph/super.h b/fs/ceph/super.h index 619fe719968f..f5cabefa98dc 100644 --- a/fs/ceph/super.h +++ b/fs/ceph/super.h @@ -293,7 +293,7 @@ struct ceph_inode_info { /* held references to caps */ int i_pin_ref; - int i_rd_ref, i_rdcache_ref, i_wr_ref; + int i_rd_ref, i_rdcache_ref, i_wr_ref, i_wb_ref; int i_wrbuffer_ref, i_wrbuffer_ref_head; u32 i_shared_gen; /* increment each time we get FILE_SHARED */ u32 i_rdcache_gen; /* incremented each time we get FILE_CACHE. */ @@ -506,7 +506,7 @@ static inline int __ceph_caps_dirty(struct ceph_inode_info *ci) { return ci->i_dirty_caps | ci->i_flushing_caps; } -extern void __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask); +extern int __ceph_mark_dirty_caps(struct ceph_inode_info *ci, int mask); extern int ceph_caps_revoking(struct ceph_inode_info *ci, int mask); extern int __ceph_caps_used(struct ceph_inode_info *ci); diff --git a/fs/ceph/xattr.c b/fs/ceph/xattr.c index 8c9eba6ef9df..f2b628696180 100644 --- a/fs/ceph/xattr.c +++ b/fs/ceph/xattr.c @@ -703,6 +703,7 @@ int ceph_setxattr(struct dentry *dentry, const char *name, struct ceph_inode_xattr *xattr = NULL; int issued; int required_blob_size; + int dirty; if (ceph_snap(inode) != CEPH_NOSNAP) return -EROFS; @@ -763,11 +764,12 @@ retry: dout("setxattr %p issued %s\n", inode, ceph_cap_string(issued)); err = __set_xattr(ci, newname, name_len, newval, val_len, 1, 1, 1, &xattr); - __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL); ci->i_xattrs.dirty = true; inode->i_ctime = CURRENT_TIME; spin_unlock(&inode->i_lock); - + if (dirty) + __mark_inode_dirty(inode, dirty); return err; do_sync: @@ -810,6 +812,7 @@ int ceph_removexattr(struct dentry *dentry, const char *name) struct ceph_vxattr_cb *vxattrs = ceph_inode_vxattrs(inode); int issued; int err; + int dirty; if (ceph_snap(inode) != CEPH_NOSNAP) return -EROFS; @@ -833,12 +836,13 @@ int ceph_removexattr(struct dentry *dentry, const char *name) goto do_sync; err = __remove_xattr_by_name(ceph_inode(inode), name); - __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL); + dirty = __ceph_mark_dirty_caps(ci, CEPH_CAP_XATTR_EXCL); ci->i_xattrs.dirty = true; inode->i_ctime = CURRENT_TIME; spin_unlock(&inode->i_lock); - + if (dirty) + __mark_inode_dirty(inode, dirty); return err; do_sync: spin_unlock(&inode->i_lock); diff --git a/fs/cifs/AUTHORS b/fs/cifs/AUTHORS index 7f7fa3c302af..ea940b1db77b 100644 --- a/fs/cifs/AUTHORS +++ b/fs/cifs/AUTHORS @@ -35,7 +35,7 @@ Adrian Bunk (kcalloc cleanups) Miklos Szeredi Kazeon team for various fixes especially for 2.4 version. Asser Ferno (Change Notify support) -Shaggy (Dave Kleikamp) for inumerable small fs suggestions and some good cleanup +Shaggy (Dave Kleikamp) for innumerable small fs suggestions and some good cleanup Gunter Kukkukk (testing and suggestions for support of old servers) Igor Mammedov (DFS support) Jeff Layton (many, many fixes, as well as great work on the cifs Kerberos code) diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index 7cb0f7f847e4..1cd4c3a1862d 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -7,6 +7,7 @@ config CIFS select CRYPTO_MD5 select CRYPTO_HMAC select CRYPTO_ARC4 + select CRYPTO_DES help This is the client VFS module for the Common Internet File System (CIFS) protocol which is the successor to the Server Message Block @@ -152,16 +153,8 @@ config CIFS_ACL Allows to fetch CIFS/NTFS ACL from the server. The DACL blob is handed over to the application/caller. -config CIFS_EXPERIMENTAL - bool "CIFS Experimental Features (EXPERIMENTAL)" +config CIFS_NFSD_EXPORT + bool "Allow nfsd to export CIFS file system (EXPERIMENTAL)" depends on CIFS && EXPERIMENTAL help - Enables cifs features under testing. These features are - experimental and currently include DFS support and directory - change notification ie fcntl(F_DNOTIFY), as well as the upcall - mechanism which will be used for Kerberos session negotiation - and uid remapping. Some of these features also may depend on - setting a value of 1 to the pseudo-file /proc/fs/cifs/Experimental - (which is disabled by default). See the file fs/cifs/README - for more details. If unsure, say N. - + Allows NFS server to export a CIFS mounted share (nfsd over cifs) diff --git a/fs/cifs/Makefile b/fs/cifs/Makefile index d87558448e3d..005d524c3a4a 100644 --- a/fs/cifs/Makefile +++ b/fs/cifs/Makefile @@ -4,7 +4,7 @@ obj-$(CONFIG_CIFS) += cifs.o cifs-y := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o \ - link.o misc.o netmisc.o smbdes.o smbencrypt.o transport.o asn1.o \ + link.o misc.o netmisc.o smbencrypt.o transport.o asn1.o \ cifs_unicode.o nterr.o xattr.o cifsencrypt.o \ readdir.o ioctl.o sess.o export.o diff --git a/fs/cifs/README b/fs/cifs/README index fe1683590828..c5c2c5e5f0f2 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -457,6 +457,9 @@ A partial list of the supported mount options follows: otherwise - read from the server. All written data are stored in the cache, but if the client doesn't have Exclusive Oplock, it writes the data to the server. + rwpidforward Forward pid of a process who opened a file to any read or write + operation on that file. This prevent applications like WINE + from failing on read and write if we use mandatory brlock style. acl Allow setfacl and getfacl to manage posix ACLs if server supports them. (default) noacl Do not allow setfacl and getfacl calls on this mount @@ -685,22 +688,6 @@ LinuxExtensionsEnabled If set to one then the client will attempt to support and want to map the uid and gid fields to values supplied at mount (rather than the actual values, then set this to zero. (default 1) -Experimental When set to 1 used to enable certain experimental - features (currently enables multipage writes - when signing is enabled, the multipage write - performance enhancement was disabled when - signing turned on in case buffer was modified - just before it was sent, also this flag will - be used to use the new experimental directory change - notification code). When set to 2 enables - an additional experimental feature, "raw ntlmssp" - session establishment support (which allows - specifying "sec=ntlmssp" on mount). The Linux cifs - module will use ntlmv2 authentication encapsulated - in "raw ntlmssp" (not using SPNEGO) when - "sec=ntlmssp" is specified on mount. - This support also requires building cifs with - the CONFIG_CIFS_EXPERIMENTAL configuration flag. These experimental features and tracing can be enabled by changing flags in /proc/fs/cifs (after the cifs module has been installed or built into the @@ -720,18 +707,6 @@ the start of smb requests and responses can be enabled via: echo 1 > /proc/fs/cifs/traceSMB -Two other experimental features are under development. To test these -requires enabling CONFIG_CIFS_EXPERIMENTAL - - cifsacl support needed to retrieve approximated mode bits based on - the contents on the CIFS ACL. - - lease support: cifs will check the oplock state before calling into - the vfs to see if we can grant a lease on a file. - - DNOTIFY fcntl: needed for support of directory change - notification and perhaps later for file leases) - Per share (per client mount) statistics are available in /proc/fs/cifs/Stats if the kernel was configured with cifs statistics enabled. The statistics represent the number of successful (ie non-zero return code from the server) diff --git a/fs/cifs/cache.c b/fs/cifs/cache.c index e654dfd092c3..dd8584d35a14 100644 --- a/fs/cifs/cache.c +++ b/fs/cifs/cache.c @@ -50,7 +50,7 @@ void cifs_fscache_unregister(void) */ struct cifs_server_key { uint16_t family; /* address family */ - uint16_t port; /* IP port */ + __be16 port; /* IP port */ union { struct in_addr ipv4_addr; struct in6_addr ipv6_addr; @@ -146,7 +146,7 @@ static char *extract_sharename(const char *treename) static uint16_t cifs_super_get_key(const void *cookie_netfs_data, void *buffer, uint16_t maxbuf) { - const struct cifsTconInfo *tcon = cookie_netfs_data; + const struct cifs_tcon *tcon = cookie_netfs_data; char *sharename; uint16_t len; @@ -173,7 +173,7 @@ cifs_fscache_super_get_aux(const void *cookie_netfs_data, void *buffer, uint16_t maxbuf) { struct cifs_fscache_super_auxdata auxdata; - const struct cifsTconInfo *tcon = cookie_netfs_data; + const struct cifs_tcon *tcon = cookie_netfs_data; memset(&auxdata, 0, sizeof(auxdata)); auxdata.resource_id = tcon->resource_id; @@ -192,7 +192,7 @@ fscache_checkaux cifs_fscache_super_check_aux(void *cookie_netfs_data, uint16_t datalen) { struct cifs_fscache_super_auxdata auxdata; - const struct cifsTconInfo *tcon = cookie_netfs_data; + const struct cifs_tcon *tcon = cookie_netfs_data; if (datalen != sizeof(auxdata)) return FSCACHE_CHECKAUX_OBSOLETE; diff --git a/fs/cifs/cifs_debug.c b/fs/cifs/cifs_debug.c index 65829d32128c..2fe3cf13b2e9 100644 --- a/fs/cifs/cifs_debug.c +++ b/fs/cifs/cifs_debug.c @@ -63,7 +63,7 @@ void cifs_dump_detail(struct smb_hdr *smb) cERROR(1, "Cmd: %d Err: 0x%x Flags: 0x%x Flgs2: 0x%x Mid: %d Pid: %d", smb->Command, smb->Status.CifsError, smb->Flags, smb->Flags2, smb->Mid, smb->Pid); - cERROR(1, "smb buf %p len %d", smb, smbCalcSize_LE(smb)); + cERROR(1, "smb buf %p len %d", smb, smbCalcSize(smb)); } @@ -110,8 +110,8 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) struct list_head *tmp1, *tmp2, *tmp3; struct mid_q_entry *mid_entry; struct TCP_Server_Info *server; - struct cifsSesInfo *ses; - struct cifsTconInfo *tcon; + struct cifs_ses *ses; + struct cifs_tcon *tcon; int i, j; __u32 dev_type; @@ -152,7 +152,7 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) tcp_ses_list); i++; list_for_each(tmp2, &server->smb_ses_list) { - ses = list_entry(tmp2, struct cifsSesInfo, + ses = list_entry(tmp2, struct cifs_ses, smb_ses_list); if ((ses->serverDomain == NULL) || (ses->serverOS == NULL) || @@ -171,7 +171,7 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) seq_printf(m, "TCP status: %d\n\tLocal Users To " "Server: %d SecMode: 0x%x Req On Wire: %d", server->tcpStatus, server->srv_count, - server->secMode, + server->sec_mode, atomic_read(&server->inFlight)); #ifdef CONFIG_CIFS_STATS2 @@ -183,7 +183,7 @@ static int cifs_debug_data_proc_show(struct seq_file *m, void *v) seq_puts(m, "\n\tShares:"); j = 0; list_for_each(tmp3, &ses->tcon_list) { - tcon = list_entry(tmp3, struct cifsTconInfo, + tcon = list_entry(tmp3, struct cifs_tcon, tcon_list); ++j; dev_type = le32_to_cpu(tcon->fsDevInfo.DeviceType); @@ -256,8 +256,8 @@ static ssize_t cifs_stats_proc_write(struct file *file, int rc; struct list_head *tmp1, *tmp2, *tmp3; struct TCP_Server_Info *server; - struct cifsSesInfo *ses; - struct cifsTconInfo *tcon; + struct cifs_ses *ses; + struct cifs_tcon *tcon; rc = get_user(c, buffer); if (rc) @@ -273,11 +273,11 @@ static ssize_t cifs_stats_proc_write(struct file *file, server = list_entry(tmp1, struct TCP_Server_Info, tcp_ses_list); list_for_each(tmp2, &server->smb_ses_list) { - ses = list_entry(tmp2, struct cifsSesInfo, + ses = list_entry(tmp2, struct cifs_ses, smb_ses_list); list_for_each(tmp3, &ses->tcon_list) { tcon = list_entry(tmp3, - struct cifsTconInfo, + struct cifs_tcon, tcon_list); atomic_set(&tcon->num_smbs_sent, 0); atomic_set(&tcon->num_writes, 0); @@ -312,8 +312,8 @@ static int cifs_stats_proc_show(struct seq_file *m, void *v) int i; struct list_head *tmp1, *tmp2, *tmp3; struct TCP_Server_Info *server; - struct cifsSesInfo *ses; - struct cifsTconInfo *tcon; + struct cifs_ses *ses; + struct cifs_tcon *tcon; seq_printf(m, "Resources in use\nCIFS Session: %d\n", @@ -346,11 +346,11 @@ static int cifs_stats_proc_show(struct seq_file *m, void *v) server = list_entry(tmp1, struct TCP_Server_Info, tcp_ses_list); list_for_each(tmp2, &server->smb_ses_list) { - ses = list_entry(tmp2, struct cifsSesInfo, + ses = list_entry(tmp2, struct cifs_ses, smb_ses_list); list_for_each(tmp3, &ses->tcon_list) { tcon = list_entry(tmp3, - struct cifsTconInfo, + struct cifs_tcon, tcon_list); i++; seq_printf(m, "\n%d) %s", i, tcon->treeName); @@ -423,7 +423,6 @@ static const struct file_operations cifs_lookup_cache_proc_fops; static const struct file_operations traceSMB_proc_fops; static const struct file_operations cifs_multiuser_mount_proc_fops; static const struct file_operations cifs_security_flags_proc_fops; -static const struct file_operations cifs_experimental_proc_fops; static const struct file_operations cifs_linux_ext_proc_fops; void @@ -441,8 +440,6 @@ cifs_proc_init(void) proc_create("cifsFYI", 0, proc_fs_cifs, &cifsFYI_proc_fops); proc_create("traceSMB", 0, proc_fs_cifs, &traceSMB_proc_fops); proc_create("OplockEnabled", 0, proc_fs_cifs, &cifs_oplock_proc_fops); - proc_create("Experimental", 0, proc_fs_cifs, - &cifs_experimental_proc_fops); proc_create("LinuxExtensionsEnabled", 0, proc_fs_cifs, &cifs_linux_ext_proc_fops); proc_create("MultiuserMount", 0, proc_fs_cifs, @@ -469,7 +466,6 @@ cifs_proc_clean(void) remove_proc_entry("OplockEnabled", proc_fs_cifs); remove_proc_entry("SecurityFlags", proc_fs_cifs); remove_proc_entry("LinuxExtensionsEnabled", proc_fs_cifs); - remove_proc_entry("Experimental", proc_fs_cifs); remove_proc_entry("LookupCacheEnabled", proc_fs_cifs); remove_proc_entry("fs/cifs", NULL); } @@ -550,45 +546,6 @@ static const struct file_operations cifs_oplock_proc_fops = { .write = cifs_oplock_proc_write, }; -static int cifs_experimental_proc_show(struct seq_file *m, void *v) -{ - seq_printf(m, "%d\n", experimEnabled); - return 0; -} - -static int cifs_experimental_proc_open(struct inode *inode, struct file *file) -{ - return single_open(file, cifs_experimental_proc_show, NULL); -} - -static ssize_t cifs_experimental_proc_write(struct file *file, - const char __user *buffer, size_t count, loff_t *ppos) -{ - char c; - int rc; - - rc = get_user(c, buffer); - if (rc) - return rc; - if (c == '0' || c == 'n' || c == 'N') - experimEnabled = 0; - else if (c == '1' || c == 'y' || c == 'Y') - experimEnabled = 1; - else if (c == '2') - experimEnabled = 2; - - return count; -} - -static const struct file_operations cifs_experimental_proc_fops = { - .owner = THIS_MODULE, - .open = cifs_experimental_proc_open, - .read = seq_read, - .llseek = seq_lseek, - .release = single_release, - .write = cifs_experimental_proc_write, -}; - static int cifs_linux_ext_proc_show(struct seq_file *m, void *v) { seq_printf(m, "%d\n", linuxExtEnabled); diff --git a/fs/cifs/cifs_dfs_ref.c b/fs/cifs/cifs_dfs_ref.c index 0a265ad9e426..8d8f28c94c0f 100644 --- a/fs/cifs/cifs_dfs_ref.c +++ b/fs/cifs/cifs_dfs_ref.c @@ -53,7 +53,7 @@ void cifs_dfs_release_automount_timer(void) * * Extracts sharename form full UNC. * i.e. strips from UNC trailing path that is not part of share - * name and fixup missing '\' in the begining of DFS node refferal + * name and fixup missing '\' in the beginning of DFS node refferal * if necessary. * Returns pointer to share name on success or ERR_PTR on error. * Caller is responsible for freeing returned string. @@ -272,7 +272,7 @@ static struct vfsmount *cifs_dfs_do_automount(struct dentry *mntpt) struct dfs_info3_param *referrals = NULL; unsigned int num_referrals = 0; struct cifs_sb_info *cifs_sb; - struct cifsSesInfo *ses; + struct cifs_ses *ses; char *full_path; int xid, i; int rc; diff --git a/fs/cifs/cifs_fs_sb.h b/fs/cifs/cifs_fs_sb.h index ac51cd2d33ae..ffb1459dc6ec 100644 --- a/fs/cifs/cifs_fs_sb.h +++ b/fs/cifs/cifs_fs_sb.h @@ -41,6 +41,7 @@ #define CIFS_MOUNT_MF_SYMLINKS 0x10000 /* Minshall+French Symlinks enabled */ #define CIFS_MOUNT_MULTIUSER 0x20000 /* multiuser mount */ #define CIFS_MOUNT_STRICT_IO 0x40000 /* strict cache mode */ +#define CIFS_MOUNT_RWPIDFORWARD 0x80000 /* use pid forwarding for rw */ struct cifs_sb_info { struct rb_root tlink_tree; @@ -56,11 +57,7 @@ struct cifs_sb_info { mode_t mnt_file_mode; mode_t mnt_dir_mode; unsigned int mnt_cifs_flags; - int prepathlen; - char *prepath; /* relative path under the share to mount to */ -#ifdef CONFIG_CIFS_DFS_UPCALL - char *mountdata; /* mount options received at mount time */ -#endif + char *mountdata; /* options received at mount time or via DFS refs */ struct backing_dev_info bdi; struct delayed_work prune_tlinks; }; diff --git a/fs/cifs/cifs_spnego.c b/fs/cifs/cifs_spnego.c index 4dfba8283165..2272fd5fe5b7 100644 --- a/fs/cifs/cifs_spnego.c +++ b/fs/cifs/cifs_spnego.c @@ -95,7 +95,7 @@ struct key_type cifs_spnego_key_type = { /* get a key struct with a SPNEGO security blob, suitable for session setup */ struct key * -cifs_get_spnego_key(struct cifsSesInfo *sesInfo) +cifs_get_spnego_key(struct cifs_ses *sesInfo) { struct TCP_Server_Info *server = sesInfo->server; struct sockaddr_in *sa = (struct sockaddr_in *) &server->dstaddr; @@ -113,7 +113,7 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) MAX_MECH_STR_LEN + UID_KEY_LEN + (sizeof(uid_t) * 2) + CREDUID_KEY_LEN + (sizeof(uid_t) * 2) + - USER_KEY_LEN + strlen(sesInfo->userName) + + USER_KEY_LEN + strlen(sesInfo->user_name) + PID_KEY_LEN + (sizeof(pid_t) * 2) + 1; spnego_key = ERR_PTR(-ENOMEM); @@ -153,7 +153,7 @@ cifs_get_spnego_key(struct cifsSesInfo *sesInfo) sprintf(dp, ";creduid=0x%x", sesInfo->cred_uid); dp = description + strlen(description); - sprintf(dp, ";user=%s", sesInfo->userName); + sprintf(dp, ";user=%s", sesInfo->user_name); dp = description + strlen(description); sprintf(dp, ";pid=0x%x", current->pid); diff --git a/fs/cifs/cifs_spnego.h b/fs/cifs/cifs_spnego.h index e4041ec4d712..31bef9ee078b 100644 --- a/fs/cifs/cifs_spnego.h +++ b/fs/cifs/cifs_spnego.h @@ -41,7 +41,7 @@ struct cifs_spnego_msg { #ifdef __KERNEL__ extern struct key_type cifs_spnego_key_type; -extern struct key *cifs_get_spnego_key(struct cifsSesInfo *sesInfo); +extern struct key *cifs_get_spnego_key(struct cifs_ses *sesInfo); #endif /* KERNEL */ #endif /* _CIFS_SPNEGO_H */ diff --git a/fs/cifs/cifs_unicode.c b/fs/cifs/cifs_unicode.c index fc0fd4fde306..1b2e180b018d 100644 --- a/fs/cifs/cifs_unicode.c +++ b/fs/cifs/cifs_unicode.c @@ -90,7 +90,7 @@ cifs_mapchar(char *target, const __u16 src_char, const struct nls_table *cp, case UNI_COLON: *target = ':'; break; - case UNI_ASTERIK: + case UNI_ASTERISK: *target = '*'; break; case UNI_QUESTION: @@ -264,40 +264,41 @@ cifs_strndup_from_ucs(const char *src, const int maxlen, const bool is_unicode, * names are little endian 16 bit Unicode on the wire */ int -cifsConvertToUCS(__le16 *target, const char *source, int maxlen, +cifsConvertToUCS(__le16 *target, const char *source, int srclen, const struct nls_table *cp, int mapChars) { int i, j, charlen; - int len_remaining = maxlen; char src_char; - __u16 temp; + __le16 dst_char; + wchar_t tmp; if (!mapChars) return cifs_strtoUCS(target, source, PATH_MAX, cp); - for (i = 0, j = 0; i < maxlen; j++) { + for (i = 0, j = 0; i < srclen; j++) { src_char = source[i]; + charlen = 1; switch (src_char) { case 0: - put_unaligned_le16(0, &target[j]); + put_unaligned(0, &target[j]); goto ctoUCS_out; case ':': - temp = UNI_COLON; + dst_char = cpu_to_le16(UNI_COLON); break; case '*': - temp = UNI_ASTERIK; + dst_char = cpu_to_le16(UNI_ASTERISK); break; case '?': - temp = UNI_QUESTION; + dst_char = cpu_to_le16(UNI_QUESTION); break; case '<': - temp = UNI_LESSTHAN; + dst_char = cpu_to_le16(UNI_LESSTHAN); break; case '>': - temp = UNI_GRTRTHAN; + dst_char = cpu_to_le16(UNI_GRTRTHAN); break; case '|': - temp = UNI_PIPE; + dst_char = cpu_to_le16(UNI_PIPE); break; /* * FIXME: We can not handle remapping backslash (UNI_SLASH) @@ -305,28 +306,24 @@ cifsConvertToUCS(__le16 *target, const char *source, int maxlen, * as they use backslash as separator. */ default: - charlen = cp->char2uni(source+i, len_remaining, - &temp); + charlen = cp->char2uni(source + i, srclen - i, &tmp); + dst_char = cpu_to_le16(tmp); + /* * if no match, use question mark, which at least in * some cases serves as wild card */ if (charlen < 1) { - temp = 0x003f; + dst_char = cpu_to_le16(0x003f); charlen = 1; } - len_remaining -= charlen; - /* - * character may take more than one byte in the source - * string, but will take exactly two bytes in the - * target string - */ - i += charlen; - continue; } - put_unaligned_le16(temp, &target[j]); - i++; /* move to next char in source string */ - len_remaining--; + /* + * character may take more than one byte in the source string, + * but will take exactly two bytes in the target string + */ + i += charlen; + put_unaligned(dst_char, &target[j]); } ctoUCS_out: diff --git a/fs/cifs/cifs_unicode.h b/fs/cifs/cifs_unicode.h index 7fe6b52df507..6d02fd560566 100644 --- a/fs/cifs/cifs_unicode.h +++ b/fs/cifs/cifs_unicode.h @@ -44,7 +44,7 @@ * reserved symbols (along with \ and /), otherwise illegal to store * in filenames in NTFS */ -#define UNI_ASTERIK (__u16) ('*' + 0xF000) +#define UNI_ASTERISK (__u16) ('*' + 0xF000) #define UNI_QUESTION (__u16) ('?' + 0xF000) #define UNI_COLON (__u16) (':' + 0xF000) #define UNI_GRTRTHAN (__u16) ('>' + 0xF000) @@ -82,6 +82,9 @@ int cifs_strtoUCS(__le16 *, const char *, int, const struct nls_table *); char *cifs_strndup_from_ucs(const char *src, const int maxlen, const bool is_unicode, const struct nls_table *codepage); +extern int cifsConvertToUCS(__le16 *target, const char *source, int maxlen, + const struct nls_table *cp, int mapChars); + #endif /* diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index beeebf194234..21de1d6d5849 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -23,77 +23,405 @@ #include <linux/fs.h> #include <linux/slab.h> +#include <linux/string.h> +#include <linux/keyctl.h> +#include <linux/key-type.h> +#include <keys/user-type.h> #include "cifspdu.h" #include "cifsglob.h" #include "cifsacl.h" #include "cifsproto.h" #include "cifs_debug.h" - -static struct cifs_wksid wksidarr[NUM_WK_SIDS] = { - {{1, 0, {0, 0, 0, 0, 0, 0}, {0, 0, 0, 0, 0} }, "null user"}, - {{1, 1, {0, 0, 0, 0, 0, 1}, {0, 0, 0, 0, 0} }, "nobody"}, - {{1, 1, {0, 0, 0, 0, 0, 5}, {__constant_cpu_to_le32(11), 0, 0, 0, 0} }, "net-users"}, - {{1, 1, {0, 0, 0, 0, 0, 5}, {__constant_cpu_to_le32(18), 0, 0, 0, 0} }, "sys"}, - {{1, 2, {0, 0, 0, 0, 0, 5}, {__constant_cpu_to_le32(32), __constant_cpu_to_le32(544), 0, 0, 0} }, "root"}, - {{1, 2, {0, 0, 0, 0, 0, 5}, {__constant_cpu_to_le32(32), __constant_cpu_to_le32(545), 0, 0, 0} }, "users"}, - {{1, 2, {0, 0, 0, 0, 0, 5}, {__constant_cpu_to_le32(32), __constant_cpu_to_le32(546), 0, 0, 0} }, "guest"} } -; - - /* security id for everyone/world system group */ static const struct cifs_sid sid_everyone = { 1, 1, {0, 0, 0, 0, 0, 1}, {0} }; /* security id for Authenticated Users system group */ static const struct cifs_sid sid_authusers = { - 1, 1, {0, 0, 0, 0, 0, 5}, {11} }; + 1, 1, {0, 0, 0, 0, 0, 5}, {__constant_cpu_to_le32(11)} }; /* group users */ static const struct cifs_sid sid_user = {1, 2 , {0, 0, 0, 0, 0, 5}, {} }; +const struct cred *root_cred; -int match_sid(struct cifs_sid *ctsid) +static void +shrink_idmap_tree(struct rb_root *root, int nr_to_scan, int *nr_rem, + int *nr_del) { - int i, j; - int num_subauth, num_sat, num_saw; - struct cifs_sid *cwsid; + struct rb_node *node; + struct rb_node *tmp; + struct cifs_sid_id *psidid; + + node = rb_first(root); + while (node) { + tmp = node; + node = rb_next(tmp); + psidid = rb_entry(tmp, struct cifs_sid_id, rbnode); + if (nr_to_scan == 0 || *nr_del == nr_to_scan) + ++(*nr_rem); + else { + if (time_after(jiffies, psidid->time + SID_MAP_EXPIRE) + && psidid->refcount == 0) { + rb_erase(tmp, root); + ++(*nr_del); + } else + ++(*nr_rem); + } + } +} - if (!ctsid) - return -1; +/* + * Run idmap cache shrinker. + */ +static int +cifs_idmap_shrinker(struct shrinker *shrink, struct shrink_control *sc) +{ + int nr_to_scan = sc->nr_to_scan; + int nr_del = 0; + int nr_rem = 0; + struct rb_root *root; + + root = &uidtree; + spin_lock(&siduidlock); + shrink_idmap_tree(root, nr_to_scan, &nr_rem, &nr_del); + spin_unlock(&siduidlock); + + root = &gidtree; + spin_lock(&sidgidlock); + shrink_idmap_tree(root, nr_to_scan, &nr_rem, &nr_del); + spin_unlock(&sidgidlock); + + return nr_rem; +} - for (i = 0; i < NUM_WK_SIDS; ++i) { - cwsid = &(wksidarr[i].cifssid); +static struct shrinker cifs_shrinker = { + .shrink = cifs_idmap_shrinker, + .seeks = DEFAULT_SEEKS, +}; - /* compare the revision */ - if (ctsid->revision != cwsid->revision) - continue; +static int +cifs_idmap_key_instantiate(struct key *key, const void *data, size_t datalen) +{ + char *payload; + + payload = kmalloc(datalen, GFP_KERNEL); + if (!payload) + return -ENOMEM; - /* compare all of the six auth values */ - for (j = 0; j < 6; ++j) { - if (ctsid->authority[j] != cwsid->authority[j]) - break; + memcpy(payload, data, datalen); + key->payload.data = payload; + return 0; +} + +static inline void +cifs_idmap_key_destroy(struct key *key) +{ + kfree(key->payload.data); +} + +struct key_type cifs_idmap_key_type = { + .name = "cifs.idmap", + .instantiate = cifs_idmap_key_instantiate, + .destroy = cifs_idmap_key_destroy, + .describe = user_describe, + .match = user_match, +}; + +static void +sid_to_str(struct cifs_sid *sidptr, char *sidstr) +{ + int i; + unsigned long saval; + char *strptr; + + strptr = sidstr; + + sprintf(strptr, "%s", "S"); + strptr = sidstr + strlen(sidstr); + + sprintf(strptr, "-%d", sidptr->revision); + strptr = sidstr + strlen(sidstr); + + for (i = 0; i < 6; ++i) { + if (sidptr->authority[i]) { + sprintf(strptr, "-%d", sidptr->authority[i]); + strptr = sidstr + strlen(sidstr); } - if (j < 6) - continue; /* all of the auth values did not match */ - - /* compare all of the subauth values if any */ - num_sat = ctsid->num_subauth; - num_saw = cwsid->num_subauth; - num_subauth = num_sat < num_saw ? num_sat : num_saw; - if (num_subauth) { - for (j = 0; j < num_subauth; ++j) { - if (ctsid->sub_auth[j] != cwsid->sub_auth[j]) - break; - } - if (j < num_subauth) - continue; /* all sub_auth values do not match */ + } + + for (i = 0; i < sidptr->num_subauth; ++i) { + saval = le32_to_cpu(sidptr->sub_auth[i]); + sprintf(strptr, "-%ld", saval); + strptr = sidstr + strlen(sidstr); + } +} + +static void +id_rb_insert(struct rb_root *root, struct cifs_sid *sidptr, + struct cifs_sid_id **psidid, char *typestr) +{ + int rc; + char *strptr; + struct rb_node *node = root->rb_node; + struct rb_node *parent = NULL; + struct rb_node **linkto = &(root->rb_node); + struct cifs_sid_id *lsidid; + + while (node) { + lsidid = rb_entry(node, struct cifs_sid_id, rbnode); + parent = node; + rc = compare_sids(sidptr, &((lsidid)->sid)); + if (rc > 0) { + linkto = &(node->rb_left); + node = node->rb_left; + } else if (rc < 0) { + linkto = &(node->rb_right); + node = node->rb_right; } + } + + memcpy(&(*psidid)->sid, sidptr, sizeof(struct cifs_sid)); + (*psidid)->time = jiffies - (SID_MAP_RETRY + 1); + (*psidid)->refcount = 0; + + sprintf((*psidid)->sidstr, "%s", typestr); + strptr = (*psidid)->sidstr + strlen((*psidid)->sidstr); + sid_to_str(&(*psidid)->sid, strptr); + + clear_bit(SID_ID_PENDING, &(*psidid)->state); + clear_bit(SID_ID_MAPPED, &(*psidid)->state); + + rb_link_node(&(*psidid)->rbnode, parent, linkto); + rb_insert_color(&(*psidid)->rbnode, root); +} - cFYI(1, "matching sid: %s\n", wksidarr[i].sidname); - return 0; /* sids compare/match */ +static struct cifs_sid_id * +id_rb_search(struct rb_root *root, struct cifs_sid *sidptr) +{ + int rc; + struct rb_node *node = root->rb_node; + struct cifs_sid_id *lsidid; + + while (node) { + lsidid = rb_entry(node, struct cifs_sid_id, rbnode); + rc = compare_sids(sidptr, &((lsidid)->sid)); + if (rc > 0) { + node = node->rb_left; + } else if (rc < 0) { + node = node->rb_right; + } else /* node found */ + return lsidid; } - cFYI(1, "No matching sid"); - return -1; + return NULL; +} + +static int +sidid_pending_wait(void *unused) +{ + schedule(); + return signal_pending(current) ? -ERESTARTSYS : 0; +} + +static int +sid_to_id(struct cifs_sb_info *cifs_sb, struct cifs_sid *psid, + struct cifs_fattr *fattr, uint sidtype) +{ + int rc; + unsigned long cid; + struct key *idkey; + const struct cred *saved_cred; + struct cifs_sid_id *psidid, *npsidid; + struct rb_root *cidtree; + spinlock_t *cidlock; + + if (sidtype == SIDOWNER) { + cid = cifs_sb->mnt_uid; /* default uid, in case upcall fails */ + cidlock = &siduidlock; + cidtree = &uidtree; + } else if (sidtype == SIDGROUP) { + cid = cifs_sb->mnt_gid; /* default gid, in case upcall fails */ + cidlock = &sidgidlock; + cidtree = &gidtree; + } else + return -ENOENT; + + spin_lock(cidlock); + psidid = id_rb_search(cidtree, psid); + + if (!psidid) { /* node does not exist, allocate one & attempt adding */ + spin_unlock(cidlock); + npsidid = kzalloc(sizeof(struct cifs_sid_id), GFP_KERNEL); + if (!npsidid) + return -ENOMEM; + + npsidid->sidstr = kmalloc(SIDLEN, GFP_KERNEL); + if (!npsidid->sidstr) { + kfree(npsidid); + return -ENOMEM; + } + + spin_lock(cidlock); + psidid = id_rb_search(cidtree, psid); + if (psidid) { /* node happened to get inserted meanwhile */ + ++psidid->refcount; + spin_unlock(cidlock); + kfree(npsidid->sidstr); + kfree(npsidid); + } else { + psidid = npsidid; + id_rb_insert(cidtree, psid, &psidid, + sidtype == SIDOWNER ? "os:" : "gs:"); + ++psidid->refcount; + spin_unlock(cidlock); + } + } else { + ++psidid->refcount; + spin_unlock(cidlock); + } + + /* + * If we are here, it is safe to access psidid and its fields + * since a reference was taken earlier while holding the spinlock. + * A reference on the node is put without holding the spinlock + * and it is OK to do so in this case, shrinker will not erase + * this node until all references are put and we do not access + * any fields of the node after a reference is put . + */ + if (test_bit(SID_ID_MAPPED, &psidid->state)) { + cid = psidid->id; + psidid->time = jiffies; /* update ts for accessing */ + goto sid_to_id_out; + } + + if (time_after(psidid->time + SID_MAP_RETRY, jiffies)) + goto sid_to_id_out; + + if (!test_and_set_bit(SID_ID_PENDING, &psidid->state)) { + saved_cred = override_creds(root_cred); + idkey = request_key(&cifs_idmap_key_type, psidid->sidstr, ""); + if (IS_ERR(idkey)) + cFYI(1, "%s: Can't map SID to an id", __func__); + else { + cid = *(unsigned long *)idkey->payload.value; + psidid->id = cid; + set_bit(SID_ID_MAPPED, &psidid->state); + key_put(idkey); + kfree(psidid->sidstr); + } + revert_creds(saved_cred); + psidid->time = jiffies; /* update ts for accessing */ + clear_bit(SID_ID_PENDING, &psidid->state); + wake_up_bit(&psidid->state, SID_ID_PENDING); + } else { + rc = wait_on_bit(&psidid->state, SID_ID_PENDING, + sidid_pending_wait, TASK_INTERRUPTIBLE); + if (rc) { + cFYI(1, "%s: sidid_pending_wait interrupted %d", + __func__, rc); + --psidid->refcount; /* decremented without spinlock */ + return rc; + } + if (test_bit(SID_ID_MAPPED, &psidid->state)) + cid = psidid->id; + } + +sid_to_id_out: + --psidid->refcount; /* decremented without spinlock */ + if (sidtype == SIDOWNER) + fattr->cf_uid = cid; + else + fattr->cf_gid = cid; + + return 0; +} + +int +init_cifs_idmap(void) +{ + struct cred *cred; + struct key *keyring; + int ret; + + cFYI(1, "Registering the %s key type\n", cifs_idmap_key_type.name); + + /* create an override credential set with a special thread keyring in + * which requests are cached + * + * this is used to prevent malicious redirections from being installed + * with add_key(). + */ + cred = prepare_kernel_cred(NULL); + if (!cred) + return -ENOMEM; + + keyring = key_alloc(&key_type_keyring, ".cifs_idmap", 0, 0, cred, + (KEY_POS_ALL & ~KEY_POS_SETATTR) | + KEY_USR_VIEW | KEY_USR_READ, + KEY_ALLOC_NOT_IN_QUOTA); + if (IS_ERR(keyring)) { + ret = PTR_ERR(keyring); + goto failed_put_cred; + } + + ret = key_instantiate_and_link(keyring, NULL, 0, NULL, NULL); + if (ret < 0) + goto failed_put_key; + + ret = register_key_type(&cifs_idmap_key_type); + if (ret < 0) + goto failed_put_key; + + /* instruct request_key() to use this special keyring as a cache for + * the results it looks up */ + cred->thread_keyring = keyring; + cred->jit_keyring = KEY_REQKEY_DEFL_THREAD_KEYRING; + root_cred = cred; + + spin_lock_init(&siduidlock); + uidtree = RB_ROOT; + spin_lock_init(&sidgidlock); + gidtree = RB_ROOT; + + register_shrinker(&cifs_shrinker); + + cFYI(1, "cifs idmap keyring: %d\n", key_serial(keyring)); + return 0; + +failed_put_key: + key_put(keyring); +failed_put_cred: + put_cred(cred); + return ret; +} + +void +exit_cifs_idmap(void) +{ + key_revoke(root_cred->thread_keyring); + unregister_key_type(&cifs_idmap_key_type); + put_cred(root_cred); + unregister_shrinker(&cifs_shrinker); + cFYI(1, "Unregistered %s key type\n", cifs_idmap_key_type.name); +} + +void +cifs_destroy_idmaptrees(void) +{ + struct rb_root *root; + struct rb_node *node; + + root = &uidtree; + spin_lock(&siduidlock); + while ((node = rb_first(root))) + rb_erase(node, root); + spin_unlock(&siduidlock); + + root = &gidtree; + spin_lock(&sidgidlock); + while ((node = rb_first(root))) + rb_erase(node, root); + spin_unlock(&sidgidlock); } /* if the two SIDs (roughly equivalent to a UUID for a user or group) are @@ -104,16 +432,24 @@ int compare_sids(const struct cifs_sid *ctsid, const struct cifs_sid *cwsid) int num_subauth, num_sat, num_saw; if ((!ctsid) || (!cwsid)) - return 0; + return 1; /* compare the revision */ - if (ctsid->revision != cwsid->revision) - return 0; + if (ctsid->revision != cwsid->revision) { + if (ctsid->revision > cwsid->revision) + return 1; + else + return -1; + } /* compare all of the six auth values */ for (i = 0; i < 6; ++i) { - if (ctsid->authority[i] != cwsid->authority[i]) - return 0; + if (ctsid->authority[i] != cwsid->authority[i]) { + if (ctsid->authority[i] > cwsid->authority[i]) + return 1; + else + return -1; + } } /* compare all of the subauth values if any */ @@ -122,12 +458,17 @@ int compare_sids(const struct cifs_sid *ctsid, const struct cifs_sid *cwsid) num_subauth = num_sat < num_saw ? num_sat : num_saw; if (num_subauth) { for (i = 0; i < num_subauth; ++i) { - if (ctsid->sub_auth[i] != cwsid->sub_auth[i]) - return 0; + if (ctsid->sub_auth[i] != cwsid->sub_auth[i]) { + if (le32_to_cpu(ctsid->sub_auth[i]) > + le32_to_cpu(cwsid->sub_auth[i])) + return 1; + else + return -1; + } } } - return 1; /* sids compare/match */ + return 0; /* sids compare/match */ } @@ -382,22 +723,22 @@ static void parse_dacl(struct cifs_acl *pdacl, char *end_of_acl, #ifdef CONFIG_CIFS_DEBUG2 dump_ace(ppace[i], end_of_acl); #endif - if (compare_sids(&(ppace[i]->sid), pownersid)) + if (compare_sids(&(ppace[i]->sid), pownersid) == 0) access_flags_to_mode(ppace[i]->access_req, ppace[i]->type, &fattr->cf_mode, &user_mask); - if (compare_sids(&(ppace[i]->sid), pgrpsid)) + if (compare_sids(&(ppace[i]->sid), pgrpsid) == 0) access_flags_to_mode(ppace[i]->access_req, ppace[i]->type, &fattr->cf_mode, &group_mask); - if (compare_sids(&(ppace[i]->sid), &sid_everyone)) + if (compare_sids(&(ppace[i]->sid), &sid_everyone) == 0) access_flags_to_mode(ppace[i]->access_req, ppace[i]->type, &fattr->cf_mode, &other_mask); - if (compare_sids(&(ppace[i]->sid), &sid_authusers)) + if (compare_sids(&(ppace[i]->sid), &sid_authusers) == 0) access_flags_to_mode(ppace[i]->access_req, ppace[i]->type, &fattr->cf_mode, @@ -475,10 +816,10 @@ static int parse_sid(struct cifs_sid *psid, char *end_of_acl) /* Convert CIFS ACL to POSIX form */ -static int parse_sec_desc(struct cifs_ntsd *pntsd, int acl_len, - struct cifs_fattr *fattr) +static int parse_sec_desc(struct cifs_sb_info *cifs_sb, + struct cifs_ntsd *pntsd, int acl_len, struct cifs_fattr *fattr) { - int rc; + int rc = 0; struct cifs_sid *owner_sid_ptr, *group_sid_ptr; struct cifs_acl *dacl_ptr; /* no need for SACL ptr */ char *end_of_acl = ((char *)pntsd) + acl_len; @@ -500,12 +841,26 @@ static int parse_sec_desc(struct cifs_ntsd *pntsd, int acl_len, le32_to_cpu(pntsd->sacloffset), dacloffset); /* cifs_dump_mem("owner_sid: ", owner_sid_ptr, 64); */ rc = parse_sid(owner_sid_ptr, end_of_acl); - if (rc) + if (rc) { + cFYI(1, "%s: Error %d parsing Owner SID", __func__, rc); + return rc; + } + rc = sid_to_id(cifs_sb, owner_sid_ptr, fattr, SIDOWNER); + if (rc) { + cFYI(1, "%s: Error %d mapping Owner SID to uid", __func__, rc); return rc; + } rc = parse_sid(group_sid_ptr, end_of_acl); - if (rc) + if (rc) { + cFYI(1, "%s: Error %d mapping Owner SID to gid", __func__, rc); return rc; + } + rc = sid_to_id(cifs_sb, group_sid_ptr, fattr, SIDGROUP); + if (rc) { + cFYI(1, "%s: Error %d mapping Group SID to gid", __func__, rc); + return rc; + } if (dacloffset) parse_dacl(dacl_ptr, end_of_acl, owner_sid_ptr, @@ -520,7 +875,7 @@ static int parse_sec_desc(struct cifs_ntsd *pntsd, int acl_len, memcpy((void *)(&(cifscred->gsid)), (void *)group_sid_ptr, sizeof(struct cifs_sid)); */ - return 0; + return rc; } @@ -592,7 +947,7 @@ static struct cifs_ntsd *get_cifs_acl_by_path(struct cifs_sb_info *cifs_sb, int oplock = 0; int xid, rc; __u16 fid; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct tcon_link *tlink = cifs_sb_tlink(cifs_sb); if (IS_ERR(tlink)) @@ -660,7 +1015,7 @@ static int set_cifs_acl_by_path(struct cifs_sb_info *cifs_sb, const char *path, int oplock = 0; int xid, rc; __u16 fid; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct tcon_link *tlink = cifs_sb_tlink(cifs_sb); if (IS_ERR(tlink)) @@ -688,7 +1043,7 @@ out: } /* Set an ACL on the server */ -static int set_cifs_acl(struct cifs_ntsd *pnntsd, __u32 acllen, +int set_cifs_acl(struct cifs_ntsd *pnntsd, __u32 acllen, struct inode *inode, const char *path) { struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); @@ -727,7 +1082,7 @@ cifs_acl_to_fattr(struct cifs_sb_info *cifs_sb, struct cifs_fattr *fattr, rc = PTR_ERR(pntsd); cERROR(1, "%s: error %d getting sec desc", __func__, rc); } else { - rc = parse_sec_desc(pntsd, acllen, fattr); + rc = parse_sec_desc(cifs_sb, pntsd, acllen, fattr); kfree(pntsd); if (rc) cERROR(1, "parse sec desc failed rc = %d", rc); diff --git a/fs/cifs/cifsacl.h b/fs/cifs/cifsacl.h index c4ae7d036563..5c902c7ce524 100644 --- a/fs/cifs/cifsacl.h +++ b/fs/cifs/cifsacl.h @@ -39,6 +39,15 @@ #define ACCESS_ALLOWED 0 #define ACCESS_DENIED 1 +#define SIDOWNER 1 +#define SIDGROUP 2 +#define SIDLEN 150 /* S- 1 revision- 6 authorities- max 5 sub authorities */ + +#define SID_ID_MAPPED 0 +#define SID_ID_PENDING 1 +#define SID_MAP_EXPIRE (3600 * HZ) /* map entry expires after one hour */ +#define SID_MAP_RETRY (300 * HZ) /* wait 5 minutes for next attempt to map */ + struct cifs_ntsd { __le16 revision; /* revision level */ __le16 type; @@ -74,7 +83,21 @@ struct cifs_wksid { char sidname[SIDNAMELENGTH]; } __attribute__((packed)); -extern int match_sid(struct cifs_sid *); +struct cifs_sid_id { + unsigned int refcount; /* increment with spinlock, decrement without */ + unsigned long id; + unsigned long time; + unsigned long state; + char *sidstr; + struct rb_node rbnode; + struct cifs_sid sid; +}; + +#ifdef __KERNEL__ +extern struct key_type cifs_idmap_key_type; +extern const struct cred *root_cred; +#endif /* KERNEL */ + extern int compare_sids(const struct cifs_sid *, const struct cifs_sid *); #endif /* _CIFSACL_H */ diff --git a/fs/cifs/cifsencrypt.c b/fs/cifs/cifsencrypt.c index a51585f9852b..dfbd9f1f373d 100644 --- a/fs/cifs/cifsencrypt.c +++ b/fs/cifs/cifsencrypt.c @@ -30,12 +30,13 @@ #include <linux/ctype.h> #include <linux/random.h> -/* Calculate and return the CIFS signature based on the mac key and SMB PDU */ -/* the 16 byte signature must be allocated by the caller */ -/* Note we only use the 1st eight bytes */ -/* Note that the smb header signature field on input contains the - sequence number before this function is called */ - +/* + * Calculate and return the CIFS signature based on the mac key and SMB PDU. + * The 16 byte signature must be allocated by the caller. Note we only use the + * 1st eight bytes and that the smb header signature field on input contains + * the sequence number before this function is called. Also, this function + * should be called with the server->srv_mutex held. + */ static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, struct TCP_Server_Info *server, char *signature) { @@ -59,7 +60,7 @@ static int cifs_calculate_signature(const struct smb_hdr *cifs_pdu, server->session_key.response, server->session_key.len); crypto_shash_update(&server->secmech.sdescmd5->shash, - cifs_pdu->Protocol, cifs_pdu->smb_buf_length); + cifs_pdu->Protocol, be32_to_cpu(cifs_pdu->smb_buf_length)); rc = crypto_shash_final(&server->secmech.sdescmd5->shash, signature); @@ -209,8 +210,10 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, cpu_to_le32(expected_sequence_number); cifs_pdu->Signature.Sequence.Reserved = 0; + mutex_lock(&server->srv_mutex); rc = cifs_calculate_signature(cifs_pdu, server, what_we_think_sig_should_be); + mutex_unlock(&server->srv_mutex); if (rc) return rc; @@ -226,7 +229,7 @@ int cifs_verify_signature(struct smb_hdr *cifs_pdu, } /* first calculate 24 bytes ntlm response and then 16 byte session key */ -int setup_ntlm_response(struct cifsSesInfo *ses) +int setup_ntlm_response(struct cifs_ses *ses) { int rc = 0; unsigned int temp_len = CIFS_SESS_KEY_SIZE + CIFS_AUTH_RESP_SIZE; @@ -265,10 +268,11 @@ int setup_ntlm_response(struct cifsSesInfo *ses) } #ifdef CONFIG_CIFS_WEAK_PW_HASH -void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, +int calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, char *lnm_session_key) { int i; + int rc; char password_with_pad[CIFS_ENCPWD_SIZE]; memset(password_with_pad, 0, CIFS_ENCPWD_SIZE); @@ -279,7 +283,7 @@ void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, memset(lnm_session_key, 0, CIFS_SESS_KEY_SIZE); memcpy(lnm_session_key, password_with_pad, CIFS_ENCPWD_SIZE); - return; + return 0; } /* calculate old style session key */ @@ -296,10 +300,9 @@ void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, for (i = 0; i < CIFS_ENCPWD_SIZE; i++) password_with_pad[i] = toupper(password_with_pad[i]); - SMBencrypt(password_with_pad, cryptkey, lnm_session_key); + rc = SMBencrypt(password_with_pad, cryptkey, lnm_session_key); - /* clear password before we return/free memory */ - memset(password_with_pad, 0, CIFS_ENCPWD_SIZE); + return rc; } #endif /* CIFS_WEAK_PW_HASH */ @@ -309,7 +312,7 @@ void calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, * Allocate domain name which gets freed when session struct is deallocated. */ static int -build_avpair_blob(struct cifsSesInfo *ses, const struct nls_table *nls_cp) +build_avpair_blob(struct cifs_ses *ses, const struct nls_table *nls_cp) { unsigned int dlen; unsigned int wlen; @@ -397,7 +400,7 @@ build_avpair_blob(struct cifsSesInfo *ses, const struct nls_table *nls_cp) * about target string i.e. for some, just user name might suffice. */ static int -find_domain_name(struct cifsSesInfo *ses, const struct nls_table *nls_cp) +find_domain_name(struct cifs_ses *ses, const struct nls_table *nls_cp) { unsigned int attrsize; unsigned int type; @@ -442,7 +445,7 @@ find_domain_name(struct cifsSesInfo *ses, const struct nls_table *nls_cp) return 0; } -static int calc_ntlmv2_hash(struct cifsSesInfo *ses, char *ntlmv2_hash, +static int calc_ntlmv2_hash(struct cifs_ses *ses, char *ntlmv2_hash, const struct nls_table *nls_cp) { int rc = 0; @@ -469,15 +472,15 @@ static int calc_ntlmv2_hash(struct cifsSesInfo *ses, char *ntlmv2_hash, return rc; } - /* convert ses->userName to unicode and uppercase */ - len = strlen(ses->userName); + /* convert ses->user_name to unicode and uppercase */ + len = strlen(ses->user_name); user = kmalloc(2 + (len * 2), GFP_KERNEL); if (user == NULL) { cERROR(1, "calc_ntlmv2_hash: user mem alloc failure\n"); rc = -ENOMEM; goto calc_exit_2; } - len = cifs_strtoUCS((__le16 *)user, ses->userName, len, nls_cp); + len = cifs_strtoUCS((__le16 *)user, ses->user_name, len, nls_cp); UniStrupr(user); crypto_shash_update(&ses->server->secmech.sdeschmacmd5->shash, @@ -524,7 +527,7 @@ calc_exit_2: } static int -CalcNTLMv2_response(const struct cifsSesInfo *ses, char *ntlmv2_hash) +CalcNTLMv2_response(const struct cifs_ses *ses, char *ntlmv2_hash) { int rc; unsigned int offset = CIFS_SESS_KEY_SIZE + 8; @@ -560,7 +563,7 @@ CalcNTLMv2_response(const struct cifsSesInfo *ses, char *ntlmv2_hash) int -setup_ntlmv2_rsp(struct cifsSesInfo *ses, const struct nls_table *nls_cp) +setup_ntlmv2_rsp(struct cifs_ses *ses, const struct nls_table *nls_cp) { int rc; int baselen; @@ -646,7 +649,7 @@ setup_ntlmv2_rsp_ret: } int -calc_seckey(struct cifsSesInfo *ses) +calc_seckey(struct cifs_ses *ses) { int rc; struct crypto_blkcipher *tfm_arc4; diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index f2970136d17d..989442dcfb45 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -53,7 +53,6 @@ int cifsFYI = 0; int cifsERROR = 1; int traceSMB = 0; unsigned int oplockEnabled = 1; -unsigned int experimEnabled = 0; unsigned int linuxExtEnabled = 1; unsigned int lookupCacheEnabled = 1; unsigned int multiuser_mount = 0; @@ -105,52 +104,25 @@ cifs_sb_deactive(struct super_block *sb) } static int -cifs_read_super(struct super_block *sb, void *data, +cifs_read_super(struct super_block *sb, struct smb_vol *volume_info, const char *devname, int silent) { struct inode *inode; struct cifs_sb_info *cifs_sb; int rc = 0; - /* BB should we make this contingent on mount parm? */ - sb->s_flags |= MS_NODIRATIME | MS_NOATIME; - sb->s_fs_info = kzalloc(sizeof(struct cifs_sb_info), GFP_KERNEL); cifs_sb = CIFS_SB(sb); - if (cifs_sb == NULL) - return -ENOMEM; spin_lock_init(&cifs_sb->tlink_tree_lock); cifs_sb->tlink_tree = RB_ROOT; rc = bdi_setup_and_register(&cifs_sb->bdi, "cifs", BDI_CAP_MAP_COPY); - if (rc) { - kfree(cifs_sb); + if (rc) return rc; - } -#ifdef CONFIG_CIFS_DFS_UPCALL - /* copy mount params to sb for use in submounts */ - /* BB: should we move this after the mount so we - * do not have to do the copy on failed mounts? - * BB: May be it is better to do simple copy before - * complex operation (mount), and in case of fail - * just exit instead of doing mount and attempting - * undo it if this copy fails?*/ - if (data) { - int len = strlen(data); - cifs_sb->mountdata = kzalloc(len + 1, GFP_KERNEL); - if (cifs_sb->mountdata == NULL) { - bdi_destroy(&cifs_sb->bdi); - kfree(sb->s_fs_info); - sb->s_fs_info = NULL; - return -ENOMEM; - } - strncpy(cifs_sb->mountdata, data, len + 1); - cifs_sb->mountdata[len] = '\0'; - } -#endif + cifs_sb->bdi.ra_pages = default_backing_dev_info.ra_pages; - rc = cifs_mount(sb, cifs_sb, data, devname); + rc = cifs_mount(sb, cifs_sb, volume_info, devname); if (rc) { if (!silent) @@ -163,7 +135,7 @@ cifs_read_super(struct super_block *sb, void *data, sb->s_bdi = &cifs_sb->bdi; sb->s_blocksize = CIFS_MAX_MSGSIZE; sb->s_blocksize_bits = 14; /* default 2**14 = CIFS_MAX_MSGSIZE */ - inode = cifs_root_iget(sb, ROOT_I); + inode = cifs_root_iget(sb); if (IS_ERR(inode)) { rc = PTR_ERR(inode); @@ -184,12 +156,12 @@ cifs_read_super(struct super_block *sb, void *data, else sb->s_d_op = &cifs_dentry_ops; -#ifdef CONFIG_CIFS_EXPERIMENTAL +#ifdef CIFS_NFSD_EXPORT if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) { cFYI(1, "export ops supported"); sb->s_export_op = &cifs_export_ops; } -#endif /* EXPERIMENTAL */ +#endif /* CIFS_NFSD_EXPORT */ return 0; @@ -201,17 +173,7 @@ out_no_root: cifs_umount(sb, cifs_sb); out_mount_failed: - if (cifs_sb) { -#ifdef CONFIG_CIFS_DFS_UPCALL - if (cifs_sb->mountdata) { - kfree(cifs_sb->mountdata); - cifs_sb->mountdata = NULL; - } -#endif - unload_nls(cifs_sb->local_nls); - bdi_destroy(&cifs_sb->bdi); - kfree(cifs_sb); - } + bdi_destroy(&cifs_sb->bdi); return rc; } @@ -231,12 +193,10 @@ cifs_put_super(struct super_block *sb) rc = cifs_umount(sb, cifs_sb); if (rc) cERROR(1, "cifs_umount failed with return code %d", rc); -#ifdef CONFIG_CIFS_DFS_UPCALL if (cifs_sb->mountdata) { kfree(cifs_sb->mountdata); cifs_sb->mountdata = NULL; } -#endif unload_nls(cifs_sb->local_nls); bdi_destroy(&cifs_sb->bdi); @@ -248,7 +208,7 @@ cifs_statfs(struct dentry *dentry, struct kstatfs *buf) { struct super_block *sb = dentry->d_sb; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); - struct cifsTconInfo *tcon = cifs_sb_master_tcon(cifs_sb); + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); int rc = -EOPNOTSUPP; int xid; @@ -401,7 +361,7 @@ static int cifs_show_options(struct seq_file *s, struct vfsmount *m) { struct cifs_sb_info *cifs_sb = CIFS_SB(m->mnt_sb); - struct cifsTconInfo *tcon = cifs_sb_master_tcon(cifs_sb); + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); struct sockaddr *srcaddr; srcaddr = (struct sockaddr *)&tcon->ses->server->srcaddr; @@ -409,8 +369,8 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER) seq_printf(s, ",multiuser"); - else if (tcon->ses->userName) - seq_printf(s, ",username=%s", tcon->ses->userName); + else if (tcon->ses->user_name) + seq_printf(s, ",username=%s", tcon->ses->user_name); if (tcon->ses->domainName) seq_printf(s, ",domain=%s", tcon->ses->domainName); @@ -455,14 +415,20 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) seq_printf(s, ",nocase"); if (tcon->retry) seq_printf(s, ",hard"); - if (cifs_sb->prepath) - seq_printf(s, ",prepath=%s", cifs_sb->prepath); + if (tcon->unix_ext) + seq_printf(s, ",unix"); + else + seq_printf(s, ",nounix"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) seq_printf(s, ",posixpaths"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SET_UID) seq_printf(s, ",setuids"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_SERVER_INUM) seq_printf(s, ",serverino"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD) + seq_printf(s, ",rwpidforward"); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) + seq_printf(s, ",forcemand"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_DIRECT_IO) seq_printf(s, ",directio"); if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NO_XATTR) @@ -495,7 +461,7 @@ cifs_show_options(struct seq_file *s, struct vfsmount *m) static void cifs_umount_begin(struct super_block *sb) { struct cifs_sb_info *cifs_sb = CIFS_SB(sb); - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; if (cifs_sb == NULL) return; @@ -570,29 +536,189 @@ static const struct super_operations cifs_super_ops = { #endif }; +/* + * Get root dentry from superblock according to prefix path mount option. + * Return dentry with refcount + 1 on success and NULL otherwise. + */ +static struct dentry * +cifs_get_root(struct smb_vol *vol, struct super_block *sb) +{ + int xid, rc; + struct inode *inode; + struct qstr name; + struct dentry *dparent = NULL, *dchild = NULL, *alias; + struct cifs_sb_info *cifs_sb = CIFS_SB(sb); + unsigned int i, full_len, len; + char *full_path = NULL, *pstart; + char sep; + + full_path = cifs_build_path_to_root(vol, cifs_sb, + cifs_sb_master_tcon(cifs_sb)); + if (full_path == NULL) + return NULL; + + cFYI(1, "Get root dentry for %s", full_path); + + xid = GetXid(); + sep = CIFS_DIR_SEP(cifs_sb); + dparent = dget(sb->s_root); + full_len = strlen(full_path); + full_path[full_len] = sep; + pstart = full_path + 1; + + for (i = 1, len = 0; i <= full_len; i++) { + if (full_path[i] != sep || !len) { + len++; + continue; + } + + full_path[i] = 0; + cFYI(1, "get dentry for %s", pstart); + + name.name = pstart; + name.len = len; + name.hash = full_name_hash(pstart, len); + dchild = d_lookup(dparent, &name); + if (dchild == NULL) { + cFYI(1, "not exists"); + dchild = d_alloc(dparent, &name); + if (dchild == NULL) { + dput(dparent); + dparent = NULL; + goto out; + } + } + + cFYI(1, "get inode"); + if (dchild->d_inode == NULL) { + cFYI(1, "not exists"); + inode = NULL; + if (cifs_sb_master_tcon(CIFS_SB(sb))->unix_ext) + rc = cifs_get_inode_info_unix(&inode, full_path, + sb, xid); + else + rc = cifs_get_inode_info(&inode, full_path, + NULL, sb, xid, NULL); + if (rc) { + dput(dchild); + dput(dparent); + dparent = NULL; + goto out; + } + alias = d_materialise_unique(dchild, inode); + if (alias != NULL) { + dput(dchild); + if (IS_ERR(alias)) { + dput(dparent); + dparent = NULL; + goto out; + } + dchild = alias; + } + } + cFYI(1, "parent %p, child %p", dparent, dchild); + + dput(dparent); + dparent = dchild; + len = 0; + pstart = full_path + i + 1; + full_path[i] = sep; + } +out: + _FreeXid(xid); + kfree(full_path); + return dparent; +} + static struct dentry * cifs_do_mount(struct file_system_type *fs_type, - int flags, const char *dev_name, void *data) + int flags, const char *dev_name, void *data) { int rc; struct super_block *sb; - - sb = sget(fs_type, NULL, set_anon_super, NULL); + struct cifs_sb_info *cifs_sb; + struct smb_vol *volume_info; + struct cifs_mnt_data mnt_data; + struct dentry *root; cFYI(1, "Devname: %s flags: %d ", dev_name, flags); - if (IS_ERR(sb)) - return ERR_CAST(sb); + rc = cifs_setup_volume_info(&volume_info, (char *)data, dev_name); + if (rc) + return ERR_PTR(rc); + + cifs_sb = kzalloc(sizeof(struct cifs_sb_info), GFP_KERNEL); + if (cifs_sb == NULL) { + root = ERR_PTR(-ENOMEM); + goto out; + } + + cifs_setup_cifs_sb(volume_info, cifs_sb); + + mnt_data.vol = volume_info; + mnt_data.cifs_sb = cifs_sb; + mnt_data.flags = flags; + + sb = sget(fs_type, cifs_match_super, set_anon_super, &mnt_data); + if (IS_ERR(sb)) { + root = ERR_CAST(sb); + goto out_cifs_sb; + } + + if (sb->s_fs_info) { + cFYI(1, "Use existing superblock"); + goto out_shared; + } + + /* + * Copy mount params for use in submounts. Better to do + * the copy here and deal with the error before cleanup gets + * complicated post-mount. + */ + cifs_sb->mountdata = kstrndup(data, PAGE_SIZE, GFP_KERNEL); + if (cifs_sb->mountdata == NULL) { + root = ERR_PTR(-ENOMEM); + goto out_super; + } sb->s_flags = flags; + /* BB should we make this contingent on mount parm? */ + sb->s_flags |= MS_NODIRATIME | MS_NOATIME; + sb->s_fs_info = cifs_sb; - rc = cifs_read_super(sb, data, dev_name, flags & MS_SILENT ? 1 : 0); + rc = cifs_read_super(sb, volume_info, dev_name, + flags & MS_SILENT ? 1 : 0); if (rc) { - deactivate_locked_super(sb); - return ERR_PTR(rc); + root = ERR_PTR(rc); + goto out_super; } + sb->s_flags |= MS_ACTIVE; - return dget(sb->s_root); + + root = cifs_get_root(volume_info, sb); + if (root == NULL) + goto out_super; + + cFYI(1, "dentry root is: %p", root); + goto out; + +out_shared: + root = cifs_get_root(volume_info, sb); + if (root) + cFYI(1, "dentry root is: %p", root); + goto out; + +out_super: + kfree(cifs_sb->mountdata); + deactivate_locked_super(sb); + +out_cifs_sb: + unload_nls(cifs_sb->local_nls); + kfree(cifs_sb); + +out: + cifs_cleanup_volume_info(&volume_info); + return root; } static ssize_t cifs_file_aio_write(struct kiocb *iocb, const struct iovec *iov, @@ -618,16 +744,31 @@ static loff_t cifs_llseek(struct file *file, loff_t offset, int origin) { /* origin == SEEK_END => we must revalidate the cached file length */ if (origin == SEEK_END) { - int retval; - - /* some applications poll for the file length in this strange - way so we must seek to end on non-oplocked files by - setting the revalidate time to zero */ - CIFS_I(file->f_path.dentry->d_inode)->time = 0; - - retval = cifs_revalidate_file(file); - if (retval < 0) - return (loff_t)retval; + int rc; + struct inode *inode = file->f_path.dentry->d_inode; + + /* + * We need to be sure that all dirty pages are written and the + * server has the newest file length. + */ + if (!CIFS_I(inode)->clientCanCacheRead && inode->i_mapping && + inode->i_mapping->nrpages != 0) { + rc = filemap_fdatawait(inode->i_mapping); + if (rc) { + mapping_set_error(inode->i_mapping, rc); + return rc; + } + } + /* + * Some applications poll for the file length in this strange + * way so we must seek to end on non-oplocked files by + * setting the revalidate time to zero. + */ + CIFS_I(inode)->time = 0; + + rc = cifs_revalidate_file_attr(file); + if (rc < 0) + return (loff_t)rc; } return generic_file_llseek_unlocked(file, offset, origin); } @@ -760,10 +901,11 @@ const struct file_operations cifs_file_strict_ops = { }; const struct file_operations cifs_file_direct_ops = { - /* no aio, no readv - - BB reevaluate whether they can be done with directio, no cache */ - .read = cifs_user_read, - .write = cifs_user_write, + /* BB reevaluate whether they can be done with directio, no cache */ + .read = do_sync_read, + .write = do_sync_write, + .aio_read = cifs_user_readv, + .aio_write = cifs_user_writev, .open = cifs_open, .release = cifs_close, .lock = cifs_lock, @@ -815,10 +957,11 @@ const struct file_operations cifs_file_strict_nobrl_ops = { }; const struct file_operations cifs_file_direct_nobrl_ops = { - /* no mmap, no aio, no readv - - BB reevaluate whether they can be done with directio, no cache */ - .read = cifs_user_read, - .write = cifs_user_write, + /* BB reevaluate whether they can be done with directio, no cache */ + .read = do_sync_read, + .write = do_sync_write, + .aio_read = cifs_user_readv, + .aio_write = cifs_user_writev, .open = cifs_open, .release = cifs_close, .fsync = cifs_fsync, @@ -981,10 +1124,10 @@ init_cifs(void) int rc = 0; cifs_proc_init(); INIT_LIST_HEAD(&cifs_tcp_ses_list); -#ifdef CONFIG_CIFS_EXPERIMENTAL +#ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* unused temporarily */ INIT_LIST_HEAD(&GlobalDnotifyReqList); INIT_LIST_HEAD(&GlobalDnotifyRsp_Q); -#endif +#endif /* was needed for dnotify, and will be needed for inotify when VFS fix */ /* * Initialize Global counters */ @@ -1033,22 +1176,33 @@ init_cifs(void) if (rc) goto out_destroy_mids; - rc = register_filesystem(&cifs_fs_type); - if (rc) - goto out_destroy_request_bufs; #ifdef CONFIG_CIFS_UPCALL rc = register_key_type(&cifs_spnego_key_type); if (rc) - goto out_unregister_filesystem; -#endif + goto out_destroy_request_bufs; +#endif /* CONFIG_CIFS_UPCALL */ + +#ifdef CONFIG_CIFS_ACL + rc = init_cifs_idmap(); + if (rc) + goto out_register_key_type; +#endif /* CONFIG_CIFS_ACL */ + + rc = register_filesystem(&cifs_fs_type); + if (rc) + goto out_init_cifs_idmap; return 0; -#ifdef CONFIG_CIFS_UPCALL -out_unregister_filesystem: - unregister_filesystem(&cifs_fs_type); +out_init_cifs_idmap: +#ifdef CONFIG_CIFS_ACL + exit_cifs_idmap(); +out_register_key_type: #endif +#ifdef CONFIG_CIFS_UPCALL + unregister_key_type(&cifs_spnego_key_type); out_destroy_request_bufs: +#endif cifs_destroy_request_bufs(); out_destroy_mids: cifs_destroy_mids(); @@ -1070,6 +1224,10 @@ exit_cifs(void) #ifdef CONFIG_CIFS_DFS_UPCALL cifs_dfs_release_automount_timer(); #endif +#ifdef CONFIG_CIFS_ACL + cifs_destroy_idmaptrees(); + exit_cifs_idmap(); +#endif #ifdef CONFIG_CIFS_UPCALL unregister_key_type(&cifs_spnego_key_type); #endif diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index a9371b6578c0..64313f778ebf 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -47,7 +47,7 @@ extern void cifs_sb_deactive(struct super_block *sb); /* Functions related to inodes */ extern const struct inode_operations cifs_dir_inode_ops; -extern struct inode *cifs_root_iget(struct super_block *, unsigned long); +extern struct inode *cifs_root_iget(struct super_block *); extern int cifs_create(struct inode *, struct dentry *, int, struct nameidata *); extern struct dentry *cifs_lookup(struct inode *, struct dentry *, @@ -59,9 +59,11 @@ extern int cifs_mkdir(struct inode *, struct dentry *, int); extern int cifs_rmdir(struct inode *, struct dentry *); extern int cifs_rename(struct inode *, struct dentry *, struct inode *, struct dentry *); +extern int cifs_revalidate_file_attr(struct file *filp); +extern int cifs_revalidate_dentry_attr(struct dentry *); extern int cifs_revalidate_file(struct file *filp); extern int cifs_revalidate_dentry(struct dentry *); -extern void cifs_invalidate_mapping(struct inode *inode); +extern int cifs_invalidate_mapping(struct inode *inode); extern int cifs_getattr(struct vfsmount *, struct dentry *, struct kstat *); extern int cifs_setattr(struct dentry *, struct iattr *); @@ -80,12 +82,12 @@ extern const struct file_operations cifs_file_strict_nobrl_ops; extern int cifs_open(struct inode *inode, struct file *file); extern int cifs_close(struct inode *inode, struct file *file); extern int cifs_closedir(struct inode *inode, struct file *file); -extern ssize_t cifs_user_read(struct file *file, char __user *read_data, - size_t read_size, loff_t *poffset); +extern ssize_t cifs_user_readv(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos); extern ssize_t cifs_strict_readv(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos); -extern ssize_t cifs_user_write(struct file *file, const char __user *write_data, - size_t write_size, loff_t *poffset); +extern ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos); extern ssize_t cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos); extern int cifs_lock(struct file *, int, struct file_lock *); @@ -123,9 +125,9 @@ extern ssize_t cifs_getxattr(struct dentry *, const char *, void *, size_t); extern ssize_t cifs_listxattr(struct dentry *, char *, size_t); extern long cifs_ioctl(struct file *filep, unsigned int cmd, unsigned long arg); -#ifdef CONFIG_CIFS_EXPERIMENTAL +#ifdef CIFS_NFSD_EXPORT extern const struct export_operations cifs_export_ops; -#endif /* EXPERIMENTAL */ +#endif /* CIFS_NFSD_EXPORT */ -#define CIFS_VERSION "1.71" +#define CIFS_VERSION "1.72" #endif /* _CIFSFS_H */ diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index 17afb0fbcaed..6255fa812c7a 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -37,10 +37,9 @@ #define MAX_TREE_SIZE (2 + MAX_SERVER_SIZE + 1 + MAX_SHARE_SIZE + 1) #define MAX_SERVER_SIZE 15 -#define MAX_SHARE_SIZE 64 /* used to be 20, this should still be enough */ -#define MAX_USERNAME_SIZE 32 /* 32 is to allow for 15 char names + null - termination then *2 for unicode versions */ -#define MAX_PASSWORD_SIZE 512 /* max for windows seems to be 256 wide chars */ +#define MAX_SHARE_SIZE 80 +#define MAX_USERNAME_SIZE 256 /* reasonable maximum for current servers */ +#define MAX_PASSWORD_SIZE 512 /* max for windows seems to be 256 wide chars */ #define CIFS_MIN_RCV_POOL 4 @@ -92,7 +91,8 @@ enum statusEnum { CifsNew = 0, CifsGood, CifsExiting, - CifsNeedReconnect + CifsNeedReconnect, + CifsNeedNegotiate }; enum securityEnum { @@ -155,6 +155,81 @@ struct cifs_cred { ***************************************************************** */ +struct smb_vol { + char *username; + char *password; + char *domainname; + char *UNC; + char *UNCip; + char *iocharset; /* local code page for mapping to and from Unicode */ + char source_rfc1001_name[RFC1001_NAME_LEN_WITH_NULL]; /* clnt nb name */ + char target_rfc1001_name[RFC1001_NAME_LEN_WITH_NULL]; /* srvr nb name */ + uid_t cred_uid; + uid_t linux_uid; + gid_t linux_gid; + mode_t file_mode; + mode_t dir_mode; + unsigned secFlg; + bool retry:1; + bool intr:1; + bool setuids:1; + bool override_uid:1; + bool override_gid:1; + bool dynperm:1; + bool noperm:1; + bool no_psx_acl:1; /* set if posix acl support should be disabled */ + bool cifs_acl:1; + bool no_xattr:1; /* set if xattr (EA) support should be disabled*/ + bool server_ino:1; /* use inode numbers from server ie UniqueId */ + bool direct_io:1; + bool strict_io:1; /* strict cache behavior */ + bool remap:1; /* set to remap seven reserved chars in filenames */ + bool posix_paths:1; /* unset to not ask for posix pathnames. */ + bool no_linux_ext:1; + bool sfu_emul:1; + bool nullauth:1; /* attempt to authenticate with null user */ + bool nocase:1; /* request case insensitive filenames */ + bool nobrl:1; /* disable sending byte range locks to srv */ + bool mand_lock:1; /* send mandatory not posix byte range lock reqs */ + bool seal:1; /* request transport encryption on share */ + bool nodfs:1; /* Do not request DFS, even if available */ + bool local_lease:1; /* check leases only on local system, not remote */ + bool noblocksnd:1; + bool noautotune:1; + bool nostrictsync:1; /* do not force expensive SMBflush on every sync */ + bool fsc:1; /* enable fscache */ + bool mfsymlinks:1; /* use Minshall+French Symlinks */ + bool multiuser:1; + bool rwpidforward:1; /* pid forward for read/write operations */ + unsigned int rsize; + unsigned int wsize; + bool sockopt_tcp_nodelay:1; + unsigned short int port; + unsigned long actimeo; /* attribute cache timeout (jiffies) */ + char *prepath; + struct sockaddr_storage srcaddr; /* allow binding to a local IP */ + struct nls_table *local_nls; +}; + +#define CIFS_MOUNT_MASK (CIFS_MOUNT_NO_PERM | CIFS_MOUNT_SET_UID | \ + CIFS_MOUNT_SERVER_INUM | CIFS_MOUNT_DIRECT_IO | \ + CIFS_MOUNT_NO_XATTR | CIFS_MOUNT_MAP_SPECIAL_CHR | \ + CIFS_MOUNT_UNX_EMUL | CIFS_MOUNT_NO_BRL | \ + CIFS_MOUNT_CIFS_ACL | CIFS_MOUNT_OVERR_UID | \ + CIFS_MOUNT_OVERR_GID | CIFS_MOUNT_DYNPERM | \ + CIFS_MOUNT_NOPOSIXBRL | CIFS_MOUNT_NOSSYNC | \ + CIFS_MOUNT_FSCACHE | CIFS_MOUNT_MF_SYMLINKS | \ + CIFS_MOUNT_MULTIUSER | CIFS_MOUNT_STRICT_IO) + +#define CIFS_MS_MASK (MS_RDONLY | MS_MANDLOCK | MS_NOEXEC | MS_NOSUID | \ + MS_NODEV | MS_SYNCHRONOUS) + +struct cifs_mnt_data { + struct cifs_sb_info *cifs_sb; + struct smb_vol *vol; + int flags; +}; + struct TCP_Server_Info { struct list_head tcp_ses_list; struct list_head smb_ses_list; @@ -179,7 +254,7 @@ struct TCP_Server_Info { struct mutex srv_mutex; struct task_struct *tsk; char server_GUID[16]; - char secMode; + char sec_mode; bool session_estab; /* mark when very first sess is established */ u16 dialect; /* dialect index that server chose */ enum securityEnum secType; @@ -254,7 +329,7 @@ static inline void cifs_set_net_ns(struct TCP_Server_Info *srv, struct net *net) /* * Session structure. One of these for each uid session with a particular host */ -struct cifsSesInfo { +struct cifs_ses { struct list_head smb_ses_list; struct list_head tcon_list; struct mutex session_mutex; @@ -274,7 +349,8 @@ struct cifsSesInfo { int capabilities; char serverName[SERVER_NAME_LEN_WITH_NULL * 2]; /* BB make bigger for TCP names - will ipv6 and sctp addresses fit? */ - char userName[MAX_USERNAME_SIZE + 1]; + char *user_name; /* must not be null except during init of sess + and after mount option parsing we fill it */ char *domainName; char *password; struct session_key auth_key; @@ -293,11 +369,11 @@ struct cifsSesInfo { * there is one of these for each connection to a resource on a particular * session */ -struct cifsTconInfo { +struct cifs_tcon { struct list_head tcon_list; int tc_count; struct list_head openFileList; - struct cifsSesInfo *ses; /* pointer to session associated with */ + struct cifs_ses *ses; /* pointer to session associated with */ char treeName[MAX_TREE_SIZE + 1]; /* UNC name of resource in ASCII */ char *nativeFileSystem; char *password; /* for share-level security */ @@ -379,12 +455,12 @@ struct tcon_link { #define TCON_LINK_IN_TREE 2 unsigned long tl_time; atomic_t tl_count; - struct cifsTconInfo *tl_tcon; + struct cifs_tcon *tl_tcon; }; extern struct tcon_link *cifs_sb_tlink(struct cifs_sb_info *cifs_sb); -static inline struct cifsTconInfo * +static inline struct cifs_tcon * tlink_tcon(struct tcon_link *tlink) { return tlink->tl_tcon; @@ -401,7 +477,7 @@ cifs_get_tlink(struct tcon_link *tlink) } /* This function is always expected to succeed */ -extern struct cifsTconInfo *cifs_sb_master_tcon(struct cifs_sb_info *cifs_sb); +extern struct cifs_tcon *cifs_sb_master_tcon(struct cifs_sb_info *cifs_sb); /* * This info hangs off the cifsFileInfo structure, pointed to by llist. @@ -454,6 +530,14 @@ struct cifsFileInfo { struct work_struct oplock_break; /* work for oplock breaks */ }; +struct cifs_io_parms { + __u16 netfid; + __u32 pid; + __u64 offset; + unsigned int length; + struct cifs_tcon *tcon; +}; + /* * Take a reference on the file private data. Must be called with * cifs_file_list_lock held. @@ -508,10 +592,30 @@ static inline char CIFS_DIR_SEP(const struct cifs_sb_info *cifs_sb) return '\\'; } +static inline void +convert_delimiter(char *path, char delim) +{ + int i; + char old_delim; + + if (path == NULL) + return; + + if (delim == '/') + old_delim = '\\'; + else + old_delim = '/'; + + for (i = 0; path[i] != '\0'; i++) { + if (path[i] == old_delim) + path[i] = delim; + } +} + #ifdef CONFIG_CIFS_STATS #define cifs_stats_inc atomic_inc -static inline void cifs_stats_bytes_written(struct cifsTconInfo *tcon, +static inline void cifs_stats_bytes_written(struct cifs_tcon *tcon, unsigned int bytes) { if (bytes) { @@ -521,7 +625,7 @@ static inline void cifs_stats_bytes_written(struct cifsTconInfo *tcon, } } -static inline void cifs_stats_bytes_read(struct cifsTconInfo *tcon, +static inline void cifs_stats_bytes_read(struct cifs_tcon *tcon, unsigned int bytes) { spin_lock(&tcon->stat_lock); @@ -542,9 +646,8 @@ struct mid_q_entry; * This is the prototype for the mid callback function. When creating one, * take special care to avoid deadlocks. Things to bear in mind: * - * - it will be called by cifsd - * - the GlobalMid_Lock will be held - * - the mid will be removed from the pending_mid_q list + * - it will be called by cifsd, with no locks held + * - the mid will be removed from any lists */ typedef void (mid_callback_t)(struct mid_q_entry *mid); @@ -572,7 +675,7 @@ struct mid_q_entry { struct oplock_q_entry { struct list_head qhead; struct inode *pinode; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; __u16 netfid; }; @@ -655,6 +758,7 @@ static inline void free_dfs_info_array(struct dfs_info3_param *param, #define MID_RESPONSE_RECEIVED 4 #define MID_RETRY_NEEDED 8 /* session closed while this request out */ #define MID_RESPONSE_MALFORMED 0x10 +#define MID_SHUTDOWN 0x20 /* Types of response buffer returned from SendReceive2 */ #define CIFS_NO_BUFFER 0 /* Response buffer not returned */ @@ -780,10 +884,12 @@ GLOBAL_EXTERN spinlock_t cifs_tcp_ses_lock; */ GLOBAL_EXTERN spinlock_t cifs_file_list_lock; +#ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* unused temporarily */ /* Outstanding dir notify requests */ GLOBAL_EXTERN struct list_head GlobalDnotifyReqList; /* DirNotify response queue */ GLOBAL_EXTERN struct list_head GlobalDnotifyRsp_Q; +#endif /* was needed for dnotify, and will be needed for inotify when VFS fix */ /* * Global transaction id (XID) information @@ -817,7 +923,6 @@ GLOBAL_EXTERN unsigned int multiuser_mount; /* if enabled allows new sessions have the uid/password or Kerberos credential or equivalent for current user */ GLOBAL_EXTERN unsigned int oplockEnabled; -GLOBAL_EXTERN unsigned int experimEnabled; GLOBAL_EXTERN unsigned int lookupCacheEnabled; GLOBAL_EXTERN unsigned int global_secflags; /* if on, session setup sent with more secure ntlmssp2 challenge/resp */ @@ -831,6 +936,11 @@ GLOBAL_EXTERN unsigned int cifs_max_pending; /* MAX requests at once to server*/ /* reconnect after this many failed echo attempts */ GLOBAL_EXTERN unsigned short echo_retries; +GLOBAL_EXTERN struct rb_root uidtree; +GLOBAL_EXTERN struct rb_root gidtree; +GLOBAL_EXTERN spinlock_t siduidlock; +GLOBAL_EXTERN spinlock_t sidgidlock; + void cifs_oplock_break(struct work_struct *work); void cifs_oplock_break_get(struct cifsFileInfo *cfile); void cifs_oplock_break_put(struct cifsFileInfo *cfile); diff --git a/fs/cifs/cifspdu.h b/fs/cifs/cifspdu.h index b5c8cc5d7a7f..de3aa285de03 100644 --- a/fs/cifs/cifspdu.h +++ b/fs/cifs/cifspdu.h @@ -397,9 +397,9 @@ #define GETU32(var) (*((__u32 *)var)) /* BB check for endian issues */ struct smb_hdr { - __u32 smb_buf_length; /* big endian on wire *//* BB length is only two - or three bytes - with one or two byte type preceding it that are - zero - we could mask the type byte off just in case BB */ + __be32 smb_buf_length; /* BB length is only two (rarely three) bytes, + with one or two byte "type" preceding it that will be + zero - we could mask the type byte off */ __u8 Protocol[4]; __u8 Command; union { @@ -428,43 +428,28 @@ struct smb_hdr { __u8 WordCount; } __attribute__((packed)); -/* given a pointer to an smb_hdr retrieve a char pointer to the byte count */ -#define BCC(smb_var) ((unsigned char *)(smb_var) + sizeof(struct smb_hdr) + \ - (2 * (smb_var)->WordCount)) +/* given a pointer to an smb_hdr, retrieve a void pointer to the ByteCount */ +static inline void * +BCC(struct smb_hdr *smb) +{ + return (void *)smb + sizeof(*smb) + 2 * smb->WordCount; +} /* given a pointer to an smb_hdr retrieve the pointer to the byte area */ #define pByteArea(smb_var) (BCC(smb_var) + 2) -/* get the converted ByteCount for a SMB packet and return it */ -static inline __u16 -get_bcc(struct smb_hdr *hdr) -{ - __u16 *bc_ptr = (__u16 *)BCC(hdr); - - return get_unaligned(bc_ptr); -} - /* get the unconverted ByteCount for a SMB packet and return it */ static inline __u16 -get_bcc_le(struct smb_hdr *hdr) +get_bcc(struct smb_hdr *hdr) { __le16 *bc_ptr = (__le16 *)BCC(hdr); return get_unaligned_le16(bc_ptr); } -/* set the ByteCount for a SMB packet in host-byte order */ -static inline void -put_bcc(__u16 count, struct smb_hdr *hdr) -{ - __u16 *bc_ptr = (__u16 *)BCC(hdr); - - put_unaligned(count, bc_ptr); -} - /* set the ByteCount for a SMB packet in little-endian */ static inline void -put_bcc_le(__u16 count, struct smb_hdr *hdr) +put_bcc(__u16 count, struct smb_hdr *hdr) { __le16 *bc_ptr = (__le16 *)BCC(hdr); diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 8096f27ad9a8..953f84413c77 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -53,9 +53,13 @@ do { \ cFYI(1, "CIFS VFS: leaving %s (xid = %d) rc = %d", \ __func__, curr_xid, (int)rc); \ } while (0) +extern int init_cifs_idmap(void); +extern void exit_cifs_idmap(void); +extern void cifs_destroy_idmaptrees(void); extern char *build_path_from_dentry(struct dentry *); -extern char *cifs_build_path_to_root(struct cifs_sb_info *cifs_sb, - struct cifsTconInfo *tcon); +extern char *cifs_build_path_to_root(struct smb_vol *vol, + struct cifs_sb_info *cifs_sb, + struct cifs_tcon *tcon); extern char *build_wildcard_path_from_dentry(struct dentry *direntry); extern char *cifs_compose_mount_options(const char *sb_mountdata, const char *fullpath, const struct dfs_info3_param *ref, @@ -64,20 +68,22 @@ extern char *cifs_compose_mount_options(const char *sb_mountdata, extern struct mid_q_entry *AllocMidQEntry(const struct smb_hdr *smb_buffer, struct TCP_Server_Info *server); extern void DeleteMidQEntry(struct mid_q_entry *midEntry); -extern int cifs_call_async(struct TCP_Server_Info *server, - struct smb_hdr *in_buf, mid_callback_t *callback, - void *cbdata); -extern int SendReceive(const unsigned int /* xid */ , struct cifsSesInfo *, +extern int cifs_call_async(struct TCP_Server_Info *server, struct kvec *iov, + unsigned int nvec, mid_callback_t *callback, + void *cbdata, bool ignore_pend); +extern int SendReceive(const unsigned int /* xid */ , struct cifs_ses *, struct smb_hdr * /* input */ , struct smb_hdr * /* out */ , int * /* bytes returned */ , const int long_op); -extern int SendReceiveNoRsp(const unsigned int xid, struct cifsSesInfo *ses, +extern int SendReceiveNoRsp(const unsigned int xid, struct cifs_ses *ses, struct smb_hdr *in_buf, int flags); -extern int SendReceive2(const unsigned int /* xid */ , struct cifsSesInfo *, +extern int cifs_check_receive(struct mid_q_entry *mid, + struct TCP_Server_Info *server, bool log_error); +extern int SendReceive2(const unsigned int /* xid */ , struct cifs_ses *, struct kvec *, int /* nvec to send */, int * /* type of buf returned */ , const int flags); extern int SendReceiveBlockingLock(const unsigned int xid, - struct cifsTconInfo *ptcon, + struct cifs_tcon *ptcon, struct smb_hdr *in_buf , struct smb_hdr *out_buf, int *bytes_returned); @@ -90,21 +96,20 @@ extern void cifs_update_eof(struct cifsInodeInfo *cifsi, loff_t offset, extern struct cifsFileInfo *find_writable_file(struct cifsInodeInfo *, bool); extern struct cifsFileInfo *find_readable_file(struct cifsInodeInfo *, bool); extern unsigned int smbCalcSize(struct smb_hdr *ptr); -extern unsigned int smbCalcSize_LE(struct smb_hdr *ptr); extern int decode_negTokenInit(unsigned char *security_blob, int length, struct TCP_Server_Info *server); extern int cifs_convert_address(struct sockaddr *dst, const char *src, int len); extern int cifs_set_port(struct sockaddr *addr, const unsigned short int port); extern int cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, const unsigned short int port); -extern int map_smb_to_linux_error(struct smb_hdr *smb, int logErr); +extern int map_smb_to_linux_error(struct smb_hdr *smb, bool logErr); extern void header_assemble(struct smb_hdr *, char /* command */ , - const struct cifsTconInfo *, int /* length of + const struct cifs_tcon *, int /* length of fixed section (word count) in two byte units */); extern int small_smb_init_no_tc(const int smb_cmd, const int wct, - struct cifsSesInfo *ses, + struct cifs_ses *ses, void **request_buf); -extern int CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, +extern int CIFS_SessSetup(unsigned int xid, struct cifs_ses *ses, const struct nls_table *nls_cp); extern __u16 GetNextMid(struct TCP_Server_Info *server); extern struct timespec cifs_NTtimeToUnix(__le64 utc_nanoseconds_since_1601); @@ -143,103 +148,111 @@ extern int cifs_acl_to_fattr(struct cifs_sb_info *cifs_sb, extern int mode_to_cifs_acl(struct inode *inode, const char *path, __u64); extern struct cifs_ntsd *get_cifs_acl(struct cifs_sb_info *, struct inode *, const char *, u32 *); +extern int set_cifs_acl(struct cifs_ntsd *, __u32, struct inode *, + const char *); -extern int cifs_mount(struct super_block *, struct cifs_sb_info *, char *, - const char *); +extern void cifs_setup_cifs_sb(struct smb_vol *pvolume_info, + struct cifs_sb_info *cifs_sb); +extern int cifs_match_super(struct super_block *, void *); +extern void cifs_cleanup_volume_info(struct smb_vol **pvolume_info); +extern int cifs_setup_volume_info(struct smb_vol **pvolume_info, + char *mount_data, const char *devname); +extern int cifs_mount(struct super_block *, struct cifs_sb_info *, + struct smb_vol *, const char *); extern int cifs_umount(struct super_block *, struct cifs_sb_info *); extern void cifs_dfs_release_automount_timer(void); void cifs_proc_init(void); void cifs_proc_clean(void); extern int cifs_negotiate_protocol(unsigned int xid, - struct cifsSesInfo *ses); -extern int cifs_setup_session(unsigned int xid, struct cifsSesInfo *ses, + struct cifs_ses *ses); +extern int cifs_setup_session(unsigned int xid, struct cifs_ses *ses, struct nls_table *nls_info); -extern int CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses); +extern int CIFSSMBNegotiate(unsigned int xid, struct cifs_ses *ses); -extern int CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, - const char *tree, struct cifsTconInfo *tcon, +extern int CIFSTCon(unsigned int xid, struct cifs_ses *ses, + const char *tree, struct cifs_tcon *tcon, const struct nls_table *); -extern int CIFSFindFirst(const int xid, struct cifsTconInfo *tcon, +extern int CIFSFindFirst(const int xid, struct cifs_tcon *tcon, const char *searchName, const struct nls_table *nls_codepage, __u16 *searchHandle, struct cifs_search_info *psrch_inf, int map, const char dirsep); -extern int CIFSFindNext(const int xid, struct cifsTconInfo *tcon, +extern int CIFSFindNext(const int xid, struct cifs_tcon *tcon, __u16 searchHandle, struct cifs_search_info *psrch_inf); -extern int CIFSFindClose(const int, struct cifsTconInfo *tcon, +extern int CIFSFindClose(const int, struct cifs_tcon *tcon, const __u16 search_handle); -extern int CIFSSMBQFileInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBQFileInfo(const int xid, struct cifs_tcon *tcon, u16 netfid, FILE_ALL_INFO *pFindData); -extern int CIFSSMBQPathInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBQPathInfo(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, FILE_ALL_INFO *findData, int legacy /* whether to use old info level */, const struct nls_table *nls_codepage, int remap); -extern int SMBQueryInformation(const int xid, struct cifsTconInfo *tcon, +extern int SMBQueryInformation(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, FILE_ALL_INFO *findData, const struct nls_table *nls_codepage, int remap); -extern int CIFSSMBUnixQFileInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBUnixQFileInfo(const int xid, struct cifs_tcon *tcon, u16 netfid, FILE_UNIX_BASIC_INFO *pFindData); extern int CIFSSMBUnixQPathInfo(const int xid, - struct cifsTconInfo *tcon, + struct cifs_tcon *tcon, const unsigned char *searchName, FILE_UNIX_BASIC_INFO *pFindData, const struct nls_table *nls_codepage, int remap); -extern int CIFSGetDFSRefer(const int xid, struct cifsSesInfo *ses, +extern int CIFSGetDFSRefer(const int xid, struct cifs_ses *ses, const unsigned char *searchName, struct dfs_info3_param **target_nodes, unsigned int *number_of_nodes_in_array, const struct nls_table *nls_codepage, int remap); -extern int get_dfs_path(int xid, struct cifsSesInfo *pSesInfo, +extern int get_dfs_path(int xid, struct cifs_ses *pSesInfo, const char *old_path, const struct nls_table *nls_codepage, unsigned int *pnum_referrals, struct dfs_info3_param **preferrals, int remap); -extern void reset_cifs_unix_caps(int xid, struct cifsTconInfo *tcon, +extern void reset_cifs_unix_caps(int xid, struct cifs_tcon *tcon, struct super_block *sb, struct smb_vol *vol); -extern int CIFSSMBQFSInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBQFSInfo(const int xid, struct cifs_tcon *tcon, struct kstatfs *FSData); -extern int SMBOldQFSInfo(const int xid, struct cifsTconInfo *tcon, +extern int SMBOldQFSInfo(const int xid, struct cifs_tcon *tcon, struct kstatfs *FSData); -extern int CIFSSMBSetFSUnixInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetFSUnixInfo(const int xid, struct cifs_tcon *tcon, __u64 cap); extern int CIFSSMBQFSAttributeInfo(const int xid, - struct cifsTconInfo *tcon); -extern int CIFSSMBQFSDeviceInfo(const int xid, struct cifsTconInfo *tcon); -extern int CIFSSMBQFSUnixInfo(const int xid, struct cifsTconInfo *tcon); -extern int CIFSSMBQFSPosixInfo(const int xid, struct cifsTconInfo *tcon, + struct cifs_tcon *tcon); +extern int CIFSSMBQFSDeviceInfo(const int xid, struct cifs_tcon *tcon); +extern int CIFSSMBQFSUnixInfo(const int xid, struct cifs_tcon *tcon); +extern int CIFSSMBQFSPosixInfo(const int xid, struct cifs_tcon *tcon, struct kstatfs *FSData); -extern int CIFSSMBSetPathInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetPathInfo(const int xid, struct cifs_tcon *tcon, const char *fileName, const FILE_BASIC_INFO *data, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBSetFileInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetFileInfo(const int xid, struct cifs_tcon *tcon, const FILE_BASIC_INFO *data, __u16 fid, __u32 pid_of_opener); -extern int CIFSSMBSetFileDisposition(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetFileDisposition(const int xid, struct cifs_tcon *tcon, bool delete_file, __u16 fid, __u32 pid_of_opener); #if 0 -extern int CIFSSMBSetAttrLegacy(int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetAttrLegacy(int xid, struct cifs_tcon *tcon, char *fileName, __u16 dos_attributes, const struct nls_table *nls_codepage); #endif /* possibly unneeded function */ -extern int CIFSSMBSetEOF(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetEOF(const int xid, struct cifs_tcon *tcon, const char *fileName, __u64 size, bool setAllocationSizeFlag, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBSetFileSize(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetFileSize(const int xid, struct cifs_tcon *tcon, __u64 size, __u16 fileHandle, __u32 opener_pid, bool AllocSizeFlag); @@ -253,121 +266,116 @@ struct cifs_unix_set_info_args { dev_t device; }; -extern int CIFSSMBUnixSetFileInfo(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBUnixSetFileInfo(const int xid, struct cifs_tcon *tcon, const struct cifs_unix_set_info_args *args, u16 fid, u32 pid_of_opener); -extern int CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *pTcon, +extern int CIFSSMBUnixSetPathInfo(const int xid, struct cifs_tcon *pTcon, char *fileName, const struct cifs_unix_set_info_args *args, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBMkDir(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBMkDir(const int xid, struct cifs_tcon *tcon, const char *newName, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBRmDir(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBRmDir(const int xid, struct cifs_tcon *tcon, const char *name, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSPOSIXDelFile(const int xid, struct cifsTconInfo *tcon, +extern int CIFSPOSIXDelFile(const int xid, struct cifs_tcon *tcon, const char *name, __u16 type, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBDelFile(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBDelFile(const int xid, struct cifs_tcon *tcon, const char *name, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBRename(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBRename(const int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBRenameOpenFile(const int xid, struct cifsTconInfo *pTcon, +extern int CIFSSMBRenameOpenFile(const int xid, struct cifs_tcon *pTcon, int netfid, const char *target_name, const struct nls_table *nls_codepage, int remap_special_chars); extern int CIFSCreateHardLink(const int xid, - struct cifsTconInfo *tcon, + struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage, int remap_special_chars); extern int CIFSUnixCreateHardLink(const int xid, - struct cifsTconInfo *tcon, + struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage, int remap_special_chars); extern int CIFSUnixCreateSymLink(const int xid, - struct cifsTconInfo *tcon, + struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage); extern int CIFSSMBUnixQuerySymLink(const int xid, - struct cifsTconInfo *tcon, + struct cifs_tcon *tcon, const unsigned char *searchName, char **syminfo, const struct nls_table *nls_codepage); +#ifdef CONFIG_CIFS_SYMLINK_EXPERIMENTAL extern int CIFSSMBQueryReparseLinkInfo(const int xid, - struct cifsTconInfo *tcon, + struct cifs_tcon *tcon, const unsigned char *searchName, char *symlinkinfo, const int buflen, __u16 fid, const struct nls_table *nls_codepage); - -extern int CIFSSMBOpen(const int xid, struct cifsTconInfo *tcon, +#endif /* temporarily unused until cifs_symlink fixed */ +extern int CIFSSMBOpen(const int xid, struct cifs_tcon *tcon, const char *fileName, const int disposition, const int access_flags, const int omode, __u16 *netfid, int *pOplock, FILE_ALL_INFO *, const struct nls_table *nls_codepage, int remap); -extern int SMBLegacyOpen(const int xid, struct cifsTconInfo *tcon, +extern int SMBLegacyOpen(const int xid, struct cifs_tcon *tcon, const char *fileName, const int disposition, const int access_flags, const int omode, __u16 *netfid, int *pOplock, FILE_ALL_INFO *, const struct nls_table *nls_codepage, int remap); -extern int CIFSPOSIXCreate(const int xid, struct cifsTconInfo *tcon, +extern int CIFSPOSIXCreate(const int xid, struct cifs_tcon *tcon, u32 posix_flags, __u64 mode, __u16 *netfid, FILE_UNIX_BASIC_INFO *pRetData, __u32 *pOplock, const char *name, const struct nls_table *nls_codepage, int remap); -extern int CIFSSMBClose(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBClose(const int xid, struct cifs_tcon *tcon, const int smb_file_id); -extern int CIFSSMBFlush(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBFlush(const int xid, struct cifs_tcon *tcon, const int smb_file_id); -extern int CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, - const int netfid, unsigned int count, - const __u64 lseek, unsigned int *nbytes, char **buf, +extern int CIFSSMBRead(const int xid, struct cifs_io_parms *io_parms, + unsigned int *nbytes, char **buf, int *return_buf_type); -extern int CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, - const int netfid, const unsigned int count, - const __u64 lseek, unsigned int *nbytes, - const char *buf, const char __user *ubuf, +extern int CIFSSMBWrite(const int xid, struct cifs_io_parms *io_parms, + unsigned int *nbytes, const char *buf, + const char __user *ubuf, const int long_op); +extern int CIFSSMBWrite2(const int xid, struct cifs_io_parms *io_parms, + unsigned int *nbytes, struct kvec *iov, const int nvec, const int long_op); -extern int CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, - const int netfid, const unsigned int count, - const __u64 offset, unsigned int *nbytes, - struct kvec *iov, const int nvec, const int long_op); -extern int CIFSGetSrvInodeNumber(const int xid, struct cifsTconInfo *tcon, +extern int CIFSGetSrvInodeNumber(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, __u64 *inode_number, const struct nls_table *nls_codepage, int remap_special_chars); -extern int cifsConvertToUCS(__le16 *target, const char *source, int maxlen, - const struct nls_table *cp, int mapChars); -extern int CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBLock(const int xid, struct cifs_tcon *tcon, const __u16 netfid, const __u64 len, const __u64 offset, const __u32 numUnlock, const __u32 numLock, const __u8 lockType, const bool waitFlag, const __u8 oplock_level); -extern int CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBPosixLock(const int xid, struct cifs_tcon *tcon, const __u16 smb_file_id, const int get_flag, const __u64 len, struct file_lock *, const __u16 lock_type, const bool waitFlag); -extern int CIFSSMBTDis(const int xid, struct cifsTconInfo *tcon); +extern int CIFSSMBTDis(const int xid, struct cifs_tcon *tcon); extern int CIFSSMBEcho(struct TCP_Server_Info *server); -extern int CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses); +extern int CIFSSMBLogoff(const int xid, struct cifs_ses *ses); -extern struct cifsSesInfo *sesInfoAlloc(void); -extern void sesInfoFree(struct cifsSesInfo *); -extern struct cifsTconInfo *tconInfoAlloc(void); -extern void tconInfoFree(struct cifsTconInfo *); +extern struct cifs_ses *sesInfoAlloc(void); +extern void sesInfoFree(struct cifs_ses *); +extern struct cifs_tcon *tconInfoAlloc(void); +extern void tconInfoFree(struct cifs_tcon *); extern int cifs_sign_smb(struct smb_hdr *, struct TCP_Server_Info *, __u32 *); extern int cifs_sign_smb2(struct kvec *iov, int n_vec, struct TCP_Server_Info *, @@ -376,49 +384,51 @@ extern int cifs_verify_signature(struct smb_hdr *, struct TCP_Server_Info *server, __u32 expected_sequence_number); extern int SMBNTencrypt(unsigned char *, unsigned char *, unsigned char *); -extern int setup_ntlm_response(struct cifsSesInfo *); -extern int setup_ntlmv2_rsp(struct cifsSesInfo *, const struct nls_table *); +extern int setup_ntlm_response(struct cifs_ses *); +extern int setup_ntlmv2_rsp(struct cifs_ses *, const struct nls_table *); extern int cifs_crypto_shash_allocate(struct TCP_Server_Info *); extern void cifs_crypto_shash_release(struct TCP_Server_Info *); -extern int calc_seckey(struct cifsSesInfo *); +extern int calc_seckey(struct cifs_ses *); #ifdef CONFIG_CIFS_WEAK_PW_HASH -extern void calc_lanman_hash(const char *password, const char *cryptkey, +extern int calc_lanman_hash(const char *password, const char *cryptkey, bool encrypt, char *lnm_session_key); #endif /* CIFS_WEAK_PW_HASH */ +#ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* unused temporarily */ +extern int CIFSSMBNotify(const int xid, struct cifs_tcon *tcon, + const int notify_subdirs, const __u16 netfid, + __u32 filter, struct file *file, int multishot, + const struct nls_table *nls_codepage); +#endif /* was needed for dnotify, and will be needed for inotify when VFS fix */ extern int CIFSSMBCopy(int xid, - struct cifsTconInfo *source_tcon, + struct cifs_tcon *source_tcon, const char *fromName, const __u16 target_tid, const char *toName, const int flags, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBNotify(const int xid, struct cifsTconInfo *tcon, - const int notify_subdirs, const __u16 netfid, - __u32 filter, struct file *file, int multishot, - const struct nls_table *nls_codepage); -extern ssize_t CIFSSMBQAllEAs(const int xid, struct cifsTconInfo *tcon, +extern ssize_t CIFSSMBQAllEAs(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, const unsigned char *ea_name, char *EAData, size_t bufsize, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBSetEA(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetEA(const int xid, struct cifs_tcon *tcon, const char *fileName, const char *ea_name, const void *ea_value, const __u16 ea_value_len, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBGetCIFSACL(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBGetCIFSACL(const int xid, struct cifs_tcon *tcon, __u16 fid, struct cifs_ntsd **acl_inf, __u32 *buflen); -extern int CIFSSMBSetCIFSACL(const int, struct cifsTconInfo *, __u16, +extern int CIFSSMBSetCIFSACL(const int, struct cifs_tcon *, __u16, struct cifs_ntsd *, __u32); -extern int CIFSSMBGetPosixACL(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBGetPosixACL(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, char *acl_inf, const int buflen, const int acl_type, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSSMBSetPosixACL(const int xid, struct cifsTconInfo *tcon, +extern int CIFSSMBSetPosixACL(const int xid, struct cifs_tcon *tcon, const unsigned char *fileName, const char *local_acl, const int buflen, const int acl_type, const struct nls_table *nls_codepage, int remap_special_chars); -extern int CIFSGetExtAttr(const int xid, struct cifsTconInfo *tcon, +extern int CIFSGetExtAttr(const int xid, struct cifs_tcon *tcon, const int netfid, __u64 *pExtAttrBits, __u64 *pMask); extern void cifs_autodisable_serverino(struct cifs_sb_info *cifs_sb); extern bool CIFSCouldBeMFSymlink(const struct cifs_fattr *fattr); @@ -427,9 +437,24 @@ extern int CIFSCheckMFSymlink(struct cifs_fattr *fattr, struct cifs_sb_info *cifs_sb, int xid); extern int mdfour(unsigned char *, unsigned char *, int); extern int E_md4hash(const unsigned char *passwd, unsigned char *p16); -extern void SMBencrypt(unsigned char *passwd, const unsigned char *c8, - unsigned char *p24); -extern void E_P16(unsigned char *p14, unsigned char *p16); -extern void E_P24(unsigned char *p21, const unsigned char *c8, +extern int SMBencrypt(unsigned char *passwd, const unsigned char *c8, unsigned char *p24); + +/* asynchronous write support */ +struct cifs_writedata { + struct kref refcount; + enum writeback_sync_modes sync_mode; + struct work_struct work; + struct cifsFileInfo *cfile; + __u64 offset; + unsigned int bytes; + int result; + unsigned int nr_pages; + struct page *pages[1]; +}; + +int cifs_async_writev(struct cifs_writedata *wdata); +struct cifs_writedata *cifs_writedata_alloc(unsigned int nr_pages); +void cifs_writedata_release(struct kref *refcount); + #endif /* _CIFSPROTO_H */ diff --git a/fs/cifs/cifssmb.c b/fs/cifs/cifssmb.c index 904aa47e3515..1a9fe7f816d1 100644 --- a/fs/cifs/cifssmb.c +++ b/fs/cifs/cifssmb.c @@ -32,6 +32,7 @@ #include <linux/vfs.h> #include <linux/slab.h> #include <linux/posix_acl_xattr.h> +#include <linux/pagemap.h> #include <asm/uaccess.h> #include "cifspdu.h" #include "cifsglob.h" @@ -84,7 +85,7 @@ static struct { /* Mark as invalid, all open files on tree connections since they were closed when session to server was lost */ -static void mark_open_files_invalid(struct cifsTconInfo *pTcon) +static void mark_open_files_invalid(struct cifs_tcon *pTcon) { struct cifsFileInfo *open_file = NULL; struct list_head *tmp; @@ -104,10 +105,10 @@ static void mark_open_files_invalid(struct cifsTconInfo *pTcon) /* reconnect the socket, tcon, and smb session if needed */ static int -cifs_reconnect_tcon(struct cifsTconInfo *tcon, int smb_command) +cifs_reconnect_tcon(struct cifs_tcon *tcon, int smb_command) { int rc = 0; - struct cifsSesInfo *ses; + struct cifs_ses *ses; struct TCP_Server_Info *server; struct nls_table *nls_codepage; @@ -142,9 +143,9 @@ cifs_reconnect_tcon(struct cifsTconInfo *tcon, int smb_command) */ while (server->tcpStatus == CifsNeedReconnect) { wait_event_interruptible_timeout(server->response_q, - (server->tcpStatus == CifsGood), 10 * HZ); + (server->tcpStatus != CifsNeedReconnect), 10 * HZ); - /* is TCP session is reestablished now ?*/ + /* are we still trying to reconnect? */ if (server->tcpStatus != CifsNeedReconnect) break; @@ -226,7 +227,7 @@ out: SMB information in the SMB header. If the return code is zero, this function must have filled in request_buf pointer */ static int -small_smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, +small_smb_init(int smb_command, int wct, struct cifs_tcon *tcon, void **request_buf) { int rc; @@ -252,7 +253,7 @@ small_smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, int small_smb_init_no_tc(const int smb_command, const int wct, - struct cifsSesInfo *ses, void **request_buf) + struct cifs_ses *ses, void **request_buf) { int rc; struct smb_hdr *buffer; @@ -278,7 +279,7 @@ small_smb_init_no_tc(const int smb_command, const int wct, /* If the return code is zero, this function must fill in request_buf pointer */ static int -__smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, +__smb_init(int smb_command, int wct, struct cifs_tcon *tcon, void **request_buf, void **response_buf) { *request_buf = cifs_buf_get(); @@ -304,7 +305,7 @@ __smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, /* If the return code is zero, this function must fill in request_buf pointer */ static int -smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, +smb_init(int smb_command, int wct, struct cifs_tcon *tcon, void **request_buf, void **response_buf) { int rc; @@ -317,7 +318,7 @@ smb_init(int smb_command, int wct, struct cifsTconInfo *tcon, } static int -smb_init_no_reconnect(int smb_command, int wct, struct cifsTconInfo *tcon, +smb_init_no_reconnect(int smb_command, int wct, struct cifs_tcon *tcon, void **request_buf, void **response_buf) { if (tcon->ses->need_reconnect || tcon->need_reconnect) @@ -339,12 +340,13 @@ static int validate_t2(struct smb_t2_rsp *pSMB) get_unaligned_le16(&pSMB->t2_rsp.DataOffset) > 1024) goto vt2_err; - /* check that bcc is at least as big as parms + data */ - /* check that bcc is less than negotiated smb buffer */ total_size = get_unaligned_le16(&pSMB->t2_rsp.ParameterCount); if (total_size >= 512) goto vt2_err; + /* check that bcc is at least as big as parms + data, and that it is + * less than negotiated smb buffer + */ total_size += get_unaligned_le16(&pSMB->t2_rsp.DataCount); if (total_size > get_bcc(&pSMB->hdr) || total_size >= CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) @@ -357,8 +359,15 @@ vt2_err: return -EINVAL; } +static inline void inc_rfc1001_len(void *pSMB, int count) +{ + struct smb_hdr *hdr = (struct smb_hdr *)pSMB; + + be32_add_cpu(&hdr->smb_buf_length, count); +} + int -CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) +CIFSSMBNegotiate(unsigned int xid, struct cifs_ses *ses) { NEGOTIATE_REQ *pSMB; NEGOTIATE_RSP *pSMBr; @@ -409,7 +418,7 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) count += strlen(protocols[i].name) + 1; /* null at end of source and target buffers anyway */ } - pSMB->hdr.smb_buf_length += count; + inc_rfc1001_len(pSMB, count); pSMB->ByteCount = cpu_to_le16(count); rc = SendReceive(xid, ses, (struct smb_hdr *) pSMB, @@ -442,7 +451,7 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) rc = -EOPNOTSUPP; goto neg_err_exit; } - server->secMode = (__u8)le16_to_cpu(rsp->SecurityMode); + server->sec_mode = (__u8)le16_to_cpu(rsp->SecurityMode); server->maxReq = le16_to_cpu(rsp->MaxMpxCount); server->maxBuf = min((__u32)le16_to_cpu(rsp->MaxBufSize), (__u32)CIFSMaxBufSize + MAX_CIFS_HDR_SIZE); @@ -496,7 +505,7 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) cpu_to_le16(CIFS_CRYPTO_KEY_SIZE)) { memcpy(ses->server->cryptkey, rsp->EncryptionKey, CIFS_CRYPTO_KEY_SIZE); - } else if (server->secMode & SECMODE_PW_ENCRYPT) { + } else if (server->sec_mode & SECMODE_PW_ENCRYPT) { rc = -EIO; /* need cryptkey unless plain text */ goto neg_err_exit; } @@ -518,11 +527,11 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) goto neg_err_exit; } /* else wct == 17 NTLM */ - server->secMode = pSMBr->SecurityMode; - if ((server->secMode & SECMODE_USER) == 0) + server->sec_mode = pSMBr->SecurityMode; + if ((server->sec_mode & SECMODE_USER) == 0) cFYI(1, "share mode security"); - if ((server->secMode & SECMODE_PW_ENCRYPT) == 0) + if ((server->sec_mode & SECMODE_PW_ENCRYPT) == 0) #ifdef CONFIG_CIFS_WEAK_PW_HASH if ((secFlags & CIFSSEC_MAY_PLNTXT) == 0) #endif /* CIFS_WEAK_PW_HASH */ @@ -541,10 +550,6 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) server->secType = RawNTLMSSP; else if (secFlags & CIFSSEC_MAY_LANMAN) server->secType = LANMAN; -/* #ifdef CONFIG_CIFS_EXPERIMENTAL - else if (secFlags & CIFSSEC_MAY_PLNTXT) - server->secType = ?? -#endif */ else { rc = -EOPNOTSUPP; cERROR(1, "Invalid security type"); @@ -566,19 +571,11 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) if (pSMBr->EncryptionKeyLength == CIFS_CRYPTO_KEY_SIZE) { memcpy(ses->server->cryptkey, pSMBr->u.EncryptionKey, CIFS_CRYPTO_KEY_SIZE); - } else if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC) - && (pSMBr->EncryptionKeyLength == 0)) { + } else if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC || + server->capabilities & CAP_EXTENDED_SECURITY) && + (pSMBr->EncryptionKeyLength == 0)) { /* decode security blob */ - } else if (server->secMode & SECMODE_PW_ENCRYPT) { - rc = -EIO; /* no crypt key only if plain text pwd */ - goto neg_err_exit; - } - - /* BB might be helpful to save off the domain of server here */ - - if ((pSMBr->hdr.Flags2 & SMBFLG2_EXT_SEC) && - (server->capabilities & CAP_EXTENDED_SECURITY)) { - count = pSMBr->ByteCount; + count = get_bcc(&pSMBr->hdr); if (count < 16) { rc = -EIO; goto neg_err_exit; @@ -620,6 +617,9 @@ CIFSSMBNegotiate(unsigned int xid, struct cifsSesInfo *ses) } else rc = -EOPNOTSUPP; } + } else if (server->sec_mode & SECMODE_PW_ENCRYPT) { + rc = -EIO; /* no crypt key only if plain text pwd */ + goto neg_err_exit; } else server->capabilities &= ~CAP_EXTENDED_SECURITY; @@ -630,27 +630,27 @@ signing_check: /* MUST_SIGN already includes the MAY_SIGN FLAG so if this is zero it means that signing is disabled */ cFYI(1, "Signing disabled"); - if (server->secMode & SECMODE_SIGN_REQUIRED) { + if (server->sec_mode & SECMODE_SIGN_REQUIRED) { cERROR(1, "Server requires " "packet signing to be enabled in " "/proc/fs/cifs/SecurityFlags."); rc = -EOPNOTSUPP; } - server->secMode &= + server->sec_mode &= ~(SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED); } else if ((secFlags & CIFSSEC_MUST_SIGN) == CIFSSEC_MUST_SIGN) { /* signing required */ cFYI(1, "Must sign - secFlags 0x%x", secFlags); - if ((server->secMode & + if ((server->sec_mode & (SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED)) == 0) { cERROR(1, "signing required but server lacks support"); rc = -EOPNOTSUPP; } else - server->secMode |= SECMODE_SIGN_REQUIRED; + server->sec_mode |= SECMODE_SIGN_REQUIRED; } else { /* signing optional ie CIFSSEC_MAY_SIGN */ - if ((server->secMode & SECMODE_SIGN_REQUIRED) == 0) - server->secMode &= + if ((server->sec_mode & SECMODE_SIGN_REQUIRED) == 0) + server->sec_mode &= ~(SECMODE_SIGN_ENABLED | SECMODE_SIGN_REQUIRED); } @@ -662,7 +662,7 @@ neg_err_exit: } int -CIFSSMBTDis(const int xid, struct cifsTconInfo *tcon) +CIFSSMBTDis(const int xid, struct cifs_tcon *tcon) { struct smb_hdr *smb_buffer; int rc = 0; @@ -721,6 +721,7 @@ CIFSSMBEcho(struct TCP_Server_Info *server) { ECHO_REQ *smb; int rc = 0; + struct kvec iov; cFYI(1, "In echo request"); @@ -729,15 +730,16 @@ CIFSSMBEcho(struct TCP_Server_Info *server) return rc; /* set up echo request */ - smb->hdr.Tid = cpu_to_le16(0xffff); + smb->hdr.Tid = 0xffff; smb->hdr.WordCount = 1; put_unaligned_le16(1, &smb->EchoCount); - put_bcc_le(1, &smb->hdr); + put_bcc(1, &smb->hdr); smb->Data[0] = 'a'; - smb->hdr.smb_buf_length += 3; + inc_rfc1001_len(smb, 3); + iov.iov_base = smb; + iov.iov_len = be32_to_cpu(smb->hdr.smb_buf_length) + 4; - rc = cifs_call_async(server, (struct smb_hdr *)smb, - cifs_echo_callback, server); + rc = cifs_call_async(server, &iov, 1, cifs_echo_callback, server, true); if (rc) cFYI(1, "Echo request failed: %d", rc); @@ -747,7 +749,7 @@ CIFSSMBEcho(struct TCP_Server_Info *server) } int -CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses) +CIFSSMBLogoff(const int xid, struct cifs_ses *ses) { LOGOFF_ANDX_REQ *pSMB; int rc = 0; @@ -774,7 +776,7 @@ CIFSSMBLogoff(const int xid, struct cifsSesInfo *ses) pSMB->hdr.Mid = GetNextMid(ses->server); - if (ses->server->secMode & + if (ses->server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) pSMB->hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; @@ -794,7 +796,7 @@ session_already_dead: } int -CIFSPOSIXDelFile(const int xid, struct cifsTconInfo *tcon, const char *fileName, +CIFSPOSIXDelFile(const int xid, struct cifs_tcon *tcon, const char *fileName, __u16 type, const struct nls_table *nls_codepage, int remap) { TRANSACTION2_SPI_REQ *pSMB = NULL; @@ -852,7 +854,7 @@ PsxDelete: pSMB->TotalParameterCount = pSMB->ParameterCount; pSMB->InformationLevel = cpu_to_le16(SMB_POSIX_UNLINK); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -869,7 +871,7 @@ PsxDelete: } int -CIFSSMBDelFile(const int xid, struct cifsTconInfo *tcon, const char *fileName, +CIFSSMBDelFile(const int xid, struct cifs_tcon *tcon, const char *fileName, const struct nls_table *nls_codepage, int remap) { DELETE_FILE_REQ *pSMB = NULL; @@ -898,7 +900,7 @@ DelFileRetry: pSMB->SearchAttributes = cpu_to_le16(ATTR_READONLY | ATTR_HIDDEN | ATTR_SYSTEM); pSMB->BufferFormat = 0x04; - pSMB->hdr.smb_buf_length += name_len + 1; + inc_rfc1001_len(pSMB, name_len + 1); pSMB->ByteCount = cpu_to_le16(name_len + 1); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -914,7 +916,7 @@ DelFileRetry: } int -CIFSSMBRmDir(const int xid, struct cifsTconInfo *tcon, const char *dirName, +CIFSSMBRmDir(const int xid, struct cifs_tcon *tcon, const char *dirName, const struct nls_table *nls_codepage, int remap) { DELETE_DIRECTORY_REQ *pSMB = NULL; @@ -942,7 +944,7 @@ RmDirRetry: } pSMB->BufferFormat = 0x04; - pSMB->hdr.smb_buf_length += name_len + 1; + inc_rfc1001_len(pSMB, name_len + 1); pSMB->ByteCount = cpu_to_le16(name_len + 1); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -957,7 +959,7 @@ RmDirRetry: } int -CIFSSMBMkDir(const int xid, struct cifsTconInfo *tcon, +CIFSSMBMkDir(const int xid, struct cifs_tcon *tcon, const char *name, const struct nls_table *nls_codepage, int remap) { int rc = 0; @@ -985,7 +987,7 @@ MkDirRetry: } pSMB->BufferFormat = 0x04; - pSMB->hdr.smb_buf_length += name_len + 1; + inc_rfc1001_len(pSMB, name_len + 1); pSMB->ByteCount = cpu_to_le16(name_len + 1); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -1000,7 +1002,7 @@ MkDirRetry: } int -CIFSPOSIXCreate(const int xid, struct cifsTconInfo *tcon, __u32 posix_flags, +CIFSPOSIXCreate(const int xid, struct cifs_tcon *tcon, __u32 posix_flags, __u64 mode, __u16 *netfid, FILE_UNIX_BASIC_INFO *pRetData, __u32 *pOplock, const char *name, const struct nls_table *nls_codepage, int remap) @@ -1063,7 +1065,7 @@ PsxCreat: pSMB->TotalParameterCount = pSMB->ParameterCount; pSMB->InformationLevel = cpu_to_le16(SMB_POSIX_OPEN); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -1075,7 +1077,7 @@ PsxCreat: cFYI(1, "copying inode info"); rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < sizeof(OPEN_PSX_RSP))) { + if (rc || get_bcc(&pSMBr->hdr) < sizeof(OPEN_PSX_RSP)) { rc = -EIO; /* bad smb */ goto psx_create_err; } @@ -1096,7 +1098,7 @@ PsxCreat: pRetData->Type = cpu_to_le32(-1); /* unknown */ cFYI(DBG2, "unknown type"); } else { - if (pSMBr->ByteCount < sizeof(OPEN_PSX_RSP) + if (get_bcc(&pSMBr->hdr) < sizeof(OPEN_PSX_RSP) + sizeof(FILE_UNIX_BASIC_INFO)) { cERROR(1, "Open response data too small"); pRetData->Type = cpu_to_le32(-1); @@ -1166,7 +1168,7 @@ access_flags_to_smbopen_mode(const int access_flags) } int -SMBLegacyOpen(const int xid, struct cifsTconInfo *tcon, +SMBLegacyOpen(const int xid, struct cifs_tcon *tcon, const char *fileName, const int openDisposition, const int access_flags, const int create_options, __u16 *netfid, int *pOplock, FILE_ALL_INFO *pfile_info, @@ -1228,7 +1230,7 @@ OldOpenRetry: pSMB->Sattr = cpu_to_le16(ATTR_HIDDEN | ATTR_SYSTEM | ATTR_DIRECTORY); pSMB->OpenFunction = cpu_to_le16(convert_disposition(openDisposition)); count += name_len; - pSMB->hdr.smb_buf_length += count; + inc_rfc1001_len(pSMB, count); pSMB->ByteCount = cpu_to_le16(count); /* long_op set to 1 to allow for oplock break timeouts */ @@ -1273,7 +1275,7 @@ OldOpenRetry: } int -CIFSSMBOpen(const int xid, struct cifsTconInfo *tcon, +CIFSSMBOpen(const int xid, struct cifs_tcon *tcon, const char *fileName, const int openDisposition, const int access_flags, const int create_options, __u16 *netfid, int *pOplock, FILE_ALL_INFO *pfile_info, @@ -1341,7 +1343,7 @@ openRetry: SECURITY_CONTEXT_TRACKING | SECURITY_EFFECTIVE_ONLY; count += name_len; - pSMB->hdr.smb_buf_length += count; + inc_rfc1001_len(pSMB, count); pSMB->ByteCount = cpu_to_le16(count); /* long_op set to 1 to allow for oplock break timeouts */ @@ -1375,8 +1377,7 @@ openRetry: } int -CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, - const unsigned int count, const __u64 lseek, unsigned int *nbytes, +CIFSSMBRead(const int xid, struct cifs_io_parms *io_parms, unsigned int *nbytes, char **buf, int *pbuf_type) { int rc = -EACCES; @@ -1386,13 +1387,18 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, int wct; int resp_buf_type = 0; struct kvec iov[1]; + __u32 pid = io_parms->pid; + __u16 netfid = io_parms->netfid; + __u64 offset = io_parms->offset; + struct cifs_tcon *tcon = io_parms->tcon; + unsigned int count = io_parms->length; cFYI(1, "Reading %d bytes on fid %d", count, netfid); if (tcon->ses->capabilities & CAP_LARGE_FILES) wct = 12; else { wct = 10; /* old style read */ - if ((lseek >> 32) > 0) { + if ((offset >> 32) > 0) { /* can not handle this big offset for old */ return -EIO; } @@ -1403,15 +1409,18 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, if (rc) return rc; + pSMB->hdr.Pid = cpu_to_le16((__u16)pid); + pSMB->hdr.PidHigh = cpu_to_le16((__u16)(pid >> 16)); + /* tcon and ses pointer are checked in smb_init */ if (tcon->ses->server == NULL) return -ECONNABORTED; pSMB->AndXCommand = 0xFF; /* none */ pSMB->Fid = netfid; - pSMB->OffsetLow = cpu_to_le32(lseek & 0xFFFFFFFF); + pSMB->OffsetLow = cpu_to_le32(offset & 0xFFFFFFFF); if (wct == 12) - pSMB->OffsetHigh = cpu_to_le32(lseek >> 32); + pSMB->OffsetHigh = cpu_to_le32(offset >> 32); pSMB->Remaining = 0; pSMB->MaxCount = cpu_to_le16(count & 0xFFFF); @@ -1426,7 +1435,7 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, } iov[0].iov_base = (char *)pSMB; - iov[0].iov_len = pSMB->hdr.smb_buf_length + 4; + iov[0].iov_len = be32_to_cpu(pSMB->hdr.smb_buf_length) + 4; rc = SendReceive2(xid, tcon->ses, iov, 1 /* num iovecs */, &resp_buf_type, CIFS_LOG_ERROR); cifs_stats_inc(&tcon->num_reads); @@ -1480,9 +1489,8 @@ CIFSSMBRead(const int xid, struct cifsTconInfo *tcon, const int netfid, int -CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, - const int netfid, const unsigned int count, - const __u64 offset, unsigned int *nbytes, const char *buf, +CIFSSMBWrite(const int xid, struct cifs_io_parms *io_parms, + unsigned int *nbytes, const char *buf, const char __user *ubuf, const int long_op) { int rc = -EACCES; @@ -1491,6 +1499,11 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, int bytes_returned, wct; __u32 bytes_sent; __u16 byte_count; + __u32 pid = io_parms->pid; + __u16 netfid = io_parms->netfid; + __u64 offset = io_parms->offset; + struct cifs_tcon *tcon = io_parms->tcon; + unsigned int count = io_parms->length; *nbytes = 0; @@ -1512,6 +1525,10 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, (void **) &pSMBr); if (rc) return rc; + + pSMB->hdr.Pid = cpu_to_le16((__u16)pid); + pSMB->hdr.PidHigh = cpu_to_le16((__u16)(pid >> 16)); + /* tcon and ses pointer are checked in smb_init */ if (tcon->ses->server == NULL) return -ECONNABORTED; @@ -1560,7 +1577,7 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, pSMB->DataLengthLow = cpu_to_le16(bytes_sent & 0xFFFF); pSMB->DataLengthHigh = cpu_to_le16(bytes_sent >> 16); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); if (wct == 14) pSMB->ByteCount = cpu_to_le16(byte_count); @@ -1598,17 +1615,259 @@ CIFSSMBWrite(const int xid, struct cifsTconInfo *tcon, return rc; } +void +cifs_writedata_release(struct kref *refcount) +{ + struct cifs_writedata *wdata = container_of(refcount, + struct cifs_writedata, refcount); + + if (wdata->cfile) + cifsFileInfo_put(wdata->cfile); + + kfree(wdata); +} + +/* + * Write failed with a retryable error. Resend the write request. It's also + * possible that the page was redirtied so re-clean the page. + */ +static void +cifs_writev_requeue(struct cifs_writedata *wdata) +{ + int i, rc; + struct inode *inode = wdata->cfile->dentry->d_inode; + + for (i = 0; i < wdata->nr_pages; i++) { + lock_page(wdata->pages[i]); + clear_page_dirty_for_io(wdata->pages[i]); + } + + do { + rc = cifs_async_writev(wdata); + } while (rc == -EAGAIN); + + for (i = 0; i < wdata->nr_pages; i++) { + if (rc != 0) + SetPageError(wdata->pages[i]); + unlock_page(wdata->pages[i]); + } + + mapping_set_error(inode->i_mapping, rc); + kref_put(&wdata->refcount, cifs_writedata_release); +} + +static void +cifs_writev_complete(struct work_struct *work) +{ + struct cifs_writedata *wdata = container_of(work, + struct cifs_writedata, work); + struct inode *inode = wdata->cfile->dentry->d_inode; + int i = 0; + + if (wdata->result == 0) { + cifs_update_eof(CIFS_I(inode), wdata->offset, wdata->bytes); + cifs_stats_bytes_written(tlink_tcon(wdata->cfile->tlink), + wdata->bytes); + } else if (wdata->sync_mode == WB_SYNC_ALL && wdata->result == -EAGAIN) + return cifs_writev_requeue(wdata); + + for (i = 0; i < wdata->nr_pages; i++) { + struct page *page = wdata->pages[i]; + if (wdata->result == -EAGAIN) + __set_page_dirty_nobuffers(page); + else if (wdata->result < 0) + SetPageError(page); + end_page_writeback(page); + page_cache_release(page); + } + if (wdata->result != -EAGAIN) + mapping_set_error(inode->i_mapping, wdata->result); + kref_put(&wdata->refcount, cifs_writedata_release); +} + +struct cifs_writedata * +cifs_writedata_alloc(unsigned int nr_pages) +{ + struct cifs_writedata *wdata; + + /* this would overflow */ + if (nr_pages == 0) { + cERROR(1, "%s: called with nr_pages == 0!", __func__); + return NULL; + } + + /* writedata + number of page pointers */ + wdata = kzalloc(sizeof(*wdata) + + sizeof(struct page *) * (nr_pages - 1), GFP_NOFS); + if (wdata != NULL) { + INIT_WORK(&wdata->work, cifs_writev_complete); + kref_init(&wdata->refcount); + } + return wdata; +} + +/* + * Check the midState and signature on received buffer (if any), and queue the + * workqueue completion task. + */ +static void +cifs_writev_callback(struct mid_q_entry *mid) +{ + struct cifs_writedata *wdata = mid->callback_data; + struct cifs_tcon *tcon = tlink_tcon(wdata->cfile->tlink); + unsigned int written; + WRITE_RSP *smb = (WRITE_RSP *)mid->resp_buf; + + switch (mid->midState) { + case MID_RESPONSE_RECEIVED: + wdata->result = cifs_check_receive(mid, tcon->ses->server, 0); + if (wdata->result != 0) + break; + + written = le16_to_cpu(smb->CountHigh); + written <<= 16; + written += le16_to_cpu(smb->Count); + /* + * Mask off high 16 bits when bytes written as returned + * by the server is greater than bytes requested by the + * client. OS/2 servers are known to set incorrect + * CountHigh values. + */ + if (written > wdata->bytes) + written &= 0xFFFF; + + if (written < wdata->bytes) + wdata->result = -ENOSPC; + else + wdata->bytes = written; + break; + case MID_REQUEST_SUBMITTED: + case MID_RETRY_NEEDED: + wdata->result = -EAGAIN; + break; + default: + wdata->result = -EIO; + break; + } + + queue_work(system_nrt_wq, &wdata->work); + DeleteMidQEntry(mid); + atomic_dec(&tcon->ses->server->inFlight); + wake_up(&tcon->ses->server->request_q); +} + +/* cifs_async_writev - send an async write, and set up mid to handle result */ +int +cifs_async_writev(struct cifs_writedata *wdata) +{ + int i, rc = -EACCES; + WRITE_REQ *smb = NULL; + int wct; + struct cifs_tcon *tcon = tlink_tcon(wdata->cfile->tlink); + struct inode *inode = wdata->cfile->dentry->d_inode; + struct kvec *iov = NULL; + + if (tcon->ses->capabilities & CAP_LARGE_FILES) { + wct = 14; + } else { + wct = 12; + if (wdata->offset >> 32 > 0) { + /* can not handle big offset for old srv */ + return -EIO; + } + } + + rc = small_smb_init(SMB_COM_WRITE_ANDX, wct, tcon, (void **)&smb); + if (rc) + goto async_writev_out; + + /* 1 iov per page + 1 for header */ + iov = kzalloc((wdata->nr_pages + 1) * sizeof(*iov), GFP_NOFS); + if (iov == NULL) { + rc = -ENOMEM; + goto async_writev_out; + } + + smb->hdr.Pid = cpu_to_le16((__u16)wdata->cfile->pid); + smb->hdr.PidHigh = cpu_to_le16((__u16)(wdata->cfile->pid >> 16)); + + smb->AndXCommand = 0xFF; /* none */ + smb->Fid = wdata->cfile->netfid; + smb->OffsetLow = cpu_to_le32(wdata->offset & 0xFFFFFFFF); + if (wct == 14) + smb->OffsetHigh = cpu_to_le32(wdata->offset >> 32); + smb->Reserved = 0xFFFFFFFF; + smb->WriteMode = 0; + smb->Remaining = 0; + + smb->DataOffset = + cpu_to_le16(offsetof(struct smb_com_write_req, Data) - 4); + + /* 4 for RFC1001 length + 1 for BCC */ + iov[0].iov_len = be32_to_cpu(smb->hdr.smb_buf_length) + 4 + 1; + iov[0].iov_base = smb; + + /* marshal up the pages into iov array */ + wdata->bytes = 0; + for (i = 0; i < wdata->nr_pages; i++) { + iov[i + 1].iov_len = min(inode->i_size - + page_offset(wdata->pages[i]), + (loff_t)PAGE_CACHE_SIZE); + iov[i + 1].iov_base = kmap(wdata->pages[i]); + wdata->bytes += iov[i + 1].iov_len; + } + + cFYI(1, "async write at %llu %u bytes", wdata->offset, wdata->bytes); + + smb->DataLengthLow = cpu_to_le16(wdata->bytes & 0xFFFF); + smb->DataLengthHigh = cpu_to_le16(wdata->bytes >> 16); + + if (wct == 14) { + inc_rfc1001_len(&smb->hdr, wdata->bytes + 1); + put_bcc(wdata->bytes + 1, &smb->hdr); + } else { + /* wct == 12 */ + struct smb_com_writex_req *smbw = + (struct smb_com_writex_req *)smb; + inc_rfc1001_len(&smbw->hdr, wdata->bytes + 5); + put_bcc(wdata->bytes + 5, &smbw->hdr); + iov[0].iov_len += 4; /* pad bigger by four bytes */ + } + + kref_get(&wdata->refcount); + rc = cifs_call_async(tcon->ses->server, iov, wdata->nr_pages + 1, + cifs_writev_callback, wdata, false); + + if (rc == 0) + cifs_stats_inc(&tcon->num_writes); + else + kref_put(&wdata->refcount, cifs_writedata_release); + + /* send is done, unmap pages */ + for (i = 0; i < wdata->nr_pages; i++) + kunmap(wdata->pages[i]); + +async_writev_out: + cifs_small_buf_release(smb); + kfree(iov); + return rc; +} + int -CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, - const int netfid, const unsigned int count, - const __u64 offset, unsigned int *nbytes, struct kvec *iov, - int n_vec, const int long_op) +CIFSSMBWrite2(const int xid, struct cifs_io_parms *io_parms, + unsigned int *nbytes, struct kvec *iov, int n_vec, + const int long_op) { int rc = -EACCES; WRITE_REQ *pSMB = NULL; int wct; int smb_hdr_len; int resp_buf_type = 0; + __u32 pid = io_parms->pid; + __u16 netfid = io_parms->netfid; + __u64 offset = io_parms->offset; + struct cifs_tcon *tcon = io_parms->tcon; + unsigned int count = io_parms->length; *nbytes = 0; @@ -1626,6 +1885,10 @@ CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, rc = small_smb_init(SMB_COM_WRITE_ANDX, wct, tcon, (void **) &pSMB); if (rc) return rc; + + pSMB->hdr.Pid = cpu_to_le16((__u16)pid); + pSMB->hdr.PidHigh = cpu_to_le16((__u16)(pid >> 16)); + /* tcon and ses pointer are checked in smb_init */ if (tcon->ses->server == NULL) return -ECONNABORTED; @@ -1644,11 +1907,12 @@ CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, pSMB->DataLengthLow = cpu_to_le16(count & 0xFFFF); pSMB->DataLengthHigh = cpu_to_le16(count >> 16); - smb_hdr_len = pSMB->hdr.smb_buf_length + 1; /* hdr + 1 byte pad */ + /* header + 1 byte pad */ + smb_hdr_len = be32_to_cpu(pSMB->hdr.smb_buf_length) + 1; if (wct == 14) - pSMB->hdr.smb_buf_length += count+1; + inc_rfc1001_len(pSMB, count + 1); else /* wct == 12 */ - pSMB->hdr.smb_buf_length += count+5; /* smb data starts later */ + inc_rfc1001_len(pSMB, count + 5); /* smb data starts later */ if (wct == 14) pSMB->ByteCount = cpu_to_le16(count + 1); else /* wct == 12 */ /* bigger pad, smaller smb hdr, keep offset ok */ { @@ -1700,7 +1964,7 @@ CIFSSMBWrite2(const int xid, struct cifsTconInfo *tcon, int -CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, +CIFSSMBLock(const int xid, struct cifs_tcon *tcon, const __u16 smb_file_id, const __u64 len, const __u64 offset, const __u32 numUnlock, const __u32 numLock, const __u8 lockType, @@ -1748,7 +2012,7 @@ CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, /* oplock break */ count = 0; } - pSMB->hdr.smb_buf_length += count; + inc_rfc1001_len(pSMB, count); pSMB->ByteCount = cpu_to_le16(count); if (waitFlag) { @@ -1770,7 +2034,7 @@ CIFSSMBLock(const int xid, struct cifsTconInfo *tcon, } int -CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, +CIFSSMBPosixLock(const int xid, struct cifs_tcon *tcon, const __u16 smb_file_id, const int get_flag, const __u64 len, struct file_lock *pLockData, const __u16 lock_type, const bool waitFlag) @@ -1839,14 +2103,14 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, pSMB->Fid = smb_file_id; pSMB->InformationLevel = cpu_to_le16(SMB_SET_POSIX_LOCK); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); if (waitFlag) { rc = SendReceiveBlockingLock(xid, tcon, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned); } else { iov[0].iov_base = (char *)pSMB; - iov[0].iov_len = pSMB->hdr.smb_buf_length + 4; + iov[0].iov_len = be32_to_cpu(pSMB->hdr.smb_buf_length) + 4; rc = SendReceive2(xid, tcon->ses, iov, 1 /* num iovecs */, &resp_buf_type, timeout); pSMB = NULL; /* request buf already freed by SendReceive2. Do @@ -1862,7 +2126,7 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, __u16 data_count; rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < sizeof(struct cifs_posix_lock))) { + if (rc || get_bcc(&pSMBr->hdr) < sizeof(*parm_data)) { rc = -EIO; /* bad smb */ goto plk_err_exit; } @@ -1884,10 +2148,10 @@ CIFSSMBPosixLock(const int xid, struct cifsTconInfo *tcon, __constant_cpu_to_le16(CIFS_WRLCK)) pLockData->fl_type = F_WRLCK; - pLockData->fl_start = parm_data->start; - pLockData->fl_end = parm_data->start + - parm_data->length - 1; - pLockData->fl_pid = parm_data->pid; + pLockData->fl_start = le64_to_cpu(parm_data->start); + pLockData->fl_end = pLockData->fl_start + + le64_to_cpu(parm_data->length) - 1; + pLockData->fl_pid = le32_to_cpu(parm_data->pid); } } @@ -1908,7 +2172,7 @@ plk_err_exit: int -CIFSSMBClose(const int xid, struct cifsTconInfo *tcon, int smb_file_id) +CIFSSMBClose(const int xid, struct cifs_tcon *tcon, int smb_file_id) { int rc = 0; CLOSE_REQ *pSMB = NULL; @@ -1941,7 +2205,7 @@ CIFSSMBClose(const int xid, struct cifsTconInfo *tcon, int smb_file_id) } int -CIFSSMBFlush(const int xid, struct cifsTconInfo *tcon, int smb_file_id) +CIFSSMBFlush(const int xid, struct cifs_tcon *tcon, int smb_file_id) { int rc = 0; FLUSH_REQ *pSMB = NULL; @@ -1962,7 +2226,7 @@ CIFSSMBFlush(const int xid, struct cifsTconInfo *tcon, int smb_file_id) } int -CIFSSMBRename(const int xid, struct cifsTconInfo *tcon, +CIFSSMBRename(const int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage, int remap) { @@ -2012,7 +2276,7 @@ renameRetry: } count = 1 /* 1st signature byte */ + name_len + name_len2; - pSMB->hdr.smb_buf_length += count; + inc_rfc1001_len(pSMB, count); pSMB->ByteCount = cpu_to_le16(count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -2029,7 +2293,7 @@ renameRetry: return rc; } -int CIFSSMBRenameOpenFile(const int xid, struct cifsTconInfo *pTcon, +int CIFSSMBRenameOpenFile(const int xid, struct cifs_tcon *pTcon, int netfid, const char *target_name, const struct nls_table *nls_codepage, int remap) { @@ -2092,7 +2356,7 @@ int CIFSSMBRenameOpenFile(const int xid, struct cifsTconInfo *pTcon, pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_RENAME_INFORMATION); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, pTcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -2109,7 +2373,7 @@ int CIFSSMBRenameOpenFile(const int xid, struct cifsTconInfo *pTcon, } int -CIFSSMBCopy(const int xid, struct cifsTconInfo *tcon, const char *fromName, +CIFSSMBCopy(const int xid, struct cifs_tcon *tcon, const char *fromName, const __u16 target_tid, const char *toName, const int flags, const struct nls_table *nls_codepage, int remap) { @@ -2159,7 +2423,7 @@ copyRetry: } count = 1 /* 1st signature byte */ + name_len + name_len2; - pSMB->hdr.smb_buf_length += count; + inc_rfc1001_len(pSMB, count); pSMB->ByteCount = cpu_to_le16(count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -2177,7 +2441,7 @@ copyRetry: } int -CIFSUnixCreateSymLink(const int xid, struct cifsTconInfo *tcon, +CIFSUnixCreateSymLink(const int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage) { @@ -2249,7 +2513,7 @@ createSymLinkRetry: pSMB->DataOffset = cpu_to_le16(offset); pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_UNIX_LINK); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -2266,7 +2530,7 @@ createSymLinkRetry: } int -CIFSUnixCreateHardLink(const int xid, struct cifsTconInfo *tcon, +CIFSUnixCreateHardLink(const int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage, int remap) { @@ -2335,7 +2599,7 @@ createHardLinkRetry: pSMB->DataOffset = cpu_to_le16(offset); pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_UNIX_HLINK); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -2351,7 +2615,7 @@ createHardLinkRetry: } int -CIFSCreateHardLink(const int xid, struct cifsTconInfo *tcon, +CIFSCreateHardLink(const int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage, int remap) { @@ -2406,7 +2670,7 @@ winCreateHardLinkRetry: } count = 1 /* string type byte */ + name_len + name_len2; - pSMB->hdr.smb_buf_length += count; + inc_rfc1001_len(pSMB, count); pSMB->ByteCount = cpu_to_le16(count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -2423,7 +2687,7 @@ winCreateHardLinkRetry: } int -CIFSSMBUnixQuerySymLink(const int xid, struct cifsTconInfo *tcon, +CIFSSMBUnixQuerySymLink(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, char **symlinkinfo, const struct nls_table *nls_codepage) { @@ -2477,7 +2741,7 @@ querySymLinkRetry: pSMB->ParameterCount = pSMB->TotalParameterCount; pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FILE_UNIX_LINK); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -2489,7 +2753,7 @@ querySymLinkRetry: rc = validate_t2((struct smb_t2_rsp *)pSMBr); /* BB also check enough total bytes returned */ - if (rc || (pSMBr->ByteCount < 2)) + if (rc || get_bcc(&pSMBr->hdr) < 2) rc = -EIO; else { bool is_unicode; @@ -2516,9 +2780,19 @@ querySymLinkRetry: return rc; } -#ifdef CONFIG_CIFS_EXPERIMENTAL +#ifdef CONFIG_CIFS_SYMLINK_EXPERIMENTAL +/* + * Recent Windows versions now create symlinks more frequently + * and they use the "reparse point" mechanism below. We can of course + * do symlinks nicely to Samba and other servers which support the + * CIFS Unix Extensions and we can also do SFU symlinks and "client only" + * "MF" symlinks optionally, but for recent Windows we really need to + * reenable the code below and fix the cifs_symlink callers to handle this. + * In the interim this code has been moved to its own config option so + * it is not compiled in by default until callers fixed up and more tested. + */ int -CIFSSMBQueryReparseLinkInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBQueryReparseLinkInfo(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, char *symlinkinfo, const int buflen, __u16 fid, const struct nls_table *nls_codepage) @@ -2561,14 +2835,14 @@ CIFSSMBQueryReparseLinkInfo(const int xid, struct cifsTconInfo *tcon, } else { /* decode response */ __u32 data_offset = le32_to_cpu(pSMBr->DataOffset); __u32 data_count = le32_to_cpu(pSMBr->DataCount); - if ((pSMBr->ByteCount < 2) || (data_offset > 512)) { - /* BB also check enough total bytes returned */ + if (get_bcc(&pSMBr->hdr) < 2 || data_offset > 512) { + /* BB also check enough total bytes returned */ rc = -EIO; /* bad smb */ goto qreparse_out; } if (data_count && (data_count < 2048)) { char *end_of_smb = 2 /* sizeof byte count */ + - pSMBr->ByteCount + (char *)&pSMBr->ByteCount; + get_bcc(&pSMBr->hdr) + (char *)&pSMBr->ByteCount; struct reparse_data *reparse_buf = (struct reparse_data *) @@ -2618,7 +2892,7 @@ qreparse_out: return rc; } -#endif /* CIFS_EXPERIMENTAL */ +#endif /* CIFS_SYMLINK_EXPERIMENTAL */ /* BB temporarily unused */ #ifdef CONFIG_CIFS_POSIX @@ -2756,7 +3030,7 @@ static __u16 ACL_to_cifs_posix(char *parm_data, const char *pACL, } int -CIFSSMBGetPosixACL(const int xid, struct cifsTconInfo *tcon, +CIFSSMBGetPosixACL(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, char *acl_inf, const int buflen, const int acl_type, const struct nls_table *nls_codepage, int remap) @@ -2814,7 +3088,7 @@ queryAclRetry: pSMB->ParameterCount = pSMB->TotalParameterCount; pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_POSIX_ACL); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -2826,8 +3100,8 @@ queryAclRetry: /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 2)) /* BB also check enough total bytes returned */ + if (rc || get_bcc(&pSMBr->hdr) < 2) rc = -EIO; /* bad smb */ else { __u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset); @@ -2844,7 +3118,7 @@ queryAclRetry: } int -CIFSSMBSetPosixACL(const int xid, struct cifsTconInfo *tcon, +CIFSSMBSetPosixACL(const int xid, struct cifs_tcon *tcon, const unsigned char *fileName, const char *local_acl, const int buflen, const int acl_type, @@ -2908,7 +3182,7 @@ setAclRetry: pSMB->ParameterCount = cpu_to_le16(params); pSMB->TotalParameterCount = pSMB->ParameterCount; pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -2924,7 +3198,7 @@ setACLerrorExit: /* BB fix tabs in this function FIXME BB */ int -CIFSGetExtAttr(const int xid, struct cifsTconInfo *tcon, +CIFSGetExtAttr(const int xid, struct cifs_tcon *tcon, const int netfid, __u64 *pExtAttrBits, __u64 *pMask) { int rc = 0; @@ -2966,7 +3240,7 @@ GetExtAttrRetry: pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_ATTR_FLAGS); pSMB->Pad = 0; pSMB->Fid = netfid; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->t2.ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -2976,8 +3250,8 @@ GetExtAttrRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 2)) /* BB also check enough total bytes returned */ + if (rc || get_bcc(&pSMBr->hdr) < 2) /* If rc should we check for EOPNOSUPP and disable the srvino flag? or in caller? */ rc = -EIO; /* bad smb */ @@ -3017,7 +3291,7 @@ GetExtAttrOut: */ static int smb_init_nttransact(const __u16 sub_command, const int setup_count, - const int parm_len, struct cifsTconInfo *tcon, + const int parm_len, struct cifs_tcon *tcon, void **ret_buf) { int rc; @@ -3052,6 +3326,7 @@ validate_ntransact(char *buf, char **ppparm, char **ppdata, char *end_of_smb; __u32 data_count, data_offset, parm_count, parm_offset; struct smb_com_ntransact_rsp *pSMBr; + u16 bcc; *pdatalen = 0; *pparmlen = 0; @@ -3061,8 +3336,8 @@ validate_ntransact(char *buf, char **ppparm, char **ppdata, pSMBr = (struct smb_com_ntransact_rsp *)buf; - /* ByteCount was converted from little endian in SendReceive */ - end_of_smb = 2 /* sizeof byte count */ + pSMBr->ByteCount + + bcc = get_bcc(&pSMBr->hdr); + end_of_smb = 2 /* sizeof byte count */ + bcc + (char *)&pSMBr->ByteCount; data_offset = le32_to_cpu(pSMBr->DataOffset); @@ -3088,7 +3363,7 @@ validate_ntransact(char *buf, char **ppparm, char **ppdata, *ppdata, data_count, (data_count + *ppdata), end_of_smb, pSMBr); return -EINVAL; - } else if (parm_count + data_count > pSMBr->ByteCount) { + } else if (parm_count + data_count > bcc) { cFYI(1, "parm count and data count larger than SMB"); return -EINVAL; } @@ -3099,7 +3374,7 @@ validate_ntransact(char *buf, char **ppparm, char **ppdata, /* Get Security Descriptor (by handle) from remote server for a file or dir */ int -CIFSSMBGetCIFSACL(const int xid, struct cifsTconInfo *tcon, __u16 fid, +CIFSSMBGetCIFSACL(const int xid, struct cifs_tcon *tcon, __u16 fid, struct cifs_ntsd **acl_inf, __u32 *pbuflen) { int rc = 0; @@ -3124,9 +3399,9 @@ CIFSSMBGetCIFSACL(const int xid, struct cifsTconInfo *tcon, __u16 fid, pSMB->AclFlags = cpu_to_le32(CIFS_ACL_OWNER | CIFS_ACL_GROUP | CIFS_ACL_DACL); pSMB->ByteCount = cpu_to_le16(11); /* 3 bytes pad + 8 bytes parm */ - pSMB->hdr.smb_buf_length += 11; + inc_rfc1001_len(pSMB, 11); iov[0].iov_base = (char *)pSMB; - iov[0].iov_len = pSMB->hdr.smb_buf_length + 4; + iov[0].iov_len = be32_to_cpu(pSMB->hdr.smb_buf_length) + 4; rc = SendReceive2(xid, tcon->ses, iov, 1 /* num iovec */, &buf_type, 0); @@ -3191,7 +3466,7 @@ qsec_out: } int -CIFSSMBSetCIFSACL(const int xid, struct cifsTconInfo *tcon, __u16 fid, +CIFSSMBSetCIFSACL(const int xid, struct cifs_tcon *tcon, __u16 fid, struct cifs_ntsd *pntsd, __u32 acllen) { __u16 byte_count, param_count, data_count, param_offset, data_offset; @@ -3235,10 +3510,9 @@ setCifsAclRetry: memcpy((char *) &pSMBr->hdr.Protocol + data_offset, (char *) pntsd, acllen); - pSMB->hdr.smb_buf_length += (byte_count + data_count); - + inc_rfc1001_len(pSMB, byte_count + data_count); } else - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -3258,7 +3532,7 @@ setCifsAclRetry: /* Legacy Query Path Information call for lookup to old servers such as Win9x/WinME */ -int SMBQueryInformation(const int xid, struct cifsTconInfo *tcon, +int SMBQueryInformation(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, FILE_ALL_INFO *pFinfo, const struct nls_table *nls_codepage, int remap) @@ -3289,7 +3563,7 @@ QInfRetry: } pSMB->BufferFormat = 0x04; name_len++; /* account for buffer type byte */ - pSMB->hdr.smb_buf_length += (__u16) name_len; + inc_rfc1001_len(pSMB, (__u16)name_len); pSMB->ByteCount = cpu_to_le16(name_len); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -3326,7 +3600,7 @@ QInfRetry: } int -CIFSSMBQFileInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBQFileInfo(const int xid, struct cifs_tcon *tcon, u16 netfid, FILE_ALL_INFO *pFindData) { struct smb_t2_qfi_req *pSMB = NULL; @@ -3364,7 +3638,7 @@ QFileInfoRetry: pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FILE_ALL_INFO); pSMB->Pad = 0; pSMB->Fid = netfid; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -3375,7 +3649,7 @@ QFileInfoRetry: if (rc) /* BB add auto retry on EOPNOTSUPP? */ rc = -EIO; - else if (pSMBr->ByteCount < 40) + else if (get_bcc(&pSMBr->hdr) < 40) rc = -EIO; /* bad smb */ else if (pFindData) { __u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset); @@ -3393,7 +3667,7 @@ QFileInfoRetry: } int -CIFSSMBQPathInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBQPathInfo(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, FILE_ALL_INFO *pFindData, int legacy /* old style infolevel */, @@ -3451,7 +3725,7 @@ QPathInfoRetry: else pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FILE_ALL_INFO); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -3463,9 +3737,9 @@ QPathInfoRetry: if (rc) /* BB add auto retry on EOPNOTSUPP? */ rc = -EIO; - else if (!legacy && (pSMBr->ByteCount < 40)) + else if (!legacy && get_bcc(&pSMBr->hdr) < 40) rc = -EIO; /* bad smb */ - else if (legacy && (pSMBr->ByteCount < 24)) + else if (legacy && get_bcc(&pSMBr->hdr) < 24) rc = -EIO; /* 24 or 26 expected but we do not read last field */ else if (pFindData) { @@ -3494,7 +3768,7 @@ QPathInfoRetry: } int -CIFSSMBUnixQFileInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBUnixQFileInfo(const int xid, struct cifs_tcon *tcon, u16 netfid, FILE_UNIX_BASIC_INFO *pFindData) { struct smb_t2_qfi_req *pSMB = NULL; @@ -3532,7 +3806,7 @@ UnixQFileInfoRetry: pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FILE_UNIX_BASIC); pSMB->Pad = 0; pSMB->Fid = netfid; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -3541,7 +3815,7 @@ UnixQFileInfoRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < sizeof(FILE_UNIX_BASIC_INFO))) { + if (rc || get_bcc(&pSMBr->hdr) < sizeof(FILE_UNIX_BASIC_INFO)) { cERROR(1, "Malformed FILE_UNIX_BASIC_INFO response.\n" "Unix Extensions can be disabled on mount " "by specifying the nosfu mount option."); @@ -3563,7 +3837,7 @@ UnixQFileInfoRetry: } int -CIFSSMBUnixQPathInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBUnixQPathInfo(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, FILE_UNIX_BASIC_INFO *pFindData, const struct nls_table *nls_codepage, int remap) @@ -3617,7 +3891,7 @@ UnixQPathInfoRetry: pSMB->ParameterCount = pSMB->TotalParameterCount; pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FILE_UNIX_BASIC); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -3627,7 +3901,7 @@ UnixQPathInfoRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < sizeof(FILE_UNIX_BASIC_INFO))) { + if (rc || get_bcc(&pSMBr->hdr) < sizeof(FILE_UNIX_BASIC_INFO)) { cERROR(1, "Malformed FILE_UNIX_BASIC_INFO response.\n" "Unix Extensions can be disabled on mount " "by specifying the nosfu mount option."); @@ -3649,7 +3923,7 @@ UnixQPathInfoRetry: /* xid, tcon, searchName and codepage are input parms, rest are returned */ int -CIFSFindFirst(const int xid, struct cifsTconInfo *tcon, +CIFSFindFirst(const int xid, struct cifs_tcon *tcon, const char *searchName, const struct nls_table *nls_codepage, __u16 *pnetfid, @@ -3731,7 +4005,7 @@ findFirstRetry: /* BB what should we set StorageType to? Does it matter? BB */ pSMB->SearchStorageType = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -3797,7 +4071,7 @@ findFirstRetry: return rc; } -int CIFSFindNext(const int xid, struct cifsTconInfo *tcon, +int CIFSFindNext(const int xid, struct cifs_tcon *tcon, __u16 searchHandle, struct cifs_search_info *psrch_inf) { TRANSACTION2_FNEXT_REQ *pSMB = NULL; @@ -3860,7 +4134,7 @@ int CIFSFindNext(const int xid, struct cifsTconInfo *tcon, byte_count = params + 1 /* pad */ ; pSMB->TotalParameterCount = cpu_to_le16(params); pSMB->ParameterCount = pSMB->TotalParameterCount; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -3935,7 +4209,7 @@ FNext2_err_exit: } int -CIFSFindClose(const int xid, struct cifsTconInfo *tcon, +CIFSFindClose(const int xid, struct cifs_tcon *tcon, const __u16 searchHandle) { int rc = 0; @@ -3967,7 +4241,7 @@ CIFSFindClose(const int xid, struct cifsTconInfo *tcon, } int -CIFSGetSrvInodeNumber(const int xid, struct cifsTconInfo *tcon, +CIFSGetSrvInodeNumber(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, __u64 *inode_number, const struct nls_table *nls_codepage, int remap) @@ -4022,7 +4296,7 @@ GetInodeNumberRetry: pSMB->ParameterCount = pSMB->TotalParameterCount; pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FILE_INTERNAL_INFO); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4032,8 +4306,8 @@ GetInodeNumberRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 2)) /* BB also check enough total bytes returned */ + if (rc || get_bcc(&pSMBr->hdr) < 2) /* If rc should we check for EOPNOSUPP and disable the srvino flag? or in caller? */ rc = -EIO; /* bad smb */ @@ -4169,7 +4443,7 @@ parse_DFS_referrals_exit: } int -CIFSGetDFSRefer(const int xid, struct cifsSesInfo *ses, +CIFSGetDFSRefer(const int xid, struct cifs_ses *ses, const unsigned char *searchName, struct dfs_info3_param **target_nodes, unsigned int *num_of_nodes, @@ -4218,7 +4492,7 @@ getDFSRetry: } if (ses->server) { - if (ses->server->secMode & + if (ses->server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) pSMB->hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; } @@ -4246,7 +4520,7 @@ getDFSRetry: pSMB->ParameterCount = cpu_to_le16(params); pSMB->TotalParameterCount = pSMB->ParameterCount; pSMB->MaxReferralLevel = cpu_to_le16(3); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, ses, (struct smb_hdr *) pSMB, @@ -4258,13 +4532,13 @@ getDFSRetry: rc = validate_t2((struct smb_t2_rsp *)pSMBr); /* BB Also check if enough total bytes returned? */ - if (rc || (pSMBr->ByteCount < 17)) { + if (rc || get_bcc(&pSMBr->hdr) < 17) { rc = -EIO; /* bad smb */ goto GetDFSRefExit; } cFYI(1, "Decoding GetDFSRefer response BCC: %d Offset %d", - pSMBr->ByteCount, + get_bcc(&pSMBr->hdr), le16_to_cpu(pSMBr->t2.DataOffset)); /* parse returned result into more usable form */ @@ -4283,7 +4557,7 @@ GetDFSRefExit: /* Query File System Info such as free space to old servers such as Win 9x */ int -SMBOldQFSInfo(const int xid, struct cifsTconInfo *tcon, struct kstatfs *FSData) +SMBOldQFSInfo(const int xid, struct cifs_tcon *tcon, struct kstatfs *FSData) { /* level 0x01 SMB_QUERY_FILE_SYSTEM_INFO */ TRANSACTION2_QFSI_REQ *pSMB = NULL; @@ -4320,7 +4594,7 @@ oldQFSInfoRetry: pSMB->Reserved3 = 0; pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_FS_INFORMATION); pSMB->InformationLevel = cpu_to_le16(SMB_INFO_ALLOCATION); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4330,12 +4604,12 @@ oldQFSInfoRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 18)) + if (rc || get_bcc(&pSMBr->hdr) < 18) rc = -EIO; /* bad smb */ else { __u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset); cFYI(1, "qfsinf resp BCC: %d Offset %d", - pSMBr->ByteCount, data_offset); + get_bcc(&pSMBr->hdr), data_offset); response_data = (FILE_SYSTEM_ALLOC_INFO *) (((char *) &pSMBr->hdr.Protocol) + data_offset); @@ -4362,7 +4636,7 @@ oldQFSInfoRetry: } int -CIFSSMBQFSInfo(const int xid, struct cifsTconInfo *tcon, struct kstatfs *FSData) +CIFSSMBQFSInfo(const int xid, struct cifs_tcon *tcon, struct kstatfs *FSData) { /* level 0x103 SMB_QUERY_FILE_SYSTEM_INFO */ TRANSACTION2_QFSI_REQ *pSMB = NULL; @@ -4399,7 +4673,7 @@ QFSInfoRetry: pSMB->Reserved3 = 0; pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_FS_INFORMATION); pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FS_SIZE_INFO); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4409,7 +4683,7 @@ QFSInfoRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 24)) + if (rc || get_bcc(&pSMBr->hdr) < 24) rc = -EIO; /* bad smb */ else { __u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset); @@ -4441,7 +4715,7 @@ QFSInfoRetry: } int -CIFSSMBQFSAttributeInfo(const int xid, struct cifsTconInfo *tcon) +CIFSSMBQFSAttributeInfo(const int xid, struct cifs_tcon *tcon) { /* level 0x105 SMB_QUERY_FILE_SYSTEM_INFO */ TRANSACTION2_QFSI_REQ *pSMB = NULL; @@ -4479,7 +4753,7 @@ QFSAttributeRetry: pSMB->Reserved3 = 0; pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_FS_INFORMATION); pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FS_ATTRIBUTE_INFO); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4489,7 +4763,7 @@ QFSAttributeRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 13)) { + if (rc || get_bcc(&pSMBr->hdr) < 13) { /* BB also check if enough bytes returned */ rc = -EIO; /* bad smb */ } else { @@ -4511,7 +4785,7 @@ QFSAttributeRetry: } int -CIFSSMBQFSDeviceInfo(const int xid, struct cifsTconInfo *tcon) +CIFSSMBQFSDeviceInfo(const int xid, struct cifs_tcon *tcon) { /* level 0x104 SMB_QUERY_FILE_SYSTEM_INFO */ TRANSACTION2_QFSI_REQ *pSMB = NULL; @@ -4550,7 +4824,7 @@ QFSDeviceRetry: pSMB->Reserved3 = 0; pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_FS_INFORMATION); pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_FS_DEVICE_INFO); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4560,7 +4834,8 @@ QFSDeviceRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < sizeof(FILE_SYSTEM_DEVICE_INFO))) + if (rc || get_bcc(&pSMBr->hdr) < + sizeof(FILE_SYSTEM_DEVICE_INFO)) rc = -EIO; /* bad smb */ else { __u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset); @@ -4581,7 +4856,7 @@ QFSDeviceRetry: } int -CIFSSMBQFSUnixInfo(const int xid, struct cifsTconInfo *tcon) +CIFSSMBQFSUnixInfo(const int xid, struct cifs_tcon *tcon) { /* level 0x200 SMB_QUERY_CIFS_UNIX_INFO */ TRANSACTION2_QFSI_REQ *pSMB = NULL; @@ -4619,7 +4894,7 @@ QFSUnixRetry: pSMB->Reserved3 = 0; pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_FS_INFORMATION); pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_CIFS_UNIX_INFO); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4629,7 +4904,7 @@ QFSUnixRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 13)) { + if (rc || get_bcc(&pSMBr->hdr) < 13) { rc = -EIO; /* bad smb */ } else { __u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset); @@ -4651,7 +4926,7 @@ QFSUnixRetry: } int -CIFSSMBSetFSUnixInfo(const int xid, struct cifsTconInfo *tcon, __u64 cap) +CIFSSMBSetFSUnixInfo(const int xid, struct cifs_tcon *tcon, __u64 cap) { /* level 0x200 SMB_SET_CIFS_UNIX_INFO */ TRANSACTION2_SETFSI_REQ *pSMB = NULL; @@ -4702,7 +4977,7 @@ SETFSUnixRetry: pSMB->ClientUnixMinor = cpu_to_le16(CIFS_UNIX_MINOR_VERSION); pSMB->ClientUnixCap = cpu_to_le64(cap); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4725,7 +5000,7 @@ SETFSUnixRetry: int -CIFSSMBQFSPosixInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBQFSPosixInfo(const int xid, struct cifs_tcon *tcon, struct kstatfs *FSData) { /* level 0x201 SMB_QUERY_CIFS_POSIX_INFO */ @@ -4764,7 +5039,7 @@ QFSPosixRetry: pSMB->Reserved3 = 0; pSMB->SubCommand = cpu_to_le16(TRANS2_QUERY_FS_INFORMATION); pSMB->InformationLevel = cpu_to_le16(SMB_QUERY_POSIX_FS_INFO); - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4774,7 +5049,7 @@ QFSPosixRetry: } else { /* decode response */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 13)) { + if (rc || get_bcc(&pSMBr->hdr) < 13) { rc = -EIO; /* bad smb */ } else { __u16 data_offset = le16_to_cpu(pSMBr->t2.DataOffset); @@ -4818,7 +5093,7 @@ QFSPosixRetry: in Samba which this routine can run into */ int -CIFSSMBSetEOF(const int xid, struct cifsTconInfo *tcon, const char *fileName, +CIFSSMBSetEOF(const int xid, struct cifs_tcon *tcon, const char *fileName, __u64 size, bool SetAllocation, const struct nls_table *nls_codepage, int remap) { @@ -4890,7 +5165,7 @@ SetEOFRetry: pSMB->ParameterCount = cpu_to_le16(params); pSMB->TotalParameterCount = pSMB->ParameterCount; pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); parm_data->FileSize = cpu_to_le64(size); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -4907,7 +5182,7 @@ SetEOFRetry: } int -CIFSSMBSetFileSize(const int xid, struct cifsTconInfo *tcon, __u64 size, +CIFSSMBSetFileSize(const int xid, struct cifs_tcon *tcon, __u64 size, __u16 fid, __u32 pid_of_opener, bool SetAllocation) { struct smb_com_transaction2_sfi_req *pSMB = NULL; @@ -4969,7 +5244,7 @@ CIFSSMBSetFileSize(const int xid, struct cifsTconInfo *tcon, __u64 size, cpu_to_le16(SMB_SET_FILE_END_OF_FILE_INFO); } pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceiveNoRsp(xid, tcon->ses, (struct smb_hdr *) pSMB, 0); if (rc) { @@ -4989,7 +5264,7 @@ CIFSSMBSetFileSize(const int xid, struct cifsTconInfo *tcon, __u64 size, time and resort to the original setpathinfo level which takes the ancient DOS time format with 2 second granularity */ int -CIFSSMBSetFileInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBSetFileInfo(const int xid, struct cifs_tcon *tcon, const FILE_BASIC_INFO *data, __u16 fid, __u32 pid_of_opener) { struct smb_com_transaction2_sfi_req *pSMB = NULL; @@ -5037,7 +5312,7 @@ CIFSSMBSetFileInfo(const int xid, struct cifsTconInfo *tcon, else pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_BASIC_INFO); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); memcpy(data_offset, data, sizeof(FILE_BASIC_INFO)); rc = SendReceiveNoRsp(xid, tcon->ses, (struct smb_hdr *) pSMB, 0); @@ -5051,7 +5326,7 @@ CIFSSMBSetFileInfo(const int xid, struct cifsTconInfo *tcon, } int -CIFSSMBSetFileDisposition(const int xid, struct cifsTconInfo *tcon, +CIFSSMBSetFileDisposition(const int xid, struct cifs_tcon *tcon, bool delete_file, __u16 fid, __u32 pid_of_opener) { struct smb_com_transaction2_sfi_req *pSMB = NULL; @@ -5096,7 +5371,7 @@ CIFSSMBSetFileDisposition(const int xid, struct cifsTconInfo *tcon, pSMB->Fid = fid; pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_DISPOSITION_INFO); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); *data_offset = delete_file ? 1 : 0; rc = SendReceiveNoRsp(xid, tcon->ses, (struct smb_hdr *) pSMB, 0); @@ -5107,7 +5382,7 @@ CIFSSMBSetFileDisposition(const int xid, struct cifsTconInfo *tcon, } int -CIFSSMBSetPathInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBSetPathInfo(const int xid, struct cifs_tcon *tcon, const char *fileName, const FILE_BASIC_INFO *data, const struct nls_table *nls_codepage, int remap) { @@ -5169,7 +5444,7 @@ SetTimesRetry: else pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_BASIC_INFO); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); memcpy(data_offset, data, sizeof(FILE_BASIC_INFO)); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -5191,7 +5466,7 @@ SetTimesRetry: handling it anyway and NT4 was what we thought it would be needed for Do not delete it until we prove whether needed for Win9x though */ int -CIFSSMBSetAttrLegacy(int xid, struct cifsTconInfo *tcon, char *fileName, +CIFSSMBSetAttrLegacy(int xid, struct cifs_tcon *tcon, char *fileName, __u16 dos_attrs, const struct nls_table *nls_codepage) { SETATTR_REQ *pSMB = NULL; @@ -5221,7 +5496,7 @@ SetAttrLgcyRetry: } pSMB->attr = cpu_to_le16(dos_attrs); pSMB->BufferFormat = 0x04; - pSMB->hdr.smb_buf_length += name_len + 1; + inc_rfc1001_len(pSMB, name_len + 1); pSMB->ByteCount = cpu_to_le16(name_len + 1); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -5247,7 +5522,7 @@ cifs_fill_unix_set_info(FILE_UNIX_BASIC_INFO *data_offset, * Samba server ignores set of file size to zero due to bugs in some * older clients, but we should be precise - we use SetFileSize to * set file size and do not want to truncate file size to zero - * accidently as happened on one Samba server beta by putting + * accidentally as happened on one Samba server beta by putting * zero instead of -1 here */ data_offset->EndOfFile = cpu_to_le64(NO_CHANGE_64); @@ -5279,7 +5554,7 @@ cifs_fill_unix_set_info(FILE_UNIX_BASIC_INFO *data_offset, } int -CIFSSMBUnixSetFileInfo(const int xid, struct cifsTconInfo *tcon, +CIFSSMBUnixSetFileInfo(const int xid, struct cifs_tcon *tcon, const struct cifs_unix_set_info_args *args, u16 fid, u32 pid_of_opener) { @@ -5326,7 +5601,7 @@ CIFSSMBUnixSetFileInfo(const int xid, struct cifsTconInfo *tcon, pSMB->Fid = fid; pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_UNIX_BASIC); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); cifs_fill_unix_set_info(data_offset, args); @@ -5342,7 +5617,7 @@ CIFSSMBUnixSetFileInfo(const int xid, struct cifsTconInfo *tcon, } int -CIFSSMBUnixSetPathInfo(const int xid, struct cifsTconInfo *tcon, char *fileName, +CIFSSMBUnixSetPathInfo(const int xid, struct cifs_tcon *tcon, char *fileName, const struct cifs_unix_set_info_args *args, const struct nls_table *nls_codepage, int remap) { @@ -5402,7 +5677,7 @@ setPermsRetry: pSMB->TotalDataCount = pSMB->DataCount; pSMB->InformationLevel = cpu_to_le16(SMB_SET_FILE_UNIX_BASIC); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); cifs_fill_unix_set_info(data_offset, args); @@ -5418,79 +5693,6 @@ setPermsRetry: return rc; } -int CIFSSMBNotify(const int xid, struct cifsTconInfo *tcon, - const int notify_subdirs, const __u16 netfid, - __u32 filter, struct file *pfile, int multishot, - const struct nls_table *nls_codepage) -{ - int rc = 0; - struct smb_com_transaction_change_notify_req *pSMB = NULL; - struct smb_com_ntransaction_change_notify_rsp *pSMBr = NULL; - struct dir_notify_req *dnotify_req; - int bytes_returned; - - cFYI(1, "In CIFSSMBNotify for file handle %d", (int)netfid); - rc = smb_init(SMB_COM_NT_TRANSACT, 23, tcon, (void **) &pSMB, - (void **) &pSMBr); - if (rc) - return rc; - - pSMB->TotalParameterCount = 0 ; - pSMB->TotalDataCount = 0; - pSMB->MaxParameterCount = cpu_to_le32(2); - /* BB find exact data count max from sess structure BB */ - pSMB->MaxDataCount = 0; /* same in little endian or be */ -/* BB VERIFY verify which is correct for above BB */ - pSMB->MaxDataCount = cpu_to_le32((tcon->ses->server->maxBuf - - MAX_CIFS_HDR_SIZE) & 0xFFFFFF00); - - pSMB->MaxSetupCount = 4; - pSMB->Reserved = 0; - pSMB->ParameterOffset = 0; - pSMB->DataCount = 0; - pSMB->DataOffset = 0; - pSMB->SetupCount = 4; /* single byte does not need le conversion */ - pSMB->SubCommand = cpu_to_le16(NT_TRANSACT_NOTIFY_CHANGE); - pSMB->ParameterCount = pSMB->TotalParameterCount; - if (notify_subdirs) - pSMB->WatchTree = 1; /* one byte - no le conversion needed */ - pSMB->Reserved2 = 0; - pSMB->CompletionFilter = cpu_to_le32(filter); - pSMB->Fid = netfid; /* file handle always le */ - pSMB->ByteCount = 0; - - rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, - (struct smb_hdr *)pSMBr, &bytes_returned, - CIFS_ASYNC_OP); - if (rc) { - cFYI(1, "Error in Notify = %d", rc); - } else { - /* Add file to outstanding requests */ - /* BB change to kmem cache alloc */ - dnotify_req = kmalloc( - sizeof(struct dir_notify_req), - GFP_KERNEL); - if (dnotify_req) { - dnotify_req->Pid = pSMB->hdr.Pid; - dnotify_req->PidHigh = pSMB->hdr.PidHigh; - dnotify_req->Mid = pSMB->hdr.Mid; - dnotify_req->Tid = pSMB->hdr.Tid; - dnotify_req->Uid = pSMB->hdr.Uid; - dnotify_req->netfid = netfid; - dnotify_req->pfile = pfile; - dnotify_req->filter = filter; - dnotify_req->multishot = multishot; - spin_lock(&GlobalMid_Lock); - list_add_tail(&dnotify_req->lhead, - &GlobalDnotifyReqList); - spin_unlock(&GlobalMid_Lock); - } else - rc = -ENOMEM; - } - cifs_buf_release(pSMB); - return rc; -} - #ifdef CONFIG_CIFS_XATTR /* * Do a path-based QUERY_ALL_EAS call and parse the result. This is a common @@ -5502,7 +5704,7 @@ int CIFSSMBNotify(const int xid, struct cifsTconInfo *tcon, * the data isn't copied to it, but the length is returned. */ ssize_t -CIFSSMBQAllEAs(const int xid, struct cifsTconInfo *tcon, +CIFSSMBQAllEAs(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, const unsigned char *ea_name, char *EAData, size_t buf_size, const struct nls_table *nls_codepage, int remap) @@ -5560,7 +5762,7 @@ QAllEAsRetry: pSMB->ParameterCount = pSMB->TotalParameterCount; pSMB->InformationLevel = cpu_to_le16(SMB_INFO_QUERY_ALL_EAS); pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, @@ -5576,7 +5778,7 @@ QAllEAsRetry: of these trans2 responses */ rc = validate_t2((struct smb_t2_rsp *)pSMBr); - if (rc || (pSMBr->ByteCount < 4)) { + if (rc || get_bcc(&pSMBr->hdr) < 4) { rc = -EIO; /* bad smb */ goto QAllEAsOut; } @@ -5683,7 +5885,7 @@ QAllEAsOut: } int -CIFSSMBSetEA(const int xid, struct cifsTconInfo *tcon, const char *fileName, +CIFSSMBSetEA(const int xid, struct cifs_tcon *tcon, const char *fileName, const char *ea_name, const void *ea_value, const __u16 ea_value_len, const struct nls_table *nls_codepage, int remap) @@ -5773,7 +5975,7 @@ SetEARetry: pSMB->ParameterCount = cpu_to_le16(params); pSMB->TotalParameterCount = pSMB->ParameterCount; pSMB->Reserved4 = 0; - pSMB->hdr.smb_buf_length += byte_count; + inc_rfc1001_len(pSMB, byte_count); pSMB->ByteCount = cpu_to_le16(byte_count); rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, (struct smb_hdr *) pSMBr, &bytes_returned, 0); @@ -5787,5 +5989,99 @@ SetEARetry: return rc; } - #endif + +#ifdef CONFIG_CIFS_DNOTIFY_EXPERIMENTAL /* BB unused temporarily */ +/* + * Years ago the kernel added a "dnotify" function for Samba server, + * to allow network clients (such as Windows) to display updated + * lists of files in directory listings automatically when + * files are added by one user when another user has the + * same directory open on their desktop. The Linux cifs kernel + * client hooked into the kernel side of this interface for + * the same reason, but ironically when the VFS moved from + * "dnotify" to "inotify" it became harder to plug in Linux + * network file system clients (the most obvious use case + * for notify interfaces is when multiple users can update + * the contents of the same directory - exactly what network + * file systems can do) although the server (Samba) could + * still use it. For the short term we leave the worker + * function ifdeffed out (below) until inotify is fixed + * in the VFS to make it easier to plug in network file + * system clients. If inotify turns out to be permanently + * incompatible for network fs clients, we could instead simply + * expose this config flag by adding a future cifs (and smb2) notify ioctl. + */ +int CIFSSMBNotify(const int xid, struct cifs_tcon *tcon, + const int notify_subdirs, const __u16 netfid, + __u32 filter, struct file *pfile, int multishot, + const struct nls_table *nls_codepage) +{ + int rc = 0; + struct smb_com_transaction_change_notify_req *pSMB = NULL; + struct smb_com_ntransaction_change_notify_rsp *pSMBr = NULL; + struct dir_notify_req *dnotify_req; + int bytes_returned; + + cFYI(1, "In CIFSSMBNotify for file handle %d", (int)netfid); + rc = smb_init(SMB_COM_NT_TRANSACT, 23, tcon, (void **) &pSMB, + (void **) &pSMBr); + if (rc) + return rc; + + pSMB->TotalParameterCount = 0 ; + pSMB->TotalDataCount = 0; + pSMB->MaxParameterCount = cpu_to_le32(2); + /* BB find exact data count max from sess structure BB */ + pSMB->MaxDataCount = 0; /* same in little endian or be */ +/* BB VERIFY verify which is correct for above BB */ + pSMB->MaxDataCount = cpu_to_le32((tcon->ses->server->maxBuf - + MAX_CIFS_HDR_SIZE) & 0xFFFFFF00); + + pSMB->MaxSetupCount = 4; + pSMB->Reserved = 0; + pSMB->ParameterOffset = 0; + pSMB->DataCount = 0; + pSMB->DataOffset = 0; + pSMB->SetupCount = 4; /* single byte does not need le conversion */ + pSMB->SubCommand = cpu_to_le16(NT_TRANSACT_NOTIFY_CHANGE); + pSMB->ParameterCount = pSMB->TotalParameterCount; + if (notify_subdirs) + pSMB->WatchTree = 1; /* one byte - no le conversion needed */ + pSMB->Reserved2 = 0; + pSMB->CompletionFilter = cpu_to_le32(filter); + pSMB->Fid = netfid; /* file handle always le */ + pSMB->ByteCount = 0; + + rc = SendReceive(xid, tcon->ses, (struct smb_hdr *) pSMB, + (struct smb_hdr *)pSMBr, &bytes_returned, + CIFS_ASYNC_OP); + if (rc) { + cFYI(1, "Error in Notify = %d", rc); + } else { + /* Add file to outstanding requests */ + /* BB change to kmem cache alloc */ + dnotify_req = kmalloc( + sizeof(struct dir_notify_req), + GFP_KERNEL); + if (dnotify_req) { + dnotify_req->Pid = pSMB->hdr.Pid; + dnotify_req->PidHigh = pSMB->hdr.PidHigh; + dnotify_req->Mid = pSMB->hdr.Mid; + dnotify_req->Tid = pSMB->hdr.Tid; + dnotify_req->Uid = pSMB->hdr.Uid; + dnotify_req->netfid = netfid; + dnotify_req->pfile = pfile; + dnotify_req->filter = filter; + dnotify_req->multishot = multishot; + spin_lock(&GlobalMid_Lock); + list_add_tail(&dnotify_req->lhead, + &GlobalDnotifyReqList); + spin_unlock(&GlobalMid_Lock); + } else + rc = -ENOMEM; + } + cifs_buf_release(pSMB); + return rc; +} +#endif /* was needed for dnotify, and will be needed for inotify when VFS fix */ diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 8d6c17ab593d..6d88b82537c3 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -57,61 +57,6 @@ extern mempool_t *cifs_req_poolp; -struct smb_vol { - char *username; - char *password; - char *domainname; - char *UNC; - char *UNCip; - char *iocharset; /* local code page for mapping to and from Unicode */ - char source_rfc1001_name[RFC1001_NAME_LEN_WITH_NULL]; /* clnt nb name */ - char target_rfc1001_name[RFC1001_NAME_LEN_WITH_NULL]; /* srvr nb name */ - uid_t cred_uid; - uid_t linux_uid; - gid_t linux_gid; - mode_t file_mode; - mode_t dir_mode; - unsigned secFlg; - bool retry:1; - bool intr:1; - bool setuids:1; - bool override_uid:1; - bool override_gid:1; - bool dynperm:1; - bool noperm:1; - bool no_psx_acl:1; /* set if posix acl support should be disabled */ - bool cifs_acl:1; - bool no_xattr:1; /* set if xattr (EA) support should be disabled*/ - bool server_ino:1; /* use inode numbers from server ie UniqueId */ - bool direct_io:1; - bool strict_io:1; /* strict cache behavior */ - bool remap:1; /* set to remap seven reserved chars in filenames */ - bool posix_paths:1; /* unset to not ask for posix pathnames. */ - bool no_linux_ext:1; - bool sfu_emul:1; - bool nullauth:1; /* attempt to authenticate with null user */ - bool nocase:1; /* request case insensitive filenames */ - bool nobrl:1; /* disable sending byte range locks to srv */ - bool mand_lock:1; /* send mandatory not posix byte range lock reqs */ - bool seal:1; /* request transport encryption on share */ - bool nodfs:1; /* Do not request DFS, even if available */ - bool local_lease:1; /* check leases only on local system, not remote */ - bool noblocksnd:1; - bool noautotune:1; - bool nostrictsync:1; /* do not force expensive SMBflush on every sync */ - bool fsc:1; /* enable fscache */ - bool mfsymlinks:1; /* use Minshall+French Symlinks */ - bool multiuser:1; - unsigned int rsize; - unsigned int wsize; - bool sockopt_tcp_nodelay:1; - unsigned short int port; - unsigned long actimeo; /* attribute cache timeout (jiffies) */ - char *prepath; - struct sockaddr_storage srcaddr; /* allow binding to a local IP */ - struct nls_table *local_nls; -}; - /* FIXME: should these be tunable? */ #define TLINK_ERROR_EXPIRE (1 * HZ) #define TLINK_IDLE_EXPIRE (600 * HZ) @@ -134,9 +79,10 @@ cifs_reconnect(struct TCP_Server_Info *server) { int rc = 0; struct list_head *tmp, *tmp2; - struct cifsSesInfo *ses; - struct cifsTconInfo *tcon; + struct cifs_ses *ses; + struct cifs_tcon *tcon; struct mid_q_entry *mid_entry; + struct list_head retry_list; spin_lock(&GlobalMid_Lock); if (server->tcpStatus == CifsExiting) { @@ -156,11 +102,11 @@ cifs_reconnect(struct TCP_Server_Info *server) cFYI(1, "%s: marking sessions and tcons for reconnect", __func__); spin_lock(&cifs_tcp_ses_lock); list_for_each(tmp, &server->smb_ses_list) { - ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); + ses = list_entry(tmp, struct cifs_ses, smb_ses_list); ses->need_reconnect = true; ses->ipc_tid = 0; list_for_each(tmp2, &ses->tcon_list) { - tcon = list_entry(tmp2, struct cifsTconInfo, tcon_list); + tcon = list_entry(tmp2, struct cifs_tcon, tcon_list); tcon->need_reconnect = true; } } @@ -188,19 +134,25 @@ cifs_reconnect(struct TCP_Server_Info *server) mutex_unlock(&server->srv_mutex); /* mark submitted MIDs for retry and issue callback */ - cFYI(1, "%s: issuing mid callbacks", __func__); + INIT_LIST_HEAD(&retry_list); + cFYI(1, "%s: moving mids to private list", __func__); spin_lock(&GlobalMid_Lock); list_for_each_safe(tmp, tmp2, &server->pending_mid_q) { mid_entry = list_entry(tmp, struct mid_q_entry, qhead); if (mid_entry->midState == MID_REQUEST_SUBMITTED) mid_entry->midState = MID_RETRY_NEEDED; + list_move(&mid_entry->qhead, &retry_list); + } + spin_unlock(&GlobalMid_Lock); + + cFYI(1, "%s: issuing mid callbacks", __func__); + list_for_each_safe(tmp, tmp2, &retry_list) { + mid_entry = list_entry(tmp, struct mid_q_entry, qhead); list_del_init(&mid_entry->qhead); mid_entry->callback(mid_entry); } - spin_unlock(&GlobalMid_Lock); - while ((server->tcpStatus != CifsExiting) && - (server->tcpStatus != CifsGood)) { + while (server->tcpStatus == CifsNeedReconnect) { try_to_freeze(); /* we should try only the port we connected to before */ @@ -212,7 +164,7 @@ cifs_reconnect(struct TCP_Server_Info *server) atomic_inc(&tcpSesReconnectCount); spin_lock(&GlobalMid_Lock); if (server->tcpStatus != CifsExiting) - server->tcpStatus = CifsGood; + server->tcpStatus = CifsNeedNegotiate; spin_unlock(&GlobalMid_Lock); } } @@ -248,24 +200,24 @@ static int check2ndT2(struct smb_hdr *pSMB, unsigned int maxBufSize) total_data_size = get_unaligned_le16(&pSMBt->t2_rsp.TotalDataCount); data_in_this_rsp = get_unaligned_le16(&pSMBt->t2_rsp.DataCount); - remaining = total_data_size - data_in_this_rsp; - - if (remaining == 0) + if (total_data_size == data_in_this_rsp) return 0; - else if (remaining < 0) { + else if (total_data_size < data_in_this_rsp) { cFYI(1, "total data %d smaller than data in frame %d", total_data_size, data_in_this_rsp); return -EINVAL; - } else { - cFYI(1, "missing %d bytes from transact2, check next response", - remaining); - if (total_data_size > maxBufSize) { - cERROR(1, "TotalDataSize %d is over maximum buffer %d", - total_data_size, maxBufSize); - return -EINVAL; - } - return remaining; } + + remaining = total_data_size - data_in_this_rsp; + + cFYI(1, "missing %d bytes from transact2, check next response", + remaining); + if (total_data_size > maxBufSize) { + cERROR(1, "TotalDataSize %d is over maximum buffer %d", + total_data_size, maxBufSize); + return -EINVAL; + } + return remaining; } static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) @@ -275,7 +227,8 @@ static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) char *data_area_of_target; char *data_area_of_buf2; int remaining; - __u16 byte_count, total_data_size, total_in_buf, total_in_buf2; + unsigned int byte_count, total_in_buf; + __u16 total_data_size, total_in_buf2; total_data_size = get_unaligned_le16(&pSMBt->t2_rsp.TotalDataCount); @@ -288,7 +241,7 @@ static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) remaining = total_data_size - total_in_buf; if (remaining < 0) - return -EINVAL; + return -EPROTO; if (remaining == 0) /* nothing to do, ignore */ return 0; @@ -309,19 +262,28 @@ static int coalesce_t2(struct smb_hdr *psecond, struct smb_hdr *pTargetSMB) data_area_of_target += total_in_buf; /* copy second buffer into end of first buffer */ - memcpy(data_area_of_target, data_area_of_buf2, total_in_buf2); total_in_buf += total_in_buf2; + /* is the result too big for the field? */ + if (total_in_buf > USHRT_MAX) + return -EPROTO; put_unaligned_le16(total_in_buf, &pSMBt->t2_rsp.DataCount); - byte_count = get_bcc_le(pTargetSMB); - byte_count += total_in_buf2; - put_bcc_le(byte_count, pTargetSMB); - byte_count = pTargetSMB->smb_buf_length; + /* fix up the BCC */ + byte_count = get_bcc(pTargetSMB); byte_count += total_in_buf2; + /* is the result too big for the field? */ + if (byte_count > USHRT_MAX) + return -EPROTO; + put_bcc(byte_count, pTargetSMB); - /* BB also add check that we are not beyond maximum buffer size */ + byte_count = be32_to_cpu(pTargetSMB->smb_buf_length); + byte_count += total_in_buf2; + /* don't allow buffer to overflow */ + if (byte_count > CIFSMaxBufSize) + return -ENOBUFS; + pTargetSMB->smb_buf_length = cpu_to_be32(byte_count); - pTargetSMB->smb_buf_length = byte_count; + memcpy(data_area_of_target, data_area_of_buf2, total_in_buf2); if (remaining == total_in_buf2) { cFYI(1, "found the last secondary response"); @@ -421,7 +383,7 @@ cifs_demultiplex_thread(struct TCP_Server_Info *server) pdu_length = 4; /* enough to get RFC1001 header */ incomplete_rcv: - if (echo_retries > 0 && + if (echo_retries > 0 && server->tcpStatus == CifsGood && time_after(jiffies, server->lstrp + (echo_retries * SMB_ECHO_INTERVAL))) { cERROR(1, "Server %s has not responded in %d seconds. " @@ -486,8 +448,7 @@ incomplete_rcv: /* Note that FC 1001 length is big endian on the wire, but we convert it here so it is always manipulated as host byte order */ - pdu_length = be32_to_cpu((__force __be32)smb_buffer->smb_buf_length); - smb_buffer->smb_buf_length = pdu_length; + pdu_length = be32_to_cpu(smb_buffer->smb_buf_length); cFYI(1, "rfc1002 length 0x%x", pdu_length+4); @@ -608,63 +569,67 @@ incomplete_rcv: list_for_each_safe(tmp, tmp2, &server->pending_mid_q) { mid_entry = list_entry(tmp, struct mid_q_entry, qhead); - if ((mid_entry->mid == smb_buffer->Mid) && - (mid_entry->midState == MID_REQUEST_SUBMITTED) && - (mid_entry->command == smb_buffer->Command)) { - if (length == 0 && - check2ndT2(smb_buffer, server->maxBuf) > 0) { - /* We have a multipart transact2 resp */ - isMultiRsp = true; - if (mid_entry->resp_buf) { - /* merge response - fix up 1st*/ - if (coalesce_t2(smb_buffer, - mid_entry->resp_buf)) { - mid_entry->multiRsp = - true; - break; - } else { - /* all parts received */ - mid_entry->multiEnd = - true; - goto multi_t2_fnd; - } + if (mid_entry->mid != smb_buffer->Mid || + mid_entry->midState != MID_REQUEST_SUBMITTED || + mid_entry->command != smb_buffer->Command) { + mid_entry = NULL; + continue; + } + + if (length == 0 && + check2ndT2(smb_buffer, server->maxBuf) > 0) { + /* We have a multipart transact2 resp */ + isMultiRsp = true; + if (mid_entry->resp_buf) { + /* merge response - fix up 1st*/ + length = coalesce_t2(smb_buffer, + mid_entry->resp_buf); + if (length > 0) { + length = 0; + mid_entry->multiRsp = true; + break; } else { - if (!isLargeBuf) { - cERROR(1, "1st trans2 resp needs bigbuf"); - /* BB maybe we can fix this up, switch - to already allocated large buffer? */ - } else { - /* Have first buffer */ - mid_entry->resp_buf = - smb_buffer; - mid_entry->largeBuf = - true; - bigbuf = NULL; - } + /* all parts received or + * packet is malformed + */ + mid_entry->multiEnd = true; + goto multi_t2_fnd; + } + } else { + if (!isLargeBuf) { + /* + * FIXME: switch to already + * allocated largebuf? + */ + cERROR(1, "1st trans2 resp " + "needs bigbuf"); + } else { + /* Have first buffer */ + mid_entry->resp_buf = + smb_buffer; + mid_entry->largeBuf = true; + bigbuf = NULL; } - break; } - mid_entry->resp_buf = smb_buffer; - mid_entry->largeBuf = isLargeBuf; + break; + } + mid_entry->resp_buf = smb_buffer; + mid_entry->largeBuf = isLargeBuf; multi_t2_fnd: - if (length == 0) - mid_entry->midState = - MID_RESPONSE_RECEIVED; - else - mid_entry->midState = - MID_RESPONSE_MALFORMED; + if (length == 0) + mid_entry->midState = MID_RESPONSE_RECEIVED; + else + mid_entry->midState = MID_RESPONSE_MALFORMED; #ifdef CONFIG_CIFS_STATS2 - mid_entry->when_received = jiffies; + mid_entry->when_received = jiffies; #endif - list_del_init(&mid_entry->qhead); - mid_entry->callback(mid_entry); - break; - } - mid_entry = NULL; + list_del_init(&mid_entry->qhead); + break; } spin_unlock(&GlobalMid_Lock); if (mid_entry != NULL) { + mid_entry->callback(mid_entry); /* Was previous buf put in mpx struct for multi-rsp? */ if (!isMultiRsp) { /* smb buffer will be freed by user thread */ @@ -722,21 +687,31 @@ multi_t2_fnd: sock_release(csocket); server->ssocket = NULL; } - /* buffer usuallly freed in free_mid - need to free it here on exit */ + /* buffer usually freed in free_mid - need to free it here on exit */ cifs_buf_release(bigbuf); if (smallbuf) /* no sense logging a debug message if NULL */ cifs_small_buf_release(smallbuf); if (!list_empty(&server->pending_mid_q)) { + struct list_head dispose_list; + + INIT_LIST_HEAD(&dispose_list); spin_lock(&GlobalMid_Lock); list_for_each_safe(tmp, tmp2, &server->pending_mid_q) { mid_entry = list_entry(tmp, struct mid_q_entry, qhead); - cFYI(1, "Clearing Mid 0x%x - issuing callback", - mid_entry->mid); + cFYI(1, "Clearing mid 0x%x", mid_entry->mid); + mid_entry->midState = MID_SHUTDOWN; + list_move(&mid_entry->qhead, &dispose_list); + } + spin_unlock(&GlobalMid_Lock); + + /* now walk dispose list and issue callbacks */ + list_for_each_safe(tmp, tmp2, &dispose_list) { + mid_entry = list_entry(tmp, struct mid_q_entry, qhead); + cFYI(1, "Callback mid 0x%x", mid_entry->mid); list_del_init(&mid_entry->qhead); mid_entry->callback(mid_entry); } - spin_unlock(&GlobalMid_Lock); /* 1/8th of sec is more than enough time for them to exit */ msleep(125); } @@ -805,11 +780,11 @@ extract_hostname(const char *unc) } static int -cifs_parse_mount_options(char *options, const char *devname, +cifs_parse_mount_options(const char *mountdata, const char *devname, struct smb_vol *vol) { - char *value; - char *data; + char *value, *data, *end; + char *mountdata_copy, *options; unsigned int temp_len, i, j; char separator[2]; short int override_uid = -1; @@ -849,9 +824,15 @@ cifs_parse_mount_options(char *options, const char *devname, vol->actimeo = CIFS_DEF_ACTIMEO; - if (!options) - return 1; + if (!mountdata) + goto cifs_parse_mount_err; + mountdata_copy = kstrndup(mountdata, PAGE_SIZE, GFP_KERNEL); + if (!mountdata_copy) + goto cifs_parse_mount_err; + + options = mountdata_copy; + end = options + strlen(options); if (strncmp(options, "sep=", 4) == 0) { if (options[4] != 0) { separator[0] = options[4]; @@ -876,16 +857,22 @@ cifs_parse_mount_options(char *options, const char *devname, if (!value) { printk(KERN_WARNING "CIFS: invalid or missing username\n"); - return 1; /* needs_arg; */ + goto cifs_parse_mount_err; } else if (!*value) { /* null user, ie anonymous, authentication */ vol->nullauth = 1; } - if (strnlen(value, 200) < 200) { - vol->username = value; + if (strnlen(value, MAX_USERNAME_SIZE) < + MAX_USERNAME_SIZE) { + vol->username = kstrdup(value, GFP_KERNEL); + if (!vol->username) { + printk(KERN_WARNING "CIFS: no memory " + "for username\n"); + goto cifs_parse_mount_err; + } } else { printk(KERN_WARNING "CIFS: username too long\n"); - return 1; + goto cifs_parse_mount_err; } } else if (strnicmp(data, "pass", 4) == 0) { if (!value) { @@ -916,6 +903,7 @@ cifs_parse_mount_options(char *options, const char *devname, the only illegal character in a password is null */ if ((value[temp_len] == 0) && + (value + temp_len < end) && (value[temp_len+1] == separator[0])) { /* reinsert comma */ value[temp_len] = separator[0]; @@ -948,7 +936,7 @@ cifs_parse_mount_options(char *options, const char *devname, if (vol->password == NULL) { printk(KERN_WARNING "CIFS: no memory " "for password\n"); - return 1; + goto cifs_parse_mount_err; } for (i = 0, j = 0; i < temp_len; i++, j++) { vol->password[j] = value[i]; @@ -964,7 +952,7 @@ cifs_parse_mount_options(char *options, const char *devname, if (vol->password == NULL) { printk(KERN_WARNING "CIFS: no memory " "for password\n"); - return 1; + goto cifs_parse_mount_err; } strcpy(vol->password, value); } @@ -974,11 +962,16 @@ cifs_parse_mount_options(char *options, const char *devname, vol->UNCip = NULL; } else if (strnlen(value, INET6_ADDRSTRLEN) < INET6_ADDRSTRLEN) { - vol->UNCip = value; + vol->UNCip = kstrdup(value, GFP_KERNEL); + if (!vol->UNCip) { + printk(KERN_WARNING "CIFS: no memory " + "for UNC IP\n"); + goto cifs_parse_mount_err; + } } else { printk(KERN_WARNING "CIFS: ip address " "too long\n"); - return 1; + goto cifs_parse_mount_err; } } else if (strnicmp(data, "sec", 3) == 0) { if (!value || !*value) { @@ -991,7 +984,7 @@ cifs_parse_mount_options(char *options, const char *devname, /* vol->secFlg |= CIFSSEC_MUST_SEAL | CIFSSEC_MAY_KRB5; */ cERROR(1, "Krb5 cifs privacy not supported"); - return 1; + goto cifs_parse_mount_err; } else if (strnicmp(value, "krb5", 4) == 0) { vol->secFlg |= CIFSSEC_MAY_KRB5; } else if (strnicmp(value, "ntlmsspi", 8) == 0) { @@ -1021,7 +1014,16 @@ cifs_parse_mount_options(char *options, const char *devname, vol->nullauth = 1; } else { cERROR(1, "bad security option: %s", value); - return 1; + goto cifs_parse_mount_err; + } + } else if (strnicmp(data, "vers", 3) == 0) { + if (!value || !*value) { + cERROR(1, "no protocol version specified" + " after vers= mount option"); + } else if ((strnicmp(value, "cifs", 4) == 0) || + (strnicmp(value, "1", 1) == 0)) { + /* this is the default */ + continue; } } else if ((strnicmp(data, "unc", 3) == 0) || (strnicmp(data, "target", 6) == 0) @@ -1029,12 +1031,12 @@ cifs_parse_mount_options(char *options, const char *devname, if (!value || !*value) { printk(KERN_WARNING "CIFS: invalid path to " "network resource\n"); - return 1; /* needs_arg; */ + goto cifs_parse_mount_err; } if ((temp_len = strnlen(value, 300)) < 300) { vol->UNC = kmalloc(temp_len+1, GFP_KERNEL); if (vol->UNC == NULL) - return 1; + goto cifs_parse_mount_err; strcpy(vol->UNC, value); if (strncmp(vol->UNC, "//", 2) == 0) { vol->UNC[0] = '\\'; @@ -1043,27 +1045,32 @@ cifs_parse_mount_options(char *options, const char *devname, printk(KERN_WARNING "CIFS: UNC Path does not begin " "with // or \\\\ \n"); - return 1; + goto cifs_parse_mount_err; } } else { printk(KERN_WARNING "CIFS: UNC name too long\n"); - return 1; + goto cifs_parse_mount_err; } } else if ((strnicmp(data, "domain", 3) == 0) || (strnicmp(data, "workgroup", 5) == 0)) { if (!value || !*value) { printk(KERN_WARNING "CIFS: invalid domain name\n"); - return 1; /* needs_arg; */ + goto cifs_parse_mount_err; } /* BB are there cases in which a comma can be valid in a domain name and need special handling? */ if (strnlen(value, 256) < 256) { - vol->domainname = value; + vol->domainname = kstrdup(value, GFP_KERNEL); + if (!vol->domainname) { + printk(KERN_WARNING "CIFS: no memory " + "for domainname\n"); + goto cifs_parse_mount_err; + } cFYI(1, "Domain name set"); } else { printk(KERN_WARNING "CIFS: domain name too " "long\n"); - return 1; + goto cifs_parse_mount_err; } } else if (strnicmp(data, "srcaddr", 7) == 0) { vol->srcaddr.ss_family = AF_UNSPEC; @@ -1071,7 +1078,7 @@ cifs_parse_mount_options(char *options, const char *devname, if (!value || !*value) { printk(KERN_WARNING "CIFS: srcaddr value" " not specified.\n"); - return 1; /* needs_arg; */ + goto cifs_parse_mount_err; } i = cifs_convert_address((struct sockaddr *)&vol->srcaddr, value, strlen(value)); @@ -1079,20 +1086,20 @@ cifs_parse_mount_options(char *options, const char *devname, printk(KERN_WARNING "CIFS: Could not parse" " srcaddr: %s\n", value); - return 1; + goto cifs_parse_mount_err; } } else if (strnicmp(data, "prefixpath", 10) == 0) { if (!value || !*value) { printk(KERN_WARNING "CIFS: invalid path prefix\n"); - return 1; /* needs_argument */ + goto cifs_parse_mount_err; } if ((temp_len = strnlen(value, 1024)) < 1024) { if (value[0] != '/') temp_len++; /* missing leading slash */ vol->prepath = kmalloc(temp_len+1, GFP_KERNEL); if (vol->prepath == NULL) - return 1; + goto cifs_parse_mount_err; if (value[0] != '/') { vol->prepath[0] = '/'; strcpy(vol->prepath+1, value); @@ -1101,24 +1108,33 @@ cifs_parse_mount_options(char *options, const char *devname, cFYI(1, "prefix path %s", vol->prepath); } else { printk(KERN_WARNING "CIFS: prefix too long\n"); - return 1; + goto cifs_parse_mount_err; } } else if (strnicmp(data, "iocharset", 9) == 0) { if (!value || !*value) { printk(KERN_WARNING "CIFS: invalid iocharset " "specified\n"); - return 1; /* needs_arg; */ + goto cifs_parse_mount_err; } if (strnlen(value, 65) < 65) { - if (strnicmp(value, "default", 7)) - vol->iocharset = value; + if (strnicmp(value, "default", 7)) { + vol->iocharset = kstrdup(value, + GFP_KERNEL); + + if (!vol->iocharset) { + printk(KERN_WARNING "CIFS: no " + "memory for" + "charset\n"); + goto cifs_parse_mount_err; + } + } /* if iocharset not set then load_nls_default is used by caller */ cFYI(1, "iocharset set to %s", value); } else { printk(KERN_WARNING "CIFS: iocharset name " "too long.\n"); - return 1; + goto cifs_parse_mount_err; } } else if (!strnicmp(data, "uid", 3) && value && *value) { vol->linux_uid = simple_strtoul(value, &value, 0); @@ -1231,7 +1247,7 @@ cifs_parse_mount_options(char *options, const char *devname, if (vol->actimeo > CIFS_MAX_ACTIMEO) { cERROR(1, "CIFS: attribute cache" "timeout too large"); - return 1; + goto cifs_parse_mount_err; } } } else if (strnicmp(data, "credentials", 4) == 0) { @@ -1343,6 +1359,8 @@ cifs_parse_mount_options(char *options, const char *devname, vol->server_ino = 1; } else if (strnicmp(data, "noserverino", 9) == 0) { vol->server_ino = 0; + } else if (strnicmp(data, "rwpidforward", 4) == 0) { + vol->rwpidforward = 1; } else if (strnicmp(data, "cifsacl", 7) == 0) { vol->cifs_acl = 1; } else if (strnicmp(data, "nocifsacl", 9) == 0) { @@ -1375,7 +1393,7 @@ cifs_parse_mount_options(char *options, const char *devname, #ifndef CONFIG_CIFS_FSCACHE cERROR(1, "FS-Cache support needs CONFIG_CIFS_FSCACHE" "kernel config option set"); - return 1; + goto cifs_parse_mount_err; #endif vol->fsc = true; } else if (strnicmp(data, "mfsymlinks", 10) == 0) { @@ -1390,12 +1408,12 @@ cifs_parse_mount_options(char *options, const char *devname, if (devname == NULL) { printk(KERN_WARNING "CIFS: Missing UNC name for mount " "target\n"); - return 1; + goto cifs_parse_mount_err; } if ((temp_len = strnlen(devname, 300)) < 300) { vol->UNC = kmalloc(temp_len+1, GFP_KERNEL); if (vol->UNC == NULL) - return 1; + goto cifs_parse_mount_err; strcpy(vol->UNC, devname); if (strncmp(vol->UNC, "//", 2) == 0) { vol->UNC[0] = '\\'; @@ -1403,21 +1421,21 @@ cifs_parse_mount_options(char *options, const char *devname, } else if (strncmp(vol->UNC, "\\\\", 2) != 0) { printk(KERN_WARNING "CIFS: UNC Path does not " "begin with // or \\\\ \n"); - return 1; + goto cifs_parse_mount_err; } value = strpbrk(vol->UNC+2, "/\\"); if (value) *value = '\\'; } else { printk(KERN_WARNING "CIFS: UNC name too long\n"); - return 1; + goto cifs_parse_mount_err; } } if (vol->multiuser && !(vol->secFlg & CIFSSEC_MAY_KRB5)) { cERROR(1, "Multiuser mounts currently require krb5 " "authentication!"); - return 1; + goto cifs_parse_mount_err; } if (vol->UNCip == NULL) @@ -1435,7 +1453,12 @@ cifs_parse_mount_options(char *options, const char *devname, printk(KERN_NOTICE "CIFS: ignoring forcegid mount option " "specified with no gid= option.\n"); + kfree(mountdata_copy); return 0; + +cifs_parse_mount_err: + kfree(mountdata_copy); + return 1; } /** Returns true if srcaddr isn't specified and rhs isn't @@ -1472,7 +1495,7 @@ srcip_matches(struct sockaddr *srcaddr, struct sockaddr *rhs) static bool match_port(struct TCP_Server_Info *server, struct sockaddr *addr) { - unsigned short int port, *sport; + __be16 port, *sport; switch (addr->sa_family) { case AF_INET: @@ -1572,18 +1595,37 @@ match_security(struct TCP_Server_Info *server, struct smb_vol *vol) return false; } - /* now check if signing mode is acceptible */ + /* now check if signing mode is acceptable */ if ((secFlags & CIFSSEC_MAY_SIGN) == 0 && - (server->secMode & SECMODE_SIGN_REQUIRED)) + (server->sec_mode & SECMODE_SIGN_REQUIRED)) return false; else if (((secFlags & CIFSSEC_MUST_SIGN) == CIFSSEC_MUST_SIGN) && - (server->secMode & + (server->sec_mode & (SECMODE_SIGN_ENABLED|SECMODE_SIGN_REQUIRED)) == 0) return false; return true; } +static int match_server(struct TCP_Server_Info *server, struct sockaddr *addr, + struct smb_vol *vol) +{ + if (!net_eq(cifs_net_ns(server), current->nsproxy->net_ns)) + return 0; + + if (!match_address(server, addr, + (struct sockaddr *)&vol->srcaddr)) + return 0; + + if (!match_port(server, addr)) + return 0; + + if (!match_security(server, vol)) + return 0; + + return 1; +} + static struct TCP_Server_Info * cifs_find_tcp_session(struct sockaddr *addr, struct smb_vol *vol) { @@ -1591,17 +1633,7 @@ cifs_find_tcp_session(struct sockaddr *addr, struct smb_vol *vol) spin_lock(&cifs_tcp_ses_lock); list_for_each_entry(server, &cifs_tcp_ses_list, tcp_ses_list) { - if (!net_eq(cifs_net_ns(server), current->nsproxy->net_ns)) - continue; - - if (!match_address(server, addr, - (struct sockaddr *)&vol->srcaddr)) - continue; - - if (!match_port(server, addr)) - continue; - - if (!match_security(server, vol)) + if (!match_server(server, addr, vol)) continue; ++server->srv_count; @@ -1765,6 +1797,7 @@ cifs_get_tcp_session(struct smb_vol *volume_info) module_put(THIS_MODULE); goto out_err_crypto_release; } + tcp_ses->tcpStatus = CifsNeedNegotiate; /* thread spawned, put it on the list */ spin_lock(&cifs_tcp_ses_lock); @@ -1794,30 +1827,39 @@ out_err: return ERR_PTR(rc); } -static struct cifsSesInfo * +static int match_session(struct cifs_ses *ses, struct smb_vol *vol) +{ + switch (ses->server->secType) { + case Kerberos: + if (vol->cred_uid != ses->cred_uid) + return 0; + break; + default: + /* anything else takes username/password */ + if (ses->user_name == NULL) + return 0; + if (strncmp(ses->user_name, vol->username, + MAX_USERNAME_SIZE)) + return 0; + if (strlen(vol->username) != 0 && + ses->password != NULL && + strncmp(ses->password, + vol->password ? vol->password : "", + MAX_PASSWORD_SIZE)) + return 0; + } + return 1; +} + +static struct cifs_ses * cifs_find_smb_ses(struct TCP_Server_Info *server, struct smb_vol *vol) { - struct cifsSesInfo *ses; + struct cifs_ses *ses; spin_lock(&cifs_tcp_ses_lock); list_for_each_entry(ses, &server->smb_ses_list, smb_ses_list) { - switch (server->secType) { - case Kerberos: - if (vol->cred_uid != ses->cred_uid) - continue; - break; - default: - /* anything else takes username/password */ - if (strncmp(ses->userName, vol->username, - MAX_USERNAME_SIZE)) - continue; - if (strlen(vol->username) != 0 && - ses->password != NULL && - strncmp(ses->password, - vol->password ? vol->password : "", - MAX_PASSWORD_SIZE)) - continue; - } + if (!match_session(ses, vol)) + continue; ++ses->ses_count; spin_unlock(&cifs_tcp_ses_lock); return ses; @@ -1827,7 +1869,7 @@ cifs_find_smb_ses(struct TCP_Server_Info *server, struct smb_vol *vol) } static void -cifs_put_smb_ses(struct cifsSesInfo *ses) +cifs_put_smb_ses(struct cifs_ses *ses) { int xid; struct TCP_Server_Info *server = ses->server; @@ -1851,11 +1893,13 @@ cifs_put_smb_ses(struct cifsSesInfo *ses) cifs_put_tcp_session(server); } -static struct cifsSesInfo * +static bool warned_on_ntlm; /* globals init to false automatically */ + +static struct cifs_ses * cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) { int rc = -ENOMEM, xid; - struct cifsSesInfo *ses; + struct cifs_ses *ses; struct sockaddr_in *addr = (struct sockaddr_in *)&server->dstaddr; struct sockaddr_in6 *addr6 = (struct sockaddr_in6 *)&server->dstaddr; @@ -1906,9 +1950,11 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) else sprintf(ses->serverName, "%pI4", &addr->sin_addr); - if (volume_info->username) - strncpy(ses->userName, volume_info->username, - MAX_USERNAME_SIZE); + if (volume_info->username) { + ses->user_name = kstrdup(volume_info->username, GFP_KERNEL); + if (!ses->user_name) + goto get_ses_fail; + } /* volume_info->password freed at unmount */ if (volume_info->password) { @@ -1923,6 +1969,15 @@ cifs_get_smb_ses(struct TCP_Server_Info *server, struct smb_vol *volume_info) } ses->cred_uid = volume_info->cred_uid; ses->linux_uid = volume_info->linux_uid; + + /* ntlmv2 is much stronger than ntlm security, and has been broadly + supported for many years, time to update default security mechanism */ + if ((volume_info->secFlg == 0) && warned_on_ntlm == false) { + warned_on_ntlm = true; + cERROR(1, "default security mechanism requested. The default " + "security mechanism will be upgraded from ntlm to " + "ntlmv2 in kernel release 2.6.41"); + } ses->overrideSecFlg = volume_info->secFlg; mutex_lock(&ses->session_mutex); @@ -1947,20 +2002,26 @@ get_ses_fail: return ERR_PTR(rc); } -static struct cifsTconInfo * -cifs_find_tcon(struct cifsSesInfo *ses, const char *unc) +static int match_tcon(struct cifs_tcon *tcon, const char *unc) +{ + if (tcon->tidStatus == CifsExiting) + return 0; + if (strncmp(tcon->treeName, unc, MAX_TREE_SIZE)) + return 0; + return 1; +} + +static struct cifs_tcon * +cifs_find_tcon(struct cifs_ses *ses, const char *unc) { struct list_head *tmp; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; spin_lock(&cifs_tcp_ses_lock); list_for_each(tmp, &ses->tcon_list) { - tcon = list_entry(tmp, struct cifsTconInfo, tcon_list); - if (tcon->tidStatus == CifsExiting) + tcon = list_entry(tmp, struct cifs_tcon, tcon_list); + if (!match_tcon(tcon, unc)) continue; - if (strncmp(tcon->treeName, unc, MAX_TREE_SIZE)) - continue; - ++tcon->tc_count; spin_unlock(&cifs_tcp_ses_lock); return tcon; @@ -1970,10 +2031,10 @@ cifs_find_tcon(struct cifsSesInfo *ses, const char *unc) } static void -cifs_put_tcon(struct cifsTconInfo *tcon) +cifs_put_tcon(struct cifs_tcon *tcon) { int xid; - struct cifsSesInfo *ses = tcon->ses; + struct cifs_ses *ses = tcon->ses; cFYI(1, "%s: tc_count=%d\n", __func__, tcon->tc_count); spin_lock(&cifs_tcp_ses_lock); @@ -1994,11 +2055,11 @@ cifs_put_tcon(struct cifsTconInfo *tcon) cifs_put_smb_ses(ses); } -static struct cifsTconInfo * -cifs_get_tcon(struct cifsSesInfo *ses, struct smb_vol *volume_info) +static struct cifs_tcon * +cifs_get_tcon(struct cifs_ses *ses, struct smb_vol *volume_info) { int rc, xid; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; tcon = cifs_find_tcon(ses, volume_info->UNC); if (tcon) { @@ -2087,8 +2148,102 @@ cifs_put_tlink(struct tcon_link *tlink) return; } +static inline struct tcon_link * +cifs_sb_master_tlink(struct cifs_sb_info *cifs_sb); + +static int +compare_mount_options(struct super_block *sb, struct cifs_mnt_data *mnt_data) +{ + struct cifs_sb_info *old = CIFS_SB(sb); + struct cifs_sb_info *new = mnt_data->cifs_sb; + + if ((sb->s_flags & CIFS_MS_MASK) != (mnt_data->flags & CIFS_MS_MASK)) + return 0; + + if ((old->mnt_cifs_flags & CIFS_MOUNT_MASK) != + (new->mnt_cifs_flags & CIFS_MOUNT_MASK)) + return 0; + + if (old->rsize != new->rsize) + return 0; + + /* + * We want to share sb only if we don't specify wsize or specified wsize + * is greater or equal than existing one. + */ + if (new->wsize && new->wsize < old->wsize) + return 0; + + if (old->mnt_uid != new->mnt_uid || old->mnt_gid != new->mnt_gid) + return 0; + + if (old->mnt_file_mode != new->mnt_file_mode || + old->mnt_dir_mode != new->mnt_dir_mode) + return 0; + + if (strcmp(old->local_nls->charset, new->local_nls->charset)) + return 0; + + if (old->actimeo != new->actimeo) + return 0; + + return 1; +} + int -get_dfs_path(int xid, struct cifsSesInfo *pSesInfo, const char *old_path, +cifs_match_super(struct super_block *sb, void *data) +{ + struct cifs_mnt_data *mnt_data = (struct cifs_mnt_data *)data; + struct smb_vol *volume_info; + struct cifs_sb_info *cifs_sb; + struct TCP_Server_Info *tcp_srv; + struct cifs_ses *ses; + struct cifs_tcon *tcon; + struct tcon_link *tlink; + struct sockaddr_storage addr; + int rc = 0; + + memset(&addr, 0, sizeof(struct sockaddr_storage)); + + spin_lock(&cifs_tcp_ses_lock); + cifs_sb = CIFS_SB(sb); + tlink = cifs_get_tlink(cifs_sb_master_tlink(cifs_sb)); + if (IS_ERR(tlink)) { + spin_unlock(&cifs_tcp_ses_lock); + return rc; + } + tcon = tlink_tcon(tlink); + ses = tcon->ses; + tcp_srv = ses->server; + + volume_info = mnt_data->vol; + + if (!volume_info->UNCip || !volume_info->UNC) + goto out; + + rc = cifs_fill_sockaddr((struct sockaddr *)&addr, + volume_info->UNCip, + strlen(volume_info->UNCip), + volume_info->port); + if (!rc) + goto out; + + if (!match_server(tcp_srv, (struct sockaddr *)&addr, volume_info) || + !match_session(ses, volume_info) || + !match_tcon(tcon, volume_info->UNC)) { + rc = 0; + goto out; + } + + rc = compare_mount_options(sb, mnt_data); +out: + cifs_put_tlink(tlink); + spin_unlock(&cifs_tcp_ses_lock); + return rc; +} + +int +get_dfs_path(int xid, struct cifs_ses *pSesInfo, const char *old_path, const struct nls_table *nls_codepage, unsigned int *pnum_referrals, struct dfs_info3_param **preferrals, int remap) { @@ -2249,7 +2404,7 @@ ip_rfc1001_connect(struct TCP_Server_Info *server) smb_buf = (struct smb_hdr *)ses_init_buf; /* sizeof RFC1002_SESSION_REQUEST with no scope */ - smb_buf->smb_buf_length = 0x81000044; + smb_buf->smb_buf_length = cpu_to_be32(0x81000044); rc = smb_send(server, smb_buf, 0x44); kfree(ses_init_buf); /* @@ -2276,7 +2431,7 @@ static int generic_ip_connect(struct TCP_Server_Info *server) { int rc = 0; - unsigned short int sport; + __be16 sport; int slen, sfamily; struct socket *socket = server->ssocket; struct sockaddr *saddr; @@ -2361,7 +2516,7 @@ generic_ip_connect(struct TCP_Server_Info *server) static int ip_connect(struct TCP_Server_Info *server) { - unsigned short int *sport; + __be16 *sport; struct sockaddr_in6 *addr6 = (struct sockaddr_in6 *)&server->dstaddr; struct sockaddr_in *addr = (struct sockaddr_in *)&server->dstaddr; @@ -2387,7 +2542,7 @@ ip_connect(struct TCP_Server_Info *server) return generic_ip_connect(server); } -void reset_cifs_unix_caps(int xid, struct cifsTconInfo *tcon, +void reset_cifs_unix_caps(int xid, struct cifs_tcon *tcon, struct super_block *sb, struct smb_vol *vol_info) { /* if we are reconnecting then should we check to see if @@ -2416,7 +2571,7 @@ void reset_cifs_unix_caps(int xid, struct cifsTconInfo *tcon, if (!CIFSSMBQFSUnixInfo(xid, tcon)) { __u64 cap = le64_to_cpu(tcon->fsUnixInfo.Capability); - + cFYI(1, "unix caps which server supports %lld", cap); /* check for reconnect case in which we do not want to change the mount behavior if we can avoid it */ if (vol_info == NULL) { @@ -2434,6 +2589,9 @@ void reset_cifs_unix_caps(int xid, struct cifsTconInfo *tcon, } } + if (cap & CIFS_UNIX_TRANSPORT_ENCRYPTION_MANDATORY_CAP) + cERROR(1, "per-share encryption not supported yet"); + cap &= CIFS_UNIX_CAP_MASK; if (vol_info && vol_info->no_psx_acl) cap &= ~CIFS_UNIX_POSIX_ACL_CAP; @@ -2452,12 +2610,6 @@ void reset_cifs_unix_caps(int xid, struct cifsTconInfo *tcon, CIFS_MOUNT_POSIX_PATHS; } - /* We might be setting the path sep back to a different - form if we are reconnecting and the server switched its - posix path capability for this share */ - if (sb && (CIFS_SB(sb)->prepathlen > 0)) - CIFS_SB(sb)->prepath[0] = CIFS_DIR_SEP(CIFS_SB(sb)); - if (sb && (CIFS_SB(sb)->rsize > 127 * 1024)) { if ((cap & CIFS_UNIX_LARGE_READ_CAP) == 0) { CIFS_SB(sb)->rsize = 127 * 1024; @@ -2482,6 +2634,10 @@ void reset_cifs_unix_caps(int xid, struct cifsTconInfo *tcon, cFYI(1, "very large read cap"); if (cap & CIFS_UNIX_LARGE_WRITE_CAP) cFYI(1, "very large write cap"); + if (cap & CIFS_UNIX_TRANSPORT_ENCRYPTION_CAP) + cFYI(1, "transport encryption cap"); + if (cap & CIFS_UNIX_TRANSPORT_ENCRYPTION_MANDATORY_CAP) + cFYI(1, "mandatory transport encryption cap"); #endif /* CIFS_DEBUG2 */ if (CIFSSMBSetFSUnixInfo(xid, tcon, cap)) { if (vol_info == NULL) { @@ -2498,28 +2654,8 @@ void reset_cifs_unix_caps(int xid, struct cifsTconInfo *tcon, } } -static void -convert_delimiter(char *path, char delim) -{ - int i; - char old_delim; - - if (path == NULL) - return; - - if (delim == '/') - old_delim = '\\'; - else - old_delim = '/'; - - for (i = 0; path[i] != '\0'; i++) { - if (path[i] == old_delim) - path[i] = delim; - } -} - -static void setup_cifs_sb(struct smb_vol *pvolume_info, - struct cifs_sb_info *cifs_sb) +void cifs_setup_cifs_sb(struct smb_vol *pvolume_info, + struct cifs_sb_info *cifs_sb) { INIT_DELAYED_WORK(&cifs_sb->prune_tlinks, cifs_prune_tlinks); @@ -2533,40 +2669,19 @@ static void setup_cifs_sb(struct smb_vol *pvolume_info, else /* default */ cifs_sb->rsize = CIFSMaxBufSize; - if (pvolume_info->wsize > PAGEVEC_SIZE * PAGE_CACHE_SIZE) { - cERROR(1, "wsize %d too large, using 4096 instead", - pvolume_info->wsize); - cifs_sb->wsize = 4096; - } else if (pvolume_info->wsize) - cifs_sb->wsize = pvolume_info->wsize; - else - cifs_sb->wsize = min_t(const int, - PAGEVEC_SIZE * PAGE_CACHE_SIZE, - 127*1024); - /* old default of CIFSMaxBufSize was too small now - that SMB Write2 can send multiple pages in kvec. - RFC1001 does not describe what happens when frame - bigger than 128K is sent so use that as max in - conjunction with 52K kvec constraint on arch with 4K - page size */ - if (cifs_sb->rsize < 2048) { cifs_sb->rsize = 2048; /* Windows ME may prefer this */ cFYI(1, "readsize set to minimum: 2048"); } - /* calculate prepath */ - cifs_sb->prepath = pvolume_info->prepath; - if (cifs_sb->prepath) { - cifs_sb->prepathlen = strlen(cifs_sb->prepath); - /* we can not convert the / to \ in the path - separators in the prefixpath yet because we do not - know (until reset_cifs_unix_caps is called later) - whether POSIX PATH CAP is available. We normalize - the / to \ after reset_cifs_unix_caps is called */ - pvolume_info->prepath = NULL; - } else - cifs_sb->prepathlen = 0; + + /* + * Temporarily set wsize for matching superblock. If we end up using + * new sb then cifs_negotiate_wsize will later negotiate it downward + * if needed. + */ + cifs_sb->wsize = pvolume_info->wsize; + cifs_sb->mnt_uid = pvolume_info->linux_uid; cifs_sb->mnt_gid = pvolume_info->linux_gid; cifs_sb->mnt_file_mode = pvolume_info->file_mode; @@ -2575,6 +2690,7 @@ static void setup_cifs_sb(struct smb_vol *pvolume_info, cifs_sb->mnt_file_mode, cifs_sb->mnt_dir_mode); cifs_sb->actimeo = pvolume_info->actimeo; + cifs_sb->local_nls = pvolume_info->local_nls; if (pvolume_info->noperm) cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NO_PERM; @@ -2594,6 +2710,8 @@ static void setup_cifs_sb(struct smb_vol *pvolume_info, cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NOSSYNC; if (pvolume_info->mand_lock) cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_NOPOSIXBRL; + if (pvolume_info->rwpidforward) + cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_RWPIDFORWARD; if (pvolume_info->cifs_acl) cifs_sb->mnt_cifs_flags |= CIFS_MOUNT_CIFS_ACL; if (pvolume_info->override_uid) @@ -2627,8 +2745,55 @@ static void setup_cifs_sb(struct smb_vol *pvolume_info, "mount option supported"); } +/* + * When the server supports very large writes via POSIX extensions, we can + * allow up to 2^24 - PAGE_CACHE_SIZE. + * + * Note that this might make for "interesting" allocation problems during + * writeback however (as we have to allocate an array of pointers for the + * pages). A 16M write means ~32kb page array with PAGE_CACHE_SIZE == 4096. + */ +#define CIFS_MAX_WSIZE ((1<<24) - PAGE_CACHE_SIZE) + +/* + * When the server doesn't allow large posix writes, default to a wsize of + * 128k - PAGE_CACHE_SIZE -- one page less than the largest frame size + * described in RFC1001. This allows space for the header without going over + * that by default. + */ +#define CIFS_MAX_RFC1001_WSIZE (128 * 1024 - PAGE_CACHE_SIZE) + +/* + * The default wsize is 1M. find_get_pages seems to return a maximum of 256 + * pages in a single call. With PAGE_CACHE_SIZE == 4k, this means we can fill + * a single wsize request with a single call. + */ +#define CIFS_DEFAULT_WSIZE (1024 * 1024) + +static unsigned int +cifs_negotiate_wsize(struct cifs_tcon *tcon, struct smb_vol *pvolume_info) +{ + __u64 unix_cap = le64_to_cpu(tcon->fsUnixInfo.Capability); + struct TCP_Server_Info *server = tcon->ses->server; + unsigned int wsize = pvolume_info->wsize ? pvolume_info->wsize : + CIFS_DEFAULT_WSIZE; + + /* can server support 24-bit write sizes? (via UNIX extensions) */ + if (!tcon->unix_ext || !(unix_cap & CIFS_UNIX_LARGE_WRITE_CAP)) + wsize = min_t(unsigned int, wsize, CIFS_MAX_RFC1001_WSIZE); + + /* no CAP_LARGE_WRITE_X? Limit it to 16 bits */ + if (!(server->capabilities & CAP_LARGE_WRITE_X)) + wsize = min_t(unsigned int, wsize, USHRT_MAX); + + /* hard limit of CIFS_MAX_WSIZE */ + wsize = min_t(unsigned int, wsize, CIFS_MAX_WSIZE); + + return wsize; +} + static int -is_path_accessible(int xid, struct cifsTconInfo *tcon, +is_path_accessible(int xid, struct cifs_tcon *tcon, struct cifs_sb_info *cifs_sb, const char *full_path) { int rc; @@ -2642,12 +2807,17 @@ is_path_accessible(int xid, struct cifsTconInfo *tcon, 0 /* not legacy */, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + + if (rc == -EOPNOTSUPP || rc == -EINVAL) + rc = SMBQueryInformation(xid, tcon, full_path, pfile_info, + cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & + CIFS_MOUNT_MAP_SPECIAL_CHR); kfree(pfile_info); return rc; } -static void -cleanup_volume_info(struct smb_vol **pvolume_info) +void +cifs_cleanup_volume_info(struct smb_vol **pvolume_info) { struct smb_vol *volume_info; @@ -2655,8 +2825,12 @@ cleanup_volume_info(struct smb_vol **pvolume_info) return; volume_info = *pvolume_info; + kfree(volume_info->username); kzfree(volume_info->password); kfree(volume_info->UNC); + kfree(volume_info->UNCip); + kfree(volume_info->domainname); + kfree(volume_info->iocharset); kfree(volume_info->prepath); kfree(volume_info); *pvolume_info = NULL; @@ -2673,55 +2847,78 @@ build_unc_path_to_root(const struct smb_vol *volume_info, char *full_path; int unc_len = strnlen(volume_info->UNC, MAX_TREE_SIZE + 1); - full_path = kmalloc(unc_len + cifs_sb->prepathlen + 1, GFP_KERNEL); + full_path = kmalloc(unc_len + 1, GFP_KERNEL); if (full_path == NULL) return ERR_PTR(-ENOMEM); strncpy(full_path, volume_info->UNC, unc_len); - if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) { - int i; - for (i = 0; i < unc_len; i++) { - if (full_path[i] == '\\') - full_path[i] = '/'; - } - } - - if (cifs_sb->prepathlen) - strncpy(full_path + unc_len, cifs_sb->prepath, - cifs_sb->prepathlen); - - full_path[unc_len + cifs_sb->prepathlen] = 0; /* add trailing null */ + full_path[unc_len] = 0; /* add trailing null */ + convert_delimiter(full_path, CIFS_DIR_SEP(cifs_sb)); return full_path; } -#endif -int -cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, - char *mount_data_global, const char *devname) +/* + * Perform a dfs referral query for a share and (optionally) prefix + * + * If a referral is found, cifs_sb->mountdata will be (re-)allocated + * to a string containing updated options for the submount. Otherwise it + * will be left untouched. + * + * Returns the rc from get_dfs_path to the caller, which can be used to + * determine whether there were referrals. + */ +static int +expand_dfs_referral(int xid, struct cifs_ses *pSesInfo, + struct smb_vol *volume_info, struct cifs_sb_info *cifs_sb, + int check_prefix) { int rc; - int xid; - struct smb_vol *volume_info; - struct cifsSesInfo *pSesInfo; - struct cifsTconInfo *tcon; - struct TCP_Server_Info *srvTcp; - char *full_path; - char *mount_data = mount_data_global; - struct tcon_link *tlink; -#ifdef CONFIG_CIFS_DFS_UPCALL - struct dfs_info3_param *referrals = NULL; unsigned int num_referrals = 0; - int referral_walks_count = 0; -try_mount_again: + struct dfs_info3_param *referrals = NULL; + char *full_path = NULL, *ref_path = NULL, *mdata = NULL; + + full_path = build_unc_path_to_root(volume_info, cifs_sb); + if (IS_ERR(full_path)) + return PTR_ERR(full_path); + + /* For DFS paths, skip the first '\' of the UNC */ + ref_path = check_prefix ? full_path + 1 : volume_info->UNC + 1; + + rc = get_dfs_path(xid, pSesInfo , ref_path, cifs_sb->local_nls, + &num_referrals, &referrals, + cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + + if (!rc && num_referrals > 0) { + char *fake_devname = NULL; + + mdata = cifs_compose_mount_options(cifs_sb->mountdata, + full_path + 1, referrals, + &fake_devname); + + free_dfs_info_array(referrals, num_referrals); + kfree(fake_devname); + + if (cifs_sb->mountdata != NULL) + kfree(cifs_sb->mountdata); + + if (IS_ERR(mdata)) { + rc = PTR_ERR(mdata); + mdata = NULL; + } + cifs_sb->mountdata = mdata; + } + kfree(full_path); + return rc; +} #endif - rc = 0; - tcon = NULL; - pSesInfo = NULL; - srvTcp = NULL; - full_path = NULL; - tlink = NULL; - xid = GetXid(); +int cifs_setup_volume_info(struct smb_vol **pvolume_info, char *mount_data, + const char *devname) +{ + struct smb_vol *volume_info; + int rc = 0; + + *pvolume_info = NULL; volume_info = kzalloc(sizeof(struct smb_vol), GFP_KERNEL); if (!volume_info) { @@ -2729,7 +2926,8 @@ try_mount_again: goto out; } - if (cifs_parse_mount_options(mount_data, devname, volume_info)) { + if (cifs_parse_mount_options(mount_data, devname, + volume_info)) { rc = -EINVAL; goto out; } @@ -2761,7 +2959,46 @@ try_mount_again: goto out; } } - cifs_sb->local_nls = volume_info->local_nls; + + *pvolume_info = volume_info; + return rc; +out: + cifs_cleanup_volume_info(&volume_info); + return rc; +} + +int +cifs_mount(struct super_block *sb, struct cifs_sb_info *cifs_sb, + struct smb_vol *volume_info, const char *devname) +{ + int rc = 0; + int xid; + struct cifs_ses *pSesInfo; + struct cifs_tcon *tcon; + struct TCP_Server_Info *srvTcp; + char *full_path; + struct tcon_link *tlink; +#ifdef CONFIG_CIFS_DFS_UPCALL + int referral_walks_count = 0; +try_mount_again: + /* cleanup activities if we're chasing a referral */ + if (referral_walks_count) { + if (tcon) + cifs_put_tcon(tcon); + else if (pSesInfo) + cifs_put_smb_ses(pSesInfo); + + cifs_cleanup_volume_info(&volume_info); + FreeXid(xid); + } +#endif + tcon = NULL; + pSesInfo = NULL; + srvTcp = NULL; + full_path = NULL; + tlink = NULL; + + xid = GetXid(); /* get a reference to a tcp session */ srvTcp = cifs_get_tcp_session(volume_info); @@ -2778,7 +3015,6 @@ try_mount_again: goto mount_fail_check; } - setup_cifs_sb(volume_info, cifs_sb); if (pSesInfo->capabilities & CAP_LARGE_FILES) sb->s_maxbytes = MAX_LFS_FILESIZE; else @@ -2795,40 +3031,59 @@ try_mount_again: goto remote_path_check; } - /* do not care if following two calls succeed - informational */ - if (!tcon->ipc) { - CIFSSMBQFSDeviceInfo(xid, tcon); - CIFSSMBQFSAttributeInfo(xid, tcon); - } - /* tell server which Unix caps we support */ - if (tcon->ses->capabilities & CAP_UNIX) + if (tcon->ses->capabilities & CAP_UNIX) { /* reset of caps checks mount to see if unix extensions disabled for just this mount */ reset_cifs_unix_caps(xid, tcon, sb, volume_info); - else + if ((tcon->ses->server->tcpStatus == CifsNeedReconnect) && + (le64_to_cpu(tcon->fsUnixInfo.Capability) & + CIFS_UNIX_TRANSPORT_ENCRYPTION_MANDATORY_CAP)) { + rc = -EACCES; + goto mount_fail_check; + } + } else tcon->unix_ext = 0; /* server does not support them */ - /* convert forward to back slashes in prepath here if needed */ - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) == 0) - convert_delimiter(cifs_sb->prepath, CIFS_DIR_SEP(cifs_sb)); + /* do not care if following two calls succeed - informational */ + if (!tcon->ipc) { + CIFSSMBQFSDeviceInfo(xid, tcon); + CIFSSMBQFSAttributeInfo(xid, tcon); + } if ((tcon->unix_ext == 0) && (cifs_sb->rsize > (1024 * 127))) { cifs_sb->rsize = 1024 * 127; cFYI(DBG2, "no very large read support, rsize now 127K"); } - if (!(tcon->ses->capabilities & CAP_LARGE_WRITE_X)) - cifs_sb->wsize = min(cifs_sb->wsize, - (tcon->ses->server->maxBuf - MAX_CIFS_HDR_SIZE)); if (!(tcon->ses->capabilities & CAP_LARGE_READ_X)) cifs_sb->rsize = min(cifs_sb->rsize, (tcon->ses->server->maxBuf - MAX_CIFS_HDR_SIZE)); + cifs_sb->wsize = cifs_negotiate_wsize(tcon, volume_info); + remote_path_check: - /* check if a whole path (including prepath) is not remote */ - if (!rc && cifs_sb->prepathlen && tcon) { +#ifdef CONFIG_CIFS_DFS_UPCALL + /* + * Perform an unconditional check for whether there are DFS + * referrals for this path without prefix, to provide support + * for DFS referrals from w2k8 servers which don't seem to respond + * with PATH_NOT_COVERED to requests that include the prefix. + * Chase the referral if found, otherwise continue normally. + */ + if (referral_walks_count == 0) { + int refrc = expand_dfs_referral(xid, pSesInfo, volume_info, + cifs_sb, false); + if (!refrc) { + referral_walks_count++; + goto try_mount_again; + } + } +#endif + + /* check if a whole path is not remote */ + if (!rc && tcon) { /* build_path_to_root works only when we have a valid tcon */ - full_path = cifs_build_path_to_root(cifs_sb, tcon); + full_path = cifs_build_path_to_root(volume_info, cifs_sb, tcon); if (full_path == NULL) { rc = -ENOMEM; goto mount_fail_check; @@ -2854,50 +3109,15 @@ remote_path_check: rc = -ELOOP; goto mount_fail_check; } - /* convert forward to back slashes in prepath here if needed */ - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_POSIX_PATHS) == 0) - convert_delimiter(cifs_sb->prepath, - CIFS_DIR_SEP(cifs_sb)); - full_path = build_unc_path_to_root(volume_info, cifs_sb); - if (IS_ERR(full_path)) { - rc = PTR_ERR(full_path); - goto mount_fail_check; - } - - cFYI(1, "Getting referral for: %s", full_path); - rc = get_dfs_path(xid, pSesInfo , full_path + 1, - cifs_sb->local_nls, &num_referrals, &referrals, - cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); - if (!rc && num_referrals > 0) { - char *fake_devname = NULL; - - if (mount_data != mount_data_global) - kfree(mount_data); - mount_data = cifs_compose_mount_options( - cifs_sb->mountdata, full_path + 1, - referrals, &fake_devname); + rc = expand_dfs_referral(xid, pSesInfo, volume_info, cifs_sb, + true); - free_dfs_info_array(referrals, num_referrals); - kfree(fake_devname); - kfree(full_path); - - if (IS_ERR(mount_data)) { - rc = PTR_ERR(mount_data); - mount_data = NULL; - goto mount_fail_check; - } - - if (tcon) - cifs_put_tcon(tcon); - else if (pSesInfo) - cifs_put_smb_ses(pSesInfo); - - cleanup_volume_info(&volume_info); + if (!rc) { referral_walks_count++; - FreeXid(xid); goto try_mount_again; } + goto mount_fail_check; #else /* No DFS support, return error on mount */ rc = -EOPNOTSUPP; #endif @@ -2930,10 +3150,8 @@ remote_path_check: mount_fail_check: /* on error free sesinfo and tcon struct if needed */ if (rc) { - if (mount_data != mount_data_global) - kfree(mount_data); /* If find_unc succeeded then rc == 0 so we can not end */ - /* up accidently freeing someone elses tcon struct */ + /* up accidentally freeing someone elses tcon struct */ if (tcon) cifs_put_tcon(tcon); else if (pSesInfo) @@ -2949,14 +3167,13 @@ mount_fail_check: password will be freed at unmount time) */ out: /* zero out password before freeing */ - cleanup_volume_info(&volume_info); FreeXid(xid); return rc; } int -CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, - const char *tree, struct cifsTconInfo *tcon, +CIFSTCon(unsigned int xid, struct cifs_ses *ses, + const char *tree, struct cifs_tcon *tcon, const struct nls_table *nls_codepage) { struct smb_hdr *smb_buffer; @@ -2988,7 +3205,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, pSMB->AndXCommand = 0xFF; pSMB->Flags = cpu_to_le16(TCON_EXTENDED_SECINFO); bcc_ptr = &pSMB->Password[0]; - if ((ses->server->secMode) & SECMODE_USER) { + if ((ses->server->sec_mode) & SECMODE_USER) { pSMB->PasswordLength = cpu_to_le16(1); /* minimum */ *bcc_ptr = 0; /* password is null byte */ bcc_ptr++; /* skip password */ @@ -3005,7 +3222,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, if ((global_secflags & CIFSSEC_MAY_LANMAN) && (ses->server->secType == LANMAN)) calc_lanman_hash(tcon->password, ses->server->cryptkey, - ses->server->secMode & + ses->server->sec_mode & SECMODE_PW_ENCRYPT ? true : false, bcc_ptr); else @@ -3021,7 +3238,7 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, } } - if (ses->server->secMode & + if (ses->server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) smb_buffer->Flags2 |= SMBFLG2_SECURITY_SIGNATURE; @@ -3047,7 +3264,8 @@ CIFSTCon(unsigned int xid, struct cifsSesInfo *ses, bcc_ptr += strlen("?????"); bcc_ptr += 1; count = bcc_ptr - &pSMB->Password[0]; - pSMB->hdr.smb_buf_length += count; + pSMB->hdr.smb_buf_length = cpu_to_be32(be32_to_cpu( + pSMB->hdr.smb_buf_length) + count); pSMB->ByteCount = cpu_to_le16(count); rc = SendReceive(xid, ses, smb_buffer, smb_buffer_response, &length, @@ -3116,7 +3334,6 @@ cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb) struct rb_root *root = &cifs_sb->tlink_tree; struct rb_node *node; struct tcon_link *tlink; - char *tmp; cancel_delayed_work_sync(&cifs_sb->prune_tlinks); @@ -3133,15 +3350,10 @@ cifs_umount(struct super_block *sb, struct cifs_sb_info *cifs_sb) } spin_unlock(&cifs_sb->tlink_tree_lock); - tmp = cifs_sb->prepath; - cifs_sb->prepathlen = 0; - cifs_sb->prepath = NULL; - kfree(tmp); - return 0; } -int cifs_negotiate_protocol(unsigned int xid, struct cifsSesInfo *ses) +int cifs_negotiate_protocol(unsigned int xid, struct cifs_ses *ses) { int rc = 0; struct TCP_Server_Info *server = ses->server; @@ -3171,7 +3383,7 @@ int cifs_negotiate_protocol(unsigned int xid, struct cifsSesInfo *ses) } -int cifs_setup_session(unsigned int xid, struct cifsSesInfo *ses, +int cifs_setup_session(unsigned int xid, struct cifs_ses *ses, struct nls_table *nls_info) { int rc = 0; @@ -3183,7 +3395,7 @@ int cifs_setup_session(unsigned int xid, struct cifsSesInfo *ses, ses->capabilities &= (~CAP_UNIX); cFYI(1, "Security Mode: 0x%x Capabilities: 0x%x TimeAdjust: %d", - server->secMode, server->capabilities, server->timeAdj); + server->sec_mode, server->capabilities, server->timeAdj); rc = CIFS_SessSetup(xid, ses, nls_info); if (rc) { @@ -3215,14 +3427,16 @@ int cifs_setup_session(unsigned int xid, struct cifsSesInfo *ses, return rc; } -static struct cifsTconInfo * +static struct cifs_tcon * cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid) { - struct cifsTconInfo *master_tcon = cifs_sb_master_tcon(cifs_sb); - struct cifsSesInfo *ses; - struct cifsTconInfo *tcon = NULL; + struct cifs_tcon *master_tcon = cifs_sb_master_tcon(cifs_sb); + struct cifs_ses *ses; + struct cifs_tcon *tcon = NULL; struct smb_vol *vol_info; - char username[MAX_USERNAME_SIZE + 1]; + char username[28]; /* big enough for "krb50x" + hex of ULONG_MAX 6+16 */ + /* We used to have this as MAX_USERNAME which is */ + /* way too big now (256 instead of 32) */ vol_info = kzalloc(sizeof(*vol_info), GFP_KERNEL); if (vol_info == NULL) { @@ -3251,7 +3465,7 @@ cifs_construct_tcon(struct cifs_sb_info *cifs_sb, uid_t fsuid) ses = cifs_get_smb_ses(master_tcon->ses->server, vol_info); if (IS_ERR(ses)) { - tcon = (struct cifsTconInfo *)ses; + tcon = (struct cifs_tcon *)ses; cifs_put_tcp_session(master_tcon->ses->server); goto out; } @@ -3276,7 +3490,7 @@ cifs_sb_master_tlink(struct cifs_sb_info *cifs_sb) return cifs_sb->master_tlink; } -struct cifsTconInfo * +struct cifs_tcon * cifs_sb_master_tcon(struct cifs_sb_info *cifs_sb) { return tlink_tcon(cifs_sb_master_tlink(cifs_sb)); diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index dd5f22918c33..81914df47ef1 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -50,12 +50,11 @@ build_path_from_dentry(struct dentry *direntry) { struct dentry *temp; int namelen; - int pplen; int dfsplen; char *full_path; char dirsep; struct cifs_sb_info *cifs_sb = CIFS_SB(direntry->d_sb); - struct cifsTconInfo *tcon = cifs_sb_master_tcon(cifs_sb); + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); if (direntry == NULL) return NULL; /* not much we can do if dentry is freed and @@ -63,13 +62,12 @@ build_path_from_dentry(struct dentry *direntry) when the server crashed */ dirsep = CIFS_DIR_SEP(cifs_sb); - pplen = cifs_sb->prepathlen; if (tcon->Flags & SMB_SHARE_IS_IN_DFS) dfsplen = strnlen(tcon->treeName, MAX_TREE_SIZE + 1); else dfsplen = 0; cifs_bp_rename_retry: - namelen = pplen + dfsplen; + namelen = dfsplen; for (temp = direntry; !IS_ROOT(temp);) { namelen += (1 + temp->d_name.len); temp = temp->d_parent; @@ -100,7 +98,7 @@ cifs_bp_rename_retry: return NULL; } } - if (namelen != pplen + dfsplen) { + if (namelen != dfsplen) { cERROR(1, "did not end path lookup where expected namelen is %d", namelen); /* presumably this is only possible if racing with a rename @@ -126,7 +124,6 @@ cifs_bp_rename_retry: } } } - strncpy(full_path + dfsplen, CIFS_SB(direntry->d_sb)->prepath, pplen); return full_path; } @@ -152,7 +149,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, __u16 fileHandle; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; char *full_path = NULL; FILE_ALL_INFO *buf = NULL; struct inode *newinode = NULL; @@ -189,7 +186,7 @@ cifs_create(struct inode *inode, struct dentry *direntry, int mode, inode->i_sb, mode, oflags, &oplock, &fileHandle, xid); /* EIO could indicate that (posix open) operation is not supported, despite what server claimed in capability - negotation. EREMOTE indicates DFS junction, which is not + negotiation. EREMOTE indicates DFS junction, which is not handled in posix open */ if (rc == 0) { @@ -356,7 +353,8 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, int mode, int xid; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; + struct cifs_io_parms io_parms; char *full_path = NULL; struct inode *newinode = NULL; int oplock = 0; @@ -439,16 +437,19 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, int mode, * timestamps in, but we can reuse it safely */ pdev = (struct win_dev *)buf; + io_parms.netfid = fileHandle; + io_parms.pid = current->tgid; + io_parms.tcon = pTcon; + io_parms.offset = 0; + io_parms.length = sizeof(struct win_dev); if (S_ISCHR(mode)) { memcpy(pdev->type, "IntxCHR", 8); pdev->major = cpu_to_le64(MAJOR(device_number)); pdev->minor = cpu_to_le64(MINOR(device_number)); - rc = CIFSSMBWrite(xid, pTcon, - fileHandle, - sizeof(struct win_dev), - 0, &bytes_written, (char *)pdev, + rc = CIFSSMBWrite(xid, &io_parms, + &bytes_written, (char *)pdev, NULL, 0); } else if (S_ISBLK(mode)) { memcpy(pdev->type, "IntxBLK", 8); @@ -456,10 +457,8 @@ int cifs_mknod(struct inode *inode, struct dentry *direntry, int mode, cpu_to_le64(MAJOR(device_number)); pdev->minor = cpu_to_le64(MINOR(device_number)); - rc = CIFSSMBWrite(xid, pTcon, - fileHandle, - sizeof(struct win_dev), - 0, &bytes_written, (char *)pdev, + rc = CIFSSMBWrite(xid, &io_parms, + &bytes_written, (char *)pdev, NULL, 0); } /* else if (S_ISFIFO) */ CIFSSMBClose(xid, pTcon, fileHandle); @@ -486,7 +485,7 @@ cifs_lookup(struct inode *parent_dir_inode, struct dentry *direntry, bool posix_open = false; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct cifsFileInfo *cfile; struct inode *newInode = NULL; char *full_path = NULL; diff --git a/fs/cifs/export.c b/fs/cifs/export.c index 993f82045bf6..55d87ac52000 100644 --- a/fs/cifs/export.c +++ b/fs/cifs/export.c @@ -45,7 +45,7 @@ #include "cifs_debug.h" #include "cifsfs.h" -#ifdef CONFIG_CIFS_EXPERIMENTAL +#ifdef CIFS_NFSD_EXPORT static struct dentry *cifs_get_parent(struct dentry *dentry) { /* BB need to add code here eventually to enable export via NFSD */ @@ -63,5 +63,5 @@ const struct export_operations cifs_export_ops = { .encode_fs = */ }; -#endif /* EXPERIMENTAL */ +#endif /* CIFS_NFSD_EXPORT */ diff --git a/fs/cifs/file.c b/fs/cifs/file.c index c27d236738fc..bb71471a4d9d 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -114,7 +114,7 @@ int cifs_posix_open(char *full_path, struct inode **pinode, struct cifs_sb_info *cifs_sb = CIFS_SB(sb); struct cifs_fattr fattr; struct tcon_link *tlink; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; cFYI(1, "posix open %s", full_path); @@ -168,7 +168,7 @@ posix_open_ret: static int cifs_nt_open(char *full_path, struct inode *inode, struct cifs_sb_info *cifs_sb, - struct cifsTconInfo *tcon, unsigned int f_flags, __u32 *poplock, + struct cifs_tcon *tcon, unsigned int f_flags, __u32 *poplock, __u16 *pnetfid, int xid) { int rc; @@ -285,7 +285,7 @@ cifs_new_fileinfo(__u16 fileHandle, struct file *file, void cifsFileInfo_put(struct cifsFileInfo *cifs_file) { struct inode *inode = cifs_file->dentry->d_inode; - struct cifsTconInfo *tcon = tlink_tcon(cifs_file->tlink); + struct cifs_tcon *tcon = tlink_tcon(cifs_file->tlink); struct cifsInodeInfo *cifsi = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsLockInfo *li, *tmp; @@ -343,7 +343,7 @@ int cifs_open(struct inode *inode, struct file *file) int xid; __u32 oplock; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct tcon_link *tlink; struct cifsFileInfo *pCifsFile = NULL; char *full_path = NULL; @@ -457,7 +457,7 @@ static int cifs_reopen_file(struct cifsFileInfo *pCifsFile, bool can_flush) int xid; __u32 oplock; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct cifsInodeInfo *pCifsInode; struct inode *inode; char *full_path = NULL; @@ -575,8 +575,10 @@ reopen_error_exit: int cifs_close(struct inode *inode, struct file *file) { - cifsFileInfo_put(file->private_data); - file->private_data = NULL; + if (file->private_data != NULL) { + cifsFileInfo_put(file->private_data); + file->private_data = NULL; + } /* return code from the ->release op is always ignored */ return 0; @@ -594,7 +596,7 @@ int cifs_closedir(struct inode *inode, struct file *file) xid = GetXid(); if (pCFileStruct) { - struct cifsTconInfo *pTcon = tlink_tcon(pCFileStruct->tlink); + struct cifs_tcon *pTcon = tlink_tcon(pCFileStruct->tlink); cFYI(1, "Freeing private data in close dir"); spin_lock(&cifs_file_list_lock); @@ -651,7 +653,7 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) __u64 length; bool wait_flag = false; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; __u16 netfid; __u8 lockType = LOCKING_ANDX_LARGE_FILES; bool posix_locking = 0; @@ -723,8 +725,8 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) else posix_lock_type = CIFS_WRLCK; rc = CIFSSMBPosixLock(xid, tcon, netfid, 1 /* get */, - length, pfLock, - posix_lock_type, wait_flag); + length, pfLock, posix_lock_type, + wait_flag); FreeXid(xid); return rc; } @@ -795,8 +797,8 @@ int cifs_lock(struct file *file, int cmd, struct file_lock *pfLock) posix_lock_type = CIFS_UNLCK; rc = CIFSSMBPosixLock(xid, tcon, netfid, 0 /* set */, - length, pfLock, - posix_lock_type, wait_flag); + length, pfLock, posix_lock_type, + wait_flag); } else { struct cifsFileInfo *fid = file->private_data; @@ -855,96 +857,7 @@ cifs_update_eof(struct cifsInodeInfo *cifsi, loff_t offset, cifsi->server_eof = end_of_write; } -ssize_t cifs_user_write(struct file *file, const char __user *write_data, - size_t write_size, loff_t *poffset) -{ - struct inode *inode = file->f_path.dentry->d_inode; - int rc = 0; - unsigned int bytes_written = 0; - unsigned int total_written; - struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *pTcon; - int xid; - struct cifsFileInfo *open_file; - struct cifsInodeInfo *cifsi = CIFS_I(inode); - - cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); - - /* cFYI(1, " write %d bytes to offset %lld of %s", write_size, - *poffset, file->f_path.dentry->d_name.name); */ - - if (file->private_data == NULL) - return -EBADF; - - open_file = file->private_data; - pTcon = tlink_tcon(open_file->tlink); - - rc = generic_write_checks(file, poffset, &write_size, 0); - if (rc) - return rc; - - xid = GetXid(); - - for (total_written = 0; write_size > total_written; - total_written += bytes_written) { - rc = -EAGAIN; - while (rc == -EAGAIN) { - if (file->private_data == NULL) { - /* file has been closed on us */ - FreeXid(xid); - /* if we have gotten here we have written some data - and blocked, and the file has been freed on us while - we blocked so return what we managed to write */ - return total_written; - } - if (open_file->invalidHandle) { - /* we could deadlock if we called - filemap_fdatawait from here so tell - reopen_file not to flush data to server - now */ - rc = cifs_reopen_file(open_file, false); - if (rc != 0) - break; - } - - rc = CIFSSMBWrite(xid, pTcon, - open_file->netfid, - min_t(const int, cifs_sb->wsize, - write_size - total_written), - *poffset, &bytes_written, - NULL, write_data + total_written, 0); - } - if (rc || (bytes_written == 0)) { - if (total_written) - break; - else { - FreeXid(xid); - return rc; - } - } else { - cifs_update_eof(cifsi, *poffset, bytes_written); - *poffset += bytes_written; - } - } - - cifs_stats_bytes_written(pTcon, total_written); - -/* Do not update local mtime - server will set its actual value on write - * inode->i_ctime = inode->i_mtime = - * current_fs_time(inode->i_sb);*/ - if (total_written > 0) { - spin_lock(&inode->i_lock); - if (*poffset > inode->i_size) - i_size_write(inode, *poffset); - spin_unlock(&inode->i_lock); - } - mark_inode_dirty_sync(inode); - - FreeXid(xid); - return total_written; -} - -static ssize_t cifs_write(struct cifsFileInfo *open_file, +static ssize_t cifs_write(struct cifsFileInfo *open_file, __u32 pid, const char *write_data, size_t write_size, loff_t *poffset) { @@ -952,10 +865,11 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, unsigned int bytes_written = 0; unsigned int total_written; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; int xid; struct dentry *dentry = open_file->dentry; struct cifsInodeInfo *cifsi = CIFS_I(dentry->d_inode); + struct cifs_io_parms io_parms; cifs_sb = CIFS_SB(dentry->d_sb); @@ -970,6 +884,9 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, total_written += bytes_written) { rc = -EAGAIN; while (rc == -EAGAIN) { + struct kvec iov[2]; + unsigned int len; + if (open_file->invalidHandle) { /* we could deadlock if we called filemap_fdatawait from here so tell @@ -979,31 +896,19 @@ static ssize_t cifs_write(struct cifsFileInfo *open_file, if (rc != 0) break; } - if (experimEnabled || (pTcon->ses->server && - ((pTcon->ses->server->secMode & - (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) - == 0))) { - struct kvec iov[2]; - unsigned int len; - - len = min((size_t)cifs_sb->wsize, - write_size - total_written); - /* iov[0] is reserved for smb header */ - iov[1].iov_base = (char *)write_data + - total_written; - iov[1].iov_len = len; - rc = CIFSSMBWrite2(xid, pTcon, - open_file->netfid, len, - *poffset, &bytes_written, - iov, 1, 0); - } else - rc = CIFSSMBWrite(xid, pTcon, - open_file->netfid, - min_t(const int, cifs_sb->wsize, - write_size - total_written), - *poffset, &bytes_written, - write_data + total_written, - NULL, 0); + + len = min((size_t)cifs_sb->wsize, + write_size - total_written); + /* iov[0] is reserved for smb header */ + iov[1].iov_base = (char *)write_data + total_written; + iov[1].iov_len = len; + io_parms.netfid = open_file->netfid; + io_parms.pid = pid; + io_parms.tcon = pTcon; + io_parms.offset = *poffset; + io_parms.length = len; + rc = CIFSSMBWrite2(xid, &io_parms, &bytes_written, iov, + 1, 0); } if (rc || (bytes_written == 0)) { if (total_written) @@ -1172,8 +1077,8 @@ static int cifs_partialpagewrite(struct page *page, unsigned from, unsigned to) open_file = find_writable_file(CIFS_I(mapping->host), false); if (open_file) { - bytes_written = cifs_write(open_file, write_data, - to - from, &offset); + bytes_written = cifs_write(open_file, open_file->pid, + write_data, to - from, &offset); cifsFileInfo_put(open_file); /* Does mm or vfs already set times? */ inode->i_atime = inode->i_mtime = current_fs_time(inode->i_sb); @@ -1193,64 +1098,20 @@ static int cifs_partialpagewrite(struct page *page, unsigned from, unsigned to) static int cifs_writepages(struct address_space *mapping, struct writeback_control *wbc) { - unsigned int bytes_to_write; - unsigned int bytes_written; - struct cifs_sb_info *cifs_sb; - int done = 0; - pgoff_t end; - pgoff_t index; - int range_whole = 0; - struct kvec *iov; - int len; - int n_iov = 0; - pgoff_t next; - int nr_pages; - __u64 offset = 0; - struct cifsFileInfo *open_file; - struct cifsTconInfo *tcon; - struct cifsInodeInfo *cifsi = CIFS_I(mapping->host); + struct cifs_sb_info *cifs_sb = CIFS_SB(mapping->host->i_sb); + bool done = false, scanned = false, range_whole = false; + pgoff_t end, index; + struct cifs_writedata *wdata; struct page *page; - struct pagevec pvec; int rc = 0; - int scanned = 0; - int xid; - - cifs_sb = CIFS_SB(mapping->host->i_sb); /* - * If wsize is smaller that the page cache size, default to writing + * If wsize is smaller than the page cache size, default to writing * one page at a time via cifs_writepage */ if (cifs_sb->wsize < PAGE_CACHE_SIZE) return generic_writepages(mapping, wbc); - iov = kmalloc(32 * sizeof(struct kvec), GFP_KERNEL); - if (iov == NULL) - return generic_writepages(mapping, wbc); - - /* - * if there's no open file, then this is likely to fail too, - * but it'll at least handle the return. Maybe it should be - * a BUG() instead? - */ - open_file = find_writable_file(CIFS_I(mapping->host), false); - if (!open_file) { - kfree(iov); - return generic_writepages(mapping, wbc); - } - - tcon = tlink_tcon(open_file->tlink); - if (!experimEnabled && tcon->ses->server->secMode & - (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) { - cifsFileInfo_put(open_file); - kfree(iov); - return generic_writepages(mapping, wbc); - } - cifsFileInfo_put(open_file); - - xid = GetXid(); - - pagevec_init(&pvec, 0); if (wbc->range_cyclic) { index = mapping->writeback_index; /* Start from prev offset */ end = -1; @@ -1258,24 +1119,49 @@ static int cifs_writepages(struct address_space *mapping, index = wbc->range_start >> PAGE_CACHE_SHIFT; end = wbc->range_end >> PAGE_CACHE_SHIFT; if (wbc->range_start == 0 && wbc->range_end == LLONG_MAX) - range_whole = 1; - scanned = 1; + range_whole = true; + scanned = true; } retry: - while (!done && (index <= end) && - (nr_pages = pagevec_lookup_tag(&pvec, mapping, &index, - PAGECACHE_TAG_DIRTY, - min(end - index, (pgoff_t)PAGEVEC_SIZE - 1) + 1))) { - int first; - unsigned int i; - - first = -1; - next = 0; - n_iov = 0; - bytes_to_write = 0; - - for (i = 0; i < nr_pages; i++) { - page = pvec.pages[i]; + while (!done && index <= end) { + unsigned int i, nr_pages, found_pages; + pgoff_t next = 0, tofind; + struct page **pages; + + tofind = min((cifs_sb->wsize / PAGE_CACHE_SIZE) - 1, + end - index) + 1; + + wdata = cifs_writedata_alloc((unsigned int)tofind); + if (!wdata) { + rc = -ENOMEM; + break; + } + + /* + * find_get_pages_tag seems to return a max of 256 on each + * iteration, so we must call it several times in order to + * fill the array or the wsize is effectively limited to + * 256 * PAGE_CACHE_SIZE. + */ + found_pages = 0; + pages = wdata->pages; + do { + nr_pages = find_get_pages_tag(mapping, &index, + PAGECACHE_TAG_DIRTY, + tofind, pages); + found_pages += nr_pages; + tofind -= nr_pages; + pages += nr_pages; + } while (nr_pages && tofind && index <= end); + + if (found_pages == 0) { + kref_put(&wdata->refcount, cifs_writedata_release); + break; + } + + nr_pages = 0; + for (i = 0; i < found_pages; i++) { + page = wdata->pages[i]; /* * At this point we hold neither mapping->tree_lock nor * lock on the page itself: the page may be truncated or @@ -1284,7 +1170,7 @@ retry: * mapping */ - if (first < 0) + if (nr_pages == 0) lock_page(page); else if (!trylock_page(page)) break; @@ -1295,7 +1181,7 @@ retry: } if (!wbc->range_cyclic && page->index > end) { - done = 1; + done = true; unlock_page(page); break; } @@ -1322,125 +1208,96 @@ retry: set_page_writeback(page); if (page_offset(page) >= mapping->host->i_size) { - done = 1; + done = true; unlock_page(page); end_page_writeback(page); break; } - /* - * BB can we get rid of this? pages are held by pvec - */ - page_cache_get(page); + wdata->pages[i] = page; + next = page->index + 1; + ++nr_pages; + } - len = min(mapping->host->i_size - page_offset(page), - (loff_t)PAGE_CACHE_SIZE); + /* reset index to refind any pages skipped */ + if (nr_pages == 0) + index = wdata->pages[0]->index + 1; - /* reserve iov[0] for the smb header */ - n_iov++; - iov[n_iov].iov_base = kmap(page); - iov[n_iov].iov_len = len; - bytes_to_write += len; + /* put any pages we aren't going to use */ + for (i = nr_pages; i < found_pages; i++) { + page_cache_release(wdata->pages[i]); + wdata->pages[i] = NULL; + } - if (first < 0) { - first = i; - offset = page_offset(page); - } - next = page->index + 1; - if (bytes_to_write + PAGE_CACHE_SIZE > cifs_sb->wsize) - break; + /* nothing to write? */ + if (nr_pages == 0) { + kref_put(&wdata->refcount, cifs_writedata_release); + continue; } - if (n_iov) { -retry_write: - open_file = find_writable_file(CIFS_I(mapping->host), - false); - if (!open_file) { - cERROR(1, "No writable handles for inode"); - rc = -EBADF; - } else { - rc = CIFSSMBWrite2(xid, tcon, open_file->netfid, - bytes_to_write, offset, - &bytes_written, iov, n_iov, - 0); - cifsFileInfo_put(open_file); - } - cFYI(1, "Write2 rc=%d, wrote=%u", rc, bytes_written); + wdata->sync_mode = wbc->sync_mode; + wdata->nr_pages = nr_pages; + wdata->offset = page_offset(wdata->pages[0]); - /* - * For now, treat a short write as if nothing got - * written. A zero length write however indicates - * ENOSPC or EFBIG. We have no way to know which - * though, so call it ENOSPC for now. EFBIG would - * get translated to AS_EIO anyway. - * - * FIXME: make it take into account the data that did - * get written - */ - if (rc == 0) { - if (bytes_written == 0) - rc = -ENOSPC; - else if (bytes_written < bytes_to_write) - rc = -EAGAIN; + do { + if (wdata->cfile != NULL) + cifsFileInfo_put(wdata->cfile); + wdata->cfile = find_writable_file(CIFS_I(mapping->host), + false); + if (!wdata->cfile) { + cERROR(1, "No writable handles for inode"); + rc = -EBADF; + break; } + rc = cifs_async_writev(wdata); + } while (wbc->sync_mode == WB_SYNC_ALL && rc == -EAGAIN); - /* retry on data-integrity flush */ - if (wbc->sync_mode == WB_SYNC_ALL && rc == -EAGAIN) - goto retry_write; - - /* fix the stats and EOF */ - if (bytes_written > 0) { - cifs_stats_bytes_written(tcon, bytes_written); - cifs_update_eof(cifsi, offset, bytes_written); - } + for (i = 0; i < nr_pages; ++i) + unlock_page(wdata->pages[i]); - for (i = 0; i < n_iov; i++) { - page = pvec.pages[first + i]; - /* on retryable write error, redirty page */ + /* send failure -- clean up the mess */ + if (rc != 0) { + for (i = 0; i < nr_pages; ++i) { if (rc == -EAGAIN) - redirty_page_for_writepage(wbc, page); - else if (rc != 0) - SetPageError(page); - kunmap(page); - unlock_page(page); - end_page_writeback(page); - page_cache_release(page); + redirty_page_for_writepage(wbc, + wdata->pages[i]); + else + SetPageError(wdata->pages[i]); + end_page_writeback(wdata->pages[i]); + page_cache_release(wdata->pages[i]); } - if (rc != -EAGAIN) mapping_set_error(mapping, rc); - else - rc = 0; + } + kref_put(&wdata->refcount, cifs_writedata_release); - if ((wbc->nr_to_write -= n_iov) <= 0) - done = 1; - index = next; - } else - /* Need to re-find the pages we skipped */ - index = pvec.pages[0]->index + 1; + wbc->nr_to_write -= nr_pages; + if (wbc->nr_to_write <= 0) + done = true; - pagevec_release(&pvec); + index = next; } + if (!scanned && !done) { /* * We hit the last page and there is more work to be done: wrap * back to the start of the file */ - scanned = 1; + scanned = true; index = 0; goto retry; } + if (wbc->range_cyclic || (range_whole && wbc->nr_to_write > 0)) mapping->writeback_index = index; - FreeXid(xid); - kfree(iov); return rc; } -static int cifs_writepage(struct page *page, struct writeback_control *wbc) +static int +cifs_writepage_locked(struct page *page, struct writeback_control *wbc) { - int rc = -EFAULT; + int rc; int xid; xid = GetXid(); @@ -1460,21 +1317,43 @@ static int cifs_writepage(struct page *page, struct writeback_control *wbc) * to fail to update with the state of the page correctly. */ set_page_writeback(page); +retry_write: rc = cifs_partialpagewrite(page, 0, PAGE_CACHE_SIZE); - SetPageUptodate(page); /* BB add check for error and Clearuptodate? */ - unlock_page(page); + if (rc == -EAGAIN && wbc->sync_mode == WB_SYNC_ALL) + goto retry_write; + else if (rc == -EAGAIN) + redirty_page_for_writepage(wbc, page); + else if (rc != 0) + SetPageError(page); + else + SetPageUptodate(page); end_page_writeback(page); page_cache_release(page); FreeXid(xid); return rc; } +static int cifs_writepage(struct page *page, struct writeback_control *wbc) +{ + int rc = cifs_writepage_locked(page, wbc); + unlock_page(page); + return rc; +} + static int cifs_write_end(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned copied, struct page *page, void *fsdata) { int rc; struct inode *inode = mapping->host; + struct cifsFileInfo *cfile = file->private_data; + struct cifs_sb_info *cifs_sb = CIFS_SB(cfile->dentry->d_sb); + __u32 pid; + + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD) + pid = cfile->pid; + else + pid = current->tgid; cFYI(1, "write_end for page %p from pos %lld with %d bytes", page, pos, copied); @@ -1498,8 +1377,7 @@ static int cifs_write_end(struct file *file, struct address_space *mapping, /* BB check if anything else missing out of ppw such as updating last write time */ page_data = kmap(page); - rc = cifs_write(file->private_data, page_data + offset, - copied, &pos); + rc = cifs_write(cfile, pid, page_data + offset, copied, &pos); /* if (rc < 0) should we set writebehind rc? */ kunmap(page); @@ -1527,7 +1405,7 @@ int cifs_strict_fsync(struct file *file, int datasync) { int xid; int rc = 0; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct cifsFileInfo *smbfile = file->private_data; struct inode *inode = file->f_path.dentry->d_inode; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); @@ -1537,8 +1415,13 @@ int cifs_strict_fsync(struct file *file, int datasync) cFYI(1, "Sync file - name: %s datasync: 0x%x", file->f_path.dentry->d_name.name, datasync); - if (!CIFS_I(inode)->clientCanCacheRead) - cifs_invalidate_mapping(inode); + if (!CIFS_I(inode)->clientCanCacheRead) { + rc = cifs_invalidate_mapping(inode); + if (rc) { + cFYI(1, "rc: %d during invalidate phase", rc); + rc = 0; /* don't care about it in fsync */ + } + } tcon = tlink_tcon(smbfile->tlink); if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOSSYNC)) @@ -1552,7 +1435,7 @@ int cifs_fsync(struct file *file, int datasync) { int xid; int rc = 0; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct cifsFileInfo *smbfile = file->private_data; struct cifs_sb_info *cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); @@ -1643,9 +1526,11 @@ cifs_iovec_write(struct file *file, const struct iovec *iov, struct iov_iter it; struct inode *inode; struct cifsFileInfo *open_file; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct cifs_sb_info *cifs_sb; + struct cifs_io_parms io_parms; int xid, rc; + __u32 pid; len = iov_length(iov, nr_segs); if (!len) @@ -1677,6 +1562,12 @@ cifs_iovec_write(struct file *file, const struct iovec *iov, xid = GetXid(); open_file = file->private_data; + + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD) + pid = open_file->pid; + else + pid = current->tgid; + pTcon = tlink_tcon(open_file->tlink); inode = file->f_path.dentry->d_inode; @@ -1703,9 +1594,13 @@ cifs_iovec_write(struct file *file, const struct iovec *iov, if (rc != 0) break; } - rc = CIFSSMBWrite2(xid, pTcon, open_file->netfid, - cur_len, *poffset, &written, - to_send, npages, 0); + io_parms.netfid = open_file->netfid; + io_parms.pid = pid; + io_parms.tcon = pTcon; + io_parms.offset = *poffset; + io_parms.length = cur_len; + rc = CIFSSMBWrite2(xid, &io_parms, &written, to_send, + npages, 0); } while (rc == -EAGAIN); for (i = 0; i < npages; i++) @@ -1744,7 +1639,7 @@ cifs_iovec_write(struct file *file, const struct iovec *iov, return total_written; } -static ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov, +ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { ssize_t written; @@ -1798,10 +1693,12 @@ cifs_iovec_read(struct file *file, const struct iovec *iov, size_t len, cur_len; int iov_offset = 0; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct cifsFileInfo *open_file; struct smb_com_read_rsp *pSMBr; + struct cifs_io_parms io_parms; char *read_data; + __u32 pid; if (!nr_segs) return 0; @@ -1816,6 +1713,11 @@ cifs_iovec_read(struct file *file, const struct iovec *iov, open_file = file->private_data; pTcon = tlink_tcon(open_file->tlink); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD) + pid = open_file->pid; + else + pid = current->tgid; + if ((file->f_flags & O_ACCMODE) == O_WRONLY) cFYI(1, "attempting read on write only file instance"); @@ -1831,8 +1733,12 @@ cifs_iovec_read(struct file *file, const struct iovec *iov, if (rc != 0) break; } - rc = CIFSSMBRead(xid, pTcon, open_file->netfid, - cur_len, *poffset, &bytes_read, + io_parms.netfid = open_file->netfid; + io_parms.pid = pid; + io_parms.tcon = pTcon; + io_parms.offset = *poffset; + io_parms.length = len; + rc = CIFSSMBRead(xid, &io_parms, &bytes_read, &read_data, &buf_type); pSMBr = (struct smb_com_read_rsp *)read_data; if (read_data) { @@ -1867,17 +1773,7 @@ cifs_iovec_read(struct file *file, const struct iovec *iov, return total_read; } -ssize_t cifs_user_read(struct file *file, char __user *read_data, - size_t read_size, loff_t *poffset) -{ - struct iovec iov; - iov.iov_base = read_data; - iov.iov_len = read_size; - - return cifs_iovec_read(file, &iov, 1, poffset); -} - -static ssize_t cifs_user_readv(struct kiocb *iocb, const struct iovec *iov, +ssize_t cifs_user_readv(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { ssize_t read; @@ -1919,11 +1815,13 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, unsigned int total_read; unsigned int current_read_size; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; int xid; char *current_offset; struct cifsFileInfo *open_file; + struct cifs_io_parms io_parms; int buf_type = CIFS_NO_BUFFER; + __u32 pid; xid = GetXid(); cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); @@ -1936,6 +1834,11 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, open_file = file->private_data; pTcon = tlink_tcon(open_file->tlink); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD) + pid = open_file->pid; + else + pid = current->tgid; + if ((file->f_flags & O_ACCMODE) == O_WRONLY) cFYI(1, "attempting read on write only file instance"); @@ -1958,11 +1861,13 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, if (rc != 0) break; } - rc = CIFSSMBRead(xid, pTcon, - open_file->netfid, - current_read_size, *poffset, - &bytes_read, ¤t_offset, - &buf_type); + io_parms.netfid = open_file->netfid; + io_parms.pid = pid; + io_parms.tcon = pTcon; + io_parms.offset = *poffset; + io_parms.length = current_read_size; + rc = CIFSSMBRead(xid, &io_parms, &bytes_read, + ¤t_offset, &buf_type); } if (rc || (bytes_read == 0)) { if (total_read) { @@ -1980,6 +1885,24 @@ static ssize_t cifs_read(struct file *file, char *read_data, size_t read_size, return total_read; } +/* + * If the page is mmap'ed into a process' page tables, then we need to make + * sure that it doesn't change while being written back. + */ +static int +cifs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) +{ + struct page *page = vmf->page; + + lock_page(page); + return VM_FAULT_LOCKED; +} + +static struct vm_operations_struct cifs_file_vm_ops = { + .fault = filemap_fault, + .page_mkwrite = cifs_page_mkwrite, +}; + int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma) { int rc, xid; @@ -1987,10 +1910,15 @@ int cifs_file_strict_mmap(struct file *file, struct vm_area_struct *vma) xid = GetXid(); - if (!CIFS_I(inode)->clientCanCacheRead) - cifs_invalidate_mapping(inode); + if (!CIFS_I(inode)->clientCanCacheRead) { + rc = cifs_invalidate_mapping(inode); + if (rc) + return rc; + } rc = generic_file_mmap(file, vma); + if (rc == 0) + vma->vm_ops = &cifs_file_vm_ops; FreeXid(xid); return rc; } @@ -2007,6 +1935,8 @@ int cifs_file_mmap(struct file *file, struct vm_area_struct *vma) return rc; } rc = generic_file_mmap(file, vma); + if (rc == 0) + vma->vm_ops = &cifs_file_vm_ops; FreeXid(xid); return rc; } @@ -2068,13 +1998,15 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, loff_t offset; struct page *page; struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; unsigned int bytes_read = 0; unsigned int read_size, i; char *smb_read_data = NULL; struct smb_com_read_rsp *pSMBr; struct cifsFileInfo *open_file; + struct cifs_io_parms io_parms; int buf_type = CIFS_NO_BUFFER; + __u32 pid; xid = GetXid(); if (file->private_data == NULL) { @@ -2096,6 +2028,11 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, goto read_complete; cFYI(DBG2, "rpages: num pages %d", num_pages); + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_RWPIDFORWARD) + pid = open_file->pid; + else + pid = current->tgid; + for (i = 0; i < num_pages; ) { unsigned contig_pages; struct page *tmp_page; @@ -2137,12 +2074,13 @@ static int cifs_readpages(struct file *file, struct address_space *mapping, if (rc != 0) break; } - - rc = CIFSSMBRead(xid, pTcon, - open_file->netfid, - read_size, offset, - &bytes_read, &smb_read_data, - &buf_type); + io_parms.netfid = open_file->netfid; + io_parms.pid = pid; + io_parms.tcon = pTcon; + io_parms.offset = offset; + io_parms.length = read_size; + rc = CIFSSMBRead(xid, &io_parms, &bytes_read, + &smb_read_data, &buf_type); /* BB more RC checks ? */ if (rc == -EAGAIN) { if (smb_read_data) { @@ -2411,6 +2349,27 @@ static void cifs_invalidate_page(struct page *page, unsigned long offset) cifs_fscache_invalidate_page(page, &cifsi->vfs_inode); } +static int cifs_launder_page(struct page *page) +{ + int rc = 0; + loff_t range_start = page_offset(page); + loff_t range_end = range_start + (loff_t)(PAGE_CACHE_SIZE - 1); + struct writeback_control wbc = { + .sync_mode = WB_SYNC_ALL, + .nr_to_write = 0, + .range_start = range_start, + .range_end = range_end, + }; + + cFYI(1, "Launder page: %p", page); + + if (clear_page_dirty_for_io(page)) + rc = cifs_writepage_locked(page, &wbc); + + cifs_fscache_invalidate_page(page, page->mapping->host); + return rc; +} + void cifs_oplock_break(struct work_struct *work) { struct cifsFileInfo *cfile = container_of(work, struct cifsFileInfo, @@ -2482,7 +2441,7 @@ const struct address_space_operations cifs_addr_ops = { .set_page_dirty = __set_page_dirty_nobuffers, .releasepage = cifs_release_page, .invalidatepage = cifs_invalidate_page, - /* .direct_IO = */ + .launder_page = cifs_launder_page, }; /* @@ -2499,5 +2458,5 @@ const struct address_space_operations cifs_addr_ops_smallbuf = { .set_page_dirty = __set_page_dirty_nobuffers, .releasepage = cifs_release_page, .invalidatepage = cifs_invalidate_page, - /* .direct_IO = */ + .launder_page = cifs_launder_page, }; diff --git a/fs/cifs/fscache.c b/fs/cifs/fscache.c index 297a43d0ff7f..d368a47ba5eb 100644 --- a/fs/cifs/fscache.c +++ b/fs/cifs/fscache.c @@ -40,7 +40,7 @@ void cifs_fscache_release_client_cookie(struct TCP_Server_Info *server) server->fscache = NULL; } -void cifs_fscache_get_super_cookie(struct cifsTconInfo *tcon) +void cifs_fscache_get_super_cookie(struct cifs_tcon *tcon) { struct TCP_Server_Info *server = tcon->ses->server; @@ -51,7 +51,7 @@ void cifs_fscache_get_super_cookie(struct cifsTconInfo *tcon) server->fscache, tcon->fscache); } -void cifs_fscache_release_super_cookie(struct cifsTconInfo *tcon) +void cifs_fscache_release_super_cookie(struct cifs_tcon *tcon) { cFYI(1, "CIFS: releasing superblock cookie (0x%p)", tcon->fscache); fscache_relinquish_cookie(tcon->fscache, 0); @@ -62,7 +62,7 @@ static void cifs_fscache_enable_inode_cookie(struct inode *inode) { struct cifsInodeInfo *cifsi = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); - struct cifsTconInfo *tcon = cifs_sb_master_tcon(cifs_sb); + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); if (cifsi->fscache) return; diff --git a/fs/cifs/fscache.h b/fs/cifs/fscache.h index 31b88ec2341e..63539323e0b9 100644 --- a/fs/cifs/fscache.h +++ b/fs/cifs/fscache.h @@ -40,8 +40,8 @@ extern void cifs_fscache_unregister(void); */ extern void cifs_fscache_get_client_cookie(struct TCP_Server_Info *); extern void cifs_fscache_release_client_cookie(struct TCP_Server_Info *); -extern void cifs_fscache_get_super_cookie(struct cifsTconInfo *); -extern void cifs_fscache_release_super_cookie(struct cifsTconInfo *); +extern void cifs_fscache_get_super_cookie(struct cifs_tcon *); +extern void cifs_fscache_release_super_cookie(struct cifs_tcon *); extern void cifs_fscache_release_inode_cookie(struct inode *); extern void cifs_fscache_set_inode_cookie(struct inode *, struct file *); @@ -99,9 +99,9 @@ static inline void cifs_fscache_get_client_cookie(struct TCP_Server_Info *server) {} static inline void cifs_fscache_release_client_cookie(struct TCP_Server_Info *server) {} -static inline void cifs_fscache_get_super_cookie(struct cifsTconInfo *tcon) {} +static inline void cifs_fscache_get_super_cookie(struct cifs_tcon *tcon) {} static inline void -cifs_fscache_release_super_cookie(struct cifsTconInfo *tcon) {} +cifs_fscache_release_super_cookie(struct cifs_tcon *tcon) {} static inline void cifs_fscache_release_inode_cookie(struct inode *inode) {} static inline void cifs_fscache_set_inode_cookie(struct inode *inode, diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index 8852470b4fbb..9b018c8334fa 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -295,7 +295,7 @@ int cifs_get_file_info_unix(struct file *filp) struct inode *inode = filp->f_path.dentry->d_inode; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsFileInfo *cfile = filp->private_data; - struct cifsTconInfo *tcon = tlink_tcon(cfile->tlink); + struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); xid = GetXid(); rc = CIFSSMBUnixQFileInfo(xid, tcon, cfile->netfid, &find_data); @@ -318,7 +318,7 @@ int cifs_get_inode_info_unix(struct inode **pinode, int rc; FILE_UNIX_BASIC_INFO find_data; struct cifs_fattr fattr; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct tcon_link *tlink; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); @@ -373,7 +373,8 @@ cifs_sfu_type(struct cifs_fattr *fattr, const unsigned char *path, int oplock = 0; __u16 netfid; struct tcon_link *tlink; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; + struct cifs_io_parms io_parms; char buf[24]; unsigned int bytes_read; char *pbuf; @@ -405,9 +406,13 @@ cifs_sfu_type(struct cifs_fattr *fattr, const unsigned char *path, if (rc == 0) { int buf_type = CIFS_NO_BUFFER; /* Read header */ - rc = CIFSSMBRead(xid, tcon, netfid, - 24 /* length */, 0 /* offset */, - &bytes_read, &pbuf, &buf_type); + io_parms.netfid = netfid; + io_parms.pid = current->tgid; + io_parms.tcon = tcon; + io_parms.offset = 0; + io_parms.length = 24; + rc = CIFSSMBRead(xid, &io_parms, &bytes_read, &pbuf, + &buf_type); if ((rc == 0) && (bytes_read >= 8)) { if (memcmp("IntxBLK", pbuf, 8) == 0) { cFYI(1, "Block device"); @@ -468,7 +473,7 @@ static int cifs_sfu_mode(struct cifs_fattr *fattr, const unsigned char *path, char ea_value[4]; __u32 mode; struct tcon_link *tlink; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; tlink = cifs_sb_tlink(cifs_sb); if (IS_ERR(tlink)) @@ -502,7 +507,7 @@ static void cifs_all_info_to_fattr(struct cifs_fattr *fattr, FILE_ALL_INFO *info, struct cifs_sb_info *cifs_sb, bool adjust_tz) { - struct cifsTconInfo *tcon = cifs_sb_master_tcon(cifs_sb); + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); memset(fattr, 0, sizeof(*fattr)); fattr->cf_cifsattrs = le32_to_cpu(info->Attributes); @@ -553,7 +558,7 @@ int cifs_get_file_info(struct file *filp) struct inode *inode = filp->f_path.dentry->d_inode; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct cifsFileInfo *cfile = filp->private_data; - struct cifsTconInfo *tcon = tlink_tcon(cfile->tlink); + struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); xid = GetXid(); rc = CIFSSMBQFileInfo(xid, tcon, cfile->netfid, &find_data); @@ -590,7 +595,7 @@ int cifs_get_inode_info(struct inode **pinode, struct super_block *sb, int xid, const __u16 *pfid) { int rc = 0, tmprc; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct tcon_link *tlink; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); char *buf = NULL; @@ -735,10 +740,10 @@ static const struct inode_operations cifs_ipc_inode_ops = { .lookup = cifs_lookup, }; -char *cifs_build_path_to_root(struct cifs_sb_info *cifs_sb, - struct cifsTconInfo *tcon) +char *cifs_build_path_to_root(struct smb_vol *vol, struct cifs_sb_info *cifs_sb, + struct cifs_tcon *tcon) { - int pplen = cifs_sb->prepathlen; + int pplen = vol->prepath ? strlen(vol->prepath) : 0; int dfsplen; char *full_path = NULL; @@ -772,7 +777,7 @@ char *cifs_build_path_to_root(struct cifs_sb_info *cifs_sb, } } } - strncpy(full_path + dfsplen, cifs_sb->prepath, pplen); + strncpy(full_path + dfsplen, vol->prepath, pplen); full_path[dfsplen + pplen] = 0; /* add trailing null */ return full_path; } @@ -878,25 +883,19 @@ retry_iget5_locked: } /* gets root inode */ -struct inode *cifs_root_iget(struct super_block *sb, unsigned long ino) +struct inode *cifs_root_iget(struct super_block *sb) { int xid; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); struct inode *inode = NULL; long rc; - char *full_path; - struct cifsTconInfo *tcon = cifs_sb_master_tcon(cifs_sb); - - full_path = cifs_build_path_to_root(cifs_sb, tcon); - if (full_path == NULL) - return ERR_PTR(-ENOMEM); + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); xid = GetXid(); if (tcon->unix_ext) - rc = cifs_get_inode_info_unix(&inode, full_path, sb, xid); + rc = cifs_get_inode_info_unix(&inode, "", sb, xid); else - rc = cifs_get_inode_info(&inode, full_path, NULL, sb, - xid, NULL); + rc = cifs_get_inode_info(&inode, "", NULL, sb, xid, NULL); if (!inode) { inode = ERR_PTR(rc); @@ -922,7 +921,6 @@ struct inode *cifs_root_iget(struct super_block *sb, unsigned long ino) } out: - kfree(full_path); /* can not call macro FreeXid here since in a void func * TODO: This is no longer true */ @@ -943,7 +941,7 @@ cifs_set_file_info(struct inode *inode, struct iattr *attrs, int xid, struct cifsInodeInfo *cifsInode = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink = NULL; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; FILE_BASIC_INFO info_buf; if (attrs == NULL) @@ -1061,7 +1059,7 @@ cifs_rename_pending_delete(char *full_path, struct dentry *dentry, int xid) struct cifsInodeInfo *cifsInode = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; __u32 dosattr, origattr; FILE_BASIC_INFO *info_buf = NULL; @@ -1179,7 +1177,7 @@ int cifs_unlink(struct inode *dir, struct dentry *dentry) struct super_block *sb = dir->i_sb; struct cifs_sb_info *cifs_sb = CIFS_SB(sb); struct tcon_link *tlink; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; struct iattr *attrs = NULL; __u32 dosattr = 0, origattr = 0; @@ -1277,7 +1275,7 @@ int cifs_mkdir(struct inode *inode, struct dentry *direntry, int mode) int xid; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; char *full_path = NULL; struct inode *newinode = NULL; struct cifs_fattr fattr; @@ -1455,7 +1453,7 @@ int cifs_rmdir(struct inode *inode, struct dentry *direntry) int xid; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; char *full_path = NULL; struct cifsInodeInfo *cifsInode; @@ -1512,7 +1510,7 @@ cifs_do_rename(int xid, struct dentry *from_dentry, const char *fromPath, { struct cifs_sb_info *cifs_sb = CIFS_SB(from_dentry->d_sb); struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; __u16 srcfid; int oplock, rc; @@ -1564,7 +1562,7 @@ int cifs_rename(struct inode *source_dir, struct dentry *source_dentry, char *toName = NULL; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; FILE_UNIX_BASIC_INFO *info_buf_source = NULL; FILE_UNIX_BASIC_INFO *info_buf_target; int xid, rc, tmprc; @@ -1683,71 +1681,70 @@ cifs_inode_needs_reval(struct inode *inode) /* * Zap the cache. Called when invalid_mapping flag is set. */ -void +int cifs_invalidate_mapping(struct inode *inode) { - int rc; + int rc = 0; struct cifsInodeInfo *cifs_i = CIFS_I(inode); cifs_i->invalid_mapping = false; - /* write back any cached data */ if (inode->i_mapping && inode->i_mapping->nrpages != 0) { - rc = filemap_write_and_wait(inode->i_mapping); - mapping_set_error(inode->i_mapping, rc); + rc = invalidate_inode_pages2(inode->i_mapping); + if (rc) { + cERROR(1, "%s: could not invalidate inode %p", __func__, + inode); + cifs_i->invalid_mapping = true; + } } - invalidate_remote_inode(inode); + cifs_fscache_reset_inode_cookie(inode); + return rc; } -int cifs_revalidate_file(struct file *filp) +int cifs_revalidate_file_attr(struct file *filp) { int rc = 0; struct inode *inode = filp->f_path.dentry->d_inode; struct cifsFileInfo *cfile = (struct cifsFileInfo *) filp->private_data; if (!cifs_inode_needs_reval(inode)) - goto check_inval; + return rc; if (tlink_tcon(cfile->tlink)->unix_ext) rc = cifs_get_file_info_unix(filp); else rc = cifs_get_file_info(filp); -check_inval: - if (CIFS_I(inode)->invalid_mapping) - cifs_invalidate_mapping(inode); - return rc; } -/* revalidate a dentry's inode attributes */ -int cifs_revalidate_dentry(struct dentry *dentry) +int cifs_revalidate_dentry_attr(struct dentry *dentry) { int xid; int rc = 0; - char *full_path = NULL; struct inode *inode = dentry->d_inode; struct super_block *sb = dentry->d_sb; + char *full_path = NULL; if (inode == NULL) return -ENOENT; - xid = GetXid(); - if (!cifs_inode_needs_reval(inode)) - goto check_inval; + return rc; + + xid = GetXid(); /* can not safely grab the rename sem here if rename calls revalidate since that would deadlock */ full_path = build_path_from_dentry(dentry); if (full_path == NULL) { rc = -ENOMEM; - goto check_inval; + goto out; } - cFYI(1, "Revalidate: %s inode 0x%p count %d dentry: 0x%p d_time %ld " - "jiffies %ld", full_path, inode, inode->i_count.counter, + cFYI(1, "Update attributes: %s inode 0x%p count %d dentry: 0x%p d_time " + "%ld jiffies %ld", full_path, inode, inode->i_count.counter, dentry, dentry->d_time, jiffies); if (cifs_sb_master_tcon(CIFS_SB(sb))->unix_ext) @@ -1756,41 +1753,83 @@ int cifs_revalidate_dentry(struct dentry *dentry) rc = cifs_get_inode_info(&inode, full_path, NULL, sb, xid, NULL); -check_inval: - if (CIFS_I(inode)->invalid_mapping) - cifs_invalidate_mapping(inode); - +out: kfree(full_path); FreeXid(xid); return rc; } +int cifs_revalidate_file(struct file *filp) +{ + int rc; + struct inode *inode = filp->f_path.dentry->d_inode; + + rc = cifs_revalidate_file_attr(filp); + if (rc) + return rc; + + if (CIFS_I(inode)->invalid_mapping) + rc = cifs_invalidate_mapping(inode); + return rc; +} + +/* revalidate a dentry's inode attributes */ +int cifs_revalidate_dentry(struct dentry *dentry) +{ + int rc; + struct inode *inode = dentry->d_inode; + + rc = cifs_revalidate_dentry_attr(dentry); + if (rc) + return rc; + + if (CIFS_I(inode)->invalid_mapping) + rc = cifs_invalidate_mapping(inode); + return rc; +} + int cifs_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) { struct cifs_sb_info *cifs_sb = CIFS_SB(dentry->d_sb); - struct cifsTconInfo *tcon = cifs_sb_master_tcon(cifs_sb); - int err = cifs_revalidate_dentry(dentry); - - if (!err) { - generic_fillattr(dentry->d_inode, stat); - stat->blksize = CIFS_MAX_MSGSIZE; - stat->ino = CIFS_I(dentry->d_inode)->uniqueid; + struct cifs_tcon *tcon = cifs_sb_master_tcon(cifs_sb); + struct inode *inode = dentry->d_inode; + int rc; - /* - * If on a multiuser mount without unix extensions, and the - * admin hasn't overridden them, set the ownership to the - * fsuid/fsgid of the current process. - */ - if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER) && - !tcon->unix_ext) { - if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID)) - stat->uid = current_fsuid(); - if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID)) - stat->gid = current_fsgid(); + /* + * We need to be sure that all dirty pages are written and the server + * has actual ctime, mtime and file length. + */ + if (!CIFS_I(inode)->clientCanCacheRead && inode->i_mapping && + inode->i_mapping->nrpages != 0) { + rc = filemap_fdatawait(inode->i_mapping); + if (rc) { + mapping_set_error(inode->i_mapping, rc); + return rc; } } - return err; + + rc = cifs_revalidate_dentry_attr(dentry); + if (rc) + return rc; + + generic_fillattr(inode, stat); + stat->blksize = CIFS_MAX_MSGSIZE; + stat->ino = CIFS_I(inode)->uniqueid; + + /* + * If on a multiuser mount without unix extensions, and the admin hasn't + * overridden them, set the ownership to the fsuid/fsgid of the current + * process. + */ + if ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MULTIUSER) && + !tcon->unix_ext) { + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID)) + stat->uid = current_fsuid(); + if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_GID)) + stat->gid = current_fsgid(); + } + return rc; } static int cifs_truncate_page(struct address_space *mapping, loff_t from) @@ -1831,7 +1870,8 @@ cifs_set_file_size(struct inode *inode, struct iattr *attrs, struct cifsInodeInfo *cifsInode = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink = NULL; - struct cifsTconInfo *pTcon = NULL; + struct cifs_tcon *pTcon = NULL; + struct cifs_io_parms io_parms; /* * To avoid spurious oplock breaks from server, in the case of @@ -1853,8 +1893,14 @@ cifs_set_file_size(struct inode *inode, struct iattr *attrs, cFYI(1, "SetFSize for attrs rc = %d", rc); if ((rc == -EINVAL) || (rc == -EOPNOTSUPP)) { unsigned int bytes_written; - rc = CIFSSMBWrite(xid, pTcon, nfid, 0, attrs->ia_size, - &bytes_written, NULL, NULL, 1); + + io_parms.netfid = nfid; + io_parms.pid = npid; + io_parms.tcon = pTcon; + io_parms.offset = 0; + io_parms.length = attrs->ia_size; + rc = CIFSSMBWrite(xid, &io_parms, &bytes_written, + NULL, NULL, 1); cFYI(1, "Wrt seteof rc %d", rc); } } else @@ -1889,10 +1935,15 @@ cifs_set_file_size(struct inode *inode, struct iattr *attrs, CIFS_MOUNT_MAP_SPECIAL_CHR); if (rc == 0) { unsigned int bytes_written; - rc = CIFSSMBWrite(xid, pTcon, netfid, 0, - attrs->ia_size, - &bytes_written, NULL, - NULL, 1); + + io_parms.netfid = netfid; + io_parms.pid = current->tgid; + io_parms.tcon = pTcon; + io_parms.offset = 0; + io_parms.length = attrs->ia_size; + rc = CIFSSMBWrite(xid, &io_parms, + &bytes_written, + NULL, NULL, 1); cFYI(1, "wrt seteof rc %d", rc); CIFSSMBClose(xid, pTcon, netfid); } @@ -1920,7 +1971,7 @@ cifs_setattr_unix(struct dentry *direntry, struct iattr *attrs) struct cifsInodeInfo *cifsInode = CIFS_I(inode); struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct cifs_unix_set_info_args *args = NULL; struct cifsFileInfo *open_file; @@ -2206,7 +2257,7 @@ cifs_setattr(struct dentry *direntry, struct iattr *attrs) { struct inode *inode = direntry->d_inode; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); - struct cifsTconInfo *pTcon = cifs_sb_master_tcon(cifs_sb); + struct cifs_tcon *pTcon = cifs_sb_master_tcon(cifs_sb); if (pTcon->unix_ext) return cifs_setattr_unix(direntry, attrs); diff --git a/fs/cifs/ioctl.c b/fs/cifs/ioctl.c index 0c98672d0122..4221b5e48a42 100644 --- a/fs/cifs/ioctl.c +++ b/fs/cifs/ioctl.c @@ -38,7 +38,7 @@ long cifs_ioctl(struct file *filep, unsigned int command, unsigned long arg) struct cifs_sb_info *cifs_sb; #ifdef CONFIG_CIFS_POSIX struct cifsFileInfo *pSMBFile = filep->private_data; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; __u64 ExtAttrBits = 0; __u64 ExtAttrMask = 0; __u64 caps; diff --git a/fs/cifs/link.c b/fs/cifs/link.c index e8804d373404..556b1a0b54de 100644 --- a/fs/cifs/link.c +++ b/fs/cifs/link.c @@ -175,7 +175,7 @@ CIFSFormatMFSymlink(u8 *buf, unsigned int buf_len, const char *link_str) } static int -CIFSCreateMFSymLink(const int xid, struct cifsTconInfo *tcon, +CIFSCreateMFSymLink(const int xid, struct cifs_tcon *tcon, const char *fromName, const char *toName, const struct nls_table *nls_codepage, int remap) { @@ -184,6 +184,7 @@ CIFSCreateMFSymLink(const int xid, struct cifsTconInfo *tcon, __u16 netfid = 0; u8 *buf; unsigned int bytes_written = 0; + struct cifs_io_parms io_parms; buf = kmalloc(CIFS_MF_SYMLINK_FILE_SIZE, GFP_KERNEL); if (!buf) @@ -203,10 +204,13 @@ CIFSCreateMFSymLink(const int xid, struct cifsTconInfo *tcon, return rc; } - rc = CIFSSMBWrite(xid, tcon, netfid, - CIFS_MF_SYMLINK_FILE_SIZE /* length */, - 0 /* offset */, - &bytes_written, buf, NULL, 0); + io_parms.netfid = netfid; + io_parms.pid = current->tgid; + io_parms.tcon = tcon; + io_parms.offset = 0; + io_parms.length = CIFS_MF_SYMLINK_FILE_SIZE; + + rc = CIFSSMBWrite(xid, &io_parms, &bytes_written, buf, NULL, 0); CIFSSMBClose(xid, tcon, netfid); kfree(buf); if (rc != 0) @@ -219,7 +223,7 @@ CIFSCreateMFSymLink(const int xid, struct cifsTconInfo *tcon, } static int -CIFSQueryMFSymLink(const int xid, struct cifsTconInfo *tcon, +CIFSQueryMFSymLink(const int xid, struct cifs_tcon *tcon, const unsigned char *searchName, char **symlinkinfo, const struct nls_table *nls_codepage, int remap) { @@ -231,6 +235,7 @@ CIFSQueryMFSymLink(const int xid, struct cifsTconInfo *tcon, unsigned int bytes_read = 0; int buf_type = CIFS_NO_BUFFER; unsigned int link_len = 0; + struct cifs_io_parms io_parms; FILE_ALL_INFO file_info; rc = CIFSSMBOpen(xid, tcon, searchName, FILE_OPEN, GENERIC_READ, @@ -239,7 +244,7 @@ CIFSQueryMFSymLink(const int xid, struct cifsTconInfo *tcon, if (rc != 0) return rc; - if (file_info.EndOfFile != CIFS_MF_SYMLINK_FILE_SIZE) { + if (file_info.EndOfFile != cpu_to_le64(CIFS_MF_SYMLINK_FILE_SIZE)) { CIFSSMBClose(xid, tcon, netfid); /* it's not a symlink */ return -EINVAL; @@ -249,11 +254,13 @@ CIFSQueryMFSymLink(const int xid, struct cifsTconInfo *tcon, if (!buf) return -ENOMEM; pbuf = buf; + io_parms.netfid = netfid; + io_parms.pid = current->tgid; + io_parms.tcon = tcon; + io_parms.offset = 0; + io_parms.length = CIFS_MF_SYMLINK_FILE_SIZE; - rc = CIFSSMBRead(xid, tcon, netfid, - CIFS_MF_SYMLINK_FILE_SIZE /* length */, - 0 /* offset */, - &bytes_read, &pbuf, &buf_type); + rc = CIFSSMBRead(xid, &io_parms, &bytes_read, &pbuf, &buf_type); CIFSSMBClose(xid, tcon, netfid); if (rc != 0) { kfree(buf); @@ -291,7 +298,8 @@ CIFSCheckMFSymlink(struct cifs_fattr *fattr, int oplock = 0; __u16 netfid = 0; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; + struct cifs_io_parms io_parms; u8 *buf; char *pbuf; unsigned int bytes_read = 0; @@ -316,7 +324,7 @@ CIFSCheckMFSymlink(struct cifs_fattr *fattr, if (rc != 0) goto out; - if (file_info.EndOfFile != CIFS_MF_SYMLINK_FILE_SIZE) { + if (file_info.EndOfFile != cpu_to_le64(CIFS_MF_SYMLINK_FILE_SIZE)) { CIFSSMBClose(xid, pTcon, netfid); /* it's not a symlink */ goto out; @@ -328,11 +336,13 @@ CIFSCheckMFSymlink(struct cifs_fattr *fattr, goto out; } pbuf = buf; + io_parms.netfid = netfid; + io_parms.pid = current->tgid; + io_parms.tcon = pTcon; + io_parms.offset = 0; + io_parms.length = CIFS_MF_SYMLINK_FILE_SIZE; - rc = CIFSSMBRead(xid, pTcon, netfid, - CIFS_MF_SYMLINK_FILE_SIZE /* length */, - 0 /* offset */, - &bytes_read, &pbuf, &buf_type); + rc = CIFSSMBRead(xid, &io_parms, &bytes_read, &pbuf, &buf_type); CIFSSMBClose(xid, pTcon, netfid); if (rc != 0) { kfree(buf); @@ -370,7 +380,7 @@ cifs_hardlink(struct dentry *old_file, struct inode *inode, char *toName = NULL; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct cifsInodeInfo *cifsInode; tlink = cifs_sb_tlink(cifs_sb); @@ -445,7 +455,7 @@ cifs_follow_link(struct dentry *direntry, struct nameidata *nd) char *target_path = NULL; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink = NULL; - struct cifsTconInfo *tcon; + struct cifs_tcon *tcon; xid = GetXid(); @@ -518,7 +528,7 @@ cifs_symlink(struct inode *inode, struct dentry *direntry, const char *symname) int xid; struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; char *full_path = NULL; struct inode *newinode = NULL; diff --git a/fs/cifs/misc.c b/fs/cifs/misc.c index 2a930a752a78..03a1f491d39b 100644 --- a/fs/cifs/misc.c +++ b/fs/cifs/misc.c @@ -67,12 +67,12 @@ _FreeXid(unsigned int xid) spin_unlock(&GlobalMid_Lock); } -struct cifsSesInfo * +struct cifs_ses * sesInfoAlloc(void) { - struct cifsSesInfo *ret_buf; + struct cifs_ses *ret_buf; - ret_buf = kzalloc(sizeof(struct cifsSesInfo), GFP_KERNEL); + ret_buf = kzalloc(sizeof(struct cifs_ses), GFP_KERNEL); if (ret_buf) { atomic_inc(&sesInfoAllocCount); ret_buf->status = CifsNew; @@ -85,7 +85,7 @@ sesInfoAlloc(void) } void -sesInfoFree(struct cifsSesInfo *buf_to_free) +sesInfoFree(struct cifs_ses *buf_to_free) { if (buf_to_free == NULL) { cFYI(1, "Null buffer passed to sesInfoFree"); @@ -100,15 +100,16 @@ sesInfoFree(struct cifsSesInfo *buf_to_free) memset(buf_to_free->password, 0, strlen(buf_to_free->password)); kfree(buf_to_free->password); } + kfree(buf_to_free->user_name); kfree(buf_to_free->domainName); kfree(buf_to_free); } -struct cifsTconInfo * +struct cifs_tcon * tconInfoAlloc(void) { - struct cifsTconInfo *ret_buf; - ret_buf = kzalloc(sizeof(struct cifsTconInfo), GFP_KERNEL); + struct cifs_tcon *ret_buf; + ret_buf = kzalloc(sizeof(struct cifs_tcon), GFP_KERNEL); if (ret_buf) { atomic_inc(&tconInfoAllocCount); ret_buf->tidStatus = CifsNew; @@ -123,7 +124,7 @@ tconInfoAlloc(void) } void -tconInfoFree(struct cifsTconInfo *buf_to_free) +tconInfoFree(struct cifs_tcon *buf_to_free) { if (buf_to_free == NULL) { cFYI(1, "Null buffer passed to tconInfoFree"); @@ -294,21 +295,19 @@ __u16 GetNextMid(struct TCP_Server_Info *server) case it is responsbility of caller to set the mid */ void header_assemble(struct smb_hdr *buffer, char smb_command /* command */ , - const struct cifsTconInfo *treeCon, int word_count + const struct cifs_tcon *treeCon, int word_count /* length of fixed section (word count) in two byte units */) { struct list_head *temp_item; - struct cifsSesInfo *ses; + struct cifs_ses *ses; char *temp = (char *) buffer; memset(temp, 0, 256); /* bigger than MAX_CIFS_HDR_SIZE */ - buffer->smb_buf_length = + buffer->smb_buf_length = cpu_to_be32( (2 * word_count) + sizeof(struct smb_hdr) - 4 /* RFC 1001 length field does not count */ + - 2 /* for bcc field itself */ ; - /* Note that this is the only network field that has to be converted - to big endian and it is done just before we send it */ + 2 /* for bcc field itself */) ; buffer->Protocol[0] = 0xFF; buffer->Protocol[1] = 'S'; @@ -360,7 +359,7 @@ header_assemble(struct smb_hdr *buffer, char smb_command /* command */ , "did not match tcon uid"); spin_lock(&cifs_tcp_ses_lock); list_for_each(temp_item, &treeCon->ses->server->smb_ses_list) { - ses = list_entry(temp_item, struct cifsSesInfo, smb_ses_list); + ses = list_entry(temp_item, struct cifs_ses, smb_ses_list); if (ses->linux_uid == current_fsuid()) { if (ses->server == treeCon->ses->server) { cFYI(1, "found matching uid substitute right smb_uid"); @@ -381,7 +380,7 @@ header_assemble(struct smb_hdr *buffer, char smb_command /* command */ , if (treeCon->nocase) buffer->Flags |= SMBFLG_CASELESS; if ((treeCon->ses) && (treeCon->ses->server)) - if (treeCon->ses->server->secMode & + if (treeCon->ses->server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) buffer->Flags2 |= SMBFLG2_SECURITY_SIGNATURE; } @@ -423,7 +422,7 @@ check_smb_hdr(struct smb_hdr *smb, __u16 mid) int checkSMB(struct smb_hdr *smb, __u16 mid, unsigned int length) { - __u32 len = smb->smb_buf_length; + __u32 len = be32_to_cpu(smb->smb_buf_length); __u32 clc_len; /* calculated length */ cFYI(0, "checkSMB Length: 0x%x, smb_buf_length: 0x%x", length, len); @@ -463,7 +462,7 @@ checkSMB(struct smb_hdr *smb, __u16 mid, unsigned int length) if (check_smb_hdr(smb, mid)) return 1; - clc_len = smbCalcSize_LE(smb); + clc_len = smbCalcSize(smb); if (4 + len != length) { cERROR(1, "Length read does not match RFC1001 length %d", @@ -508,8 +507,8 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) { struct smb_com_lock_req *pSMB = (struct smb_com_lock_req *)buf; struct list_head *tmp, *tmp1, *tmp2; - struct cifsSesInfo *ses; - struct cifsTconInfo *tcon; + struct cifs_ses *ses; + struct cifs_tcon *tcon; struct cifsInodeInfo *pCifsInode; struct cifsFileInfo *netfile; @@ -520,7 +519,7 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) (struct smb_com_transaction_change_notify_rsp *)buf; struct file_notify_information *pnotify; __u32 data_offset = 0; - if (pSMBr->ByteCount > sizeof(struct file_notify_information)) { + if (get_bcc(buf) > sizeof(struct file_notify_information)) { data_offset = le32_to_cpu(pSMBr->DataOffset); pnotify = (struct file_notify_information *) @@ -567,9 +566,9 @@ is_valid_oplock_break(struct smb_hdr *buf, struct TCP_Server_Info *srv) /* look up tcon based on tid & uid */ spin_lock(&cifs_tcp_ses_lock); list_for_each(tmp, &srv->smb_ses_list) { - ses = list_entry(tmp, struct cifsSesInfo, smb_ses_list); + ses = list_entry(tmp, struct cifs_ses, smb_ses_list); list_for_each(tmp1, &ses->tcon_list) { - tcon = list_entry(tmp1, struct cifsTconInfo, tcon_list); + tcon = list_entry(tmp1, struct cifs_tcon, tcon_list); if (tcon->tid != buf->Tid) continue; diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index 79f641eeda30..73e47e84b61a 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -836,7 +836,7 @@ ntstatus_to_dos(__u32 ntstatus, __u8 *eclass, __u16 *ecode) } int -map_smb_to_linux_error(struct smb_hdr *smb, int logErr) +map_smb_to_linux_error(struct smb_hdr *smb, bool logErr) { unsigned int i; int rc = -EIO; /* if transport error smb error may not be set */ @@ -919,13 +919,6 @@ smbCalcSize(struct smb_hdr *ptr) 2 /* size of the bcc field */ + get_bcc(ptr)); } -unsigned int -smbCalcSize_LE(struct smb_hdr *ptr) -{ - return (sizeof(struct smb_hdr) + (2 * ptr->WordCount) + - 2 /* size of the bcc field */ + get_bcc_le(ptr)); -} - /* The following are taken from fs/ntfs/util.c */ #define NTFS_TIME_OFFSET ((u64)(369*365 + 89) * 24 * 3600 * 10000000) diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index f8e4cd2a7912..6751e745bbc6 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -195,7 +195,7 @@ int get_symlink_reparse_path(char *full_path, struct cifs_sb_info *cifs_sb, int len; int oplock = 0; int rc; - struct cifsTconInfo *ptcon = cifs_sb_tcon(cifs_sb); + struct cifs_tcon *ptcon = cifs_sb_tcon(cifs_sb); char *tmpbuffer; rc = CIFSSMBOpen(xid, ptcon, full_path, FILE_OPEN, GENERIC_READ, @@ -223,7 +223,7 @@ static int initiate_cifs_search(const int xid, struct file *file) struct cifsFileInfo *cifsFile; struct cifs_sb_info *cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); struct tcon_link *tlink = NULL; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; if (file->private_data == NULL) { tlink = cifs_sb_tlink(cifs_sb); @@ -496,7 +496,7 @@ static int cifs_save_resume_key(const char *current_entry, assume that they are located in the findfirst return buffer.*/ /* We start counting in the buffer with entry 2 and increment for every entry (do not increment for . or .. entry) */ -static int find_cifs_entry(const int xid, struct cifsTconInfo *pTcon, +static int find_cifs_entry(const int xid, struct cifs_tcon *pTcon, struct file *file, char **ppCurrentEntry, int *num_to_ret) { int rc = 0; @@ -764,7 +764,7 @@ int cifs_readdir(struct file *file, void *direntry, filldir_t filldir) { int rc = 0; int xid, i; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct cifsFileInfo *cifsFile = NULL; char *current_entry; int num_to_fill = 0; diff --git a/fs/cifs/sess.c b/fs/cifs/sess.c index 16765703131b..3892ab817a36 100644 --- a/fs/cifs/sess.c +++ b/fs/cifs/sess.c @@ -37,13 +37,13 @@ * the socket has been reestablished (so we know whether to use vc 0). * Called while holding the cifs_tcp_ses_lock, so do not block */ -static bool is_first_ses_reconnect(struct cifsSesInfo *ses) +static bool is_first_ses_reconnect(struct cifs_ses *ses) { struct list_head *tmp; - struct cifsSesInfo *tmp_ses; + struct cifs_ses *tmp_ses; list_for_each(tmp, &ses->server->smb_ses_list) { - tmp_ses = list_entry(tmp, struct cifsSesInfo, + tmp_ses = list_entry(tmp, struct cifs_ses, smb_ses_list); if (tmp_ses->need_reconnect == false) return false; @@ -61,11 +61,11 @@ static bool is_first_ses_reconnect(struct cifsSesInfo *ses) * any vc but zero (some servers reset the connection on vcnum zero) * */ -static __le16 get_next_vcnum(struct cifsSesInfo *ses) +static __le16 get_next_vcnum(struct cifs_ses *ses) { __u16 vcnum = 0; struct list_head *tmp; - struct cifsSesInfo *tmp_ses; + struct cifs_ses *tmp_ses; __u16 max_vcs = ses->server->max_vcs; __u16 i; int free_vc_found = 0; @@ -87,7 +87,7 @@ static __le16 get_next_vcnum(struct cifsSesInfo *ses) free_vc_found = 1; list_for_each(tmp, &ses->server->smb_ses_list) { - tmp_ses = list_entry(tmp, struct cifsSesInfo, + tmp_ses = list_entry(tmp, struct cifs_ses, smb_ses_list); if (tmp_ses->vcnum == i) { free_vc_found = 0; @@ -114,7 +114,7 @@ get_vc_num_exit: return cpu_to_le16(vcnum); } -static __u32 cifs_ssetup_hdr(struct cifsSesInfo *ses, SESSION_SETUP_ANDX *pSMB) +static __u32 cifs_ssetup_hdr(struct cifs_ses *ses, SESSION_SETUP_ANDX *pSMB) { __u32 capabilities = 0; @@ -136,7 +136,7 @@ static __u32 cifs_ssetup_hdr(struct cifsSesInfo *ses, SESSION_SETUP_ANDX *pSMB) capabilities = CAP_LARGE_FILES | CAP_NT_SMBS | CAP_LEVEL_II_OPLOCKS | CAP_LARGE_WRITE_X | CAP_LARGE_READ_X; - if (ses->server->secMode & + if (ses->server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) pSMB->req.hdr.Flags2 |= SMBFLG2_SECURITY_SIGNATURE; @@ -181,7 +181,7 @@ unicode_oslm_strings(char **pbcc_area, const struct nls_table *nls_cp) *pbcc_area = bcc_ptr; } -static void unicode_domain_string(char **pbcc_area, struct cifsSesInfo *ses, +static void unicode_domain_string(char **pbcc_area, struct cifs_ses *ses, const struct nls_table *nls_cp) { char *bcc_ptr = *pbcc_area; @@ -204,7 +204,7 @@ static void unicode_domain_string(char **pbcc_area, struct cifsSesInfo *ses, } -static void unicode_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, +static void unicode_ssetup_strings(char **pbcc_area, struct cifs_ses *ses, const struct nls_table *nls_cp) { char *bcc_ptr = *pbcc_area; @@ -219,12 +219,12 @@ static void unicode_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, bcc_ptr++; } */ /* copy user */ - if (ses->userName == NULL) { + if (ses->user_name == NULL) { /* null user mount */ *bcc_ptr = 0; *(bcc_ptr+1) = 0; } else { - bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->userName, + bytes_ret = cifs_strtoUCS((__le16 *) bcc_ptr, ses->user_name, MAX_USERNAME_SIZE, nls_cp); } bcc_ptr += 2 * bytes_ret; @@ -236,7 +236,7 @@ static void unicode_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, *pbcc_area = bcc_ptr; } -static void ascii_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, +static void ascii_ssetup_strings(char **pbcc_area, struct cifs_ses *ses, const struct nls_table *nls_cp) { char *bcc_ptr = *pbcc_area; @@ -244,12 +244,11 @@ static void ascii_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, /* copy user */ /* BB what about null user mounts - check that we do this BB */ /* copy user */ - if (ses->userName == NULL) { - /* BB what about null user mounts - check that we do this BB */ - } else { - strncpy(bcc_ptr, ses->userName, MAX_USERNAME_SIZE); - } - bcc_ptr += strnlen(ses->userName, MAX_USERNAME_SIZE); + if (ses->user_name != NULL) + strncpy(bcc_ptr, ses->user_name, MAX_USERNAME_SIZE); + /* else null user mount */ + + bcc_ptr += strnlen(ses->user_name, MAX_USERNAME_SIZE); *bcc_ptr = 0; bcc_ptr++; /* account for null termination */ @@ -277,7 +276,7 @@ static void ascii_ssetup_strings(char **pbcc_area, struct cifsSesInfo *ses, } static void -decode_unicode_ssetup(char **pbcc_area, __u16 bleft, struct cifsSesInfo *ses, +decode_unicode_ssetup(char **pbcc_area, int bleft, struct cifs_ses *ses, const struct nls_table *nls_cp) { int len; @@ -285,19 +284,6 @@ decode_unicode_ssetup(char **pbcc_area, __u16 bleft, struct cifsSesInfo *ses, cFYI(1, "bleft %d", bleft); - /* - * Windows servers do not always double null terminate their final - * Unicode string. Check to see if there are an uneven number of bytes - * left. If so, then add an extra NULL pad byte to the end of the - * response. - * - * See section 2.7.2 in "Implementing CIFS" for details - */ - if (bleft % 2) { - data[bleft] = 0; - ++bleft; - } - kfree(ses->serverOS); ses->serverOS = cifs_strndup_from_ucs(data, bleft, true, nls_cp); cFYI(1, "serverOS=%s", ses->serverOS); @@ -324,7 +310,7 @@ decode_unicode_ssetup(char **pbcc_area, __u16 bleft, struct cifsSesInfo *ses, } static int decode_ascii_ssetup(char **pbcc_area, __u16 bleft, - struct cifsSesInfo *ses, + struct cifs_ses *ses, const struct nls_table *nls_cp) { int rc = 0; @@ -378,7 +364,7 @@ static int decode_ascii_ssetup(char **pbcc_area, __u16 bleft, } static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, - struct cifsSesInfo *ses) + struct cifs_ses *ses) { unsigned int tioffset; /* challenge message target info area */ unsigned int tilen; /* challenge message target info area length */ @@ -405,8 +391,8 @@ static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, /* BB spec says that if AvId field of MsvAvTimestamp is populated then we must set the MIC field of the AUTHENTICATE_MESSAGE */ ses->ntlmssp->server_flags = le32_to_cpu(pblob->NegotiateFlags); - tioffset = cpu_to_le16(pblob->TargetInfoArray.BufferOffset); - tilen = cpu_to_le16(pblob->TargetInfoArray.Length); + tioffset = le32_to_cpu(pblob->TargetInfoArray.BufferOffset); + tilen = le16_to_cpu(pblob->TargetInfoArray.Length); if (tilen) { ses->auth_key.response = kmalloc(tilen, GFP_KERNEL); if (!ses->auth_key.response) { @@ -425,7 +411,7 @@ static int decode_ntlmssp_challenge(char *bcc_ptr, int blob_len, /* We do not malloc the blob, it is passed in pbuffer, because it is fixed size, and small, making this approach cleaner */ static void build_ntlmssp_negotiate_blob(unsigned char *pbuffer, - struct cifsSesInfo *ses) + struct cifs_ses *ses) { NEGOTIATE_MESSAGE *sec_blob = (NEGOTIATE_MESSAGE *)pbuffer; __u32 flags; @@ -438,7 +424,7 @@ static void build_ntlmssp_negotiate_blob(unsigned char *pbuffer, flags = NTLMSSP_NEGOTIATE_56 | NTLMSSP_REQUEST_TARGET | NTLMSSP_NEGOTIATE_128 | NTLMSSP_NEGOTIATE_UNICODE | NTLMSSP_NEGOTIATE_NTLM | NTLMSSP_NEGOTIATE_EXTENDED_SEC; - if (ses->server->secMode & + if (ses->server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) { flags |= NTLMSSP_NEGOTIATE_SIGN; if (!ses->server->session_estab) @@ -463,7 +449,7 @@ static void build_ntlmssp_negotiate_blob(unsigned char *pbuffer, This function returns the length of the data in the blob */ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, u16 *buflen, - struct cifsSesInfo *ses, + struct cifs_ses *ses, const struct nls_table *nls_cp) { int rc; @@ -478,10 +464,10 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, NTLMSSP_REQUEST_TARGET | NTLMSSP_NEGOTIATE_TARGET_INFO | NTLMSSP_NEGOTIATE_128 | NTLMSSP_NEGOTIATE_UNICODE | NTLMSSP_NEGOTIATE_NTLM | NTLMSSP_NEGOTIATE_EXTENDED_SEC; - if (ses->server->secMode & + if (ses->server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) flags |= NTLMSSP_NEGOTIATE_SIGN; - if (ses->server->secMode & SECMODE_SIGN_REQUIRED) + if (ses->server->sec_mode & SECMODE_SIGN_REQUIRED) flags |= NTLMSSP_NEGOTIATE_ALWAYS_SIGN; tmp = pbuffer + sizeof(AUTHENTICATE_MESSAGE); @@ -523,14 +509,14 @@ static int build_ntlmssp_auth_blob(unsigned char *pbuffer, tmp += len; } - if (ses->userName == NULL) { + if (ses->user_name == NULL) { sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - pbuffer); sec_blob->UserName.Length = 0; sec_blob->UserName.MaximumLength = 0; tmp += 2; } else { int len; - len = cifs_strtoUCS((__le16 *)tmp, ses->userName, + len = cifs_strtoUCS((__le16 *)tmp, ses->user_name, MAX_USERNAME_SIZE, nls_cp); len *= 2; /* unicode is 2 bytes each */ sec_blob->UserName.BufferOffset = cpu_to_le32(tmp - pbuffer); @@ -565,7 +551,7 @@ setup_ntlmv2_ret: } int -CIFS_SessSetup(unsigned int xid, struct cifsSesInfo *ses, +CIFS_SessSetup(unsigned int xid, struct cifs_ses *ses, const struct nls_table *nls_cp) { int rc = 0; @@ -635,7 +621,7 @@ ssetup_ntlmssp_authenticate: and rest of bcc area. This allows us to avoid a large buffer 17K allocation */ iov[0].iov_base = (char *)pSMB; - iov[0].iov_len = smb_buf->smb_buf_length + 4; + iov[0].iov_len = be32_to_cpu(smb_buf->smb_buf_length) + 4; /* setting this here allows the code at the end of the function to free the request buffer if there's an error */ @@ -670,8 +656,8 @@ ssetup_ntlmssp_authenticate: * to use challenge/response method (i.e. Password bit is 1). */ - calc_lanman_hash(ses->password, ses->server->cryptkey, - ses->server->secMode & SECMODE_PW_ENCRYPT ? + rc = calc_lanman_hash(ses->password, ses->server->cryptkey, + ses->server->sec_mode & SECMODE_PW_ENCRYPT ? true : false, lnm_session_key); ses->flags |= CIFS_SES_LANMAN; @@ -873,9 +859,10 @@ ssetup_ntlmssp_authenticate: iov[2].iov_len = (long) bcc_ptr - (long) str_area; count = iov[1].iov_len + iov[2].iov_len; - smb_buf->smb_buf_length += count; + smb_buf->smb_buf_length = + cpu_to_be32(be32_to_cpu(smb_buf->smb_buf_length) + count); - put_bcc_le(count, smb_buf); + put_bcc(count, smb_buf); rc = SendReceive2(xid, ses, iov, 3 /* num_iovecs */, &resp_buf_type, CIFS_LOG_ERROR); @@ -930,7 +917,9 @@ ssetup_ntlmssp_authenticate: } /* BB check if Unicode and decode strings */ - if (smb_buf->Flags2 & SMBFLG2_UNICODE) { + if (bytes_remaining == 0) { + /* no string area to decode, do nothing */ + } else if (smb_buf->Flags2 & SMBFLG2_UNICODE) { /* unicode string area must be word-aligned */ if (((unsigned long) bcc_ptr - (unsigned long) smb_buf) % 2) { ++bcc_ptr; diff --git a/fs/cifs/smbdes.c b/fs/cifs/smbdes.c deleted file mode 100644 index 04721485925d..000000000000 --- a/fs/cifs/smbdes.c +++ /dev/null @@ -1,418 +0,0 @@ -/* - Unix SMB/Netbios implementation. - Version 1.9. - - a partial implementation of DES designed for use in the - SMB authentication protocol - - Copyright (C) Andrew Tridgell 1998 - Modified by Steve French (sfrench@us.ibm.com) 2002,2004 - - This program is free software; you can redistribute it and/or modify - it under the terms of the GNU General Public License as published by - the Free Software Foundation; either version 2 of the License, or - (at your option) any later version. - - This program is distributed in the hope that it will be useful, - but WITHOUT ANY WARRANTY; without even the implied warranty of - MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the - GNU General Public License for more details. - - You should have received a copy of the GNU General Public License - along with this program; if not, write to the Free Software - Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. -*/ - -/* NOTES: - - This code makes no attempt to be fast! In fact, it is a very - slow implementation - - This code is NOT a complete DES implementation. It implements only - the minimum necessary for SMB authentication, as used by all SMB - products (including every copy of Microsoft Windows95 ever sold) - - In particular, it can only do a unchained forward DES pass. This - means it is not possible to use this code for encryption/decryption - of data, instead it is only useful as a "hash" algorithm. - - There is no entry point into this code that allows normal DES operation. - - I believe this means that this code does not come under ITAR - regulations but this is NOT a legal opinion. If you are concerned - about the applicability of ITAR regulations to this code then you - should confirm it for yourself (and maybe let me know if you come - up with a different answer to the one above) -*/ -#include <linux/slab.h> -#define uchar unsigned char - -static uchar perm1[56] = { 57, 49, 41, 33, 25, 17, 9, - 1, 58, 50, 42, 34, 26, 18, - 10, 2, 59, 51, 43, 35, 27, - 19, 11, 3, 60, 52, 44, 36, - 63, 55, 47, 39, 31, 23, 15, - 7, 62, 54, 46, 38, 30, 22, - 14, 6, 61, 53, 45, 37, 29, - 21, 13, 5, 28, 20, 12, 4 -}; - -static uchar perm2[48] = { 14, 17, 11, 24, 1, 5, - 3, 28, 15, 6, 21, 10, - 23, 19, 12, 4, 26, 8, - 16, 7, 27, 20, 13, 2, - 41, 52, 31, 37, 47, 55, - 30, 40, 51, 45, 33, 48, - 44, 49, 39, 56, 34, 53, - 46, 42, 50, 36, 29, 32 -}; - -static uchar perm3[64] = { 58, 50, 42, 34, 26, 18, 10, 2, - 60, 52, 44, 36, 28, 20, 12, 4, - 62, 54, 46, 38, 30, 22, 14, 6, - 64, 56, 48, 40, 32, 24, 16, 8, - 57, 49, 41, 33, 25, 17, 9, 1, - 59, 51, 43, 35, 27, 19, 11, 3, - 61, 53, 45, 37, 29, 21, 13, 5, - 63, 55, 47, 39, 31, 23, 15, 7 -}; - -static uchar perm4[48] = { 32, 1, 2, 3, 4, 5, - 4, 5, 6, 7, 8, 9, - 8, 9, 10, 11, 12, 13, - 12, 13, 14, 15, 16, 17, - 16, 17, 18, 19, 20, 21, - 20, 21, 22, 23, 24, 25, - 24, 25, 26, 27, 28, 29, - 28, 29, 30, 31, 32, 1 -}; - -static uchar perm5[32] = { 16, 7, 20, 21, - 29, 12, 28, 17, - 1, 15, 23, 26, - 5, 18, 31, 10, - 2, 8, 24, 14, - 32, 27, 3, 9, - 19, 13, 30, 6, - 22, 11, 4, 25 -}; - -static uchar perm6[64] = { 40, 8, 48, 16, 56, 24, 64, 32, - 39, 7, 47, 15, 55, 23, 63, 31, - 38, 6, 46, 14, 54, 22, 62, 30, - 37, 5, 45, 13, 53, 21, 61, 29, - 36, 4, 44, 12, 52, 20, 60, 28, - 35, 3, 43, 11, 51, 19, 59, 27, - 34, 2, 42, 10, 50, 18, 58, 26, - 33, 1, 41, 9, 49, 17, 57, 25 -}; - -static uchar sc[16] = { 1, 1, 2, 2, 2, 2, 2, 2, 1, 2, 2, 2, 2, 2, 2, 1 }; - -static uchar sbox[8][4][16] = { - {{14, 4, 13, 1, 2, 15, 11, 8, 3, 10, 6, 12, 5, 9, 0, 7}, - {0, 15, 7, 4, 14, 2, 13, 1, 10, 6, 12, 11, 9, 5, 3, 8}, - {4, 1, 14, 8, 13, 6, 2, 11, 15, 12, 9, 7, 3, 10, 5, 0}, - {15, 12, 8, 2, 4, 9, 1, 7, 5, 11, 3, 14, 10, 0, 6, 13} }, - - {{15, 1, 8, 14, 6, 11, 3, 4, 9, 7, 2, 13, 12, 0, 5, 10}, - {3, 13, 4, 7, 15, 2, 8, 14, 12, 0, 1, 10, 6, 9, 11, 5}, - {0, 14, 7, 11, 10, 4, 13, 1, 5, 8, 12, 6, 9, 3, 2, 15}, - {13, 8, 10, 1, 3, 15, 4, 2, 11, 6, 7, 12, 0, 5, 14, 9} }, - - {{10, 0, 9, 14, 6, 3, 15, 5, 1, 13, 12, 7, 11, 4, 2, 8}, - {13, 7, 0, 9, 3, 4, 6, 10, 2, 8, 5, 14, 12, 11, 15, 1}, - {13, 6, 4, 9, 8, 15, 3, 0, 11, 1, 2, 12, 5, 10, 14, 7}, - {1, 10, 13, 0, 6, 9, 8, 7, 4, 15, 14, 3, 11, 5, 2, 12} }, - - {{7, 13, 14, 3, 0, 6, 9, 10, 1, 2, 8, 5, 11, 12, 4, 15}, - {13, 8, 11, 5, 6, 15, 0, 3, 4, 7, 2, 12, 1, 10, 14, 9}, - {10, 6, 9, 0, 12, 11, 7, 13, 15, 1, 3, 14, 5, 2, 8, 4}, - {3, 15, 0, 6, 10, 1, 13, 8, 9, 4, 5, 11, 12, 7, 2, 14} }, - - {{2, 12, 4, 1, 7, 10, 11, 6, 8, 5, 3, 15, 13, 0, 14, 9}, - {14, 11, 2, 12, 4, 7, 13, 1, 5, 0, 15, 10, 3, 9, 8, 6}, - {4, 2, 1, 11, 10, 13, 7, 8, 15, 9, 12, 5, 6, 3, 0, 14}, - {11, 8, 12, 7, 1, 14, 2, 13, 6, 15, 0, 9, 10, 4, 5, 3} }, - - {{12, 1, 10, 15, 9, 2, 6, 8, 0, 13, 3, 4, 14, 7, 5, 11}, - {10, 15, 4, 2, 7, 12, 9, 5, 6, 1, 13, 14, 0, 11, 3, 8}, - {9, 14, 15, 5, 2, 8, 12, 3, 7, 0, 4, 10, 1, 13, 11, 6}, - {4, 3, 2, 12, 9, 5, 15, 10, 11, 14, 1, 7, 6, 0, 8, 13} }, - - {{4, 11, 2, 14, 15, 0, 8, 13, 3, 12, 9, 7, 5, 10, 6, 1}, - {13, 0, 11, 7, 4, 9, 1, 10, 14, 3, 5, 12, 2, 15, 8, 6}, - {1, 4, 11, 13, 12, 3, 7, 14, 10, 15, 6, 8, 0, 5, 9, 2}, - {6, 11, 13, 8, 1, 4, 10, 7, 9, 5, 0, 15, 14, 2, 3, 12} }, - - {{13, 2, 8, 4, 6, 15, 11, 1, 10, 9, 3, 14, 5, 0, 12, 7}, - {1, 15, 13, 8, 10, 3, 7, 4, 12, 5, 6, 11, 0, 14, 9, 2}, - {7, 11, 4, 1, 9, 12, 14, 2, 0, 6, 10, 13, 15, 3, 5, 8}, - {2, 1, 14, 7, 4, 10, 8, 13, 15, 12, 9, 0, 3, 5, 6, 11} } -}; - -static void -permute(char *out, char *in, uchar *p, int n) -{ - int i; - for (i = 0; i < n; i++) - out[i] = in[p[i] - 1]; -} - -static void -lshift(char *d, int count, int n) -{ - char out[64]; - int i; - for (i = 0; i < n; i++) - out[i] = d[(i + count) % n]; - for (i = 0; i < n; i++) - d[i] = out[i]; -} - -static void -concat(char *out, char *in1, char *in2, int l1, int l2) -{ - while (l1--) - *out++ = *in1++; - while (l2--) - *out++ = *in2++; -} - -static void -xor(char *out, char *in1, char *in2, int n) -{ - int i; - for (i = 0; i < n; i++) - out[i] = in1[i] ^ in2[i]; -} - -static void -dohash(char *out, char *in, char *key, int forw) -{ - int i, j, k; - char *pk1; - char c[28]; - char d[28]; - char *cd; - char (*ki)[48]; - char *pd1; - char l[32], r[32]; - char *rl; - - /* Have to reduce stack usage */ - pk1 = kmalloc(56+56+64+64, GFP_KERNEL); - if (pk1 == NULL) - return; - - ki = kmalloc(16*48, GFP_KERNEL); - if (ki == NULL) { - kfree(pk1); - return; - } - - cd = pk1 + 56; - pd1 = cd + 56; - rl = pd1 + 64; - - permute(pk1, key, perm1, 56); - - for (i = 0; i < 28; i++) - c[i] = pk1[i]; - for (i = 0; i < 28; i++) - d[i] = pk1[i + 28]; - - for (i = 0; i < 16; i++) { - lshift(c, sc[i], 28); - lshift(d, sc[i], 28); - - concat(cd, c, d, 28, 28); - permute(ki[i], cd, perm2, 48); - } - - permute(pd1, in, perm3, 64); - - for (j = 0; j < 32; j++) { - l[j] = pd1[j]; - r[j] = pd1[j + 32]; - } - - for (i = 0; i < 16; i++) { - char *er; /* er[48] */ - char *erk; /* erk[48] */ - char b[8][6]; - char *cb; /* cb[32] */ - char *pcb; /* pcb[32] */ - char *r2; /* r2[32] */ - - er = kmalloc(48+48+32+32+32, GFP_KERNEL); - if (er == NULL) { - kfree(pk1); - kfree(ki); - return; - } - erk = er+48; - cb = erk+48; - pcb = cb+32; - r2 = pcb+32; - - permute(er, r, perm4, 48); - - xor(erk, er, ki[forw ? i : 15 - i], 48); - - for (j = 0; j < 8; j++) - for (k = 0; k < 6; k++) - b[j][k] = erk[j * 6 + k]; - - for (j = 0; j < 8; j++) { - int m, n; - m = (b[j][0] << 1) | b[j][5]; - - n = (b[j][1] << 3) | (b[j][2] << 2) | (b[j][3] << - 1) | b[j][4]; - - for (k = 0; k < 4; k++) - b[j][k] = - (sbox[j][m][n] & (1 << (3 - k))) ? 1 : 0; - } - - for (j = 0; j < 8; j++) - for (k = 0; k < 4; k++) - cb[j * 4 + k] = b[j][k]; - permute(pcb, cb, perm5, 32); - - xor(r2, l, pcb, 32); - - for (j = 0; j < 32; j++) - l[j] = r[j]; - - for (j = 0; j < 32; j++) - r[j] = r2[j]; - - kfree(er); - } - - concat(rl, r, l, 32, 32); - - permute(out, rl, perm6, 64); - kfree(pk1); - kfree(ki); -} - -static void -str_to_key(unsigned char *str, unsigned char *key) -{ - int i; - - key[0] = str[0] >> 1; - key[1] = ((str[0] & 0x01) << 6) | (str[1] >> 2); - key[2] = ((str[1] & 0x03) << 5) | (str[2] >> 3); - key[3] = ((str[2] & 0x07) << 4) | (str[3] >> 4); - key[4] = ((str[3] & 0x0F) << 3) | (str[4] >> 5); - key[5] = ((str[4] & 0x1F) << 2) | (str[5] >> 6); - key[6] = ((str[5] & 0x3F) << 1) | (str[6] >> 7); - key[7] = str[6] & 0x7F; - for (i = 0; i < 8; i++) - key[i] = (key[i] << 1); -} - -static void -smbhash(unsigned char *out, const unsigned char *in, unsigned char *key, - int forw) -{ - int i; - char *outb; /* outb[64] */ - char *inb; /* inb[64] */ - char *keyb; /* keyb[64] */ - unsigned char key2[8]; - - outb = kmalloc(64 * 3, GFP_KERNEL); - if (outb == NULL) - return; - - inb = outb + 64; - keyb = inb + 64; - - str_to_key(key, key2); - - for (i = 0; i < 64; i++) { - inb[i] = (in[i / 8] & (1 << (7 - (i % 8)))) ? 1 : 0; - keyb[i] = (key2[i / 8] & (1 << (7 - (i % 8)))) ? 1 : 0; - outb[i] = 0; - } - - dohash(outb, inb, keyb, forw); - - for (i = 0; i < 8; i++) - out[i] = 0; - - for (i = 0; i < 64; i++) { - if (outb[i]) - out[i / 8] |= (1 << (7 - (i % 8))); - } - kfree(outb); -} - -void -E_P16(unsigned char *p14, unsigned char *p16) -{ - unsigned char sp8[8] = - { 0x4b, 0x47, 0x53, 0x21, 0x40, 0x23, 0x24, 0x25 }; - smbhash(p16, sp8, p14, 1); - smbhash(p16 + 8, sp8, p14 + 7, 1); -} - -void -E_P24(unsigned char *p21, const unsigned char *c8, unsigned char *p24) -{ - smbhash(p24, c8, p21, 1); - smbhash(p24 + 8, c8, p21 + 7, 1); - smbhash(p24 + 16, c8, p21 + 14, 1); -} - -#if 0 /* currently unused */ -static void -D_P16(unsigned char *p14, unsigned char *in, unsigned char *out) -{ - smbhash(out, in, p14, 0); - smbhash(out + 8, in + 8, p14 + 7, 0); -} - -static void -E_old_pw_hash(unsigned char *p14, unsigned char *in, unsigned char *out) -{ - smbhash(out, in, p14, 1); - smbhash(out + 8, in + 8, p14 + 7, 1); -} -/* these routines are currently unneeded, but may be - needed later */ -void -cred_hash1(unsigned char *out, unsigned char *in, unsigned char *key) -{ - unsigned char buf[8]; - - smbhash(buf, in, key, 1); - smbhash(out, buf, key + 9, 1); -} - -void -cred_hash2(unsigned char *out, unsigned char *in, unsigned char *key) -{ - unsigned char buf[8]; - static unsigned char key2[8]; - - smbhash(buf, in, key, 1); - key2[0] = key[7]; - smbhash(out, buf, key2, 1); -} - -void -cred_hash3(unsigned char *out, unsigned char *in, unsigned char *key, int forw) -{ - static unsigned char key2[8]; - - smbhash(out, in, key, forw); - key2[0] = key[7]; - smbhash(out + 8, in + 8, key2, forw); -} -#endif /* unneeded routines */ diff --git a/fs/cifs/smbencrypt.c b/fs/cifs/smbencrypt.c index b5041c849981..1525d5e662b6 100644 --- a/fs/cifs/smbencrypt.c +++ b/fs/cifs/smbencrypt.c @@ -47,6 +47,88 @@ #define SSVALX(buf,pos,val) (CVAL(buf,pos)=(val)&0xFF,CVAL(buf,pos+1)=(val)>>8) #define SSVAL(buf,pos,val) SSVALX((buf),(pos),((__u16)(val))) +static void +str_to_key(unsigned char *str, unsigned char *key) +{ + int i; + + key[0] = str[0] >> 1; + key[1] = ((str[0] & 0x01) << 6) | (str[1] >> 2); + key[2] = ((str[1] & 0x03) << 5) | (str[2] >> 3); + key[3] = ((str[2] & 0x07) << 4) | (str[3] >> 4); + key[4] = ((str[3] & 0x0F) << 3) | (str[4] >> 5); + key[5] = ((str[4] & 0x1F) << 2) | (str[5] >> 6); + key[6] = ((str[5] & 0x3F) << 1) | (str[6] >> 7); + key[7] = str[6] & 0x7F; + for (i = 0; i < 8; i++) + key[i] = (key[i] << 1); +} + +static int +smbhash(unsigned char *out, const unsigned char *in, unsigned char *key) +{ + int rc; + unsigned char key2[8]; + struct crypto_blkcipher *tfm_des; + struct scatterlist sgin, sgout; + struct blkcipher_desc desc; + + str_to_key(key, key2); + + tfm_des = crypto_alloc_blkcipher("ecb(des)", 0, CRYPTO_ALG_ASYNC); + if (IS_ERR(tfm_des)) { + rc = PTR_ERR(tfm_des); + cERROR(1, "could not allocate des crypto API\n"); + goto smbhash_err; + } + + desc.tfm = tfm_des; + + crypto_blkcipher_setkey(tfm_des, key2, 8); + + sg_init_one(&sgin, in, 8); + sg_init_one(&sgout, out, 8); + + rc = crypto_blkcipher_encrypt(&desc, &sgout, &sgin, 8); + if (rc) { + cERROR(1, "could not encrypt crypt key rc: %d\n", rc); + crypto_free_blkcipher(tfm_des); + goto smbhash_err; + } + +smbhash_err: + return rc; +} + +static int +E_P16(unsigned char *p14, unsigned char *p16) +{ + int rc; + unsigned char sp8[8] = + { 0x4b, 0x47, 0x53, 0x21, 0x40, 0x23, 0x24, 0x25 }; + + rc = smbhash(p16, sp8, p14); + if (rc) + return rc; + rc = smbhash(p16 + 8, sp8, p14 + 7); + return rc; +} + +static int +E_P24(unsigned char *p21, const unsigned char *c8, unsigned char *p24) +{ + int rc; + + rc = smbhash(p24, c8, p21); + if (rc) + return rc; + rc = smbhash(p24 + 8, c8, p21 + 7); + if (rc) + return rc; + rc = smbhash(p24 + 16, c8, p21 + 14); + return rc; +} + /* produce a md4 message digest from data of length n bytes */ int mdfour(unsigned char *md4_hash, unsigned char *link_str, int link_len) @@ -87,40 +169,30 @@ mdfour_err: return rc; } -/* Does the des encryption from the NT or LM MD4 hash. */ -static void -SMBOWFencrypt(unsigned char passwd[16], const unsigned char *c8, - unsigned char p24[24]) -{ - unsigned char p21[21]; - - memset(p21, '\0', 21); - - memcpy(p21, passwd, 16); - E_P24(p21, c8, p24); -} - /* This implements the X/Open SMB password encryption It takes a password, a 8 byte "crypt key" and puts 24 bytes of encrypted password into p24 */ /* Note that password must be uppercased and null terminated */ -void +int SMBencrypt(unsigned char *passwd, const unsigned char *c8, unsigned char *p24) { - unsigned char p14[15], p21[21]; + int rc; + unsigned char p14[14], p16[16], p21[21]; - memset(p21, '\0', 21); memset(p14, '\0', 14); - strncpy((char *) p14, (char *) passwd, 14); + memset(p16, '\0', 16); + memset(p21, '\0', 21); -/* strupper((char *)p14); *//* BB at least uppercase the easy range */ - E_P16(p14, p21); + memcpy(p14, passwd, 14); + rc = E_P16(p14, p16); + if (rc) + return rc; - SMBOWFencrypt(p21, c8, p24); + memcpy(p21, p16, 16); + rc = E_P24(p21, c8, p24); - memset(p14, 0, 15); - memset(p21, 0, 21); + return rc; } /* Routines for Windows NT MD4 Hash functions. */ @@ -279,16 +351,18 @@ int SMBNTencrypt(unsigned char *passwd, unsigned char *c8, unsigned char *p24) { int rc; - unsigned char p21[21]; + unsigned char p16[16], p21[21]; + memset(p16, '\0', 16); memset(p21, '\0', 21); - rc = E_md4hash(passwd, p21); + rc = E_md4hash(passwd, p16); if (rc) { cFYI(1, "%s Can't generate NT hash, error: %d", __func__, rc); return rc; } - SMBOWFencrypt(p21, c8, p24); + memcpy(p21, p16, 16); + rc = E_P24(p21, c8, p24); return rc; } diff --git a/fs/cifs/transport.c b/fs/cifs/transport.c index 46d8756f2b24..147aa22c3c3a 100644 --- a/fs/cifs/transport.c +++ b/fs/cifs/transport.c @@ -129,7 +129,7 @@ smb_sendv(struct TCP_Server_Info *server, struct kvec *iov, int n_vec) unsigned int len = iov[0].iov_len; unsigned int total_len; int first_vec = 0; - unsigned int smb_buf_length = smb_buffer->smb_buf_length; + unsigned int smb_buf_length = be32_to_cpu(smb_buffer->smb_buf_length); struct socket *ssocket = server->ssocket; if (ssocket == NULL) @@ -144,17 +144,10 @@ smb_sendv(struct TCP_Server_Info *server, struct kvec *iov, int n_vec) else smb_msg.msg_flags = MSG_NOSIGNAL; - /* smb header is converted in header_assemble. bcc and rest of SMB word - area, and byte area if necessary, is converted to littleendian in - cifssmb.c and RFC1001 len is converted to bigendian in smb_send - Flags2 is converted in SendReceive */ - - total_len = 0; for (i = 0; i < n_vec; i++) total_len += iov[i].iov_len; - smb_buffer->smb_buf_length = cpu_to_be32(smb_buffer->smb_buf_length); cFYI(1, "Sending smb: total_len %d", total_len); dump_smb(smb_buffer, len); @@ -243,7 +236,7 @@ smb_sendv(struct TCP_Server_Info *server, struct kvec *iov, int n_vec) /* Don't want to modify the buffer as a side effect of this call. */ - smb_buffer->smb_buf_length = smb_buf_length; + smb_buffer->smb_buf_length = cpu_to_be32(smb_buf_length); return rc; } @@ -302,7 +295,7 @@ static int wait_for_free_request(struct TCP_Server_Info *server, return 0; } -static int allocate_mid(struct cifsSesInfo *ses, struct smb_hdr *in_buf, +static int allocate_mid(struct cifs_ses *ses, struct smb_hdr *in_buf, struct mid_q_entry **ppmidQ) { if (ses->server->tcpStatus == CifsExiting) { @@ -349,22 +342,24 @@ wait_for_response(struct TCP_Server_Info *server, struct mid_q_entry *midQ) * the result. Caller is responsible for dealing with timeouts. */ int -cifs_call_async(struct TCP_Server_Info *server, struct smb_hdr *in_buf, - mid_callback_t *callback, void *cbdata) +cifs_call_async(struct TCP_Server_Info *server, struct kvec *iov, + unsigned int nvec, mid_callback_t *callback, void *cbdata, + bool ignore_pend) { int rc; struct mid_q_entry *mid; + struct smb_hdr *hdr = (struct smb_hdr *)iov[0].iov_base; - rc = wait_for_free_request(server, CIFS_ASYNC_OP); + rc = wait_for_free_request(server, ignore_pend ? CIFS_ASYNC_OP : 0); if (rc) return rc; /* enable signing if server requires it */ - if (server->secMode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) - in_buf->Flags2 |= SMBFLG2_SECURITY_SIGNATURE; + if (server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) + hdr->Flags2 |= SMBFLG2_SECURITY_SIGNATURE; mutex_lock(&server->srv_mutex); - mid = AllocMidQEntry(in_buf, server); + mid = AllocMidQEntry(hdr, server); if (mid == NULL) { mutex_unlock(&server->srv_mutex); return -ENOMEM; @@ -375,7 +370,7 @@ cifs_call_async(struct TCP_Server_Info *server, struct smb_hdr *in_buf, list_add_tail(&mid->qhead, &server->pending_mid_q); spin_unlock(&GlobalMid_Lock); - rc = cifs_sign_smb(in_buf, server, &mid->sequence_number); + rc = cifs_sign_smb2(iov, nvec, server, &mid->sequence_number); if (rc) { mutex_unlock(&server->srv_mutex); goto out_err; @@ -387,7 +382,7 @@ cifs_call_async(struct TCP_Server_Info *server, struct smb_hdr *in_buf, #ifdef CONFIG_CIFS_STATS2 atomic_inc(&server->inSend); #endif - rc = smb_send(server, in_buf, in_buf->smb_buf_length); + rc = smb_sendv(server, iov, nvec); #ifdef CONFIG_CIFS_STATS2 atomic_dec(&server->inSend); mid->when_sent = jiffies; @@ -414,7 +409,7 @@ out_err: * */ int -SendReceiveNoRsp(const unsigned int xid, struct cifsSesInfo *ses, +SendReceiveNoRsp(const unsigned int xid, struct cifs_ses *ses, struct smb_hdr *in_buf, int flags) { int rc; @@ -422,7 +417,7 @@ SendReceiveNoRsp(const unsigned int xid, struct cifsSesInfo *ses, int resp_buf_type; iov[0].iov_base = (char *)in_buf; - iov[0].iov_len = in_buf->smb_buf_length + 4; + iov[0].iov_len = be32_to_cpu(in_buf->smb_buf_length) + 4; flags |= CIFS_NO_RESP; rc = SendReceive2(xid, ses, iov, 1, &resp_buf_type, flags); cFYI(DBG2, "SendRcvNoRsp flags %d rc %d", flags, rc); @@ -431,7 +426,7 @@ SendReceiveNoRsp(const unsigned int xid, struct cifsSesInfo *ses, } static int -sync_mid_result(struct mid_q_entry *mid, struct TCP_Server_Info *server) +cifs_sync_mid_result(struct mid_q_entry *mid, struct TCP_Server_Info *server) { int rc = 0; @@ -439,28 +434,21 @@ sync_mid_result(struct mid_q_entry *mid, struct TCP_Server_Info *server) mid->mid, mid->midState); spin_lock(&GlobalMid_Lock); - /* ensure that it's no longer on the pending_mid_q */ - list_del_init(&mid->qhead); - switch (mid->midState) { case MID_RESPONSE_RECEIVED: spin_unlock(&GlobalMid_Lock); return rc; - case MID_REQUEST_SUBMITTED: - /* socket is going down, reject all calls */ - if (server->tcpStatus == CifsExiting) { - cERROR(1, "%s: canceling mid=%d cmd=0x%x state=%d", - __func__, mid->mid, mid->command, mid->midState); - rc = -EHOSTDOWN; - break; - } case MID_RETRY_NEEDED: rc = -EAGAIN; break; case MID_RESPONSE_MALFORMED: rc = -EIO; break; + case MID_SHUTDOWN: + rc = -EHOSTDOWN; + break; default: + list_del_init(&mid->qhead); cERROR(1, "%s: invalid mid state mid=%d state=%d", __func__, mid->mid, mid->midState); rc = -EIO; @@ -488,10 +476,10 @@ send_nt_cancel(struct TCP_Server_Info *server, struct smb_hdr *in_buf, int rc = 0; /* -4 for RFC1001 length and +2 for BCC field */ - in_buf->smb_buf_length = sizeof(struct smb_hdr) - 4 + 2; + in_buf->smb_buf_length = cpu_to_be32(sizeof(struct smb_hdr) - 4 + 2); in_buf->Command = SMB_COM_NT_CANCEL; in_buf->WordCount = 0; - put_bcc_le(0, in_buf); + put_bcc(0, in_buf); mutex_lock(&server->srv_mutex); rc = cifs_sign_smb(in_buf, server, &mid->sequence_number); @@ -499,7 +487,7 @@ send_nt_cancel(struct TCP_Server_Info *server, struct smb_hdr *in_buf, mutex_unlock(&server->srv_mutex); return rc; } - rc = smb_send(server, in_buf, in_buf->smb_buf_length); + rc = smb_send(server, in_buf, be32_to_cpu(in_buf->smb_buf_length)); mutex_unlock(&server->srv_mutex); cFYI(1, "issued NT_CANCEL for mid %u, rc = %d", @@ -509,13 +497,31 @@ send_nt_cancel(struct TCP_Server_Info *server, struct smb_hdr *in_buf, } int -SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, +cifs_check_receive(struct mid_q_entry *mid, struct TCP_Server_Info *server, + bool log_error) +{ + dump_smb(mid->resp_buf, + min_t(u32, 92, be32_to_cpu(mid->resp_buf->smb_buf_length))); + + /* convert the length into a more usable form */ + if (server->sec_mode & (SECMODE_SIGN_REQUIRED | SECMODE_SIGN_ENABLED)) { + /* FIXME: add code to kill session */ + if (cifs_verify_signature(mid->resp_buf, server, + mid->sequence_number + 1) != 0) + cERROR(1, "Unexpected SMB signature"); + } + + /* BB special case reconnect tid and uid here? */ + return map_smb_to_linux_error(mid->resp_buf, log_error); +} + +int +SendReceive2(const unsigned int xid, struct cifs_ses *ses, struct kvec *iov, int n_vec, int *pRespBufType /* ret */, const int flags) { int rc = 0; int long_op; - unsigned int receive_len; struct mid_q_entry *midQ; struct smb_hdr *in_buf = iov[0].iov_base; @@ -605,66 +611,31 @@ SendReceive2(const unsigned int xid, struct cifsSesInfo *ses, cifs_small_buf_release(in_buf); - rc = sync_mid_result(midQ, ses->server); + rc = cifs_sync_mid_result(midQ, ses->server); if (rc != 0) { atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); return rc; } - receive_len = midQ->resp_buf->smb_buf_length; - - if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { - cERROR(1, "Frame too large received. Length: %d Xid: %d", - receive_len, xid); + if (!midQ->resp_buf || midQ->midState != MID_RESPONSE_RECEIVED) { rc = -EIO; + cFYI(1, "Bad MID state?"); goto out; } - /* rcvd frame is ok */ - - if (midQ->resp_buf && - (midQ->midState == MID_RESPONSE_RECEIVED)) { - - iov[0].iov_base = (char *)midQ->resp_buf; - if (midQ->largeBuf) - *pRespBufType = CIFS_LARGE_BUFFER; - else - *pRespBufType = CIFS_SMALL_BUFFER; - iov[0].iov_len = receive_len + 4; - - dump_smb(midQ->resp_buf, 80); - /* convert the length into a more usable form */ - if ((receive_len > 24) && - (ses->server->secMode & (SECMODE_SIGN_REQUIRED | - SECMODE_SIGN_ENABLED))) { - rc = cifs_verify_signature(midQ->resp_buf, - ses->server, - midQ->sequence_number+1); - if (rc) { - cERROR(1, "Unexpected SMB signature"); - /* BB FIXME add code to kill session */ - } - } + iov[0].iov_base = (char *)midQ->resp_buf; + iov[0].iov_len = be32_to_cpu(midQ->resp_buf->smb_buf_length) + 4; + if (midQ->largeBuf) + *pRespBufType = CIFS_LARGE_BUFFER; + else + *pRespBufType = CIFS_SMALL_BUFFER; - /* BB special case reconnect tid and uid here? */ - rc = map_smb_to_linux_error(midQ->resp_buf, - flags & CIFS_LOG_ERROR); - - /* convert ByteCount if necessary */ - if (receive_len >= sizeof(struct smb_hdr) - 4 - /* do not count RFC1001 header */ + - (2 * midQ->resp_buf->WordCount) + 2 /* bcc */ ) - put_bcc(get_bcc_le(midQ->resp_buf), midQ->resp_buf); - if ((flags & CIFS_NO_RESP) == 0) - midQ->resp_buf = NULL; /* mark it so buf will - not be freed by - delete_mid */ - } else { - rc = -EIO; - cFYI(1, "Bad MID state?"); - } + rc = cifs_check_receive(midQ, ses->server, flags & CIFS_LOG_ERROR); + /* mark it so buf will not be freed by delete_mid */ + if ((flags & CIFS_NO_RESP) == 0) + midQ->resp_buf = NULL; out: delete_mid(midQ); atomic_dec(&ses->server->inFlight); @@ -674,12 +645,11 @@ out: } int -SendReceive(const unsigned int xid, struct cifsSesInfo *ses, +SendReceive(const unsigned int xid, struct cifs_ses *ses, struct smb_hdr *in_buf, struct smb_hdr *out_buf, int *pbytes_returned, const int long_op) { int rc = 0; - unsigned int receive_len; struct mid_q_entry *midQ; if (ses == NULL) { @@ -698,9 +668,10 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, to the same server. We may make this configurable later or use ses->maxReq */ - if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { + if (be32_to_cpu(in_buf->smb_buf_length) > CIFSMaxBufSize + + MAX_CIFS_HDR_SIZE - 4) { cERROR(1, "Illegal length, greater than maximum frame, %d", - in_buf->smb_buf_length); + be32_to_cpu(in_buf->smb_buf_length)); return -EIO; } @@ -733,7 +704,7 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, #ifdef CONFIG_CIFS_STATS2 atomic_inc(&ses->server->inSend); #endif - rc = smb_send(ses->server, in_buf, in_buf->smb_buf_length); + rc = smb_send(ses->server, in_buf, be32_to_cpu(in_buf->smb_buf_length)); #ifdef CONFIG_CIFS_STATS2 atomic_dec(&ses->server->inSend); midQ->when_sent = jiffies; @@ -761,60 +732,23 @@ SendReceive(const unsigned int xid, struct cifsSesInfo *ses, spin_unlock(&GlobalMid_Lock); } - rc = sync_mid_result(midQ, ses->server); + rc = cifs_sync_mid_result(midQ, ses->server); if (rc != 0) { atomic_dec(&ses->server->inFlight); wake_up(&ses->server->request_q); return rc; } - receive_len = midQ->resp_buf->smb_buf_length; - - if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { - cERROR(1, "Frame too large received. Length: %d Xid: %d", - receive_len, xid); - rc = -EIO; - goto out; - } - - /* rcvd frame is ok */ - - if (midQ->resp_buf && out_buf - && (midQ->midState == MID_RESPONSE_RECEIVED)) { - out_buf->smb_buf_length = receive_len; - memcpy((char *)out_buf + 4, - (char *)midQ->resp_buf + 4, - receive_len); - - dump_smb(out_buf, 92); - /* convert the length into a more usable form */ - if ((receive_len > 24) && - (ses->server->secMode & (SECMODE_SIGN_REQUIRED | - SECMODE_SIGN_ENABLED))) { - rc = cifs_verify_signature(out_buf, - ses->server, - midQ->sequence_number+1); - if (rc) { - cERROR(1, "Unexpected SMB signature"); - /* BB FIXME add code to kill session */ - } - } - - *pbytes_returned = out_buf->smb_buf_length; - - /* BB special case reconnect tid and uid here? */ - rc = map_smb_to_linux_error(out_buf, 0 /* no log */ ); - - /* convert ByteCount if necessary */ - if (receive_len >= sizeof(struct smb_hdr) - 4 - /* do not count RFC1001 header */ + - (2 * out_buf->WordCount) + 2 /* bcc */ ) - put_bcc(get_bcc_le(midQ->resp_buf), midQ->resp_buf); - } else { + if (!midQ->resp_buf || !out_buf || + midQ->midState != MID_RESPONSE_RECEIVED) { rc = -EIO; cERROR(1, "Bad MID state?"); + goto out; } + *pbytes_returned = be32_to_cpu(midQ->resp_buf->smb_buf_length); + memcpy(out_buf, midQ->resp_buf, *pbytes_returned + 4); + rc = cifs_check_receive(midQ, ses->server, 0); out: delete_mid(midQ); atomic_dec(&ses->server->inFlight); @@ -827,12 +761,12 @@ out: blocking lock to return. */ static int -send_lock_cancel(const unsigned int xid, struct cifsTconInfo *tcon, +send_lock_cancel(const unsigned int xid, struct cifs_tcon *tcon, struct smb_hdr *in_buf, struct smb_hdr *out_buf) { int bytes_returned; - struct cifsSesInfo *ses = tcon->ses; + struct cifs_ses *ses = tcon->ses; LOCK_REQ *pSMB = (LOCK_REQ *)in_buf; /* We just modify the current in_buf to change @@ -849,15 +783,14 @@ send_lock_cancel(const unsigned int xid, struct cifsTconInfo *tcon, } int -SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, +SendReceiveBlockingLock(const unsigned int xid, struct cifs_tcon *tcon, struct smb_hdr *in_buf, struct smb_hdr *out_buf, int *pbytes_returned) { int rc = 0; int rstart = 0; - unsigned int receive_len; struct mid_q_entry *midQ; - struct cifsSesInfo *ses; + struct cifs_ses *ses; if (tcon == NULL || tcon->ses == NULL) { cERROR(1, "Null smb session"); @@ -877,9 +810,10 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, to the same server. We may make this configurable later or use ses->maxReq */ - if (in_buf->smb_buf_length > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE - 4) { + if (be32_to_cpu(in_buf->smb_buf_length) > CIFSMaxBufSize + + MAX_CIFS_HDR_SIZE - 4) { cERROR(1, "Illegal length, greater than maximum frame, %d", - in_buf->smb_buf_length); + be32_to_cpu(in_buf->smb_buf_length)); return -EIO; } @@ -910,7 +844,7 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, #ifdef CONFIG_CIFS_STATS2 atomic_inc(&ses->server->inSend); #endif - rc = smb_send(ses->server, in_buf, in_buf->smb_buf_length); + rc = smb_send(ses->server, in_buf, be32_to_cpu(in_buf->smb_buf_length)); #ifdef CONFIG_CIFS_STATS2 atomic_dec(&ses->server->inSend); midQ->when_sent = jiffies; @@ -973,56 +907,20 @@ SendReceiveBlockingLock(const unsigned int xid, struct cifsTconInfo *tcon, rstart = 1; } - rc = sync_mid_result(midQ, ses->server); + rc = cifs_sync_mid_result(midQ, ses->server); if (rc != 0) return rc; - receive_len = midQ->resp_buf->smb_buf_length; - if (receive_len > CIFSMaxBufSize + MAX_CIFS_HDR_SIZE) { - cERROR(1, "Frame too large received. Length: %d Xid: %d", - receive_len, xid); - rc = -EIO; - goto out; - } - /* rcvd frame is ok */ - - if ((out_buf == NULL) || (midQ->midState != MID_RESPONSE_RECEIVED)) { + if (out_buf == NULL || midQ->midState != MID_RESPONSE_RECEIVED) { rc = -EIO; cERROR(1, "Bad MID state?"); goto out; } - out_buf->smb_buf_length = receive_len; - memcpy((char *)out_buf + 4, - (char *)midQ->resp_buf + 4, - receive_len); - - dump_smb(out_buf, 92); - /* convert the length into a more usable form */ - if ((receive_len > 24) && - (ses->server->secMode & (SECMODE_SIGN_REQUIRED | - SECMODE_SIGN_ENABLED))) { - rc = cifs_verify_signature(out_buf, - ses->server, - midQ->sequence_number+1); - if (rc) { - cERROR(1, "Unexpected SMB signature"); - /* BB FIXME add code to kill session */ - } - } - - *pbytes_returned = out_buf->smb_buf_length; - - /* BB special case reconnect tid and uid here? */ - rc = map_smb_to_linux_error(out_buf, 0 /* no log */ ); - - /* convert ByteCount if necessary */ - if (receive_len >= sizeof(struct smb_hdr) - 4 - /* do not count RFC1001 header */ + - (2 * out_buf->WordCount) + 2 /* bcc */ ) - put_bcc(get_bcc_le(out_buf), out_buf); - + *pbytes_returned = be32_to_cpu(midQ->resp_buf->smb_buf_length); + memcpy(out_buf, midQ->resp_buf, *pbytes_returned + 4); + rc = cifs_check_receive(midQ, ses->server, 0); out: delete_mid(midQ); if (rstart && rc == -EACCES) diff --git a/fs/cifs/xattr.c b/fs/cifs/xattr.c index eae2a1491608..2a22fb2989e4 100644 --- a/fs/cifs/xattr.c +++ b/fs/cifs/xattr.c @@ -49,7 +49,7 @@ int cifs_removexattr(struct dentry *direntry, const char *ea_name) int xid; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct super_block *sb; char *full_path = NULL; @@ -109,9 +109,10 @@ int cifs_setxattr(struct dentry *direntry, const char *ea_name, int xid; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct super_block *sb; char *full_path; + struct cifs_ntsd *pacl; if (direntry == NULL) return -EIO; @@ -166,6 +167,25 @@ int cifs_setxattr(struct dentry *direntry, const char *ea_name, rc = CIFSSMBSetEA(xid, pTcon, full_path, ea_name, ea_value, (__u16)value_size, cifs_sb->local_nls, cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); + } else if (strncmp(ea_name, CIFS_XATTR_CIFS_ACL, + strlen(CIFS_XATTR_CIFS_ACL)) == 0) { + pacl = kmalloc(value_size, GFP_KERNEL); + if (!pacl) { + cFYI(1, "%s: Can't allocate memory for ACL", + __func__); + rc = -ENOMEM; + } else { +#ifdef CONFIG_CIFS_ACL + memcpy(pacl, ea_value, value_size); + rc = set_cifs_acl(pacl, value_size, + direntry->d_inode, full_path); + if (rc == 0) /* force revalidate of the inode */ + CIFS_I(direntry->d_inode)->time = 0; + kfree(pacl); +#else + cFYI(1, "Set CIFS ACL not supported yet"); +#endif /* CONFIG_CIFS_ACL */ + } } else { int temp; temp = strncmp(ea_name, POSIX_ACL_XATTR_ACCESS, @@ -220,7 +240,7 @@ ssize_t cifs_getxattr(struct dentry *direntry, const char *ea_name, int xid; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct super_block *sb; char *full_path; @@ -352,7 +372,7 @@ ssize_t cifs_listxattr(struct dentry *direntry, char *data, size_t buf_size) int xid; struct cifs_sb_info *cifs_sb; struct tcon_link *tlink; - struct cifsTconInfo *pTcon; + struct cifs_tcon *pTcon; struct super_block *sb; char *full_path; diff --git a/fs/compat.c b/fs/compat.c index 72fe6cda9108..0ea00832de23 100644 --- a/fs/compat.c +++ b/fs/compat.c @@ -1306,241 +1306,6 @@ compat_sys_openat(unsigned int dfd, const char __user *filename, int flags, int return do_sys_open(dfd, filename, flags, mode); } -/* - * compat_count() counts the number of arguments/envelopes. It is basically - * a copy of count() from fs/exec.c, except that it works with 32 bit argv - * and envp pointers. - */ -static int compat_count(compat_uptr_t __user *argv, int max) -{ - int i = 0; - - if (argv != NULL) { - for (;;) { - compat_uptr_t p; - - if (get_user(p, argv)) - return -EFAULT; - if (!p) - break; - argv++; - if (i++ >= max) - return -E2BIG; - - if (fatal_signal_pending(current)) - return -ERESTARTNOHAND; - cond_resched(); - } - } - return i; -} - -/* - * compat_copy_strings() is basically a copy of copy_strings() from fs/exec.c - * except that it works with 32 bit argv and envp pointers. - */ -static int compat_copy_strings(int argc, compat_uptr_t __user *argv, - struct linux_binprm *bprm) -{ - struct page *kmapped_page = NULL; - char *kaddr = NULL; - unsigned long kpos = 0; - int ret; - - while (argc-- > 0) { - compat_uptr_t str; - int len; - unsigned long pos; - - if (get_user(str, argv+argc) || - !(len = strnlen_user(compat_ptr(str), MAX_ARG_STRLEN))) { - ret = -EFAULT; - goto out; - } - - if (len > MAX_ARG_STRLEN) { - ret = -E2BIG; - goto out; - } - - /* We're going to work our way backwords. */ - pos = bprm->p; - str += len; - bprm->p -= len; - - while (len > 0) { - int offset, bytes_to_copy; - - if (fatal_signal_pending(current)) { - ret = -ERESTARTNOHAND; - goto out; - } - cond_resched(); - - offset = pos % PAGE_SIZE; - if (offset == 0) - offset = PAGE_SIZE; - - bytes_to_copy = offset; - if (bytes_to_copy > len) - bytes_to_copy = len; - - offset -= bytes_to_copy; - pos -= bytes_to_copy; - str -= bytes_to_copy; - len -= bytes_to_copy; - - if (!kmapped_page || kpos != (pos & PAGE_MASK)) { - struct page *page; - - page = get_arg_page(bprm, pos, 1); - if (!page) { - ret = -E2BIG; - goto out; - } - - if (kmapped_page) { - flush_kernel_dcache_page(kmapped_page); - kunmap(kmapped_page); - put_page(kmapped_page); - } - kmapped_page = page; - kaddr = kmap(kmapped_page); - kpos = pos & PAGE_MASK; - flush_cache_page(bprm->vma, kpos, - page_to_pfn(kmapped_page)); - } - if (copy_from_user(kaddr+offset, compat_ptr(str), - bytes_to_copy)) { - ret = -EFAULT; - goto out; - } - } - } - ret = 0; -out: - if (kmapped_page) { - flush_kernel_dcache_page(kmapped_page); - kunmap(kmapped_page); - put_page(kmapped_page); - } - return ret; -} - -/* - * compat_do_execve() is mostly a copy of do_execve(), with the exception - * that it processes 32 bit argv and envp pointers. - */ -int compat_do_execve(char * filename, - compat_uptr_t __user *argv, - compat_uptr_t __user *envp, - struct pt_regs * regs) -{ - struct linux_binprm *bprm; - struct file *file; - struct files_struct *displaced; - bool clear_in_exec; - int retval; - - retval = unshare_files(&displaced); - if (retval) - goto out_ret; - - retval = -ENOMEM; - bprm = kzalloc(sizeof(*bprm), GFP_KERNEL); - if (!bprm) - goto out_files; - - retval = prepare_bprm_creds(bprm); - if (retval) - goto out_free; - - retval = check_unsafe_exec(bprm); - if (retval < 0) - goto out_free; - clear_in_exec = retval; - current->in_execve = 1; - - file = open_exec(filename); - retval = PTR_ERR(file); - if (IS_ERR(file)) - goto out_unmark; - - sched_exec(); - - bprm->file = file; - bprm->filename = filename; - bprm->interp = filename; - - retval = bprm_mm_init(bprm); - if (retval) - goto out_file; - - bprm->argc = compat_count(argv, MAX_ARG_STRINGS); - if ((retval = bprm->argc) < 0) - goto out; - - bprm->envc = compat_count(envp, MAX_ARG_STRINGS); - if ((retval = bprm->envc) < 0) - goto out; - - retval = prepare_binprm(bprm); - if (retval < 0) - goto out; - - retval = copy_strings_kernel(1, &bprm->filename, bprm); - if (retval < 0) - goto out; - - bprm->exec = bprm->p; - retval = compat_copy_strings(bprm->envc, envp, bprm); - if (retval < 0) - goto out; - - retval = compat_copy_strings(bprm->argc, argv, bprm); - if (retval < 0) - goto out; - - retval = search_binary_handler(bprm, regs); - if (retval < 0) - goto out; - - /* execve succeeded */ - current->fs->in_exec = 0; - current->in_execve = 0; - acct_update_integrals(current); - free_bprm(bprm); - if (displaced) - put_files_struct(displaced); - return retval; - -out: - if (bprm->mm) { - acct_arg_size(bprm, 0); - mmput(bprm->mm); - } - -out_file: - if (bprm->file) { - allow_write_access(bprm->file); - fput(bprm->file); - } - -out_unmark: - if (clear_in_exec) - current->fs->in_exec = 0; - current->in_execve = 0; - -out_free: - free_bprm(bprm); - -out_files: - if (displaced) - reset_files_struct(displaced); -out_ret: - return retval; -} - #define __COMPAT_NFDBITS (8 * sizeof(compat_ulong_t)) static int poll_select_copy_remaining(struct timespec *end_time, void __user *p, diff --git a/fs/configfs/dir.c b/fs/configfs/dir.c index 90ff3cb10de3..9a37a9b6de3a 100644 --- a/fs/configfs/dir.c +++ b/fs/configfs/dir.c @@ -53,11 +53,14 @@ DEFINE_SPINLOCK(configfs_dirent_lock); static void configfs_d_iput(struct dentry * dentry, struct inode * inode) { - struct configfs_dirent * sd = dentry->d_fsdata; + struct configfs_dirent *sd = dentry->d_fsdata; if (sd) { BUG_ON(sd->s_dentry != dentry); + /* Coordinate with configfs_readdir */ + spin_lock(&configfs_dirent_lock); sd->s_dentry = NULL; + spin_unlock(&configfs_dirent_lock); configfs_put(sd); } iput(inode); @@ -689,7 +692,8 @@ static int create_default_group(struct config_group *parent_group, sd = child->d_fsdata; sd->s_type |= CONFIGFS_USET_DEFAULT; } else { - d_delete(child); + BUG_ON(child->d_inode); + d_drop(child); dput(child); } } @@ -990,7 +994,7 @@ static int configfs_dump(struct configfs_dirent *sd, int level) * This describes these functions and their helpers. * * Allow another kernel system to depend on a config_item. If this - * happens, the item cannot go away until the dependant can live without + * happens, the item cannot go away until the dependent can live without * it. The idea is to give client modules as simple an interface as * possible. When a system asks them to depend on an item, they just * call configfs_depend_item(). If the item is live and the client @@ -1545,7 +1549,7 @@ static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir struct configfs_dirent * parent_sd = dentry->d_fsdata; struct configfs_dirent *cursor = filp->private_data; struct list_head *p, *q = &cursor->s_sibling; - ino_t ino; + ino_t ino = 0; int i = filp->f_pos; switch (i) { @@ -1573,6 +1577,7 @@ static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir struct configfs_dirent *next; const char * name; int len; + struct inode *inode = NULL; next = list_entry(p, struct configfs_dirent, s_sibling); @@ -1581,9 +1586,28 @@ static int configfs_readdir(struct file * filp, void * dirent, filldir_t filldir name = configfs_get_name(next); len = strlen(name); - if (next->s_dentry) - ino = next->s_dentry->d_inode->i_ino; - else + + /* + * We'll have a dentry and an inode for + * PINNED items and for open attribute + * files. We lock here to prevent a race + * with configfs_d_iput() clearing + * s_dentry before calling iput(). + * + * Why do we go to the trouble? If + * someone has an attribute file open, + * the inode number should match until + * they close it. Beyond that, we don't + * care. + */ + spin_lock(&configfs_dirent_lock); + dentry = next->s_dentry; + if (dentry) + inode = dentry->d_inode; + if (inode) + ino = inode->i_ino; + spin_unlock(&configfs_dirent_lock); + if (!inode) ino = iunique(configfs_sb, 2); if (filldir(dirent, name, len, filp->f_pos, ino, @@ -1683,7 +1707,8 @@ int configfs_register_subsystem(struct configfs_subsystem *subsys) err = configfs_attach_group(sd->s_element, &group->cg_item, dentry); if (err) { - d_delete(dentry); + BUG_ON(dentry->d_inode); + d_drop(dentry); dput(dentry); } else { spin_lock(&configfs_dirent_lock); diff --git a/fs/dcache.c b/fs/dcache.c index ad25c4cec7d5..37f72ee5bf7c 100644 --- a/fs/dcache.c +++ b/fs/dcache.c @@ -35,6 +35,7 @@ #include <linux/hardirq.h> #include <linux/bit_spinlock.h> #include <linux/rculist_bl.h> +#include <linux/prefetch.h> #include "internal.h" /* @@ -99,12 +100,9 @@ static struct kmem_cache *dentry_cache __read_mostly; static unsigned int d_hash_mask __read_mostly; static unsigned int d_hash_shift __read_mostly; -struct dcache_hash_bucket { - struct hlist_bl_head head; -}; -static struct dcache_hash_bucket *dentry_hashtable __read_mostly; +static struct hlist_bl_head *dentry_hashtable __read_mostly; -static inline struct dcache_hash_bucket *d_hash(struct dentry *parent, +static inline struct hlist_bl_head *d_hash(struct dentry *parent, unsigned long hash) { hash += ((unsigned long) parent ^ GOLDEN_RATIO_PRIME) / L1_CACHE_BYTES; @@ -112,16 +110,6 @@ static inline struct dcache_hash_bucket *d_hash(struct dentry *parent, return dentry_hashtable + (hash & D_HASHMASK); } -static inline void spin_lock_bucket(struct dcache_hash_bucket *b) -{ - bit_spin_lock(0, (unsigned long *)&b->head.first); -} - -static inline void spin_unlock_bucket(struct dcache_hash_bucket *b) -{ - __bit_spin_unlock(0, (unsigned long *)&b->head.first); -} - /* Statistics gathering. */ struct dentry_stat_t dentry_stat = { .age_limit = 45, @@ -167,8 +155,8 @@ static void d_free(struct dentry *dentry) if (dentry->d_op && dentry->d_op->d_release) dentry->d_op->d_release(dentry); - /* if dentry was never inserted into hash, immediate free is OK */ - if (hlist_bl_unhashed(&dentry->d_hash)) + /* if dentry was never visible to RCU, immediate free is OK */ + if (!(dentry->d_flags & DCACHE_RCUACCESS)) __d_free(&dentry->d_u.d_rcu); else call_rcu(&dentry->d_u.d_rcu, __d_free); @@ -330,28 +318,19 @@ static struct dentry *d_kill(struct dentry *dentry, struct dentry *parent) */ void __d_drop(struct dentry *dentry) { - if (!(dentry->d_flags & DCACHE_UNHASHED)) { - if (unlikely(dentry->d_flags & DCACHE_DISCONNECTED)) { - bit_spin_lock(0, - (unsigned long *)&dentry->d_sb->s_anon.first); - dentry->d_flags |= DCACHE_UNHASHED; - hlist_bl_del_init(&dentry->d_hash); - __bit_spin_unlock(0, - (unsigned long *)&dentry->d_sb->s_anon.first); - } else { - struct dcache_hash_bucket *b; + if (!d_unhashed(dentry)) { + struct hlist_bl_head *b; + if (unlikely(dentry->d_flags & DCACHE_DISCONNECTED)) + b = &dentry->d_sb->s_anon; + else b = d_hash(dentry->d_parent, dentry->d_name.hash); - spin_lock_bucket(b); - /* - * We may not actually need to put DCACHE_UNHASHED - * manipulations under the hash lock, but follow - * the principle of least surprise. - */ - dentry->d_flags |= DCACHE_UNHASHED; - hlist_bl_del_rcu(&dentry->d_hash); - spin_unlock_bucket(b); - dentry_rcuwalk_barrier(dentry); - } + + hlist_bl_lock(b); + __hlist_bl_del(&dentry->d_hash); + dentry->d_hash.pprev = NULL; + hlist_bl_unlock(b); + + dentry_rcuwalk_barrier(dentry); } } EXPORT_SYMBOL(__d_drop); @@ -1241,7 +1220,7 @@ void shrink_dcache_parent(struct dentry * parent) EXPORT_SYMBOL(shrink_dcache_parent); /* - * Scan `nr' dentries and return the number which remain. + * Scan `sc->nr_slab_to_reclaim' dentries and return the number which remain. * * We need to avoid reentering the filesystem if the caller is performing a * GFP_NOFS allocation attempt. One example deadlock is: @@ -1252,8 +1231,12 @@ EXPORT_SYMBOL(shrink_dcache_parent); * * In this case we return -1 to tell the caller that we baled. */ -static int shrink_dcache_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) +static int shrink_dcache_memory(struct shrinker *shrink, + struct shrink_control *sc) { + int nr = sc->nr_to_scan; + gfp_t gfp_mask = sc->gfp_mask; + if (nr) { if (!(gfp_mask & __GFP_FS)) return -1; @@ -1304,7 +1287,7 @@ struct dentry *d_alloc(struct dentry * parent, const struct qstr *name) dname[name->len] = 0; dentry->d_count = 1; - dentry->d_flags = DCACHE_UNHASHED; + dentry->d_flags = 0; spin_lock_init(&dentry->d_lock); seqcount_init(&dentry->d_seq); dentry->d_inode = NULL; @@ -1606,10 +1589,9 @@ struct dentry *d_obtain_alias(struct inode *inode) tmp->d_inode = inode; tmp->d_flags |= DCACHE_DISCONNECTED; list_add(&tmp->d_alias, &inode->i_dentry); - bit_spin_lock(0, (unsigned long *)&tmp->d_sb->s_anon.first); - tmp->d_flags &= ~DCACHE_UNHASHED; + hlist_bl_lock(&tmp->d_sb->s_anon); hlist_bl_add_head(&tmp->d_hash, &tmp->d_sb->s_anon); - __bit_spin_unlock(0, (unsigned long *)&tmp->d_sb->s_anon.first); + hlist_bl_unlock(&tmp->d_sb->s_anon); spin_unlock(&tmp->d_lock); spin_unlock(&inode->i_lock); security_d_instantiate(tmp, inode); @@ -1789,7 +1771,7 @@ struct dentry *__d_lookup_rcu(struct dentry *parent, struct qstr *name, unsigned int len = name->len; unsigned int hash = name->hash; const unsigned char *str = name->name; - struct dcache_hash_bucket *b = d_hash(parent, hash); + struct hlist_bl_head *b = d_hash(parent, hash); struct hlist_bl_node *node; struct dentry *dentry; @@ -1813,7 +1795,7 @@ struct dentry *__d_lookup_rcu(struct dentry *parent, struct qstr *name, * * See Documentation/filesystems/path-lookup.txt for more details. */ - hlist_bl_for_each_entry_rcu(dentry, node, &b->head, d_hash) { + hlist_bl_for_each_entry_rcu(dentry, node, b, d_hash) { struct inode *i; const char *tname; int tlen; @@ -1908,7 +1890,7 @@ struct dentry *__d_lookup(struct dentry *parent, struct qstr *name) unsigned int len = name->len; unsigned int hash = name->hash; const unsigned char *str = name->name; - struct dcache_hash_bucket *b = d_hash(parent, hash); + struct hlist_bl_head *b = d_hash(parent, hash); struct hlist_bl_node *node; struct dentry *found = NULL; struct dentry *dentry; @@ -1935,7 +1917,7 @@ struct dentry *__d_lookup(struct dentry *parent, struct qstr *name) */ rcu_read_lock(); - hlist_bl_for_each_entry_rcu(dentry, node, &b->head, d_hash) { + hlist_bl_for_each_entry_rcu(dentry, node, b, d_hash) { const char *tname; int tlen; @@ -2086,13 +2068,13 @@ again: } EXPORT_SYMBOL(d_delete); -static void __d_rehash(struct dentry * entry, struct dcache_hash_bucket *b) +static void __d_rehash(struct dentry * entry, struct hlist_bl_head *b) { BUG_ON(!d_unhashed(entry)); - spin_lock_bucket(b); - entry->d_flags &= ~DCACHE_UNHASHED; - hlist_bl_add_head_rcu(&entry->d_hash, &b->head); - spin_unlock_bucket(b); + hlist_bl_lock(b); + entry->d_flags |= DCACHE_RCUACCESS; + hlist_bl_add_head_rcu(&entry->d_hash, b); + hlist_bl_unlock(b); } static void _d_rehash(struct dentry * entry) @@ -2131,7 +2113,7 @@ EXPORT_SYMBOL(d_rehash); */ void dentry_update_name_case(struct dentry *dentry, struct qstr *name) { - BUG_ON(!mutex_is_locked(&dentry->d_inode->i_mutex)); + BUG_ON(!mutex_is_locked(&dentry->d_parent->d_inode->i_mutex)); BUG_ON(dentry->d_name.len != name->len); /* d_lookup gives this */ spin_lock(&dentry->d_lock); @@ -3025,7 +3007,7 @@ static void __init dcache_init_early(void) dentry_hashtable = alloc_large_system_hash("Dentry cache", - sizeof(struct dcache_hash_bucket), + sizeof(struct hlist_bl_head), dhash_entries, 13, HASH_EARLY, @@ -3034,7 +3016,7 @@ static void __init dcache_init_early(void) 0); for (loop = 0; loop < (1 << d_hash_shift); loop++) - INIT_HLIST_BL_HEAD(&dentry_hashtable[loop].head); + INIT_HLIST_BL_HEAD(dentry_hashtable + loop); } static void __init dcache_init(void) @@ -3057,7 +3039,7 @@ static void __init dcache_init(void) dentry_hashtable = alloc_large_system_hash("Dentry cache", - sizeof(struct dcache_hash_bucket), + sizeof(struct hlist_bl_head), dhash_entries, 13, 0, @@ -3066,7 +3048,7 @@ static void __init dcache_init(void) 0); for (loop = 0; loop < (1 << d_hash_shift); loop++) - INIT_HLIST_BL_HEAD(&dentry_hashtable[loop].head); + INIT_HLIST_BL_HEAD(dentry_hashtable + loop); } /* SLAB cache for __getname() consumers */ diff --git a/fs/debugfs/file.c b/fs/debugfs/file.c index 89d394d8fe24..90f76575c056 100644 --- a/fs/debugfs/file.c +++ b/fs/debugfs/file.c @@ -428,26 +428,17 @@ static ssize_t write_file_bool(struct file *file, const char __user *user_buf, size_t count, loff_t *ppos) { char buf[32]; - int buf_size; + size_t buf_size; + bool bv; u32 *val = file->private_data; buf_size = min(count, (sizeof(buf)-1)); if (copy_from_user(buf, user_buf, buf_size)) return -EFAULT; - switch (buf[0]) { - case 'y': - case 'Y': - case '1': - *val = 1; - break; - case 'n': - case 'N': - case '0': - *val = 0; - break; - } - + if (strtobool(buf, &bv) == 0) + *val = bv; + return count; } diff --git a/fs/dlm/config.c b/fs/dlm/config.c index 0d329ff8ed4c..9b026ea8baa9 100644 --- a/fs/dlm/config.c +++ b/fs/dlm/config.c @@ -100,6 +100,7 @@ struct dlm_cluster { unsigned int cl_log_debug; unsigned int cl_protocol; unsigned int cl_timewarn_cs; + unsigned int cl_waitwarn_us; }; enum { @@ -114,6 +115,7 @@ enum { CLUSTER_ATTR_LOG_DEBUG, CLUSTER_ATTR_PROTOCOL, CLUSTER_ATTR_TIMEWARN_CS, + CLUSTER_ATTR_WAITWARN_US, }; struct cluster_attribute { @@ -166,6 +168,7 @@ CLUSTER_ATTR(scan_secs, 1); CLUSTER_ATTR(log_debug, 0); CLUSTER_ATTR(protocol, 0); CLUSTER_ATTR(timewarn_cs, 1); +CLUSTER_ATTR(waitwarn_us, 0); static struct configfs_attribute *cluster_attrs[] = { [CLUSTER_ATTR_TCP_PORT] = &cluster_attr_tcp_port.attr, @@ -179,6 +182,7 @@ static struct configfs_attribute *cluster_attrs[] = { [CLUSTER_ATTR_LOG_DEBUG] = &cluster_attr_log_debug.attr, [CLUSTER_ATTR_PROTOCOL] = &cluster_attr_protocol.attr, [CLUSTER_ATTR_TIMEWARN_CS] = &cluster_attr_timewarn_cs.attr, + [CLUSTER_ATTR_WAITWARN_US] = &cluster_attr_waitwarn_us.attr, NULL, }; @@ -439,6 +443,7 @@ static struct config_group *make_cluster(struct config_group *g, cl->cl_log_debug = dlm_config.ci_log_debug; cl->cl_protocol = dlm_config.ci_protocol; cl->cl_timewarn_cs = dlm_config.ci_timewarn_cs; + cl->cl_waitwarn_us = dlm_config.ci_waitwarn_us; space_list = &sps->ss_group; comm_list = &cms->cs_group; @@ -986,6 +991,7 @@ int dlm_our_addr(struct sockaddr_storage *addr, int num) #define DEFAULT_LOG_DEBUG 0 #define DEFAULT_PROTOCOL 0 #define DEFAULT_TIMEWARN_CS 500 /* 5 sec = 500 centiseconds */ +#define DEFAULT_WAITWARN_US 0 struct dlm_config_info dlm_config = { .ci_tcp_port = DEFAULT_TCP_PORT, @@ -998,6 +1004,7 @@ struct dlm_config_info dlm_config = { .ci_scan_secs = DEFAULT_SCAN_SECS, .ci_log_debug = DEFAULT_LOG_DEBUG, .ci_protocol = DEFAULT_PROTOCOL, - .ci_timewarn_cs = DEFAULT_TIMEWARN_CS + .ci_timewarn_cs = DEFAULT_TIMEWARN_CS, + .ci_waitwarn_us = DEFAULT_WAITWARN_US }; diff --git a/fs/dlm/config.h b/fs/dlm/config.h index 4f1d6fce58c5..dd0ce24d5a80 100644 --- a/fs/dlm/config.h +++ b/fs/dlm/config.h @@ -28,6 +28,7 @@ struct dlm_config_info { int ci_log_debug; int ci_protocol; int ci_timewarn_cs; + int ci_waitwarn_us; }; extern struct dlm_config_info dlm_config; diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index b94204913011..0262451eb9c6 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -209,6 +209,7 @@ struct dlm_args { #define DLM_IFL_WATCH_TIMEWARN 0x00400000 #define DLM_IFL_TIMEOUT_CANCEL 0x00800000 #define DLM_IFL_DEADLOCK_CANCEL 0x01000000 +#define DLM_IFL_STUB_MS 0x02000000 /* magic number for m_flags */ #define DLM_IFL_USER 0x00000001 #define DLM_IFL_ORPHAN 0x00000002 @@ -245,6 +246,7 @@ struct dlm_lkb { int8_t lkb_wait_type; /* type of reply waiting for */ int8_t lkb_wait_count; + int lkb_wait_nodeid; /* for debugging */ struct list_head lkb_idtbl_list; /* lockspace lkbtbl */ struct list_head lkb_statequeue; /* rsb g/c/w list */ @@ -254,6 +256,7 @@ struct dlm_lkb { struct list_head lkb_ownqueue; /* list of locks for a process */ struct list_head lkb_time_list; ktime_t lkb_timestamp; + ktime_t lkb_wait_time; unsigned long lkb_timeout_cs; struct dlm_callback lkb_callbacks[DLM_CALLBACKS_SIZE]; diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index 04b8c449303f..f71d0b5abd95 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -519,7 +519,7 @@ static void toss_rsb(struct kref *kref) } } -/* When all references to the rsb are gone it's transfered to +/* When all references to the rsb are gone it's transferred to the tossed list for later disposal. */ static void put_rsb(struct dlm_rsb *r) @@ -799,10 +799,84 @@ static int msg_reply_type(int mstype) return -1; } +static int nodeid_warned(int nodeid, int num_nodes, int *warned) +{ + int i; + + for (i = 0; i < num_nodes; i++) { + if (!warned[i]) { + warned[i] = nodeid; + return 0; + } + if (warned[i] == nodeid) + return 1; + } + return 0; +} + +void dlm_scan_waiters(struct dlm_ls *ls) +{ + struct dlm_lkb *lkb; + ktime_t zero = ktime_set(0, 0); + s64 us; + s64 debug_maxus = 0; + u32 debug_scanned = 0; + u32 debug_expired = 0; + int num_nodes = 0; + int *warned = NULL; + + if (!dlm_config.ci_waitwarn_us) + return; + + mutex_lock(&ls->ls_waiters_mutex); + + list_for_each_entry(lkb, &ls->ls_waiters, lkb_wait_reply) { + if (ktime_equal(lkb->lkb_wait_time, zero)) + continue; + + debug_scanned++; + + us = ktime_to_us(ktime_sub(ktime_get(), lkb->lkb_wait_time)); + + if (us < dlm_config.ci_waitwarn_us) + continue; + + lkb->lkb_wait_time = zero; + + debug_expired++; + if (us > debug_maxus) + debug_maxus = us; + + if (!num_nodes) { + num_nodes = ls->ls_num_nodes; + warned = kmalloc(GFP_KERNEL, num_nodes * sizeof(int)); + if (warned) + memset(warned, 0, num_nodes * sizeof(int)); + } + if (!warned) + continue; + if (nodeid_warned(lkb->lkb_wait_nodeid, num_nodes, warned)) + continue; + + log_error(ls, "waitwarn %x %lld %d us check connection to " + "node %d", lkb->lkb_id, (long long)us, + dlm_config.ci_waitwarn_us, lkb->lkb_wait_nodeid); + } + mutex_unlock(&ls->ls_waiters_mutex); + + if (warned) + kfree(warned); + + if (debug_expired) + log_debug(ls, "scan_waiters %u warn %u over %d us max %lld us", + debug_scanned, debug_expired, + dlm_config.ci_waitwarn_us, (long long)debug_maxus); +} + /* add/remove lkb from global waiters list of lkb's waiting for a reply from a remote node */ -static int add_to_waiters(struct dlm_lkb *lkb, int mstype) +static int add_to_waiters(struct dlm_lkb *lkb, int mstype, int to_nodeid) { struct dlm_ls *ls = lkb->lkb_resource->res_ls; int error = 0; @@ -842,6 +916,8 @@ static int add_to_waiters(struct dlm_lkb *lkb, int mstype) lkb->lkb_wait_count++; lkb->lkb_wait_type = mstype; + lkb->lkb_wait_time = ktime_get(); + lkb->lkb_wait_nodeid = to_nodeid; /* for debugging */ hold_lkb(lkb); list_add(&lkb->lkb_wait_reply, &ls->ls_waiters); out: @@ -961,10 +1037,10 @@ static int remove_from_waiters_ms(struct dlm_lkb *lkb, struct dlm_message *ms) struct dlm_ls *ls = lkb->lkb_resource->res_ls; int error; - if (ms != &ls->ls_stub_ms) + if (ms->m_flags != DLM_IFL_STUB_MS) mutex_lock(&ls->ls_waiters_mutex); error = _remove_from_waiters(lkb, ms->m_type, ms); - if (ms != &ls->ls_stub_ms) + if (ms->m_flags != DLM_IFL_STUB_MS) mutex_unlock(&ls->ls_waiters_mutex); return error; } @@ -1157,6 +1233,16 @@ void dlm_adjust_timeouts(struct dlm_ls *ls) list_for_each_entry(lkb, &ls->ls_timeout, lkb_time_list) lkb->lkb_timestamp = ktime_add_us(lkb->lkb_timestamp, adj_us); mutex_unlock(&ls->ls_timeout_mutex); + + if (!dlm_config.ci_waitwarn_us) + return; + + mutex_lock(&ls->ls_waiters_mutex); + list_for_each_entry(lkb, &ls->ls_waiters, lkb_wait_reply) { + if (ktime_to_us(lkb->lkb_wait_time)) + lkb->lkb_wait_time = ktime_get(); + } + mutex_unlock(&ls->ls_waiters_mutex); } /* lkb is master or local copy */ @@ -1376,14 +1462,8 @@ static void grant_lock_pending(struct dlm_rsb *r, struct dlm_lkb *lkb) ALTPR/ALTCW: our rqmode may have been changed to PR or CW to become compatible with other granted locks */ -static void munge_demoted(struct dlm_lkb *lkb, struct dlm_message *ms) +static void munge_demoted(struct dlm_lkb *lkb) { - if (ms->m_type != DLM_MSG_CONVERT_REPLY) { - log_print("munge_demoted %x invalid reply type %d", - lkb->lkb_id, ms->m_type); - return; - } - if (lkb->lkb_rqmode == DLM_LOCK_IV || lkb->lkb_grmode == DLM_LOCK_IV) { log_print("munge_demoted %x invalid modes gr %d rq %d", lkb->lkb_id, lkb->lkb_grmode, lkb->lkb_rqmode); @@ -2844,12 +2924,12 @@ static int send_common(struct dlm_rsb *r, struct dlm_lkb *lkb, int mstype) struct dlm_mhandle *mh; int to_nodeid, error; - error = add_to_waiters(lkb, mstype); + to_nodeid = r->res_nodeid; + + error = add_to_waiters(lkb, mstype, to_nodeid); if (error) return error; - to_nodeid = r->res_nodeid; - error = create_message(r, lkb, to_nodeid, mstype, &ms, &mh); if (error) goto fail; @@ -2880,9 +2960,9 @@ static int send_convert(struct dlm_rsb *r, struct dlm_lkb *lkb) /* down conversions go without a reply from the master */ if (!error && down_conversion(lkb)) { remove_from_waiters(lkb, DLM_MSG_CONVERT_REPLY); + r->res_ls->ls_stub_ms.m_flags = DLM_IFL_STUB_MS; r->res_ls->ls_stub_ms.m_type = DLM_MSG_CONVERT_REPLY; r->res_ls->ls_stub_ms.m_result = 0; - r->res_ls->ls_stub_ms.m_flags = lkb->lkb_flags; __receive_convert_reply(r, lkb, &r->res_ls->ls_stub_ms); } @@ -2951,12 +3031,12 @@ static int send_lookup(struct dlm_rsb *r, struct dlm_lkb *lkb) struct dlm_mhandle *mh; int to_nodeid, error; - error = add_to_waiters(lkb, DLM_MSG_LOOKUP); + to_nodeid = dlm_dir_nodeid(r); + + error = add_to_waiters(lkb, DLM_MSG_LOOKUP, to_nodeid); if (error) return error; - to_nodeid = dlm_dir_nodeid(r); - error = create_message(r, NULL, to_nodeid, DLM_MSG_LOOKUP, &ms, &mh); if (error) goto fail; @@ -3070,6 +3150,9 @@ static void receive_flags(struct dlm_lkb *lkb, struct dlm_message *ms) static void receive_flags_reply(struct dlm_lkb *lkb, struct dlm_message *ms) { + if (ms->m_flags == DLM_IFL_STUB_MS) + return; + lkb->lkb_sbflags = ms->m_sbflags; lkb->lkb_flags = (lkb->lkb_flags & 0xFFFF0000) | (ms->m_flags & 0x0000FFFF); @@ -3612,7 +3695,7 @@ static void __receive_convert_reply(struct dlm_rsb *r, struct dlm_lkb *lkb, /* convert was queued on remote master */ receive_flags_reply(lkb, ms); if (is_demoted(lkb)) - munge_demoted(lkb, ms); + munge_demoted(lkb); del_lkb(r, lkb); add_lkb(r, lkb, DLM_LKSTS_CONVERT); add_timeout(lkb); @@ -3622,7 +3705,7 @@ static void __receive_convert_reply(struct dlm_rsb *r, struct dlm_lkb *lkb, /* convert was granted on remote master */ receive_flags_reply(lkb, ms); if (is_demoted(lkb)) - munge_demoted(lkb, ms); + munge_demoted(lkb); grant_lock_pc(r, lkb, ms); queue_cast(r, lkb, 0); break; @@ -3996,15 +4079,17 @@ void dlm_receive_buffer(union dlm_packet *p, int nodeid) dlm_put_lockspace(ls); } -static void recover_convert_waiter(struct dlm_ls *ls, struct dlm_lkb *lkb) +static void recover_convert_waiter(struct dlm_ls *ls, struct dlm_lkb *lkb, + struct dlm_message *ms_stub) { if (middle_conversion(lkb)) { hold_lkb(lkb); - ls->ls_stub_ms.m_type = DLM_MSG_CONVERT_REPLY; - ls->ls_stub_ms.m_result = -EINPROGRESS; - ls->ls_stub_ms.m_flags = lkb->lkb_flags; - ls->ls_stub_ms.m_header.h_nodeid = lkb->lkb_nodeid; - _receive_convert_reply(lkb, &ls->ls_stub_ms); + memset(ms_stub, 0, sizeof(struct dlm_message)); + ms_stub->m_flags = DLM_IFL_STUB_MS; + ms_stub->m_type = DLM_MSG_CONVERT_REPLY; + ms_stub->m_result = -EINPROGRESS; + ms_stub->m_header.h_nodeid = lkb->lkb_nodeid; + _receive_convert_reply(lkb, ms_stub); /* Same special case as in receive_rcom_lock_args() */ lkb->lkb_grmode = DLM_LOCK_IV; @@ -4045,13 +4130,27 @@ static int waiter_needs_recovery(struct dlm_ls *ls, struct dlm_lkb *lkb) void dlm_recover_waiters_pre(struct dlm_ls *ls) { struct dlm_lkb *lkb, *safe; + struct dlm_message *ms_stub; int wait_type, stub_unlock_result, stub_cancel_result; + ms_stub = kmalloc(GFP_KERNEL, sizeof(struct dlm_message)); + if (!ms_stub) { + log_error(ls, "dlm_recover_waiters_pre no mem"); + return; + } + mutex_lock(&ls->ls_waiters_mutex); list_for_each_entry_safe(lkb, safe, &ls->ls_waiters, lkb_wait_reply) { - log_debug(ls, "pre recover waiter lkid %x type %d flags %x", - lkb->lkb_id, lkb->lkb_wait_type, lkb->lkb_flags); + + /* exclude debug messages about unlocks because there can be so + many and they aren't very interesting */ + + if (lkb->lkb_wait_type != DLM_MSG_UNLOCK) { + log_debug(ls, "recover_waiter %x nodeid %d " + "msg %d to %d", lkb->lkb_id, lkb->lkb_nodeid, + lkb->lkb_wait_type, lkb->lkb_wait_nodeid); + } /* all outstanding lookups, regardless of destination will be resent after recovery is done */ @@ -4097,26 +4196,28 @@ void dlm_recover_waiters_pre(struct dlm_ls *ls) break; case DLM_MSG_CONVERT: - recover_convert_waiter(ls, lkb); + recover_convert_waiter(ls, lkb, ms_stub); break; case DLM_MSG_UNLOCK: hold_lkb(lkb); - ls->ls_stub_ms.m_type = DLM_MSG_UNLOCK_REPLY; - ls->ls_stub_ms.m_result = stub_unlock_result; - ls->ls_stub_ms.m_flags = lkb->lkb_flags; - ls->ls_stub_ms.m_header.h_nodeid = lkb->lkb_nodeid; - _receive_unlock_reply(lkb, &ls->ls_stub_ms); + memset(ms_stub, 0, sizeof(struct dlm_message)); + ms_stub->m_flags = DLM_IFL_STUB_MS; + ms_stub->m_type = DLM_MSG_UNLOCK_REPLY; + ms_stub->m_result = stub_unlock_result; + ms_stub->m_header.h_nodeid = lkb->lkb_nodeid; + _receive_unlock_reply(lkb, ms_stub); dlm_put_lkb(lkb); break; case DLM_MSG_CANCEL: hold_lkb(lkb); - ls->ls_stub_ms.m_type = DLM_MSG_CANCEL_REPLY; - ls->ls_stub_ms.m_result = stub_cancel_result; - ls->ls_stub_ms.m_flags = lkb->lkb_flags; - ls->ls_stub_ms.m_header.h_nodeid = lkb->lkb_nodeid; - _receive_cancel_reply(lkb, &ls->ls_stub_ms); + memset(ms_stub, 0, sizeof(struct dlm_message)); + ms_stub->m_flags = DLM_IFL_STUB_MS; + ms_stub->m_type = DLM_MSG_CANCEL_REPLY; + ms_stub->m_result = stub_cancel_result; + ms_stub->m_header.h_nodeid = lkb->lkb_nodeid; + _receive_cancel_reply(lkb, ms_stub); dlm_put_lkb(lkb); break; @@ -4127,6 +4228,7 @@ void dlm_recover_waiters_pre(struct dlm_ls *ls) schedule(); } mutex_unlock(&ls->ls_waiters_mutex); + kfree(ms_stub); } static struct dlm_lkb *find_resend_waiter(struct dlm_ls *ls) @@ -4191,8 +4293,8 @@ int dlm_recover_waiters_post(struct dlm_ls *ls) ou = is_overlap_unlock(lkb); err = 0; - log_debug(ls, "recover_waiters_post %x type %d flags %x %s", - lkb->lkb_id, mstype, lkb->lkb_flags, r->res_name); + log_debug(ls, "recover_waiter %x nodeid %d msg %d r_nodeid %d", + lkb->lkb_id, lkb->lkb_nodeid, mstype, r->res_nodeid); /* At this point we assume that we won't get a reply to any previous op or overlap op on this lock. First, do a big diff --git a/fs/dlm/lock.h b/fs/dlm/lock.h index 88e93c80cc22..265017a7c3e7 100644 --- a/fs/dlm/lock.h +++ b/fs/dlm/lock.h @@ -24,6 +24,7 @@ int dlm_put_lkb(struct dlm_lkb *lkb); void dlm_scan_rsbs(struct dlm_ls *ls); int dlm_lock_recovery_try(struct dlm_ls *ls); void dlm_unlock_recovery(struct dlm_ls *ls); +void dlm_scan_waiters(struct dlm_ls *ls); void dlm_scan_timeout(struct dlm_ls *ls); void dlm_adjust_timeouts(struct dlm_ls *ls); diff --git a/fs/dlm/lockspace.c b/fs/dlm/lockspace.c index f994a7dfda85..14cbf4099753 100644 --- a/fs/dlm/lockspace.c +++ b/fs/dlm/lockspace.c @@ -243,7 +243,6 @@ static struct dlm_ls *find_ls_to_scan(void) static int dlm_scand(void *data) { struct dlm_ls *ls; - int timeout_jiffies = dlm_config.ci_scan_secs * HZ; while (!kthread_should_stop()) { ls = find_ls_to_scan(); @@ -252,13 +251,14 @@ static int dlm_scand(void *data) ls->ls_scan_time = jiffies; dlm_scan_rsbs(ls); dlm_scan_timeout(ls); + dlm_scan_waiters(ls); dlm_unlock_recovery(ls); } else { ls->ls_scan_time += HZ; } - } else { - schedule_timeout_interruptible(timeout_jiffies); + continue; } + schedule_timeout_interruptible(dlm_config.ci_scan_secs * HZ); } return 0; } diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index bffa1e73b9a9..5e2c71f05e46 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -810,7 +810,7 @@ static int tcp_accept_from_sock(struct connection *con) /* * Add it to the active queue in case we got data - * beween processing the accept adding the socket + * between processing the accept adding the socket * to the read_sockets list */ if (!test_and_set_bit(CF_READ_PENDING, &addcon->flags)) diff --git a/fs/dlm/main.c b/fs/dlm/main.c index b80e0aa3cfa5..5a59efa0bb46 100644 --- a/fs/dlm/main.c +++ b/fs/dlm/main.c @@ -50,7 +50,7 @@ static int __init init_dlm(void) if (error) goto out_netlink; - printk("DLM (built %s %s) installed\n", __DATE__, __TIME__); + printk("DLM installed\n"); return 0; diff --git a/fs/dlm/plock.c b/fs/dlm/plock.c index 30d8b85febbf..e2b878004364 100644 --- a/fs/dlm/plock.c +++ b/fs/dlm/plock.c @@ -71,6 +71,36 @@ static void send_op(struct plock_op *op) wake_up(&send_wq); } +/* If a process was killed while waiting for the only plock on a file, + locks_remove_posix will not see any lock on the file so it won't + send an unlock-close to us to pass on to userspace to clean up the + abandoned waiter. So, we have to insert the unlock-close when the + lock call is interrupted. */ + +static void do_unlock_close(struct dlm_ls *ls, u64 number, + struct file *file, struct file_lock *fl) +{ + struct plock_op *op; + + op = kzalloc(sizeof(*op), GFP_NOFS); + if (!op) + return; + + op->info.optype = DLM_PLOCK_OP_UNLOCK; + op->info.pid = fl->fl_pid; + op->info.fsid = ls->ls_global_id; + op->info.number = number; + op->info.start = 0; + op->info.end = OFFSET_MAX; + if (fl->fl_lmops && fl->fl_lmops->fl_grant) + op->info.owner = (__u64) fl->fl_pid; + else + op->info.owner = (__u64)(long) fl->fl_owner; + + op->info.flags |= DLM_PLOCK_FL_CLOSE; + send_op(op); +} + int dlm_posix_lock(dlm_lockspace_t *lockspace, u64 number, struct file *file, int cmd, struct file_lock *fl) { @@ -114,9 +144,19 @@ int dlm_posix_lock(dlm_lockspace_t *lockspace, u64 number, struct file *file, send_op(op); - if (xop->callback == NULL) - wait_event(recv_wq, (op->done != 0)); - else { + if (xop->callback == NULL) { + rv = wait_event_killable(recv_wq, (op->done != 0)); + if (rv == -ERESTARTSYS) { + log_debug(ls, "dlm_posix_lock: wait killed %llx", + (unsigned long long)number); + spin_lock(&ops_lock); + list_del(&op->list); + spin_unlock(&ops_lock); + kfree(xop); + do_unlock_close(ls, number, file, fl); + goto out; + } + } else { rv = FILE_LOCK_DEFERRED; goto out; } @@ -233,6 +273,13 @@ int dlm_posix_unlock(dlm_lockspace_t *lockspace, u64 number, struct file *file, else op->info.owner = (__u64)(long) fl->fl_owner; + if (fl->fl_flags & FL_CLOSE) { + op->info.flags |= DLM_PLOCK_FL_CLOSE; + send_op(op); + rv = 0; + goto out; + } + send_op(op); wait_event(recv_wq, (op->done != 0)); @@ -334,7 +381,10 @@ static ssize_t dev_read(struct file *file, char __user *u, size_t count, spin_lock(&ops_lock); if (!list_empty(&send_list)) { op = list_entry(send_list.next, struct plock_op, list); - list_move(&op->list, &recv_list); + if (op->info.flags & DLM_PLOCK_FL_CLOSE) + list_del(&op->list); + else + list_move(&op->list, &recv_list); memcpy(&info, &op->info, sizeof(info)); } spin_unlock(&ops_lock); @@ -342,6 +392,13 @@ static ssize_t dev_read(struct file *file, char __user *u, size_t count, if (!op) return -EAGAIN; + /* there is no need to get a reply from userspace for unlocks + that were generated by the vfs cleaning up for a close + (the process did not make an unlock call). */ + + if (op->info.flags & DLM_PLOCK_FL_CLOSE) + kfree(op); + if (copy_to_user(u, &info, sizeof(info))) return -EFAULT; return sizeof(info); diff --git a/fs/dlm/recover.c b/fs/dlm/recover.c index eda43f362616..14638235f7b2 100644 --- a/fs/dlm/recover.c +++ b/fs/dlm/recover.c @@ -304,7 +304,7 @@ static void set_master_lkbs(struct dlm_rsb *r) } /* - * Propogate the new master nodeid to locks + * Propagate the new master nodeid to locks * The NEW_MASTER flag tells dlm_recover_locks() which rsb's to consider. * The NEW_MASTER2 flag tells recover_lvb() and set_locks_purged() which * rsb's to consider. diff --git a/fs/dlm/user.c b/fs/dlm/user.c index d5ab3fe7c198..e96bf3e9be88 100644 --- a/fs/dlm/user.c +++ b/fs/dlm/user.c @@ -611,7 +611,6 @@ static ssize_t device_write(struct file *file, const char __user *buf, out_sig: sigprocmask(SIG_SETMASK, &tmpsig, NULL); - recalc_sigpending(); out_free: kfree(kbuf); return error; diff --git a/fs/drop_caches.c b/fs/drop_caches.c index 98b77c89494c..c00e055b6282 100644 --- a/fs/drop_caches.c +++ b/fs/drop_caches.c @@ -40,9 +40,12 @@ static void drop_pagecache_sb(struct super_block *sb, void *unused) static void drop_slab(void) { int nr_objects; + struct shrink_control shrink = { + .gfp_mask = GFP_KERNEL, + }; do { - nr_objects = shrink_slab(1000, GFP_KERNEL, 1000); + nr_objects = shrink_slab(&shrink, 1000, 1000); } while (nr_objects > 10); } diff --git a/fs/ecryptfs/crypto.c b/fs/ecryptfs/crypto.c index d2a70a4561f9..b8d5c8091024 100644 --- a/fs/ecryptfs/crypto.c +++ b/fs/ecryptfs/crypto.c @@ -1452,6 +1452,25 @@ static void set_default_header_data(struct ecryptfs_crypt_stat *crypt_stat) crypt_stat->metadata_size = ECRYPTFS_MINIMUM_HEADER_EXTENT_SIZE; } +void ecryptfs_i_size_init(const char *page_virt, struct inode *inode) +{ + struct ecryptfs_mount_crypt_stat *mount_crypt_stat; + struct ecryptfs_crypt_stat *crypt_stat; + u64 file_size; + + crypt_stat = &ecryptfs_inode_to_private(inode)->crypt_stat; + mount_crypt_stat = + &ecryptfs_superblock_to_private(inode->i_sb)->mount_crypt_stat; + if (mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED) { + file_size = i_size_read(ecryptfs_inode_to_lower(inode)); + if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) + file_size += crypt_stat->metadata_size; + } else + file_size = get_unaligned_be64(page_virt); + i_size_write(inode, (loff_t)file_size); + crypt_stat->flags |= ECRYPTFS_I_SIZE_INITIALIZED; +} + /** * ecryptfs_read_headers_virt * @page_virt: The virtual address into which to read the headers @@ -1482,6 +1501,8 @@ static int ecryptfs_read_headers_virt(char *page_virt, rc = -EINVAL; goto out; } + if (!(crypt_stat->flags & ECRYPTFS_I_SIZE_INITIALIZED)) + ecryptfs_i_size_init(page_virt, ecryptfs_dentry->d_inode); offset += MAGIC_ECRYPTFS_MARKER_SIZE_BYTES; rc = ecryptfs_process_flags(crypt_stat, (page_virt + offset), &bytes_read); diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index bd3cafd0949d..e70282775e2c 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -269,6 +269,7 @@ struct ecryptfs_crypt_stat { #define ECRYPTFS_ENCFN_USE_MOUNT_FNEK 0x00000800 #define ECRYPTFS_ENCFN_USE_FEK 0x00001000 #define ECRYPTFS_UNLINK_SIGS 0x00002000 +#define ECRYPTFS_I_SIZE_INITIALIZED 0x00004000 u32 flags; unsigned int file_version; size_t iv_bytes; @@ -295,6 +296,8 @@ struct ecryptfs_crypt_stat { struct ecryptfs_inode_info { struct inode vfs_inode; struct inode *wii_inode; + struct mutex lower_file_mutex; + atomic_t lower_file_count; struct file *lower_file; struct ecryptfs_crypt_stat crypt_stat; }; @@ -626,6 +629,7 @@ struct ecryptfs_open_req { int ecryptfs_interpose(struct dentry *hidden_dentry, struct dentry *this_dentry, struct super_block *sb, u32 flags); +void ecryptfs_i_size_init(const char *page_virt, struct inode *inode); int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, struct dentry *lower_dentry, struct inode *ecryptfs_dir_inode); @@ -757,7 +761,8 @@ int ecryptfs_privileged_open(struct file **lower_file, struct dentry *lower_dentry, struct vfsmount *lower_mnt, const struct cred *cred); -int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry); +int ecryptfs_get_lower_file(struct dentry *ecryptfs_dentry); +void ecryptfs_put_lower_file(struct inode *inode); int ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, size_t *packet_size, diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index cedc913d11ba..566e5472f78c 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -191,10 +191,10 @@ static int ecryptfs_open(struct inode *inode, struct file *file) | ECRYPTFS_ENCRYPTED); } mutex_unlock(&crypt_stat->cs_mutex); - rc = ecryptfs_init_persistent_file(ecryptfs_dentry); + rc = ecryptfs_get_lower_file(ecryptfs_dentry); if (rc) { printk(KERN_ERR "%s: Error attempting to initialize " - "the persistent file for the dentry with name " + "the lower file for the dentry with name " "[%s]; rc = [%d]\n", __func__, ecryptfs_dentry->d_name.name, rc); goto out_free; @@ -202,9 +202,9 @@ static int ecryptfs_open(struct inode *inode, struct file *file) if ((ecryptfs_inode_to_private(inode)->lower_file->f_flags & O_ACCMODE) == O_RDONLY && (file->f_flags & O_ACCMODE) != O_RDONLY) { rc = -EPERM; - printk(KERN_WARNING "%s: Lower persistent file is RO; eCryptfs " + printk(KERN_WARNING "%s: Lower file is RO; eCryptfs " "file must hence be opened RO\n", __func__); - goto out_free; + goto out_put; } ecryptfs_set_file_lower( file, ecryptfs_inode_to_private(inode)->lower_file); @@ -232,10 +232,11 @@ static int ecryptfs_open(struct inode *inode, struct file *file) "Plaintext passthrough mode is not " "enabled; returning -EIO\n"); mutex_unlock(&crypt_stat->cs_mutex); - goto out_free; + goto out_put; } rc = 0; - crypt_stat->flags &= ~(ECRYPTFS_ENCRYPTED); + crypt_stat->flags &= ~(ECRYPTFS_I_SIZE_INITIALIZED + | ECRYPTFS_ENCRYPTED); mutex_unlock(&crypt_stat->cs_mutex); goto out; } @@ -245,6 +246,8 @@ static int ecryptfs_open(struct inode *inode, struct file *file) "[0x%.16lx] size: [0x%.16llx]\n", inode, inode->i_ino, (unsigned long long)i_size_read(inode)); goto out; +out_put: + ecryptfs_put_lower_file(inode); out_free: kmem_cache_free(ecryptfs_file_info_cache, ecryptfs_file_to_private(file)); @@ -254,17 +257,13 @@ out: static int ecryptfs_flush(struct file *file, fl_owner_t td) { - int rc = 0; - struct file *lower_file = NULL; - - lower_file = ecryptfs_file_to_lower(file); - if (lower_file->f_op && lower_file->f_op->flush) - rc = lower_file->f_op->flush(lower_file, td); - return rc; + return file->f_mode & FMODE_WRITE + ? filemap_write_and_wait(file->f_mapping) : 0; } static int ecryptfs_release(struct inode *inode, struct file *file) { + ecryptfs_put_lower_file(inode); kmem_cache_free(ecryptfs_file_info_cache, ecryptfs_file_to_private(file)); return 0; diff --git a/fs/ecryptfs/inode.c b/fs/ecryptfs/inode.c index f99051b7adab..94ab3c06317a 100644 --- a/fs/ecryptfs/inode.c +++ b/fs/ecryptfs/inode.c @@ -168,19 +168,18 @@ static int ecryptfs_initialize_file(struct dentry *ecryptfs_dentry) "context; rc = [%d]\n", rc); goto out; } - rc = ecryptfs_init_persistent_file(ecryptfs_dentry); + rc = ecryptfs_get_lower_file(ecryptfs_dentry); if (rc) { printk(KERN_ERR "%s: Error attempting to initialize " - "the persistent file for the dentry with name " + "the lower file for the dentry with name " "[%s]; rc = [%d]\n", __func__, ecryptfs_dentry->d_name.name, rc); goto out; } rc = ecryptfs_write_metadata(ecryptfs_dentry); - if (rc) { + if (rc) printk(KERN_ERR "Error writing headers; rc = [%d]\n", rc); - goto out; - } + ecryptfs_put_lower_file(ecryptfs_dentry->d_inode); out: return rc; } @@ -226,11 +225,9 @@ int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, struct dentry *lower_dir_dentry; struct vfsmount *lower_mnt; struct inode *lower_inode; - struct ecryptfs_mount_crypt_stat *mount_crypt_stat; struct ecryptfs_crypt_stat *crypt_stat; char *page_virt = NULL; - u64 file_size; - int rc = 0; + int put_lower = 0, rc = 0; lower_dir_dentry = lower_dentry->d_parent; lower_mnt = mntget(ecryptfs_dentry_to_lower_mnt( @@ -277,14 +274,15 @@ int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, rc = -ENOMEM; goto out; } - rc = ecryptfs_init_persistent_file(ecryptfs_dentry); + rc = ecryptfs_get_lower_file(ecryptfs_dentry); if (rc) { printk(KERN_ERR "%s: Error attempting to initialize " - "the persistent file for the dentry with name " + "the lower file for the dentry with name " "[%s]; rc = [%d]\n", __func__, ecryptfs_dentry->d_name.name, rc); goto out_free_kmem; } + put_lower = 1; crypt_stat = &ecryptfs_inode_to_private( ecryptfs_dentry->d_inode)->crypt_stat; /* TODO: lock for crypt_stat comparison */ @@ -302,18 +300,7 @@ int ecryptfs_lookup_and_interpose_lower(struct dentry *ecryptfs_dentry, } crypt_stat->flags |= ECRYPTFS_METADATA_IN_XATTR; } - mount_crypt_stat = &ecryptfs_superblock_to_private( - ecryptfs_dentry->d_sb)->mount_crypt_stat; - if (mount_crypt_stat->flags & ECRYPTFS_ENCRYPTED_VIEW_ENABLED) { - if (crypt_stat->flags & ECRYPTFS_METADATA_IN_XATTR) - file_size = (crypt_stat->metadata_size - + i_size_read(lower_dentry->d_inode)); - else - file_size = i_size_read(lower_dentry->d_inode); - } else { - file_size = get_unaligned_be64(page_virt); - } - i_size_write(ecryptfs_dentry->d_inode, (loff_t)file_size); + ecryptfs_i_size_init(page_virt, ecryptfs_dentry->d_inode); out_free_kmem: kmem_cache_free(ecryptfs_header_cache_2, page_virt); goto out; @@ -322,6 +309,8 @@ out_put: mntput(lower_mnt); d_drop(ecryptfs_dentry); out: + if (put_lower) + ecryptfs_put_lower_file(ecryptfs_dentry->d_inode); return rc; } @@ -538,8 +527,8 @@ static int ecryptfs_rmdir(struct inode *dir, struct dentry *dentry) dget(lower_dentry); rc = vfs_rmdir(lower_dir_dentry->d_inode, lower_dentry); dput(lower_dentry); - if (!rc) - d_delete(lower_dentry); + if (!rc && dentry->d_inode) + clear_nlink(dentry->d_inode); fsstack_copy_attr_times(dir, lower_dir_dentry->d_inode); dir->i_nlink = lower_dir_dentry->d_inode->i_nlink; unlock_dir(lower_dir_dentry); @@ -610,8 +599,8 @@ ecryptfs_rename(struct inode *old_dir, struct dentry *old_dentry, fsstack_copy_attr_all(old_dir, lower_old_dir_dentry->d_inode); out_lock: unlock_rename(lower_old_dir_dentry, lower_new_dir_dentry); - dput(lower_new_dentry->d_parent); - dput(lower_old_dentry->d_parent); + dput(lower_new_dir_dentry); + dput(lower_old_dir_dentry); dput(lower_new_dentry); dput(lower_old_dentry); return rc; @@ -759,8 +748,11 @@ static int truncate_upper(struct dentry *dentry, struct iattr *ia, if (unlikely((ia->ia_size == i_size))) { lower_ia->ia_valid &= ~ATTR_SIZE; - goto out; + return 0; } + rc = ecryptfs_get_lower_file(dentry); + if (rc) + return rc; crypt_stat = &ecryptfs_inode_to_private(dentry->d_inode)->crypt_stat; /* Switch on growing or shrinking file */ if (ia->ia_size > i_size) { @@ -838,6 +830,7 @@ static int truncate_upper(struct dentry *dentry, struct iattr *ia, lower_ia->ia_valid &= ~ATTR_SIZE; } out: + ecryptfs_put_lower_file(inode); return rc; } @@ -913,7 +906,13 @@ static int ecryptfs_setattr(struct dentry *dentry, struct iattr *ia) mount_crypt_stat = &ecryptfs_superblock_to_private( dentry->d_sb)->mount_crypt_stat; + rc = ecryptfs_get_lower_file(dentry); + if (rc) { + mutex_unlock(&crypt_stat->cs_mutex); + goto out; + } rc = ecryptfs_read_metadata(dentry); + ecryptfs_put_lower_file(inode); if (rc) { if (!(mount_crypt_stat->flags & ECRYPTFS_PLAINTEXT_PASSTHROUGH_ENABLED)) { @@ -927,10 +926,17 @@ static int ecryptfs_setattr(struct dentry *dentry, struct iattr *ia) goto out; } rc = 0; - crypt_stat->flags &= ~(ECRYPTFS_ENCRYPTED); + crypt_stat->flags &= ~(ECRYPTFS_I_SIZE_INITIALIZED + | ECRYPTFS_ENCRYPTED); } } mutex_unlock(&crypt_stat->cs_mutex); + if (S_ISREG(inode->i_mode)) { + rc = filemap_write_and_wait(inode->i_mapping); + if (rc) + goto out; + fsstack_copy_attr_all(inode, lower_inode); + } memcpy(&lower_ia, ia, sizeof(lower_ia)); if (ia->ia_valid & ATTR_FILE) lower_ia.ia_file = ecryptfs_file_to_lower(ia->ia_file); diff --git a/fs/ecryptfs/keystore.c b/fs/ecryptfs/keystore.c index 03e609c45012..27a7fefb83eb 100644 --- a/fs/ecryptfs/keystore.c +++ b/fs/ecryptfs/keystore.c @@ -599,8 +599,8 @@ struct ecryptfs_write_tag_70_packet_silly_stack { struct mutex *tfm_mutex; char *block_aligned_filename; struct ecryptfs_auth_tok *auth_tok; - struct scatterlist src_sg; - struct scatterlist dst_sg; + struct scatterlist src_sg[2]; + struct scatterlist dst_sg[2]; struct blkcipher_desc desc; char iv[ECRYPTFS_MAX_IV_BYTES]; char hash[ECRYPTFS_TAG_70_DIGEST_SIZE]; @@ -816,23 +816,21 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, memcpy(&s->block_aligned_filename[s->num_rand_bytes], filename, filename_size); rc = virt_to_scatterlist(s->block_aligned_filename, - s->block_aligned_filename_size, &s->src_sg, 1); - if (rc != 1) { + s->block_aligned_filename_size, s->src_sg, 2); + if (rc < 1) { printk(KERN_ERR "%s: Internal error whilst attempting to " - "convert filename memory to scatterlist; " - "expected rc = 1; got rc = [%d]. " + "convert filename memory to scatterlist; rc = [%d]. " "block_aligned_filename_size = [%zd]\n", __func__, rc, s->block_aligned_filename_size); goto out_release_free_unlock; } rc = virt_to_scatterlist(&dest[s->i], s->block_aligned_filename_size, - &s->dst_sg, 1); - if (rc != 1) { + s->dst_sg, 2); + if (rc < 1) { printk(KERN_ERR "%s: Internal error whilst attempting to " "convert encrypted filename memory to scatterlist; " - "expected rc = 1; got rc = [%d]. " - "block_aligned_filename_size = [%zd]\n", __func__, rc, - s->block_aligned_filename_size); + "rc = [%d]. block_aligned_filename_size = [%zd]\n", + __func__, rc, s->block_aligned_filename_size); goto out_release_free_unlock; } /* The characters in the first block effectively do the job @@ -855,7 +853,7 @@ ecryptfs_write_tag_70_packet(char *dest, size_t *remaining_bytes, mount_crypt_stat->global_default_fn_cipher_key_bytes); goto out_release_free_unlock; } - rc = crypto_blkcipher_encrypt_iv(&s->desc, &s->dst_sg, &s->src_sg, + rc = crypto_blkcipher_encrypt_iv(&s->desc, s->dst_sg, s->src_sg, s->block_aligned_filename_size); if (rc) { printk(KERN_ERR "%s: Error attempting to encrypt filename; " @@ -891,8 +889,8 @@ struct ecryptfs_parse_tag_70_packet_silly_stack { struct mutex *tfm_mutex; char *decrypted_filename; struct ecryptfs_auth_tok *auth_tok; - struct scatterlist src_sg; - struct scatterlist dst_sg; + struct scatterlist src_sg[2]; + struct scatterlist dst_sg[2]; struct blkcipher_desc desc; char fnek_sig_hex[ECRYPTFS_SIG_SIZE_HEX + 1]; char iv[ECRYPTFS_MAX_IV_BYTES]; @@ -1008,13 +1006,12 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, } mutex_lock(s->tfm_mutex); rc = virt_to_scatterlist(&data[(*packet_size)], - s->block_aligned_filename_size, &s->src_sg, 1); - if (rc != 1) { + s->block_aligned_filename_size, s->src_sg, 2); + if (rc < 1) { printk(KERN_ERR "%s: Internal error whilst attempting to " "convert encrypted filename memory to scatterlist; " - "expected rc = 1; got rc = [%d]. " - "block_aligned_filename_size = [%zd]\n", __func__, rc, - s->block_aligned_filename_size); + "rc = [%d]. block_aligned_filename_size = [%zd]\n", + __func__, rc, s->block_aligned_filename_size); goto out_unlock; } (*packet_size) += s->block_aligned_filename_size; @@ -1028,13 +1025,12 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, goto out_unlock; } rc = virt_to_scatterlist(s->decrypted_filename, - s->block_aligned_filename_size, &s->dst_sg, 1); - if (rc != 1) { + s->block_aligned_filename_size, s->dst_sg, 2); + if (rc < 1) { printk(KERN_ERR "%s: Internal error whilst attempting to " "convert decrypted filename memory to scatterlist; " - "expected rc = 1; got rc = [%d]. " - "block_aligned_filename_size = [%zd]\n", __func__, rc, - s->block_aligned_filename_size); + "rc = [%d]. block_aligned_filename_size = [%zd]\n", + __func__, rc, s->block_aligned_filename_size); goto out_free_unlock; } /* The characters in the first block effectively do the job of @@ -1065,7 +1061,7 @@ ecryptfs_parse_tag_70_packet(char **filename, size_t *filename_size, mount_crypt_stat->global_default_fn_cipher_key_bytes); goto out_free_unlock; } - rc = crypto_blkcipher_decrypt_iv(&s->desc, &s->dst_sg, &s->src_sg, + rc = crypto_blkcipher_decrypt_iv(&s->desc, s->dst_sg, s->src_sg, s->block_aligned_filename_size); if (rc) { printk(KERN_ERR "%s: Error attempting to decrypt filename; " diff --git a/fs/ecryptfs/kthread.c b/fs/ecryptfs/kthread.c index 0851ab6980f5..69f994a7d524 100644 --- a/fs/ecryptfs/kthread.c +++ b/fs/ecryptfs/kthread.c @@ -44,7 +44,7 @@ static struct task_struct *ecryptfs_kthread; * @ignored: ignored * * The eCryptfs kernel thread that has the responsibility of getting - * the lower persistent file with RW permissions. + * the lower file with RW permissions. * * Returns zero on success; non-zero otherwise */ @@ -141,8 +141,8 @@ int ecryptfs_privileged_open(struct file **lower_file, int rc = 0; /* Corresponding dput() and mntput() are done when the - * persistent file is fput() when the eCryptfs inode is - * destroyed. */ + * lower file is fput() when all eCryptfs files for the inode are + * released. */ dget(lower_dentry); mntget(lower_mnt); flags |= IS_RDONLY(lower_dentry->d_inode) ? O_RDONLY : O_RDWR; diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index c27c0ecf90bc..89b93389af8e 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -96,7 +96,7 @@ void __ecryptfs_printk(const char *fmt, ...) } /** - * ecryptfs_init_persistent_file + * ecryptfs_init_lower_file * @ecryptfs_dentry: Fully initialized eCryptfs dentry object, with * the lower dentry and the lower mount set * @@ -104,42 +104,70 @@ void __ecryptfs_printk(const char *fmt, ...) * inode. All I/O operations to the lower inode occur through that * file. When the first eCryptfs dentry that interposes with the first * lower dentry for that inode is created, this function creates the - * persistent file struct and associates it with the eCryptfs - * inode. When the eCryptfs inode is destroyed, the file is closed. + * lower file struct and associates it with the eCryptfs + * inode. When all eCryptfs files associated with the inode are released, the + * file is closed. * - * The persistent file will be opened with read/write permissions, if + * The lower file will be opened with read/write permissions, if * possible. Otherwise, it is opened read-only. * - * This function does nothing if a lower persistent file is already + * This function does nothing if a lower file is already * associated with the eCryptfs inode. * * Returns zero on success; non-zero otherwise */ -int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry) +static int ecryptfs_init_lower_file(struct dentry *dentry, + struct file **lower_file) { const struct cred *cred = current_cred(); - struct ecryptfs_inode_info *inode_info = - ecryptfs_inode_to_private(ecryptfs_dentry->d_inode); - int rc = 0; + struct dentry *lower_dentry = ecryptfs_dentry_to_lower(dentry); + struct vfsmount *lower_mnt = ecryptfs_dentry_to_lower_mnt(dentry); + int rc; - if (!inode_info->lower_file) { - struct dentry *lower_dentry; - struct vfsmount *lower_mnt = - ecryptfs_dentry_to_lower_mnt(ecryptfs_dentry); + rc = ecryptfs_privileged_open(lower_file, lower_dentry, lower_mnt, + cred); + if (rc) { + printk(KERN_ERR "Error opening lower file " + "for lower_dentry [0x%p] and lower_mnt [0x%p]; " + "rc = [%d]\n", lower_dentry, lower_mnt, rc); + (*lower_file) = NULL; + } + return rc; +} - lower_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry); - rc = ecryptfs_privileged_open(&inode_info->lower_file, - lower_dentry, lower_mnt, cred); - if (rc) { - printk(KERN_ERR "Error opening lower persistent file " - "for lower_dentry [0x%p] and lower_mnt [0x%p]; " - "rc = [%d]\n", lower_dentry, lower_mnt, rc); - inode_info->lower_file = NULL; - } +int ecryptfs_get_lower_file(struct dentry *dentry) +{ + struct ecryptfs_inode_info *inode_info = + ecryptfs_inode_to_private(dentry->d_inode); + int count, rc = 0; + + mutex_lock(&inode_info->lower_file_mutex); + count = atomic_inc_return(&inode_info->lower_file_count); + if (WARN_ON_ONCE(count < 1)) + rc = -EINVAL; + else if (count == 1) { + rc = ecryptfs_init_lower_file(dentry, + &inode_info->lower_file); + if (rc) + atomic_set(&inode_info->lower_file_count, 0); } + mutex_unlock(&inode_info->lower_file_mutex); return rc; } +void ecryptfs_put_lower_file(struct inode *inode) +{ + struct ecryptfs_inode_info *inode_info; + + inode_info = ecryptfs_inode_to_private(inode); + if (atomic_dec_and_mutex_lock(&inode_info->lower_file_count, + &inode_info->lower_file_mutex)) { + fput(inode_info->lower_file); + inode_info->lower_file = NULL; + mutex_unlock(&inode_info->lower_file_mutex); + } +} + static struct inode *ecryptfs_get_inode(struct inode *lower_inode, struct super_block *sb) { @@ -276,7 +304,7 @@ static void ecryptfs_init_mount_crypt_stat( /** * ecryptfs_parse_options * @sb: The ecryptfs super block - * @options: The options pased to the kernel + * @options: The options passed to the kernel * * Parse mount options: * debug=N - ecryptfs_verbosity level for debug output @@ -840,7 +868,7 @@ static int __init ecryptfs_init(void) } rc = ecryptfs_init_messaging(); if (rc) { - printk(KERN_ERR "Failure occured while attempting to " + printk(KERN_ERR "Failure occurred while attempting to " "initialize the communications channel to " "ecryptfsd\n"); goto out_destroy_kthread; diff --git a/fs/ecryptfs/super.c b/fs/ecryptfs/super.c index bacc882e1ae4..245b517bf1b6 100644 --- a/fs/ecryptfs/super.c +++ b/fs/ecryptfs/super.c @@ -55,6 +55,8 @@ static struct inode *ecryptfs_alloc_inode(struct super_block *sb) if (unlikely(!inode_info)) goto out; ecryptfs_init_crypt_stat(&inode_info->crypt_stat); + mutex_init(&inode_info->lower_file_mutex); + atomic_set(&inode_info->lower_file_count, 0); inode_info->lower_file = NULL; inode = &inode_info->vfs_inode; out: @@ -77,8 +79,7 @@ static void ecryptfs_i_callback(struct rcu_head *head) * * This is used during the final destruction of the inode. All * allocation of memory related to the inode, including allocated - * memory in the crypt_stat struct, will be released here. This - * function also fput()'s the persistent file for the lower inode. + * memory in the crypt_stat struct, will be released here. * There should be no chance that this deallocation will be missed. */ static void ecryptfs_destroy_inode(struct inode *inode) @@ -86,16 +87,7 @@ static void ecryptfs_destroy_inode(struct inode *inode) struct ecryptfs_inode_info *inode_info; inode_info = ecryptfs_inode_to_private(inode); - if (inode_info->lower_file) { - struct dentry *lower_dentry = - inode_info->lower_file->f_dentry; - - BUG_ON(!lower_dentry); - if (lower_dentry->d_inode) { - fput(inode_info->lower_file); - inode_info->lower_file = NULL; - } - } + BUG_ON(inode_info->lower_file); ecryptfs_destroy_crypt_stat(&inode_info->crypt_stat); call_rcu(&inode->i_rcu, ecryptfs_i_callback); } diff --git a/fs/eventpoll.c b/fs/eventpoll.c index ed38801b57a7..f9cfd168fbe2 100644 --- a/fs/eventpoll.c +++ b/fs/eventpoll.c @@ -181,7 +181,7 @@ struct eventpoll { /* * This is a single linked list that chains all the "struct epitem" that - * happened while transfering ready events to userspace w/out + * happened while transferring ready events to userspace w/out * holding ->lock. */ struct epitem *ovflist; @@ -606,7 +606,7 @@ static void ep_free(struct eventpoll *ep) * We do not need to hold "ep->mtx" here because the epoll file * is on the way to be removed and no one has references to it * anymore. The only hit might come from eventpoll_release_file() but - * holding "epmutex" is sufficent here. + * holding "epmutex" is sufficient here. */ mutex_lock(&epmutex); @@ -720,7 +720,7 @@ void eventpoll_release_file(struct file *file) /* * We don't want to get "file->f_lock" because it is not * necessary. It is not necessary because we're in the "struct file" - * cleanup path, and this means that noone is using this file anymore. + * cleanup path, and this means that no one is using this file anymore. * So, for example, epoll_ctl() cannot hit here since if we reach this * point, the file counter already went to zero and fget() would fail. * The only hit might come from ep_free() but by holding the mutex @@ -1112,7 +1112,7 @@ static int ep_send_events_proc(struct eventpoll *ep, struct list_head *head, * Trigger mode, we need to insert back inside * the ready list, so that the next call to * epoll_wait() will check again the events - * availability. At this point, noone can insert + * availability. At this point, no one can insert * into ep->rdllist besides us. The epoll_ctl() * callers are locked out by * ep_scan_ready_list() holding "mtx" and the diff --git a/fs/exec.c b/fs/exec.c index 5e62d26a4fec..ea5f748906a8 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -42,7 +42,6 @@ #include <linux/pid_namespace.h> #include <linux/module.h> #include <linux/namei.h> -#include <linux/proc_fs.h> #include <linux/mount.h> #include <linux/security.h> #include <linux/syscalls.h> @@ -55,6 +54,7 @@ #include <linux/fs_struct.h> #include <linux/pipe_fs_i.h> #include <linux/oom.h> +#include <linux/compat.h> #include <asm/uaccess.h> #include <asm/mmu_context.h> @@ -166,8 +166,13 @@ out: } #ifdef CONFIG_MMU - -void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) +/* + * The nascent bprm->mm is not visible until exec_mmap() but it can + * use a lot of memory, account these pages in current->mm temporary + * for oom_badness()->get_mm_rss(). Once exec succeeds or fails, we + * change the counter back via acct_arg_size(0). + */ +static void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) { struct mm_struct *mm = current->mm; long diff = (long)(pages - bprm->vma_pages); @@ -186,7 +191,7 @@ void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) #endif } -struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, +static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, int write) { struct page *page; @@ -194,7 +199,7 @@ struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, #ifdef CONFIG_STACK_GROWSUP if (write) { - ret = expand_stack_downwards(bprm->vma, pos); + ret = expand_downwards(bprm->vma, pos); if (ret < 0) return NULL; } @@ -305,11 +310,11 @@ static bool valid_arg_len(struct linux_binprm *bprm, long len) #else -void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) +static inline void acct_arg_size(struct linux_binprm *bprm, unsigned long pages) { } -struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, +static struct page *get_arg_page(struct linux_binprm *bprm, unsigned long pos, int write) { struct page *page; @@ -398,22 +403,56 @@ err: return err; } +struct user_arg_ptr { +#ifdef CONFIG_COMPAT + bool is_compat; +#endif + union { + const char __user *const __user *native; +#ifdef CONFIG_COMPAT + compat_uptr_t __user *compat; +#endif + } ptr; +}; + +static const char __user *get_user_arg_ptr(struct user_arg_ptr argv, int nr) +{ + const char __user *native; + +#ifdef CONFIG_COMPAT + if (unlikely(argv.is_compat)) { + compat_uptr_t compat; + + if (get_user(compat, argv.ptr.compat + nr)) + return ERR_PTR(-EFAULT); + + return compat_ptr(compat); + } +#endif + + if (get_user(native, argv.ptr.native + nr)) + return ERR_PTR(-EFAULT); + + return native; +} + /* * count() counts the number of strings in array ARGV. */ -static int count(const char __user * const __user * argv, int max) +static int count(struct user_arg_ptr argv, int max) { int i = 0; - if (argv != NULL) { + if (argv.ptr.native != NULL) { for (;;) { - const char __user * p; + const char __user *p = get_user_arg_ptr(argv, i); - if (get_user(p, argv)) - return -EFAULT; if (!p) break; - argv++; + + if (IS_ERR(p)) + return -EFAULT; + if (i++ >= max) return -E2BIG; @@ -430,7 +469,7 @@ static int count(const char __user * const __user * argv, int max) * processes's memory to the new process's stack. The call to get_user_pages() * ensures the destination page is created and not swapped out. */ -static int copy_strings(int argc, const char __user *const __user *argv, +static int copy_strings(int argc, struct user_arg_ptr argv, struct linux_binprm *bprm) { struct page *kmapped_page = NULL; @@ -443,16 +482,18 @@ static int copy_strings(int argc, const char __user *const __user *argv, int len; unsigned long pos; - if (get_user(str, argv+argc) || - !(len = strnlen_user(str, MAX_ARG_STRLEN))) { - ret = -EFAULT; + ret = -EFAULT; + str = get_user_arg_ptr(argv, argc); + if (IS_ERR(str)) goto out; - } - if (!valid_arg_len(bprm, len)) { - ret = -E2BIG; + len = strnlen_user(str, MAX_ARG_STRLEN); + if (!len) + goto out; + + ret = -E2BIG; + if (!valid_arg_len(bprm, len)) goto out; - } /* We're going to work our way backwords. */ pos = bprm->p; @@ -519,14 +560,19 @@ out: /* * Like copy_strings, but get argv and its values from kernel memory. */ -int copy_strings_kernel(int argc, const char *const *argv, +int copy_strings_kernel(int argc, const char *const *__argv, struct linux_binprm *bprm) { int r; mm_segment_t oldfs = get_fs(); + struct user_arg_ptr argv = { + .ptr.native = (const char __user *const __user *)__argv, + }; + set_fs(KERNEL_DS); - r = copy_strings(argc, (const char __user *const __user *)argv, bprm); + r = copy_strings(argc, argv, bprm); set_fs(oldfs); + return r; } EXPORT_SYMBOL(copy_strings_kernel); @@ -553,7 +599,7 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift) unsigned long length = old_end - old_start; unsigned long new_start = old_start - shift; unsigned long new_end = old_end - shift; - struct mmu_gather *tlb; + struct mmu_gather tlb; BUG_ON(new_start > new_end); @@ -579,12 +625,12 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift) return -ENOMEM; lru_add_drain(); - tlb = tlb_gather_mmu(mm, 0); + tlb_gather_mmu(&tlb, mm, 0); if (new_end > old_start) { /* * when the old and new regions overlap clear from new_end. */ - free_pgd_range(tlb, new_end, old_end, new_end, + free_pgd_range(&tlb, new_end, old_end, new_end, vma->vm_next ? vma->vm_next->vm_start : 0); } else { /* @@ -593,10 +639,10 @@ static int shift_arg_pages(struct vm_area_struct *vma, unsigned long shift) * have constraints on va-space that make this illegal (IA64) - * for the others its just a little faster. */ - free_pgd_range(tlb, old_start, old_end, new_end, + free_pgd_range(&tlb, old_start, old_end, new_end, vma->vm_next ? vma->vm_next->vm_start : 0); } - tlb_finish_mmu(tlb, new_end, old_end); + tlb_finish_mmu(&tlb, new_end, old_end); /* * Shrink the vma to just the new range. Always succeeds. @@ -1004,6 +1050,7 @@ char *get_task_comm(char *buf, struct task_struct *tsk) task_unlock(tsk); return buf; } +EXPORT_SYMBOL_GPL(get_task_comm); void set_task_comm(struct task_struct *tsk, char *buf) { @@ -1379,10 +1426,10 @@ EXPORT_SYMBOL(search_binary_handler); /* * sys_execve() executes a new program. */ -int do_execve(const char * filename, - const char __user *const __user *argv, - const char __user *const __user *envp, - struct pt_regs * regs) +static int do_execve_common(const char *filename, + struct user_arg_ptr argv, + struct user_arg_ptr envp, + struct pt_regs *regs) { struct linux_binprm *bprm; struct file *file; @@ -1489,6 +1536,34 @@ out_ret: return retval; } +int do_execve(const char *filename, + const char __user *const __user *__argv, + const char __user *const __user *__envp, + struct pt_regs *regs) +{ + struct user_arg_ptr argv = { .ptr.native = __argv }; + struct user_arg_ptr envp = { .ptr.native = __envp }; + return do_execve_common(filename, argv, envp, regs); +} + +#ifdef CONFIG_COMPAT +int compat_do_execve(char *filename, + compat_uptr_t __user *__argv, + compat_uptr_t __user *__envp, + struct pt_regs *regs) +{ + struct user_arg_ptr argv = { + .is_compat = true, + .ptr.compat = __argv, + }; + struct user_arg_ptr envp = { + .is_compat = true, + .ptr.compat = __envp, + }; + return do_execve_common(filename, argv, envp, regs); +} +#endif + void set_binfmt(struct linux_binfmt *new) { struct mm_struct *mm = current->mm; @@ -1548,6 +1623,41 @@ expand_fail: return ret; } +static int cn_print_exe_file(struct core_name *cn) +{ + struct file *exe_file; + char *pathbuf, *path, *p; + int ret; + + exe_file = get_mm_exe_file(current->mm); + if (!exe_file) + return cn_printf(cn, "(unknown)"); + + pathbuf = kmalloc(PATH_MAX, GFP_TEMPORARY); + if (!pathbuf) { + ret = -ENOMEM; + goto put_exe_file; + } + + path = d_path(&exe_file->f_path, pathbuf, PATH_MAX); + if (IS_ERR(path)) { + ret = PTR_ERR(path); + goto free_buf; + } + + for (p = path; *p; p++) + if (*p == '/') + *p = '!'; + + ret = cn_printf(cn, "%s", path); + +free_buf: + kfree(pathbuf); +put_exe_file: + fput(exe_file); + return ret; +} + /* format_corename will inspect the pattern parameter, and output a * name into corename, which must have space for at least * CORENAME_MAX_SIZE bytes plus one byte for the zero terminator. @@ -1619,6 +1729,9 @@ static int format_corename(struct core_name *cn, long signr) case 'e': err = cn_printf(cn, "%s", current->comm); break; + case 'E': + err = cn_print_exe_file(cn); + break; /* core limit size */ case 'c': err = cn_printf(cn, "%lu", @@ -1659,6 +1772,7 @@ static int zap_process(struct task_struct *start, int exit_code) t = start; do { + task_clear_group_stop_pending(t); if (t != current && t->mm) { sigaddset(&t->pending.signal, SIGKILL); signal_wake_up(t, 1); diff --git a/fs/exofs/common.h b/fs/exofs/common.h index 5e74ad3d4009..3bbd46956d77 100644 --- a/fs/exofs/common.h +++ b/fs/exofs/common.h @@ -115,7 +115,7 @@ struct exofs_sb_stats { * Describes the raid used in the FS. It is part of the device table. * This here is taken from the pNFS-objects definition. In exofs we * use one raid policy through-out the filesystem. (NOTE: the funny - * alignment at begining. We take care of it at exofs_device_table. + * alignment at beginning. We take care of it at exofs_device_table. */ struct exofs_dt_data_map { __le32 cb_num_comps; @@ -136,7 +136,7 @@ struct exofs_dt_device_info { u8 systemid[OSD_SYSTEMID_LEN]; __le64 long_name_offset; /* If !0 then offset-in-file */ __le32 osdname_len; /* */ - u8 osdname[44]; /* Embbeded, Ususally an asci uuid */ + u8 osdname[44]; /* Embbeded, Usually an asci uuid */ } __packed; /* diff --git a/fs/ext2/balloc.c b/fs/ext2/balloc.c index 0d06f4e75699..8f44cef1b3ef 100644 --- a/fs/ext2/balloc.c +++ b/fs/ext2/balloc.c @@ -850,7 +850,7 @@ static int find_next_reservable_window( rsv_window_remove(sb, my_rsv); /* - * Let's book the whole avaliable window for now. We will check the + * Let's book the whole available window for now. We will check the * disk bitmap later and then, if there are free blocks then we adjust * the window size if it's larger than requested. * Otherwise, we will remove this node from the tree next time @@ -1357,9 +1357,9 @@ retry_alloc: goto allocated; } /* - * We may end up a bogus ealier ENOSPC error due to + * We may end up a bogus earlier ENOSPC error due to * filesystem is "full" of reservations, but - * there maybe indeed free blocks avaliable on disk + * there maybe indeed free blocks available on disk * In this case, we just forget about the reservations * just do block allocation as without reservations. */ diff --git a/fs/ext2/inode.c b/fs/ext2/inode.c index c47f706878b5..788e09a07f7e 100644 --- a/fs/ext2/inode.c +++ b/fs/ext2/inode.c @@ -305,7 +305,7 @@ static ext2_fsblk_t ext2_find_near(struct inode *inode, Indirect *ind) return ind->bh->b_blocknr; /* - * It is going to be refered from inode itself? OK, just put it into + * It is going to be referred from inode itself? OK, just put it into * the same cylinder group then. */ bg_start = ext2_group_first_block_no(inode->i_sb, ei->i_block_group); @@ -913,7 +913,7 @@ static inline int all_zeroes(__le32 *p, __le32 *q) * * When we do truncate() we may have to clean the ends of several indirect * blocks but leave the blocks themselves alive. Block is partially - * truncated if some data below the new i_size is refered from it (and + * truncated if some data below the new i_size is referred from it (and * it is on the path to the first completely truncated data block, indeed). * We have to free the top of that path along with everything to the right * of the path. Since no allocation past the truncation point is possible @@ -990,7 +990,7 @@ no_top: * @p: array of block numbers * @q: points immediately past the end of array * - * We are freeing all blocks refered from that array (numbers are + * We are freeing all blocks referred from that array (numbers are * stored as little-endian 32-bit) and updating @inode->i_blocks * appropriately. */ @@ -1030,7 +1030,7 @@ static inline void ext2_free_data(struct inode *inode, __le32 *p, __le32 *q) * @q: pointer immediately past the end of array * @depth: depth of the branches to free * - * We are freeing all blocks refered from these branches (numbers are + * We are freeing all blocks referred from these branches (numbers are * stored as little-endian 32-bit) and updating @inode->i_blocks * appropriately. */ diff --git a/fs/ext2/super.c b/fs/ext2/super.c index 7731695e65d9..1dd62ed35b85 100644 --- a/fs/ext2/super.c +++ b/fs/ext2/super.c @@ -898,7 +898,8 @@ static int ext2_fill_super(struct super_block *sb, void *data, int silent) brelse(bh); if (!sb_set_blocksize(sb, blocksize)) { - ext2_msg(sb, KERN_ERR, "error: blocksize is too small"); + ext2_msg(sb, KERN_ERR, + "error: bad blocksize %d", blocksize); goto failed_sbi; } @@ -1382,7 +1383,7 @@ static struct dentry *ext2_mount(struct file_system_type *fs_type, /* Read data from quotafile - avoid pagecache and such because we cannot afford * acquiring the locks... As quota files are never truncated and quota code - * itself serializes the operations (and noone else should touch the files) + * itself serializes the operations (and no one else should touch the files) * we don't have to be afraid of races */ static ssize_t ext2_quota_read(struct super_block *sb, int type, char *data, size_t len, loff_t off) diff --git a/fs/ext2/xattr.c b/fs/ext2/xattr.c index c2e4dce984d2..529970617a21 100644 --- a/fs/ext2/xattr.c +++ b/fs/ext2/xattr.c @@ -35,7 +35,7 @@ * +------------------+ * * The block header is followed by multiple entry descriptors. These entry - * descriptors are variable in size, and alligned to EXT2_XATTR_PAD + * descriptors are variable in size, and aligned to EXT2_XATTR_PAD * byte boundaries. The entry descriptors are sorted by attribute name, * so that two extended attribute blocks can be compared efficiently. * diff --git a/fs/ext3/balloc.c b/fs/ext3/balloc.c index 153242187fce..fe52297e31ad 100644 --- a/fs/ext3/balloc.c +++ b/fs/ext3/balloc.c @@ -590,7 +590,7 @@ do_more: BUFFER_TRACE(debug_bh, "Deleted!"); if (!bh2jh(bitmap_bh)->b_committed_data) BUFFER_TRACE(debug_bh, - "No commited data in bitmap"); + "No committed data in bitmap"); BUFFER_TRACE2(debug_bh, bitmap_bh, "bitmap"); __brelse(debug_bh); } @@ -1063,7 +1063,7 @@ static int find_next_reservable_window( rsv_window_remove(sb, my_rsv); /* - * Let's book the whole avaliable window for now. We will check the + * Let's book the whole available window for now. We will check the * disk bitmap later and then, if there are free blocks then we adjust * the window size if it's larger than requested. * Otherwise, we will remove this node from the tree next time @@ -1456,7 +1456,7 @@ static int ext3_has_free_blocks(struct ext3_sb_info *sbi) * * ext3_should_retry_alloc() is called when ENOSPC is returned, and if * it is profitable to retry the operation, this function will wait - * for the current or commiting transaction to complete, and then + * for the current or committing transaction to complete, and then * return TRUE. * * if the total number of retries exceed three times, return FALSE. @@ -1632,9 +1632,9 @@ retry_alloc: goto allocated; } /* - * We may end up a bogus ealier ENOSPC error due to + * We may end up a bogus earlier ENOSPC error due to * filesystem is "full" of reservations, but - * there maybe indeed free blocks avaliable on disk + * there maybe indeed free blocks available on disk * In this case, we just forget about the reservations * just do block allocation as without reservations. */ diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index fe2541d250e4..3451d23c3bae 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -2055,7 +2055,7 @@ static inline int all_zeroes(__le32 *p, __le32 *q) * * When we do truncate() we may have to clean the ends of several * indirect blocks but leave the blocks themselves alive. Block is - * partially truncated if some data below the new i_size is refered + * partially truncated if some data below the new i_size is referred * from it (and it is on the path to the first completely truncated * data block, indeed). We have to free the top of that path along * with everything to the right of the path. Since no allocation @@ -2184,7 +2184,7 @@ static void ext3_clear_blocks(handle_t *handle, struct inode *inode, * @first: array of block numbers * @last: points immediately past the end of array * - * We are freeing all blocks refered from that array (numbers are stored as + * We are freeing all blocks referred from that array (numbers are stored as * little-endian 32-bit) and updating @inode->i_blocks appropriately. * * We accumulate contiguous runs of blocks to free. Conveniently, if these @@ -2272,7 +2272,7 @@ static void ext3_free_data(handle_t *handle, struct inode *inode, * @last: pointer immediately past the end of array * @depth: depth of the branches to free * - * We are freeing all blocks refered from these branches (numbers are + * We are freeing all blocks referred from these branches (numbers are * stored as little-endian 32-bit) and updating @inode->i_blocks * appropriately. */ @@ -3291,7 +3291,7 @@ static int ext3_writepage_trans_blocks(struct inode *inode) if (ext3_should_journal_data(inode)) ret = 3 * (bpp + indirects) + 2; else - ret = 2 * (bpp + indirects) + 2; + ret = 2 * (bpp + indirects) + indirects + 2; #ifdef CONFIG_QUOTA /* We know that structure was already allocated during dquot_initialize so @@ -3392,7 +3392,7 @@ int ext3_mark_inode_dirty(handle_t *handle, struct inode *inode) * so would cause a commit on atime updates, which we don't bother doing. * We handle synchronous inodes at the highest possible level. */ -void ext3_dirty_inode(struct inode *inode) +void ext3_dirty_inode(struct inode *inode, int flags) { handle_t *current_handle = ext3_journal_current_handle(); handle_t *handle; diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 32f3b8695859..34b6d9bfc48a 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -1416,10 +1416,19 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, frame->at = entries; frame->bh = bh; bh = bh2; + /* + * Mark buffers dirty here so that if do_split() fails we write a + * consistent set of buffers to disk. + */ + ext3_journal_dirty_metadata(handle, frame->bh); + ext3_journal_dirty_metadata(handle, bh); de = do_split(handle,dir, &bh, frame, &hinfo, &retval); - dx_release (frames); - if (!(de)) + if (!de) { + ext3_mark_inode_dirty(handle, dir); + dx_release(frames); return retval; + } + dx_release(frames); return add_dirent_to_buf(handle, dentry, inode, de, bh); } @@ -2189,6 +2198,7 @@ static int ext3_symlink (struct inode * dir, handle_t *handle; struct inode * inode; int l, err, retries = 0; + int credits; l = strlen(symname)+1; if (l > dir->i_sb->s_blocksize) @@ -2196,10 +2206,26 @@ static int ext3_symlink (struct inode * dir, dquot_initialize(dir); + if (l > EXT3_N_BLOCKS * 4) { + /* + * For non-fast symlinks, we just allocate inode and put it on + * orphan list in the first transaction => we need bitmap, + * group descriptor, sb, inode block, quota blocks. + */ + credits = 4 + EXT3_MAXQUOTAS_INIT_BLOCKS(dir->i_sb); + } else { + /* + * Fast symlink. We have to add entry to directory + * (EXT3_DATA_TRANS_BLOCKS + EXT3_INDEX_EXTRA_TRANS_BLOCKS), + * allocate new inode (bitmap, group descriptor, inode block, + * quota blocks, sb is already counted in previous macros). + */ + credits = EXT3_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT3_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT3_MAXQUOTAS_INIT_BLOCKS(dir->i_sb); + } retry: - handle = ext3_journal_start(dir, EXT3_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT3_INDEX_EXTRA_TRANS_BLOCKS + 5 + - EXT3_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); + handle = ext3_journal_start(dir, credits); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2211,21 +2237,45 @@ retry: if (IS_ERR(inode)) goto out_stop; - if (l > sizeof (EXT3_I(inode)->i_data)) { + if (l > EXT3_N_BLOCKS * 4) { inode->i_op = &ext3_symlink_inode_operations; ext3_set_aops(inode); /* - * page_symlink() calls into ext3_prepare/commit_write. - * We have a transaction open. All is sweetness. It also sets - * i_size in generic_commit_write(). + * We cannot call page_symlink() with transaction started + * because it calls into ext3_write_begin() which acquires page + * lock which ranks below transaction start (and it can also + * wait for journal commit if we are running out of space). So + * we have to stop transaction now and restart it when symlink + * contents is written. + * + * To keep fs consistent in case of crash, we have to put inode + * to orphan list in the mean time. */ + drop_nlink(inode); + err = ext3_orphan_add(handle, inode); + ext3_journal_stop(handle); + if (err) + goto err_drop_inode; err = __page_symlink(inode, symname, l, 1); + if (err) + goto err_drop_inode; + /* + * Now inode is being linked into dir (EXT3_DATA_TRANS_BLOCKS + * + EXT3_INDEX_EXTRA_TRANS_BLOCKS), inode is also modified + */ + handle = ext3_journal_start(dir, + EXT3_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT3_INDEX_EXTRA_TRANS_BLOCKS + 1); + if (IS_ERR(handle)) { + err = PTR_ERR(handle); + goto err_drop_inode; + } + inc_nlink(inode); + err = ext3_orphan_del(handle, inode); if (err) { + ext3_journal_stop(handle); drop_nlink(inode); - unlock_new_inode(inode); - ext3_mark_inode_dirty(handle, inode); - iput (inode); - goto out_stop; + goto err_drop_inode; } } else { inode->i_op = &ext3_fast_symlink_inode_operations; @@ -2239,6 +2289,10 @@ out_stop: if (err == -ENOSPC && ext3_should_retry_alloc(dir->i_sb, &retries)) goto retry; return err; +err_drop_inode: + unlock_new_inode(inode); + iput(inode); + return err; } static int ext3_link (struct dentry * old_dentry, diff --git a/fs/ext3/resize.c b/fs/ext3/resize.c index 108b142e11ed..7916e4ce166a 100644 --- a/fs/ext3/resize.c +++ b/fs/ext3/resize.c @@ -1009,7 +1009,7 @@ int ext3_group_extend(struct super_block *sb, struct ext3_super_block *es, if (test_opt(sb, DEBUG)) printk(KERN_DEBUG "EXT3-fs: extending last group from "E3FSBLK - " upto "E3FSBLK" blocks\n", + " up to "E3FSBLK" blocks\n", o_blocks_count, n_blocks_count); if (n_blocks_count == 0 || n_blocks_count == o_blocks_count) diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 071689f86e18..aad153ef6b78 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -36,6 +36,7 @@ #include <linux/quotaops.h> #include <linux/seq_file.h> #include <linux/log2.h> +#include <linux/cleancache.h> #include <asm/uaccess.h> @@ -1367,6 +1368,7 @@ static int ext3_setup_super(struct super_block *sb, struct ext3_super_block *es, } else { ext3_msg(sb, KERN_INFO, "using internal journal"); } + cleancache_init_fs(sb); return res; } @@ -2925,7 +2927,7 @@ static int ext3_quota_on(struct super_block *sb, int type, int format_id, /* Read data from quotafile - avoid pagecache and such because we cannot afford * acquiring the locks... As quota files are never truncated and quota code - * itself serializes the operations (and noone else should touch the files) + * itself serializes the operations (and no one else should touch the files) * we don't have to be afraid of races */ static ssize_t ext3_quota_read(struct super_block *sb, int type, char *data, size_t len, loff_t off) diff --git a/fs/ext4/Makefile b/fs/ext4/Makefile index c947e36eda6c..04109460ba9e 100644 --- a/fs/ext4/Makefile +++ b/fs/ext4/Makefile @@ -6,7 +6,8 @@ obj-$(CONFIG_EXT4_FS) += ext4.o ext4-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o page-io.o \ ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o \ - ext4_jbd2.o migrate.o mballoc.o block_validity.o move_extent.o + ext4_jbd2.o migrate.o mballoc.o block_validity.o move_extent.o \ + mmp.o ext4-$(CONFIG_EXT4_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o ext4-$(CONFIG_EXT4_FS_POSIX_ACL) += acl.o diff --git a/fs/ext4/balloc.c b/fs/ext4/balloc.c index 97b970e7dd13..264f6949511e 100644 --- a/fs/ext4/balloc.c +++ b/fs/ext4/balloc.c @@ -362,130 +362,6 @@ ext4_read_block_bitmap(struct super_block *sb, ext4_group_t block_group) } /** - * ext4_add_groupblocks() -- Add given blocks to an existing group - * @handle: handle to this transaction - * @sb: super block - * @block: start physcial block to add to the block group - * @count: number of blocks to free - * - * This marks the blocks as free in the bitmap. We ask the - * mballoc to reload the buddy after this by setting group - * EXT4_GROUP_INFO_NEED_INIT_BIT flag - */ -void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, - ext4_fsblk_t block, unsigned long count) -{ - struct buffer_head *bitmap_bh = NULL; - struct buffer_head *gd_bh; - ext4_group_t block_group; - ext4_grpblk_t bit; - unsigned int i; - struct ext4_group_desc *desc; - struct ext4_sb_info *sbi = EXT4_SB(sb); - int err = 0, ret, blk_free_count; - ext4_grpblk_t blocks_freed; - struct ext4_group_info *grp; - - ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1); - - ext4_get_group_no_and_offset(sb, block, &block_group, &bit); - grp = ext4_get_group_info(sb, block_group); - /* - * Check to see if we are freeing blocks across a group - * boundary. - */ - if (bit + count > EXT4_BLOCKS_PER_GROUP(sb)) { - goto error_return; - } - bitmap_bh = ext4_read_block_bitmap(sb, block_group); - if (!bitmap_bh) - goto error_return; - desc = ext4_get_group_desc(sb, block_group, &gd_bh); - if (!desc) - goto error_return; - - if (in_range(ext4_block_bitmap(sb, desc), block, count) || - in_range(ext4_inode_bitmap(sb, desc), block, count) || - in_range(block, ext4_inode_table(sb, desc), sbi->s_itb_per_group) || - in_range(block + count - 1, ext4_inode_table(sb, desc), - sbi->s_itb_per_group)) { - ext4_error(sb, "Adding blocks in system zones - " - "Block = %llu, count = %lu", - block, count); - goto error_return; - } - - /* - * We are about to add blocks to the bitmap, - * so we need undo access. - */ - BUFFER_TRACE(bitmap_bh, "getting undo access"); - err = ext4_journal_get_undo_access(handle, bitmap_bh); - if (err) - goto error_return; - - /* - * We are about to modify some metadata. Call the journal APIs - * to unshare ->b_data if a currently-committing transaction is - * using it - */ - BUFFER_TRACE(gd_bh, "get_write_access"); - err = ext4_journal_get_write_access(handle, gd_bh); - if (err) - goto error_return; - /* - * make sure we don't allow a parallel init on other groups in the - * same buddy cache - */ - down_write(&grp->alloc_sem); - for (i = 0, blocks_freed = 0; i < count; i++) { - BUFFER_TRACE(bitmap_bh, "clear bit"); - if (!ext4_clear_bit_atomic(ext4_group_lock_ptr(sb, block_group), - bit + i, bitmap_bh->b_data)) { - ext4_error(sb, "bit already cleared for block %llu", - (ext4_fsblk_t)(block + i)); - BUFFER_TRACE(bitmap_bh, "bit already cleared"); - } else { - blocks_freed++; - } - } - ext4_lock_group(sb, block_group); - blk_free_count = blocks_freed + ext4_free_blks_count(sb, desc); - ext4_free_blks_set(sb, desc, blk_free_count); - desc->bg_checksum = ext4_group_desc_csum(sbi, block_group, desc); - ext4_unlock_group(sb, block_group); - percpu_counter_add(&sbi->s_freeblocks_counter, blocks_freed); - - if (sbi->s_log_groups_per_flex) { - ext4_group_t flex_group = ext4_flex_group(sbi, block_group); - atomic_add(blocks_freed, - &sbi->s_flex_groups[flex_group].free_blocks); - } - /* - * request to reload the buddy with the - * new bitmap information - */ - set_bit(EXT4_GROUP_INFO_NEED_INIT_BIT, &(grp->bb_state)); - grp->bb_free += blocks_freed; - up_write(&grp->alloc_sem); - - /* We dirtied the bitmap block */ - BUFFER_TRACE(bitmap_bh, "dirtied bitmap block"); - err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); - - /* And the group descriptor block */ - BUFFER_TRACE(gd_bh, "dirtied group descriptor block"); - ret = ext4_handle_dirty_metadata(handle, NULL, gd_bh); - if (!err) - err = ret; - -error_return: - brelse(bitmap_bh); - ext4_std_error(sb, err); - return; -} - -/** * ext4_has_free_blocks() * @sbi: in-core super block structure. * @nblocks: number of needed blocks @@ -493,7 +369,8 @@ error_return: * Check if filesystem has nblocks free & available for allocation. * On success return 1, return 0 on failure. */ -static int ext4_has_free_blocks(struct ext4_sb_info *sbi, s64 nblocks) +static int ext4_has_free_blocks(struct ext4_sb_info *sbi, + s64 nblocks, unsigned int flags) { s64 free_blocks, dirty_blocks, root_blocks; struct percpu_counter *fbc = &sbi->s_freeblocks_counter; @@ -507,11 +384,6 @@ static int ext4_has_free_blocks(struct ext4_sb_info *sbi, s64 nblocks) EXT4_FREEBLOCKS_WATERMARK) { free_blocks = percpu_counter_sum_positive(fbc); dirty_blocks = percpu_counter_sum_positive(dbc); - if (dirty_blocks < 0) { - printk(KERN_CRIT "Dirty block accounting " - "went wrong %lld\n", - (long long)dirty_blocks); - } } /* Check whether we have space after * accounting for current dirty blocks & root reserved blocks. @@ -522,7 +394,9 @@ static int ext4_has_free_blocks(struct ext4_sb_info *sbi, s64 nblocks) /* Hm, nope. Are (enough) root reserved blocks available? */ if (sbi->s_resuid == current_fsuid() || ((sbi->s_resgid != 0) && in_group_p(sbi->s_resgid)) || - capable(CAP_SYS_RESOURCE)) { + capable(CAP_SYS_RESOURCE) || + (flags & EXT4_MB_USE_ROOT_BLOCKS)) { + if (free_blocks >= (nblocks + dirty_blocks)) return 1; } @@ -531,9 +405,9 @@ static int ext4_has_free_blocks(struct ext4_sb_info *sbi, s64 nblocks) } int ext4_claim_free_blocks(struct ext4_sb_info *sbi, - s64 nblocks) + s64 nblocks, unsigned int flags) { - if (ext4_has_free_blocks(sbi, nblocks)) { + if (ext4_has_free_blocks(sbi, nblocks, flags)) { percpu_counter_add(&sbi->s_dirtyblocks_counter, nblocks); return 0; } else @@ -547,14 +421,14 @@ int ext4_claim_free_blocks(struct ext4_sb_info *sbi, * * ext4_should_retry_alloc() is called when ENOSPC is returned, and if * it is profitable to retry the operation, this function will wait - * for the current or commiting transaction to complete, and then + * for the current or committing transaction to complete, and then * return TRUE. * * if the total number of retries exceed three times, return FALSE. */ int ext4_should_retry_alloc(struct super_block *sb, int *retries) { - if (!ext4_has_free_blocks(EXT4_SB(sb), 1) || + if (!ext4_has_free_blocks(EXT4_SB(sb), 1, 0) || (*retries)++ > 3 || !EXT4_SB(sb)->s_journal) return 0; @@ -577,7 +451,8 @@ int ext4_should_retry_alloc(struct super_block *sb, int *retries) * error stores in errp pointer */ ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode, - ext4_fsblk_t goal, unsigned long *count, int *errp) + ext4_fsblk_t goal, unsigned int flags, + unsigned long *count, int *errp) { struct ext4_allocation_request ar; ext4_fsblk_t ret; @@ -587,6 +462,7 @@ ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode, ar.inode = inode; ar.goal = goal; ar.len = count ? *count : 1; + ar.flags = flags; ret = ext4_mb_new_blocks(handle, &ar, errp); if (count) diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 4daaf2b753f4..1921392cd708 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -108,7 +108,8 @@ typedef unsigned int ext4_group_t; #define EXT4_MB_DELALLOC_RESERVED 0x0400 /* We are doing stream allocation */ #define EXT4_MB_STREAM_ALLOC 0x0800 - +/* Use reserved root blocks if needed */ +#define EXT4_MB_USE_ROOT_BLOCKS 0x1000 struct ext4_allocation_request { /* target inode for block we're allocating */ @@ -209,6 +210,8 @@ struct ext4_io_submit { */ #define EXT4_BAD_INO 1 /* Bad blocks inode */ #define EXT4_ROOT_INO 2 /* Root inode */ +#define EXT4_USR_QUOTA_INO 3 /* User quota inode */ +#define EXT4_GRP_QUOTA_INO 4 /* Group quota inode */ #define EXT4_BOOT_LOADER_INO 5 /* Boot loader inode */ #define EXT4_UNDEL_DIR_INO 6 /* Undelete directory inode */ #define EXT4_RESIZE_INO 7 /* Reserved group descriptors inode */ @@ -512,6 +515,10 @@ struct ext4_new_group_data { /* Convert extent to initialized after IO complete */ #define EXT4_GET_BLOCKS_IO_CONVERT_EXT (EXT4_GET_BLOCKS_CONVERT|\ EXT4_GET_BLOCKS_CREATE_UNINIT_EXT) + /* Punch out blocks of an extent */ +#define EXT4_GET_BLOCKS_PUNCH_OUT_EXT 0x0020 + /* Don't normalize allocation size (used for fallocate) */ +#define EXT4_GET_BLOCKS_NO_NORMALIZE 0x0040 /* * Flags used by ext4_free_blocks @@ -1028,7 +1035,7 @@ struct ext4_super_block { __le16 s_want_extra_isize; /* New inodes should reserve # bytes */ __le32 s_flags; /* Miscellaneous flags */ __le16 s_raid_stride; /* RAID stride */ - __le16 s_mmp_interval; /* # seconds to wait in MMP checking */ + __le16 s_mmp_update_interval; /* # seconds to wait in MMP checking */ __le64 s_mmp_block; /* Block for multi-mount protection */ __le32 s_raid_stripe_width; /* blocks on all data disks (N*stride)*/ __u8 s_log_groups_per_flex; /* FLEX_BG group size */ @@ -1144,6 +1151,9 @@ struct ext4_sb_info { unsigned long s_ext_blocks; unsigned long s_ext_extents; #endif + /* ext4 extent cache stats */ + unsigned long extent_cache_hits; + unsigned long extent_cache_misses; /* for buddy allocator */ struct ext4_group_info ***s_group_info; @@ -1201,6 +1211,9 @@ struct ext4_sb_info { struct ext4_li_request *s_li_request; /* Wait multiplier for lazy initialization thread */ unsigned int s_li_wait_mult; + + /* Kernel thread for multiple mount protection */ + struct task_struct *s_mmp_tsk; }; static inline struct ext4_sb_info *EXT4_SB(struct super_block *sb) @@ -1338,6 +1351,7 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei) #define EXT4_FEATURE_RO_COMPAT_GDT_CSUM 0x0010 #define EXT4_FEATURE_RO_COMPAT_DIR_NLINK 0x0020 #define EXT4_FEATURE_RO_COMPAT_EXTRA_ISIZE 0x0040 +#define EXT4_FEATURE_RO_COMPAT_QUOTA 0x0100 #define EXT4_FEATURE_INCOMPAT_COMPRESSION 0x0001 #define EXT4_FEATURE_INCOMPAT_FILETYPE 0x0002 @@ -1351,13 +1365,29 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei) #define EXT4_FEATURE_INCOMPAT_EA_INODE 0x0400 /* EA in inode */ #define EXT4_FEATURE_INCOMPAT_DIRDATA 0x1000 /* data in dirent */ +#define EXT2_FEATURE_COMPAT_SUPP EXT4_FEATURE_COMPAT_EXT_ATTR +#define EXT2_FEATURE_INCOMPAT_SUPP (EXT4_FEATURE_INCOMPAT_FILETYPE| \ + EXT4_FEATURE_INCOMPAT_META_BG) +#define EXT2_FEATURE_RO_COMPAT_SUPP (EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER| \ + EXT4_FEATURE_RO_COMPAT_LARGE_FILE| \ + EXT4_FEATURE_RO_COMPAT_BTREE_DIR) + +#define EXT3_FEATURE_COMPAT_SUPP EXT4_FEATURE_COMPAT_EXT_ATTR +#define EXT3_FEATURE_INCOMPAT_SUPP (EXT4_FEATURE_INCOMPAT_FILETYPE| \ + EXT4_FEATURE_INCOMPAT_RECOVER| \ + EXT4_FEATURE_INCOMPAT_META_BG) +#define EXT3_FEATURE_RO_COMPAT_SUPP (EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER| \ + EXT4_FEATURE_RO_COMPAT_LARGE_FILE| \ + EXT4_FEATURE_RO_COMPAT_BTREE_DIR) + #define EXT4_FEATURE_COMPAT_SUPP EXT2_FEATURE_COMPAT_EXT_ATTR #define EXT4_FEATURE_INCOMPAT_SUPP (EXT4_FEATURE_INCOMPAT_FILETYPE| \ EXT4_FEATURE_INCOMPAT_RECOVER| \ EXT4_FEATURE_INCOMPAT_META_BG| \ EXT4_FEATURE_INCOMPAT_EXTENTS| \ EXT4_FEATURE_INCOMPAT_64BIT| \ - EXT4_FEATURE_INCOMPAT_FLEX_BG) + EXT4_FEATURE_INCOMPAT_FLEX_BG| \ + EXT4_FEATURE_INCOMPAT_MMP) #define EXT4_FEATURE_RO_COMPAT_SUPP (EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER| \ EXT4_FEATURE_RO_COMPAT_LARGE_FILE| \ EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \ @@ -1590,12 +1620,6 @@ void ext4_get_group_no_and_offset(struct super_block *sb, ext4_fsblk_t blocknr, */ struct ext4_lazy_init { unsigned long li_state; - - wait_queue_head_t li_wait_daemon; - wait_queue_head_t li_wait_task; - struct timer_list li_timer; - struct task_struct *li_task; - struct list_head li_request_list; struct mutex li_list_mtx; }; @@ -1615,6 +1639,67 @@ struct ext4_features { }; /* + * This structure will be used for multiple mount protection. It will be + * written into the block number saved in the s_mmp_block field in the + * superblock. Programs that check MMP should assume that if + * SEQ_FSCK (or any unknown code above SEQ_MAX) is present then it is NOT safe + * to use the filesystem, regardless of how old the timestamp is. + */ +#define EXT4_MMP_MAGIC 0x004D4D50U /* ASCII for MMP */ +#define EXT4_MMP_SEQ_CLEAN 0xFF4D4D50U /* mmp_seq value for clean unmount */ +#define EXT4_MMP_SEQ_FSCK 0xE24D4D50U /* mmp_seq value when being fscked */ +#define EXT4_MMP_SEQ_MAX 0xE24D4D4FU /* maximum valid mmp_seq value */ + +struct mmp_struct { + __le32 mmp_magic; /* Magic number for MMP */ + __le32 mmp_seq; /* Sequence no. updated periodically */ + + /* + * mmp_time, mmp_nodename & mmp_bdevname are only used for information + * purposes and do not affect the correctness of the algorithm + */ + __le64 mmp_time; /* Time last updated */ + char mmp_nodename[64]; /* Node which last updated MMP block */ + char mmp_bdevname[32]; /* Bdev which last updated MMP block */ + + /* + * mmp_check_interval is used to verify if the MMP block has been + * updated on the block device. The value is updated based on the + * maximum time to write the MMP block during an update cycle. + */ + __le16 mmp_check_interval; + + __le16 mmp_pad1; + __le32 mmp_pad2[227]; +}; + +/* arguments passed to the mmp thread */ +struct mmpd_data { + struct buffer_head *bh; /* bh from initial read_mmp_block() */ + struct super_block *sb; /* super block of the fs */ +}; + +/* + * Check interval multiplier + * The MMP block is written every update interval and initially checked every + * update interval x the multiplier (the value is then adapted based on the + * write latency). The reason is that writes can be delayed under load and we + * don't want readers to incorrectly assume that the filesystem is no longer + * in use. + */ +#define EXT4_MMP_CHECK_MULT 2UL + +/* + * Minimum interval for MMP checking in seconds. + */ +#define EXT4_MMP_MIN_CHECK_INTERVAL 5UL + +/* + * Maximum interval for MMP checking in seconds. + */ +#define EXT4_MMP_MAX_CHECK_INTERVAL 300UL + +/* * Function prototypes */ @@ -1638,10 +1723,12 @@ extern int ext4_bg_has_super(struct super_block *sb, ext4_group_t group); extern unsigned long ext4_bg_num_gdb(struct super_block *sb, ext4_group_t group); extern ext4_fsblk_t ext4_new_meta_blocks(handle_t *handle, struct inode *inode, - ext4_fsblk_t goal, unsigned long *count, int *errp); -extern int ext4_claim_free_blocks(struct ext4_sb_info *sbi, s64 nblocks); -extern void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, - ext4_fsblk_t block, unsigned long count); + ext4_fsblk_t goal, + unsigned int flags, + unsigned long *count, + int *errp); +extern int ext4_claim_free_blocks(struct ext4_sb_info *sbi, + s64 nblocks, unsigned int flags); extern ext4_fsblk_t ext4_count_free_blocks(struct super_block *); extern void ext4_check_blocks_bitmap(struct super_block *); extern struct ext4_group_desc * ext4_get_group_desc(struct super_block * sb, @@ -1706,6 +1793,8 @@ extern void ext4_free_blocks(handle_t *handle, struct inode *inode, unsigned long count, int flags); extern int ext4_mb_add_groupinfo(struct super_block *sb, ext4_group_t i, struct ext4_group_desc *desc); +extern void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, + ext4_fsblk_t block, unsigned long count); extern int ext4_trim_fs(struct super_block *, struct fstrim_range *); /* inode.c */ @@ -1724,11 +1813,12 @@ extern int ext4_getattr(struct vfsmount *mnt, struct dentry *dentry, extern void ext4_evict_inode(struct inode *); extern void ext4_clear_inode(struct inode *); extern int ext4_sync_inode(handle_t *, struct inode *); -extern void ext4_dirty_inode(struct inode *); +extern void ext4_dirty_inode(struct inode *, int); extern int ext4_change_inode_journal_flag(struct inode *, int); extern int ext4_get_inode_loc(struct inode *, struct ext4_iloc *); extern int ext4_can_truncate(struct inode *inode); extern void ext4_truncate(struct inode *); +extern int ext4_punch_hole(struct file *file, loff_t offset, loff_t length); extern int ext4_truncate_restart_trans(handle_t *, struct inode *, int nblocks); extern void ext4_set_inode_flags(struct inode *); extern void ext4_get_inode_flags(struct ext4_inode_info *); @@ -1738,6 +1828,8 @@ extern int ext4_writepage_trans_blocks(struct inode *); extern int ext4_chunk_trans_blocks(struct inode *, int nrblocks); extern int ext4_block_truncate_page(handle_t *handle, struct address_space *mapping, loff_t from); +extern int ext4_block_zero_page_range(handle_t *handle, + struct address_space *mapping, loff_t from, loff_t length); extern int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf); extern qsize_t *ext4_get_reserved_space(struct inode *inode); extern void ext4_da_update_reserve_space(struct inode *inode, @@ -1788,6 +1880,10 @@ extern void __ext4_warning(struct super_block *, const char *, unsigned int, __LINE__, ## message) extern void ext4_msg(struct super_block *, const char *, const char *, ...) __attribute__ ((format (printf, 3, 4))); +extern void __dump_mmp_msg(struct super_block *, struct mmp_struct *mmp, + const char *, unsigned int, const char *); +#define dump_mmp_msg(sb, mmp, msg) __dump_mmp_msg(sb, mmp, __func__, \ + __LINE__, msg) extern void __ext4_grp_locked_error(const char *, unsigned int, \ struct super_block *, ext4_group_t, \ unsigned long, ext4_fsblk_t, \ @@ -2064,6 +2160,8 @@ extern int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, extern int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, int flags); extern void ext4_ext_truncate(struct inode *); +extern int ext4_ext_punch_hole(struct file *file, loff_t offset, + loff_t length); extern void ext4_ext_init(struct super_block *); extern void ext4_ext_release(struct super_block *); extern long ext4_fallocate(struct file *file, int mode, loff_t offset, @@ -2092,6 +2190,9 @@ extern int ext4_bio_write_page(struct ext4_io_submit *io, int len, struct writeback_control *wbc); +/* mmp.c */ +extern int ext4_multi_mount_protect(struct super_block *, ext4_fsblk_t); + /* BH_Uninit flag: blocks are allocated but uninitialized on disk */ enum ext4_state_bits { BH_Uninit /* blocks are allocated but uninitialized on disk */ diff --git a/fs/ext4/ext4_jbd2.c b/fs/ext4/ext4_jbd2.c index 6e272ef6ba96..f5240aa15601 100644 --- a/fs/ext4/ext4_jbd2.c +++ b/fs/ext4/ext4_jbd2.c @@ -6,20 +6,6 @@ #include <trace/events/ext4.h> -int __ext4_journal_get_undo_access(const char *where, unsigned int line, - handle_t *handle, struct buffer_head *bh) -{ - int err = 0; - - if (ext4_handle_valid(handle)) { - err = jbd2_journal_get_undo_access(handle, bh); - if (err) - ext4_journal_abort_handle(where, line, __func__, bh, - handle, err); - } - return err; -} - int __ext4_journal_get_write_access(const char *where, unsigned int line, handle_t *handle, struct buffer_head *bh) { diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index e25e99bf7ee1..bb85757689b6 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -86,8 +86,8 @@ #ifdef CONFIG_QUOTA /* Amount of blocks needed for quota update - we know that the structure was - * allocated so we need to update only inode+data */ -#define EXT4_QUOTA_TRANS_BLOCKS(sb) (test_opt(sb, QUOTA) ? 2 : 0) + * allocated so we need to update only data block */ +#define EXT4_QUOTA_TRANS_BLOCKS(sb) (test_opt(sb, QUOTA) ? 1 : 0) /* Amount of blocks needed for quota insert/delete - we do some block writes * but inode, sb and group updates are done only once */ #define EXT4_QUOTA_INIT_BLOCKS(sb) (test_opt(sb, QUOTA) ? (DQUOT_INIT_ALLOC*\ @@ -126,9 +126,6 @@ void ext4_journal_abort_handle(const char *caller, unsigned int line, const char *err_fn, struct buffer_head *bh, handle_t *handle, int err); -int __ext4_journal_get_undo_access(const char *where, unsigned int line, - handle_t *handle, struct buffer_head *bh); - int __ext4_journal_get_write_access(const char *where, unsigned int line, handle_t *handle, struct buffer_head *bh); @@ -146,8 +143,6 @@ int __ext4_handle_dirty_metadata(const char *where, unsigned int line, int __ext4_handle_dirty_super(const char *where, unsigned int line, handle_t *handle, struct super_block *sb); -#define ext4_journal_get_undo_access(handle, bh) \ - __ext4_journal_get_undo_access(__func__, __LINE__, (handle), (bh)) #define ext4_journal_get_write_access(handle, bh) \ __ext4_journal_get_write_access(__func__, __LINE__, (handle), (bh)) #define ext4_forget(handle, is_metadata, inode, bh, block_nr) \ diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index dd2cb5076ff9..5199bac7fc62 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -46,6 +46,13 @@ #include <trace/events/ext4.h> +static int ext4_split_extent(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path, + struct ext4_map_blocks *map, + int split_flag, + int flags); + static int ext4_ext_truncate_extend_restart(handle_t *handle, struct inode *inode, int needed) @@ -192,12 +199,13 @@ static ext4_fsblk_t ext4_ext_find_goal(struct inode *inode, static ext4_fsblk_t ext4_ext_new_meta_block(handle_t *handle, struct inode *inode, struct ext4_ext_path *path, - struct ext4_extent *ex, int *err) + struct ext4_extent *ex, int *err, unsigned int flags) { ext4_fsblk_t goal, newblock; goal = ext4_ext_find_goal(inode, path, le32_to_cpu(ex->ee_block)); - newblock = ext4_new_meta_blocks(handle, inode, goal, NULL, err); + newblock = ext4_new_meta_blocks(handle, inode, goal, flags, + NULL, err); return newblock; } @@ -474,9 +482,43 @@ static void ext4_ext_show_leaf(struct inode *inode, struct ext4_ext_path *path) } ext_debug("\n"); } + +static void ext4_ext_show_move(struct inode *inode, struct ext4_ext_path *path, + ext4_fsblk_t newblock, int level) +{ + int depth = ext_depth(inode); + struct ext4_extent *ex; + + if (depth != level) { + struct ext4_extent_idx *idx; + idx = path[level].p_idx; + while (idx <= EXT_MAX_INDEX(path[level].p_hdr)) { + ext_debug("%d: move %d:%llu in new index %llu\n", level, + le32_to_cpu(idx->ei_block), + ext4_idx_pblock(idx), + newblock); + idx++; + } + + return; + } + + ex = path[depth].p_ext; + while (ex <= EXT_MAX_EXTENT(path[depth].p_hdr)) { + ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n", + le32_to_cpu(ex->ee_block), + ext4_ext_pblock(ex), + ext4_ext_is_uninitialized(ex), + ext4_ext_get_actual_len(ex), + newblock); + ex++; + } +} + #else #define ext4_ext_show_path(inode, path) #define ext4_ext_show_leaf(inode, path) +#define ext4_ext_show_move(inode, path, newblock, level) #endif void ext4_ext_drop_refs(struct ext4_ext_path *path) @@ -792,14 +834,14 @@ static int ext4_ext_insert_index(handle_t *handle, struct inode *inode, * - initializes subtree */ static int ext4_ext_split(handle_t *handle, struct inode *inode, - struct ext4_ext_path *path, - struct ext4_extent *newext, int at) + unsigned int flags, + struct ext4_ext_path *path, + struct ext4_extent *newext, int at) { struct buffer_head *bh = NULL; int depth = ext_depth(inode); struct ext4_extent_header *neh; struct ext4_extent_idx *fidx; - struct ext4_extent *ex; int i = at, k, m, a; ext4_fsblk_t newblock, oldblock; __le32 border; @@ -847,7 +889,7 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, ext_debug("allocate %d blocks for indexes/leaf\n", depth - at); for (a = 0; a < depth - at; a++) { newblock = ext4_ext_new_meta_block(handle, inode, path, - newext, &err); + newext, &err, flags); if (newblock == 0) goto cleanup; ablocks[a] = newblock; @@ -876,7 +918,6 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, neh->eh_max = cpu_to_le16(ext4_ext_space_block(inode, 0)); neh->eh_magic = EXT4_EXT_MAGIC; neh->eh_depth = 0; - ex = EXT_FIRST_EXTENT(neh); /* move remainder of path[depth] to the new leaf */ if (unlikely(path[depth].p_hdr->eh_entries != @@ -888,25 +929,12 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, goto cleanup; } /* start copy from next extent */ - /* TODO: we could do it by single memmove */ - m = 0; - path[depth].p_ext++; - while (path[depth].p_ext <= - EXT_MAX_EXTENT(path[depth].p_hdr)) { - ext_debug("move %d:%llu:[%d]%d in new leaf %llu\n", - le32_to_cpu(path[depth].p_ext->ee_block), - ext4_ext_pblock(path[depth].p_ext), - ext4_ext_is_uninitialized(path[depth].p_ext), - ext4_ext_get_actual_len(path[depth].p_ext), - newblock); - /*memmove(ex++, path[depth].p_ext++, - sizeof(struct ext4_extent)); - neh->eh_entries++;*/ - path[depth].p_ext++; - m++; - } + m = EXT_MAX_EXTENT(path[depth].p_hdr) - path[depth].p_ext++; + ext4_ext_show_move(inode, path, newblock, depth); if (m) { - memmove(ex, path[depth].p_ext-m, sizeof(struct ext4_extent)*m); + struct ext4_extent *ex; + ex = EXT_FIRST_EXTENT(neh); + memmove(ex, path[depth].p_ext, sizeof(struct ext4_extent) * m); le16_add_cpu(&neh->eh_entries, m); } @@ -968,12 +996,8 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, ext_debug("int.index at %d (block %llu): %u -> %llu\n", i, newblock, le32_to_cpu(border), oldblock); - /* copy indexes */ - m = 0; - path[i].p_idx++; - ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx, - EXT_MAX_INDEX(path[i].p_hdr)); + /* move remainder of path[i] to the new index block */ if (unlikely(EXT_MAX_INDEX(path[i].p_hdr) != EXT_LAST_INDEX(path[i].p_hdr))) { EXT4_ERROR_INODE(inode, @@ -982,20 +1006,13 @@ static int ext4_ext_split(handle_t *handle, struct inode *inode, err = -EIO; goto cleanup; } - while (path[i].p_idx <= EXT_MAX_INDEX(path[i].p_hdr)) { - ext_debug("%d: move %d:%llu in new index %llu\n", i, - le32_to_cpu(path[i].p_idx->ei_block), - ext4_idx_pblock(path[i].p_idx), - newblock); - /*memmove(++fidx, path[i].p_idx++, - sizeof(struct ext4_extent_idx)); - neh->eh_entries++; - BUG_ON(neh->eh_entries > neh->eh_max);*/ - path[i].p_idx++; - m++; - } + /* start copy indexes */ + m = EXT_MAX_INDEX(path[i].p_hdr) - path[i].p_idx++; + ext_debug("cur 0x%p, last 0x%p\n", path[i].p_idx, + EXT_MAX_INDEX(path[i].p_hdr)); + ext4_ext_show_move(inode, path, newblock, i); if (m) { - memmove(++fidx, path[i].p_idx - m, + memmove(++fidx, path[i].p_idx, sizeof(struct ext4_extent_idx) * m); le16_add_cpu(&neh->eh_entries, m); } @@ -1056,8 +1073,9 @@ cleanup: * just created block */ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, - struct ext4_ext_path *path, - struct ext4_extent *newext) + unsigned int flags, + struct ext4_ext_path *path, + struct ext4_extent *newext) { struct ext4_ext_path *curp = path; struct ext4_extent_header *neh; @@ -1065,7 +1083,8 @@ static int ext4_ext_grow_indepth(handle_t *handle, struct inode *inode, ext4_fsblk_t newblock; int err = 0; - newblock = ext4_ext_new_meta_block(handle, inode, path, newext, &err); + newblock = ext4_ext_new_meta_block(handle, inode, path, + newext, &err, flags); if (newblock == 0) return err; @@ -1140,8 +1159,9 @@ out: * if no free index is found, then it requests in-depth growing. */ static int ext4_ext_create_new_leaf(handle_t *handle, struct inode *inode, - struct ext4_ext_path *path, - struct ext4_extent *newext) + unsigned int flags, + struct ext4_ext_path *path, + struct ext4_extent *newext) { struct ext4_ext_path *curp; int depth, i, err = 0; @@ -1161,7 +1181,7 @@ repeat: if (EXT_HAS_FREE_INDEX(curp)) { /* if we found index with free entry, then use that * entry: create all needed subtree and add new leaf */ - err = ext4_ext_split(handle, inode, path, newext, i); + err = ext4_ext_split(handle, inode, flags, path, newext, i); if (err) goto out; @@ -1174,7 +1194,8 @@ repeat: err = PTR_ERR(path); } else { /* tree is full, time to grow in depth */ - err = ext4_ext_grow_indepth(handle, inode, path, newext); + err = ext4_ext_grow_indepth(handle, inode, flags, + path, newext); if (err) goto out; @@ -1563,7 +1584,7 @@ ext4_can_extents_be_merged(struct inode *inode, struct ext4_extent *ex1, * Returns 0 if the extents (ex and ex+1) were _not_ merged and returns * 1 if they got merged. */ -static int ext4_ext_try_to_merge(struct inode *inode, +static int ext4_ext_try_to_merge_right(struct inode *inode, struct ext4_ext_path *path, struct ext4_extent *ex) { @@ -1603,6 +1624,31 @@ static int ext4_ext_try_to_merge(struct inode *inode, } /* + * This function tries to merge the @ex extent to neighbours in the tree. + * return 1 if merge left else 0. + */ +static int ext4_ext_try_to_merge(struct inode *inode, + struct ext4_ext_path *path, + struct ext4_extent *ex) { + struct ext4_extent_header *eh; + unsigned int depth; + int merge_done = 0; + int ret = 0; + + depth = ext_depth(inode); + BUG_ON(path[depth].p_hdr == NULL); + eh = path[depth].p_hdr; + + if (ex > EXT_FIRST_EXTENT(eh)) + merge_done = ext4_ext_try_to_merge_right(inode, path, ex - 1); + + if (!merge_done) + ret = ext4_ext_try_to_merge_right(inode, path, ex); + + return ret; +} + +/* * check if a portion of the "newext" extent overlaps with an * existing extent. * @@ -1668,6 +1714,7 @@ int ext4_ext_insert_extent(handle_t *handle, struct inode *inode, int depth, len, err; ext4_lblk_t next; unsigned uninitialized = 0; + int flags = 0; if (unlikely(ext4_ext_get_actual_len(newext) == 0)) { EXT4_ERROR_INODE(inode, "ext4_ext_get_actual_len(newext) == 0"); @@ -1729,7 +1776,7 @@ repeat: BUG_ON(npath->p_depth != path->p_depth); eh = npath[depth].p_hdr; if (le16_to_cpu(eh->eh_entries) < le16_to_cpu(eh->eh_max)) { - ext_debug("next leaf isnt full(%d)\n", + ext_debug("next leaf isn't full(%d)\n", le16_to_cpu(eh->eh_entries)); path = npath; goto repeat; @@ -1742,7 +1789,9 @@ repeat: * There is no free space in the found leaf. * We're gonna add a new leaf in the tree. */ - err = ext4_ext_create_new_leaf(handle, inode, path, newext); + if (flag & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) + flags = EXT4_MB_USE_ROOT_BLOCKS; + err = ext4_ext_create_new_leaf(handle, inode, flags, path, newext); if (err) goto cleanup; depth = ext_depth(inode); @@ -2003,13 +2052,25 @@ ext4_ext_put_gap_in_cache(struct inode *inode, struct ext4_ext_path *path, } /* + * ext4_ext_in_cache() + * Checks to see if the given block is in the cache. + * If it is, the cached extent is stored in the given + * cache extent pointer. If the cached extent is a hole, + * this routine should be used instead of + * ext4_ext_in_cache if the calling function needs to + * know the size of the hole. + * + * @inode: The files inode + * @block: The block to look for in the cache + * @ex: Pointer where the cached extent will be stored + * if it contains block + * * Return 0 if cache is invalid; 1 if the cache is valid */ -static int -ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block, - struct ext4_extent *ex) -{ +static int ext4_ext_check_cache(struct inode *inode, ext4_lblk_t block, + struct ext4_ext_cache *ex){ struct ext4_ext_cache *cex; + struct ext4_sb_info *sbi; int ret = 0; /* @@ -2017,26 +2078,60 @@ ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block, */ spin_lock(&EXT4_I(inode)->i_block_reservation_lock); cex = &EXT4_I(inode)->i_cached_extent; + sbi = EXT4_SB(inode->i_sb); /* has cache valid data? */ if (cex->ec_len == 0) goto errout; if (in_range(block, cex->ec_block, cex->ec_len)) { - ex->ee_block = cpu_to_le32(cex->ec_block); - ext4_ext_store_pblock(ex, cex->ec_start); - ex->ee_len = cpu_to_le16(cex->ec_len); + memcpy(ex, cex, sizeof(struct ext4_ext_cache)); ext_debug("%u cached by %u:%u:%llu\n", block, cex->ec_block, cex->ec_len, cex->ec_start); ret = 1; } errout: + if (!ret) + sbi->extent_cache_misses++; + else + sbi->extent_cache_hits++; spin_unlock(&EXT4_I(inode)->i_block_reservation_lock); return ret; } /* + * ext4_ext_in_cache() + * Checks to see if the given block is in the cache. + * If it is, the cached extent is stored in the given + * extent pointer. + * + * @inode: The files inode + * @block: The block to look for in the cache + * @ex: Pointer where the cached extent will be stored + * if it contains block + * + * Return 0 if cache is invalid; 1 if the cache is valid + */ +static int +ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block, + struct ext4_extent *ex) +{ + struct ext4_ext_cache cex; + int ret = 0; + + if (ext4_ext_check_cache(inode, block, &cex)) { + ex->ee_block = cpu_to_le32(cex.ec_block); + ext4_ext_store_pblock(ex, cex.ec_start); + ex->ee_len = cpu_to_le16(cex.ec_len); + ret = 1; + } + + return ret; +} + + +/* * ext4_ext_rm_idx: * removes index from the index block. * It's used in truncate case only, thus all requests are for @@ -2163,8 +2258,16 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode, ext4_free_blocks(handle, inode, NULL, start, num, flags); } else if (from == le32_to_cpu(ex->ee_block) && to <= le32_to_cpu(ex->ee_block) + ee_len - 1) { - printk(KERN_INFO "strange request: removal %u-%u from %u:%u\n", - from, to, le32_to_cpu(ex->ee_block), ee_len); + /* head removal */ + ext4_lblk_t num; + ext4_fsblk_t start; + + num = to - from; + start = ext4_ext_pblock(ex); + + ext_debug("free first %u blocks starting %llu\n", num, start); + ext4_free_blocks(handle, inode, 0, start, num, flags); + } else { printk(KERN_INFO "strange request: removal(2) " "%u-%u from %u:%u\n", @@ -2173,9 +2276,22 @@ static int ext4_remove_blocks(handle_t *handle, struct inode *inode, return 0; } + +/* + * ext4_ext_rm_leaf() Removes the extents associated with the + * blocks appearing between "start" and "end", and splits the extents + * if "start" and "end" appear in the same extent + * + * @handle: The journal handle + * @inode: The files inode + * @path: The path to the leaf + * @start: The first block to remove + * @end: The last block to remove + */ static int ext4_ext_rm_leaf(handle_t *handle, struct inode *inode, - struct ext4_ext_path *path, ext4_lblk_t start) + struct ext4_ext_path *path, ext4_lblk_t start, + ext4_lblk_t end) { int err = 0, correct_index = 0; int depth = ext_depth(inode), credits; @@ -2186,6 +2302,7 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode, unsigned short ex_ee_len; unsigned uninitialized = 0; struct ext4_extent *ex; + struct ext4_map_blocks map; /* the header must be checked already in ext4_ext_remove_space() */ ext_debug("truncate since %u in leaf\n", start); @@ -2215,31 +2332,95 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode, path[depth].p_ext = ex; a = ex_ee_block > start ? ex_ee_block : start; - b = ex_ee_block + ex_ee_len - 1 < EXT_MAX_BLOCK ? - ex_ee_block + ex_ee_len - 1 : EXT_MAX_BLOCK; + b = ex_ee_block+ex_ee_len - 1 < end ? + ex_ee_block+ex_ee_len - 1 : end; ext_debug(" border %u:%u\n", a, b); - if (a != ex_ee_block && b != ex_ee_block + ex_ee_len - 1) { - block = 0; - num = 0; - BUG(); + /* If this extent is beyond the end of the hole, skip it */ + if (end <= ex_ee_block) { + ex--; + ex_ee_block = le32_to_cpu(ex->ee_block); + ex_ee_len = ext4_ext_get_actual_len(ex); + continue; + } else if (a != ex_ee_block && + b != ex_ee_block + ex_ee_len - 1) { + /* + * If this is a truncate, then this condition should + * never happen because at least one of the end points + * needs to be on the edge of the extent. + */ + if (end == EXT_MAX_BLOCK) { + ext_debug(" bad truncate %u:%u\n", + start, end); + block = 0; + num = 0; + err = -EIO; + goto out; + } + /* + * else this is a hole punch, so the extent needs to + * be split since neither edge of the hole is on the + * extent edge + */ + else{ + map.m_pblk = ext4_ext_pblock(ex); + map.m_lblk = ex_ee_block; + map.m_len = b - ex_ee_block; + + err = ext4_split_extent(handle, + inode, path, &map, 0, + EXT4_GET_BLOCKS_PUNCH_OUT_EXT | + EXT4_GET_BLOCKS_PRE_IO); + + if (err < 0) + goto out; + + ex_ee_len = ext4_ext_get_actual_len(ex); + + b = ex_ee_block+ex_ee_len - 1 < end ? + ex_ee_block+ex_ee_len - 1 : end; + + /* Then remove tail of this extent */ + block = ex_ee_block; + num = a - block; + } } else if (a != ex_ee_block) { /* remove tail of the extent */ block = ex_ee_block; num = a - block; } else if (b != ex_ee_block + ex_ee_len - 1) { /* remove head of the extent */ - block = a; - num = b - a; - /* there is no "make a hole" API yet */ - BUG(); + block = b; + num = ex_ee_block + ex_ee_len - b; + + /* + * If this is a truncate, this condition + * should never happen + */ + if (end == EXT_MAX_BLOCK) { + ext_debug(" bad truncate %u:%u\n", + start, end); + err = -EIO; + goto out; + } } else { /* remove whole extent: excellent! */ block = ex_ee_block; num = 0; - BUG_ON(a != ex_ee_block); - BUG_ON(b != ex_ee_block + ex_ee_len - 1); + if (a != ex_ee_block) { + ext_debug(" bad truncate %u:%u\n", + start, end); + err = -EIO; + goto out; + } + + if (b != ex_ee_block + ex_ee_len - 1) { + ext_debug(" bad truncate %u:%u\n", + start, end); + err = -EIO; + goto out; + } } /* @@ -2270,7 +2451,13 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode, if (num == 0) { /* this extent is removed; mark slot entirely unused */ ext4_ext_store_pblock(ex, 0); - le16_add_cpu(&eh->eh_entries, -1); + } else if (block != ex_ee_block) { + /* + * If this was a head removal, then we need to update + * the physical block since it is now at a different + * location + */ + ext4_ext_store_pblock(ex, ext4_ext_pblock(ex) + (b-a)); } ex->ee_block = cpu_to_le32(block); @@ -2286,6 +2473,27 @@ ext4_ext_rm_leaf(handle_t *handle, struct inode *inode, if (err) goto out; + /* + * If the extent was completely released, + * we need to remove it from the leaf + */ + if (num == 0) { + if (end != EXT_MAX_BLOCK) { + /* + * For hole punching, we need to scoot all the + * extents up when an extent is removed so that + * we dont have blank extents in the middle + */ + memmove(ex, ex+1, (EXT_LAST_EXTENT(eh) - ex) * + sizeof(struct ext4_extent)); + + /* Now get rid of the one at the end */ + memset(EXT_LAST_EXTENT(eh), 0, + sizeof(struct ext4_extent)); + } + le16_add_cpu(&eh->eh_entries, -1); + } + ext_debug("new extent: %u:%u:%llu\n", block, num, ext4_ext_pblock(ex)); ex--; @@ -2326,7 +2534,8 @@ ext4_ext_more_to_rm(struct ext4_ext_path *path) return 1; } -static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start) +static int ext4_ext_remove_space(struct inode *inode, ext4_lblk_t start, + ext4_lblk_t end) { struct super_block *sb = inode->i_sb; int depth = ext_depth(inode); @@ -2365,7 +2574,8 @@ again: while (i >= 0 && err == 0) { if (i == depth) { /* this is leaf block */ - err = ext4_ext_rm_leaf(handle, inode, path, start); + err = ext4_ext_rm_leaf(handle, inode, path, + start, end); /* root level has p_bh == NULL, brelse() eats this */ brelse(path[i].p_bh); path[i].p_bh = NULL; @@ -2529,11 +2739,200 @@ static int ext4_ext_zeroout(struct inode *inode, struct ext4_extent *ex) return ret; } +/* + * used by extent splitting. + */ +#define EXT4_EXT_MAY_ZEROOUT 0x1 /* safe to zeroout if split fails \ + due to ENOSPC */ +#define EXT4_EXT_MARK_UNINIT1 0x2 /* mark first half uninitialized */ +#define EXT4_EXT_MARK_UNINIT2 0x4 /* mark second half uninitialized */ + +/* + * ext4_split_extent_at() splits an extent at given block. + * + * @handle: the journal handle + * @inode: the file inode + * @path: the path to the extent + * @split: the logical block where the extent is splitted. + * @split_flags: indicates if the extent could be zeroout if split fails, and + * the states(init or uninit) of new extents. + * @flags: flags used to insert new extent to extent tree. + * + * + * Splits extent [a, b] into two extents [a, @split) and [@split, b], states + * of which are deterimined by split_flag. + * + * There are two cases: + * a> the extent are splitted into two extent. + * b> split is not needed, and just mark the extent. + * + * return 0 on success. + */ +static int ext4_split_extent_at(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path, + ext4_lblk_t split, + int split_flag, + int flags) +{ + ext4_fsblk_t newblock; + ext4_lblk_t ee_block; + struct ext4_extent *ex, newex, orig_ex; + struct ext4_extent *ex2 = NULL; + unsigned int ee_len, depth; + int err = 0; + + ext_debug("ext4_split_extents_at: inode %lu, logical" + "block %llu\n", inode->i_ino, (unsigned long long)split); + + ext4_ext_show_leaf(inode, path); + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + newblock = split - ee_block + ext4_ext_pblock(ex); + + BUG_ON(split < ee_block || split >= (ee_block + ee_len)); + + err = ext4_ext_get_access(handle, inode, path + depth); + if (err) + goto out; + + if (split == ee_block) { + /* + * case b: block @split is the block that the extent begins with + * then we just change the state of the extent, and splitting + * is not needed. + */ + if (split_flag & EXT4_EXT_MARK_UNINIT2) + ext4_ext_mark_uninitialized(ex); + else + ext4_ext_mark_initialized(ex); + + if (!(flags & EXT4_GET_BLOCKS_PRE_IO)) + ext4_ext_try_to_merge(inode, path, ex); + + err = ext4_ext_dirty(handle, inode, path + depth); + goto out; + } + + /* case a */ + memcpy(&orig_ex, ex, sizeof(orig_ex)); + ex->ee_len = cpu_to_le16(split - ee_block); + if (split_flag & EXT4_EXT_MARK_UNINIT1) + ext4_ext_mark_uninitialized(ex); + + /* + * path may lead to new leaf, not to original leaf any more + * after ext4_ext_insert_extent() returns, + */ + err = ext4_ext_dirty(handle, inode, path + depth); + if (err) + goto fix_extent_len; + + ex2 = &newex; + ex2->ee_block = cpu_to_le32(split); + ex2->ee_len = cpu_to_le16(ee_len - (split - ee_block)); + ext4_ext_store_pblock(ex2, newblock); + if (split_flag & EXT4_EXT_MARK_UNINIT2) + ext4_ext_mark_uninitialized(ex2); + + err = ext4_ext_insert_extent(handle, inode, path, &newex, flags); + if (err == -ENOSPC && (EXT4_EXT_MAY_ZEROOUT & split_flag)) { + err = ext4_ext_zeroout(inode, &orig_ex); + if (err) + goto fix_extent_len; + /* update the extent length and mark as initialized */ + ex->ee_len = cpu_to_le32(ee_len); + ext4_ext_try_to_merge(inode, path, ex); + err = ext4_ext_dirty(handle, inode, path + depth); + goto out; + } else if (err) + goto fix_extent_len; + +out: + ext4_ext_show_leaf(inode, path); + return err; + +fix_extent_len: + ex->ee_len = orig_ex.ee_len; + ext4_ext_dirty(handle, inode, path + depth); + return err; +} + +/* + * ext4_split_extents() splits an extent and mark extent which is covered + * by @map as split_flags indicates + * + * It may result in splitting the extent into multiple extents (upto three) + * There are three possibilities: + * a> There is no split required + * b> Splits in two extents: Split is happening at either end of the extent + * c> Splits in three extents: Somone is splitting in middle of the extent + * + */ +static int ext4_split_extent(handle_t *handle, + struct inode *inode, + struct ext4_ext_path *path, + struct ext4_map_blocks *map, + int split_flag, + int flags) +{ + ext4_lblk_t ee_block; + struct ext4_extent *ex; + unsigned int ee_len, depth; + int err = 0; + int uninitialized; + int split_flag1, flags1; + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); + uninitialized = ext4_ext_is_uninitialized(ex); + + if (map->m_lblk + map->m_len < ee_block + ee_len) { + split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ? + EXT4_EXT_MAY_ZEROOUT : 0; + flags1 = flags | EXT4_GET_BLOCKS_PRE_IO; + if (uninitialized) + split_flag1 |= EXT4_EXT_MARK_UNINIT1 | + EXT4_EXT_MARK_UNINIT2; + err = ext4_split_extent_at(handle, inode, path, + map->m_lblk + map->m_len, split_flag1, flags1); + if (err) + goto out; + } + + ext4_ext_drop_refs(path); + path = ext4_ext_find_extent(inode, map->m_lblk, path); + if (IS_ERR(path)) + return PTR_ERR(path); + + if (map->m_lblk >= ee_block) { + split_flag1 = split_flag & EXT4_EXT_MAY_ZEROOUT ? + EXT4_EXT_MAY_ZEROOUT : 0; + if (uninitialized) + split_flag1 |= EXT4_EXT_MARK_UNINIT1; + if (split_flag & EXT4_EXT_MARK_UNINIT2) + split_flag1 |= EXT4_EXT_MARK_UNINIT2; + err = ext4_split_extent_at(handle, inode, path, + map->m_lblk, split_flag1, flags); + if (err) + goto out; + } + + ext4_ext_show_leaf(inode, path); +out: + return err ? err : map->m_len; +} + #define EXT4_EXT_ZERO_LEN 7 /* * This function is called by ext4_ext_map_blocks() if someone tries to write * to an uninitialized extent. It may result in splitting the uninitialized - * extent into multiple extents (upto three - one initialized and two + * extent into multiple extents (up to three - one initialized and two * uninitialized). * There are three possibilities: * a> There is no split required: Entire extent should be initialized @@ -2545,17 +2944,13 @@ static int ext4_ext_convert_to_initialized(handle_t *handle, struct ext4_map_blocks *map, struct ext4_ext_path *path) { - struct ext4_extent *ex, newex, orig_ex; - struct ext4_extent *ex1 = NULL; - struct ext4_extent *ex2 = NULL; - struct ext4_extent *ex3 = NULL; - struct ext4_extent_header *eh; + struct ext4_map_blocks split_map; + struct ext4_extent zero_ex; + struct ext4_extent *ex; ext4_lblk_t ee_block, eof_block; unsigned int allocated, ee_len, depth; - ext4_fsblk_t newblock; int err = 0; - int ret = 0; - int may_zeroout; + int split_flag = 0; ext_debug("ext4_ext_convert_to_initialized: inode %lu, logical" "block %llu, max_blocks %u\n", inode->i_ino, @@ -2567,280 +2962,86 @@ static int ext4_ext_convert_to_initialized(handle_t *handle, eof_block = map->m_lblk + map->m_len; depth = ext_depth(inode); - eh = path[depth].p_hdr; ex = path[depth].p_ext; ee_block = le32_to_cpu(ex->ee_block); ee_len = ext4_ext_get_actual_len(ex); allocated = ee_len - (map->m_lblk - ee_block); - newblock = map->m_lblk - ee_block + ext4_ext_pblock(ex); - - ex2 = ex; - orig_ex.ee_block = ex->ee_block; - orig_ex.ee_len = cpu_to_le16(ee_len); - ext4_ext_store_pblock(&orig_ex, ext4_ext_pblock(ex)); + WARN_ON(map->m_lblk < ee_block); /* * It is safe to convert extent to initialized via explicit * zeroout only if extent is fully insde i_size or new_size. */ - may_zeroout = ee_block + ee_len <= eof_block; + split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0; - err = ext4_ext_get_access(handle, inode, path + depth); - if (err) - goto out; /* If extent has less than 2*EXT4_EXT_ZERO_LEN zerout directly */ - if (ee_len <= 2*EXT4_EXT_ZERO_LEN && may_zeroout) { - err = ext4_ext_zeroout(inode, &orig_ex); + if (ee_len <= 2*EXT4_EXT_ZERO_LEN && + (EXT4_EXT_MAY_ZEROOUT & split_flag)) { + err = ext4_ext_zeroout(inode, ex); if (err) - goto fix_extent_len; - /* update the extent length and mark as initialized */ - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - /* zeroed the full extent */ - return allocated; - } - - /* ex1: ee_block to map->m_lblk - 1 : uninitialized */ - if (map->m_lblk > ee_block) { - ex1 = ex; - ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block); - ext4_ext_mark_uninitialized(ex1); - ex2 = &newex; - } - /* - * for sanity, update the length of the ex2 extent before - * we insert ex3, if ex1 is NULL. This is to avoid temporary - * overlap of blocks. - */ - if (!ex1 && allocated > map->m_len) - ex2->ee_len = cpu_to_le16(map->m_len); - /* ex3: to ee_block + ee_len : uninitialised */ - if (allocated > map->m_len) { - unsigned int newdepth; - /* If extent has less than EXT4_EXT_ZERO_LEN zerout directly */ - if (allocated <= EXT4_EXT_ZERO_LEN && may_zeroout) { - /* - * map->m_lblk == ee_block is handled by the zerouout - * at the beginning. - * Mark first half uninitialized. - * Mark second half initialized and zero out the - * initialized extent - */ - ex->ee_block = orig_ex.ee_block; - ex->ee_len = cpu_to_le16(ee_len - allocated); - ext4_ext_mark_uninitialized(ex); - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - - ex3 = &newex; - ex3->ee_block = cpu_to_le32(map->m_lblk); - ext4_ext_store_pblock(ex3, newblock); - ex3->ee_len = cpu_to_le16(allocated); - err = ext4_ext_insert_extent(handle, inode, path, - ex3, 0); - if (err == -ENOSPC) { - err = ext4_ext_zeroout(inode, &orig_ex); - if (err) - goto fix_extent_len; - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, - ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - /* blocks available from map->m_lblk */ - return allocated; - - } else if (err) - goto fix_extent_len; - - /* - * We need to zero out the second half because - * an fallocate request can update file size and - * converting the second half to initialized extent - * implies that we can leak some junk data to user - * space. - */ - err = ext4_ext_zeroout(inode, ex3); - if (err) { - /* - * We should actually mark the - * second half as uninit and return error - * Insert would have changed the extent - */ - depth = ext_depth(inode); - ext4_ext_drop_refs(path); - path = ext4_ext_find_extent(inode, map->m_lblk, - path); - if (IS_ERR(path)) { - err = PTR_ERR(path); - return err; - } - /* get the second half extent details */ - ex = path[depth].p_ext; - err = ext4_ext_get_access(handle, inode, - path + depth); - if (err) - return err; - ext4_ext_mark_uninitialized(ex); - ext4_ext_dirty(handle, inode, path + depth); - return err; - } - - /* zeroed the second half */ - return allocated; - } - ex3 = &newex; - ex3->ee_block = cpu_to_le32(map->m_lblk + map->m_len); - ext4_ext_store_pblock(ex3, newblock + map->m_len); - ex3->ee_len = cpu_to_le16(allocated - map->m_len); - ext4_ext_mark_uninitialized(ex3); - err = ext4_ext_insert_extent(handle, inode, path, ex3, 0); - if (err == -ENOSPC && may_zeroout) { - err = ext4_ext_zeroout(inode, &orig_ex); - if (err) - goto fix_extent_len; - /* update the extent length and mark as initialized */ - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - /* zeroed the full extent */ - /* blocks available from map->m_lblk */ - return allocated; - - } else if (err) - goto fix_extent_len; - /* - * The depth, and hence eh & ex might change - * as part of the insert above. - */ - newdepth = ext_depth(inode); - /* - * update the extent length after successful insert of the - * split extent - */ - ee_len -= ext4_ext_get_actual_len(ex3); - orig_ex.ee_len = cpu_to_le16(ee_len); - may_zeroout = ee_block + ee_len <= eof_block; - - depth = newdepth; - ext4_ext_drop_refs(path); - path = ext4_ext_find_extent(inode, map->m_lblk, path); - if (IS_ERR(path)) { - err = PTR_ERR(path); goto out; - } - eh = path[depth].p_hdr; - ex = path[depth].p_ext; - if (ex2 != &newex) - ex2 = ex; err = ext4_ext_get_access(handle, inode, path + depth); if (err) goto out; - - allocated = map->m_len; - - /* If extent has less than EXT4_EXT_ZERO_LEN and we are trying - * to insert a extent in the middle zerout directly - * otherwise give the extent a chance to merge to left - */ - if (le16_to_cpu(orig_ex.ee_len) <= EXT4_EXT_ZERO_LEN && - map->m_lblk != ee_block && may_zeroout) { - err = ext4_ext_zeroout(inode, &orig_ex); - if (err) - goto fix_extent_len; - /* update the extent length and mark as initialized */ - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - /* zero out the first half */ - /* blocks available from map->m_lblk */ - return allocated; - } - } - /* - * If there was a change of depth as part of the - * insertion of ex3 above, we need to update the length - * of the ex1 extent again here - */ - if (ex1 && ex1 != ex) { - ex1 = ex; - ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block); - ext4_ext_mark_uninitialized(ex1); - ex2 = &newex; - } - /* ex2: map->m_lblk to map->m_lblk + maxblocks-1 : initialised */ - ex2->ee_block = cpu_to_le32(map->m_lblk); - ext4_ext_store_pblock(ex2, newblock); - ex2->ee_len = cpu_to_le16(allocated); - if (ex2 != ex) - goto insert; - /* - * New (initialized) extent starts from the first block - * in the current extent. i.e., ex2 == ex - * We have to see if it can be merged with the extent - * on the left. - */ - if (ex2 > EXT_FIRST_EXTENT(eh)) { - /* - * To merge left, pass "ex2 - 1" to try_to_merge(), - * since it merges towards right _only_. - */ - ret = ext4_ext_try_to_merge(inode, path, ex2 - 1); - if (ret) { - err = ext4_ext_correct_indexes(handle, inode, path); - if (err) - goto out; - depth = ext_depth(inode); - ex2--; - } + ext4_ext_mark_initialized(ex); + ext4_ext_try_to_merge(inode, path, ex); + err = ext4_ext_dirty(handle, inode, path + depth); + goto out; } + /* - * Try to Merge towards right. This might be required - * only when the whole extent is being written to. - * i.e. ex2 == ex and ex3 == NULL. + * four cases: + * 1. split the extent into three extents. + * 2. split the extent into two extents, zeroout the first half. + * 3. split the extent into two extents, zeroout the second half. + * 4. split the extent into two extents with out zeroout. */ - if (!ex3) { - ret = ext4_ext_try_to_merge(inode, path, ex2); - if (ret) { - err = ext4_ext_correct_indexes(handle, inode, path); + split_map.m_lblk = map->m_lblk; + split_map.m_len = map->m_len; + + if (allocated > map->m_len) { + if (allocated <= EXT4_EXT_ZERO_LEN && + (EXT4_EXT_MAY_ZEROOUT & split_flag)) { + /* case 3 */ + zero_ex.ee_block = + cpu_to_le32(map->m_lblk); + zero_ex.ee_len = cpu_to_le16(allocated); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(ex) + map->m_lblk - ee_block); + err = ext4_ext_zeroout(inode, &zero_ex); if (err) goto out; + split_map.m_lblk = map->m_lblk; + split_map.m_len = allocated; + } else if ((map->m_lblk - ee_block + map->m_len < + EXT4_EXT_ZERO_LEN) && + (EXT4_EXT_MAY_ZEROOUT & split_flag)) { + /* case 2 */ + if (map->m_lblk != ee_block) { + zero_ex.ee_block = ex->ee_block; + zero_ex.ee_len = cpu_to_le16(map->m_lblk - + ee_block); + ext4_ext_store_pblock(&zero_ex, + ext4_ext_pblock(ex)); + err = ext4_ext_zeroout(inode, &zero_ex); + if (err) + goto out; + } + + split_map.m_lblk = ee_block; + split_map.m_len = map->m_lblk - ee_block + map->m_len; + allocated = map->m_len; } } - /* Mark modified extent as dirty */ - err = ext4_ext_dirty(handle, inode, path + depth); - goto out; -insert: - err = ext4_ext_insert_extent(handle, inode, path, &newex, 0); - if (err == -ENOSPC && may_zeroout) { - err = ext4_ext_zeroout(inode, &orig_ex); - if (err) - goto fix_extent_len; - /* update the extent length and mark as initialized */ - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - /* zero out the first half */ - return allocated; - } else if (err) - goto fix_extent_len; + + allocated = ext4_split_extent(handle, inode, path, + &split_map, split_flag, 0); + if (allocated < 0) + err = allocated; + out: - ext4_ext_show_leaf(inode, path); return err ? err : allocated; - -fix_extent_len: - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_mark_uninitialized(ex); - ext4_ext_dirty(handle, inode, path + depth); - return err; } /* @@ -2871,15 +3072,11 @@ static int ext4_split_unwritten_extents(handle_t *handle, struct ext4_ext_path *path, int flags) { - struct ext4_extent *ex, newex, orig_ex; - struct ext4_extent *ex1 = NULL; - struct ext4_extent *ex2 = NULL; - struct ext4_extent *ex3 = NULL; - ext4_lblk_t ee_block, eof_block; - unsigned int allocated, ee_len, depth; - ext4_fsblk_t newblock; - int err = 0; - int may_zeroout; + ext4_lblk_t eof_block; + ext4_lblk_t ee_block; + struct ext4_extent *ex; + unsigned int ee_len; + int split_flag = 0, depth; ext_debug("ext4_split_unwritten_extents: inode %lu, logical" "block %llu, max_blocks %u\n", inode->i_ino, @@ -2889,156 +3086,22 @@ static int ext4_split_unwritten_extents(handle_t *handle, inode->i_sb->s_blocksize_bits; if (eof_block < map->m_lblk + map->m_len) eof_block = map->m_lblk + map->m_len; - - depth = ext_depth(inode); - ex = path[depth].p_ext; - ee_block = le32_to_cpu(ex->ee_block); - ee_len = ext4_ext_get_actual_len(ex); - allocated = ee_len - (map->m_lblk - ee_block); - newblock = map->m_lblk - ee_block + ext4_ext_pblock(ex); - - ex2 = ex; - orig_ex.ee_block = ex->ee_block; - orig_ex.ee_len = cpu_to_le16(ee_len); - ext4_ext_store_pblock(&orig_ex, ext4_ext_pblock(ex)); - /* * It is safe to convert extent to initialized via explicit * zeroout only if extent is fully insde i_size or new_size. */ - may_zeroout = ee_block + ee_len <= eof_block; - - /* - * If the uninitialized extent begins at the same logical - * block where the write begins, and the write completely - * covers the extent, then we don't need to split it. - */ - if ((map->m_lblk == ee_block) && (allocated <= map->m_len)) - return allocated; - - err = ext4_ext_get_access(handle, inode, path + depth); - if (err) - goto out; - /* ex1: ee_block to map->m_lblk - 1 : uninitialized */ - if (map->m_lblk > ee_block) { - ex1 = ex; - ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block); - ext4_ext_mark_uninitialized(ex1); - ex2 = &newex; - } - /* - * for sanity, update the length of the ex2 extent before - * we insert ex3, if ex1 is NULL. This is to avoid temporary - * overlap of blocks. - */ - if (!ex1 && allocated > map->m_len) - ex2->ee_len = cpu_to_le16(map->m_len); - /* ex3: to ee_block + ee_len : uninitialised */ - if (allocated > map->m_len) { - unsigned int newdepth; - ex3 = &newex; - ex3->ee_block = cpu_to_le32(map->m_lblk + map->m_len); - ext4_ext_store_pblock(ex3, newblock + map->m_len); - ex3->ee_len = cpu_to_le16(allocated - map->m_len); - ext4_ext_mark_uninitialized(ex3); - err = ext4_ext_insert_extent(handle, inode, path, ex3, flags); - if (err == -ENOSPC && may_zeroout) { - err = ext4_ext_zeroout(inode, &orig_ex); - if (err) - goto fix_extent_len; - /* update the extent length and mark as initialized */ - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - /* zeroed the full extent */ - /* blocks available from map->m_lblk */ - return allocated; - - } else if (err) - goto fix_extent_len; - /* - * The depth, and hence eh & ex might change - * as part of the insert above. - */ - newdepth = ext_depth(inode); - /* - * update the extent length after successful insert of the - * split extent - */ - ee_len -= ext4_ext_get_actual_len(ex3); - orig_ex.ee_len = cpu_to_le16(ee_len); - may_zeroout = ee_block + ee_len <= eof_block; - - depth = newdepth; - ext4_ext_drop_refs(path); - path = ext4_ext_find_extent(inode, map->m_lblk, path); - if (IS_ERR(path)) { - err = PTR_ERR(path); - goto out; - } - ex = path[depth].p_ext; - if (ex2 != &newex) - ex2 = ex; + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_block = le32_to_cpu(ex->ee_block); + ee_len = ext4_ext_get_actual_len(ex); - err = ext4_ext_get_access(handle, inode, path + depth); - if (err) - goto out; + split_flag |= ee_block + ee_len <= eof_block ? EXT4_EXT_MAY_ZEROOUT : 0; + split_flag |= EXT4_EXT_MARK_UNINIT2; - allocated = map->m_len; - } - /* - * If there was a change of depth as part of the - * insertion of ex3 above, we need to update the length - * of the ex1 extent again here - */ - if (ex1 && ex1 != ex) { - ex1 = ex; - ex1->ee_len = cpu_to_le16(map->m_lblk - ee_block); - ext4_ext_mark_uninitialized(ex1); - ex2 = &newex; - } - /* - * ex2: map->m_lblk to map->m_lblk + map->m_len-1 : to be written - * using direct I/O, uninitialised still. - */ - ex2->ee_block = cpu_to_le32(map->m_lblk); - ext4_ext_store_pblock(ex2, newblock); - ex2->ee_len = cpu_to_le16(allocated); - ext4_ext_mark_uninitialized(ex2); - if (ex2 != ex) - goto insert; - /* Mark modified extent as dirty */ - err = ext4_ext_dirty(handle, inode, path + depth); - ext_debug("out here\n"); - goto out; -insert: - err = ext4_ext_insert_extent(handle, inode, path, &newex, flags); - if (err == -ENOSPC && may_zeroout) { - err = ext4_ext_zeroout(inode, &orig_ex); - if (err) - goto fix_extent_len; - /* update the extent length and mark as initialized */ - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_dirty(handle, inode, path + depth); - /* zero out the first half */ - return allocated; - } else if (err) - goto fix_extent_len; -out: - ext4_ext_show_leaf(inode, path); - return err ? err : allocated; - -fix_extent_len: - ex->ee_block = orig_ex.ee_block; - ex->ee_len = orig_ex.ee_len; - ext4_ext_store_pblock(ex, ext4_ext_pblock(&orig_ex)); - ext4_ext_mark_uninitialized(ex); - ext4_ext_dirty(handle, inode, path + depth); - return err; + flags |= EXT4_GET_BLOCKS_PRE_IO; + return ext4_split_extent(handle, inode, path, map, split_flag, flags); } + static int ext4_convert_unwritten_extents_endio(handle_t *handle, struct inode *inode, struct ext4_ext_path *path) @@ -3047,46 +3110,27 @@ static int ext4_convert_unwritten_extents_endio(handle_t *handle, struct ext4_extent_header *eh; int depth; int err = 0; - int ret = 0; depth = ext_depth(inode); eh = path[depth].p_hdr; ex = path[depth].p_ext; + ext_debug("ext4_convert_unwritten_extents_endio: inode %lu, logical" + "block %llu, max_blocks %u\n", inode->i_ino, + (unsigned long long)le32_to_cpu(ex->ee_block), + ext4_ext_get_actual_len(ex)); + err = ext4_ext_get_access(handle, inode, path + depth); if (err) goto out; /* first mark the extent as initialized */ ext4_ext_mark_initialized(ex); - /* - * We have to see if it can be merged with the extent - * on the left. - */ - if (ex > EXT_FIRST_EXTENT(eh)) { - /* - * To merge left, pass "ex - 1" to try_to_merge(), - * since it merges towards right _only_. - */ - ret = ext4_ext_try_to_merge(inode, path, ex - 1); - if (ret) { - err = ext4_ext_correct_indexes(handle, inode, path); - if (err) - goto out; - depth = ext_depth(inode); - ex--; - } - } - /* - * Try to Merge towards right. + /* note: ext4_ext_correct_indexes() isn't needed here because + * borders are not changed */ - ret = ext4_ext_try_to_merge(inode, path, ex); - if (ret) { - err = ext4_ext_correct_indexes(handle, inode, path); - if (err) - goto out; - depth = ext_depth(inode); - } + ext4_ext_try_to_merge(inode, path, ex); + /* Mark modified extent as dirty */ err = ext4_ext_dirty(handle, inode, path + depth); out: @@ -3174,7 +3218,7 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, path, flags); /* * Flag the inode(non aio case) or end_io struct (aio case) - * that this IO needs to convertion to written when IO is + * that this IO needs to conversion to written when IO is * completed */ if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) { @@ -3302,15 +3346,19 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ext4_fsblk_t newblock = 0; int err = 0, depth, ret; unsigned int allocated = 0; + unsigned int punched_out = 0; + unsigned int result = 0; struct ext4_allocation_request ar; ext4_io_end_t *io = EXT4_I(inode)->cur_aio_dio; + struct ext4_map_blocks punch_map; ext_debug("blocks %u/%u requested for inode %lu\n", map->m_lblk, map->m_len, inode->i_ino); trace_ext4_ext_map_blocks_enter(inode, map->m_lblk, map->m_len, flags); /* check in cache */ - if (ext4_ext_in_cache(inode, map->m_lblk, &newex)) { + if (ext4_ext_in_cache(inode, map->m_lblk, &newex) && + ((flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) == 0)) { if (!newex.ee_start_lo && !newex.ee_start_hi) { if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { /* @@ -3375,16 +3423,84 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ext_debug("%u fit into %u:%d -> %llu\n", map->m_lblk, ee_block, ee_len, newblock); - /* Do not put uninitialized extent in the cache */ - if (!ext4_ext_is_uninitialized(ex)) { - ext4_ext_put_in_cache(inode, ee_block, - ee_len, ee_start); - goto out; + if ((flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) == 0) { + /* + * Do not put uninitialized extent + * in the cache + */ + if (!ext4_ext_is_uninitialized(ex)) { + ext4_ext_put_in_cache(inode, ee_block, + ee_len, ee_start); + goto out; + } + ret = ext4_ext_handle_uninitialized_extents( + handle, inode, map, path, flags, + allocated, newblock); + return ret; } - ret = ext4_ext_handle_uninitialized_extents(handle, - inode, map, path, flags, allocated, - newblock); - return ret; + + /* + * Punch out the map length, but only to the + * end of the extent + */ + punched_out = allocated < map->m_len ? + allocated : map->m_len; + + /* + * Sense extents need to be converted to + * uninitialized, they must fit in an + * uninitialized extent + */ + if (punched_out > EXT_UNINIT_MAX_LEN) + punched_out = EXT_UNINIT_MAX_LEN; + + punch_map.m_lblk = map->m_lblk; + punch_map.m_pblk = newblock; + punch_map.m_len = punched_out; + punch_map.m_flags = 0; + + /* Check to see if the extent needs to be split */ + if (punch_map.m_len != ee_len || + punch_map.m_lblk != ee_block) { + + ret = ext4_split_extent(handle, inode, + path, &punch_map, 0, + EXT4_GET_BLOCKS_PUNCH_OUT_EXT | + EXT4_GET_BLOCKS_PRE_IO); + + if (ret < 0) { + err = ret; + goto out2; + } + /* + * find extent for the block at + * the start of the hole + */ + ext4_ext_drop_refs(path); + kfree(path); + + path = ext4_ext_find_extent(inode, + map->m_lblk, NULL); + if (IS_ERR(path)) { + err = PTR_ERR(path); + path = NULL; + goto out2; + } + + depth = ext_depth(inode); + ex = path[depth].p_ext; + ee_len = ext4_ext_get_actual_len(ex); + ee_block = le32_to_cpu(ex->ee_block); + ee_start = ext4_ext_pblock(ex); + + } + + ext4_ext_mark_uninitialized(ex); + + err = ext4_ext_remove_space(inode, map->m_lblk, + map->m_lblk + punched_out); + + goto out2; } } @@ -3446,6 +3562,8 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, else /* disable in-core preallocation for non-regular files */ ar.flags = 0; + if (flags & EXT4_GET_BLOCKS_NO_NORMALIZE) + ar.flags |= EXT4_MB_HINT_NOPREALLOC; newblock = ext4_mb_new_blocks(handle, &ar, &err); if (!newblock) goto out2; @@ -3460,10 +3578,10 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ext4_ext_mark_uninitialized(&newex); /* * io_end structure was created for every IO write to an - * uninitialized extent. To avoid unecessary conversion, + * uninitialized extent. To avoid unnecessary conversion, * here we flag the IO that really needs the conversion. * For non asycn direct IO case, flag the inode state - * that we need to perform convertion when IO is done. + * that we need to perform conversion when IO is done. */ if ((flags & EXT4_GET_BLOCKS_PRE_IO)) { if (io && !(io->flag & EXT4_IO_END_UNWRITTEN)) { @@ -3529,7 +3647,11 @@ out2: } trace_ext4_ext_map_blocks_exit(inode, map->m_lblk, newblock, map->m_len, err ? err : allocated); - return err ? err : allocated; + + result = (flags & EXT4_GET_BLOCKS_PUNCH_OUT_EXT) ? + punched_out : allocated; + + return err ? err : result; } void ext4_ext_truncate(struct inode *inode) @@ -3577,7 +3699,7 @@ void ext4_ext_truncate(struct inode *inode) last_block = (inode->i_size + sb->s_blocksize - 1) >> EXT4_BLOCK_SIZE_BITS(sb); - err = ext4_ext_remove_space(inode, last_block); + err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCK); /* In a multi-transaction truncate, we only make the final * transaction synchronous. @@ -3585,8 +3707,9 @@ void ext4_ext_truncate(struct inode *inode) if (IS_SYNC(inode)) ext4_handle_sync(handle); -out_stop: up_write(&EXT4_I(inode)->i_data_sem); + +out_stop: /* * If this was a simple ftruncate() and the file will remain alive, * then we need to clear up the orphan record which we created above. @@ -3651,10 +3774,6 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) struct ext4_map_blocks map; unsigned int credits, blkbits = inode->i_blkbits; - /* We only support the FALLOC_FL_KEEP_SIZE mode */ - if (mode & ~FALLOC_FL_KEEP_SIZE) - return -EOPNOTSUPP; - /* * currently supporting (pre)allocate mode for extent-based * files _only_ @@ -3662,6 +3781,13 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) return -EOPNOTSUPP; + /* Return error if mode is not supported */ + if (mode & ~(FALLOC_FL_KEEP_SIZE | FALLOC_FL_PUNCH_HOLE)) + return -EOPNOTSUPP; + + if (mode & FALLOC_FL_PUNCH_HOLE) + return ext4_punch_hole(file, offset, len); + trace_ext4_fallocate_enter(inode, offset, len, mode); map.m_lblk = offset >> blkbits; /* @@ -3691,7 +3817,8 @@ retry: break; } ret = ext4_map_blocks(handle, inode, &map, - EXT4_GET_BLOCKS_CREATE_UNINIT_EXT); + EXT4_GET_BLOCKS_CREATE_UNINIT_EXT | + EXT4_GET_BLOCKS_NO_NORMALIZE); if (ret <= 0) { #ifdef EXT4FS_DEBUG WARN_ON(ret <= 0); @@ -3822,6 +3949,7 @@ static int ext4_ext_fiemap_cb(struct inode *inode, struct ext4_ext_path *path, pgoff_t last_offset; pgoff_t offset; pgoff_t index; + pgoff_t start_index = 0; struct page **pages = NULL; struct buffer_head *bh = NULL; struct buffer_head *head = NULL; @@ -3848,39 +3976,57 @@ out: kfree(pages); return EXT_CONTINUE; } + index = 0; +next_page: /* Try to find the 1st mapped buffer. */ - end = ((__u64)pages[0]->index << PAGE_SHIFT) >> + end = ((__u64)pages[index]->index << PAGE_SHIFT) >> blksize_bits; - if (!page_has_buffers(pages[0])) + if (!page_has_buffers(pages[index])) goto out; - head = page_buffers(pages[0]); + head = page_buffers(pages[index]); if (!head) goto out; + index++; bh = head; do { - if (buffer_mapped(bh)) { + if (end >= newex->ec_block + + newex->ec_len) + /* The buffer is out of + * the request range. + */ + goto out; + + if (buffer_mapped(bh) && + end >= newex->ec_block) { + start_index = index - 1; /* get the 1st mapped buffer. */ - if (end > newex->ec_block + - newex->ec_len) - /* The buffer is out of - * the request range. - */ - goto out; goto found_mapped_buffer; } + bh = bh->b_this_page; end++; } while (bh != head); - /* No mapped buffer found. */ - goto out; + /* No mapped buffer in the range found in this page, + * We need to look up next page. + */ + if (index >= ret) { + /* There is no page left, but we need to limit + * newex->ec_len. + */ + newex->ec_len = end - newex->ec_block; + goto out; + } + goto next_page; } else { /*Find contiguous delayed buffers. */ if (ret > 0 && pages[0]->index == last_offset) head = page_buffers(pages[0]); bh = head; + index = 1; + start_index = 0; } found_mapped_buffer: @@ -3903,7 +4049,7 @@ found_mapped_buffer: end++; } while (bh != head); - for (index = 1; index < ret; index++) { + for (; index < ret; index++) { if (!page_has_buffers(pages[index])) { bh = NULL; break; @@ -3913,8 +4059,10 @@ found_mapped_buffer: bh = NULL; break; } + if (pages[index]->index != - pages[0]->index + index) { + pages[start_index]->index + index + - start_index) { /* Blocks are not contiguous. */ bh = NULL; break; @@ -4006,6 +4154,177 @@ static int ext4_xattr_fiemap(struct inode *inode, return (error < 0 ? error : 0); } +/* + * ext4_ext_punch_hole + * + * Punches a hole of "length" bytes in a file starting + * at byte "offset" + * + * @inode: The inode of the file to punch a hole in + * @offset: The starting byte offset of the hole + * @length: The length of the hole + * + * Returns the number of blocks removed or negative on err + */ +int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length) +{ + struct inode *inode = file->f_path.dentry->d_inode; + struct super_block *sb = inode->i_sb; + struct ext4_ext_cache cache_ex; + ext4_lblk_t first_block, last_block, num_blocks, iblock, max_blocks; + struct address_space *mapping = inode->i_mapping; + struct ext4_map_blocks map; + handle_t *handle; + loff_t first_block_offset, last_block_offset, block_len; + loff_t first_page, last_page, first_page_offset, last_page_offset; + int ret, credits, blocks_released, err = 0; + + first_block = (offset + sb->s_blocksize - 1) >> + EXT4_BLOCK_SIZE_BITS(sb); + last_block = (offset + length) >> EXT4_BLOCK_SIZE_BITS(sb); + + first_block_offset = first_block << EXT4_BLOCK_SIZE_BITS(sb); + last_block_offset = last_block << EXT4_BLOCK_SIZE_BITS(sb); + + first_page = (offset + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; + last_page = (offset + length) >> PAGE_CACHE_SHIFT; + + first_page_offset = first_page << PAGE_CACHE_SHIFT; + last_page_offset = last_page << PAGE_CACHE_SHIFT; + + /* + * Write out all dirty pages to avoid race conditions + * Then release them. + */ + if (mapping->nrpages && mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) { + err = filemap_write_and_wait_range(mapping, + first_page_offset == 0 ? 0 : first_page_offset-1, + last_page_offset); + + if (err) + return err; + } + + /* Now release the pages */ + if (last_page_offset > first_page_offset) { + truncate_inode_pages_range(mapping, first_page_offset, + last_page_offset-1); + } + + /* finish any pending end_io work */ + ext4_flush_completed_IO(inode); + + credits = ext4_writepage_trans_blocks(inode); + handle = ext4_journal_start(inode, credits); + if (IS_ERR(handle)) + return PTR_ERR(handle); + + err = ext4_orphan_add(handle, inode); + if (err) + goto out; + + /* + * Now we need to zero out the un block aligned data. + * If the file is smaller than a block, just + * zero out the middle + */ + if (first_block > last_block) + ext4_block_zero_page_range(handle, mapping, offset, length); + else { + /* zero out the head of the hole before the first block */ + block_len = first_block_offset - offset; + if (block_len > 0) + ext4_block_zero_page_range(handle, mapping, + offset, block_len); + + /* zero out the tail of the hole after the last block */ + block_len = offset + length - last_block_offset; + if (block_len > 0) { + ext4_block_zero_page_range(handle, mapping, + last_block_offset, block_len); + } + } + + /* If there are no blocks to remove, return now */ + if (first_block >= last_block) + goto out; + + down_write(&EXT4_I(inode)->i_data_sem); + ext4_ext_invalidate_cache(inode); + ext4_discard_preallocations(inode); + + /* + * Loop over all the blocks and identify blocks + * that need to be punched out + */ + iblock = first_block; + blocks_released = 0; + while (iblock < last_block) { + max_blocks = last_block - iblock; + num_blocks = 1; + memset(&map, 0, sizeof(map)); + map.m_lblk = iblock; + map.m_len = max_blocks; + ret = ext4_ext_map_blocks(handle, inode, &map, + EXT4_GET_BLOCKS_PUNCH_OUT_EXT); + + if (ret > 0) { + blocks_released += ret; + num_blocks = ret; + } else if (ret == 0) { + /* + * If map blocks could not find the block, + * then it is in a hole. If the hole was + * not already cached, then map blocks should + * put it in the cache. So we can get the hole + * out of the cache + */ + memset(&cache_ex, 0, sizeof(cache_ex)); + if ((ext4_ext_check_cache(inode, iblock, &cache_ex)) && + !cache_ex.ec_start) { + + /* The hole is cached */ + num_blocks = cache_ex.ec_block + + cache_ex.ec_len - iblock; + + } else { + /* The block could not be identified */ + err = -EIO; + break; + } + } else { + /* Map blocks error */ + err = ret; + break; + } + + if (num_blocks == 0) { + /* This condition should never happen */ + ext_debug("Block lookup failed"); + err = -EIO; + break; + } + + iblock += num_blocks; + } + + if (blocks_released > 0) { + ext4_ext_invalidate_cache(inode); + ext4_discard_preallocations(inode); + } + + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + + up_write(&EXT4_I(inode)->i_data_sem); + +out: + ext4_orphan_del(handle, inode); + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + return err; +} int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len) { @@ -4042,4 +4361,3 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, return error; } - diff --git a/fs/ext4/file.c b/fs/ext4/file.c index 7b80d543b89e..2c0972322009 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -272,7 +272,6 @@ const struct file_operations ext4_file_operations = { }; const struct inode_operations ext4_file_inode_operations = { - .truncate = ext4_truncate, .setattr = ext4_setattr, .getattr = ext4_getattr, #ifdef CONFIG_EXT4_FS_XATTR diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index 7f74019d6d77..ce66d2fe826c 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -36,7 +36,7 @@ static void dump_completed_IO(struct inode * inode) { -#ifdef EXT4_DEBUG +#ifdef EXT4FS_DEBUG struct list_head *cur, *before, *after; ext4_io_end_t *io, *io0, *io1; unsigned long flags; @@ -101,7 +101,7 @@ extern int ext4_flush_completed_IO(struct inode *inode) * to the work-to-be schedule is freed. * * Thus we need to keep the io structure still valid here after - * convertion finished. The io structure has a flag to + * conversion finished. The io structure has a flag to * avoid double converting from both fsync and background work * queue work. */ @@ -125,9 +125,11 @@ extern int ext4_flush_completed_IO(struct inode *inode) * the parent directory's parent as well, and so on recursively, if * they are also freshly created. */ -static void ext4_sync_parent(struct inode *inode) +static int ext4_sync_parent(struct inode *inode) { + struct writeback_control wbc; struct dentry *dentry = NULL; + int ret = 0; while (inode && ext4_test_inode_state(inode, EXT4_STATE_NEWENTRY)) { ext4_clear_inode_state(inode, EXT4_STATE_NEWENTRY); @@ -136,8 +138,17 @@ static void ext4_sync_parent(struct inode *inode) if (!dentry || !dentry->d_parent || !dentry->d_parent->d_inode) break; inode = dentry->d_parent->d_inode; - sync_mapping_buffers(inode->i_mapping); + ret = sync_mapping_buffers(inode->i_mapping); + if (ret) + break; + memset(&wbc, 0, sizeof(wbc)); + wbc.sync_mode = WB_SYNC_ALL; + wbc.nr_to_write = 0; /* only write out the inode */ + ret = sync_inode(inode, &wbc); + if (ret) + break; } + return ret; } /* @@ -161,6 +172,7 @@ int ext4_sync_file(struct file *file, int datasync) journal_t *journal = EXT4_SB(inode->i_sb)->s_journal; int ret; tid_t commit_tid; + bool needs_barrier = false; J_ASSERT(ext4_journal_current_handle() == NULL); @@ -176,7 +188,7 @@ int ext4_sync_file(struct file *file, int datasync) if (!journal) { ret = generic_file_fsync(file, datasync); if (!ret && !list_empty(&inode->i_dentry)) - ext4_sync_parent(inode); + ret = ext4_sync_parent(inode); goto out; } @@ -200,22 +212,12 @@ int ext4_sync_file(struct file *file, int datasync) } commit_tid = datasync ? ei->i_datasync_tid : ei->i_sync_tid; - if (jbd2_log_start_commit(journal, commit_tid)) { - /* - * When the journal is on a different device than the - * fs data disk, we need to issue the barrier in - * writeback mode. (In ordered mode, the jbd2 layer - * will take care of issuing the barrier. In - * data=journal, all of the data blocks are written to - * the journal device.) - */ - if (ext4_should_writeback_data(inode) && - (journal->j_fs_dev != journal->j_dev) && - (journal->j_flags & JBD2_BARRIER)) - blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, - NULL); - ret = jbd2_log_wait_commit(journal, commit_tid); - } else if (journal->j_flags & JBD2_BARRIER) + if (journal->j_flags & JBD2_BARRIER && + !jbd2_trans_will_send_data_barrier(journal, commit_tid)) + needs_barrier = true; + jbd2_log_start_commit(journal, commit_tid); + ret = jbd2_log_wait_commit(journal, commit_tid); + if (needs_barrier) blkdev_issue_flush(inode->i_sb->s_bdev, GFP_KERNEL, NULL); out: trace_ext4_sync_file_exit(inode, ret); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 1a86282b9024..a5763e3505ba 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -639,8 +639,8 @@ static int ext4_alloc_blocks(handle_t *handle, struct inode *inode, while (target > 0) { count = target; /* allocating blocks for indirect blocks and direct blocks */ - current_block = ext4_new_meta_blocks(handle, inode, - goal, &count, err); + current_block = ext4_new_meta_blocks(handle, inode, goal, + 0, &count, err); if (*err) goto failed_out; @@ -1930,7 +1930,7 @@ repeat: * We do still charge estimated metadata to the sb though; * we cannot afford to run out of free blocks. */ - if (ext4_claim_free_blocks(sbi, md_needed + 1)) { + if (ext4_claim_free_blocks(sbi, md_needed + 1, 0)) { dquot_release_reservation_block(inode, 1); if (ext4_should_retry_alloc(inode->i_sb, &retries)) { yield(); @@ -2502,6 +2502,7 @@ static int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, * for partial write. */ set_buffer_new(bh); + set_buffer_mapped(bh); } return 0; } @@ -2588,7 +2589,7 @@ static void ext4_end_io_buffer_write(struct buffer_head *bh, int uptodate); * because we should have holes filled from ext4_page_mkwrite(). We even don't * need to file the inode to the transaction's list in ordered mode because if * we are writing back data added by write(), the inode is already there and if - * we are writing back data modified via mmap(), noone guarantees in which + * we are writing back data modified via mmap(), no one guarantees in which * transaction the data will hit the disk. In case we are journaling data, we * cannot start transaction directly because transaction start ranks above page * lock so we have to do some magic. @@ -2690,7 +2691,7 @@ static int ext4_writepage(struct page *page, /* * This is called via ext4_da_writepages() to - * calulate the total number of credits to reserve to fit + * calculate the total number of credits to reserve to fit * a single extent allocation into a single transaction, * ext4_da_writpeages() will loop calling this before * the block allocation. @@ -2795,9 +2796,7 @@ static int write_cache_pages_da(struct address_space *mapping, continue; } - if (PageWriteback(page)) - wait_on_page_writeback(page); - + wait_on_page_writeback(page); BUG_ON(PageWriteback(page)); if (mpd->next_page != page->index) @@ -3304,7 +3303,7 @@ int ext4_alloc_da_blocks(struct inode *inode) * the pages by calling redirty_page_for_writepage() but that * would be ugly in the extreme. So instead we would need to * replicate parts of the code in the above functions, - * simplifying them becuase we wouldn't actually intend to + * simplifying them because we wouldn't actually intend to * write out the pages, but rather only collect contiguous * logical block extents, call the multi-block allocator, and * then update the buffer heads with the block allocations. @@ -3512,7 +3511,7 @@ retry: loff_t end = offset + iov_length(iov, nr_segs); if (end > isize) - vmtruncate(inode, isize); + ext4_truncate_failed_write(inode); } } if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) @@ -3694,7 +3693,7 @@ retry: * * The unwrritten extents will be converted to written when DIO is completed. * For async direct IO, since the IO may still pending when return, we - * set up an end_io call back function, which will do the convertion + * set up an end_io call back function, which will do the conversion * when async direct IO completed. * * If the O_DIRECT write will extend the file then add this inode to the @@ -3717,7 +3716,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, * We could direct write to holes and fallocate. * * Allocated blocks to fill the hole are marked as uninitialized - * to prevent paralel buffered read to expose the stale data + * to prevent parallel buffered read to expose the stale data * before DIO complete the data IO. * * As to previously fallocated extents, ext4 get_block @@ -3778,7 +3777,7 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, int err; /* * for non AIO case, since the IO is already - * completed, we could do the convertion right here + * completed, we could do the conversion right here */ err = ext4_convert_unwritten_extents(inode, offset, ret); @@ -3915,9 +3914,30 @@ void ext4_set_aops(struct inode *inode) int ext4_block_truncate_page(handle_t *handle, struct address_space *mapping, loff_t from) { + unsigned offset = from & (PAGE_CACHE_SIZE-1); + unsigned length; + unsigned blocksize; + struct inode *inode = mapping->host; + + blocksize = inode->i_sb->s_blocksize; + length = blocksize - (offset & (blocksize - 1)); + + return ext4_block_zero_page_range(handle, mapping, from, length); +} + +/* + * ext4_block_zero_page_range() zeros out a mapping of length 'length' + * starting from file offset 'from'. The range to be zero'd must + * be contained with in one block. If the specified range exceeds + * the end of the block it will be shortened to end of the block + * that cooresponds to 'from' + */ +int ext4_block_zero_page_range(handle_t *handle, + struct address_space *mapping, loff_t from, loff_t length) +{ ext4_fsblk_t index = from >> PAGE_CACHE_SHIFT; unsigned offset = from & (PAGE_CACHE_SIZE-1); - unsigned blocksize, length, pos; + unsigned blocksize, max, pos; ext4_lblk_t iblock; struct inode *inode = mapping->host; struct buffer_head *bh; @@ -3930,7 +3950,15 @@ int ext4_block_truncate_page(handle_t *handle, return -EINVAL; blocksize = inode->i_sb->s_blocksize; - length = blocksize - (offset & (blocksize - 1)); + max = blocksize - (offset & (blocksize - 1)); + + /* + * correct length if it does not fall between + * 'from' and the end of the block + */ + if (length > max || length < 0) + length = max; + iblock = index << (PAGE_CACHE_SHIFT - inode->i_sb->s_blocksize_bits); if (!page_has_buffers(page)) @@ -4025,7 +4053,7 @@ static inline int all_zeroes(__le32 *p, __le32 *q) * * When we do truncate() we may have to clean the ends of several * indirect blocks but leave the blocks themselves alive. Block is - * partially truncated if some data below the new i_size is refered + * partially truncated if some data below the new i_size is referred * from it (and it is on the path to the first completely truncated * data block, indeed). We have to free the top of that path along * with everything to the right of the path. Since no allocation @@ -4169,7 +4197,7 @@ out_err: * @first: array of block numbers * @last: points immediately past the end of array * - * We are freeing all blocks refered from that array (numbers are stored as + * We are freeing all blocks referred from that array (numbers are stored as * little-endian 32-bit) and updating @inode->i_blocks appropriately. * * We accumulate contiguous runs of blocks to free. Conveniently, if these @@ -4261,7 +4289,7 @@ static void ext4_free_data(handle_t *handle, struct inode *inode, * @last: pointer immediately past the end of array * @depth: depth of the branches to free * - * We are freeing all blocks refered from these branches (numbers are + * We are freeing all blocks referred from these branches (numbers are * stored as little-endian 32-bit) and updating @inode->i_blocks * appropriately. */ @@ -4379,8 +4407,6 @@ static void ext4_free_branches(handle_t *handle, struct inode *inode, int ext4_can_truncate(struct inode *inode) { - if (IS_APPEND(inode) || IS_IMMUTABLE(inode)) - return 0; if (S_ISREG(inode->i_mode)) return 1; if (S_ISDIR(inode->i_mode)) @@ -4391,6 +4417,31 @@ int ext4_can_truncate(struct inode *inode) } /* + * ext4_punch_hole: punches a hole in a file by releaseing the blocks + * associated with the given offset and length + * + * @inode: File inode + * @offset: The offset where the hole will begin + * @len: The length of the hole + * + * Returns: 0 on sucess or negative on failure + */ + +int ext4_punch_hole(struct file *file, loff_t offset, loff_t length) +{ + struct inode *inode = file->f_path.dentry->d_inode; + if (!S_ISREG(inode->i_mode)) + return -ENOTSUPP; + + if (!ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + /* TODO: Add support for non extent hole punching */ + return -ENOTSUPP; + } + + return ext4_ext_punch_hole(file, offset, length); +} + +/* * ext4_truncate() * * We block out ext4_get_block() block instantiations across the entire @@ -4429,8 +4480,8 @@ void ext4_truncate(struct inode *inode) Indirect chain[4]; Indirect *partial; __le32 nr = 0; - int n; - ext4_lblk_t last_block; + int n = 0; + ext4_lblk_t last_block, max_block; unsigned blocksize = inode->i_sb->s_blocksize; trace_ext4_truncate_enter(inode); @@ -4455,14 +4506,18 @@ void ext4_truncate(struct inode *inode) last_block = (inode->i_size + blocksize-1) >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); + max_block = (EXT4_SB(inode->i_sb)->s_bitmap_maxbytes + blocksize-1) + >> EXT4_BLOCK_SIZE_BITS(inode->i_sb); if (inode->i_size & (blocksize - 1)) if (ext4_block_truncate_page(handle, mapping, inode->i_size)) goto out_stop; - n = ext4_block_to_path(inode, last_block, offsets, NULL); - if (n == 0) - goto out_stop; /* error */ + if (last_block != max_block) { + n = ext4_block_to_path(inode, last_block, offsets, NULL); + if (n == 0) + goto out_stop; /* error */ + } /* * OK. This truncate is going to happen. We add the inode to the @@ -4493,7 +4548,13 @@ void ext4_truncate(struct inode *inode) */ ei->i_disksize = inode->i_size; - if (n == 1) { /* direct blocks */ + if (last_block == max_block) { + /* + * It is unnecessary to free any data blocks if last_block is + * equal to the indirect block limit. + */ + goto out_unlock; + } else if (n == 1) { /* direct blocks */ ext4_free_data(handle, inode, NULL, i_data+offsets[0], i_data + EXT4_NDIR_BLOCKS); goto do_indirects; @@ -4553,6 +4614,7 @@ do_indirects: ; } +out_unlock: up_write(&ei->i_data_sem); inode->i_mtime = inode->i_ctime = ext4_current_time(inode); ext4_mark_inode_dirty(handle, inode); @@ -4605,7 +4667,7 @@ static int __ext4_get_inode_loc(struct inode *inode, /* * Figure out the offset within the block group inode table */ - inodes_per_block = (EXT4_BLOCK_SIZE(sb) / EXT4_INODE_SIZE(sb)); + inodes_per_block = EXT4_SB(sb)->s_inodes_per_block; inode_offset = ((inode->i_ino - 1) % EXT4_INODES_PER_GROUP(sb)); block = ext4_inode_table(sb, gdp) + (inode_offset / inodes_per_block); @@ -5299,8 +5361,7 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) if (S_ISREG(inode->i_mode) && attr->ia_valid & ATTR_SIZE && - (attr->ia_size < inode->i_size || - (ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS)))) { + (attr->ia_size < inode->i_size)) { handle_t *handle; handle = ext4_journal_start(inode, 3); @@ -5334,14 +5395,15 @@ int ext4_setattr(struct dentry *dentry, struct iattr *attr) goto err_out; } } - /* ext4_truncate will clear the flag */ - if ((ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS))) - ext4_truncate(inode); } - if ((attr->ia_valid & ATTR_SIZE) && - attr->ia_size != i_size_read(inode)) - rc = vmtruncate(inode, attr->ia_size); + if (attr->ia_valid & ATTR_SIZE) { + if (attr->ia_size != i_size_read(inode)) { + truncate_setsize(inode, attr->ia_size); + ext4_truncate(inode); + } else if (ext4_test_inode_flag(inode, EXT4_INODE_EOFBLOCKS)) + ext4_truncate(inode); + } if (!rc) { setattr_copy(inode, attr); @@ -5398,13 +5460,12 @@ static int ext4_indirect_trans_blocks(struct inode *inode, int nrblocks, /* if nrblocks are contiguous */ if (chunk) { /* - * With N contiguous data blocks, it need at most - * N/EXT4_ADDR_PER_BLOCK(inode->i_sb) indirect blocks - * 2 dindirect blocks - * 1 tindirect block + * With N contiguous data blocks, we need at most + * N/EXT4_ADDR_PER_BLOCK(inode->i_sb) + 1 indirect blocks, + * 2 dindirect blocks, and 1 tindirect block */ - indirects = nrblocks / EXT4_ADDR_PER_BLOCK(inode->i_sb); - return indirects + 3; + return DIV_ROUND_UP(nrblocks, + EXT4_ADDR_PER_BLOCK(inode->i_sb)) + 4; } /* * if nrblocks are not contiguous, worse case, each block touch @@ -5478,7 +5539,7 @@ static int ext4_meta_trans_blocks(struct inode *inode, int nrblocks, int chunk) } /* - * Calulate the total number of credits to reserve to fit + * Calculate the total number of credits to reserve to fit * the modification of a single pages into a single transaction, * which may include multiple chunks of block allocations. * @@ -5672,7 +5733,7 @@ int ext4_mark_inode_dirty(handle_t *handle, struct inode *inode) * so would cause a commit on atime updates, which we don't bother doing. * We handle synchronous inodes at the highest possible level. */ -void ext4_dirty_inode(struct inode *inode) +void ext4_dirty_inode(struct inode *inode, int flags) { handle_t *handle; @@ -5800,15 +5861,19 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) goto out_unlock; } ret = 0; - if (PageMappedToDisk(page)) - goto out_unlock; + + lock_page(page); + wait_on_page_writeback(page); + if (PageMappedToDisk(page)) { + up_read(&inode->i_alloc_sem); + return VM_FAULT_LOCKED; + } if (page->index == size >> PAGE_CACHE_SHIFT) len = size & ~PAGE_CACHE_MASK; else len = PAGE_CACHE_SIZE; - lock_page(page); /* * return if we have all the buffers mapped. This avoid * the need to call write_begin/write_end which does a @@ -5818,8 +5883,8 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) if (page_has_buffers(page)) { if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, ext4_bh_unmapped)) { - unlock_page(page); - goto out_unlock; + up_read(&inode->i_alloc_sem); + return VM_FAULT_LOCKED; } } unlock_page(page); @@ -5839,6 +5904,16 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) if (ret < 0) goto out_unlock; ret = 0; + + /* + * write_begin/end might have created a dirty page and someone + * could wander in and start the IO. Make sure that hasn't + * happened. + */ + lock_page(page); + wait_on_page_writeback(page); + up_read(&inode->i_alloc_sem); + return VM_FAULT_LOCKED; out_unlock: if (ret) ret = VM_FAULT_SIGBUS; diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index a5837a837a8b..859f2ae8864e 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -92,7 +92,7 @@ * between CPUs. It is possible to get scheduled at this point. * * The locality group prealloc space is used looking at whether we have - * enough free space (pa_free) withing the prealloc space. + * enough free space (pa_free) within the prealloc space. * * If we can't allocate blocks via inode prealloc or/and locality group * prealloc then we look at the buddy cache. The buddy cache is represented @@ -787,6 +787,7 @@ static int ext4_mb_init_cache(struct page *page, char *incore) struct inode *inode; char *data; char *bitmap; + struct ext4_group_info *grinfo; mb_debug(1, "init page %lu\n", page->index); @@ -819,6 +820,18 @@ static int ext4_mb_init_cache(struct page *page, char *incore) if (first_group + i >= ngroups) break; + grinfo = ext4_get_group_info(sb, first_group + i); + /* + * If page is uptodate then we came here after online resize + * which added some new uninitialized group info structs, so + * we must skip all initialized uptodate buddies on the page, + * which may be currently in use by an allocating task. + */ + if (PageUptodate(page) && !EXT4_MB_GRP_NEED_INIT(grinfo)) { + bh[i] = NULL; + continue; + } + err = -EIO; desc = ext4_get_group_desc(sb, first_group + i, NULL); if (desc == NULL) @@ -871,26 +884,28 @@ static int ext4_mb_init_cache(struct page *page, char *incore) } /* wait for I/O completion */ - for (i = 0; i < groups_per_page && bh[i]; i++) - wait_on_buffer(bh[i]); + for (i = 0; i < groups_per_page; i++) + if (bh[i]) + wait_on_buffer(bh[i]); err = -EIO; - for (i = 0; i < groups_per_page && bh[i]; i++) - if (!buffer_uptodate(bh[i])) + for (i = 0; i < groups_per_page; i++) + if (bh[i] && !buffer_uptodate(bh[i])) goto out; err = 0; first_block = page->index * blocks_per_page; - /* init the page */ - memset(page_address(page), 0xff, PAGE_CACHE_SIZE); for (i = 0; i < blocks_per_page; i++) { int group; - struct ext4_group_info *grinfo; group = (first_block + i) >> 1; if (group >= ngroups) break; + if (!bh[group - first_group]) + /* skip initialized uptodate buddy */ + continue; + /* * data carry information regarding this * particular group in the format specified @@ -919,6 +934,8 @@ static int ext4_mb_init_cache(struct page *page, char *incore) * incore got set to the group block bitmap below */ ext4_lock_group(sb, group); + /* init the buddy */ + memset(data, 0xff, blocksize); ext4_mb_generate_buddy(sb, data, incore, group); ext4_unlock_group(sb, group); incore = NULL; @@ -948,7 +965,7 @@ static int ext4_mb_init_cache(struct page *page, char *incore) out: if (bh) { - for (i = 0; i < groups_per_page && bh[i]; i++) + for (i = 0; i < groups_per_page; i++) brelse(bh[i]); if (bh != &bhs) kfree(bh); @@ -957,22 +974,21 @@ out: } /* - * lock the group_info alloc_sem of all the groups - * belonging to the same buddy cache page. This - * make sure other parallel operation on the buddy - * cache doesn't happen whild holding the buddy cache - * lock + * Lock the buddy and bitmap pages. This make sure other parallel init_group + * on the same buddy page doesn't happen whild holding the buddy page lock. + * Return locked buddy and bitmap pages on e4b struct. If buddy and bitmap + * are on the same page e4b->bd_buddy_page is NULL and return value is 0. */ -static int ext4_mb_get_buddy_cache_lock(struct super_block *sb, - ext4_group_t group) +static int ext4_mb_get_buddy_page_lock(struct super_block *sb, + ext4_group_t group, struct ext4_buddy *e4b) { - int i; - int block, pnum; + struct inode *inode = EXT4_SB(sb)->s_buddy_cache; + int block, pnum, poff; int blocks_per_page; - int groups_per_page; - ext4_group_t ngroups = ext4_get_groups_count(sb); - ext4_group_t first_group; - struct ext4_group_info *grp; + struct page *page; + + e4b->bd_buddy_page = NULL; + e4b->bd_bitmap_page = NULL; blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize; /* @@ -982,57 +998,40 @@ static int ext4_mb_get_buddy_cache_lock(struct super_block *sb, */ block = group * 2; pnum = block / blocks_per_page; - first_group = pnum * blocks_per_page / 2; - - groups_per_page = blocks_per_page >> 1; - if (groups_per_page == 0) - groups_per_page = 1; - /* read all groups the page covers into the cache */ - for (i = 0; i < groups_per_page; i++) { + poff = block % blocks_per_page; + page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); + if (!page) + return -EIO; + BUG_ON(page->mapping != inode->i_mapping); + e4b->bd_bitmap_page = page; + e4b->bd_bitmap = page_address(page) + (poff * sb->s_blocksize); - if ((first_group + i) >= ngroups) - break; - grp = ext4_get_group_info(sb, first_group + i); - /* take all groups write allocation - * semaphore. This make sure there is - * no block allocation going on in any - * of that groups - */ - down_write_nested(&grp->alloc_sem, i); + if (blocks_per_page >= 2) { + /* buddy and bitmap are on the same page */ + return 0; } - return i; + + block++; + pnum = block / blocks_per_page; + poff = block % blocks_per_page; + page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); + if (!page) + return -EIO; + BUG_ON(page->mapping != inode->i_mapping); + e4b->bd_buddy_page = page; + return 0; } -static void ext4_mb_put_buddy_cache_lock(struct super_block *sb, - ext4_group_t group, int locked_group) +static void ext4_mb_put_buddy_page_lock(struct ext4_buddy *e4b) { - int i; - int block, pnum; - int blocks_per_page; - ext4_group_t first_group; - struct ext4_group_info *grp; - - blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize; - /* - * the buddy cache inode stores the block bitmap - * and buddy information in consecutive blocks. - * So for each group we need two blocks. - */ - block = group * 2; - pnum = block / blocks_per_page; - first_group = pnum * blocks_per_page / 2; - /* release locks on all the groups */ - for (i = 0; i < locked_group; i++) { - - grp = ext4_get_group_info(sb, first_group + i); - /* take all groups write allocation - * semaphore. This make sure there is - * no block allocation going on in any - * of that groups - */ - up_write(&grp->alloc_sem); + if (e4b->bd_bitmap_page) { + unlock_page(e4b->bd_bitmap_page); + page_cache_release(e4b->bd_bitmap_page); + } + if (e4b->bd_buddy_page) { + unlock_page(e4b->bd_buddy_page); + page_cache_release(e4b->bd_buddy_page); } - } /* @@ -1044,93 +1043,60 @@ static noinline_for_stack int ext4_mb_init_group(struct super_block *sb, ext4_group_t group) { - int ret = 0; - void *bitmap; - int blocks_per_page; - int block, pnum, poff; - int num_grp_locked = 0; struct ext4_group_info *this_grp; - struct ext4_sb_info *sbi = EXT4_SB(sb); - struct inode *inode = sbi->s_buddy_cache; - struct page *page = NULL, *bitmap_page = NULL; + struct ext4_buddy e4b; + struct page *page; + int ret = 0; mb_debug(1, "init group %u\n", group); - blocks_per_page = PAGE_CACHE_SIZE / sb->s_blocksize; this_grp = ext4_get_group_info(sb, group); /* * This ensures that we don't reinit the buddy cache * page which map to the group from which we are already * allocating. If we are looking at the buddy cache we would * have taken a reference using ext4_mb_load_buddy and that - * would have taken the alloc_sem lock. + * would have pinned buddy page to page cache. */ - num_grp_locked = ext4_mb_get_buddy_cache_lock(sb, group); - if (!EXT4_MB_GRP_NEED_INIT(this_grp)) { + ret = ext4_mb_get_buddy_page_lock(sb, group, &e4b); + if (ret || !EXT4_MB_GRP_NEED_INIT(this_grp)) { /* * somebody initialized the group * return without doing anything */ - ret = 0; goto err; } - /* - * the buddy cache inode stores the block bitmap - * and buddy information in consecutive blocks. - * So for each group we need two blocks. - */ - block = group * 2; - pnum = block / blocks_per_page; - poff = block % blocks_per_page; - page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); - if (page) { - BUG_ON(page->mapping != inode->i_mapping); - ret = ext4_mb_init_cache(page, NULL); - if (ret) { - unlock_page(page); - goto err; - } - unlock_page(page); - } - if (page == NULL || !PageUptodate(page)) { + + page = e4b.bd_bitmap_page; + ret = ext4_mb_init_cache(page, NULL); + if (ret) + goto err; + if (!PageUptodate(page)) { ret = -EIO; goto err; } mark_page_accessed(page); - bitmap_page = page; - bitmap = page_address(page) + (poff * sb->s_blocksize); - /* init buddy cache */ - block++; - pnum = block / blocks_per_page; - poff = block % blocks_per_page; - page = find_or_create_page(inode->i_mapping, pnum, GFP_NOFS); - if (page == bitmap_page) { + if (e4b.bd_buddy_page == NULL) { /* * If both the bitmap and buddy are in * the same page we don't need to force * init the buddy */ - unlock_page(page); - } else if (page) { - BUG_ON(page->mapping != inode->i_mapping); - ret = ext4_mb_init_cache(page, bitmap); - if (ret) { - unlock_page(page); - goto err; - } - unlock_page(page); + ret = 0; + goto err; } - if (page == NULL || !PageUptodate(page)) { + /* init buddy cache */ + page = e4b.bd_buddy_page; + ret = ext4_mb_init_cache(page, e4b.bd_bitmap); + if (ret) + goto err; + if (!PageUptodate(page)) { ret = -EIO; goto err; } mark_page_accessed(page); err: - ext4_mb_put_buddy_cache_lock(sb, group, num_grp_locked); - if (bitmap_page) - page_cache_release(bitmap_page); - if (page) - page_cache_release(page); + ext4_mb_put_buddy_page_lock(&e4b); return ret; } @@ -1164,24 +1130,8 @@ ext4_mb_load_buddy(struct super_block *sb, ext4_group_t group, e4b->bd_group = group; e4b->bd_buddy_page = NULL; e4b->bd_bitmap_page = NULL; - e4b->alloc_semp = &grp->alloc_sem; - - /* Take the read lock on the group alloc - * sem. This would make sure a parallel - * ext4_mb_init_group happening on other - * groups mapped by the page is blocked - * till we are done with allocation - */ -repeat_load_buddy: - down_read(e4b->alloc_semp); if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) { - /* we need to check for group need init flag - * with alloc_semp held so that we can be sure - * that new blocks didn't get added to the group - * when we are loading the buddy cache - */ - up_read(e4b->alloc_semp); /* * we need full data about the group * to make a good selection @@ -1189,7 +1139,6 @@ repeat_load_buddy: ret = ext4_mb_init_group(sb, group); if (ret) return ret; - goto repeat_load_buddy; } /* @@ -1273,15 +1222,14 @@ repeat_load_buddy: return 0; err: + if (page) + page_cache_release(page); if (e4b->bd_bitmap_page) page_cache_release(e4b->bd_bitmap_page); if (e4b->bd_buddy_page) page_cache_release(e4b->bd_buddy_page); e4b->bd_buddy = NULL; e4b->bd_bitmap = NULL; - - /* Done with the buddy cache */ - up_read(e4b->alloc_semp); return ret; } @@ -1291,9 +1239,6 @@ static void ext4_mb_unload_buddy(struct ext4_buddy *e4b) page_cache_release(e4b->bd_bitmap_page); if (e4b->bd_buddy_page) page_cache_release(e4b->bd_buddy_page); - /* Done with the buddy cache */ - if (e4b->alloc_semp) - up_read(e4b->alloc_semp); } @@ -1606,9 +1551,6 @@ static void ext4_mb_use_best_found(struct ext4_allocation_context *ac, get_page(ac->ac_bitmap_page); ac->ac_buddy_page = e4b->bd_buddy_page; get_page(ac->ac_buddy_page); - /* on allocation we use ac to track the held semaphore */ - ac->alloc_semp = e4b->alloc_semp; - e4b->alloc_semp = NULL; /* store last allocated for subsequent stream allocation */ if (ac->ac_flags & EXT4_MB_STREAM_ALLOC) { spin_lock(&sbi->s_md_lock); @@ -2659,7 +2601,7 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn) struct super_block *sb = journal->j_private; struct ext4_buddy e4b; struct ext4_group_info *db; - int err, ret, count = 0, count2 = 0; + int err, count = 0, count2 = 0; struct ext4_free_data *entry; struct list_head *l, *ltmp; @@ -2669,15 +2611,9 @@ static void release_blocks_on_commit(journal_t *journal, transaction_t *txn) mb_debug(1, "gonna free %u blocks in group %u (0x%p):", entry->count, entry->group, entry); - if (test_opt(sb, DISCARD)) { - ret = ext4_issue_discard(sb, entry->group, - entry->start_blk, entry->count); - if (unlikely(ret == -EOPNOTSUPP)) { - ext4_warning(sb, "discard not supported, " - "disabling"); - clear_opt(sb, DISCARD); - } - } + if (test_opt(sb, DISCARD)) + ext4_issue_discard(sb, entry->group, + entry->start_blk, entry->count); err = ext4_mb_load_buddy(sb, entry->group, &e4b); /* we expect to find existing buddy because it's pinned */ @@ -4226,15 +4162,12 @@ static int ext4_mb_release_context(struct ext4_allocation_context *ac) spin_unlock(&pa->pa_lock); } } - if (ac->alloc_semp) - up_read(ac->alloc_semp); if (pa) { /* * We want to add the pa to the right bucket. * Remove it from the list and while adding * make sure the list to which we are adding - * doesn't grow big. We need to release - * alloc_semp before calling ext4_mb_add_n_trim() + * doesn't grow big. */ if ((pa->pa_type == MB_GROUP_PA) && likely(pa->pa_free)) { spin_lock(pa->pa_obj_lock); @@ -4303,7 +4236,9 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, * there is enough free blocks to do block allocation * and verify allocation doesn't exceed the quota limits. */ - while (ar->len && ext4_claim_free_blocks(sbi, ar->len)) { + while (ar->len && + ext4_claim_free_blocks(sbi, ar->len, ar->flags)) { + /* let others to free the space */ yield(); ar->len = ar->len >> 1; @@ -4313,9 +4248,15 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, return 0; } reserv_blks = ar->len; - while (ar->len && dquot_alloc_block(ar->inode, ar->len)) { - ar->flags |= EXT4_MB_HINT_NOPREALLOC; - ar->len--; + if (ar->flags & EXT4_MB_USE_ROOT_BLOCKS) { + dquot_alloc_block_nofail(ar->inode, ar->len); + } else { + while (ar->len && + dquot_alloc_block(ar->inode, ar->len)) { + + ar->flags |= EXT4_MB_HINT_NOPREALLOC; + ar->len--; + } } inquota = ar->len; if (ar->len == 0) { @@ -4704,6 +4645,127 @@ error_return: } /** + * ext4_add_groupblocks() -- Add given blocks to an existing group + * @handle: handle to this transaction + * @sb: super block + * @block: start physcial block to add to the block group + * @count: number of blocks to free + * + * This marks the blocks as free in the bitmap and buddy. + */ +void ext4_add_groupblocks(handle_t *handle, struct super_block *sb, + ext4_fsblk_t block, unsigned long count) +{ + struct buffer_head *bitmap_bh = NULL; + struct buffer_head *gd_bh; + ext4_group_t block_group; + ext4_grpblk_t bit; + unsigned int i; + struct ext4_group_desc *desc; + struct ext4_sb_info *sbi = EXT4_SB(sb); + struct ext4_buddy e4b; + int err = 0, ret, blk_free_count; + ext4_grpblk_t blocks_freed; + struct ext4_group_info *grp; + + ext4_debug("Adding block(s) %llu-%llu\n", block, block + count - 1); + + ext4_get_group_no_and_offset(sb, block, &block_group, &bit); + grp = ext4_get_group_info(sb, block_group); + /* + * Check to see if we are freeing blocks across a group + * boundary. + */ + if (bit + count > EXT4_BLOCKS_PER_GROUP(sb)) + goto error_return; + + bitmap_bh = ext4_read_block_bitmap(sb, block_group); + if (!bitmap_bh) + goto error_return; + desc = ext4_get_group_desc(sb, block_group, &gd_bh); + if (!desc) + goto error_return; + + if (in_range(ext4_block_bitmap(sb, desc), block, count) || + in_range(ext4_inode_bitmap(sb, desc), block, count) || + in_range(block, ext4_inode_table(sb, desc), sbi->s_itb_per_group) || + in_range(block + count - 1, ext4_inode_table(sb, desc), + sbi->s_itb_per_group)) { + ext4_error(sb, "Adding blocks in system zones - " + "Block = %llu, count = %lu", + block, count); + goto error_return; + } + + BUFFER_TRACE(bitmap_bh, "getting write access"); + err = ext4_journal_get_write_access(handle, bitmap_bh); + if (err) + goto error_return; + + /* + * We are about to modify some metadata. Call the journal APIs + * to unshare ->b_data if a currently-committing transaction is + * using it + */ + BUFFER_TRACE(gd_bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, gd_bh); + if (err) + goto error_return; + + for (i = 0, blocks_freed = 0; i < count; i++) { + BUFFER_TRACE(bitmap_bh, "clear bit"); + if (!mb_test_bit(bit + i, bitmap_bh->b_data)) { + ext4_error(sb, "bit already cleared for block %llu", + (ext4_fsblk_t)(block + i)); + BUFFER_TRACE(bitmap_bh, "bit already cleared"); + } else { + blocks_freed++; + } + } + + err = ext4_mb_load_buddy(sb, block_group, &e4b); + if (err) + goto error_return; + + /* + * need to update group_info->bb_free and bitmap + * with group lock held. generate_buddy look at + * them with group lock_held + */ + ext4_lock_group(sb, block_group); + mb_clear_bits(bitmap_bh->b_data, bit, count); + mb_free_blocks(NULL, &e4b, bit, count); + blk_free_count = blocks_freed + ext4_free_blks_count(sb, desc); + ext4_free_blks_set(sb, desc, blk_free_count); + desc->bg_checksum = ext4_group_desc_csum(sbi, block_group, desc); + ext4_unlock_group(sb, block_group); + percpu_counter_add(&sbi->s_freeblocks_counter, blocks_freed); + + if (sbi->s_log_groups_per_flex) { + ext4_group_t flex_group = ext4_flex_group(sbi, block_group); + atomic_add(blocks_freed, + &sbi->s_flex_groups[flex_group].free_blocks); + } + + ext4_mb_unload_buddy(&e4b); + + /* We dirtied the bitmap block */ + BUFFER_TRACE(bitmap_bh, "dirtied bitmap block"); + err = ext4_handle_dirty_metadata(handle, NULL, bitmap_bh); + + /* And the group descriptor block */ + BUFFER_TRACE(gd_bh, "dirtied group descriptor block"); + ret = ext4_handle_dirty_metadata(handle, NULL, gd_bh); + if (!err) + err = ret; + +error_return: + brelse(bitmap_bh); + ext4_std_error(sb, err); + return; +} + +/** * ext4_trim_extent -- function to TRIM one single free extent in the group * @sb: super block for the file system * @start: starting block of the free extent in the alloc. group @@ -4715,11 +4777,10 @@ error_return: * one will allocate those blocks, mark it as used in buddy bitmap. This must * be called with under the group lock. */ -static int ext4_trim_extent(struct super_block *sb, int start, int count, - ext4_group_t group, struct ext4_buddy *e4b) +static void ext4_trim_extent(struct super_block *sb, int start, int count, + ext4_group_t group, struct ext4_buddy *e4b) { struct ext4_free_extent ex; - int ret = 0; assert_spin_locked(ext4_group_lock_ptr(sb, group)); @@ -4733,12 +4794,9 @@ static int ext4_trim_extent(struct super_block *sb, int start, int count, */ mb_mark_used(e4b, &ex); ext4_unlock_group(sb, group); - - ret = ext4_issue_discard(sb, group, start, count); - + ext4_issue_discard(sb, group, start, count); ext4_lock_group(sb, group); mb_free_blocks(NULL, e4b, start, ex.fe_len); - return ret; } /** @@ -4760,21 +4818,26 @@ static int ext4_trim_extent(struct super_block *sb, int start, int count, * the group buddy bitmap. This is done until whole group is scanned. */ static ext4_grpblk_t -ext4_trim_all_free(struct super_block *sb, struct ext4_buddy *e4b, - ext4_grpblk_t start, ext4_grpblk_t max, ext4_grpblk_t minblocks) +ext4_trim_all_free(struct super_block *sb, ext4_group_t group, + ext4_grpblk_t start, ext4_grpblk_t max, + ext4_grpblk_t minblocks) { void *bitmap; ext4_grpblk_t next, count = 0; - ext4_group_t group; - int ret = 0; + struct ext4_buddy e4b; + int ret; - BUG_ON(e4b == NULL); + ret = ext4_mb_load_buddy(sb, group, &e4b); + if (ret) { + ext4_error(sb, "Error in loading buddy " + "information for %u", group); + return ret; + } + bitmap = e4b.bd_bitmap; - bitmap = e4b->bd_bitmap; - group = e4b->bd_group; - start = (e4b->bd_info->bb_first_free > start) ? - e4b->bd_info->bb_first_free : start; ext4_lock_group(sb, group); + start = (e4b.bd_info->bb_first_free > start) ? + e4b.bd_info->bb_first_free : start; while (start < max) { start = mb_find_next_zero_bit(bitmap, max, start); @@ -4783,10 +4846,8 @@ ext4_trim_all_free(struct super_block *sb, struct ext4_buddy *e4b, next = mb_find_next_bit(bitmap, max, start); if ((next - start) >= minblocks) { - ret = ext4_trim_extent(sb, start, - next - start, group, e4b); - if (ret < 0) - break; + ext4_trim_extent(sb, start, + next - start, group, &e4b); count += next - start; } start = next + 1; @@ -4802,17 +4863,15 @@ ext4_trim_all_free(struct super_block *sb, struct ext4_buddy *e4b, ext4_lock_group(sb, group); } - if ((e4b->bd_info->bb_free - count) < minblocks) + if ((e4b.bd_info->bb_free - count) < minblocks) break; } ext4_unlock_group(sb, group); + ext4_mb_unload_buddy(&e4b); ext4_debug("trimmed %d blocks in the group %d\n", count, group); - if (ret < 0) - count = ret; - return count; } @@ -4830,11 +4889,11 @@ ext4_trim_all_free(struct super_block *sb, struct ext4_buddy *e4b, */ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range) { - struct ext4_buddy e4b; + struct ext4_group_info *grp; ext4_group_t first_group, last_group; ext4_group_t group, ngroups = ext4_get_groups_count(sb); ext4_grpblk_t cnt = 0, first_block, last_block; - uint64_t start, len, minlen, trimmed; + uint64_t start, len, minlen, trimmed = 0; ext4_fsblk_t first_data_blk = le32_to_cpu(EXT4_SB(sb)->s_es->s_first_data_block); int ret = 0; @@ -4842,7 +4901,6 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range) start = range->start >> sb->s_blocksize_bits; len = range->len >> sb->s_blocksize_bits; minlen = range->minlen >> sb->s_blocksize_bits; - trimmed = 0; if (unlikely(minlen > EXT4_BLOCKS_PER_GROUP(sb))) return -EINVAL; @@ -4863,11 +4921,12 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range) return -EINVAL; for (group = first_group; group <= last_group; group++) { - ret = ext4_mb_load_buddy(sb, group, &e4b); - if (ret) { - ext4_error(sb, "Error in loading buddy " - "information for %u", group); - break; + grp = ext4_get_group_info(sb, group); + /* We only do this if the grp has never been initialized */ + if (unlikely(EXT4_MB_GRP_NEED_INIT(grp))) { + ret = ext4_mb_init_group(sb, group); + if (ret) + break; } /* @@ -4880,16 +4939,14 @@ int ext4_trim_fs(struct super_block *sb, struct fstrim_range *range) last_block = first_block + len; len -= last_block - first_block; - if (e4b.bd_info->bb_free >= minlen) { - cnt = ext4_trim_all_free(sb, &e4b, first_block, + if (grp->bb_free >= minlen) { + cnt = ext4_trim_all_free(sb, group, first_block, last_block, minlen); if (cnt < 0) { ret = cnt; - ext4_mb_unload_buddy(&e4b); break; } } - ext4_mb_unload_buddy(&e4b); trimmed += cnt; first_block = 0; } diff --git a/fs/ext4/mballoc.h b/fs/ext4/mballoc.h index 22bd4d7f289b..20b5e7bfebd1 100644 --- a/fs/ext4/mballoc.h +++ b/fs/ext4/mballoc.h @@ -193,11 +193,6 @@ struct ext4_allocation_context { __u8 ac_op; /* operation, for history only */ struct page *ac_bitmap_page; struct page *ac_buddy_page; - /* - * pointer to the held semaphore upon successful - * block allocation - */ - struct rw_semaphore *alloc_semp; struct ext4_prealloc_space *ac_pa; struct ext4_locality_group *ac_lg; }; @@ -215,7 +210,6 @@ struct ext4_buddy { struct super_block *bd_sb; __u16 bd_blkbits; ext4_group_t bd_group; - struct rw_semaphore *alloc_semp; }; #define EXT4_MB_BITMAP(e4b) ((e4b)->bd_bitmap) #define EXT4_MB_BUDDY(e4b) ((e4b)->bd_buddy) diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index d1bafa57f483..b57b98fb44d1 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -376,7 +376,7 @@ static int ext4_ext_swap_inode_data(handle_t *handle, struct inode *inode, * We have the extent map build with the tmp inode. * Now copy the i_data across */ - ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS); + ext4_set_inode_flag(inode, EXT4_INODE_EXTENTS); memcpy(ei->i_data, tmp_ei->i_data, sizeof(ei->i_data)); /* @@ -517,7 +517,7 @@ int ext4_ext_migrate(struct inode *inode) * start with one credit accounted for * superblock modification. * - * For the tmp_inode we already have commited the + * For the tmp_inode we already have committed the * trascation that created the inode. Later as and * when we add extents we extent the journal */ diff --git a/fs/ext4/mmp.c b/fs/ext4/mmp.c new file mode 100644 index 000000000000..9bdef3f537c5 --- /dev/null +++ b/fs/ext4/mmp.c @@ -0,0 +1,351 @@ +#include <linux/fs.h> +#include <linux/random.h> +#include <linux/buffer_head.h> +#include <linux/utsname.h> +#include <linux/kthread.h> + +#include "ext4.h" + +/* + * Write the MMP block using WRITE_SYNC to try to get the block on-disk + * faster. + */ +static int write_mmp_block(struct buffer_head *bh) +{ + mark_buffer_dirty(bh); + lock_buffer(bh); + bh->b_end_io = end_buffer_write_sync; + get_bh(bh); + submit_bh(WRITE_SYNC, bh); + wait_on_buffer(bh); + if (unlikely(!buffer_uptodate(bh))) + return 1; + + return 0; +} + +/* + * Read the MMP block. It _must_ be read from disk and hence we clear the + * uptodate flag on the buffer. + */ +static int read_mmp_block(struct super_block *sb, struct buffer_head **bh, + ext4_fsblk_t mmp_block) +{ + struct mmp_struct *mmp; + + if (*bh) + clear_buffer_uptodate(*bh); + + /* This would be sb_bread(sb, mmp_block), except we need to be sure + * that the MD RAID device cache has been bypassed, and that the read + * is not blocked in the elevator. */ + if (!*bh) + *bh = sb_getblk(sb, mmp_block); + if (*bh) { + get_bh(*bh); + lock_buffer(*bh); + (*bh)->b_end_io = end_buffer_read_sync; + submit_bh(READ_SYNC, *bh); + wait_on_buffer(*bh); + if (!buffer_uptodate(*bh)) { + brelse(*bh); + *bh = NULL; + } + } + if (!*bh) { + ext4_warning(sb, "Error while reading MMP block %llu", + mmp_block); + return -EIO; + } + + mmp = (struct mmp_struct *)((*bh)->b_data); + if (le32_to_cpu(mmp->mmp_magic) != EXT4_MMP_MAGIC) + return -EINVAL; + + return 0; +} + +/* + * Dump as much information as possible to help the admin. + */ +void __dump_mmp_msg(struct super_block *sb, struct mmp_struct *mmp, + const char *function, unsigned int line, const char *msg) +{ + __ext4_warning(sb, function, line, msg); + __ext4_warning(sb, function, line, + "MMP failure info: last update time: %llu, last update " + "node: %s, last update device: %s\n", + (long long unsigned int) le64_to_cpu(mmp->mmp_time), + mmp->mmp_nodename, mmp->mmp_bdevname); +} + +/* + * kmmpd will update the MMP sequence every s_mmp_update_interval seconds + */ +static int kmmpd(void *data) +{ + struct super_block *sb = ((struct mmpd_data *) data)->sb; + struct buffer_head *bh = ((struct mmpd_data *) data)->bh; + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + struct mmp_struct *mmp; + ext4_fsblk_t mmp_block; + u32 seq = 0; + unsigned long failed_writes = 0; + int mmp_update_interval = le16_to_cpu(es->s_mmp_update_interval); + unsigned mmp_check_interval; + unsigned long last_update_time; + unsigned long diff; + int retval; + + mmp_block = le64_to_cpu(es->s_mmp_block); + mmp = (struct mmp_struct *)(bh->b_data); + mmp->mmp_time = cpu_to_le64(get_seconds()); + /* + * Start with the higher mmp_check_interval and reduce it if + * the MMP block is being updated on time. + */ + mmp_check_interval = max(EXT4_MMP_CHECK_MULT * mmp_update_interval, + EXT4_MMP_MIN_CHECK_INTERVAL); + mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval); + bdevname(bh->b_bdev, mmp->mmp_bdevname); + + memcpy(mmp->mmp_nodename, init_utsname()->sysname, + sizeof(mmp->mmp_nodename)); + + while (!kthread_should_stop()) { + if (++seq > EXT4_MMP_SEQ_MAX) + seq = 1; + + mmp->mmp_seq = cpu_to_le32(seq); + mmp->mmp_time = cpu_to_le64(get_seconds()); + last_update_time = jiffies; + + retval = write_mmp_block(bh); + /* + * Don't spew too many error messages. Print one every + * (s_mmp_update_interval * 60) seconds. + */ + if (retval && (failed_writes % 60) == 0) { + ext4_error(sb, "Error writing to MMP block"); + failed_writes++; + } + + if (!(le32_to_cpu(es->s_feature_incompat) & + EXT4_FEATURE_INCOMPAT_MMP)) { + ext4_warning(sb, "kmmpd being stopped since MMP feature" + " has been disabled."); + EXT4_SB(sb)->s_mmp_tsk = NULL; + goto failed; + } + + if (sb->s_flags & MS_RDONLY) { + ext4_warning(sb, "kmmpd being stopped since filesystem " + "has been remounted as readonly."); + EXT4_SB(sb)->s_mmp_tsk = NULL; + goto failed; + } + + diff = jiffies - last_update_time; + if (diff < mmp_update_interval * HZ) + schedule_timeout_interruptible(mmp_update_interval * + HZ - diff); + + /* + * We need to make sure that more than mmp_check_interval + * seconds have not passed since writing. If that has happened + * we need to check if the MMP block is as we left it. + */ + diff = jiffies - last_update_time; + if (diff > mmp_check_interval * HZ) { + struct buffer_head *bh_check = NULL; + struct mmp_struct *mmp_check; + + retval = read_mmp_block(sb, &bh_check, mmp_block); + if (retval) { + ext4_error(sb, "error reading MMP data: %d", + retval); + + EXT4_SB(sb)->s_mmp_tsk = NULL; + goto failed; + } + + mmp_check = (struct mmp_struct *)(bh_check->b_data); + if (mmp->mmp_seq != mmp_check->mmp_seq || + memcmp(mmp->mmp_nodename, mmp_check->mmp_nodename, + sizeof(mmp->mmp_nodename))) { + dump_mmp_msg(sb, mmp_check, + "Error while updating MMP info. " + "The filesystem seems to have been" + " multiply mounted."); + ext4_error(sb, "abort"); + goto failed; + } + put_bh(bh_check); + } + + /* + * Adjust the mmp_check_interval depending on how much time + * it took for the MMP block to be written. + */ + mmp_check_interval = max(min(EXT4_MMP_CHECK_MULT * diff / HZ, + EXT4_MMP_MAX_CHECK_INTERVAL), + EXT4_MMP_MIN_CHECK_INTERVAL); + mmp->mmp_check_interval = cpu_to_le16(mmp_check_interval); + } + + /* + * Unmount seems to be clean. + */ + mmp->mmp_seq = cpu_to_le32(EXT4_MMP_SEQ_CLEAN); + mmp->mmp_time = cpu_to_le64(get_seconds()); + + retval = write_mmp_block(bh); + +failed: + kfree(data); + brelse(bh); + return retval; +} + +/* + * Get a random new sequence number but make sure it is not greater than + * EXT4_MMP_SEQ_MAX. + */ +static unsigned int mmp_new_seq(void) +{ + u32 new_seq; + + do { + get_random_bytes(&new_seq, sizeof(u32)); + } while (new_seq > EXT4_MMP_SEQ_MAX); + + return new_seq; +} + +/* + * Protect the filesystem from being mounted more than once. + */ +int ext4_multi_mount_protect(struct super_block *sb, + ext4_fsblk_t mmp_block) +{ + struct ext4_super_block *es = EXT4_SB(sb)->s_es; + struct buffer_head *bh = NULL; + struct mmp_struct *mmp = NULL; + struct mmpd_data *mmpd_data; + u32 seq; + unsigned int mmp_check_interval = le16_to_cpu(es->s_mmp_update_interval); + unsigned int wait_time = 0; + int retval; + + if (mmp_block < le32_to_cpu(es->s_first_data_block) || + mmp_block >= ext4_blocks_count(es)) { + ext4_warning(sb, "Invalid MMP block in superblock"); + goto failed; + } + + retval = read_mmp_block(sb, &bh, mmp_block); + if (retval) + goto failed; + + mmp = (struct mmp_struct *)(bh->b_data); + + if (mmp_check_interval < EXT4_MMP_MIN_CHECK_INTERVAL) + mmp_check_interval = EXT4_MMP_MIN_CHECK_INTERVAL; + + /* + * If check_interval in MMP block is larger, use that instead of + * update_interval from the superblock. + */ + if (mmp->mmp_check_interval > mmp_check_interval) + mmp_check_interval = mmp->mmp_check_interval; + + seq = le32_to_cpu(mmp->mmp_seq); + if (seq == EXT4_MMP_SEQ_CLEAN) + goto skip; + + if (seq == EXT4_MMP_SEQ_FSCK) { + dump_mmp_msg(sb, mmp, "fsck is running on the filesystem"); + goto failed; + } + + wait_time = min(mmp_check_interval * 2 + 1, + mmp_check_interval + 60); + + /* Print MMP interval if more than 20 secs. */ + if (wait_time > EXT4_MMP_MIN_CHECK_INTERVAL * 4) + ext4_warning(sb, "MMP interval %u higher than expected, please" + " wait.\n", wait_time * 2); + + if (schedule_timeout_interruptible(HZ * wait_time) != 0) { + ext4_warning(sb, "MMP startup interrupted, failing mount\n"); + goto failed; + } + + retval = read_mmp_block(sb, &bh, mmp_block); + if (retval) + goto failed; + mmp = (struct mmp_struct *)(bh->b_data); + if (seq != le32_to_cpu(mmp->mmp_seq)) { + dump_mmp_msg(sb, mmp, + "Device is already active on another node."); + goto failed; + } + +skip: + /* + * write a new random sequence number. + */ + mmp->mmp_seq = seq = cpu_to_le32(mmp_new_seq()); + + retval = write_mmp_block(bh); + if (retval) + goto failed; + + /* + * wait for MMP interval and check mmp_seq. + */ + if (schedule_timeout_interruptible(HZ * wait_time) != 0) { + ext4_warning(sb, "MMP startup interrupted, failing mount\n"); + goto failed; + } + + retval = read_mmp_block(sb, &bh, mmp_block); + if (retval) + goto failed; + mmp = (struct mmp_struct *)(bh->b_data); + if (seq != le32_to_cpu(mmp->mmp_seq)) { + dump_mmp_msg(sb, mmp, + "Device is already active on another node."); + goto failed; + } + + mmpd_data = kmalloc(sizeof(struct mmpd_data), GFP_KERNEL); + if (!mmpd_data) { + ext4_warning(sb, "not enough memory for mmpd_data"); + goto failed; + } + mmpd_data->sb = sb; + mmpd_data->bh = bh; + + /* + * Start a kernel thread to update the MMP block periodically. + */ + EXT4_SB(sb)->s_mmp_tsk = kthread_run(kmmpd, mmpd_data, "kmmpd-%s", + bdevname(bh->b_bdev, + mmp->mmp_bdevname)); + if (IS_ERR(EXT4_SB(sb)->s_mmp_tsk)) { + EXT4_SB(sb)->s_mmp_tsk = NULL; + kfree(mmpd_data); + ext4_warning(sb, "Unable to create kmmpd thread for %s.", + sb->s_id); + goto failed; + } + + return 0; + +failed: + brelse(bh); + return 1; +} + + diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index b9f3e7862f13..2b8304bf3c50 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -876,8 +876,7 @@ move_extent_per_page(struct file *o_filp, struct inode *donor_inode, * It needs to call wait_on_page_writeback() to wait for the * writeback of the page. */ - if (PageWriteback(page)) - wait_on_page_writeback(page); + wait_on_page_writeback(page); /* Release old bh and drop refs */ try_to_release_page(page, 0); diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 67fd0b025858..b754b7721f51 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -1413,10 +1413,22 @@ static int make_indexed_dir(handle_t *handle, struct dentry *dentry, frame->at = entries; frame->bh = bh; bh = bh2; + + ext4_handle_dirty_metadata(handle, dir, frame->bh); + ext4_handle_dirty_metadata(handle, dir, bh); + de = do_split(handle,dir, &bh, frame, &hinfo, &retval); - dx_release (frames); - if (!(de)) + if (!de) { + /* + * Even if the block split failed, we have to properly write + * out all the changes we did so far. Otherwise we can end up + * with corrupted filesystem. + */ + ext4_mark_inode_dirty(handle, dir); + dx_release(frames); return retval; + } + dx_release(frames); retval = add_dirent_to_buf(handle, dentry, inode, de, bh); brelse(bh); @@ -2240,6 +2252,7 @@ static int ext4_symlink(struct inode *dir, handle_t *handle; struct inode *inode; int l, err, retries = 0; + int credits; l = strlen(symname)+1; if (l > dir->i_sb->s_blocksize) @@ -2247,10 +2260,26 @@ static int ext4_symlink(struct inode *dir, dquot_initialize(dir); + if (l > EXT4_N_BLOCKS * 4) { + /* + * For non-fast symlinks, we just allocate inode and put it on + * orphan list in the first transaction => we need bitmap, + * group descriptor, sb, inode block, quota blocks. + */ + credits = 4 + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb); + } else { + /* + * Fast symlink. We have to add entry to directory + * (EXT4_DATA_TRANS_BLOCKS + EXT4_INDEX_EXTRA_TRANS_BLOCKS), + * allocate new inode (bitmap, group descriptor, inode block, + * quota blocks, sb is already counted in previous macros). + */ + credits = EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 3 + + EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb); + } retry: - handle = ext4_journal_start(dir, EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + - EXT4_INDEX_EXTRA_TRANS_BLOCKS + 5 + - EXT4_MAXQUOTAS_INIT_BLOCKS(dir->i_sb)); + handle = ext4_journal_start(dir, credits); if (IS_ERR(handle)) return PTR_ERR(handle); @@ -2263,21 +2292,44 @@ retry: if (IS_ERR(inode)) goto out_stop; - if (l > sizeof(EXT4_I(inode)->i_data)) { + if (l > EXT4_N_BLOCKS * 4) { inode->i_op = &ext4_symlink_inode_operations; ext4_set_aops(inode); /* - * page_symlink() calls into ext4_prepare/commit_write. - * We have a transaction open. All is sweetness. It also sets - * i_size in generic_commit_write(). + * We cannot call page_symlink() with transaction started + * because it calls into ext4_write_begin() which can wait + * for transaction commit if we are running out of space + * and thus we deadlock. So we have to stop transaction now + * and restart it when symlink contents is written. + * + * To keep fs consistent in case of crash, we have to put inode + * to orphan list in the mean time. */ + drop_nlink(inode); + err = ext4_orphan_add(handle, inode); + ext4_journal_stop(handle); + if (err) + goto err_drop_inode; err = __page_symlink(inode, symname, l, 1); + if (err) + goto err_drop_inode; + /* + * Now inode is being linked into dir (EXT4_DATA_TRANS_BLOCKS + * + EXT4_INDEX_EXTRA_TRANS_BLOCKS), inode is also modified + */ + handle = ext4_journal_start(dir, + EXT4_DATA_TRANS_BLOCKS(dir->i_sb) + + EXT4_INDEX_EXTRA_TRANS_BLOCKS + 1); + if (IS_ERR(handle)) { + err = PTR_ERR(handle); + goto err_drop_inode; + } + inc_nlink(inode); + err = ext4_orphan_del(handle, inode); if (err) { + ext4_journal_stop(handle); clear_nlink(inode); - unlock_new_inode(inode); - ext4_mark_inode_dirty(handle, inode); - iput(inode); - goto out_stop; + goto err_drop_inode; } } else { /* clear the extent format for fast symlink */ @@ -2293,6 +2345,10 @@ out_stop: if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries)) goto retry; return err; +err_drop_inode: + unlock_new_inode(inode); + iput(inode); + return err; } static int ext4_link(struct dentry *old_dentry, diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index b6dbd056fcb1..7bb8f76d470a 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -203,46 +203,29 @@ static void ext4_end_bio(struct bio *bio, int error) for (i = 0; i < io_end->num_io_pages; i++) { struct page *page = io_end->pages[i]->p_page; struct buffer_head *bh, *head; - int partial_write = 0; + loff_t offset; + loff_t io_end_offset; - head = page_buffers(page); - if (error) + if (error) { SetPageError(page); - BUG_ON(!head); - if (head->b_size != PAGE_CACHE_SIZE) { - loff_t offset; - loff_t io_end_offset = io_end->offset + io_end->size; + set_bit(AS_EIO, &page->mapping->flags); + head = page_buffers(page); + BUG_ON(!head); + + io_end_offset = io_end->offset + io_end->size; offset = (sector_t) page->index << PAGE_CACHE_SHIFT; bh = head; do { if ((offset >= io_end->offset) && - (offset+bh->b_size <= io_end_offset)) { - if (error) - buffer_io_error(bh); - - } - if (buffer_delay(bh)) - partial_write = 1; - else if (!buffer_mapped(bh)) - clear_buffer_dirty(bh); - else if (buffer_dirty(bh)) - partial_write = 1; + (offset+bh->b_size <= io_end_offset)) + buffer_io_error(bh); + offset += bh->b_size; bh = bh->b_this_page; } while (bh != head); } - /* - * If this is a partial write which happened to make - * all buffers uptodate then we can optimize away a - * bogus readpage() for the next read(). Here we - * 'discover' whether the page went uptodate as a - * result of this (potentially partial) write. - */ - if (!partial_write) - SetPageUptodate(page); - put_io_page(io_end->pages[i]); } io_end->num_io_pages = 0; diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 22546ad7f0ae..cc5c157aa11d 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -38,6 +38,7 @@ #include <linux/ctype.h> #include <linux/log2.h> #include <linux/crc16.h> +#include <linux/cleancache.h> #include <asm/uaccess.h> #include <linux/kthread.h> @@ -75,11 +76,27 @@ static void ext4_write_super(struct super_block *sb); static int ext4_freeze(struct super_block *sb); static struct dentry *ext4_mount(struct file_system_type *fs_type, int flags, const char *dev_name, void *data); +static inline int ext2_feature_set_ok(struct super_block *sb); +static inline int ext3_feature_set_ok(struct super_block *sb); static int ext4_feature_set_ok(struct super_block *sb, int readonly); static void ext4_destroy_lazyinit_thread(void); static void ext4_unregister_li_request(struct super_block *sb); static void ext4_clear_request_list(void); +#if !defined(CONFIG_EXT2_FS) && !defined(CONFIG_EXT2_FS_MODULE) && defined(CONFIG_EXT4_USE_FOR_EXT23) +static struct file_system_type ext2_fs_type = { + .owner = THIS_MODULE, + .name = "ext2", + .mount = ext4_mount, + .kill_sb = kill_block_super, + .fs_flags = FS_REQUIRES_DEV, +}; +#define IS_EXT2_SB(sb) ((sb)->s_bdev->bd_holder == &ext2_fs_type) +#else +#define IS_EXT2_SB(sb) (0) +#endif + + #if !defined(CONFIG_EXT3_FS) && !defined(CONFIG_EXT3_FS_MODULE) && defined(CONFIG_EXT4_USE_FOR_EXT23) static struct file_system_type ext3_fs_type = { .owner = THIS_MODULE, @@ -242,27 +259,44 @@ static void ext4_put_nojournal(handle_t *handle) * journal_end calls result in the superblock being marked dirty, so * that sync() will call the filesystem's write_super callback if * appropriate. + * + * To avoid j_barrier hold in userspace when a user calls freeze(), + * ext4 prevents a new handle from being started by s_frozen, which + * is in an upper layer. */ handle_t *ext4_journal_start_sb(struct super_block *sb, int nblocks) { journal_t *journal; + handle_t *handle; if (sb->s_flags & MS_RDONLY) return ERR_PTR(-EROFS); - vfs_check_frozen(sb, SB_FREEZE_TRANS); - /* Special case here: if the journal has aborted behind our - * backs (eg. EIO in the commit thread), then we still need to - * take the FS itself readonly cleanly. */ journal = EXT4_SB(sb)->s_journal; - if (journal) { - if (is_journal_aborted(journal)) { - ext4_abort(sb, "Detected aborted journal"); - return ERR_PTR(-EROFS); - } - return jbd2_journal_start(journal, nblocks); + handle = ext4_journal_current_handle(); + + /* + * If a handle has been started, it should be allowed to + * finish, otherwise deadlock could happen between freeze + * and others(e.g. truncate) due to the restart of the + * journal handle if the filesystem is forzen and active + * handles are not stopped. + */ + if (!handle) + vfs_check_frozen(sb, SB_FREEZE_TRANS); + + if (!journal) + return ext4_get_nojournal(); + /* + * Special case here: if the journal has aborted behind our + * backs (eg. EIO in the commit thread), then we still need to + * take the FS itself readonly cleanly. + */ + if (is_journal_aborted(journal)) { + ext4_abort(sb, "Detected aborted journal"); + return ERR_PTR(-EROFS); } - return ext4_get_nojournal(); + return jbd2_journal_start(journal, nblocks); } /* @@ -617,7 +651,7 @@ __acquires(bitlock) * filesystem will have already been marked read/only and the * journal has been aborted. We return 1 as a hint to callers * who might what to use the return value from - * ext4_grp_locked_error() to distinguish beween the + * ext4_grp_locked_error() to distinguish between the * ERRORS_CONT and ERRORS_RO case, and perhaps return more * aggressively from the ext4 function in question, with a * more appropriate error code. @@ -789,6 +823,8 @@ static void ext4_put_super(struct super_block *sb) invalidate_bdev(sbi->journal_bdev); ext4_blkdev_remove(sbi); } + if (sbi->s_mmp_tsk) + kthread_stop(sbi->s_mmp_tsk); sb->s_fs_info = NULL; /* * Now that we are completely done shutting down the @@ -1079,7 +1115,7 @@ static int ext4_show_options(struct seq_file *seq, struct vfsmount *vfs) if (!test_opt(sb, INIT_INODE_TABLE)) seq_puts(seq, ",noinit_inode_table"); - else if (sbi->s_li_wait_mult) + else if (sbi->s_li_wait_mult != EXT4_DEF_LI_WAIT_MULT) seq_printf(seq, ",init_inode_table=%u", (unsigned) sbi->s_li_wait_mult); @@ -1170,9 +1206,7 @@ static ssize_t ext4_quota_write(struct super_block *sb, int type, const char *data, size_t len, loff_t off); static const struct dquot_operations ext4_quota_operations = { -#ifdef CONFIG_QUOTA .get_reserved_space = ext4_get_reserved_space, -#endif .write_dquot = ext4_write_dquot, .acquire_dquot = ext4_acquire_dquot, .release_dquot = ext4_release_dquot, @@ -1883,7 +1917,7 @@ static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es, ext4_msg(sb, KERN_WARNING, "warning: mounting fs with errors, " "running e2fsck is recommended"); - else if ((__s16) le16_to_cpu(es->s_max_mnt_count) >= 0 && + else if ((__s16) le16_to_cpu(es->s_max_mnt_count) > 0 && le16_to_cpu(es->s_mnt_count) >= (unsigned short) (__s16) le16_to_cpu(es->s_max_mnt_count)) ext4_msg(sb, KERN_WARNING, @@ -1915,6 +1949,7 @@ static int ext4_setup_super(struct super_block *sb, struct ext4_super_block *es, EXT4_INODES_PER_GROUP(sb), sbi->s_mount_opt, sbi->s_mount_opt2); + cleancache_init_fs(sb); return res; } @@ -2408,6 +2443,18 @@ static ssize_t lifetime_write_kbytes_show(struct ext4_attr *a, EXT4_SB(sb)->s_sectors_written_start) >> 1))); } +static ssize_t extent_cache_hits_show(struct ext4_attr *a, + struct ext4_sb_info *sbi, char *buf) +{ + return snprintf(buf, PAGE_SIZE, "%lu\n", sbi->extent_cache_hits); +} + +static ssize_t extent_cache_misses_show(struct ext4_attr *a, + struct ext4_sb_info *sbi, char *buf) +{ + return snprintf(buf, PAGE_SIZE, "%lu\n", sbi->extent_cache_misses); +} + static ssize_t inode_readahead_blks_store(struct ext4_attr *a, struct ext4_sb_info *sbi, const char *buf, size_t count) @@ -2465,6 +2512,8 @@ static struct ext4_attr ext4_attr_##name = __ATTR(name, mode, show, store) EXT4_RO_ATTR(delayed_allocation_blocks); EXT4_RO_ATTR(session_write_kbytes); EXT4_RO_ATTR(lifetime_write_kbytes); +EXT4_RO_ATTR(extent_cache_hits); +EXT4_RO_ATTR(extent_cache_misses); EXT4_ATTR_OFFSET(inode_readahead_blks, 0644, sbi_ui_show, inode_readahead_blks_store, s_inode_readahead_blks); EXT4_RW_ATTR_SBI_UI(inode_goal, s_inode_goal); @@ -2480,6 +2529,8 @@ static struct attribute *ext4_attrs[] = { ATTR_LIST(delayed_allocation_blocks), ATTR_LIST(session_write_kbytes), ATTR_LIST(lifetime_write_kbytes), + ATTR_LIST(extent_cache_hits), + ATTR_LIST(extent_cache_misses), ATTR_LIST(inode_readahead_blks), ATTR_LIST(inode_goal), ATTR_LIST(mb_stats), @@ -2642,12 +2693,6 @@ static void print_daily_error_info(unsigned long arg) mod_timer(&sbi->s_err_report, jiffies + 24*60*60*HZ); /* Once a day */ } -static void ext4_lazyinode_timeout(unsigned long data) -{ - struct task_struct *p = (struct task_struct *)data; - wake_up_process(p); -} - /* Find next suitable group and run ext4_init_inode_table */ static int ext4_run_li_request(struct ext4_li_request *elr) { @@ -2679,11 +2724,8 @@ static int ext4_run_li_request(struct ext4_li_request *elr) ret = ext4_init_inode_table(sb, group, elr->lr_timeout ? 0 : 1); if (elr->lr_timeout == 0) { - timeout = jiffies - timeout; - if (elr->lr_sbi->s_li_wait_mult) - timeout *= elr->lr_sbi->s_li_wait_mult; - else - timeout *= 20; + timeout = (jiffies - timeout) * + elr->lr_sbi->s_li_wait_mult; elr->lr_timeout = timeout; } elr->lr_next_sched = jiffies + elr->lr_timeout; @@ -2695,7 +2737,7 @@ static int ext4_run_li_request(struct ext4_li_request *elr) /* * Remove lr_request from the list_request and free the - * request tructure. Should be called with li_list_mtx held + * request structure. Should be called with li_list_mtx held */ static void ext4_remove_li_request(struct ext4_li_request *elr) { @@ -2713,14 +2755,16 @@ static void ext4_remove_li_request(struct ext4_li_request *elr) static void ext4_unregister_li_request(struct super_block *sb) { - struct ext4_li_request *elr = EXT4_SB(sb)->s_li_request; - - if (!ext4_li_info) + mutex_lock(&ext4_li_mtx); + if (!ext4_li_info) { + mutex_unlock(&ext4_li_mtx); return; + } mutex_lock(&ext4_li_info->li_list_mtx); - ext4_remove_li_request(elr); + ext4_remove_li_request(EXT4_SB(sb)->s_li_request); mutex_unlock(&ext4_li_info->li_list_mtx); + mutex_unlock(&ext4_li_mtx); } static struct task_struct *ext4_lazyinit_task; @@ -2739,17 +2783,10 @@ static int ext4_lazyinit_thread(void *arg) struct ext4_lazy_init *eli = (struct ext4_lazy_init *)arg; struct list_head *pos, *n; struct ext4_li_request *elr; - unsigned long next_wakeup; - DEFINE_WAIT(wait); + unsigned long next_wakeup, cur; BUG_ON(NULL == eli); - eli->li_timer.data = (unsigned long)current; - eli->li_timer.function = ext4_lazyinode_timeout; - - eli->li_task = current; - wake_up(&eli->li_wait_task); - cont_thread: while (true) { next_wakeup = MAX_JIFFY_OFFSET; @@ -2780,19 +2817,15 @@ cont_thread: if (freezing(current)) refrigerator(); - if ((time_after_eq(jiffies, next_wakeup)) || + cur = jiffies; + if ((time_after_eq(cur, next_wakeup)) || (MAX_JIFFY_OFFSET == next_wakeup)) { cond_resched(); continue; } - eli->li_timer.expires = next_wakeup; - add_timer(&eli->li_timer); - prepare_to_wait(&eli->li_wait_daemon, &wait, - TASK_INTERRUPTIBLE); - if (time_before(jiffies, next_wakeup)) - schedule(); - finish_wait(&eli->li_wait_daemon, &wait); + schedule_timeout_interruptible(next_wakeup - cur); + if (kthread_should_stop()) { ext4_clear_request_list(); goto exit_thread; @@ -2816,12 +2849,7 @@ exit_thread: goto cont_thread; } mutex_unlock(&eli->li_list_mtx); - del_timer_sync(&ext4_li_info->li_timer); - eli->li_task = NULL; - wake_up(&eli->li_wait_task); - kfree(ext4_li_info); - ext4_lazyinit_task = NULL; ext4_li_info = NULL; mutex_unlock(&ext4_li_mtx); @@ -2849,7 +2877,6 @@ static int ext4_run_lazyinit_thread(void) if (IS_ERR(ext4_lazyinit_task)) { int err = PTR_ERR(ext4_lazyinit_task); ext4_clear_request_list(); - del_timer_sync(&ext4_li_info->li_timer); kfree(ext4_li_info); ext4_li_info = NULL; printk(KERN_CRIT "EXT4: error %d creating inode table " @@ -2858,8 +2885,6 @@ static int ext4_run_lazyinit_thread(void) return err; } ext4_li_info->li_state |= EXT4_LAZYINIT_RUNNING; - - wait_event(ext4_li_info->li_wait_task, ext4_li_info->li_task != NULL); return 0; } @@ -2894,13 +2919,9 @@ static int ext4_li_info_new(void) if (!eli) return -ENOMEM; - eli->li_task = NULL; INIT_LIST_HEAD(&eli->li_request_list); mutex_init(&eli->li_list_mtx); - init_waitqueue_head(&eli->li_wait_daemon); - init_waitqueue_head(&eli->li_wait_task); - init_timer(&eli->li_timer); eli->li_state |= EXT4_LAZYINIT_QUIT; ext4_li_info = eli; @@ -2943,20 +2964,19 @@ static int ext4_register_li_request(struct super_block *sb, ext4_group_t ngroups = EXT4_SB(sb)->s_groups_count; int ret = 0; - if (sbi->s_li_request != NULL) + if (sbi->s_li_request != NULL) { + /* + * Reset timeout so it can be computed again, because + * s_li_wait_mult might have changed. + */ + sbi->s_li_request->lr_timeout = 0; return 0; + } if (first_not_zeroed == ngroups || (sb->s_flags & MS_RDONLY) || - !test_opt(sb, INIT_INODE_TABLE)) { - sbi->s_li_request = NULL; - return 0; - } - - if (first_not_zeroed == ngroups) { - sbi->s_li_request = NULL; + !test_opt(sb, INIT_INODE_TABLE)) return 0; - } elr = ext4_li_request_new(sb, first_not_zeroed); if (!elr) @@ -2975,6 +2995,12 @@ static int ext4_register_li_request(struct super_block *sb, mutex_unlock(&ext4_li_info->li_list_mtx); sbi->s_li_request = elr; + /* + * set elr to NULL here since it has been inserted to + * the request_list and the removal and free of it is + * handled by ext4_clear_request_list from now on. + */ + elr = NULL; if (!(ext4_li_info->li_state & EXT4_LAZYINIT_RUNNING)) { ret = ext4_run_lazyinit_thread(); @@ -3143,6 +3169,12 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) ((def_mount_opts & EXT4_DEFM_NODELALLOC) == 0)) set_opt(sb, DELALLOC); + /* + * set default s_li_wait_mult for lazyinit, for the case there is + * no mount option specified. + */ + sbi->s_li_wait_mult = EXT4_DEF_LI_WAIT_MULT; + if (!parse_options((char *) sbi->s_es->s_mount_opts, sb, &journal_devnum, &journal_ioprio, NULL, 0)) { ext4_msg(sb, KERN_WARNING, @@ -3164,6 +3196,28 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) "feature flags set on rev 0 fs, " "running e2fsck is recommended"); + if (IS_EXT2_SB(sb)) { + if (ext2_feature_set_ok(sb)) + ext4_msg(sb, KERN_INFO, "mounting ext2 file system " + "using the ext4 subsystem"); + else { + ext4_msg(sb, KERN_ERR, "couldn't mount as ext2 due " + "to feature incompatibilities"); + goto failed_mount; + } + } + + if (IS_EXT3_SB(sb)) { + if (ext3_feature_set_ok(sb)) + ext4_msg(sb, KERN_INFO, "mounting ext3 file system " + "using the ext4 subsystem"); + else { + ext4_msg(sb, KERN_ERR, "couldn't mount as ext3 due " + "to feature incompatibilities"); + goto failed_mount; + } + } + /* * Check feature flags regardless of the revision level, since we * previously didn't change the revision level when setting the flags, @@ -3385,6 +3439,10 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) get_random_bytes(&sbi->s_next_generation, sizeof(u32)); spin_lock_init(&sbi->s_next_gen_lock); + init_timer(&sbi->s_err_report); + sbi->s_err_report.function = print_daily_error_info; + sbi->s_err_report.data = (unsigned long) sb; + err = percpu_counter_init(&sbi->s_freeblocks_counter, ext4_count_free_blocks(sb)); if (!err) { @@ -3432,6 +3490,11 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER)); + if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_MMP) && + !(sb->s_flags & MS_RDONLY)) + if (ext4_multi_mount_protect(sb, le64_to_cpu(es->s_mmp_block))) + goto failed_mount3; + /* * The first inode we look at is the journal inode. Don't try * root first: it may be modified in the journal! @@ -3447,7 +3510,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) goto failed_mount_wq; } else { clear_opt(sb, DATA_FLAGS); - set_opt(sb, WRITEBACK_DATA); sbi->s_journal = NULL; needs_recovery = 0; goto no_journal; @@ -3646,9 +3708,6 @@ no_journal: "Opts: %s%s%s", descr, sbi->s_es->s_mount_opts, *sbi->s_es->s_mount_opts ? "; " : "", orig_data); - init_timer(&sbi->s_err_report); - sbi->s_err_report.function = print_daily_error_info; - sbi->s_err_report.data = (unsigned long) sb; if (es->s_error_count) mod_timer(&sbi->s_err_report, jiffies + 300*HZ); /* 5 minutes */ @@ -3672,6 +3731,7 @@ failed_mount_wq: sbi->s_journal = NULL; } failed_mount3: + del_timer(&sbi->s_err_report); if (sbi->s_flex_groups) { if (is_vmalloc_addr(sbi->s_flex_groups)) vfree(sbi->s_flex_groups); @@ -3682,6 +3742,8 @@ failed_mount3: percpu_counter_destroy(&sbi->s_freeinodes_counter); percpu_counter_destroy(&sbi->s_dirs_counter); percpu_counter_destroy(&sbi->s_dirtyblocks_counter); + if (sbi->s_mmp_tsk) + kthread_stop(sbi->s_mmp_tsk); failed_mount2: for (i = 0; i < db_count; i++) brelse(sbi->s_group_desc[i]); @@ -4138,6 +4200,11 @@ static int ext4_sync_fs(struct super_block *sb, int wait) /* * LVM calls this function before a (read-only) snapshot is created. This * gives us a chance to flush the journal completely and mark the fs clean. + * + * Note that only this function cannot bring a filesystem to be in a clean + * state independently, because ext4 prevents a new handle from being started + * by @sb->s_frozen, which stays in an upper layer. It thus needs help from + * the upper layer. */ static int ext4_freeze(struct super_block *sb) { @@ -4212,7 +4279,7 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) int enable_quota = 0; ext4_group_t g; unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO; - int err; + int err = 0; #ifdef CONFIG_QUOTA int i; #endif @@ -4338,6 +4405,13 @@ static int ext4_remount(struct super_block *sb, int *flags, char *data) goto restore_opts; if (!ext4_setup_super(sb, es, 0)) sb->s_flags &= ~MS_RDONLY; + if (EXT4_HAS_INCOMPAT_FEATURE(sb, + EXT4_FEATURE_INCOMPAT_MMP)) + if (ext4_multi_mount_protect(sb, + le64_to_cpu(es->s_mmp_block))) { + err = -EROFS; + goto restore_opts; + } enable_quota = 1; } } @@ -4402,6 +4476,7 @@ static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf) struct ext4_sb_info *sbi = EXT4_SB(sb); struct ext4_super_block *es = sbi->s_es; u64 fsid; + s64 bfree; if (test_opt(sb, MINIX_DF)) { sbi->s_overhead_last = 0; @@ -4445,8 +4520,10 @@ static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_type = EXT4_SUPER_MAGIC; buf->f_bsize = sb->s_blocksize; buf->f_blocks = ext4_blocks_count(es) - sbi->s_overhead_last; - buf->f_bfree = percpu_counter_sum_positive(&sbi->s_freeblocks_counter) - + bfree = percpu_counter_sum_positive(&sbi->s_freeblocks_counter) - percpu_counter_sum_positive(&sbi->s_dirtyblocks_counter); + /* prevent underflow in case that few free space is available */ + buf->f_bfree = max_t(s64, bfree, 0); buf->f_bavail = buf->f_bfree - ext4_r_blocks_count(es); if (buf->f_bfree < ext4_r_blocks_count(es)) buf->f_bavail = 0; @@ -4614,17 +4691,33 @@ static int ext4_quota_on(struct super_block *sb, int type, int format_id, static int ext4_quota_off(struct super_block *sb, int type) { + struct inode *inode = sb_dqopt(sb)->files[type]; + handle_t *handle; + /* Force all delayed allocation blocks to be allocated. * Caller already holds s_umount sem */ if (test_opt(sb, DELALLOC)) sync_filesystem(sb); + if (!inode) + goto out; + + /* Update modification times of quota files when userspace can + * start looking at them */ + handle = ext4_journal_start(inode, 1); + if (IS_ERR(handle)) + goto out; + inode->i_mtime = inode->i_ctime = CURRENT_TIME; + ext4_mark_inode_dirty(handle, inode); + ext4_journal_stop(handle); + +out: return dquot_quota_off(sb, type); } /* Read data from quotafile - avoid pagecache and such because we cannot afford * acquiring the locks... As quota files are never truncated and quota code - * itself serializes the operations (and noone else should touch the files) + * itself serializes the operations (and no one else should touch the files) * we don't have to be afraid of races */ static ssize_t ext4_quota_read(struct super_block *sb, int type, char *data, size_t len, loff_t off) @@ -4714,9 +4807,8 @@ out: if (inode->i_size < off + len) { i_size_write(inode, off + len); EXT4_I(inode)->i_disksize = inode->i_size; + ext4_mark_inode_dirty(handle, inode); } - inode->i_mtime = inode->i_ctime = CURRENT_TIME; - ext4_mark_inode_dirty(handle, inode); mutex_unlock(&inode->i_mutex); return len; } @@ -4730,14 +4822,6 @@ static struct dentry *ext4_mount(struct file_system_type *fs_type, int flags, } #if !defined(CONFIG_EXT2_FS) && !defined(CONFIG_EXT2_FS_MODULE) && defined(CONFIG_EXT4_USE_FOR_EXT23) -static struct file_system_type ext2_fs_type = { - .owner = THIS_MODULE, - .name = "ext2", - .mount = ext4_mount, - .kill_sb = kill_block_super, - .fs_flags = FS_REQUIRES_DEV, -}; - static inline void register_as_ext2(void) { int err = register_filesystem(&ext2_fs_type); @@ -4750,10 +4834,22 @@ static inline void unregister_as_ext2(void) { unregister_filesystem(&ext2_fs_type); } + +static inline int ext2_feature_set_ok(struct super_block *sb) +{ + if (EXT4_HAS_INCOMPAT_FEATURE(sb, ~EXT2_FEATURE_INCOMPAT_SUPP)) + return 0; + if (sb->s_flags & MS_RDONLY) + return 1; + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, ~EXT2_FEATURE_RO_COMPAT_SUPP)) + return 0; + return 1; +} MODULE_ALIAS("ext2"); #else static inline void register_as_ext2(void) { } static inline void unregister_as_ext2(void) { } +static inline int ext2_feature_set_ok(struct super_block *sb) { return 0; } #endif #if !defined(CONFIG_EXT3_FS) && !defined(CONFIG_EXT3_FS_MODULE) && defined(CONFIG_EXT4_USE_FOR_EXT23) @@ -4769,10 +4865,24 @@ static inline void unregister_as_ext3(void) { unregister_filesystem(&ext3_fs_type); } + +static inline int ext3_feature_set_ok(struct super_block *sb) +{ + if (EXT4_HAS_INCOMPAT_FEATURE(sb, ~EXT3_FEATURE_INCOMPAT_SUPP)) + return 0; + if (!EXT4_HAS_COMPAT_FEATURE(sb, EXT4_FEATURE_COMPAT_HAS_JOURNAL)) + return 0; + if (sb->s_flags & MS_RDONLY) + return 1; + if (EXT4_HAS_RO_COMPAT_FEATURE(sb, ~EXT3_FEATURE_RO_COMPAT_SUPP)) + return 0; + return 1; +} MODULE_ALIAS("ext3"); #else static inline void register_as_ext3(void) { } static inline void unregister_as_ext3(void) { } +static inline int ext3_feature_set_ok(struct super_block *sb) { return 0; } #endif static struct file_system_type ext4_fs_type = { @@ -4856,8 +4966,8 @@ static int __init ext4_init_fs(void) err = init_inodecache(); if (err) goto out1; - register_as_ext2(); register_as_ext3(); + register_as_ext2(); err = register_filesystem(&ext4_fs_type); if (err) goto out; diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index b545ca1c459c..c757adc97250 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -820,8 +820,8 @@ inserted: if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) goal = goal & EXT4_MAX_BLOCK_FILE_PHYS; - block = ext4_new_meta_blocks(handle, inode, - goal, NULL, &error); + block = ext4_new_meta_blocks(handle, inode, goal, 0, + NULL, &error); if (error) goto cleanup; diff --git a/fs/fat/cache.c b/fs/fat/cache.c index ae8200f84e39..1cc7038e273d 100644 --- a/fs/fat/cache.c +++ b/fs/fat/cache.c @@ -151,6 +151,13 @@ static void fat_cache_add(struct inode *inode, struct fat_cache_id *new) spin_unlock(&MSDOS_I(inode)->cache_lru_lock); tmp = fat_cache_alloc(inode); + if (!tmp) { + spin_lock(&MSDOS_I(inode)->cache_lru_lock); + MSDOS_I(inode)->nr_caches--; + spin_unlock(&MSDOS_I(inode)->cache_lru_lock); + return; + } + spin_lock(&MSDOS_I(inode)->cache_lru_lock); cache = fat_cache_merge(inode, new); if (cache != NULL) { diff --git a/fs/fat/dir.c b/fs/fat/dir.c index ee42b9e0b16a..4ad64732cbce 100644 --- a/fs/fat/dir.c +++ b/fs/fat/dir.c @@ -98,7 +98,7 @@ next: *bh = sb_bread(sb, phys); if (*bh == NULL) { - printk(KERN_ERR "FAT: Directory bread(block %llu) failed\n", + fat_msg(sb, KERN_ERR, "Directory bread(block %llu) failed", (llu)phys); /* skip this block */ *pos = (iblock + 1) << sb->s_blocksize_bits; @@ -136,9 +136,10 @@ static inline int fat_get_entry(struct inode *dir, loff_t *pos, * but ignore that right now. * Ahem... Stack smashing in ring 0 isn't fun. Fixed. */ -static int uni16_to_x8(unsigned char *ascii, const wchar_t *uni, int len, - int uni_xlate, struct nls_table *nls) +static int uni16_to_x8(struct super_block *sb, unsigned char *ascii, + const wchar_t *uni, int len, struct nls_table *nls) { + int uni_xlate = MSDOS_SB(sb)->options.unicode_xlate; const wchar_t *ip; wchar_t ec; unsigned char *op; @@ -166,23 +167,23 @@ static int uni16_to_x8(unsigned char *ascii, const wchar_t *uni, int len, } if (unlikely(*ip)) { - printk(KERN_WARNING "FAT: filename was truncated while " - "converting."); + fat_msg(sb, KERN_WARNING, "filename was truncated while " + "converting."); } *op = 0; return (op - ascii); } -static inline int fat_uni_to_x8(struct msdos_sb_info *sbi, const wchar_t *uni, +static inline int fat_uni_to_x8(struct super_block *sb, const wchar_t *uni, unsigned char *buf, int size) { + struct msdos_sb_info *sbi = MSDOS_SB(sb); if (sbi->options.utf8) return utf16s_to_utf8s(uni, FAT_MAX_UNI_CHARS, UTF16_HOST_ENDIAN, buf, size); else - return uni16_to_x8(buf, uni, size, sbi->options.unicode_xlate, - sbi->nls_io); + return uni16_to_x8(sb, buf, uni, size, sbi->nls_io); } static inline int @@ -419,7 +420,7 @@ parse_record: /* Compare shortname */ bufuname[last_u] = 0x0000; - len = fat_uni_to_x8(sbi, bufuname, bufname, sizeof(bufname)); + len = fat_uni_to_x8(sb, bufuname, bufname, sizeof(bufname)); if (fat_name_match(sbi, name, name_len, bufname, len)) goto found; @@ -428,7 +429,7 @@ parse_record: int size = PATH_MAX - FAT_MAX_UNI_SIZE; /* Compare longname */ - len = fat_uni_to_x8(sbi, unicode, longname, size); + len = fat_uni_to_x8(sb, unicode, longname, size); if (fat_name_match(sbi, name, name_len, longname, len)) goto found; } @@ -545,7 +546,7 @@ parse_record: if (nr_slots) { void *longname = unicode + FAT_MAX_UNI_CHARS; int size = PATH_MAX - FAT_MAX_UNI_SIZE; - int len = fat_uni_to_x8(sbi, unicode, longname, size); + int len = fat_uni_to_x8(sb, unicode, longname, size); fill_name = longname; fill_len = len; @@ -621,7 +622,7 @@ parse_record: if (isvfat) { bufuname[j] = 0x0000; - i = fat_uni_to_x8(sbi, bufuname, bufname, sizeof(bufname)); + i = fat_uni_to_x8(sb, bufuname, bufname, sizeof(bufname)); } if (nr_slots) { /* hack for fat_ioctl_filldir() */ @@ -979,6 +980,7 @@ static int __fat_remove_entries(struct inode *dir, loff_t pos, int nr_slots) int fat_remove_entries(struct inode *dir, struct fat_slot_info *sinfo) { + struct super_block *sb = dir->i_sb; struct msdos_dir_entry *de; struct buffer_head *bh; int err = 0, nr_slots; @@ -1013,8 +1015,8 @@ int fat_remove_entries(struct inode *dir, struct fat_slot_info *sinfo) */ err = __fat_remove_entries(dir, sinfo->slot_off, nr_slots); if (err) { - printk(KERN_WARNING - "FAT: Couldn't remove the long name slots\n"); + fat_msg(sb, KERN_WARNING, + "Couldn't remove the long name slots"); } } @@ -1265,7 +1267,7 @@ int fat_add_entries(struct inode *dir, void *slots, int nr_slots, if (sbi->fat_bits != 32) goto error; } else if (MSDOS_I(dir)->i_start == 0) { - printk(KERN_ERR "FAT: Corrupted directory (i_pos %lld)\n", + fat_msg(sb, KERN_ERR, "Corrupted directory (i_pos %lld)", MSDOS_I(dir)->i_pos); err = -EIO; goto error; diff --git a/fs/fat/fat.h b/fs/fat/fat.h index f50408901f7e..8276cc282dec 100644 --- a/fs/fat/fat.h +++ b/fs/fat/fat.h @@ -319,19 +319,20 @@ extern struct inode *fat_build_inode(struct super_block *sb, struct msdos_dir_entry *de, loff_t i_pos); extern int fat_sync_inode(struct inode *inode); extern int fat_fill_super(struct super_block *sb, void *data, int silent, - const struct inode_operations *fs_dir_inode_ops, - int isvfat, void (*setup)(struct super_block *)); + int isvfat, void (*setup)(struct super_block *)); extern int fat_flush_inodes(struct super_block *sb, struct inode *i1, struct inode *i2); /* fat/misc.c */ extern void -__fat_fs_error(struct super_block *s, int report, const char *fmt, ...) +__fat_fs_error(struct super_block *sb, int report, const char *fmt, ...) + __attribute__ ((format (printf, 3, 4))) __cold; +#define fat_fs_error(sb, fmt, args...) \ + __fat_fs_error(sb, 1, fmt , ## args) +#define fat_fs_error_ratelimit(sb, fmt, args...) \ + __fat_fs_error(sb, __ratelimit(&MSDOS_SB(sb)->ratelimit), fmt , ## args) +void fat_msg(struct super_block *sb, const char *level, const char *fmt, ...) __attribute__ ((format (printf, 3, 4))) __cold; -#define fat_fs_error(s, fmt, args...) \ - __fat_fs_error(s, 1, fmt , ## args) -#define fat_fs_error_ratelimit(s, fmt, args...) \ - __fat_fs_error(s, __ratelimit(&MSDOS_SB(s)->ratelimit), fmt , ## args) extern int fat_clusters_flush(struct super_block *sb); extern int fat_chain_add(struct inode *inode, int new_dclus, int nr_cluster); extern void fat_time_fat2unix(struct msdos_sb_info *sbi, struct timespec *ts, diff --git a/fs/fat/fatent.c b/fs/fat/fatent.c index b47d2c9f4fa1..2e81ac0df7e2 100644 --- a/fs/fat/fatent.c +++ b/fs/fat/fatent.c @@ -95,7 +95,7 @@ static int fat12_ent_bread(struct super_block *sb, struct fat_entry *fatent, err_brelse: brelse(bhs[0]); err: - printk(KERN_ERR "FAT: FAT read failed (blocknr %llu)\n", (llu)blocknr); + fat_msg(sb, KERN_ERR, "FAT read failed (blocknr %llu)", (llu)blocknr); return -EIO; } @@ -108,7 +108,7 @@ static int fat_ent_bread(struct super_block *sb, struct fat_entry *fatent, fatent->fat_inode = MSDOS_SB(sb)->fat_inode; fatent->bhs[0] = sb_bread(sb, blocknr); if (!fatent->bhs[0]) { - printk(KERN_ERR "FAT: FAT read failed (blocknr %llu)\n", + fat_msg(sb, KERN_ERR, "FAT read failed (blocknr %llu)", (llu)blocknr); return -EIO; } diff --git a/fs/fat/inode.c b/fs/fat/inode.c index 8d68690bdcf1..cb8d8391ac0b 100644 --- a/fs/fat/inode.c +++ b/fs/fat/inode.c @@ -581,7 +581,8 @@ static int fat_statfs(struct dentry *dentry, struct kstatfs *buf) buf->f_bavail = sbi->free_clusters; buf->f_fsid.val[0] = (u32)id; buf->f_fsid.val[1] = (u32)(id >> 32); - buf->f_namelen = sbi->options.isvfat ? FAT_LFN_LEN : 12; + buf->f_namelen = + (sbi->options.isvfat ? FAT_LFN_LEN : 12) * NLS_MAX_CHARSET_SIZE; return 0; } @@ -619,8 +620,8 @@ retry: bh = sb_bread(sb, i_pos >> sbi->dir_per_block_bits); if (!bh) { - printk(KERN_ERR "FAT: unable to read inode block " - "for updating (i_pos %lld)\n", i_pos); + fat_msg(sb, KERN_ERR, "unable to read inode block " + "for updating (i_pos %lld)", i_pos); return -EIO; } spin_lock(&sbi->inode_hash_lock); @@ -976,8 +977,8 @@ static const match_table_t vfat_tokens = { {Opt_err, NULL} }; -static int parse_options(char *options, int is_vfat, int silent, int *debug, - struct fat_mount_options *opts) +static int parse_options(struct super_block *sb, char *options, int is_vfat, + int silent, int *debug, struct fat_mount_options *opts) { char *p; substring_t args[MAX_OPT_ARGS]; @@ -1168,15 +1169,15 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, /* obsolete mount options */ case Opt_obsolate: - printk(KERN_INFO "FAT: \"%s\" option is obsolete, " - "not supported now\n", p); + fat_msg(sb, KERN_INFO, "\"%s\" option is obsolete, " + "not supported now", p); break; /* unknown option */ default: if (!silent) { - printk(KERN_ERR - "FAT: Unrecognized mount option \"%s\" " - "or missing value\n", p); + fat_msg(sb, KERN_ERR, + "Unrecognized mount option \"%s\" " + "or missing value", p); } return -EINVAL; } @@ -1185,7 +1186,7 @@ static int parse_options(char *options, int is_vfat, int silent, int *debug, out: /* UTF-8 doesn't provide FAT semantics */ if (!strcmp(opts->iocharset, "utf8")) { - printk(KERN_ERR "FAT: utf8 is not a recommended IO charset" + fat_msg(sb, KERN_ERR, "utf8 is not a recommended IO charset" " for FAT filesystems, filesystem will be " "case sensitive!\n"); } @@ -1238,8 +1239,7 @@ static int fat_read_root(struct inode *inode) /* * Read the super block of an MS-DOS FS. */ -int fat_fill_super(struct super_block *sb, void *data, int silent, - const struct inode_operations *fs_dir_inode_ops, int isvfat, +int fat_fill_super(struct super_block *sb, void *data, int silent, int isvfat, void (*setup)(struct super_block *)) { struct inode *root_inode = NULL, *fat_inode = NULL; @@ -1268,11 +1268,10 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, sb->s_magic = MSDOS_SUPER_MAGIC; sb->s_op = &fat_sops; sb->s_export_op = &fat_export_ops; - sbi->dir_ops = fs_dir_inode_ops; ratelimit_state_init(&sbi->ratelimit, DEFAULT_RATELIMIT_INTERVAL, DEFAULT_RATELIMIT_BURST); - error = parse_options(data, isvfat, silent, &debug, &sbi->options); + error = parse_options(sb, data, isvfat, silent, &debug, &sbi->options); if (error) goto out_fail; @@ -1282,20 +1281,20 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, sb_min_blocksize(sb, 512); bh = sb_bread(sb, 0); if (bh == NULL) { - printk(KERN_ERR "FAT: unable to read boot sector\n"); + fat_msg(sb, KERN_ERR, "unable to read boot sector"); goto out_fail; } b = (struct fat_boot_sector *) bh->b_data; if (!b->reserved) { if (!silent) - printk(KERN_ERR "FAT: bogus number of reserved sectors\n"); + fat_msg(sb, KERN_ERR, "bogus number of reserved sectors"); brelse(bh); goto out_invalid; } if (!b->fats) { if (!silent) - printk(KERN_ERR "FAT: bogus number of FAT structure\n"); + fat_msg(sb, KERN_ERR, "bogus number of FAT structure"); brelse(bh); goto out_invalid; } @@ -1308,7 +1307,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, media = b->media; if (!fat_valid_media(media)) { if (!silent) - printk(KERN_ERR "FAT: invalid media value (0x%02x)\n", + fat_msg(sb, KERN_ERR, "invalid media value (0x%02x)", media); brelse(bh); goto out_invalid; @@ -1318,7 +1317,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, || (logical_sector_size < 512) || (logical_sector_size > 4096)) { if (!silent) - printk(KERN_ERR "FAT: bogus logical sector size %u\n", + fat_msg(sb, KERN_ERR, "bogus logical sector size %u", logical_sector_size); brelse(bh); goto out_invalid; @@ -1326,15 +1325,15 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, sbi->sec_per_clus = b->sec_per_clus; if (!is_power_of_2(sbi->sec_per_clus)) { if (!silent) - printk(KERN_ERR "FAT: bogus sectors per cluster %u\n", + fat_msg(sb, KERN_ERR, "bogus sectors per cluster %u", sbi->sec_per_clus); brelse(bh); goto out_invalid; } if (logical_sector_size < sb->s_blocksize) { - printk(KERN_ERR "FAT: logical sector size too small for device" - " (logical sector size = %u)\n", logical_sector_size); + fat_msg(sb, KERN_ERR, "logical sector size too small for device" + " (logical sector size = %u)", logical_sector_size); brelse(bh); goto out_fail; } @@ -1342,14 +1341,14 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, brelse(bh); if (!sb_set_blocksize(sb, logical_sector_size)) { - printk(KERN_ERR "FAT: unable to set blocksize %u\n", + fat_msg(sb, KERN_ERR, "unable to set blocksize %u", logical_sector_size); goto out_fail; } bh = sb_bread(sb, 0); if (bh == NULL) { - printk(KERN_ERR "FAT: unable to read boot sector" - " (logical sector size = %lu)\n", + fat_msg(sb, KERN_ERR, "unable to read boot sector" + " (logical sector size = %lu)", sb->s_blocksize); goto out_fail; } @@ -1385,16 +1384,16 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, fsinfo_bh = sb_bread(sb, sbi->fsinfo_sector); if (fsinfo_bh == NULL) { - printk(KERN_ERR "FAT: bread failed, FSINFO block" - " (sector = %lu)\n", sbi->fsinfo_sector); + fat_msg(sb, KERN_ERR, "bread failed, FSINFO block" + " (sector = %lu)", sbi->fsinfo_sector); brelse(bh); goto out_fail; } fsinfo = (struct fat_boot_fsinfo *)fsinfo_bh->b_data; if (!IS_FSINFO(fsinfo)) { - printk(KERN_WARNING "FAT: Invalid FSINFO signature: " - "0x%08x, 0x%08x (sector = %lu)\n", + fat_msg(sb, KERN_WARNING, "Invalid FSINFO signature: " + "0x%08x, 0x%08x (sector = %lu)", le32_to_cpu(fsinfo->signature1), le32_to_cpu(fsinfo->signature2), sbi->fsinfo_sector); @@ -1415,8 +1414,8 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, sbi->dir_entries = get_unaligned_le16(&b->dir_entries); if (sbi->dir_entries & (sbi->dir_per_block - 1)) { if (!silent) - printk(KERN_ERR "FAT: bogus directroy-entries per block" - " (%u)\n", sbi->dir_entries); + fat_msg(sb, KERN_ERR, "bogus directroy-entries per block" + " (%u)", sbi->dir_entries); brelse(bh); goto out_invalid; } @@ -1438,7 +1437,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, total_clusters = min(total_clusters, fat_clusters - FAT_START_ENT); if (total_clusters > MAX_FAT(sb)) { if (!silent) - printk(KERN_ERR "FAT: count of clusters too big (%u)\n", + fat_msg(sb, KERN_ERR, "count of clusters too big (%u)", total_clusters); brelse(bh); goto out_invalid; @@ -1471,7 +1470,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, sprintf(buf, "cp%d", sbi->options.codepage); sbi->nls_disk = load_nls(buf); if (!sbi->nls_disk) { - printk(KERN_ERR "FAT: codepage %s not found\n", buf); + fat_msg(sb, KERN_ERR, "codepage %s not found", buf); goto out_fail; } @@ -1479,7 +1478,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, if (sbi->options.isvfat) { sbi->nls_io = load_nls(sbi->options.iocharset); if (!sbi->nls_io) { - printk(KERN_ERR "FAT: IO charset %s not found\n", + fat_msg(sb, KERN_ERR, "IO charset %s not found", sbi->options.iocharset); goto out_fail; } @@ -1503,7 +1502,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, insert_inode_hash(root_inode); sb->s_root = d_alloc_root(root_inode); if (!sb->s_root) { - printk(KERN_ERR "FAT: get root inode failed\n"); + fat_msg(sb, KERN_ERR, "get root inode failed"); goto out_fail; } @@ -1512,8 +1511,7 @@ int fat_fill_super(struct super_block *sb, void *data, int silent, out_invalid: error = -EINVAL; if (!silent) - printk(KERN_INFO "VFS: Can't find a valid FAT filesystem" - " on dev %s.\n", sb->s_id); + fat_msg(sb, KERN_INFO, "Can't find a valid FAT filesystem"); out_fail: if (fat_inode) diff --git a/fs/fat/misc.c b/fs/fat/misc.c index 970e682ea754..6d93360ca0cc 100644 --- a/fs/fat/misc.c +++ b/fs/fat/misc.c @@ -20,30 +20,46 @@ * In case the file system is remounted read-only, it can be made writable * again by remounting it. */ -void __fat_fs_error(struct super_block *s, int report, const char *fmt, ...) +void __fat_fs_error(struct super_block *sb, int report, const char *fmt, ...) { - struct fat_mount_options *opts = &MSDOS_SB(s)->options; + struct fat_mount_options *opts = &MSDOS_SB(sb)->options; va_list args; + struct va_format vaf; if (report) { - printk(KERN_ERR "FAT: Filesystem error (dev %s)\n", s->s_id); - - printk(KERN_ERR " "); va_start(args, fmt); - vprintk(fmt, args); + vaf.fmt = fmt; + vaf.va = &args; + printk(KERN_ERR "FAT-fs (%s): error, %pV\n", sb->s_id, &vaf); va_end(args); - printk("\n"); } if (opts->errors == FAT_ERRORS_PANIC) - panic("FAT: fs panic from previous error\n"); - else if (opts->errors == FAT_ERRORS_RO && !(s->s_flags & MS_RDONLY)) { - s->s_flags |= MS_RDONLY; - printk(KERN_ERR "FAT: Filesystem has been set read-only\n"); + panic("FAT-fs (%s): fs panic from previous error\n", sb->s_id); + else if (opts->errors == FAT_ERRORS_RO && !(sb->s_flags & MS_RDONLY)) { + sb->s_flags |= MS_RDONLY; + printk(KERN_ERR "FAT-fs (%s): Filesystem has been " + "set read-only\n", sb->s_id); } } EXPORT_SYMBOL_GPL(__fat_fs_error); +/** + * fat_msg() - print preformated FAT specific messages. Every thing what is + * not fat_fs_error() should be fat_msg(). + */ +void fat_msg(struct super_block *sb, const char *level, const char *fmt, ...) +{ + struct va_format vaf; + va_list args; + + va_start(args, fmt); + vaf.fmt = fmt; + vaf.va = &args; + printk("%sFAT-fs (%s): %pV\n", level, sb->s_id, &vaf); + va_end(args); +} + /* Flushes the number of free clusters on FAT32 */ /* XXX: Need to write one per FSINFO block. Currently only writes 1 */ int fat_clusters_flush(struct super_block *sb) @@ -57,15 +73,15 @@ int fat_clusters_flush(struct super_block *sb) bh = sb_bread(sb, sbi->fsinfo_sector); if (bh == NULL) { - printk(KERN_ERR "FAT: bread failed in fat_clusters_flush\n"); + fat_msg(sb, KERN_ERR, "bread failed in fat_clusters_flush"); return -EIO; } fsinfo = (struct fat_boot_fsinfo *)bh->b_data; /* Sanity check */ if (!IS_FSINFO(fsinfo)) { - printk(KERN_ERR "FAT: Invalid FSINFO signature: " - "0x%08x, 0x%08x (sector = %lu)\n", + fat_msg(sb, KERN_ERR, "Invalid FSINFO signature: " + "0x%08x, 0x%08x (sector = %lu)", le32_to_cpu(fsinfo->signature1), le32_to_cpu(fsinfo->signature2), sbi->fsinfo_sector); diff --git a/fs/fat/namei_msdos.c b/fs/fat/namei_msdos.c index 711499040eb6..3b222dafd15b 100644 --- a/fs/fat/namei_msdos.c +++ b/fs/fat/namei_msdos.c @@ -659,14 +659,14 @@ static const struct inode_operations msdos_dir_inode_operations = { static void setup(struct super_block *sb) { + MSDOS_SB(sb)->dir_ops = &msdos_dir_inode_operations; sb->s_d_op = &msdos_dentry_operations; sb->s_flags |= MS_NOATIME; } static int msdos_fill_super(struct super_block *sb, void *data, int silent) { - return fat_fill_super(sb, data, silent, &msdos_dir_inode_operations, - 0, setup); + return fat_fill_super(sb, data, silent, 0, setup); } static struct dentry *msdos_mount(struct file_system_type *fs_type, diff --git a/fs/fat/namei_vfat.c b/fs/fat/namei_vfat.c index adae3fb7451a..20b4ea53fdc4 100644 --- a/fs/fat/namei_vfat.c +++ b/fs/fat/namei_vfat.c @@ -1065,6 +1065,7 @@ static const struct inode_operations vfat_dir_inode_operations = { static void setup(struct super_block *sb) { + MSDOS_SB(sb)->dir_ops = &vfat_dir_inode_operations; if (MSDOS_SB(sb)->options.name_check != 's') sb->s_d_op = &vfat_ci_dentry_ops; else @@ -1073,8 +1074,7 @@ static void setup(struct super_block *sb) static int vfat_fill_super(struct super_block *sb, void *data, int silent) { - return fat_fill_super(sb, data, silent, &vfat_dir_inode_operations, - 1, setup); + return fat_fill_super(sb, data, silent, 1, setup); } static struct dentry *vfat_mount(struct file_system_type *fs_type, diff --git a/fs/fhandle.c b/fs/fhandle.c index bf93ad2bee07..6b088641f5bf 100644 --- a/fs/fhandle.c +++ b/fs/fhandle.c @@ -7,6 +7,7 @@ #include <linux/exportfs.h> #include <linux/fs_struct.h> #include <linux/fsnotify.h> +#include <linux/personality.h> #include <asm/uaccess.h> #include "internal.h" diff --git a/fs/file.c b/fs/file.c index 0be344755c02..4c6992d8f3ba 100644 --- a/fs/file.c +++ b/fs/file.c @@ -9,6 +9,7 @@ #include <linux/module.h> #include <linux/fs.h> #include <linux/mm.h> +#include <linux/mmzone.h> #include <linux/time.h> #include <linux/sched.h> #include <linux/slab.h> @@ -39,14 +40,17 @@ int sysctl_nr_open_max = 1024 * 1024; /* raised later */ */ static DEFINE_PER_CPU(struct fdtable_defer, fdtable_defer_list); -static inline void *alloc_fdmem(unsigned int size) +static void *alloc_fdmem(unsigned int size) { - void *data; - - data = kmalloc(size, GFP_KERNEL|__GFP_NOWARN); - if (data != NULL) - return data; - + /* + * Very large allocations can stress page reclaim, so fall back to + * vmalloc() if the allocation size will be considered "large" by the VM. + */ + if (size <= (PAGE_SIZE << PAGE_ALLOC_COSTLY_ORDER)) { + void *data = kmalloc(size, GFP_KERNEL|__GFP_NOWARN); + if (data != NULL) + return data; + } return vmalloc(size); } diff --git a/fs/filesystems.c b/fs/filesystems.c index 751d6b255a12..0845f84f2a5f 100644 --- a/fs/filesystems.c +++ b/fs/filesystems.c @@ -110,14 +110,13 @@ int unregister_filesystem(struct file_system_type * fs) *tmp = fs->next; fs->next = NULL; write_unlock(&file_systems_lock); + synchronize_rcu(); return 0; } tmp = &(*tmp)->next; } write_unlock(&file_systems_lock); - synchronize_rcu(); - return -EINVAL; } diff --git a/fs/freevxfs/vxfs_fshead.c b/fs/freevxfs/vxfs_fshead.c index 78948b4b1894..c9a6a94e58e9 100644 --- a/fs/freevxfs/vxfs_fshead.c +++ b/fs/freevxfs/vxfs_fshead.c @@ -164,7 +164,7 @@ vxfs_read_fshead(struct super_block *sbp) goto out_free_pfp; } if (!VXFS_ISILT(VXFS_INO(infp->vsi_stilist))) { - printk(KERN_ERR "vxfs: structual list inode is of wrong type (%x)\n", + printk(KERN_ERR "vxfs: structural list inode is of wrong type (%x)\n", VXFS_INO(infp->vsi_stilist)->vii_mode & VXFS_TYPE_MASK); goto out_iput_stilist; } diff --git a/fs/freevxfs/vxfs_inode.c b/fs/freevxfs/vxfs_inode.c index 2ba6719ac612..1a4311437a8b 100644 --- a/fs/freevxfs/vxfs_inode.c +++ b/fs/freevxfs/vxfs_inode.c @@ -272,7 +272,7 @@ vxfs_get_fake_inode(struct super_block *sbp, struct vxfs_inode_info *vip) * *ip: VFS inode * * Description: - * vxfs_put_fake_inode frees all data asssociated with @ip. + * vxfs_put_fake_inode frees all data associated with @ip. */ void vxfs_put_fake_inode(struct inode *ip) diff --git a/fs/freevxfs/vxfs_lookup.c b/fs/freevxfs/vxfs_lookup.c index 6c5131d592f0..3360f1e678ad 100644 --- a/fs/freevxfs/vxfs_lookup.c +++ b/fs/freevxfs/vxfs_lookup.c @@ -162,7 +162,7 @@ vxfs_find_entry(struct inode *ip, struct dentry *dp, struct page **ppp) /** * vxfs_inode_by_name - find inode number for dentry * @dip: directory to search in - * @dp: dentry we seach for + * @dp: dentry we search for * * Description: * vxfs_inode_by_name finds out the inode number of diff --git a/fs/freevxfs/vxfs_olt.h b/fs/freevxfs/vxfs_olt.h index d8324296486f..b7b3af502615 100644 --- a/fs/freevxfs/vxfs_olt.h +++ b/fs/freevxfs/vxfs_olt.h @@ -60,7 +60,7 @@ enum { * * The Object Location Table header is placed at the beginning of each * OLT extent. It is used to fing certain filesystem-wide metadata, e.g. - * the inital inode list, the fileset header or the device configuration. + * the initial inode list, the fileset header or the device configuration. */ struct vxfs_olt { u_int32_t olt_magic; /* magic number */ diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index b5ed541fb137..0f015a0468de 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -144,7 +144,7 @@ __bdi_start_writeback(struct backing_dev_info *bdi, long nr_pages, * * Description: * This does WB_SYNC_NONE opportunistic writeback. The IO is only - * started when this function returns, we make no guarentees on + * started when this function returns, we make no guarantees on * completion. Caller need not hold sb s_umount semaphore. * */ @@ -1007,9 +1007,6 @@ static noinline void block_dump___mark_inode_dirty(struct inode *inode) * In short, make sure you hash any inodes _before_ you start marking * them dirty. * - * This function *must* be atomic for the I_DIRTY_PAGES case - - * set_page_dirty() is called under spinlock in several places. - * * Note that for blockdevs, inode->dirtied_when represents the dirtying time of * the block-special inode (/dev/hda1) itself. And the ->dirtied_when field of * the kernel-internal blockdev inode represents the dirtying time of the @@ -1028,7 +1025,7 @@ void __mark_inode_dirty(struct inode *inode, int flags) */ if (flags & (I_DIRTY_SYNC | I_DIRTY_DATASYNC)) { if (sb->s_op->dirty_inode) - sb->s_op->dirty_inode(inode); + sb->s_op->dirty_inode(inode, flags); } /* diff --git a/fs/fscache/operation.c b/fs/fscache/operation.c index 48a18f184d50..30afdfa7aec7 100644 --- a/fs/fscache/operation.c +++ b/fs/fscache/operation.c @@ -33,8 +33,6 @@ void fscache_enqueue_operation(struct fscache_operation *op) _enter("{OBJ%x OP%x,%u}", op->object->debug_id, op->debug_id, atomic_read(&op->usage)); - fscache_set_op_state(op, "EnQ"); - ASSERT(list_empty(&op->pend_link)); ASSERT(op->processor != NULL); ASSERTCMP(op->object->state, >=, FSCACHE_OBJECT_AVAILABLE); @@ -66,8 +64,6 @@ EXPORT_SYMBOL(fscache_enqueue_operation); static void fscache_run_op(struct fscache_object *object, struct fscache_operation *op) { - fscache_set_op_state(op, "Run"); - object->n_in_progress++; if (test_and_clear_bit(FSCACHE_OP_WAITING, &op->flags)) wake_up_bit(&op->flags, FSCACHE_OP_WAITING); @@ -88,8 +84,6 @@ int fscache_submit_exclusive_op(struct fscache_object *object, _enter("{OBJ%x OP%x},", object->debug_id, op->debug_id); - fscache_set_op_state(op, "SubmitX"); - spin_lock(&object->lock); ASSERTCMP(object->n_ops, >=, object->n_in_progress); ASSERTCMP(object->n_ops, >=, object->n_exclusive); @@ -194,8 +188,6 @@ int fscache_submit_op(struct fscache_object *object, ASSERTCMP(atomic_read(&op->usage), >, 0); - fscache_set_op_state(op, "Submit"); - spin_lock(&object->lock); ASSERTCMP(object->n_ops, >=, object->n_in_progress); ASSERTCMP(object->n_ops, >=, object->n_exclusive); @@ -335,8 +327,6 @@ void fscache_put_operation(struct fscache_operation *op) if (!atomic_dec_and_test(&op->usage)) return; - fscache_set_op_state(op, "Put"); - _debug("PUT OP"); if (test_and_set_bit(FSCACHE_OP_DEAD, &op->flags)) BUG(); diff --git a/fs/fscache/page.c b/fs/fscache/page.c index 41c441c2058d..a2a5d19ece6a 100644 --- a/fs/fscache/page.c +++ b/fs/fscache/page.c @@ -155,11 +155,9 @@ static void fscache_attr_changed_op(struct fscache_operation *op) fscache_stat(&fscache_n_attr_changed_calls); if (fscache_object_is_active(object)) { - fscache_set_op_state(op, "CallFS"); fscache_stat(&fscache_n_cop_attr_changed); ret = object->cache->ops->attr_changed(object); fscache_stat_d(&fscache_n_cop_attr_changed); - fscache_set_op_state(op, "Done"); if (ret < 0) fscache_abort_object(object); } @@ -190,7 +188,6 @@ int __fscache_attr_changed(struct fscache_cookie *cookie) fscache_operation_init(op, fscache_attr_changed_op, NULL); op->flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_EXCLUSIVE); - fscache_set_op_name(op, "Attr"); spin_lock(&cookie->lock); @@ -257,7 +254,6 @@ static struct fscache_retrieval *fscache_alloc_retrieval( op->context = context; op->start_time = jiffies; INIT_LIST_HEAD(&op->to_do); - fscache_set_op_name(&op->op, "Retr"); return op; } @@ -368,7 +364,6 @@ int __fscache_read_or_alloc_page(struct fscache_cookie *cookie, _leave(" = -ENOMEM"); return -ENOMEM; } - fscache_set_op_name(&op->op, "RetrRA1"); spin_lock(&cookie->lock); @@ -487,7 +482,6 @@ int __fscache_read_or_alloc_pages(struct fscache_cookie *cookie, op = fscache_alloc_retrieval(mapping, end_io_func, context); if (!op) return -ENOMEM; - fscache_set_op_name(&op->op, "RetrRAN"); spin_lock(&cookie->lock); @@ -589,7 +583,6 @@ int __fscache_alloc_page(struct fscache_cookie *cookie, op = fscache_alloc_retrieval(page->mapping, NULL, NULL); if (!op) return -ENOMEM; - fscache_set_op_name(&op->op, "RetrAL1"); spin_lock(&cookie->lock); @@ -662,8 +655,6 @@ static void fscache_write_op(struct fscache_operation *_op) _enter("{OP%x,%d}", op->op.debug_id, atomic_read(&op->op.usage)); - fscache_set_op_state(&op->op, "GetPage"); - spin_lock(&object->lock); cookie = object->cookie; @@ -698,15 +689,12 @@ static void fscache_write_op(struct fscache_operation *_op) spin_unlock(&cookie->stores_lock); spin_unlock(&object->lock); - fscache_set_op_state(&op->op, "Store"); fscache_stat(&fscache_n_store_pages); fscache_stat(&fscache_n_cop_write_page); ret = object->cache->ops->write_page(op, page); fscache_stat_d(&fscache_n_cop_write_page); - fscache_set_op_state(&op->op, "EndWrite"); fscache_end_page_write(object, page); if (ret < 0) { - fscache_set_op_state(&op->op, "Abort"); fscache_abort_object(object); } else { fscache_enqueue_operation(&op->op); @@ -778,7 +766,6 @@ int __fscache_write_page(struct fscache_cookie *cookie, fscache_operation_init(&op->op, fscache_write_op, fscache_release_write_op); op->op.flags = FSCACHE_OP_ASYNC | (1 << FSCACHE_OP_WAITING); - fscache_set_op_name(&op->op, "Write1"); ret = radix_tree_preload(gfp & ~__GFP_HIGHMEM); if (ret < 0) diff --git a/fs/fuse/dir.c b/fs/fuse/dir.c index c6ba49bd95b3..d50160714595 100644 --- a/fs/fuse/dir.c +++ b/fs/fuse/dir.c @@ -174,7 +174,7 @@ static int fuse_dentry_revalidate(struct dentry *entry, struct nameidata *nd) if (!inode) return 0; - if (nd->flags & LOOKUP_RCU) + if (nd && (nd->flags & LOOKUP_RCU)) return -ECHILD; fc = get_fuse_conn(inode); @@ -691,6 +691,7 @@ static int fuse_rename(struct inode *olddir, struct dentry *oldent, struct fuse_rename_in inarg; struct fuse_conn *fc = get_fuse_conn(olddir); struct fuse_req *req = fuse_get_req(fc); + if (IS_ERR(req)) return PTR_ERR(req); diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 6ea00734984e..82a66466a24c 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -523,7 +523,7 @@ static int fuse_readpage(struct file *file, struct page *page) goto out; /* - * Page writeback can extend beyond the liftime of the + * Page writeback can extend beyond the lifetime of the * page-cache page, so make sure we read a properly synced * page. */ diff --git a/fs/gfs2/Makefile b/fs/gfs2/Makefile index f3d23ef4e876..86128202384f 100644 --- a/fs/gfs2/Makefile +++ b/fs/gfs2/Makefile @@ -1,9 +1,9 @@ ccflags-y := -I$(src) obj-$(CONFIG_GFS2_FS) += gfs2.o gfs2-y := acl.o bmap.o dir.o xattr.o glock.o \ - glops.o inode.o log.o lops.o main.o meta_io.o \ + glops.o log.o lops.o main.o meta_io.o \ aops.o dentry.o export.o file.o \ - ops_fstype.o ops_inode.o quota.o \ + ops_fstype.o inode.o quota.o \ recovery.o rgrp.o super.o sys.o trans.o util.o gfs2-$(CONFIG_GFS2_FS_LOCKING_DLM) += lock_dlm.o diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index c71995b111bf..802ac5eeba28 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -884,8 +884,8 @@ static int gfs2_write_end(struct file *file, struct address_space *mapping, } brelse(dibh); - gfs2_trans_end(sdp); failed: + gfs2_trans_end(sdp); if (al) { gfs2_inplace_release(ip); gfs2_quota_unlock(ip); @@ -1076,8 +1076,8 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask) bd = bh->b_private; if (bd && bd->bd_ail) goto cannot_release; - gfs2_assert_warn(sdp, !buffer_pinned(bh)); - gfs2_assert_warn(sdp, !buffer_dirty(bh)); + if (buffer_pinned(bh) || buffer_dirty(bh)) + goto not_possible; bh = bh->b_this_page; } while(bh != head); gfs2_log_unlock(sdp); @@ -1107,6 +1107,10 @@ int gfs2_releasepage(struct page *page, gfp_t gfp_mask) } while (bh != head); return try_to_free_buffers(page); + +not_possible: /* Should never happen */ + WARN_ON(buffer_dirty(bh)); + WARN_ON(buffer_pinned(bh)); cannot_release: gfs2_log_unlock(sdp); return 0; diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index ef3dc4b9fae2..e65493a8ac00 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -780,6 +780,8 @@ static int do_strip(struct gfs2_inode *ip, struct buffer_head *dibh, metadata = (height != ip->i_height - 1); if (metadata) revokes = (height) ? sdp->sd_inptrs : sdp->sd_diptrs; + else if (ip->i_depth) + revokes = sdp->sd_inptrs; if (ip != GFS2_I(sdp->sd_rindex)) error = gfs2_rindex_hold(sdp, &ip->i_alloc->al_ri_gh); @@ -1136,7 +1138,7 @@ void gfs2_trim_blocks(struct inode *inode) * earlier versions of GFS2 have a bug in the stuffed file reading * code which will result in a buffer overrun if the size is larger * than the max stuffed file size. In order to prevent this from - * occuring, such files are unstuffed, but in other cases we can + * occurring, such files are unstuffed, but in other cases we can * just update the inode size directly. * * Returns: 0 on success, or -ve on error diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 5c356d09c321..091ee4779538 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -82,12 +82,9 @@ struct qstr gfs2_qdot __read_mostly; struct qstr gfs2_qdotdot __read_mostly; -typedef int (*leaf_call_t) (struct gfs2_inode *dip, u32 index, u32 len, - u64 leaf_no, void *data); typedef int (*gfs2_dscan_t)(const struct gfs2_dirent *dent, const struct qstr *name, void *opaque); - int gfs2_dir_get_new_buffer(struct gfs2_inode *ip, u64 block, struct buffer_head **bhp) { @@ -1506,7 +1503,7 @@ struct inode *gfs2_dir_search(struct inode *dir, const struct qstr *name) inode = gfs2_inode_lookup(dir->i_sb, be16_to_cpu(dent->de_type), be64_to_cpu(dent->de_inum.no_addr), - be64_to_cpu(dent->de_inum.no_formal_ino)); + be64_to_cpu(dent->de_inum.no_formal_ino), 0); brelse(bh); return inode; } @@ -1600,7 +1597,7 @@ static int dir_new_leaf(struct inode *inode, const struct qstr *name) */ int gfs2_dir_add(struct inode *inode, const struct qstr *name, - const struct gfs2_inode *nip, unsigned type) + const struct gfs2_inode *nip) { struct gfs2_inode *ip = GFS2_I(inode); struct buffer_head *bh; @@ -1616,7 +1613,7 @@ int gfs2_dir_add(struct inode *inode, const struct qstr *name, return PTR_ERR(dent); dent = gfs2_init_dirent(inode, dent, name, bh); gfs2_inum_out(nip, dent); - dent->de_type = cpu_to_be16(type); + dent->de_type = cpu_to_be16(IF2DT(nip->i_inode.i_mode)); if (ip->i_diskflags & GFS2_DIF_EXHASH) { leaf = (struct gfs2_leaf *)bh->b_data; be16_add_cpu(&leaf->lf_entries, 1); @@ -1628,6 +1625,8 @@ int gfs2_dir_add(struct inode *inode, const struct qstr *name, gfs2_trans_add_bh(ip->i_gl, bh, 1); ip->i_entries++; ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; + if (S_ISDIR(nip->i_inode.i_mode)) + inc_nlink(&ip->i_inode); gfs2_dinode_out(ip, bh->b_data); brelse(bh); error = 0; @@ -1672,8 +1671,9 @@ int gfs2_dir_add(struct inode *inode, const struct qstr *name, * Returns: 0 on success, error code on failure */ -int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *name) +int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry) { + const struct qstr *name = &dentry->d_name; struct gfs2_dirent *dent, *prev = NULL; struct buffer_head *bh; int error; @@ -1714,6 +1714,8 @@ int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *name) gfs2_trans_add_bh(dip->i_gl, bh, 1); dip->i_entries--; dip->i_inode.i_mtime = dip->i_inode.i_ctime = CURRENT_TIME; + if (S_ISDIR(dentry->d_inode->i_mode)) + drop_nlink(&dip->i_inode); gfs2_dinode_out(dip, bh->b_data); brelse(bh); mark_inode_dirty(&dip->i_inode); @@ -1768,94 +1770,20 @@ int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, } /** - * foreach_leaf - call a function for each leaf in a directory - * @dip: the directory - * @lc: the function to call for each each - * @data: private data to pass to it - * - * Returns: errno - */ - -static int foreach_leaf(struct gfs2_inode *dip, leaf_call_t lc, void *data) -{ - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - struct buffer_head *bh; - struct gfs2_leaf *leaf; - u32 hsize, len; - u32 ht_offset, lp_offset, ht_offset_cur = -1; - u32 index = 0; - __be64 *lp; - u64 leaf_no; - int error = 0; - - hsize = 1 << dip->i_depth; - if (hsize * sizeof(u64) != i_size_read(&dip->i_inode)) { - gfs2_consist_inode(dip); - return -EIO; - } - - lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS); - if (!lp) - return -ENOMEM; - - while (index < hsize) { - lp_offset = index & (sdp->sd_hash_ptrs - 1); - ht_offset = index - lp_offset; - - if (ht_offset_cur != ht_offset) { - error = gfs2_dir_read_data(dip, (char *)lp, - ht_offset * sizeof(__be64), - sdp->sd_hash_bsize, 1); - if (error != sdp->sd_hash_bsize) { - if (error >= 0) - error = -EIO; - goto out; - } - ht_offset_cur = ht_offset; - } - - leaf_no = be64_to_cpu(lp[lp_offset]); - if (leaf_no) { - error = get_leaf(dip, leaf_no, &bh); - if (error) - goto out; - leaf = (struct gfs2_leaf *)bh->b_data; - len = 1 << (dip->i_depth - be16_to_cpu(leaf->lf_depth)); - brelse(bh); - - error = lc(dip, index, len, leaf_no, data); - if (error) - goto out; - - index = (index & ~(len - 1)) + len; - } else - index++; - } - - if (index != hsize) { - gfs2_consist_inode(dip); - error = -EIO; - } - -out: - kfree(lp); - - return error; -} - -/** * leaf_dealloc - Deallocate a directory leaf * @dip: the directory * @index: the hash table offset in the directory * @len: the number of pointers to this leaf * @leaf_no: the leaf number - * @data: not used + * @leaf_bh: buffer_head for the starting leaf + * last_dealloc: 1 if this is the final dealloc for the leaf, else 0 * * Returns: errno */ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, - u64 leaf_no, void *data) + u64 leaf_no, struct buffer_head *leaf_bh, + int last_dealloc) { struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct gfs2_leaf *tmp_leaf; @@ -1887,14 +1815,18 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, goto out_qs; /* Count the number of leaves */ + bh = leaf_bh; for (blk = leaf_no; blk; blk = nblk) { - error = get_leaf(dip, blk, &bh); - if (error) - goto out_rlist; + if (blk != leaf_no) { + error = get_leaf(dip, blk, &bh); + if (error) + goto out_rlist; + } tmp_leaf = (struct gfs2_leaf *)bh->b_data; nblk = be64_to_cpu(tmp_leaf->lf_next); - brelse(bh); + if (blk != leaf_no) + brelse(bh); gfs2_rlist_add(sdp, &rlist, blk); l_blocks++; @@ -1918,13 +1850,18 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, if (error) goto out_rg_gunlock; + bh = leaf_bh; + for (blk = leaf_no; blk; blk = nblk) { - error = get_leaf(dip, blk, &bh); - if (error) - goto out_end_trans; + if (blk != leaf_no) { + error = get_leaf(dip, blk, &bh); + if (error) + goto out_end_trans; + } tmp_leaf = (struct gfs2_leaf *)bh->b_data; nblk = be64_to_cpu(tmp_leaf->lf_next); - brelse(bh); + if (blk != leaf_no) + brelse(bh); gfs2_free_meta(dip, blk, 1); gfs2_add_inode_blocks(&dip->i_inode, -1); @@ -1942,6 +1879,10 @@ static int leaf_dealloc(struct gfs2_inode *dip, u32 index, u32 len, goto out_end_trans; gfs2_trans_add_bh(dip->i_gl, dibh, 1); + /* On the last dealloc, make this a regular file in case we crash. + (We don't want to free these blocks a second time.) */ + if (last_dealloc) + dip->i_inode.i_mode = S_IFREG; gfs2_dinode_out(dip, dibh->b_data); brelse(dibh); @@ -1975,29 +1916,67 @@ int gfs2_dir_exhash_dealloc(struct gfs2_inode *dip) { struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct buffer_head *bh; - int error; + struct gfs2_leaf *leaf; + u32 hsize, len; + u32 ht_offset, lp_offset, ht_offset_cur = -1; + u32 index = 0, next_index; + __be64 *lp; + u64 leaf_no; + int error = 0, last; - /* Dealloc on-disk leaves to FREEMETA state */ - error = foreach_leaf(dip, leaf_dealloc, NULL); - if (error) - return error; + hsize = 1 << dip->i_depth; + if (hsize * sizeof(u64) != i_size_read(&dip->i_inode)) { + gfs2_consist_inode(dip); + return -EIO; + } - /* Make this a regular file in case we crash. - (We don't want to free these blocks a second time.) */ + lp = kmalloc(sdp->sd_hash_bsize, GFP_NOFS); + if (!lp) + return -ENOMEM; - error = gfs2_trans_begin(sdp, RES_DINODE, 0); - if (error) - return error; + while (index < hsize) { + lp_offset = index & (sdp->sd_hash_ptrs - 1); + ht_offset = index - lp_offset; - error = gfs2_meta_inode_buffer(dip, &bh); - if (!error) { - gfs2_trans_add_bh(dip->i_gl, bh, 1); - ((struct gfs2_dinode *)bh->b_data)->di_mode = - cpu_to_be32(S_IFREG); - brelse(bh); + if (ht_offset_cur != ht_offset) { + error = gfs2_dir_read_data(dip, (char *)lp, + ht_offset * sizeof(__be64), + sdp->sd_hash_bsize, 1); + if (error != sdp->sd_hash_bsize) { + if (error >= 0) + error = -EIO; + goto out; + } + ht_offset_cur = ht_offset; + } + + leaf_no = be64_to_cpu(lp[lp_offset]); + if (leaf_no) { + error = get_leaf(dip, leaf_no, &bh); + if (error) + goto out; + leaf = (struct gfs2_leaf *)bh->b_data; + len = 1 << (dip->i_depth - be16_to_cpu(leaf->lf_depth)); + + next_index = (index & ~(len - 1)) + len; + last = ((next_index >= hsize) ? 1 : 0); + error = leaf_dealloc(dip, index, len, leaf_no, bh, + last); + brelse(bh); + if (error) + goto out; + index = next_index; + } else + index++; } - gfs2_trans_end(sdp); + if (index != hsize) { + gfs2_consist_inode(dip); + error = -EIO; + } + +out: + kfree(lp); return error; } diff --git a/fs/gfs2/dir.h b/fs/gfs2/dir.h index a98f644bd3df..e686af11becd 100644 --- a/fs/gfs2/dir.h +++ b/fs/gfs2/dir.h @@ -22,8 +22,8 @@ extern struct inode *gfs2_dir_search(struct inode *dir, extern int gfs2_dir_check(struct inode *dir, const struct qstr *filename, const struct gfs2_inode *ip); extern int gfs2_dir_add(struct inode *inode, const struct qstr *filename, - const struct gfs2_inode *ip, unsigned int type); -extern int gfs2_dir_del(struct gfs2_inode *dip, const struct qstr *filename); + const struct gfs2_inode *ip); +extern int gfs2_dir_del(struct gfs2_inode *dip, const struct dentry *dentry); extern int gfs2_dir_read(struct inode *inode, u64 *offset, void *opaque, filldir_t filldir); extern int gfs2_dir_mvino(struct gfs2_inode *dip, const struct qstr *filename, diff --git a/fs/gfs2/export.c b/fs/gfs2/export.c index b5a5e60df0d5..fe9945f2ff72 100644 --- a/fs/gfs2/export.c +++ b/fs/gfs2/export.c @@ -139,7 +139,7 @@ static struct dentry *gfs2_get_dentry(struct super_block *sb, struct gfs2_sbd *sdp = sb->s_fs_info; struct inode *inode; - inode = gfs2_ilookup(sb, inum->no_addr); + inode = gfs2_ilookup(sb, inum->no_addr, 0); if (inode) { if (GFS2_I(inode)->i_no_formal_ino != inum->no_formal_ino) { iput(inode); diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index b2682e073eee..a9f5cbe45cd9 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -545,18 +545,10 @@ static int gfs2_close(struct inode *inode, struct file *file) /** * gfs2_fsync - sync the dirty data for a file (across the cluster) * @file: the file that points to the dentry (we ignore this) - * @dentry: the dentry that points to the inode to sync + * @datasync: set if we can ignore timestamp changes * - * The VFS will flush "normal" data for us. We only need to worry - * about metadata here. For journaled data, we just do a log flush - * as we can't avoid it. Otherwise we can just bale out if datasync - * is set. For stuffed inodes we must flush the log in order to - * ensure that all data is on disk. - * - * The call to write_inode_now() is there to write back metadata and - * the inode itself. It does also try and write the data, but thats - * (hopefully) a no-op due to the VFS having already called filemap_fdatawrite() - * for us. + * The VFS will flush data for us. We only need to worry + * about metadata here. * * Returns: errno */ @@ -565,22 +557,20 @@ static int gfs2_fsync(struct file *file, int datasync) { struct inode *inode = file->f_mapping->host; int sync_state = inode->i_state & (I_DIRTY_SYNC|I_DIRTY_DATASYNC); - int ret = 0; - - if (gfs2_is_jdata(GFS2_I(inode))) { - gfs2_log_flush(GFS2_SB(inode), GFS2_I(inode)->i_gl); - return 0; - } + struct gfs2_inode *ip = GFS2_I(inode); + int ret; - if (sync_state != 0) { - if (!datasync) - ret = write_inode_now(inode, 0); + if (datasync) + sync_state &= ~I_DIRTY_SYNC; - if (gfs2_is_stuffed(GFS2_I(inode))) - gfs2_log_flush(GFS2_SB(inode), GFS2_I(inode)->i_gl); + if (sync_state) { + ret = sync_inode_metadata(inode, 1); + if (ret) + return ret; + gfs2_ail_flush(ip->i_gl); } - return ret; + return 0; } /** @@ -617,18 +607,51 @@ static ssize_t gfs2_file_aio_write(struct kiocb *iocb, const struct iovec *iov, return generic_file_aio_write(iocb, iov, nr_segs, pos); } -static void empty_write_end(struct page *page, unsigned from, - unsigned to) +static int empty_write_end(struct page *page, unsigned from, + unsigned to, int mode) { - struct gfs2_inode *ip = GFS2_I(page->mapping->host); + struct inode *inode = page->mapping->host; + struct gfs2_inode *ip = GFS2_I(inode); + struct buffer_head *bh; + unsigned offset, blksize = 1 << inode->i_blkbits; + pgoff_t end_index = i_size_read(inode) >> PAGE_CACHE_SHIFT; zero_user(page, from, to-from); mark_page_accessed(page); - if (!gfs2_is_writeback(ip)) - gfs2_page_add_databufs(ip, page, from, to); + if (page->index < end_index || !(mode & FALLOC_FL_KEEP_SIZE)) { + if (!gfs2_is_writeback(ip)) + gfs2_page_add_databufs(ip, page, from, to); - block_commit_write(page, from, to); + block_commit_write(page, from, to); + return 0; + } + + offset = 0; + bh = page_buffers(page); + while (offset < to) { + if (offset >= from) { + set_buffer_uptodate(bh); + mark_buffer_dirty(bh); + clear_buffer_new(bh); + write_dirty_buffer(bh, WRITE); + } + offset += blksize; + bh = bh->b_this_page; + } + + offset = 0; + bh = page_buffers(page); + while (offset < to) { + if (offset >= from) { + wait_on_buffer(bh); + if (!buffer_uptodate(bh)) + return -EIO; + } + offset += blksize; + bh = bh->b_this_page; + } + return 0; } static int needs_empty_write(sector_t block, struct inode *inode) @@ -643,7 +666,8 @@ static int needs_empty_write(sector_t block, struct inode *inode) return !buffer_mapped(&bh_map); } -static int write_empty_blocks(struct page *page, unsigned from, unsigned to) +static int write_empty_blocks(struct page *page, unsigned from, unsigned to, + int mode) { struct inode *inode = page->mapping->host; unsigned start, end, next, blksize; @@ -668,7 +692,9 @@ static int write_empty_blocks(struct page *page, unsigned from, unsigned to) gfs2_block_map); if (unlikely(ret)) return ret; - empty_write_end(page, start, end); + ret = empty_write_end(page, start, end, mode); + if (unlikely(ret)) + return ret; end = 0; } start = next; @@ -682,7 +708,9 @@ static int write_empty_blocks(struct page *page, unsigned from, unsigned to) ret = __block_write_begin(page, start, end - start, gfs2_block_map); if (unlikely(ret)) return ret; - empty_write_end(page, start, end); + ret = empty_write_end(page, start, end, mode); + if (unlikely(ret)) + return ret; } return 0; @@ -731,7 +759,7 @@ static int fallocate_chunk(struct inode *inode, loff_t offset, loff_t len, if (curr == end) to = end_offset; - error = write_empty_blocks(page, from, to); + error = write_empty_blocks(page, from, to, mode); if (!error && offset + to > inode->i_size && !(mode & FALLOC_FL_KEEP_SIZE)) { i_size_write(inode, offset + to); @@ -788,6 +816,7 @@ static long gfs2_fallocate(struct file *file, int mode, loff_t offset, loff_t bytes, max_bytes; struct gfs2_alloc *al; int error; + loff_t bsize_mask = ~((loff_t)sdp->sd_sb.sb_bsize - 1); loff_t next = (offset + len - 1) >> sdp->sd_sb.sb_bsize_shift; next = (next + 1) << sdp->sd_sb.sb_bsize_shift; @@ -795,13 +824,15 @@ static long gfs2_fallocate(struct file *file, int mode, loff_t offset, if (mode & ~FALLOC_FL_KEEP_SIZE) return -EOPNOTSUPP; - offset = (offset >> sdp->sd_sb.sb_bsize_shift) << - sdp->sd_sb.sb_bsize_shift; + offset &= bsize_mask; len = next - offset; bytes = sdp->sd_max_rg_data * sdp->sd_sb.sb_bsize / 2; if (!bytes) bytes = UINT_MAX; + bytes &= bsize_mask; + if (bytes == 0) + bytes = sdp->sd_sb.sb_bsize; gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &ip->i_gh); error = gfs2_glock_nq(&ip->i_gh); @@ -832,6 +863,9 @@ retry: if (error) { if (error == -ENOSPC && bytes > sdp->sd_sb.sb_bsize) { bytes >>= 1; + bytes &= bsize_mask; + if (bytes == 0) + bytes = sdp->sd_sb.sb_bsize; goto retry; } goto out_qunlock; diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index e2431313491f..2792a790e50b 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -93,14 +93,12 @@ static unsigned int gl_hash(const struct gfs2_sbd *sdp, static inline void spin_lock_bucket(unsigned int hash) { - struct hlist_bl_head *bl = &gl_hash_table[hash]; - bit_spin_lock(0, (unsigned long *)bl); + hlist_bl_lock(&gl_hash_table[hash]); } static inline void spin_unlock_bucket(unsigned int hash) { - struct hlist_bl_head *bl = &gl_hash_table[hash]; - __bit_spin_unlock(0, (unsigned long *)bl); + hlist_bl_unlock(&gl_hash_table[hash]); } static void gfs2_glock_dealloc(struct rcu_head *rcu) @@ -145,14 +143,9 @@ static int demote_ok(const struct gfs2_glock *gl) { const struct gfs2_glock_operations *glops = gl->gl_ops; - /* assert_spin_locked(&gl->gl_spin); */ - if (gl->gl_state == LM_ST_UNLOCKED) return 0; - if (test_bit(GLF_LFLUSH, &gl->gl_flags)) - return 0; - if ((gl->gl_name.ln_type != LM_TYPE_INODE) && - !list_empty(&gl->gl_holders)) + if (!list_empty(&gl->gl_holders)) return 0; if (glops->go_demote_ok) return glops->go_demote_ok(gl); @@ -160,6 +153,31 @@ static int demote_ok(const struct gfs2_glock *gl) } +void gfs2_glock_add_to_lru(struct gfs2_glock *gl) +{ + spin_lock(&lru_lock); + + if (!list_empty(&gl->gl_lru)) + list_del_init(&gl->gl_lru); + else + atomic_inc(&lru_count); + + list_add_tail(&gl->gl_lru, &lru_list); + set_bit(GLF_LRU, &gl->gl_flags); + spin_unlock(&lru_lock); +} + +static void gfs2_glock_remove_from_lru(struct gfs2_glock *gl) +{ + spin_lock(&lru_lock); + if (!list_empty(&gl->gl_lru)) { + list_del_init(&gl->gl_lru); + atomic_dec(&lru_count); + clear_bit(GLF_LRU, &gl->gl_flags); + } + spin_unlock(&lru_lock); +} + /** * __gfs2_glock_schedule_for_reclaim - Add a glock to the reclaim list * @gl: the glock @@ -170,24 +188,8 @@ static int demote_ok(const struct gfs2_glock *gl) static void __gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) { - if (demote_ok(gl)) { - spin_lock(&lru_lock); - - if (!list_empty(&gl->gl_lru)) - list_del_init(&gl->gl_lru); - else - atomic_inc(&lru_count); - - list_add_tail(&gl->gl_lru, &lru_list); - spin_unlock(&lru_lock); - } -} - -void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl) -{ - spin_lock(&gl->gl_spin); - __gfs2_glock_schedule_for_reclaim(gl); - spin_unlock(&gl->gl_spin); + if (demote_ok(gl)) + gfs2_glock_add_to_lru(gl); } /** @@ -219,12 +221,7 @@ void gfs2_glock_put(struct gfs2_glock *gl) spin_lock_bucket(gl->gl_hash); hlist_bl_del_rcu(&gl->gl_list); spin_unlock_bucket(gl->gl_hash); - spin_lock(&lru_lock); - if (!list_empty(&gl->gl_lru)) { - list_del_init(&gl->gl_lru); - atomic_dec(&lru_count); - } - spin_unlock(&lru_lock); + gfs2_glock_remove_from_lru(gl); GLOCK_BUG_ON(gl, !list_empty(&gl->gl_holders)); GLOCK_BUG_ON(gl, mapping && mapping->nrpages); trace_gfs2_glock_put(gl); @@ -544,11 +541,6 @@ __acquires(&gl->gl_spin) clear_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags); gfs2_glock_hold(gl); - if (target != LM_ST_UNLOCKED && (gl->gl_state == LM_ST_SHARED || - gl->gl_state == LM_ST_DEFERRED) && - !(lck_flags & (LM_FLAG_TRY | LM_FLAG_TRY_1CB))) - lck_flags |= LM_FLAG_TRY_1CB; - if (sdp->sd_lockstruct.ls_ops->lm_lock) { /* lock_dlm */ ret = sdp->sd_lockstruct.ls_ops->lm_lock(gl, target, lck_flags); @@ -650,7 +642,7 @@ static void delete_work_func(struct work_struct *work) /* Note: Unsafe to dereference ip as we don't hold right refs/locks */ if (ip) - inode = gfs2_ilookup(sdp->sd_vfs, no_addr); + inode = gfs2_ilookup(sdp->sd_vfs, no_addr, 1); else inode = gfs2_lookup_by_inum(sdp, no_addr, NULL, GFS2_BLKST_UNLINKED); if (inode && !IS_ERR(inode)) { @@ -1027,6 +1019,9 @@ int gfs2_glock_nq(struct gfs2_holder *gh) if (unlikely(test_bit(SDF_SHUTDOWN, &sdp->sd_flags))) return -EIO; + if (test_bit(GLF_LRU, &gl->gl_flags)) + gfs2_glock_remove_from_lru(gl); + spin_lock(&gl->gl_spin); add_to_queue(gh); if ((LM_FLAG_NOEXP & gh->gh_flags) && @@ -1084,7 +1079,8 @@ void gfs2_glock_dq(struct gfs2_holder *gh) !test_bit(GLF_DEMOTE, &gl->gl_flags)) fast_path = 1; } - __gfs2_glock_schedule_for_reclaim(gl); + if (!test_bit(GLF_LFLUSH, &gl->gl_flags)) + __gfs2_glock_schedule_for_reclaim(gl); trace_gfs2_glock_queue(gh, 0); spin_unlock(&gl->gl_spin); if (likely(fast_path)) @@ -1123,7 +1119,7 @@ void gfs2_glock_dq_uninit(struct gfs2_holder *gh) * @number: the lock number * @glops: the glock operations for the type of glock * @state: the state to acquire the glock in - * @flags: modifier flags for the aquisition + * @flags: modifier flags for the acquisition * @gh: the struct gfs2_holder * * Returns: errno @@ -1350,11 +1346,14 @@ void gfs2_glock_complete(struct gfs2_glock *gl, int ret) } -static int gfs2_shrink_glock_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) +static int gfs2_shrink_glock_memory(struct shrinker *shrink, + struct shrink_control *sc) { struct gfs2_glock *gl; int may_demote; int nr_skipped = 0; + int nr = sc->nr_to_scan; + gfp_t gfp_mask = sc->gfp_mask; LIST_HEAD(skipped); if (nr == 0) @@ -1367,6 +1366,7 @@ static int gfs2_shrink_glock_memory(struct shrinker *shrink, int nr, gfp_t gfp_m while(nr && !list_empty(&lru_list)) { gl = list_entry(lru_list.next, struct gfs2_glock, gl_lru); list_del_init(&gl->gl_lru); + clear_bit(GLF_LRU, &gl->gl_flags); atomic_dec(&lru_count); /* Test for being demotable */ @@ -1389,6 +1389,7 @@ static int gfs2_shrink_glock_memory(struct shrinker *shrink, int nr, gfp_t gfp_m } nr_skipped++; list_add(&gl->gl_lru, &skipped); + set_bit(GLF_LRU, &gl->gl_flags); } list_splice(&skipped, &lru_list); atomic_add(nr_skipped, &lru_count); @@ -1461,12 +1462,7 @@ static void thaw_glock(struct gfs2_glock *gl) static void clear_glock(struct gfs2_glock *gl) { - spin_lock(&lru_lock); - if (!list_empty(&gl->gl_lru)) { - list_del_init(&gl->gl_lru); - atomic_dec(&lru_count); - } - spin_unlock(&lru_lock); + gfs2_glock_remove_from_lru(gl); spin_lock(&gl->gl_spin); if (gl->gl_state != LM_ST_UNLOCKED) @@ -1601,9 +1597,11 @@ static int dump_holder(struct seq_file *seq, const struct gfs2_holder *gh) return 0; } -static const char *gflags2str(char *buf, const unsigned long *gflags) +static const char *gflags2str(char *buf, const struct gfs2_glock *gl) { + const unsigned long *gflags = &gl->gl_flags; char *p = buf; + if (test_bit(GLF_LOCK, gflags)) *p++ = 'l'; if (test_bit(GLF_DEMOTE, gflags)) @@ -1626,6 +1624,10 @@ static const char *gflags2str(char *buf, const unsigned long *gflags) *p++ = 'F'; if (test_bit(GLF_QUEUED, gflags)) *p++ = 'q'; + if (test_bit(GLF_LRU, gflags)) + *p++ = 'L'; + if (gl->gl_object) + *p++ = 'o'; *p = 0; return buf; } @@ -1660,14 +1662,15 @@ static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl) dtime *= 1000000/HZ; /* demote time in uSec */ if (!test_bit(GLF_DEMOTE, &gl->gl_flags)) dtime = 0; - gfs2_print_dbg(seq, "G: s:%s n:%u/%llx f:%s t:%s d:%s/%llu a:%d r:%d\n", + gfs2_print_dbg(seq, "G: s:%s n:%u/%llx f:%s t:%s d:%s/%llu a:%d v:%d r:%d\n", state2str(gl->gl_state), gl->gl_name.ln_type, (unsigned long long)gl->gl_name.ln_number, - gflags2str(gflags_buf, &gl->gl_flags), + gflags2str(gflags_buf, gl), state2str(gl->gl_target), state2str(gl->gl_demote_state), dtime, atomic_read(&gl->gl_ail_count), + atomic_read(&gl->gl_revokes), atomic_read(&gl->gl_ref)); list_for_each_entry(gh, &gl->gl_holders, gh_list) { diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h index aea160690e94..6b2f757b9281 100644 --- a/fs/gfs2/glock.h +++ b/fs/gfs2/glock.h @@ -225,11 +225,10 @@ static inline int gfs2_glock_nq_init(struct gfs2_glock *gl, extern void gfs2_glock_cb(struct gfs2_glock *gl, unsigned int state); extern void gfs2_glock_complete(struct gfs2_glock *gl, int ret); -extern void gfs2_reclaim_glock(struct gfs2_sbd *sdp); extern void gfs2_gl_hash_clear(struct gfs2_sbd *sdp); extern void gfs2_glock_finish_truncate(struct gfs2_inode *ip); extern void gfs2_glock_thaw(struct gfs2_sbd *sdp); -extern void gfs2_glock_schedule_for_reclaim(struct gfs2_glock *gl); +extern void gfs2_glock_add_to_lru(struct gfs2_glock *gl); extern void gfs2_glock_free(struct gfs2_glock *gl); extern int __init gfs2_glock_init(void); diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index 3754e3cbf02b..8ef70f464731 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -28,33 +28,18 @@ #include "trans.h" /** - * ail_empty_gl - remove all buffers for a given lock from the AIL + * __gfs2_ail_flush - remove all buffers for a given lock from the AIL * @gl: the glock * * None of the buffers should be dirty, locked, or pinned. */ -static void gfs2_ail_empty_gl(struct gfs2_glock *gl) +static void __gfs2_ail_flush(struct gfs2_glock *gl) { struct gfs2_sbd *sdp = gl->gl_sbd; struct list_head *head = &gl->gl_ail_list; struct gfs2_bufdata *bd; struct buffer_head *bh; - struct gfs2_trans tr; - - memset(&tr, 0, sizeof(tr)); - tr.tr_revokes = atomic_read(&gl->gl_ail_count); - - if (!tr.tr_revokes) - return; - - /* A shortened, inline version of gfs2_trans_begin() */ - tr.tr_reserved = 1 + gfs2_struct2blk(sdp, tr.tr_revokes, sizeof(u64)); - tr.tr_ip = (unsigned long)__builtin_return_address(0); - INIT_LIST_HEAD(&tr.tr_list_buf); - gfs2_log_reserve(sdp, tr.tr_reserved); - BUG_ON(current->journal_info); - current->journal_info = &tr; spin_lock(&sdp->sd_ail_lock); while (!list_empty(head)) { @@ -76,11 +61,51 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl) } gfs2_assert_withdraw(sdp, !atomic_read(&gl->gl_ail_count)); spin_unlock(&sdp->sd_ail_lock); +} + + +static void gfs2_ail_empty_gl(struct gfs2_glock *gl) +{ + struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_trans tr; + + memset(&tr, 0, sizeof(tr)); + tr.tr_revokes = atomic_read(&gl->gl_ail_count); + + if (!tr.tr_revokes) + return; + + /* A shortened, inline version of gfs2_trans_begin() */ + tr.tr_reserved = 1 + gfs2_struct2blk(sdp, tr.tr_revokes, sizeof(u64)); + tr.tr_ip = (unsigned long)__builtin_return_address(0); + INIT_LIST_HEAD(&tr.tr_list_buf); + gfs2_log_reserve(sdp, tr.tr_reserved); + BUG_ON(current->journal_info); + current->journal_info = &tr; + + __gfs2_ail_flush(gl); gfs2_trans_end(sdp); gfs2_log_flush(sdp, NULL); } +void gfs2_ail_flush(struct gfs2_glock *gl) +{ + struct gfs2_sbd *sdp = gl->gl_sbd; + unsigned int revokes = atomic_read(&gl->gl_ail_count); + int ret; + + if (!revokes) + return; + + ret = gfs2_trans_begin(sdp, 0, revokes); + if (ret) + return; + __gfs2_ail_flush(gl); + gfs2_trans_end(sdp); + gfs2_log_flush(sdp, NULL); +} + /** * rgrp_go_sync - sync out the metadata for this glock * @gl: the glock @@ -227,6 +252,119 @@ static int inode_go_demote_ok(const struct gfs2_glock *gl) } /** + * gfs2_set_nlink - Set the inode's link count based on on-disk info + * @inode: The inode in question + * @nlink: The link count + * + * If the link count has hit zero, it must never be raised, whatever the + * on-disk inode might say. When new struct inodes are created the link + * count is set to 1, so that we can safely use this test even when reading + * in on disk information for the first time. + */ + +static void gfs2_set_nlink(struct inode *inode, u32 nlink) +{ + /* + * We will need to review setting the nlink count here in the + * light of the forthcoming ro bind mount work. This is a reminder + * to do that. + */ + if ((inode->i_nlink != nlink) && (inode->i_nlink != 0)) { + if (nlink == 0) + clear_nlink(inode); + else + inode->i_nlink = nlink; + } +} + +static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) +{ + const struct gfs2_dinode *str = buf; + struct timespec atime; + u16 height, depth; + + if (unlikely(ip->i_no_addr != be64_to_cpu(str->di_num.no_addr))) + goto corrupt; + ip->i_no_formal_ino = be64_to_cpu(str->di_num.no_formal_ino); + ip->i_inode.i_mode = be32_to_cpu(str->di_mode); + ip->i_inode.i_rdev = 0; + switch (ip->i_inode.i_mode & S_IFMT) { + case S_IFBLK: + case S_IFCHR: + ip->i_inode.i_rdev = MKDEV(be32_to_cpu(str->di_major), + be32_to_cpu(str->di_minor)); + break; + }; + + ip->i_inode.i_uid = be32_to_cpu(str->di_uid); + ip->i_inode.i_gid = be32_to_cpu(str->di_gid); + gfs2_set_nlink(&ip->i_inode, be32_to_cpu(str->di_nlink)); + i_size_write(&ip->i_inode, be64_to_cpu(str->di_size)); + gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks)); + atime.tv_sec = be64_to_cpu(str->di_atime); + atime.tv_nsec = be32_to_cpu(str->di_atime_nsec); + if (timespec_compare(&ip->i_inode.i_atime, &atime) < 0) + ip->i_inode.i_atime = atime; + ip->i_inode.i_mtime.tv_sec = be64_to_cpu(str->di_mtime); + ip->i_inode.i_mtime.tv_nsec = be32_to_cpu(str->di_mtime_nsec); + ip->i_inode.i_ctime.tv_sec = be64_to_cpu(str->di_ctime); + ip->i_inode.i_ctime.tv_nsec = be32_to_cpu(str->di_ctime_nsec); + + ip->i_goal = be64_to_cpu(str->di_goal_meta); + ip->i_generation = be64_to_cpu(str->di_generation); + + ip->i_diskflags = be32_to_cpu(str->di_flags); + gfs2_set_inode_flags(&ip->i_inode); + height = be16_to_cpu(str->di_height); + if (unlikely(height > GFS2_MAX_META_HEIGHT)) + goto corrupt; + ip->i_height = (u8)height; + + depth = be16_to_cpu(str->di_depth); + if (unlikely(depth > GFS2_DIR_MAX_DEPTH)) + goto corrupt; + ip->i_depth = (u8)depth; + ip->i_entries = be32_to_cpu(str->di_entries); + + ip->i_eattr = be64_to_cpu(str->di_eattr); + if (S_ISREG(ip->i_inode.i_mode)) + gfs2_set_aops(&ip->i_inode); + + return 0; +corrupt: + gfs2_consist_inode(ip); + return -EIO; +} + +/** + * gfs2_inode_refresh - Refresh the incore copy of the dinode + * @ip: The GFS2 inode + * + * Returns: errno + */ + +int gfs2_inode_refresh(struct gfs2_inode *ip) +{ + struct buffer_head *dibh; + int error; + + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) + return error; + + if (gfs2_metatype_check(GFS2_SB(&ip->i_inode), dibh, GFS2_METATYPE_DI)) { + brelse(dibh); + return -EIO; + } + + error = gfs2_dinode_in(ip, dibh->b_data); + brelse(dibh); + clear_bit(GIF_INVALID, &ip->i_flags); + + return error; +} + +/** * inode_go_lock - operation done after an inode lock is locked by a process * @gl: the glock * @flags: @@ -385,6 +523,10 @@ static int trans_go_demote_ok(const struct gfs2_glock *gl) static void iopen_go_callback(struct gfs2_glock *gl) { struct gfs2_inode *ip = (struct gfs2_inode *)gl->gl_object; + struct gfs2_sbd *sdp = gl->gl_sbd; + + if (sdp->sd_vfs->s_flags & MS_RDONLY) + return; if (gl->gl_demote_state == LM_ST_UNLOCKED && gl->gl_state == LM_ST_SHARED && ip) { diff --git a/fs/gfs2/glops.h b/fs/gfs2/glops.h index b3aa2e3210fd..6fce409b5a50 100644 --- a/fs/gfs2/glops.h +++ b/fs/gfs2/glops.h @@ -23,4 +23,6 @@ extern const struct gfs2_glock_operations gfs2_quota_glops; extern const struct gfs2_glock_operations gfs2_journal_glops; extern const struct gfs2_glock_operations *gfs2_glops_list[]; +extern void gfs2_ail_flush(struct gfs2_glock *gl); + #endif /* __GLOPS_DOT_H__ */ diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 870a89d6d4dc..0a064e91ac70 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -20,7 +20,6 @@ #define DIO_WAIT 0x00000010 #define DIO_METADATA 0x00000020 -#define DIO_ALL 0x00000100 struct gfs2_log_operations; struct gfs2_log_element; @@ -200,6 +199,8 @@ enum { GLF_INITIAL = 10, GLF_FROZEN = 11, GLF_QUEUED = 12, + GLF_LRU = 13, + GLF_OBJECT = 14, /* Used only for tracing */ }; struct gfs2_glock { @@ -234,6 +235,7 @@ struct gfs2_glock { struct list_head gl_ail_list; atomic_t gl_ail_count; + atomic_t gl_revokes; struct delayed_work gl_work; struct work_struct gl_delete; struct rcu_head gl_rcu; @@ -374,8 +376,6 @@ struct gfs2_ail { unsigned int ai_first; struct list_head ai_ail1_list; struct list_head ai_ail2_list; - - u64 ai_sync_gen; }; struct gfs2_journal_extent { @@ -488,7 +488,6 @@ struct gfs2_sb_host { char sb_lockproto[GFS2_LOCKNAME_LEN]; char sb_locktable[GFS2_LOCKNAME_LEN]; - u8 sb_uuid[16]; }; /* @@ -654,7 +653,6 @@ struct gfs2_sbd { spinlock_t sd_ail_lock; struct list_head sd_ail1_list; struct list_head sd_ail2_list; - u64 sd_ail_sync_gen; /* Replay stuff */ diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 97d54a28776a..03e0c529063e 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -1,23 +1,25 @@ /* * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2008 Red Hat, Inc. All rights reserved. + * Copyright (C) 2004-2011 Red Hat, Inc. All rights reserved. * * This copyrighted material is made available to anyone wishing to use, * modify, copy, or redistribute it subject to the terms and conditions * of the GNU General Public License version 2. */ -#include <linux/sched.h> #include <linux/slab.h> #include <linux/spinlock.h> #include <linux/completion.h> #include <linux/buffer_head.h> +#include <linux/namei.h> +#include <linux/mm.h> +#include <linux/xattr.h> #include <linux/posix_acl.h> -#include <linux/sort.h> #include <linux/gfs2_ondisk.h> #include <linux/crc32.h> +#include <linux/fiemap.h> #include <linux/security.h> -#include <linux/time.h> +#include <asm/uaccess.h> #include "gfs2.h" #include "incore.h" @@ -26,51 +28,70 @@ #include "dir.h" #include "xattr.h" #include "glock.h" -#include "glops.h" #include "inode.h" -#include "log.h" #include "meta_io.h" #include "quota.h" #include "rgrp.h" #include "trans.h" #include "util.h" +#include "super.h" +#include "glops.h" -struct gfs2_inum_range_host { - u64 ir_start; - u64 ir_length; +struct gfs2_skip_data { + u64 no_addr; + int skipped; + int non_block; }; static int iget_test(struct inode *inode, void *opaque) { struct gfs2_inode *ip = GFS2_I(inode); - u64 *no_addr = opaque; + struct gfs2_skip_data *data = opaque; - if (ip->i_no_addr == *no_addr) + if (ip->i_no_addr == data->no_addr) { + if (data->non_block && + inode->i_state & (I_FREEING|I_CLEAR|I_WILL_FREE)) { + data->skipped = 1; + return 0; + } return 1; - + } return 0; } static int iget_set(struct inode *inode, void *opaque) { struct gfs2_inode *ip = GFS2_I(inode); - u64 *no_addr = opaque; + struct gfs2_skip_data *data = opaque; - inode->i_ino = (unsigned long)*no_addr; - ip->i_no_addr = *no_addr; + if (data->skipped) + return -ENOENT; + inode->i_ino = (unsigned long)(data->no_addr); + ip->i_no_addr = data->no_addr; return 0; } -struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr) +struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr, int non_block) { unsigned long hash = (unsigned long)no_addr; - return ilookup5(sb, hash, iget_test, &no_addr); + struct gfs2_skip_data data; + + data.no_addr = no_addr; + data.skipped = 0; + data.non_block = non_block; + return ilookup5(sb, hash, iget_test, &data); } -static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr) +static struct inode *gfs2_iget(struct super_block *sb, u64 no_addr, + int non_block) { + struct gfs2_skip_data data; unsigned long hash = (unsigned long)no_addr; - return iget5_locked(sb, hash, iget_test, iget_set, &no_addr); + + data.no_addr = no_addr; + data.skipped = 0; + data.non_block = non_block; + return iget5_locked(sb, hash, iget_test, iget_set, &data); } /** @@ -111,19 +132,20 @@ static void gfs2_set_iop(struct inode *inode) * @sb: The super block * @no_addr: The inode number * @type: The type of the inode + * non_block: Can we block on inodes that are being freed? * * Returns: A VFS inode, or an error */ struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned int type, - u64 no_addr, u64 no_formal_ino) + u64 no_addr, u64 no_formal_ino, int non_block) { struct inode *inode; struct gfs2_inode *ip; struct gfs2_glock *io_gl = NULL; int error; - inode = gfs2_iget(sb, no_addr); + inode = gfs2_iget(sb, no_addr, non_block); ip = GFS2_I(inode); if (!inode) @@ -185,11 +207,12 @@ struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, { struct super_block *sb = sdp->sd_vfs; struct gfs2_holder i_gh; - struct inode *inode; + struct inode *inode = NULL; int error; + /* Must not read in block until block type is verified */ error = gfs2_glock_nq_num(sdp, no_addr, &gfs2_inode_glops, - LM_ST_SHARED, LM_FLAG_ANY, &i_gh); + LM_ST_EXCLUSIVE, GL_SKIP, &i_gh); if (error) return ERR_PTR(error); @@ -197,7 +220,7 @@ struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, if (error) goto fail; - inode = gfs2_inode_lookup(sb, DT_UNKNOWN, no_addr, 0); + inode = gfs2_inode_lookup(sb, DT_UNKNOWN, no_addr, 0, 1); if (IS_ERR(inode)) goto fail; @@ -222,203 +245,6 @@ fail_iput: goto fail; } -static int gfs2_dinode_in(struct gfs2_inode *ip, const void *buf) -{ - const struct gfs2_dinode *str = buf; - struct timespec atime; - u16 height, depth; - - if (unlikely(ip->i_no_addr != be64_to_cpu(str->di_num.no_addr))) - goto corrupt; - ip->i_no_formal_ino = be64_to_cpu(str->di_num.no_formal_ino); - ip->i_inode.i_mode = be32_to_cpu(str->di_mode); - ip->i_inode.i_rdev = 0; - switch (ip->i_inode.i_mode & S_IFMT) { - case S_IFBLK: - case S_IFCHR: - ip->i_inode.i_rdev = MKDEV(be32_to_cpu(str->di_major), - be32_to_cpu(str->di_minor)); - break; - }; - - ip->i_inode.i_uid = be32_to_cpu(str->di_uid); - ip->i_inode.i_gid = be32_to_cpu(str->di_gid); - /* - * We will need to review setting the nlink count here in the - * light of the forthcoming ro bind mount work. This is a reminder - * to do that. - */ - ip->i_inode.i_nlink = be32_to_cpu(str->di_nlink); - i_size_write(&ip->i_inode, be64_to_cpu(str->di_size)); - gfs2_set_inode_blocks(&ip->i_inode, be64_to_cpu(str->di_blocks)); - atime.tv_sec = be64_to_cpu(str->di_atime); - atime.tv_nsec = be32_to_cpu(str->di_atime_nsec); - if (timespec_compare(&ip->i_inode.i_atime, &atime) < 0) - ip->i_inode.i_atime = atime; - ip->i_inode.i_mtime.tv_sec = be64_to_cpu(str->di_mtime); - ip->i_inode.i_mtime.tv_nsec = be32_to_cpu(str->di_mtime_nsec); - ip->i_inode.i_ctime.tv_sec = be64_to_cpu(str->di_ctime); - ip->i_inode.i_ctime.tv_nsec = be32_to_cpu(str->di_ctime_nsec); - - ip->i_goal = be64_to_cpu(str->di_goal_meta); - ip->i_generation = be64_to_cpu(str->di_generation); - - ip->i_diskflags = be32_to_cpu(str->di_flags); - gfs2_set_inode_flags(&ip->i_inode); - height = be16_to_cpu(str->di_height); - if (unlikely(height > GFS2_MAX_META_HEIGHT)) - goto corrupt; - ip->i_height = (u8)height; - - depth = be16_to_cpu(str->di_depth); - if (unlikely(depth > GFS2_DIR_MAX_DEPTH)) - goto corrupt; - ip->i_depth = (u8)depth; - ip->i_entries = be32_to_cpu(str->di_entries); - - ip->i_eattr = be64_to_cpu(str->di_eattr); - if (S_ISREG(ip->i_inode.i_mode)) - gfs2_set_aops(&ip->i_inode); - - return 0; -corrupt: - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; -} - -/** - * gfs2_inode_refresh - Refresh the incore copy of the dinode - * @ip: The GFS2 inode - * - * Returns: errno - */ - -int gfs2_inode_refresh(struct gfs2_inode *ip) -{ - struct buffer_head *dibh; - int error; - - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - return error; - - if (gfs2_metatype_check(GFS2_SB(&ip->i_inode), dibh, GFS2_METATYPE_DI)) { - brelse(dibh); - return -EIO; - } - - error = gfs2_dinode_in(ip, dibh->b_data); - brelse(dibh); - clear_bit(GIF_INVALID, &ip->i_flags); - - return error; -} - -int gfs2_dinode_dealloc(struct gfs2_inode *ip) -{ - struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); - struct gfs2_alloc *al; - struct gfs2_rgrpd *rgd; - int error; - - if (gfs2_get_inode_blocks(&ip->i_inode) != 1) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; - } - - al = gfs2_alloc_get(ip); - if (!al) - return -ENOMEM; - - error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); - if (error) - goto out; - - error = gfs2_rindex_hold(sdp, &al->al_ri_gh); - if (error) - goto out_qs; - - rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); - if (!rgd) { - gfs2_consist_inode(ip); - error = -EIO; - goto out_rindex_relse; - } - - error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, - &al->al_rgd_gh); - if (error) - goto out_rindex_relse; - - error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_STATFS + RES_QUOTA, 1); - if (error) - goto out_rg_gunlock; - - set_bit(GLF_DIRTY, &ip->i_gl->gl_flags); - set_bit(GLF_LFLUSH, &ip->i_gl->gl_flags); - - gfs2_free_di(rgd, ip); - - gfs2_trans_end(sdp); - -out_rg_gunlock: - gfs2_glock_dq_uninit(&al->al_rgd_gh); -out_rindex_relse: - gfs2_glock_dq_uninit(&al->al_ri_gh); -out_qs: - gfs2_quota_unhold(ip); -out: - gfs2_alloc_put(ip); - return error; -} - -/** - * gfs2_change_nlink - Change nlink count on inode - * @ip: The GFS2 inode - * @diff: The change in the nlink count required - * - * Returns: errno - */ -int gfs2_change_nlink(struct gfs2_inode *ip, int diff) -{ - struct buffer_head *dibh; - u32 nlink; - int error; - - BUG_ON(diff != 1 && diff != -1); - nlink = ip->i_inode.i_nlink + diff; - - /* If we are reducing the nlink count, but the new value ends up being - bigger than the old one, we must have underflowed. */ - if (diff < 0 && nlink > ip->i_inode.i_nlink) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; - } - - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - return error; - - if (diff > 0) - inc_nlink(&ip->i_inode); - else - drop_nlink(&ip->i_inode); - - ip->i_inode.i_ctime = CURRENT_TIME; - - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); - brelse(dibh); - mark_inode_dirty(&ip->i_inode); - - if (ip->i_inode.i_nlink == 0) - gfs2_unlink_di(&ip->i_inode); /* mark inode unlinked */ - - return error; -} struct inode *gfs2_lookup_simple(struct inode *dip, const char *name) { @@ -517,7 +343,7 @@ static int create_ok(struct gfs2_inode *dip, const struct qstr *name, /* Don't create entries in an unlinked directory */ if (!dip->i_inode.i_nlink) - return -EPERM; + return -ENOENT; error = gfs2_dir_check(&dip->i_inode, name, NULL); switch (error) { @@ -587,21 +413,44 @@ out: return error; } +static void gfs2_init_dir(struct buffer_head *dibh, + const struct gfs2_inode *parent) +{ + struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data; + struct gfs2_dirent *dent = (struct gfs2_dirent *)(di+1); + + gfs2_qstr2dirent(&gfs2_qdot, GFS2_DIRENT_SIZE(gfs2_qdot.len), dent); + dent->de_inum = di->di_num; /* already GFS2 endian */ + dent->de_type = cpu_to_be16(DT_DIR); + + dent = (struct gfs2_dirent *)((char*)dent + GFS2_DIRENT_SIZE(1)); + gfs2_qstr2dirent(&gfs2_qdotdot, dibh->b_size - GFS2_DIRENT_SIZE(1) - sizeof(struct gfs2_dinode), dent); + gfs2_inum_out(parent, dent); + dent->de_type = cpu_to_be16(DT_DIR); + +} + /** * init_dinode - Fill in a new dinode structure - * @dip: the directory this inode is being created in + * @dip: The directory this inode is being created in * @gl: The glock covering the new inode - * @inum: the inode number - * @mode: the file permissions - * @uid: - * @gid: + * @inum: The inode number + * @mode: The file permissions + * @uid: The uid of the new inode + * @gid: The gid of the new inode + * @generation: The generation number of the new inode + * @dev: The device number (if a device node) + * @symname: The symlink destination (if a symlink) + * @size: The inode size (ignored for directories) + * @bhp: The buffer head (returned to caller) * */ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, const struct gfs2_inum_host *inum, unsigned int mode, unsigned int uid, unsigned int gid, - const u64 *generation, dev_t dev, struct buffer_head **bhp) + const u64 *generation, dev_t dev, const char *symname, + unsigned size, struct buffer_head **bhp) { struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct gfs2_dinode *di; @@ -620,7 +469,7 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, di->di_uid = cpu_to_be32(uid); di->di_gid = cpu_to_be32(gid); di->di_nlink = 0; - di->di_size = 0; + di->di_size = cpu_to_be64(size); di->di_blocks = cpu_to_be64(1); di->di_atime = di->di_mtime = di->di_ctime = cpu_to_be64(tv.tv_sec); di->di_major = cpu_to_be32(MAJOR(dev)); @@ -628,16 +477,6 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, di->di_goal_meta = di->di_goal_data = cpu_to_be64(inum->no_addr); di->di_generation = cpu_to_be64(*generation); di->di_flags = 0; - - if (S_ISREG(mode)) { - if ((dip->i_diskflags & GFS2_DIF_INHERIT_JDATA) || - gfs2_tune_get(sdp, gt_new_files_jdata)) - di->di_flags |= cpu_to_be32(GFS2_DIF_JDATA); - } else if (S_ISDIR(mode)) { - di->di_flags |= cpu_to_be32(dip->i_diskflags & - GFS2_DIF_INHERIT_JDATA); - } - di->__pad1 = 0; di->di_payload_format = cpu_to_be32(S_ISDIR(mode) ? GFS2_FORMAT_DE : 0); di->di_height = 0; @@ -651,7 +490,26 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, di->di_mtime_nsec = cpu_to_be32(tv.tv_nsec); di->di_ctime_nsec = cpu_to_be32(tv.tv_nsec); memset(&di->di_reserved, 0, sizeof(di->di_reserved)); - + + switch(mode & S_IFMT) { + case S_IFREG: + if ((dip->i_diskflags & GFS2_DIF_INHERIT_JDATA) || + gfs2_tune_get(sdp, gt_new_files_jdata)) + di->di_flags |= cpu_to_be32(GFS2_DIF_JDATA); + break; + case S_IFDIR: + di->di_flags |= cpu_to_be32(dip->i_diskflags & + GFS2_DIF_INHERIT_JDATA); + di->di_flags |= cpu_to_be32(GFS2_DIF_JDATA); + di->di_size = cpu_to_be64(sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)); + di->di_entries = cpu_to_be32(2); + gfs2_init_dir(dibh, dip); + break; + case S_IFLNK: + memcpy(dibh->b_data + sizeof(struct gfs2_dinode), symname, size); + break; + } + set_buffer_uptodate(dibh); *bhp = dibh; @@ -659,7 +517,8 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, static int make_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, unsigned int mode, const struct gfs2_inum_host *inum, - const u64 *generation, dev_t dev, struct buffer_head **bhp) + const u64 *generation, dev_t dev, const char *symname, + unsigned int size, struct buffer_head **bhp) { struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); unsigned int uid, gid; @@ -681,7 +540,7 @@ static int make_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, if (error) goto out_quota; - init_dinode(dip, gl, inum, mode, uid, gid, generation, dev, bhp); + init_dinode(dip, gl, inum, mode, uid, gid, generation, dev, symname, size, bhp); gfs2_quota_change(dip, +1, uid, gid); gfs2_trans_end(sdp); @@ -735,14 +594,16 @@ static int link_dinode(struct gfs2_inode *dip, const struct qstr *name, goto fail_quota_locks; } - error = gfs2_dir_add(&dip->i_inode, name, ip, IF2DT(ip->i_inode.i_mode)); + error = gfs2_dir_add(&dip->i_inode, name, ip); if (error) goto fail_end_trans; error = gfs2_meta_inode_buffer(ip, &dibh); if (error) goto fail_end_trans; - ip->i_inode.i_nlink = 1; + inc_nlink(&ip->i_inode); + if (S_ISDIR(ip->i_inode.i_mode)) + inc_nlink(&ip->i_inode); gfs2_trans_add_bh(ip->i_gl, dibh, 1); gfs2_dinode_out(ip, dibh->b_data); brelse(dibh); @@ -789,27 +650,25 @@ static int gfs2_security_init(struct gfs2_inode *dip, struct gfs2_inode *ip, } /** - * gfs2_createi - Create a new inode - * @ghs: An array of two holders - * @name: The name of the new file - * @mode: the permissions on the new inode + * gfs2_create_inode - Create a new inode + * @dir: The parent directory + * @dentry: The new dentry + * @mode: The permissions on the new inode + * @dev: For device nodes, this is the device number + * @symname: For symlinks, this is the link destination + * @size: The initial size of the inode (ignored for directories) * - * @ghs[0] is an initialized holder for the directory - * @ghs[1] is the holder for the inode lock - * - * If the return value is not NULL, the glocks on both the directory and the new - * file are held. A transaction has been started and an inplace reservation - * is held, as well. - * - * Returns: An inode + * Returns: 0 on success, or error code */ -struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, - unsigned int mode, dev_t dev) +static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, + unsigned int mode, dev_t dev, const char *symname, + unsigned int size) { + const struct qstr *name = &dentry->d_name; + struct gfs2_holder ghs[2]; struct inode *inode = NULL; - struct gfs2_inode *dip = ghs->gh_gl->gl_object; - struct inode *dir = &dip->i_inode; + struct gfs2_inode *dip = GFS2_I(dir); struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct gfs2_inum_host inum = { .no_addr = 0, .no_formal_ino = 0 }; int error; @@ -817,10 +676,9 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, struct buffer_head *bh = NULL; if (!name->len || name->len > GFS2_FNAMESIZE) - return ERR_PTR(-ENAMETOOLONG); + return -ENAMETOOLONG; - gfs2_holder_reinit(LM_ST_EXCLUSIVE, 0, ghs); - error = gfs2_glock_nq(ghs); + error = gfs2_glock_nq_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); if (error) goto fail; @@ -838,12 +696,12 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, if (error) goto fail_gunlock; - error = make_dinode(dip, ghs[1].gh_gl, mode, &inum, &generation, dev, &bh); + error = make_dinode(dip, ghs[1].gh_gl, mode, &inum, &generation, dev, symname, size, &bh); if (error) goto fail_gunlock2; inode = gfs2_inode_lookup(dir->i_sb, IF2DT(mode), inum.no_addr, - inum.no_formal_ino); + inum.no_formal_ino, 0); if (IS_ERR(inode)) goto fail_gunlock2; @@ -865,18 +723,852 @@ struct inode *gfs2_createi(struct gfs2_holder *ghs, const struct qstr *name, if (bh) brelse(bh); - return inode; + + gfs2_trans_end(sdp); + if (dip->i_alloc->al_rgd) + gfs2_inplace_release(dip); + gfs2_quota_unlock(dip); + gfs2_alloc_put(dip); + gfs2_glock_dq_uninit_m(2, ghs); + mark_inode_dirty(inode); + d_instantiate(dentry, inode); + return 0; fail_gunlock2: gfs2_glock_dq_uninit(ghs + 1); if (inode && !IS_ERR(inode)) iput(inode); fail_gunlock: - gfs2_glock_dq(ghs); + gfs2_glock_dq_uninit(ghs); fail: if (bh) brelse(bh); - return ERR_PTR(error); + return error; +} + +/** + * gfs2_create - Create a file + * @dir: The directory in which to create the file + * @dentry: The dentry of the new file + * @mode: The mode of the new file + * + * Returns: errno + */ + +static int gfs2_create(struct inode *dir, struct dentry *dentry, + int mode, struct nameidata *nd) +{ + struct inode *inode; + int ret; + + for (;;) { + ret = gfs2_create_inode(dir, dentry, S_IFREG | mode, 0, NULL, 0); + if (ret != -EEXIST || (nd && (nd->flags & LOOKUP_EXCL))) + return ret; + + inode = gfs2_lookupi(dir, &dentry->d_name, 0); + if (inode) { + if (!IS_ERR(inode)) + break; + return PTR_ERR(inode); + } + } + + d_instantiate(dentry, inode); + return 0; +} + +/** + * gfs2_lookup - Look up a filename in a directory and return its inode + * @dir: The directory inode + * @dentry: The dentry of the new inode + * @nd: passed from Linux VFS, ignored by us + * + * Called by the VFS layer. Lock dir and call gfs2_lookupi() + * + * Returns: errno + */ + +static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry, + struct nameidata *nd) +{ + struct inode *inode = NULL; + + inode = gfs2_lookupi(dir, &dentry->d_name, 0); + if (inode && IS_ERR(inode)) + return ERR_CAST(inode); + + if (inode) { + struct gfs2_glock *gl = GFS2_I(inode)->i_gl; + struct gfs2_holder gh; + int error; + error = gfs2_glock_nq_init(gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); + if (error) { + iput(inode); + return ERR_PTR(error); + } + gfs2_glock_dq_uninit(&gh); + return d_splice_alias(inode, dentry); + } + d_add(dentry, inode); + + return NULL; +} + +/** + * gfs2_link - Link to a file + * @old_dentry: The inode to link + * @dir: Add link to this directory + * @dentry: The name of the link + * + * Link the inode in "old_dentry" into the directory "dir" with the + * name in "dentry". + * + * Returns: errno + */ + +static int gfs2_link(struct dentry *old_dentry, struct inode *dir, + struct dentry *dentry) +{ + struct gfs2_inode *dip = GFS2_I(dir); + struct gfs2_sbd *sdp = GFS2_SB(dir); + struct inode *inode = old_dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder ghs[2]; + struct buffer_head *dibh; + int alloc_required; + int error; + + if (S_ISDIR(inode->i_mode)) + return -EPERM; + + gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); + + error = gfs2_glock_nq(ghs); /* parent */ + if (error) + goto out_parent; + + error = gfs2_glock_nq(ghs + 1); /* child */ + if (error) + goto out_child; + + error = -ENOENT; + if (inode->i_nlink == 0) + goto out_gunlock; + + error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC, 0); + if (error) + goto out_gunlock; + + error = gfs2_dir_check(dir, &dentry->d_name, NULL); + switch (error) { + case -ENOENT: + break; + case 0: + error = -EEXIST; + default: + goto out_gunlock; + } + + error = -EINVAL; + if (!dip->i_inode.i_nlink) + goto out_gunlock; + error = -EFBIG; + if (dip->i_entries == (u32)-1) + goto out_gunlock; + error = -EPERM; + if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) + goto out_gunlock; + error = -EINVAL; + if (!ip->i_inode.i_nlink) + goto out_gunlock; + error = -EMLINK; + if (ip->i_inode.i_nlink == (u32)-1) + goto out_gunlock; + + alloc_required = error = gfs2_diradd_alloc_required(dir, &dentry->d_name); + if (error < 0) + goto out_gunlock; + error = 0; + + if (alloc_required) { + struct gfs2_alloc *al = gfs2_alloc_get(dip); + if (!al) { + error = -ENOMEM; + goto out_gunlock; + } + + error = gfs2_quota_lock_check(dip); + if (error) + goto out_alloc; + + al->al_requested = sdp->sd_max_dirres; + + error = gfs2_inplace_reserve(dip); + if (error) + goto out_gunlock_q; + + error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + + gfs2_rg_blocks(al) + + 2 * RES_DINODE + RES_STATFS + + RES_QUOTA, 0); + if (error) + goto out_ipres; + } else { + error = gfs2_trans_begin(sdp, 2 * RES_DINODE + RES_LEAF, 0); + if (error) + goto out_ipres; + } + + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) + goto out_end_trans; + + error = gfs2_dir_add(dir, &dentry->d_name, ip); + if (error) + goto out_brelse; + + gfs2_trans_add_bh(ip->i_gl, dibh, 1); + inc_nlink(&ip->i_inode); + ip->i_inode.i_ctime = CURRENT_TIME; + gfs2_dinode_out(ip, dibh->b_data); + mark_inode_dirty(&ip->i_inode); + +out_brelse: + brelse(dibh); +out_end_trans: + gfs2_trans_end(sdp); +out_ipres: + if (alloc_required) + gfs2_inplace_release(dip); +out_gunlock_q: + if (alloc_required) + gfs2_quota_unlock(dip); +out_alloc: + if (alloc_required) + gfs2_alloc_put(dip); +out_gunlock: + gfs2_glock_dq(ghs + 1); +out_child: + gfs2_glock_dq(ghs); +out_parent: + gfs2_holder_uninit(ghs); + gfs2_holder_uninit(ghs + 1); + if (!error) { + ihold(inode); + d_instantiate(dentry, inode); + mark_inode_dirty(inode); + } + return error; +} + +/* + * gfs2_unlink_ok - check to see that a inode is still in a directory + * @dip: the directory + * @name: the name of the file + * @ip: the inode + * + * Assumes that the lock on (at least) @dip is held. + * + * Returns: 0 if the parent/child relationship is correct, errno if it isn't + */ + +static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, + const struct gfs2_inode *ip) +{ + int error; + + if (IS_IMMUTABLE(&ip->i_inode) || IS_APPEND(&ip->i_inode)) + return -EPERM; + + if ((dip->i_inode.i_mode & S_ISVTX) && + dip->i_inode.i_uid != current_fsuid() && + ip->i_inode.i_uid != current_fsuid() && !capable(CAP_FOWNER)) + return -EPERM; + + if (IS_APPEND(&dip->i_inode)) + return -EPERM; + + error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0); + if (error) + return error; + + error = gfs2_dir_check(&dip->i_inode, name, ip); + if (error) + return error; + + return 0; +} + +/** + * gfs2_unlink_inode - Removes an inode from its parent dir and unlinks it + * @dip: The parent directory + * @name: The name of the entry in the parent directory + * @bh: The inode buffer for the inode to be removed + * @inode: The inode to be removed + * + * Called with all the locks and in a transaction. This will only be + * called for a directory after it has been checked to ensure it is empty. + * + * Returns: 0 on success, or an error + */ + +static int gfs2_unlink_inode(struct gfs2_inode *dip, + const struct dentry *dentry, + struct buffer_head *bh) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + int error; + + error = gfs2_dir_del(dip, dentry); + if (error) + return error; + + ip->i_entries = 0; + inode->i_ctime = CURRENT_TIME; + if (S_ISDIR(inode->i_mode)) + clear_nlink(inode); + else + drop_nlink(inode); + gfs2_trans_add_bh(ip->i_gl, bh, 1); + gfs2_dinode_out(ip, bh->b_data); + mark_inode_dirty(inode); + if (inode->i_nlink == 0) + gfs2_unlink_di(inode); + return 0; +} + + +/** + * gfs2_unlink - Unlink an inode (this does rmdir as well) + * @dir: The inode of the directory containing the inode to unlink + * @dentry: The file itself + * + * This routine uses the type of the inode as a flag to figure out + * whether this is an unlink or an rmdir. + * + * Returns: errno + */ + +static int gfs2_unlink(struct inode *dir, struct dentry *dentry) +{ + struct gfs2_inode *dip = GFS2_I(dir); + struct gfs2_sbd *sdp = GFS2_SB(dir); + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct buffer_head *bh; + struct gfs2_holder ghs[3]; + struct gfs2_rgrpd *rgd; + struct gfs2_holder ri_gh; + int error; + + error = gfs2_rindex_hold(sdp, &ri_gh); + if (error) + return error; + + gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); + + rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); + gfs2_holder_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + 2); + + + error = gfs2_glock_nq(ghs); /* parent */ + if (error) + goto out_parent; + + error = gfs2_glock_nq(ghs + 1); /* child */ + if (error) + goto out_child; + + error = -ENOENT; + if (inode->i_nlink == 0) + goto out_rgrp; + + if (S_ISDIR(inode->i_mode)) { + error = -ENOTEMPTY; + if (ip->i_entries > 2 || inode->i_nlink > 2) + goto out_rgrp; + } + + error = gfs2_glock_nq(ghs + 2); /* rgrp */ + if (error) + goto out_rgrp; + + error = gfs2_unlink_ok(dip, &dentry->d_name, ip); + if (error) + goto out_gunlock; + + error = gfs2_trans_begin(sdp, 2*RES_DINODE + 3*RES_LEAF + RES_RG_BIT, 0); + if (error) + goto out_gunlock; + + error = gfs2_meta_inode_buffer(ip, &bh); + if (error) + goto out_end_trans; + + error = gfs2_unlink_inode(dip, dentry, bh); + brelse(bh); + +out_end_trans: + gfs2_trans_end(sdp); +out_gunlock: + gfs2_glock_dq(ghs + 2); +out_rgrp: + gfs2_holder_uninit(ghs + 2); + gfs2_glock_dq(ghs + 1); +out_child: + gfs2_holder_uninit(ghs + 1); + gfs2_glock_dq(ghs); +out_parent: + gfs2_holder_uninit(ghs); + gfs2_glock_dq_uninit(&ri_gh); + return error; +} + +/** + * gfs2_symlink - Create a symlink + * @dir: The directory to create the symlink in + * @dentry: The dentry to put the symlink in + * @symname: The thing which the link points to + * + * Returns: errno + */ + +static int gfs2_symlink(struct inode *dir, struct dentry *dentry, + const char *symname) +{ + struct gfs2_sbd *sdp = GFS2_SB(dir); + unsigned int size; + + size = strlen(symname); + if (size > sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode) - 1) + return -ENAMETOOLONG; + + return gfs2_create_inode(dir, dentry, S_IFLNK | S_IRWXUGO, 0, symname, size); +} + +/** + * gfs2_mkdir - Make a directory + * @dir: The parent directory of the new one + * @dentry: The dentry of the new directory + * @mode: The mode of the new directory + * + * Returns: errno + */ + +static int gfs2_mkdir(struct inode *dir, struct dentry *dentry, int mode) +{ + return gfs2_create_inode(dir, dentry, S_IFDIR | mode, 0, NULL, 0); +} + +/** + * gfs2_mknod - Make a special file + * @dir: The directory in which the special file will reside + * @dentry: The dentry of the special file + * @mode: The mode of the special file + * @dev: The device specification of the special file + * + */ + +static int gfs2_mknod(struct inode *dir, struct dentry *dentry, int mode, + dev_t dev) +{ + return gfs2_create_inode(dir, dentry, mode, dev, NULL, 0); +} + +/* + * gfs2_ok_to_move - check if it's ok to move a directory to another directory + * @this: move this + * @to: to here + * + * Follow @to back to the root and make sure we don't encounter @this + * Assumes we already hold the rename lock. + * + * Returns: errno + */ + +static int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to) +{ + struct inode *dir = &to->i_inode; + struct super_block *sb = dir->i_sb; + struct inode *tmp; + int error = 0; + + igrab(dir); + + for (;;) { + if (dir == &this->i_inode) { + error = -EINVAL; + break; + } + if (dir == sb->s_root->d_inode) { + error = 0; + break; + } + + tmp = gfs2_lookupi(dir, &gfs2_qdotdot, 1); + if (IS_ERR(tmp)) { + error = PTR_ERR(tmp); + break; + } + + iput(dir); + dir = tmp; + } + + iput(dir); + + return error; +} + +/** + * gfs2_rename - Rename a file + * @odir: Parent directory of old file name + * @odentry: The old dentry of the file + * @ndir: Parent directory of new file name + * @ndentry: The new dentry of the file + * + * Returns: errno + */ + +static int gfs2_rename(struct inode *odir, struct dentry *odentry, + struct inode *ndir, struct dentry *ndentry) +{ + struct gfs2_inode *odip = GFS2_I(odir); + struct gfs2_inode *ndip = GFS2_I(ndir); + struct gfs2_inode *ip = GFS2_I(odentry->d_inode); + struct gfs2_inode *nip = NULL; + struct gfs2_sbd *sdp = GFS2_SB(odir); + struct gfs2_holder ghs[5], r_gh = { .gh_gl = NULL, }, ri_gh; + struct gfs2_rgrpd *nrgd; + unsigned int num_gh; + int dir_rename = 0; + int alloc_required = 0; + unsigned int x; + int error; + + if (ndentry->d_inode) { + nip = GFS2_I(ndentry->d_inode); + if (ip == nip) + return 0; + } + + error = gfs2_rindex_hold(sdp, &ri_gh); + if (error) + return error; + + if (odip != ndip) { + error = gfs2_glock_nq_init(sdp->sd_rename_gl, LM_ST_EXCLUSIVE, + 0, &r_gh); + if (error) + goto out; + + if (S_ISDIR(ip->i_inode.i_mode)) { + dir_rename = 1; + /* don't move a dirctory into it's subdir */ + error = gfs2_ok_to_move(ip, ndip); + if (error) + goto out_gunlock_r; + } + } + + num_gh = 1; + gfs2_holder_init(odip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); + if (odip != ndip) { + gfs2_holder_init(ndip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + } + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + + if (nip) { + gfs2_holder_init(nip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); + num_gh++; + /* grab the resource lock for unlink flag twiddling + * this is the case of the target file already existing + * so we unlink before doing the rename + */ + nrgd = gfs2_blk2rgrpd(sdp, nip->i_no_addr); + if (nrgd) + gfs2_holder_init(nrgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh++); + } + + for (x = 0; x < num_gh; x++) { + error = gfs2_glock_nq(ghs + x); + if (error) + goto out_gunlock; + } + + error = -ENOENT; + if (ip->i_inode.i_nlink == 0) + goto out_gunlock; + + /* Check out the old directory */ + + error = gfs2_unlink_ok(odip, &odentry->d_name, ip); + if (error) + goto out_gunlock; + + /* Check out the new directory */ + + if (nip) { + error = gfs2_unlink_ok(ndip, &ndentry->d_name, nip); + if (error) + goto out_gunlock; + + if (nip->i_inode.i_nlink == 0) { + error = -EAGAIN; + goto out_gunlock; + } + + if (S_ISDIR(nip->i_inode.i_mode)) { + if (nip->i_entries < 2) { + gfs2_consist_inode(nip); + error = -EIO; + goto out_gunlock; + } + if (nip->i_entries > 2) { + error = -ENOTEMPTY; + goto out_gunlock; + } + } + } else { + error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC, 0); + if (error) + goto out_gunlock; + + error = gfs2_dir_check(ndir, &ndentry->d_name, NULL); + switch (error) { + case -ENOENT: + error = 0; + break; + case 0: + error = -EEXIST; + default: + goto out_gunlock; + }; + + if (odip != ndip) { + if (!ndip->i_inode.i_nlink) { + error = -ENOENT; + goto out_gunlock; + } + if (ndip->i_entries == (u32)-1) { + error = -EFBIG; + goto out_gunlock; + } + if (S_ISDIR(ip->i_inode.i_mode) && + ndip->i_inode.i_nlink == (u32)-1) { + error = -EMLINK; + goto out_gunlock; + } + } + } + + /* Check out the dir to be renamed */ + + if (dir_rename) { + error = gfs2_permission(odentry->d_inode, MAY_WRITE, 0); + if (error) + goto out_gunlock; + } + + if (nip == NULL) + alloc_required = gfs2_diradd_alloc_required(ndir, &ndentry->d_name); + error = alloc_required; + if (error < 0) + goto out_gunlock; + error = 0; + + if (alloc_required) { + struct gfs2_alloc *al = gfs2_alloc_get(ndip); + if (!al) { + error = -ENOMEM; + goto out_gunlock; + } + + error = gfs2_quota_lock_check(ndip); + if (error) + goto out_alloc; + + al->al_requested = sdp->sd_max_dirres; + + error = gfs2_inplace_reserve_ri(ndip); + if (error) + goto out_gunlock_q; + + error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + + gfs2_rg_blocks(al) + + 4 * RES_DINODE + 4 * RES_LEAF + + RES_STATFS + RES_QUOTA + 4, 0); + if (error) + goto out_ipreserv; + } else { + error = gfs2_trans_begin(sdp, 4 * RES_DINODE + + 5 * RES_LEAF + 4, 0); + if (error) + goto out_gunlock; + } + + /* Remove the target file, if it exists */ + + if (nip) { + struct buffer_head *bh; + error = gfs2_meta_inode_buffer(nip, &bh); + if (error) + goto out_end_trans; + error = gfs2_unlink_inode(ndip, ndentry, bh); + brelse(bh); + } + + if (dir_rename) { + error = gfs2_dir_mvino(ip, &gfs2_qdotdot, ndip, DT_DIR); + if (error) + goto out_end_trans; + } else { + struct buffer_head *dibh; + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) + goto out_end_trans; + ip->i_inode.i_ctime = CURRENT_TIME; + gfs2_trans_add_bh(ip->i_gl, dibh, 1); + gfs2_dinode_out(ip, dibh->b_data); + brelse(dibh); + } + + error = gfs2_dir_del(odip, odentry); + if (error) + goto out_end_trans; + + error = gfs2_dir_add(ndir, &ndentry->d_name, ip); + if (error) + goto out_end_trans; + +out_end_trans: + gfs2_trans_end(sdp); +out_ipreserv: + if (alloc_required) + gfs2_inplace_release(ndip); +out_gunlock_q: + if (alloc_required) + gfs2_quota_unlock(ndip); +out_alloc: + if (alloc_required) + gfs2_alloc_put(ndip); +out_gunlock: + while (x--) { + gfs2_glock_dq(ghs + x); + gfs2_holder_uninit(ghs + x); + } +out_gunlock_r: + if (r_gh.gh_gl) + gfs2_glock_dq_uninit(&r_gh); +out: + gfs2_glock_dq_uninit(&ri_gh); + return error; +} + +/** + * gfs2_follow_link - Follow a symbolic link + * @dentry: The dentry of the link + * @nd: Data that we pass to vfs_follow_link() + * + * This can handle symlinks of any size. + * + * Returns: 0 on success or error code + */ + +static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd) +{ + struct gfs2_inode *ip = GFS2_I(dentry->d_inode); + struct gfs2_holder i_gh; + struct buffer_head *dibh; + unsigned int size; + char *buf; + int error; + + gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &i_gh); + error = gfs2_glock_nq(&i_gh); + if (error) { + gfs2_holder_uninit(&i_gh); + nd_set_link(nd, ERR_PTR(error)); + return NULL; + } + + size = (unsigned int)i_size_read(&ip->i_inode); + if (size == 0) { + gfs2_consist_inode(ip); + buf = ERR_PTR(-EIO); + goto out; + } + + error = gfs2_meta_inode_buffer(ip, &dibh); + if (error) { + buf = ERR_PTR(error); + goto out; + } + + buf = kzalloc(size + 1, GFP_NOFS); + if (!buf) + buf = ERR_PTR(-ENOMEM); + else + memcpy(buf, dibh->b_data + sizeof(struct gfs2_dinode), size); + brelse(dibh); +out: + gfs2_glock_dq_uninit(&i_gh); + nd_set_link(nd, buf); + return NULL; +} + +static void gfs2_put_link(struct dentry *dentry, struct nameidata *nd, void *p) +{ + char *s = nd_get_link(nd); + if (!IS_ERR(s)) + kfree(s); +} + +/** + * gfs2_permission - + * @inode: The inode + * @mask: The mask to be tested + * @flags: Indicates whether this is an RCU path walk or not + * + * This may be called from the VFS directly, or from within GFS2 with the + * inode locked, so we look to see if the glock is already locked and only + * lock the glock if its not already been done. + * + * Returns: errno + */ + +int gfs2_permission(struct inode *inode, int mask, unsigned int flags) +{ + struct gfs2_inode *ip; + struct gfs2_holder i_gh; + int error; + int unlock = 0; + + + ip = GFS2_I(inode); + if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { + if (flags & IPERM_FLAG_RCU) + return -ECHILD; + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); + if (error) + return error; + unlock = 1; + } + + if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode)) + error = -EACCES; + else + error = generic_permission(inode, mask, flags, gfs2_check_acl); + if (unlock) + gfs2_glock_dq_uninit(&i_gh); + + return error; } static int __gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr) @@ -902,8 +1594,6 @@ static int __gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr) * @ip: * @attr: * - * Called with a reference on the vnode. - * * Returns: errno */ @@ -923,60 +1613,280 @@ int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr) return error; } -void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) -{ - struct gfs2_dinode *str = buf; - - str->di_header.mh_magic = cpu_to_be32(GFS2_MAGIC); - str->di_header.mh_type = cpu_to_be32(GFS2_METATYPE_DI); - str->di_header.mh_format = cpu_to_be32(GFS2_FORMAT_DI); - str->di_num.no_addr = cpu_to_be64(ip->i_no_addr); - str->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino); - str->di_mode = cpu_to_be32(ip->i_inode.i_mode); - str->di_uid = cpu_to_be32(ip->i_inode.i_uid); - str->di_gid = cpu_to_be32(ip->i_inode.i_gid); - str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink); - str->di_size = cpu_to_be64(i_size_read(&ip->i_inode)); - str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); - str->di_atime = cpu_to_be64(ip->i_inode.i_atime.tv_sec); - str->di_mtime = cpu_to_be64(ip->i_inode.i_mtime.tv_sec); - str->di_ctime = cpu_to_be64(ip->i_inode.i_ctime.tv_sec); - - str->di_goal_meta = cpu_to_be64(ip->i_goal); - str->di_goal_data = cpu_to_be64(ip->i_goal); - str->di_generation = cpu_to_be64(ip->i_generation); - - str->di_flags = cpu_to_be32(ip->i_diskflags); - str->di_height = cpu_to_be16(ip->i_height); - str->di_payload_format = cpu_to_be32(S_ISDIR(ip->i_inode.i_mode) && - !(ip->i_diskflags & GFS2_DIF_EXHASH) ? - GFS2_FORMAT_DE : 0); - str->di_depth = cpu_to_be16(ip->i_depth); - str->di_entries = cpu_to_be32(ip->i_entries); - - str->di_eattr = cpu_to_be64(ip->i_eattr); - str->di_atime_nsec = cpu_to_be32(ip->i_inode.i_atime.tv_nsec); - str->di_mtime_nsec = cpu_to_be32(ip->i_inode.i_mtime.tv_nsec); - str->di_ctime_nsec = cpu_to_be32(ip->i_inode.i_ctime.tv_nsec); -} - -void gfs2_dinode_print(const struct gfs2_inode *ip) -{ - printk(KERN_INFO " no_formal_ino = %llu\n", - (unsigned long long)ip->i_no_formal_ino); - printk(KERN_INFO " no_addr = %llu\n", - (unsigned long long)ip->i_no_addr); - printk(KERN_INFO " i_size = %llu\n", - (unsigned long long)i_size_read(&ip->i_inode)); - printk(KERN_INFO " blocks = %llu\n", - (unsigned long long)gfs2_get_inode_blocks(&ip->i_inode)); - printk(KERN_INFO " i_goal = %llu\n", - (unsigned long long)ip->i_goal); - printk(KERN_INFO " i_diskflags = 0x%.8X\n", ip->i_diskflags); - printk(KERN_INFO " i_height = %u\n", ip->i_height); - printk(KERN_INFO " i_depth = %u\n", ip->i_depth); - printk(KERN_INFO " i_entries = %u\n", ip->i_entries); - printk(KERN_INFO " i_eattr = %llu\n", - (unsigned long long)ip->i_eattr); +static int setattr_chown(struct inode *inode, struct iattr *attr) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_sbd *sdp = GFS2_SB(inode); + u32 ouid, ogid, nuid, ngid; + int error; + + ouid = inode->i_uid; + ogid = inode->i_gid; + nuid = attr->ia_uid; + ngid = attr->ia_gid; + + if (!(attr->ia_valid & ATTR_UID) || ouid == nuid) + ouid = nuid = NO_QUOTA_CHANGE; + if (!(attr->ia_valid & ATTR_GID) || ogid == ngid) + ogid = ngid = NO_QUOTA_CHANGE; + + if (!gfs2_alloc_get(ip)) + return -ENOMEM; + + error = gfs2_quota_lock(ip, nuid, ngid); + if (error) + goto out_alloc; + + if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { + error = gfs2_quota_check(ip, nuid, ngid); + if (error) + goto out_gunlock_q; + } + + error = gfs2_trans_begin(sdp, RES_DINODE + 2 * RES_QUOTA, 0); + if (error) + goto out_gunlock_q; + + error = gfs2_setattr_simple(ip, attr); + if (error) + goto out_end_trans; + + if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { + u64 blocks = gfs2_get_inode_blocks(&ip->i_inode); + gfs2_quota_change(ip, -blocks, ouid, ogid); + gfs2_quota_change(ip, blocks, nuid, ngid); + } + +out_end_trans: + gfs2_trans_end(sdp); +out_gunlock_q: + gfs2_quota_unlock(ip); +out_alloc: + gfs2_alloc_put(ip); + return error; } +/** + * gfs2_setattr - Change attributes on an inode + * @dentry: The dentry which is changing + * @attr: The structure describing the change + * + * The VFS layer wants to change one or more of an inodes attributes. Write + * that change out to disk. + * + * Returns: errno + */ + +static int gfs2_setattr(struct dentry *dentry, struct iattr *attr) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder i_gh; + int error; + + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &i_gh); + if (error) + return error; + + error = -EPERM; + if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) + goto out; + + error = inode_change_ok(inode, attr); + if (error) + goto out; + + if (attr->ia_valid & ATTR_SIZE) + error = gfs2_setattr_size(inode, attr->ia_size); + else if (attr->ia_valid & (ATTR_UID | ATTR_GID)) + error = setattr_chown(inode, attr); + else if ((attr->ia_valid & ATTR_MODE) && IS_POSIXACL(inode)) + error = gfs2_acl_chmod(ip, attr); + else + error = gfs2_setattr_simple(ip, attr); + +out: + gfs2_glock_dq_uninit(&i_gh); + if (!error) + mark_inode_dirty(inode); + return error; +} + +/** + * gfs2_getattr - Read out an inode's attributes + * @mnt: The vfsmount the inode is being accessed from + * @dentry: The dentry to stat + * @stat: The inode's stats + * + * This may be called from the VFS directly, or from within GFS2 with the + * inode locked, so we look to see if the glock is already locked and only + * lock the glock if its not already been done. Note that its the NFS + * readdirplus operation which causes this to be called (from filldir) + * with the glock already held. + * + * Returns: errno + */ + +static int gfs2_getattr(struct vfsmount *mnt, struct dentry *dentry, + struct kstat *stat) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int error; + int unlock = 0; + + if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); + if (error) + return error; + unlock = 1; + } + + generic_fillattr(inode, stat); + if (unlock) + gfs2_glock_dq_uninit(&gh); + + return 0; +} + +static int gfs2_setxattr(struct dentry *dentry, const char *name, + const void *data, size_t size, int flags) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + ret = gfs2_glock_nq(&gh); + if (ret == 0) { + ret = generic_setxattr(dentry, name, data, size, flags); + gfs2_glock_dq(&gh); + } + gfs2_holder_uninit(&gh); + return ret; +} + +static ssize_t gfs2_getxattr(struct dentry *dentry, const char *name, + void *data, size_t size) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + gfs2_holder_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); + ret = gfs2_glock_nq(&gh); + if (ret == 0) { + ret = generic_getxattr(dentry, name, data, size); + gfs2_glock_dq(&gh); + } + gfs2_holder_uninit(&gh); + return ret; +} + +static int gfs2_removexattr(struct dentry *dentry, const char *name) +{ + struct inode *inode = dentry->d_inode; + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + ret = gfs2_glock_nq(&gh); + if (ret == 0) { + ret = generic_removexattr(dentry, name); + gfs2_glock_dq(&gh); + } + gfs2_holder_uninit(&gh); + return ret; +} + +static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + u64 start, u64 len) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC); + if (ret) + return ret; + + mutex_lock(&inode->i_mutex); + + ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh); + if (ret) + goto out; + + if (gfs2_is_stuffed(ip)) { + u64 phys = ip->i_no_addr << inode->i_blkbits; + u64 size = i_size_read(inode); + u32 flags = FIEMAP_EXTENT_LAST|FIEMAP_EXTENT_NOT_ALIGNED| + FIEMAP_EXTENT_DATA_INLINE; + phys += sizeof(struct gfs2_dinode); + phys += start; + if (start + len > size) + len = size - start; + if (start < size) + ret = fiemap_fill_next_extent(fieinfo, start, phys, + len, flags); + if (ret == 1) + ret = 0; + } else { + ret = __generic_block_fiemap(inode, fieinfo, start, len, + gfs2_block_map); + } + + gfs2_glock_dq_uninit(&gh); +out: + mutex_unlock(&inode->i_mutex); + return ret; +} + +const struct inode_operations gfs2_file_iops = { + .permission = gfs2_permission, + .setattr = gfs2_setattr, + .getattr = gfs2_getattr, + .setxattr = gfs2_setxattr, + .getxattr = gfs2_getxattr, + .listxattr = gfs2_listxattr, + .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, +}; + +const struct inode_operations gfs2_dir_iops = { + .create = gfs2_create, + .lookup = gfs2_lookup, + .link = gfs2_link, + .unlink = gfs2_unlink, + .symlink = gfs2_symlink, + .mkdir = gfs2_mkdir, + .rmdir = gfs2_unlink, + .mknod = gfs2_mknod, + .rename = gfs2_rename, + .permission = gfs2_permission, + .setattr = gfs2_setattr, + .getattr = gfs2_getattr, + .setxattr = gfs2_setxattr, + .getxattr = gfs2_getxattr, + .listxattr = gfs2_listxattr, + .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, +}; + +const struct inode_operations gfs2_symlink_iops = { + .readlink = generic_readlink, + .follow_link = gfs2_follow_link, + .put_link = gfs2_put_link, + .permission = gfs2_permission, + .setattr = gfs2_setattr, + .getattr = gfs2_getattr, + .setxattr = gfs2_setxattr, + .getxattr = gfs2_getxattr, + .listxattr = gfs2_listxattr, + .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, +}; + diff --git a/fs/gfs2/inode.h b/fs/gfs2/inode.h index 3e00a66e7cbd..31606076f701 100644 --- a/fs/gfs2/inode.h +++ b/fs/gfs2/inode.h @@ -97,26 +97,21 @@ err: } extern struct inode *gfs2_inode_lookup(struct super_block *sb, unsigned type, - u64 no_addr, u64 no_formal_ino); + u64 no_addr, u64 no_formal_ino, + int non_block); extern struct inode *gfs2_lookup_by_inum(struct gfs2_sbd *sdp, u64 no_addr, u64 *no_formal_ino, unsigned int blktype); -extern struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr); +extern struct inode *gfs2_ilookup(struct super_block *sb, u64 no_addr, int nonblock); extern int gfs2_inode_refresh(struct gfs2_inode *ip); -extern int gfs2_dinode_dealloc(struct gfs2_inode *inode); -extern int gfs2_change_nlink(struct gfs2_inode *ip, int diff); extern struct inode *gfs2_lookupi(struct inode *dir, const struct qstr *name, int is_root); -extern struct inode *gfs2_createi(struct gfs2_holder *ghs, - const struct qstr *name, - unsigned int mode, dev_t dev); extern int gfs2_permission(struct inode *inode, int mask, unsigned int flags); extern int gfs2_setattr_simple(struct gfs2_inode *ip, struct iattr *attr); extern struct inode *gfs2_lookup_simple(struct inode *dip, const char *name); extern void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf); -extern void gfs2_dinode_print(const struct gfs2_inode *ip); extern const struct inode_operations gfs2_file_iops; extern const struct inode_operations gfs2_dir_iops; diff --git a/fs/gfs2/log.c b/fs/gfs2/log.c index 5b102c1887fd..903115f2bb34 100644 --- a/fs/gfs2/log.c +++ b/fs/gfs2/log.c @@ -18,6 +18,7 @@ #include <linux/kthread.h> #include <linux/freezer.h> #include <linux/bio.h> +#include <linux/writeback.h> #include "gfs2.h" #include "incore.h" @@ -83,55 +84,97 @@ void gfs2_remove_from_ail(struct gfs2_bufdata *bd) /** * gfs2_ail1_start_one - Start I/O on a part of the AIL * @sdp: the filesystem - * @tr: the part of the AIL + * @wbc: The writeback control structure + * @ai: The ail structure * */ -static void gfs2_ail1_start_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai) +static int gfs2_ail1_start_one(struct gfs2_sbd *sdp, + struct writeback_control *wbc, + struct gfs2_ail *ai) __releases(&sdp->sd_ail_lock) __acquires(&sdp->sd_ail_lock) { + struct gfs2_glock *gl = NULL; + struct address_space *mapping; struct gfs2_bufdata *bd, *s; struct buffer_head *bh; - int retry; - do { - retry = 0; + list_for_each_entry_safe_reverse(bd, s, &ai->ai_ail1_list, bd_ail_st_list) { + bh = bd->bd_bh; - list_for_each_entry_safe_reverse(bd, s, &ai->ai_ail1_list, - bd_ail_st_list) { - bh = bd->bd_bh; + gfs2_assert(sdp, bd->bd_ail == ai); - gfs2_assert(sdp, bd->bd_ail == ai); + if (!buffer_busy(bh)) { + if (!buffer_uptodate(bh)) + gfs2_io_error_bh(sdp, bh); + list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list); + continue; + } - if (!buffer_busy(bh)) { - if (!buffer_uptodate(bh)) - gfs2_io_error_bh(sdp, bh); - list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list); - continue; - } + if (!buffer_dirty(bh)) + continue; + if (gl == bd->bd_gl) + continue; + gl = bd->bd_gl; + list_move(&bd->bd_ail_st_list, &ai->ai_ail1_list); + mapping = bh->b_page->mapping; + if (!mapping) + continue; + spin_unlock(&sdp->sd_ail_lock); + generic_writepages(mapping, wbc); + spin_lock(&sdp->sd_ail_lock); + if (wbc->nr_to_write <= 0) + break; + return 1; + } - if (!buffer_dirty(bh)) - continue; + return 0; +} - list_move(&bd->bd_ail_st_list, &ai->ai_ail1_list); - get_bh(bh); - spin_unlock(&sdp->sd_ail_lock); - lock_buffer(bh); - if (test_clear_buffer_dirty(bh)) { - bh->b_end_io = end_buffer_write_sync; - submit_bh(WRITE_SYNC, bh); - } else { - unlock_buffer(bh); - brelse(bh); - } - spin_lock(&sdp->sd_ail_lock); - - retry = 1; +/** + * gfs2_ail1_flush - start writeback of some ail1 entries + * @sdp: The super block + * @wbc: The writeback control structure + * + * Writes back some ail1 entries, according to the limits in the + * writeback control structure + */ + +void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc) +{ + struct list_head *head = &sdp->sd_ail1_list; + struct gfs2_ail *ai; + + trace_gfs2_ail_flush(sdp, wbc, 1); + spin_lock(&sdp->sd_ail_lock); +restart: + list_for_each_entry_reverse(ai, head, ai_list) { + if (wbc->nr_to_write <= 0) break; - } - } while (retry); + if (gfs2_ail1_start_one(sdp, wbc, ai)) + goto restart; + } + spin_unlock(&sdp->sd_ail_lock); + trace_gfs2_ail_flush(sdp, wbc, 0); +} + +/** + * gfs2_ail1_start - start writeback of all ail1 entries + * @sdp: The superblock + */ + +static void gfs2_ail1_start(struct gfs2_sbd *sdp) +{ + struct writeback_control wbc = { + .sync_mode = WB_SYNC_NONE, + .nr_to_write = LONG_MAX, + .range_start = 0, + .range_end = LLONG_MAX, + }; + + return gfs2_ail1_flush(sdp, &wbc); } /** @@ -141,7 +184,7 @@ __acquires(&sdp->sd_ail_lock) * */ -static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai, int flags) +static void gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai) { struct gfs2_bufdata *bd, *s; struct buffer_head *bh; @@ -149,76 +192,63 @@ static int gfs2_ail1_empty_one(struct gfs2_sbd *sdp, struct gfs2_ail *ai, int fl list_for_each_entry_safe_reverse(bd, s, &ai->ai_ail1_list, bd_ail_st_list) { bh = bd->bd_bh; - gfs2_assert(sdp, bd->bd_ail == ai); - - if (buffer_busy(bh)) { - if (flags & DIO_ALL) - continue; - else - break; - } - + if (buffer_busy(bh)) + continue; if (!buffer_uptodate(bh)) gfs2_io_error_bh(sdp, bh); - list_move(&bd->bd_ail_st_list, &ai->ai_ail2_list); } - return list_empty(&ai->ai_ail1_list); } -static void gfs2_ail1_start(struct gfs2_sbd *sdp) -{ - struct list_head *head; - u64 sync_gen; - struct gfs2_ail *ai; - int done = 0; - - spin_lock(&sdp->sd_ail_lock); - head = &sdp->sd_ail1_list; - if (list_empty(head)) { - spin_unlock(&sdp->sd_ail_lock); - return; - } - sync_gen = sdp->sd_ail_sync_gen++; - - while(!done) { - done = 1; - list_for_each_entry_reverse(ai, head, ai_list) { - if (ai->ai_sync_gen >= sync_gen) - continue; - ai->ai_sync_gen = sync_gen; - gfs2_ail1_start_one(sdp, ai); /* This may drop ail lock */ - done = 0; - break; - } - } - - spin_unlock(&sdp->sd_ail_lock); -} +/** + * gfs2_ail1_empty - Try to empty the ail1 lists + * @sdp: The superblock + * + * Tries to empty the ail1 lists, starting with the oldest first + */ -static int gfs2_ail1_empty(struct gfs2_sbd *sdp, int flags) +static int gfs2_ail1_empty(struct gfs2_sbd *sdp) { struct gfs2_ail *ai, *s; int ret; spin_lock(&sdp->sd_ail_lock); - list_for_each_entry_safe_reverse(ai, s, &sdp->sd_ail1_list, ai_list) { - if (gfs2_ail1_empty_one(sdp, ai, flags)) + gfs2_ail1_empty_one(sdp, ai); + if (list_empty(&ai->ai_ail1_list)) list_move(&ai->ai_list, &sdp->sd_ail2_list); - else if (!(flags & DIO_ALL)) + else break; } - ret = list_empty(&sdp->sd_ail1_list); - spin_unlock(&sdp->sd_ail_lock); return ret; } +static void gfs2_ail1_wait(struct gfs2_sbd *sdp) +{ + struct gfs2_ail *ai; + struct gfs2_bufdata *bd; + struct buffer_head *bh; + + spin_lock(&sdp->sd_ail_lock); + list_for_each_entry_reverse(ai, &sdp->sd_ail1_list, ai_list) { + list_for_each_entry(bd, &ai->ai_ail1_list, bd_ail_st_list) { + bh = bd->bd_bh; + if (!buffer_locked(bh)) + continue; + get_bh(bh); + spin_unlock(&sdp->sd_ail_lock); + wait_on_buffer(bh); + brelse(bh); + return; + } + } + spin_unlock(&sdp->sd_ail_lock); +} /** * gfs2_ail2_empty_one - Check whether or not a trans in the AIL has been synced @@ -574,7 +604,7 @@ static void log_write_header(struct gfs2_sbd *sdp, u32 flags, int pull) set_buffer_uptodate(bh); clear_buffer_dirty(bh); - gfs2_ail1_empty(sdp, 0); + gfs2_ail1_empty(sdp); tail = current_tail(sdp); lh = (struct gfs2_log_header *)bh->b_data; @@ -869,9 +899,9 @@ void gfs2_meta_syncfs(struct gfs2_sbd *sdp) gfs2_log_flush(sdp, NULL); for (;;) { gfs2_ail1_start(sdp); - if (gfs2_ail1_empty(sdp, DIO_ALL)) + gfs2_ail1_wait(sdp); + if (gfs2_ail1_empty(sdp)) break; - msleep(10); } } @@ -905,20 +935,20 @@ int gfs2_logd(void *data) preflush = atomic_read(&sdp->sd_log_pinned); if (gfs2_jrnl_flush_reqd(sdp) || t == 0) { - gfs2_ail1_empty(sdp, DIO_ALL); + gfs2_ail1_empty(sdp); gfs2_log_flush(sdp, NULL); - gfs2_ail1_empty(sdp, DIO_ALL); } if (gfs2_ail_flush_reqd(sdp)) { gfs2_ail1_start(sdp); - io_schedule(); - gfs2_ail1_empty(sdp, 0); + gfs2_ail1_wait(sdp); + gfs2_ail1_empty(sdp); gfs2_log_flush(sdp, NULL); - gfs2_ail1_empty(sdp, DIO_ALL); } - wake_up(&sdp->sd_log_waitq); + if (!gfs2_ail_flush_reqd(sdp)) + wake_up(&sdp->sd_log_waitq); + t = gfs2_tune_get(sdp, gt_logd_secs) * HZ; if (freezing(current)) refrigerator(); diff --git a/fs/gfs2/log.h b/fs/gfs2/log.h index 0d007f920234..ab0621698b73 100644 --- a/fs/gfs2/log.h +++ b/fs/gfs2/log.h @@ -12,6 +12,7 @@ #include <linux/list.h> #include <linux/spinlock.h> +#include <linux/writeback.h> #include "incore.h" /** @@ -59,6 +60,7 @@ extern struct buffer_head *gfs2_log_fake_buf(struct gfs2_sbd *sdp, extern void gfs2_log_flush(struct gfs2_sbd *sdp, struct gfs2_glock *gl); extern void gfs2_log_commit(struct gfs2_sbd *sdp, struct gfs2_trans *trans); extern void gfs2_remove_from_ail(struct gfs2_bufdata *bd); +extern void gfs2_ail1_flush(struct gfs2_sbd *sdp, struct writeback_control *wbc); extern void gfs2_log_shutdown(struct gfs2_sbd *sdp); extern void gfs2_meta_syncfs(struct gfs2_sbd *sdp); diff --git a/fs/gfs2/lops.c b/fs/gfs2/lops.c index 51d27f00ebb4..05bbb124699f 100644 --- a/fs/gfs2/lops.c +++ b/fs/gfs2/lops.c @@ -40,7 +40,7 @@ static void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) { struct gfs2_bufdata *bd; - gfs2_assert_withdraw(sdp, test_bit(SDF_JOURNAL_LIVE, &sdp->sd_flags)); + BUG_ON(!current->journal_info); clear_buffer_dirty(bh); if (test_set_buffer_pinned(bh)) @@ -65,6 +65,7 @@ static void gfs2_pin(struct gfs2_sbd *sdp, struct buffer_head *bh) * @sdp: the filesystem the buffer belongs to * @bh: The buffer to unpin * @ai: + * @flags: The inode dirty flags * */ @@ -73,10 +74,8 @@ static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh, { struct gfs2_bufdata *bd = bh->b_private; - gfs2_assert_withdraw(sdp, buffer_uptodate(bh)); - - if (!buffer_pinned(bh)) - gfs2_assert_withdraw(sdp, 0); + BUG_ON(!buffer_uptodate(bh)); + BUG_ON(!buffer_pinned(bh)); lock_buffer(bh); mark_buffer_dirty(bh); @@ -95,8 +94,7 @@ static void gfs2_unpin(struct gfs2_sbd *sdp, struct buffer_head *bh, list_add(&bd->bd_ail_st_list, &ai->ai_ail1_list); spin_unlock(&sdp->sd_ail_lock); - if (test_and_clear_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags)) - gfs2_glock_schedule_for_reclaim(bd->bd_gl); + clear_bit(GLF_LFLUSH, &bd->bd_gl->gl_flags); trace_gfs2_pin(bd, 0); unlock_buffer(bh); atomic_dec(&sdp->sd_log_pinned); @@ -322,12 +320,16 @@ static void buf_lo_after_scan(struct gfs2_jdesc *jd, int error, int pass) static void revoke_lo_add(struct gfs2_sbd *sdp, struct gfs2_log_element *le) { + struct gfs2_bufdata *bd = container_of(le, struct gfs2_bufdata, bd_le); + struct gfs2_glock *gl = bd->bd_gl; struct gfs2_trans *tr; tr = current->journal_info; tr->tr_touched = 1; tr->tr_num_revoke++; sdp->sd_log_num_revoke++; + atomic_inc(&gl->gl_revokes); + set_bit(GLF_LFLUSH, &gl->gl_flags); list_add(&le->le_list, &sdp->sd_log_le_revoke); } @@ -350,9 +352,7 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp) ld->ld_data1 = cpu_to_be32(sdp->sd_log_num_revoke); offset = sizeof(struct gfs2_log_descriptor); - while (!list_empty(head)) { - bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list); - list_del_init(&bd->bd_le.le_list); + list_for_each_entry(bd, head, bd_le.le_list) { sdp->sd_log_num_revoke--; if (offset + sizeof(u64) > sdp->sd_sb.sb_bsize) { @@ -367,8 +367,6 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp) } *(__be64 *)(bh->b_data + offset) = cpu_to_be64(bd->bd_blkno); - kmem_cache_free(gfs2_bufdata_cachep, bd); - offset += sizeof(u64); } gfs2_assert_withdraw(sdp, !sdp->sd_log_num_revoke); @@ -376,6 +374,22 @@ static void revoke_lo_before_commit(struct gfs2_sbd *sdp) submit_bh(WRITE_SYNC, bh); } +static void revoke_lo_after_commit(struct gfs2_sbd *sdp, struct gfs2_ail *ai) +{ + struct list_head *head = &sdp->sd_log_le_revoke; + struct gfs2_bufdata *bd; + struct gfs2_glock *gl; + + while (!list_empty(head)) { + bd = list_entry(head->next, struct gfs2_bufdata, bd_le.le_list); + list_del_init(&bd->bd_le.le_list); + gl = bd->bd_gl; + atomic_dec(&gl->gl_revokes); + clear_bit(GLF_LFLUSH, &gl->gl_flags); + kmem_cache_free(gfs2_bufdata_cachep, bd); + } +} + static void revoke_lo_before_scan(struct gfs2_jdesc *jd, struct gfs2_log_header_host *head, int pass) { @@ -749,6 +763,7 @@ const struct gfs2_log_operations gfs2_buf_lops = { const struct gfs2_log_operations gfs2_revoke_lops = { .lo_add = revoke_lo_add, .lo_before_commit = revoke_lo_before_commit, + .lo_after_commit = revoke_lo_after_commit, .lo_before_scan = revoke_lo_before_scan, .lo_scan_elements = revoke_lo_scan_elements, .lo_after_scan = revoke_lo_after_scan, diff --git a/fs/gfs2/main.c b/fs/gfs2/main.c index 888a5f5a1a58..c2b34cd2abe0 100644 --- a/fs/gfs2/main.c +++ b/fs/gfs2/main.c @@ -53,6 +53,7 @@ static void gfs2_init_glock_once(void *foo) INIT_LIST_HEAD(&gl->gl_lru); INIT_LIST_HEAD(&gl->gl_ail_list); atomic_set(&gl->gl_ail_count, 0); + atomic_set(&gl->gl_revokes, 0); } static void gfs2_init_gl_aspace_once(void *foo) @@ -145,7 +146,7 @@ static int __init init_gfs2_fs(void) gfs2_register_debugfs(); - printk("GFS2 (built %s %s) installed\n", __DATE__, __TIME__); + printk("GFS2 installed\n"); return 0; diff --git a/fs/gfs2/meta_io.c b/fs/gfs2/meta_io.c index 675349b5a133..747238cd9f96 100644 --- a/fs/gfs2/meta_io.c +++ b/fs/gfs2/meta_io.c @@ -31,6 +31,7 @@ #include "rgrp.h" #include "trans.h" #include "util.h" +#include "trace_gfs2.h" static int gfs2_aspace_writepage(struct page *page, struct writeback_control *wbc) { @@ -310,6 +311,7 @@ void gfs2_remove_from_journal(struct buffer_head *bh, struct gfs2_trans *tr, int struct gfs2_bufdata *bd = bh->b_private; if (test_clear_buffer_pinned(bh)) { + trace_gfs2_pin(bd, 0); atomic_dec(&sdp->sd_log_pinned); list_del_init(&bd->bd_le.le_list); if (meta) { diff --git a/fs/gfs2/meta_io.h b/fs/gfs2/meta_io.h index 6a1d9ba16411..22c526593131 100644 --- a/fs/gfs2/meta_io.h +++ b/fs/gfs2/meta_io.h @@ -77,8 +77,6 @@ struct buffer_head *gfs2_meta_ra(struct gfs2_glock *gl, u64 dblock, u32 extlen); #define buffer_busy(bh) \ ((bh)->b_state & ((1ul << BH_Dirty) | (1ul << BH_Lock) | (1ul << BH_Pinned))) -#define buffer_in_io(bh) \ -((bh)->b_state & ((1ul << BH_Dirty) | (1ul << BH_Lock))) #endif /* __DIO_DOT_H__ */ diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index 42ef24355afb..8ac9ae189b53 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -126,8 +126,10 @@ static struct gfs2_sbd *init_sbd(struct super_block *sb) * changed. */ -static int gfs2_check_sb(struct gfs2_sbd *sdp, struct gfs2_sb_host *sb, int silent) +static int gfs2_check_sb(struct gfs2_sbd *sdp, int silent) { + struct gfs2_sb_host *sb = &sdp->sd_sb; + if (sb->sb_magic != GFS2_MAGIC || sb->sb_type != GFS2_METATYPE_SB) { if (!silent) @@ -157,8 +159,10 @@ static void end_bio_io_page(struct bio *bio, int error) unlock_page(page); } -static void gfs2_sb_in(struct gfs2_sb_host *sb, const void *buf) +static void gfs2_sb_in(struct gfs2_sbd *sdp, const void *buf) { + struct gfs2_sb_host *sb = &sdp->sd_sb; + struct super_block *s = sdp->sd_vfs; const struct gfs2_sb *str = buf; sb->sb_magic = be32_to_cpu(str->sb_header.mh_magic); @@ -175,7 +179,7 @@ static void gfs2_sb_in(struct gfs2_sb_host *sb, const void *buf) memcpy(sb->sb_lockproto, str->sb_lockproto, GFS2_LOCKNAME_LEN); memcpy(sb->sb_locktable, str->sb_locktable, GFS2_LOCKNAME_LEN); - memcpy(sb->sb_uuid, str->sb_uuid, 16); + memcpy(s->s_uuid, str->sb_uuid, 16); } /** @@ -197,7 +201,7 @@ static void gfs2_sb_in(struct gfs2_sb_host *sb, const void *buf) * Returns: 0 on success or error */ -static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector) +static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector, int silent) { struct super_block *sb = sdp->sd_vfs; struct gfs2_sb *p; @@ -227,10 +231,10 @@ static int gfs2_read_super(struct gfs2_sbd *sdp, sector_t sector) return -EIO; } p = kmap(page); - gfs2_sb_in(&sdp->sd_sb, p); + gfs2_sb_in(sdp, p); kunmap(page); __free_page(page); - return 0; + return gfs2_check_sb(sdp, silent); } /** @@ -247,17 +251,13 @@ static int gfs2_read_sb(struct gfs2_sbd *sdp, int silent) unsigned int x; int error; - error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift); + error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift, silent); if (error) { if (!silent) fs_err(sdp, "can't read superblock\n"); return error; } - error = gfs2_check_sb(sdp, &sdp->sd_sb, silent); - if (error) - return error; - sdp->sd_fsb2bb_shift = sdp->sd_sb.sb_bsize_shift - GFS2_BASIC_BLOCK_SHIFT; sdp->sd_fsb2bb = 1 << sdp->sd_fsb2bb_shift; @@ -340,14 +340,10 @@ static int init_names(struct gfs2_sbd *sdp, int silent) /* Try to autodetect */ if (!proto[0] || !table[0]) { - error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift); + error = gfs2_read_super(sdp, GFS2_SB_ADDR >> sdp->sd_fsb2bb_shift, silent); if (error) return error; - error = gfs2_check_sb(sdp, &sdp->sd_sb, silent); - if (error) - goto out; - if (!proto[0]) proto = sdp->sd_sb.sb_lockproto; if (!table[0]) @@ -364,7 +360,6 @@ static int init_names(struct gfs2_sbd *sdp, int silent) while ((table = strchr(table, '/'))) *table = '_'; -out: return error; } @@ -430,7 +425,7 @@ static int gfs2_lookup_root(struct super_block *sb, struct dentry **dptr, struct dentry *dentry; struct inode *inode; - inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0); + inode = gfs2_inode_lookup(sb, DT_DIR, no_addr, 0, 0); if (IS_ERR(inode)) { fs_err(sdp, "can't read in %s inode: %ld\n", name, PTR_ERR(inode)); return PTR_ERR(inode); @@ -1119,8 +1114,7 @@ static int fill_super(struct super_block *sb, struct gfs2_args *args, int silent if (sdp->sd_args.ar_statfs_quantum) { sdp->sd_tune.gt_statfs_slow = 0; sdp->sd_tune.gt_statfs_quantum = sdp->sd_args.ar_statfs_quantum; - } - else { + } else { sdp->sd_tune.gt_statfs_slow = 1; sdp->sd_tune.gt_statfs_quantum = 30; } diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c deleted file mode 100644 index 09e436a50723..000000000000 --- a/fs/gfs2/ops_inode.c +++ /dev/null @@ -1,1344 +0,0 @@ -/* - * Copyright (C) Sistina Software, Inc. 1997-2003 All rights reserved. - * Copyright (C) 2004-2006 Red Hat, Inc. All rights reserved. - * - * This copyrighted material is made available to anyone wishing to use, - * modify, copy, or redistribute it subject to the terms and conditions - * of the GNU General Public License version 2. - */ - -#include <linux/slab.h> -#include <linux/spinlock.h> -#include <linux/completion.h> -#include <linux/buffer_head.h> -#include <linux/namei.h> -#include <linux/mm.h> -#include <linux/xattr.h> -#include <linux/posix_acl.h> -#include <linux/gfs2_ondisk.h> -#include <linux/crc32.h> -#include <linux/fiemap.h> -#include <asm/uaccess.h> - -#include "gfs2.h" -#include "incore.h" -#include "acl.h" -#include "bmap.h" -#include "dir.h" -#include "xattr.h" -#include "glock.h" -#include "inode.h" -#include "meta_io.h" -#include "quota.h" -#include "rgrp.h" -#include "trans.h" -#include "util.h" -#include "super.h" - -/** - * gfs2_create - Create a file - * @dir: The directory in which to create the file - * @dentry: The dentry of the new file - * @mode: The mode of the new file - * - * Returns: errno - */ - -static int gfs2_create(struct inode *dir, struct dentry *dentry, - int mode, struct nameidata *nd) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - for (;;) { - inode = gfs2_createi(ghs, &dentry->d_name, S_IFREG | mode, 0); - if (!IS_ERR(inode)) { - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - gfs2_glock_dq_uninit_m(2, ghs); - mark_inode_dirty(inode); - break; - } else if (PTR_ERR(inode) != -EEXIST || - (nd && nd->flags & LOOKUP_EXCL)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - inode = gfs2_lookupi(dir, &dentry->d_name, 0); - if (inode) { - if (!IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - break; - } else { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - } - } - - d_instantiate(dentry, inode); - - return 0; -} - -/** - * gfs2_lookup - Look up a filename in a directory and return its inode - * @dir: The directory inode - * @dentry: The dentry of the new inode - * @nd: passed from Linux VFS, ignored by us - * - * Called by the VFS layer. Lock dir and call gfs2_lookupi() - * - * Returns: errno - */ - -static struct dentry *gfs2_lookup(struct inode *dir, struct dentry *dentry, - struct nameidata *nd) -{ - struct inode *inode = NULL; - - inode = gfs2_lookupi(dir, &dentry->d_name, 0); - if (inode && IS_ERR(inode)) - return ERR_CAST(inode); - - if (inode) { - struct gfs2_glock *gl = GFS2_I(inode)->i_gl; - struct gfs2_holder gh; - int error; - error = gfs2_glock_nq_init(gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); - if (error) { - iput(inode); - return ERR_PTR(error); - } - gfs2_glock_dq_uninit(&gh); - return d_splice_alias(inode, dentry); - } - d_add(dentry, inode); - - return NULL; -} - -/** - * gfs2_link - Link to a file - * @old_dentry: The inode to link - * @dir: Add link to this directory - * @dentry: The name of the link - * - * Link the inode in "old_dentry" into the directory "dir" with the - * name in "dentry". - * - * Returns: errno - */ - -static int gfs2_link(struct dentry *old_dentry, struct inode *dir, - struct dentry *dentry) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct inode *inode = old_dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder ghs[2]; - int alloc_required; - int error; - - if (S_ISDIR(inode->i_mode)) - return -EPERM; - - gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); - - error = gfs2_glock_nq(ghs); /* parent */ - if (error) - goto out_parent; - - error = gfs2_glock_nq(ghs + 1); /* child */ - if (error) - goto out_child; - - error = gfs2_permission(dir, MAY_WRITE | MAY_EXEC, 0); - if (error) - goto out_gunlock; - - error = gfs2_dir_check(dir, &dentry->d_name, NULL); - switch (error) { - case -ENOENT: - break; - case 0: - error = -EEXIST; - default: - goto out_gunlock; - } - - error = -EINVAL; - if (!dip->i_inode.i_nlink) - goto out_gunlock; - error = -EFBIG; - if (dip->i_entries == (u32)-1) - goto out_gunlock; - error = -EPERM; - if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) - goto out_gunlock; - error = -EINVAL; - if (!ip->i_inode.i_nlink) - goto out_gunlock; - error = -EMLINK; - if (ip->i_inode.i_nlink == (u32)-1) - goto out_gunlock; - - alloc_required = error = gfs2_diradd_alloc_required(dir, &dentry->d_name); - if (error < 0) - goto out_gunlock; - error = 0; - - if (alloc_required) { - struct gfs2_alloc *al = gfs2_alloc_get(dip); - if (!al) { - error = -ENOMEM; - goto out_gunlock; - } - - error = gfs2_quota_lock_check(dip); - if (error) - goto out_alloc; - - al->al_requested = sdp->sd_max_dirres; - - error = gfs2_inplace_reserve(dip); - if (error) - goto out_gunlock_q; - - error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + - gfs2_rg_blocks(al) + - 2 * RES_DINODE + RES_STATFS + - RES_QUOTA, 0); - if (error) - goto out_ipres; - } else { - error = gfs2_trans_begin(sdp, 2 * RES_DINODE + RES_LEAF, 0); - if (error) - goto out_ipres; - } - - error = gfs2_dir_add(dir, &dentry->d_name, ip, IF2DT(inode->i_mode)); - if (error) - goto out_end_trans; - - error = gfs2_change_nlink(ip, +1); - -out_end_trans: - gfs2_trans_end(sdp); -out_ipres: - if (alloc_required) - gfs2_inplace_release(dip); -out_gunlock_q: - if (alloc_required) - gfs2_quota_unlock(dip); -out_alloc: - if (alloc_required) - gfs2_alloc_put(dip); -out_gunlock: - gfs2_glock_dq(ghs + 1); -out_child: - gfs2_glock_dq(ghs); -out_parent: - gfs2_holder_uninit(ghs); - gfs2_holder_uninit(ghs + 1); - if (!error) { - ihold(inode); - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - } - return error; -} - -/* - * gfs2_unlink_ok - check to see that a inode is still in a directory - * @dip: the directory - * @name: the name of the file - * @ip: the inode - * - * Assumes that the lock on (at least) @dip is held. - * - * Returns: 0 if the parent/child relationship is correct, errno if it isn't - */ - -static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, - const struct gfs2_inode *ip) -{ - int error; - - if (IS_IMMUTABLE(&ip->i_inode) || IS_APPEND(&ip->i_inode)) - return -EPERM; - - if ((dip->i_inode.i_mode & S_ISVTX) && - dip->i_inode.i_uid != current_fsuid() && - ip->i_inode.i_uid != current_fsuid() && !capable(CAP_FOWNER)) - return -EPERM; - - if (IS_APPEND(&dip->i_inode)) - return -EPERM; - - error = gfs2_permission(&dip->i_inode, MAY_WRITE | MAY_EXEC, 0); - if (error) - return error; - - error = gfs2_dir_check(&dip->i_inode, name, ip); - if (error) - return error; - - return 0; -} - -/** - * gfs2_unlink - Unlink a file - * @dir: The inode of the directory containing the file to unlink - * @dentry: The file itself - * - * Unlink a file. Call gfs2_unlinki() - * - * Returns: errno - */ - -static int gfs2_unlink(struct inode *dir, struct dentry *dentry) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); - struct gfs2_holder ghs[3]; - struct gfs2_rgrpd *rgd; - struct gfs2_holder ri_gh; - int error; - - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - return error; - - gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); - - rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); - gfs2_holder_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + 2); - - - error = gfs2_glock_nq(ghs); /* parent */ - if (error) - goto out_parent; - - error = gfs2_glock_nq(ghs + 1); /* child */ - if (error) - goto out_child; - - error = gfs2_glock_nq(ghs + 2); /* rgrp */ - if (error) - goto out_rgrp; - - error = gfs2_unlink_ok(dip, &dentry->d_name, ip); - if (error) - goto out_gunlock; - - error = gfs2_trans_begin(sdp, 2*RES_DINODE + RES_LEAF + RES_RG_BIT, 0); - if (error) - goto out_gunlock; - - error = gfs2_dir_del(dip, &dentry->d_name); - if (error) - goto out_end_trans; - - error = gfs2_change_nlink(ip, -1); - -out_end_trans: - gfs2_trans_end(sdp); -out_gunlock: - gfs2_glock_dq(ghs + 2); -out_rgrp: - gfs2_holder_uninit(ghs + 2); - gfs2_glock_dq(ghs + 1); -out_child: - gfs2_holder_uninit(ghs + 1); - gfs2_glock_dq(ghs); -out_parent: - gfs2_holder_uninit(ghs); - gfs2_glock_dq_uninit(&ri_gh); - return error; -} - -/** - * gfs2_symlink - Create a symlink - * @dir: The directory to create the symlink in - * @dentry: The dentry to put the symlink in - * @symname: The thing which the link points to - * - * Returns: errno - */ - -static int gfs2_symlink(struct inode *dir, struct dentry *dentry, - const char *symname) -{ - struct gfs2_inode *dip = GFS2_I(dir), *ip; - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - struct buffer_head *dibh; - int size; - int error; - - /* Must be stuffed with a null terminator for gfs2_follow_link() */ - size = strlen(symname); - if (size > sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode) - 1) - return -ENAMETOOLONG; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - inode = gfs2_createi(ghs, &dentry->d_name, S_IFLNK | S_IRWXUGO, 0); - if (IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - ip = ghs[1].gh_gl->gl_object; - - i_size_write(inode, size); - - error = gfs2_meta_inode_buffer(ip, &dibh); - - if (!gfs2_assert_withdraw(sdp, !error)) { - gfs2_dinode_out(ip, dibh->b_data); - memcpy(dibh->b_data + sizeof(struct gfs2_dinode), symname, - size); - brelse(dibh); - } - - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - - gfs2_glock_dq_uninit_m(2, ghs); - - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - - return 0; -} - -/** - * gfs2_mkdir - Make a directory - * @dir: The parent directory of the new one - * @dentry: The dentry of the new directory - * @mode: The mode of the new directory - * - * Returns: errno - */ - -static int gfs2_mkdir(struct inode *dir, struct dentry *dentry, int mode) -{ - struct gfs2_inode *dip = GFS2_I(dir), *ip; - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - struct buffer_head *dibh; - int error; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - inode = gfs2_createi(ghs, &dentry->d_name, S_IFDIR | mode, 0); - if (IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - ip = ghs[1].gh_gl->gl_object; - - ip->i_inode.i_nlink = 2; - i_size_write(inode, sdp->sd_sb.sb_bsize - sizeof(struct gfs2_dinode)); - ip->i_diskflags |= GFS2_DIF_JDATA; - ip->i_entries = 2; - - error = gfs2_meta_inode_buffer(ip, &dibh); - - if (!gfs2_assert_withdraw(sdp, !error)) { - struct gfs2_dinode *di = (struct gfs2_dinode *)dibh->b_data; - struct gfs2_dirent *dent = (struct gfs2_dirent *)(di+1); - - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_qstr2dirent(&gfs2_qdot, GFS2_DIRENT_SIZE(gfs2_qdot.len), dent); - dent->de_inum = di->di_num; /* already GFS2 endian */ - dent->de_type = cpu_to_be16(DT_DIR); - di->di_entries = cpu_to_be32(1); - - dent = (struct gfs2_dirent *)((char*)dent + GFS2_DIRENT_SIZE(1)); - gfs2_qstr2dirent(&gfs2_qdotdot, dibh->b_size - GFS2_DIRENT_SIZE(1) - sizeof(struct gfs2_dinode), dent); - - gfs2_inum_out(dip, dent); - dent->de_type = cpu_to_be16(DT_DIR); - - gfs2_dinode_out(ip, di); - - brelse(dibh); - } - - error = gfs2_change_nlink(dip, +1); - gfs2_assert_withdraw(sdp, !error); /* dip already pinned */ - - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - - gfs2_glock_dq_uninit_m(2, ghs); - - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - - return 0; -} - -/** - * gfs2_rmdiri - Remove a directory - * @dip: The parent directory of the directory to be removed - * @name: The name of the directory to be removed - * @ip: The GFS2 inode of the directory to be removed - * - * Assumes Glocks on dip and ip are held - * - * Returns: errno - */ - -static int gfs2_rmdiri(struct gfs2_inode *dip, const struct qstr *name, - struct gfs2_inode *ip) -{ - int error; - - if (ip->i_entries != 2) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - return -EIO; - } - - error = gfs2_dir_del(dip, name); - if (error) - return error; - - error = gfs2_change_nlink(dip, -1); - if (error) - return error; - - error = gfs2_dir_del(ip, &gfs2_qdot); - if (error) - return error; - - error = gfs2_dir_del(ip, &gfs2_qdotdot); - if (error) - return error; - - /* It looks odd, but it really should be done twice */ - error = gfs2_change_nlink(ip, -1); - if (error) - return error; - - error = gfs2_change_nlink(ip, -1); - if (error) - return error; - - return error; -} - -/** - * gfs2_rmdir - Remove a directory - * @dir: The parent directory of the directory to be removed - * @dentry: The dentry of the directory to remove - * - * Remove a directory. Call gfs2_rmdiri() - * - * Returns: errno - */ - -static int gfs2_rmdir(struct inode *dir, struct dentry *dentry) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); - struct gfs2_holder ghs[3]; - struct gfs2_rgrpd *rgd; - struct gfs2_holder ri_gh; - int error; - - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - return error; - gfs2_holder_init(dip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + 1); - - rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); - gfs2_holder_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + 2); - - error = gfs2_glock_nq(ghs); /* parent */ - if (error) - goto out_parent; - - error = gfs2_glock_nq(ghs + 1); /* child */ - if (error) - goto out_child; - - error = gfs2_glock_nq(ghs + 2); /* rgrp */ - if (error) - goto out_rgrp; - - error = gfs2_unlink_ok(dip, &dentry->d_name, ip); - if (error) - goto out_gunlock; - - if (ip->i_entries < 2) { - if (gfs2_consist_inode(ip)) - gfs2_dinode_print(ip); - error = -EIO; - goto out_gunlock; - } - if (ip->i_entries > 2) { - error = -ENOTEMPTY; - goto out_gunlock; - } - - error = gfs2_trans_begin(sdp, 2 * RES_DINODE + 3 * RES_LEAF + RES_RG_BIT, 0); - if (error) - goto out_gunlock; - - error = gfs2_rmdiri(dip, &dentry->d_name, ip); - - gfs2_trans_end(sdp); - -out_gunlock: - gfs2_glock_dq(ghs + 2); -out_rgrp: - gfs2_holder_uninit(ghs + 2); - gfs2_glock_dq(ghs + 1); -out_child: - gfs2_holder_uninit(ghs + 1); - gfs2_glock_dq(ghs); -out_parent: - gfs2_holder_uninit(ghs); - gfs2_glock_dq_uninit(&ri_gh); - return error; -} - -/** - * gfs2_mknod - Make a special file - * @dir: The directory in which the special file will reside - * @dentry: The dentry of the special file - * @mode: The mode of the special file - * @rdev: The device specification of the special file - * - */ - -static int gfs2_mknod(struct inode *dir, struct dentry *dentry, int mode, - dev_t dev) -{ - struct gfs2_inode *dip = GFS2_I(dir); - struct gfs2_sbd *sdp = GFS2_SB(dir); - struct gfs2_holder ghs[2]; - struct inode *inode; - - gfs2_holder_init(dip->i_gl, 0, 0, ghs); - - inode = gfs2_createi(ghs, &dentry->d_name, mode, dev); - if (IS_ERR(inode)) { - gfs2_holder_uninit(ghs); - return PTR_ERR(inode); - } - - gfs2_trans_end(sdp); - if (dip->i_alloc->al_rgd) - gfs2_inplace_release(dip); - gfs2_quota_unlock(dip); - gfs2_alloc_put(dip); - - gfs2_glock_dq_uninit_m(2, ghs); - - d_instantiate(dentry, inode); - mark_inode_dirty(inode); - - return 0; -} - -/* - * gfs2_ok_to_move - check if it's ok to move a directory to another directory - * @this: move this - * @to: to here - * - * Follow @to back to the root and make sure we don't encounter @this - * Assumes we already hold the rename lock. - * - * Returns: errno - */ - -static int gfs2_ok_to_move(struct gfs2_inode *this, struct gfs2_inode *to) -{ - struct inode *dir = &to->i_inode; - struct super_block *sb = dir->i_sb; - struct inode *tmp; - int error = 0; - - igrab(dir); - - for (;;) { - if (dir == &this->i_inode) { - error = -EINVAL; - break; - } - if (dir == sb->s_root->d_inode) { - error = 0; - break; - } - - tmp = gfs2_lookupi(dir, &gfs2_qdotdot, 1); - if (IS_ERR(tmp)) { - error = PTR_ERR(tmp); - break; - } - - iput(dir); - dir = tmp; - } - - iput(dir); - - return error; -} - -/** - * gfs2_rename - Rename a file - * @odir: Parent directory of old file name - * @odentry: The old dentry of the file - * @ndir: Parent directory of new file name - * @ndentry: The new dentry of the file - * - * Returns: errno - */ - -static int gfs2_rename(struct inode *odir, struct dentry *odentry, - struct inode *ndir, struct dentry *ndentry) -{ - struct gfs2_inode *odip = GFS2_I(odir); - struct gfs2_inode *ndip = GFS2_I(ndir); - struct gfs2_inode *ip = GFS2_I(odentry->d_inode); - struct gfs2_inode *nip = NULL; - struct gfs2_sbd *sdp = GFS2_SB(odir); - struct gfs2_holder ghs[5], r_gh = { .gh_gl = NULL, }, ri_gh; - struct gfs2_rgrpd *nrgd; - unsigned int num_gh; - int dir_rename = 0; - int alloc_required = 0; - unsigned int x; - int error; - - if (ndentry->d_inode) { - nip = GFS2_I(ndentry->d_inode); - if (ip == nip) - return 0; - } - - error = gfs2_rindex_hold(sdp, &ri_gh); - if (error) - return error; - - if (odip != ndip) { - error = gfs2_glock_nq_init(sdp->sd_rename_gl, LM_ST_EXCLUSIVE, - 0, &r_gh); - if (error) - goto out; - - if (S_ISDIR(ip->i_inode.i_mode)) { - dir_rename = 1; - /* don't move a dirctory into it's subdir */ - error = gfs2_ok_to_move(ip, ndip); - if (error) - goto out_gunlock_r; - } - } - - num_gh = 1; - gfs2_holder_init(odip->i_gl, LM_ST_EXCLUSIVE, 0, ghs); - if (odip != ndip) { - gfs2_holder_init(ndip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); - num_gh++; - } - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); - num_gh++; - - if (nip) { - gfs2_holder_init(nip->i_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh); - num_gh++; - /* grab the resource lock for unlink flag twiddling - * this is the case of the target file already existing - * so we unlink before doing the rename - */ - nrgd = gfs2_blk2rgrpd(sdp, nip->i_no_addr); - if (nrgd) - gfs2_holder_init(nrgd->rd_gl, LM_ST_EXCLUSIVE, 0, ghs + num_gh++); - } - - for (x = 0; x < num_gh; x++) { - error = gfs2_glock_nq(ghs + x); - if (error) - goto out_gunlock; - } - - /* Check out the old directory */ - - error = gfs2_unlink_ok(odip, &odentry->d_name, ip); - if (error) - goto out_gunlock; - - /* Check out the new directory */ - - if (nip) { - error = gfs2_unlink_ok(ndip, &ndentry->d_name, nip); - if (error) - goto out_gunlock; - - if (S_ISDIR(nip->i_inode.i_mode)) { - if (nip->i_entries < 2) { - if (gfs2_consist_inode(nip)) - gfs2_dinode_print(nip); - error = -EIO; - goto out_gunlock; - } - if (nip->i_entries > 2) { - error = -ENOTEMPTY; - goto out_gunlock; - } - } - } else { - error = gfs2_permission(ndir, MAY_WRITE | MAY_EXEC, 0); - if (error) - goto out_gunlock; - - error = gfs2_dir_check(ndir, &ndentry->d_name, NULL); - switch (error) { - case -ENOENT: - error = 0; - break; - case 0: - error = -EEXIST; - default: - goto out_gunlock; - }; - - if (odip != ndip) { - if (!ndip->i_inode.i_nlink) { - error = -EINVAL; - goto out_gunlock; - } - if (ndip->i_entries == (u32)-1) { - error = -EFBIG; - goto out_gunlock; - } - if (S_ISDIR(ip->i_inode.i_mode) && - ndip->i_inode.i_nlink == (u32)-1) { - error = -EMLINK; - goto out_gunlock; - } - } - } - - /* Check out the dir to be renamed */ - - if (dir_rename) { - error = gfs2_permission(odentry->d_inode, MAY_WRITE, 0); - if (error) - goto out_gunlock; - } - - if (nip == NULL) - alloc_required = gfs2_diradd_alloc_required(ndir, &ndentry->d_name); - error = alloc_required; - if (error < 0) - goto out_gunlock; - error = 0; - - if (alloc_required) { - struct gfs2_alloc *al = gfs2_alloc_get(ndip); - if (!al) { - error = -ENOMEM; - goto out_gunlock; - } - - error = gfs2_quota_lock_check(ndip); - if (error) - goto out_alloc; - - al->al_requested = sdp->sd_max_dirres; - - error = gfs2_inplace_reserve_ri(ndip); - if (error) - goto out_gunlock_q; - - error = gfs2_trans_begin(sdp, sdp->sd_max_dirres + - gfs2_rg_blocks(al) + - 4 * RES_DINODE + 4 * RES_LEAF + - RES_STATFS + RES_QUOTA + 4, 0); - if (error) - goto out_ipreserv; - } else { - error = gfs2_trans_begin(sdp, 4 * RES_DINODE + - 5 * RES_LEAF + 4, 0); - if (error) - goto out_gunlock; - } - - /* Remove the target file, if it exists */ - - if (nip) { - if (S_ISDIR(nip->i_inode.i_mode)) - error = gfs2_rmdiri(ndip, &ndentry->d_name, nip); - else { - error = gfs2_dir_del(ndip, &ndentry->d_name); - if (error) - goto out_end_trans; - error = gfs2_change_nlink(nip, -1); - } - if (error) - goto out_end_trans; - } - - if (dir_rename) { - error = gfs2_change_nlink(ndip, +1); - if (error) - goto out_end_trans; - error = gfs2_change_nlink(odip, -1); - if (error) - goto out_end_trans; - - error = gfs2_dir_mvino(ip, &gfs2_qdotdot, ndip, DT_DIR); - if (error) - goto out_end_trans; - } else { - struct buffer_head *dibh; - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) - goto out_end_trans; - ip->i_inode.i_ctime = CURRENT_TIME; - gfs2_trans_add_bh(ip->i_gl, dibh, 1); - gfs2_dinode_out(ip, dibh->b_data); - brelse(dibh); - } - - error = gfs2_dir_del(odip, &odentry->d_name); - if (error) - goto out_end_trans; - - error = gfs2_dir_add(ndir, &ndentry->d_name, ip, IF2DT(ip->i_inode.i_mode)); - if (error) - goto out_end_trans; - -out_end_trans: - gfs2_trans_end(sdp); -out_ipreserv: - if (alloc_required) - gfs2_inplace_release(ndip); -out_gunlock_q: - if (alloc_required) - gfs2_quota_unlock(ndip); -out_alloc: - if (alloc_required) - gfs2_alloc_put(ndip); -out_gunlock: - while (x--) { - gfs2_glock_dq(ghs + x); - gfs2_holder_uninit(ghs + x); - } -out_gunlock_r: - if (r_gh.gh_gl) - gfs2_glock_dq_uninit(&r_gh); -out: - gfs2_glock_dq_uninit(&ri_gh); - return error; -} - -/** - * gfs2_follow_link - Follow a symbolic link - * @dentry: The dentry of the link - * @nd: Data that we pass to vfs_follow_link() - * - * This can handle symlinks of any size. - * - * Returns: 0 on success or error code - */ - -static void *gfs2_follow_link(struct dentry *dentry, struct nameidata *nd) -{ - struct gfs2_inode *ip = GFS2_I(dentry->d_inode); - struct gfs2_holder i_gh; - struct buffer_head *dibh; - unsigned int x, size; - char *buf; - int error; - - gfs2_holder_init(ip->i_gl, LM_ST_SHARED, 0, &i_gh); - error = gfs2_glock_nq(&i_gh); - if (error) { - gfs2_holder_uninit(&i_gh); - nd_set_link(nd, ERR_PTR(error)); - return NULL; - } - - size = (unsigned int)i_size_read(&ip->i_inode); - if (size == 0) { - gfs2_consist_inode(ip); - buf = ERR_PTR(-EIO); - goto out; - } - - error = gfs2_meta_inode_buffer(ip, &dibh); - if (error) { - buf = ERR_PTR(error); - goto out; - } - - x = size + 1; - buf = kmalloc(x, GFP_NOFS); - if (!buf) - buf = ERR_PTR(-ENOMEM); - else - memcpy(buf, dibh->b_data + sizeof(struct gfs2_dinode), x); - brelse(dibh); -out: - gfs2_glock_dq_uninit(&i_gh); - nd_set_link(nd, buf); - return NULL; -} - -static void gfs2_put_link(struct dentry *dentry, struct nameidata *nd, void *p) -{ - char *s = nd_get_link(nd); - if (!IS_ERR(s)) - kfree(s); -} - -/** - * gfs2_permission - - * @inode: The inode - * @mask: The mask to be tested - * @flags: Indicates whether this is an RCU path walk or not - * - * This may be called from the VFS directly, or from within GFS2 with the - * inode locked, so we look to see if the glock is already locked and only - * lock the glock if its not already been done. - * - * Returns: errno - */ - -int gfs2_permission(struct inode *inode, int mask, unsigned int flags) -{ - struct gfs2_inode *ip; - struct gfs2_holder i_gh; - int error; - int unlock = 0; - - - ip = GFS2_I(inode); - if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { - if (flags & IPERM_FLAG_RCU) - return -ECHILD; - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &i_gh); - if (error) - return error; - unlock = 1; - } - - if ((mask & MAY_WRITE) && IS_IMMUTABLE(inode)) - error = -EACCES; - else - error = generic_permission(inode, mask, flags, gfs2_check_acl); - if (unlock) - gfs2_glock_dq_uninit(&i_gh); - - return error; -} - -static int setattr_chown(struct inode *inode, struct iattr *attr) -{ - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_sbd *sdp = GFS2_SB(inode); - u32 ouid, ogid, nuid, ngid; - int error; - - ouid = inode->i_uid; - ogid = inode->i_gid; - nuid = attr->ia_uid; - ngid = attr->ia_gid; - - if (!(attr->ia_valid & ATTR_UID) || ouid == nuid) - ouid = nuid = NO_QUOTA_CHANGE; - if (!(attr->ia_valid & ATTR_GID) || ogid == ngid) - ogid = ngid = NO_QUOTA_CHANGE; - - if (!gfs2_alloc_get(ip)) - return -ENOMEM; - - error = gfs2_quota_lock(ip, nuid, ngid); - if (error) - goto out_alloc; - - if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { - error = gfs2_quota_check(ip, nuid, ngid); - if (error) - goto out_gunlock_q; - } - - error = gfs2_trans_begin(sdp, RES_DINODE + 2 * RES_QUOTA, 0); - if (error) - goto out_gunlock_q; - - error = gfs2_setattr_simple(ip, attr); - if (error) - goto out_end_trans; - - if (ouid != NO_QUOTA_CHANGE || ogid != NO_QUOTA_CHANGE) { - u64 blocks = gfs2_get_inode_blocks(&ip->i_inode); - gfs2_quota_change(ip, -blocks, ouid, ogid); - gfs2_quota_change(ip, blocks, nuid, ngid); - } - -out_end_trans: - gfs2_trans_end(sdp); -out_gunlock_q: - gfs2_quota_unlock(ip); -out_alloc: - gfs2_alloc_put(ip); - return error; -} - -/** - * gfs2_setattr - Change attributes on an inode - * @dentry: The dentry which is changing - * @attr: The structure describing the change - * - * The VFS layer wants to change one or more of an inodes attributes. Write - * that change out to disk. - * - * Returns: errno - */ - -static int gfs2_setattr(struct dentry *dentry, struct iattr *attr) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder i_gh; - int error; - - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &i_gh); - if (error) - return error; - - error = -EPERM; - if (IS_IMMUTABLE(inode) || IS_APPEND(inode)) - goto out; - - error = inode_change_ok(inode, attr); - if (error) - goto out; - - if (attr->ia_valid & ATTR_SIZE) - error = gfs2_setattr_size(inode, attr->ia_size); - else if (attr->ia_valid & (ATTR_UID | ATTR_GID)) - error = setattr_chown(inode, attr); - else if ((attr->ia_valid & ATTR_MODE) && IS_POSIXACL(inode)) - error = gfs2_acl_chmod(ip, attr); - else - error = gfs2_setattr_simple(ip, attr); - -out: - gfs2_glock_dq_uninit(&i_gh); - if (!error) - mark_inode_dirty(inode); - return error; -} - -/** - * gfs2_getattr - Read out an inode's attributes - * @mnt: The vfsmount the inode is being accessed from - * @dentry: The dentry to stat - * @stat: The inode's stats - * - * This may be called from the VFS directly, or from within GFS2 with the - * inode locked, so we look to see if the glock is already locked and only - * lock the glock if its not already been done. Note that its the NFS - * readdirplus operation which causes this to be called (from filldir) - * with the glock already held. - * - * Returns: errno - */ - -static int gfs2_getattr(struct vfsmount *mnt, struct dentry *dentry, - struct kstat *stat) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int error; - int unlock = 0; - - if (gfs2_glock_is_locked_by_me(ip->i_gl) == NULL) { - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); - if (error) - return error; - unlock = 1; - } - - generic_fillattr(inode, stat); - if (unlock) - gfs2_glock_dq_uninit(&gh); - - return 0; -} - -static int gfs2_setxattr(struct dentry *dentry, const char *name, - const void *data, size_t size, int flags) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); - ret = gfs2_glock_nq(&gh); - if (ret == 0) { - ret = generic_setxattr(dentry, name, data, size, flags); - gfs2_glock_dq(&gh); - } - gfs2_holder_uninit(&gh); - return ret; -} - -static ssize_t gfs2_getxattr(struct dentry *dentry, const char *name, - void *data, size_t size) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - gfs2_holder_init(ip->i_gl, LM_ST_SHARED, LM_FLAG_ANY, &gh); - ret = gfs2_glock_nq(&gh); - if (ret == 0) { - ret = generic_getxattr(dentry, name, data, size); - gfs2_glock_dq(&gh); - } - gfs2_holder_uninit(&gh); - return ret; -} - -static int gfs2_removexattr(struct dentry *dentry, const char *name) -{ - struct inode *inode = dentry->d_inode; - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - gfs2_holder_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); - ret = gfs2_glock_nq(&gh); - if (ret == 0) { - ret = generic_removexattr(dentry, name); - gfs2_glock_dq(&gh); - } - gfs2_holder_uninit(&gh); - return ret; -} - -static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, - u64 start, u64 len) -{ - struct gfs2_inode *ip = GFS2_I(inode); - struct gfs2_holder gh; - int ret; - - ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC); - if (ret) - return ret; - - mutex_lock(&inode->i_mutex); - - ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh); - if (ret) - goto out; - - if (gfs2_is_stuffed(ip)) { - u64 phys = ip->i_no_addr << inode->i_blkbits; - u64 size = i_size_read(inode); - u32 flags = FIEMAP_EXTENT_LAST|FIEMAP_EXTENT_NOT_ALIGNED| - FIEMAP_EXTENT_DATA_INLINE; - phys += sizeof(struct gfs2_dinode); - phys += start; - if (start + len > size) - len = size - start; - if (start < size) - ret = fiemap_fill_next_extent(fieinfo, start, phys, - len, flags); - if (ret == 1) - ret = 0; - } else { - ret = __generic_block_fiemap(inode, fieinfo, start, len, - gfs2_block_map); - } - - gfs2_glock_dq_uninit(&gh); -out: - mutex_unlock(&inode->i_mutex); - return ret; -} - -const struct inode_operations gfs2_file_iops = { - .permission = gfs2_permission, - .setattr = gfs2_setattr, - .getattr = gfs2_getattr, - .setxattr = gfs2_setxattr, - .getxattr = gfs2_getxattr, - .listxattr = gfs2_listxattr, - .removexattr = gfs2_removexattr, - .fiemap = gfs2_fiemap, -}; - -const struct inode_operations gfs2_dir_iops = { - .create = gfs2_create, - .lookup = gfs2_lookup, - .link = gfs2_link, - .unlink = gfs2_unlink, - .symlink = gfs2_symlink, - .mkdir = gfs2_mkdir, - .rmdir = gfs2_rmdir, - .mknod = gfs2_mknod, - .rename = gfs2_rename, - .permission = gfs2_permission, - .setattr = gfs2_setattr, - .getattr = gfs2_getattr, - .setxattr = gfs2_setxattr, - .getxattr = gfs2_getxattr, - .listxattr = gfs2_listxattr, - .removexattr = gfs2_removexattr, - .fiemap = gfs2_fiemap, -}; - -const struct inode_operations gfs2_symlink_iops = { - .readlink = generic_readlink, - .follow_link = gfs2_follow_link, - .put_link = gfs2_put_link, - .permission = gfs2_permission, - .setattr = gfs2_setattr, - .getattr = gfs2_getattr, - .setxattr = gfs2_setxattr, - .getxattr = gfs2_getxattr, - .listxattr = gfs2_listxattr, - .removexattr = gfs2_removexattr, - .fiemap = gfs2_fiemap, -}; - diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index e23d9864c418..42e8d23bc047 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -38,6 +38,7 @@ #include <linux/sched.h> #include <linux/slab.h> +#include <linux/mm.h> #include <linux/spinlock.h> #include <linux/completion.h> #include <linux/buffer_head.h> @@ -77,19 +78,20 @@ static LIST_HEAD(qd_lru_list); static atomic_t qd_lru_count = ATOMIC_INIT(0); static DEFINE_SPINLOCK(qd_lru_lock); -int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) +int gfs2_shrink_qd_memory(struct shrinker *shrink, struct shrink_control *sc) { struct gfs2_quota_data *qd; struct gfs2_sbd *sdp; + int nr_to_scan = sc->nr_to_scan; - if (nr == 0) + if (nr_to_scan == 0) goto out; - if (!(gfp_mask & __GFP_FS)) + if (!(sc->gfp_mask & __GFP_FS)) return -1; spin_lock(&qd_lru_lock); - while (nr && !list_empty(&qd_lru_list)) { + while (nr_to_scan && !list_empty(&qd_lru_list)) { qd = list_entry(qd_lru_list.next, struct gfs2_quota_data, qd_reclaim); sdp = qd->qd_gl->gl_sbd; @@ -110,7 +112,7 @@ int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) spin_unlock(&qd_lru_lock); kmem_cache_free(gfs2_quotad_cachep, qd); spin_lock(&qd_lru_lock); - nr--; + nr_to_scan--; } spin_unlock(&qd_lru_lock); diff --git a/fs/gfs2/quota.h b/fs/gfs2/quota.h index e7d236ca48bd..90bf1c302a98 100644 --- a/fs/gfs2/quota.h +++ b/fs/gfs2/quota.h @@ -12,6 +12,7 @@ struct gfs2_inode; struct gfs2_sbd; +struct shrink_control; #define NO_QUOTA_CHANGE ((u32)-1) @@ -51,7 +52,8 @@ static inline int gfs2_quota_lock_check(struct gfs2_inode *ip) return ret; } -extern int gfs2_shrink_qd_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask); +extern int gfs2_shrink_qd_memory(struct shrinker *shrink, + struct shrink_control *sc); extern const struct quotactl_ops gfs2_quotactl_ops; #endif /* __QUOTA_DOT_H__ */ diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index cf930cd9664a..9b780df3fd54 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -78,10 +78,11 @@ static u32 rgblk_search(struct gfs2_rgrpd *rgd, u32 goal, static inline void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buf1, unsigned char *buf2, unsigned int offset, - unsigned int buflen, u32 block, + struct gfs2_bitmap *bi, u32 block, unsigned char new_state) { unsigned char *byte1, *byte2, *end, cur_state; + unsigned int buflen = bi->bi_len; const unsigned int bit = (block % GFS2_NBBY) * GFS2_BIT_SIZE; byte1 = buf1 + offset + (block / GFS2_NBBY); @@ -92,6 +93,16 @@ static inline void gfs2_setbit(struct gfs2_rgrpd *rgd, unsigned char *buf1, cur_state = (*byte1 >> bit) & GFS2_BIT_MASK; if (unlikely(!valid_change[new_state * 4 + cur_state])) { + printk(KERN_WARNING "GFS2: buf_blk = 0x%llx old_state=%d, " + "new_state=%d\n", + (unsigned long long)block, cur_state, new_state); + printk(KERN_WARNING "GFS2: rgrp=0x%llx bi_start=0x%lx\n", + (unsigned long long)rgd->rd_addr, + (unsigned long)bi->bi_start); + printk(KERN_WARNING "GFS2: bi_offset=0x%lx bi_len=0x%lx\n", + (unsigned long)bi->bi_offset, + (unsigned long)bi->bi_len); + dump_stack(); gfs2_consist_rgrpd(rgd); return; } @@ -381,6 +392,7 @@ static void clear_rgrpdi(struct gfs2_sbd *sdp) if (gl) { gl->gl_object = NULL; + gfs2_glock_add_to_lru(gl); gfs2_glock_put(gl); } @@ -945,7 +957,7 @@ static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip /* rgblk_search can return a block < goal, so we need to keep it marching forward. */ no_addr = block + rgd->rd_data0; - goal++; + goal = max(block + 1, goal + 1); if (*last_unlinked != NO_BLOCK && no_addr <= *last_unlinked) continue; if (no_addr == skip) @@ -971,7 +983,7 @@ static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip found++; /* Limit reclaim to sensible number of tasks */ - if (found > 2*NR_CPUS) + if (found > NR_CPUS) return; } @@ -1365,7 +1377,7 @@ skip: gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset, - bi->bi_len, blk, new_state); + bi, blk, new_state); goal = blk; while (*n < elen) { goal++; @@ -1375,7 +1387,7 @@ skip: GFS2_BLKST_FREE) break; gfs2_setbit(rgd, bi->bi_bh->b_data, bi->bi_clone, bi->bi_offset, - bi->bi_len, goal, new_state); + bi, goal, new_state); (*n)++; } out: @@ -1432,7 +1444,7 @@ static struct gfs2_rgrpd *rgblk_free(struct gfs2_sbd *sdp, u64 bstart, } gfs2_trans_add_bh(rgd->rd_gl, bi->bi_bh, 1); gfs2_setbit(rgd, bi->bi_bh->b_data, NULL, bi->bi_offset, - bi->bi_len, buf_blk, new_state); + bi, buf_blk, new_state); } return rgd; @@ -1617,6 +1629,10 @@ void __gfs2_free_data(struct gfs2_inode *ip, u64 bstart, u32 blen) gfs2_rgrp_out(rgd, rgd->rd_bits[0].bi_bh->b_data); gfs2_trans_add_rg(rgd); + + /* Directories keep their data in the metadata address space */ + if (ip->i_depth) + gfs2_meta_wipe(ip, bstart, blen); } /** diff --git a/fs/gfs2/super.c b/fs/gfs2/super.c index ec73ed70bae1..ed540e7018be 100644 --- a/fs/gfs2/super.c +++ b/fs/gfs2/super.c @@ -23,6 +23,7 @@ #include <linux/time.h> #include <linux/wait.h> #include <linux/writeback.h> +#include <linux/backing-dev.h> #include "gfs2.h" #include "incore.h" @@ -657,7 +658,7 @@ out: * @sdp: the file system * * This function flushes data and meta data for all machines by - * aquiring the transaction log exclusively. All journals are + * acquiring the transaction log exclusively. All journals are * ensured to be in a clean state as well. * * Returns: errno @@ -700,11 +701,47 @@ void gfs2_unfreeze_fs(struct gfs2_sbd *sdp) mutex_unlock(&sdp->sd_freeze_lock); } +void gfs2_dinode_out(const struct gfs2_inode *ip, void *buf) +{ + struct gfs2_dinode *str = buf; + + str->di_header.mh_magic = cpu_to_be32(GFS2_MAGIC); + str->di_header.mh_type = cpu_to_be32(GFS2_METATYPE_DI); + str->di_header.mh_format = cpu_to_be32(GFS2_FORMAT_DI); + str->di_num.no_addr = cpu_to_be64(ip->i_no_addr); + str->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino); + str->di_mode = cpu_to_be32(ip->i_inode.i_mode); + str->di_uid = cpu_to_be32(ip->i_inode.i_uid); + str->di_gid = cpu_to_be32(ip->i_inode.i_gid); + str->di_nlink = cpu_to_be32(ip->i_inode.i_nlink); + str->di_size = cpu_to_be64(i_size_read(&ip->i_inode)); + str->di_blocks = cpu_to_be64(gfs2_get_inode_blocks(&ip->i_inode)); + str->di_atime = cpu_to_be64(ip->i_inode.i_atime.tv_sec); + str->di_mtime = cpu_to_be64(ip->i_inode.i_mtime.tv_sec); + str->di_ctime = cpu_to_be64(ip->i_inode.i_ctime.tv_sec); + + str->di_goal_meta = cpu_to_be64(ip->i_goal); + str->di_goal_data = cpu_to_be64(ip->i_goal); + str->di_generation = cpu_to_be64(ip->i_generation); + + str->di_flags = cpu_to_be32(ip->i_diskflags); + str->di_height = cpu_to_be16(ip->i_height); + str->di_payload_format = cpu_to_be32(S_ISDIR(ip->i_inode.i_mode) && + !(ip->i_diskflags & GFS2_DIF_EXHASH) ? + GFS2_FORMAT_DE : 0); + str->di_depth = cpu_to_be16(ip->i_depth); + str->di_entries = cpu_to_be32(ip->i_entries); + + str->di_eattr = cpu_to_be64(ip->i_eattr); + str->di_atime_nsec = cpu_to_be32(ip->i_inode.i_atime.tv_nsec); + str->di_mtime_nsec = cpu_to_be32(ip->i_inode.i_mtime.tv_nsec); + str->di_ctime_nsec = cpu_to_be32(ip->i_inode.i_ctime.tv_nsec); +} /** * gfs2_write_inode - Make sure the inode is stable on the disk * @inode: The inode - * @sync: synchronous write flag + * @wbc: The writeback control structure * * Returns: errno */ @@ -713,15 +750,17 @@ static int gfs2_write_inode(struct inode *inode, struct writeback_control *wbc) { struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_sbd *sdp = GFS2_SB(inode); + struct address_space *metamapping = gfs2_glock2aspace(ip->i_gl); + struct backing_dev_info *bdi = metamapping->backing_dev_info; struct gfs2_holder gh; struct buffer_head *bh; struct timespec atime; struct gfs2_dinode *di; - int ret = 0; + int ret = -EAGAIN; - /* Check this is a "normal" inode, etc */ + /* Skip timestamp update, if this is from a memalloc */ if (current->flags & PF_MEMALLOC) - return 0; + goto do_flush; ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); if (ret) goto do_flush; @@ -745,6 +784,13 @@ do_unlock: do_flush: if (wbc->sync_mode == WB_SYNC_ALL) gfs2_log_flush(GFS2_SB(inode), ip->i_gl); + filemap_fdatawrite(metamapping); + if (bdi->dirty_exceeded) + gfs2_ail1_flush(sdp, wbc); + if (!ret && (wbc->sync_mode == WB_SYNC_ALL)) + ret = filemap_fdatawait(metamapping); + if (ret) + mark_inode_dirty_sync(inode); return ret; } @@ -874,8 +920,9 @@ restart: static int gfs2_sync_fs(struct super_block *sb, int wait) { - if (wait && sb->s_fs_info) - gfs2_log_flush(sb->s_fs_info, NULL); + struct gfs2_sbd *sdp = sb->s_fs_info; + if (wait && sdp) + gfs2_log_flush(sdp, NULL); return 0; } @@ -1308,6 +1355,78 @@ static int gfs2_show_options(struct seq_file *s, struct vfsmount *mnt) return 0; } +static void gfs2_final_release_pages(struct gfs2_inode *ip) +{ + struct inode *inode = &ip->i_inode; + struct gfs2_glock *gl = ip->i_gl; + + truncate_inode_pages(gfs2_glock2aspace(ip->i_gl), 0); + truncate_inode_pages(&inode->i_data, 0); + + if (atomic_read(&gl->gl_revokes) == 0) { + clear_bit(GLF_LFLUSH, &gl->gl_flags); + clear_bit(GLF_DIRTY, &gl->gl_flags); + } +} + +static int gfs2_dinode_dealloc(struct gfs2_inode *ip) +{ + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + struct gfs2_alloc *al; + struct gfs2_rgrpd *rgd; + int error; + + if (gfs2_get_inode_blocks(&ip->i_inode) != 1) { + gfs2_consist_inode(ip); + return -EIO; + } + + al = gfs2_alloc_get(ip); + if (!al) + return -ENOMEM; + + error = gfs2_quota_hold(ip, NO_QUOTA_CHANGE, NO_QUOTA_CHANGE); + if (error) + goto out; + + error = gfs2_rindex_hold(sdp, &al->al_ri_gh); + if (error) + goto out_qs; + + rgd = gfs2_blk2rgrpd(sdp, ip->i_no_addr); + if (!rgd) { + gfs2_consist_inode(ip); + error = -EIO; + goto out_rindex_relse; + } + + error = gfs2_glock_nq_init(rgd->rd_gl, LM_ST_EXCLUSIVE, 0, + &al->al_rgd_gh); + if (error) + goto out_rindex_relse; + + error = gfs2_trans_begin(sdp, RES_RG_BIT + RES_STATFS + RES_QUOTA, + sdp->sd_jdesc->jd_blocks); + if (error) + goto out_rg_gunlock; + + gfs2_free_di(rgd, ip); + + gfs2_final_release_pages(ip); + + gfs2_trans_end(sdp); + +out_rg_gunlock: + gfs2_glock_dq_uninit(&al->al_rgd_gh); +out_rindex_relse: + gfs2_glock_dq_uninit(&al->al_ri_gh); +out_qs: + gfs2_quota_unhold(ip); +out: + gfs2_alloc_put(ip); + return error; +} + /* * We have to (at the moment) hold the inodes main lock to cover * the gap between unlocking the shared lock on the iopen lock and @@ -1318,15 +1437,17 @@ static int gfs2_show_options(struct seq_file *s, struct vfsmount *mnt) static void gfs2_evict_inode(struct inode *inode) { - struct gfs2_sbd *sdp = inode->i_sb->s_fs_info; + struct super_block *sb = inode->i_sb; + struct gfs2_sbd *sdp = sb->s_fs_info; struct gfs2_inode *ip = GFS2_I(inode); struct gfs2_holder gh; int error; - if (inode->i_nlink) + if (inode->i_nlink || (sb->s_flags & MS_RDONLY)) goto out; - error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, 0, &gh); + /* Must not read inode block until block type has been verified */ + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, &gh); if (unlikely(error)) { gfs2_glock_dq_uninit(&ip->i_iopen_gh); goto out; @@ -1336,6 +1457,12 @@ static void gfs2_evict_inode(struct inode *inode) if (error) goto out_truncate; + if (test_bit(GIF_INVALID, &ip->i_flags)) { + error = gfs2_inode_refresh(ip); + if (error) + goto out_truncate; + } + ip->i_iopen_gh.gh_flags |= GL_NOCACHE; gfs2_glock_dq_wait(&ip->i_iopen_gh); gfs2_holder_reinit(LM_ST_EXCLUSIVE, LM_FLAG_TRY_1CB | GL_NOCACHE, &ip->i_iopen_gh); @@ -1363,15 +1490,13 @@ static void gfs2_evict_inode(struct inode *inode) } error = gfs2_dinode_dealloc(ip); - if (error) - goto out_unlock; + goto out_unlock; out_truncate: error = gfs2_trans_begin(sdp, 0, sdp->sd_jdesc->jd_blocks); if (error) goto out_unlock; - /* Needs to be done before glock release & also in a transaction */ - truncate_inode_pages(&inode->i_data, 0); + gfs2_final_release_pages(ip); gfs2_trans_end(sdp); out_unlock: @@ -1386,6 +1511,7 @@ out: end_writeback(inode); ip->i_gl->gl_object = NULL; + gfs2_glock_add_to_lru(ip->i_gl); gfs2_glock_put(ip->i_gl); ip->i_gl = NULL; if (ip->i_iopen_gh.gh_gl) { diff --git a/fs/gfs2/sys.c b/fs/gfs2/sys.c index 748ccb557c18..e20eab37bc80 100644 --- a/fs/gfs2/sys.c +++ b/fs/gfs2/sys.c @@ -81,7 +81,8 @@ static int gfs2_uuid_valid(const u8 *uuid) static ssize_t uuid_show(struct gfs2_sbd *sdp, char *buf) { - const u8 *uuid = sdp->sd_sb.sb_uuid; + struct super_block *s = sdp->sd_vfs; + const u8 *uuid = s->s_uuid; buf[0] = '\0'; if (!gfs2_uuid_valid(uuid)) return 0; @@ -616,7 +617,8 @@ static int gfs2_uevent(struct kset *kset, struct kobject *kobj, struct kobj_uevent_env *env) { struct gfs2_sbd *sdp = container_of(kobj, struct gfs2_sbd, sd_kobj); - const u8 *uuid = sdp->sd_sb.sb_uuid; + struct super_block *s = sdp->sd_vfs; + const u8 *uuid = s->s_uuid; add_uevent_var(env, "LOCKTABLE=%s", sdp->sd_table_name); add_uevent_var(env, "LOCKPROTO=%s", sdp->sd_proto_name); diff --git a/fs/gfs2/trace_gfs2.h b/fs/gfs2/trace_gfs2.h index cedb0bb96d96..5d07609ec57d 100644 --- a/fs/gfs2/trace_gfs2.h +++ b/fs/gfs2/trace_gfs2.h @@ -10,6 +10,7 @@ #include <linux/buffer_head.h> #include <linux/dlmconstants.h> #include <linux/gfs2_ondisk.h> +#include <linux/writeback.h> #include "incore.h" #include "glock.h" @@ -40,7 +41,9 @@ {(1UL << GLF_REPLY_PENDING), "r" }, \ {(1UL << GLF_INITIAL), "I" }, \ {(1UL << GLF_FROZEN), "F" }, \ - {(1UL << GLF_QUEUED), "q" }) + {(1UL << GLF_QUEUED), "q" }, \ + {(1UL << GLF_LRU), "L" }, \ + {(1UL << GLF_OBJECT), "o" }) #ifndef NUMPTY #define NUMPTY @@ -94,7 +97,7 @@ TRACE_EVENT(gfs2_glock_state_change, __entry->new_state = glock_trace_state(new_state); __entry->tgt_state = glock_trace_state(gl->gl_target); __entry->dmt_state = glock_trace_state(gl->gl_demote_state); - __entry->flags = gl->gl_flags; + __entry->flags = gl->gl_flags | (gl->gl_object ? (1UL<<GLF_OBJECT) : 0); ), TP_printk("%u,%u glock %d:%lld state %s to %s tgt:%s dmt:%s flags:%s", @@ -127,7 +130,7 @@ TRACE_EVENT(gfs2_glock_put, __entry->gltype = gl->gl_name.ln_type; __entry->glnum = gl->gl_name.ln_number; __entry->cur_state = glock_trace_state(gl->gl_state); - __entry->flags = gl->gl_flags; + __entry->flags = gl->gl_flags | (gl->gl_object ? (1UL<<GLF_OBJECT) : 0); ), TP_printk("%u,%u glock %d:%lld state %s => %s flags:%s", @@ -161,7 +164,7 @@ TRACE_EVENT(gfs2_demote_rq, __entry->glnum = gl->gl_name.ln_number; __entry->cur_state = glock_trace_state(gl->gl_state); __entry->dmt_state = glock_trace_state(gl->gl_demote_state); - __entry->flags = gl->gl_flags; + __entry->flags = gl->gl_flags | (gl->gl_object ? (1UL<<GLF_OBJECT) : 0); ), TP_printk("%u,%u glock %d:%lld demote %s to %s flags:%s", @@ -318,6 +321,33 @@ TRACE_EVENT(gfs2_log_blocks, MINOR(__entry->dev), __entry->blocks) ); +/* Writing back the AIL */ +TRACE_EVENT(gfs2_ail_flush, + + TP_PROTO(const struct gfs2_sbd *sdp, const struct writeback_control *wbc, int start), + + TP_ARGS(sdp, wbc, start), + + TP_STRUCT__entry( + __field( dev_t, dev ) + __field( int, start ) + __field( int, sync_mode ) + __field( long, nr_to_write ) + ), + + TP_fast_assign( + __entry->dev = sdp->sd_vfs->s_dev; + __entry->start = start; + __entry->sync_mode = wbc->sync_mode; + __entry->nr_to_write = wbc->nr_to_write; + ), + + TP_printk("%u,%u ail flush %s %s %ld", MAJOR(__entry->dev), + MINOR(__entry->dev), __entry->start ? "start" : "end", + __entry->sync_mode == WB_SYNC_ALL ? "all" : "none", + __entry->nr_to_write) +); + /* Section 3 - bmap * * Objectives: diff --git a/fs/hpfs/Kconfig b/fs/hpfs/Kconfig index 0c39dc3ef7d7..56bd15c5bf6c 100644 --- a/fs/hpfs/Kconfig +++ b/fs/hpfs/Kconfig @@ -1,7 +1,6 @@ config HPFS_FS tristate "OS/2 HPFS file system support" depends on BLOCK - depends on BROKEN || !PREEMPT help OS/2 is IBM's operating system for PC's, the same as Warp, and HPFS is the file system used for organizing files on OS/2 hard disk diff --git a/fs/hpfs/alloc.c b/fs/hpfs/alloc.c index 5503e2c28910..7a5eb2c718c8 100644 --- a/fs/hpfs/alloc.c +++ b/fs/hpfs/alloc.c @@ -8,8 +8,6 @@ #include "hpfs_fn.h" -static int hpfs_alloc_if_possible_nolock(struct super_block *s, secno sec); - /* * Check if a sector is allocated in bitmap * This is really slow. Turned on only if chk==2 @@ -18,9 +16,9 @@ static int hpfs_alloc_if_possible_nolock(struct super_block *s, secno sec); static int chk_if_allocated(struct super_block *s, secno sec, char *msg) { struct quad_buffer_head qbh; - unsigned *bmp; + u32 *bmp; if (!(bmp = hpfs_map_bitmap(s, sec >> 14, &qbh, "chk"))) goto fail; - if ((bmp[(sec & 0x3fff) >> 5] >> (sec & 0x1f)) & 1) { + if ((cpu_to_le32(bmp[(sec & 0x3fff) >> 5]) >> (sec & 0x1f)) & 1) { hpfs_error(s, "sector '%s' - %08x not allocated in bitmap", msg, sec); goto fail1; } @@ -28,7 +26,7 @@ static int chk_if_allocated(struct super_block *s, secno sec, char *msg) if (sec >= hpfs_sb(s)->sb_dirband_start && sec < hpfs_sb(s)->sb_dirband_start + hpfs_sb(s)->sb_dirband_size) { unsigned ssec = (sec - hpfs_sb(s)->sb_dirband_start) / 4; if (!(bmp = hpfs_map_dnode_bitmap(s, &qbh))) goto fail; - if ((bmp[ssec >> 5] >> (ssec & 0x1f)) & 1) { + if ((le32_to_cpu(bmp[ssec >> 5]) >> (ssec & 0x1f)) & 1) { hpfs_error(s, "sector '%s' - %08x not allocated in directory bitmap", msg, sec); goto fail1; } @@ -75,7 +73,6 @@ static secno alloc_in_bmp(struct super_block *s, secno near, unsigned n, unsigne hpfs_error(s, "Bad allocation size: %d", n); return 0; } - lock_super(s); if (bs != ~0x3fff) { if (!(bmp = hpfs_map_bitmap(s, near >> 14, &qbh, "aib"))) goto uls; } else { @@ -85,10 +82,6 @@ static secno alloc_in_bmp(struct super_block *s, secno near, unsigned n, unsigne ret = bs + nr; goto rt; } - /*if (!tstbits(bmp, nr + n, n + forward)) { - ret = bs + nr + n; - goto rt; - }*/ q = nr + n; b = 0; while ((a = tstbits(bmp, q, n + forward)) != 0) { q += a; @@ -105,14 +98,14 @@ static secno alloc_in_bmp(struct super_block *s, secno near, unsigned n, unsigne goto rt; } nr >>= 5; - /*for (i = nr + 1; i != nr; i++, i &= 0x1ff) {*/ + /*for (i = nr + 1; i != nr; i++, i &= 0x1ff) */ i = nr; do { - if (!bmp[i]) goto cont; - if (n + forward >= 0x3f && bmp[i] != -1) goto cont; + if (!le32_to_cpu(bmp[i])) goto cont; + if (n + forward >= 0x3f && le32_to_cpu(bmp[i]) != 0xffffffff) goto cont; q = i<<5; if (i > 0) { - unsigned k = bmp[i-1]; + unsigned k = le32_to_cpu(bmp[i-1]); while (k & 0x80000000) { q--; k <<= 1; } @@ -132,18 +125,17 @@ static secno alloc_in_bmp(struct super_block *s, secno near, unsigned n, unsigne } while (i != nr); rt: if (ret) { - if (hpfs_sb(s)->sb_chk && ((ret >> 14) != (bs >> 14) || (bmp[(ret & 0x3fff) >> 5] | ~(((1 << n) - 1) << (ret & 0x1f))) != 0xffffffff)) { + if (hpfs_sb(s)->sb_chk && ((ret >> 14) != (bs >> 14) || (le32_to_cpu(bmp[(ret & 0x3fff) >> 5]) | ~(((1 << n) - 1) << (ret & 0x1f))) != 0xffffffff)) { hpfs_error(s, "Allocation doesn't work! Wanted %d, allocated at %08x", n, ret); ret = 0; goto b; } - bmp[(ret & 0x3fff) >> 5] &= ~(((1 << n) - 1) << (ret & 0x1f)); + bmp[(ret & 0x3fff) >> 5] &= cpu_to_le32(~(((1 << n) - 1) << (ret & 0x1f))); hpfs_mark_4buffers_dirty(&qbh); } b: hpfs_brelse4(&qbh); uls: - unlock_super(s); return ret; } @@ -155,7 +147,7 @@ static secno alloc_in_bmp(struct super_block *s, secno near, unsigned n, unsigne * sectors */ -secno hpfs_alloc_sector(struct super_block *s, secno near, unsigned n, int forward, int lock) +secno hpfs_alloc_sector(struct super_block *s, secno near, unsigned n, int forward) { secno sec; int i; @@ -167,7 +159,6 @@ secno hpfs_alloc_sector(struct super_block *s, secno near, unsigned n, int forwa forward = -forward; f_p = 1; } - if (lock) hpfs_lock_creation(s); n_bmps = (sbi->sb_fs_size + 0x4000 - 1) >> 14; if (near && near < sbi->sb_fs_size) { if ((sec = alloc_in_bmp(s, near, n, f_p ? forward : forward/4))) goto ret; @@ -214,18 +205,17 @@ secno hpfs_alloc_sector(struct super_block *s, secno near, unsigned n, int forwa ret: if (sec && f_p) { for (i = 0; i < forward; i++) { - if (!hpfs_alloc_if_possible_nolock(s, sec + i + 1)) { + if (!hpfs_alloc_if_possible(s, sec + i + 1)) { hpfs_error(s, "Prealloc doesn't work! Wanted %d, allocated at %08x, can't allocate %d", forward, sec, i); sec = 0; break; } } } - if (lock) hpfs_unlock_creation(s); return sec; } -static secno alloc_in_dirband(struct super_block *s, secno near, int lock) +static secno alloc_in_dirband(struct super_block *s, secno near) { unsigned nr = near; secno sec; @@ -236,49 +226,35 @@ static secno alloc_in_dirband(struct super_block *s, secno near, int lock) nr = sbi->sb_dirband_start + sbi->sb_dirband_size - 4; nr -= sbi->sb_dirband_start; nr >>= 2; - if (lock) hpfs_lock_creation(s); sec = alloc_in_bmp(s, (~0x3fff) | nr, 1, 0); - if (lock) hpfs_unlock_creation(s); if (!sec) return 0; return ((sec & 0x3fff) << 2) + sbi->sb_dirband_start; } /* Alloc sector if it's free */ -static int hpfs_alloc_if_possible_nolock(struct super_block *s, secno sec) +int hpfs_alloc_if_possible(struct super_block *s, secno sec) { struct quad_buffer_head qbh; - unsigned *bmp; - lock_super(s); + u32 *bmp; if (!(bmp = hpfs_map_bitmap(s, sec >> 14, &qbh, "aip"))) goto end; - if (bmp[(sec & 0x3fff) >> 5] & (1 << (sec & 0x1f))) { - bmp[(sec & 0x3fff) >> 5] &= ~(1 << (sec & 0x1f)); + if (le32_to_cpu(bmp[(sec & 0x3fff) >> 5]) & (1 << (sec & 0x1f))) { + bmp[(sec & 0x3fff) >> 5] &= cpu_to_le32(~(1 << (sec & 0x1f))); hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); - unlock_super(s); return 1; } hpfs_brelse4(&qbh); end: - unlock_super(s); return 0; } -int hpfs_alloc_if_possible(struct super_block *s, secno sec) -{ - int r; - hpfs_lock_creation(s); - r = hpfs_alloc_if_possible_nolock(s, sec); - hpfs_unlock_creation(s); - return r; -} - /* Free sectors in bitmaps */ void hpfs_free_sectors(struct super_block *s, secno sec, unsigned n) { struct quad_buffer_head qbh; - unsigned *bmp; + u32 *bmp; struct hpfs_sb_info *sbi = hpfs_sb(s); /*printk("2 - ");*/ if (!n) return; @@ -286,26 +262,22 @@ void hpfs_free_sectors(struct super_block *s, secno sec, unsigned n) hpfs_error(s, "Trying to free reserved sector %08x", sec); return; } - lock_super(s); sbi->sb_max_fwd_alloc += n > 0xffff ? 0xffff : n; if (sbi->sb_max_fwd_alloc > 0xffffff) sbi->sb_max_fwd_alloc = 0xffffff; new_map: if (!(bmp = hpfs_map_bitmap(s, sec >> 14, &qbh, "free"))) { - unlock_super(s); return; } new_tst: - if ((bmp[(sec & 0x3fff) >> 5] >> (sec & 0x1f) & 1)) { + if ((le32_to_cpu(bmp[(sec & 0x3fff) >> 5]) >> (sec & 0x1f) & 1)) { hpfs_error(s, "sector %08x not allocated", sec); hpfs_brelse4(&qbh); - unlock_super(s); return; } - bmp[(sec & 0x3fff) >> 5] |= 1 << (sec & 0x1f); + bmp[(sec & 0x3fff) >> 5] |= cpu_to_le32(1 << (sec & 0x1f)); if (!--n) { hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); - unlock_super(s); return; } if (!(++sec & 0x3fff)) { @@ -327,13 +299,13 @@ int hpfs_check_free_dnodes(struct super_block *s, int n) int n_bmps = (hpfs_sb(s)->sb_fs_size + 0x4000 - 1) >> 14; int b = hpfs_sb(s)->sb_c_bitmap & 0x0fffffff; int i, j; - unsigned *bmp; + u32 *bmp; struct quad_buffer_head qbh; if ((bmp = hpfs_map_dnode_bitmap(s, &qbh))) { for (j = 0; j < 512; j++) { unsigned k; - if (!bmp[j]) continue; - for (k = bmp[j]; k; k >>= 1) if (k & 1) if (!--n) { + if (!le32_to_cpu(bmp[j])) continue; + for (k = le32_to_cpu(bmp[j]); k; k >>= 1) if (k & 1) if (!--n) { hpfs_brelse4(&qbh); return 0; } @@ -352,10 +324,10 @@ int hpfs_check_free_dnodes(struct super_block *s, int n) chk_bmp: if (bmp) { for (j = 0; j < 512; j++) { - unsigned k; - if (!bmp[j]) continue; + u32 k; + if (!le32_to_cpu(bmp[j])) continue; for (k = 0xf; k; k <<= 4) - if ((bmp[j] & k) == k) { + if ((le32_to_cpu(bmp[j]) & k) == k) { if (!--n) { hpfs_brelse4(&qbh); return 0; @@ -379,44 +351,40 @@ void hpfs_free_dnode(struct super_block *s, dnode_secno dno) hpfs_free_sectors(s, dno, 4); } else { struct quad_buffer_head qbh; - unsigned *bmp; + u32 *bmp; unsigned ssec = (dno - hpfs_sb(s)->sb_dirband_start) / 4; - lock_super(s); if (!(bmp = hpfs_map_dnode_bitmap(s, &qbh))) { - unlock_super(s); return; } - bmp[ssec >> 5] |= 1 << (ssec & 0x1f); + bmp[ssec >> 5] |= cpu_to_le32(1 << (ssec & 0x1f)); hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); - unlock_super(s); } } struct dnode *hpfs_alloc_dnode(struct super_block *s, secno near, - dnode_secno *dno, struct quad_buffer_head *qbh, - int lock) + dnode_secno *dno, struct quad_buffer_head *qbh) { struct dnode *d; if (hpfs_count_one_bitmap(s, hpfs_sb(s)->sb_dmap) > FREE_DNODES_ADD) { - if (!(*dno = alloc_in_dirband(s, near, lock))) - if (!(*dno = hpfs_alloc_sector(s, near, 4, 0, lock))) return NULL; + if (!(*dno = alloc_in_dirband(s, near))) + if (!(*dno = hpfs_alloc_sector(s, near, 4, 0))) return NULL; } else { - if (!(*dno = hpfs_alloc_sector(s, near, 4, 0, lock))) - if (!(*dno = alloc_in_dirband(s, near, lock))) return NULL; + if (!(*dno = hpfs_alloc_sector(s, near, 4, 0))) + if (!(*dno = alloc_in_dirband(s, near))) return NULL; } if (!(d = hpfs_get_4sectors(s, *dno, qbh))) { hpfs_free_dnode(s, *dno); return NULL; } memset(d, 0, 2048); - d->magic = DNODE_MAGIC; - d->first_free = 52; + d->magic = cpu_to_le32(DNODE_MAGIC); + d->first_free = cpu_to_le32(52); d->dirent[0] = 32; d->dirent[2] = 8; d->dirent[30] = 1; d->dirent[31] = 255; - d->self = *dno; + d->self = cpu_to_le32(*dno); return d; } @@ -424,16 +392,16 @@ struct fnode *hpfs_alloc_fnode(struct super_block *s, secno near, fnode_secno *f struct buffer_head **bh) { struct fnode *f; - if (!(*fno = hpfs_alloc_sector(s, near, 1, FNODE_ALLOC_FWD, 1))) return NULL; + if (!(*fno = hpfs_alloc_sector(s, near, 1, FNODE_ALLOC_FWD))) return NULL; if (!(f = hpfs_get_sector(s, *fno, bh))) { hpfs_free_sectors(s, *fno, 1); return NULL; } memset(f, 0, 512); - f->magic = FNODE_MAGIC; - f->ea_offs = 0xc4; + f->magic = cpu_to_le32(FNODE_MAGIC); + f->ea_offs = cpu_to_le16(0xc4); f->btree.n_free_nodes = 8; - f->btree.first_free = 8; + f->btree.first_free = cpu_to_le16(8); return f; } @@ -441,16 +409,16 @@ struct anode *hpfs_alloc_anode(struct super_block *s, secno near, anode_secno *a struct buffer_head **bh) { struct anode *a; - if (!(*ano = hpfs_alloc_sector(s, near, 1, ANODE_ALLOC_FWD, 1))) return NULL; + if (!(*ano = hpfs_alloc_sector(s, near, 1, ANODE_ALLOC_FWD))) return NULL; if (!(a = hpfs_get_sector(s, *ano, bh))) { hpfs_free_sectors(s, *ano, 1); return NULL; } memset(a, 0, 512); - a->magic = ANODE_MAGIC; - a->self = *ano; + a->magic = cpu_to_le32(ANODE_MAGIC); + a->self = cpu_to_le32(*ano); a->btree.n_free_nodes = 40; a->btree.n_used_nodes = 0; - a->btree.first_free = 8; + a->btree.first_free = cpu_to_le16(8); return a; } diff --git a/fs/hpfs/anode.c b/fs/hpfs/anode.c index 6a2f04bf3df0..08b503e8ed29 100644 --- a/fs/hpfs/anode.c +++ b/fs/hpfs/anode.c @@ -22,8 +22,8 @@ secno hpfs_bplus_lookup(struct super_block *s, struct inode *inode, if (hpfs_sb(s)->sb_chk) if (hpfs_stop_cycles(s, a, &c1, &c2, "hpfs_bplus_lookup")) return -1; if (btree->internal) { for (i = 0; i < btree->n_used_nodes; i++) - if (btree->u.internal[i].file_secno > sec) { - a = btree->u.internal[i].down; + if (le32_to_cpu(btree->u.internal[i].file_secno) > sec) { + a = le32_to_cpu(btree->u.internal[i].down); brelse(bh); if (!(anode = hpfs_map_anode(s, a, &bh))) return -1; btree = &anode->btree; @@ -34,18 +34,18 @@ secno hpfs_bplus_lookup(struct super_block *s, struct inode *inode, return -1; } for (i = 0; i < btree->n_used_nodes; i++) - if (btree->u.external[i].file_secno <= sec && - btree->u.external[i].file_secno + btree->u.external[i].length > sec) { - a = btree->u.external[i].disk_secno + sec - btree->u.external[i].file_secno; + if (le32_to_cpu(btree->u.external[i].file_secno) <= sec && + le32_to_cpu(btree->u.external[i].file_secno) + le32_to_cpu(btree->u.external[i].length) > sec) { + a = le32_to_cpu(btree->u.external[i].disk_secno) + sec - le32_to_cpu(btree->u.external[i].file_secno); if (hpfs_sb(s)->sb_chk) if (hpfs_chk_sectors(s, a, 1, "data")) { brelse(bh); return -1; } if (inode) { struct hpfs_inode_info *hpfs_inode = hpfs_i(inode); - hpfs_inode->i_file_sec = btree->u.external[i].file_secno; - hpfs_inode->i_disk_sec = btree->u.external[i].disk_secno; - hpfs_inode->i_n_secs = btree->u.external[i].length; + hpfs_inode->i_file_sec = le32_to_cpu(btree->u.external[i].file_secno); + hpfs_inode->i_disk_sec = le32_to_cpu(btree->u.external[i].disk_secno); + hpfs_inode->i_n_secs = le32_to_cpu(btree->u.external[i].length); } brelse(bh); return a; @@ -83,8 +83,8 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi return -1; } if (btree->internal) { - a = btree->u.internal[n].down; - btree->u.internal[n].file_secno = -1; + a = le32_to_cpu(btree->u.internal[n].down); + btree->u.internal[n].file_secno = cpu_to_le32(-1); mark_buffer_dirty(bh); brelse(bh); if (hpfs_sb(s)->sb_chk) @@ -94,15 +94,15 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi goto go_down; } if (n >= 0) { - if (btree->u.external[n].file_secno + btree->u.external[n].length != fsecno) { + if (le32_to_cpu(btree->u.external[n].file_secno) + le32_to_cpu(btree->u.external[n].length) != fsecno) { hpfs_error(s, "allocated size %08x, trying to add sector %08x, %cnode %08x", - btree->u.external[n].file_secno + btree->u.external[n].length, fsecno, + le32_to_cpu(btree->u.external[n].file_secno) + le32_to_cpu(btree->u.external[n].length), fsecno, fnod?'f':'a', node); brelse(bh); return -1; } - if (hpfs_alloc_if_possible(s, se = btree->u.external[n].disk_secno + btree->u.external[n].length)) { - btree->u.external[n].length++; + if (hpfs_alloc_if_possible(s, se = le32_to_cpu(btree->u.external[n].disk_secno) + le32_to_cpu(btree->u.external[n].length))) { + btree->u.external[n].length = cpu_to_le32(le32_to_cpu(btree->u.external[n].length) + 1); mark_buffer_dirty(bh); brelse(bh); return se; @@ -115,20 +115,20 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi } se = !fnod ? node : (node + 16384) & ~16383; } - if (!(se = hpfs_alloc_sector(s, se, 1, fsecno*ALLOC_M>ALLOC_FWD_MAX ? ALLOC_FWD_MAX : fsecno*ALLOC_M<ALLOC_FWD_MIN ? ALLOC_FWD_MIN : fsecno*ALLOC_M, 1))) { + if (!(se = hpfs_alloc_sector(s, se, 1, fsecno*ALLOC_M>ALLOC_FWD_MAX ? ALLOC_FWD_MAX : fsecno*ALLOC_M<ALLOC_FWD_MIN ? ALLOC_FWD_MIN : fsecno*ALLOC_M))) { brelse(bh); return -1; } - fs = n < 0 ? 0 : btree->u.external[n].file_secno + btree->u.external[n].length; + fs = n < 0 ? 0 : le32_to_cpu(btree->u.external[n].file_secno) + le32_to_cpu(btree->u.external[n].length); if (!btree->n_free_nodes) { - up = a != node ? anode->up : -1; + up = a != node ? le32_to_cpu(anode->up) : -1; if (!(anode = hpfs_alloc_anode(s, a, &na, &bh1))) { brelse(bh); hpfs_free_sectors(s, se, 1); return -1; } if (a == node && fnod) { - anode->up = node; + anode->up = cpu_to_le32(node); anode->btree.fnode_parent = 1; anode->btree.n_used_nodes = btree->n_used_nodes; anode->btree.first_free = btree->first_free; @@ -137,9 +137,9 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi btree->internal = 1; btree->n_free_nodes = 11; btree->n_used_nodes = 1; - btree->first_free = (char *)&(btree->u.internal[1]) - (char *)btree; - btree->u.internal[0].file_secno = -1; - btree->u.internal[0].down = na; + btree->first_free = cpu_to_le16((char *)&(btree->u.internal[1]) - (char *)btree); + btree->u.internal[0].file_secno = cpu_to_le32(-1); + btree->u.internal[0].down = cpu_to_le32(na); mark_buffer_dirty(bh); } else if (!(ranode = hpfs_alloc_anode(s, /*a*/0, &ra, &bh2))) { brelse(bh); @@ -153,15 +153,15 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi btree = &anode->btree; } btree->n_free_nodes--; n = btree->n_used_nodes++; - btree->first_free += 12; - btree->u.external[n].disk_secno = se; - btree->u.external[n].file_secno = fs; - btree->u.external[n].length = 1; + btree->first_free = cpu_to_le16(le16_to_cpu(btree->first_free) + 12); + btree->u.external[n].disk_secno = cpu_to_le32(se); + btree->u.external[n].file_secno = cpu_to_le32(fs); + btree->u.external[n].length = cpu_to_le32(1); mark_buffer_dirty(bh); brelse(bh); if ((a == node && fnod) || na == -1) return se; c2 = 0; - while (up != -1) { + while (up != (anode_secno)-1) { struct anode *new_anode; if (hpfs_sb(s)->sb_chk) if (hpfs_stop_cycles(s, up, &c1, &c2, "hpfs_add_sector_to_btree #2")) return -1; @@ -174,47 +174,47 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi } if (btree->n_free_nodes) { btree->n_free_nodes--; n = btree->n_used_nodes++; - btree->first_free += 8; - btree->u.internal[n].file_secno = -1; - btree->u.internal[n].down = na; - btree->u.internal[n-1].file_secno = fs; + btree->first_free = cpu_to_le16(le16_to_cpu(btree->first_free) + 8); + btree->u.internal[n].file_secno = cpu_to_le32(-1); + btree->u.internal[n].down = cpu_to_le32(na); + btree->u.internal[n-1].file_secno = cpu_to_le32(fs); mark_buffer_dirty(bh); brelse(bh); brelse(bh2); hpfs_free_sectors(s, ra, 1); if ((anode = hpfs_map_anode(s, na, &bh))) { - anode->up = up; + anode->up = cpu_to_le32(up); anode->btree.fnode_parent = up == node && fnod; mark_buffer_dirty(bh); brelse(bh); } return se; } - up = up != node ? anode->up : -1; - btree->u.internal[btree->n_used_nodes - 1].file_secno = /*fs*/-1; + up = up != node ? le32_to_cpu(anode->up) : -1; + btree->u.internal[btree->n_used_nodes - 1].file_secno = cpu_to_le32(/*fs*/-1); mark_buffer_dirty(bh); brelse(bh); a = na; if ((new_anode = hpfs_alloc_anode(s, a, &na, &bh))) { anode = new_anode; - /*anode->up = up != -1 ? up : ra;*/ + /*anode->up = cpu_to_le32(up != -1 ? up : ra);*/ anode->btree.internal = 1; anode->btree.n_used_nodes = 1; anode->btree.n_free_nodes = 59; - anode->btree.first_free = 16; - anode->btree.u.internal[0].down = a; - anode->btree.u.internal[0].file_secno = -1; + anode->btree.first_free = cpu_to_le16(16); + anode->btree.u.internal[0].down = cpu_to_le32(a); + anode->btree.u.internal[0].file_secno = cpu_to_le32(-1); mark_buffer_dirty(bh); brelse(bh); if ((anode = hpfs_map_anode(s, a, &bh))) { - anode->up = na; + anode->up = cpu_to_le32(na); mark_buffer_dirty(bh); brelse(bh); } } else na = a; } if ((anode = hpfs_map_anode(s, na, &bh))) { - anode->up = node; + anode->up = cpu_to_le32(node); if (fnod) anode->btree.fnode_parent = 1; mark_buffer_dirty(bh); brelse(bh); @@ -232,14 +232,14 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi } btree = &fnode->btree; } - ranode->up = node; - memcpy(&ranode->btree, btree, btree->first_free); + ranode->up = cpu_to_le32(node); + memcpy(&ranode->btree, btree, le16_to_cpu(btree->first_free)); if (fnod) ranode->btree.fnode_parent = 1; ranode->btree.n_free_nodes = (ranode->btree.internal ? 60 : 40) - ranode->btree.n_used_nodes; if (ranode->btree.internal) for (n = 0; n < ranode->btree.n_used_nodes; n++) { struct anode *unode; - if ((unode = hpfs_map_anode(s, ranode->u.internal[n].down, &bh1))) { - unode->up = ra; + if ((unode = hpfs_map_anode(s, le32_to_cpu(ranode->u.internal[n].down), &bh1))) { + unode->up = cpu_to_le32(ra); unode->btree.fnode_parent = 0; mark_buffer_dirty(bh1); brelse(bh1); @@ -248,11 +248,11 @@ secno hpfs_add_sector_to_btree(struct super_block *s, secno node, int fnod, unsi btree->internal = 1; btree->n_free_nodes = fnod ? 10 : 58; btree->n_used_nodes = 2; - btree->first_free = (char *)&btree->u.internal[2] - (char *)btree; - btree->u.internal[0].file_secno = fs; - btree->u.internal[0].down = ra; - btree->u.internal[1].file_secno = -1; - btree->u.internal[1].down = na; + btree->first_free = cpu_to_le16((char *)&btree->u.internal[2] - (char *)btree); + btree->u.internal[0].file_secno = cpu_to_le32(fs); + btree->u.internal[0].down = cpu_to_le32(ra); + btree->u.internal[1].file_secno = cpu_to_le32(-1); + btree->u.internal[1].down = cpu_to_le32(na); mark_buffer_dirty(bh); brelse(bh); mark_buffer_dirty(bh2); @@ -279,7 +279,7 @@ void hpfs_remove_btree(struct super_block *s, struct bplus_header *btree) go_down: d2 = 0; while (btree1->internal) { - ano = btree1->u.internal[pos].down; + ano = le32_to_cpu(btree1->u.internal[pos].down); if (level) brelse(bh); if (hpfs_sb(s)->sb_chk) if (hpfs_stop_cycles(s, ano, &d1, &d2, "hpfs_remove_btree #1")) @@ -290,7 +290,7 @@ void hpfs_remove_btree(struct super_block *s, struct bplus_header *btree) pos = 0; } for (i = 0; i < btree1->n_used_nodes; i++) - hpfs_free_sectors(s, btree1->u.external[i].disk_secno, btree1->u.external[i].length); + hpfs_free_sectors(s, le32_to_cpu(btree1->u.external[i].disk_secno), le32_to_cpu(btree1->u.external[i].length)); go_up: if (!level) return; brelse(bh); @@ -298,13 +298,13 @@ void hpfs_remove_btree(struct super_block *s, struct bplus_header *btree) if (hpfs_stop_cycles(s, ano, &c1, &c2, "hpfs_remove_btree #2")) return; hpfs_free_sectors(s, ano, 1); oano = ano; - ano = anode->up; + ano = le32_to_cpu(anode->up); if (--level) { if (!(anode = hpfs_map_anode(s, ano, &bh))) return; btree1 = &anode->btree; } else btree1 = btree; for (i = 0; i < btree1->n_used_nodes; i++) { - if (btree1->u.internal[i].down == oano) { + if (le32_to_cpu(btree1->u.internal[i].down) == oano) { if ((pos = i + 1) < btree1->n_used_nodes) goto go_down; else @@ -411,7 +411,7 @@ void hpfs_truncate_btree(struct super_block *s, secno f, int fno, unsigned secs) if (fno) { btree->n_free_nodes = 8; btree->n_used_nodes = 0; - btree->first_free = 8; + btree->first_free = cpu_to_le16(8); btree->internal = 0; mark_buffer_dirty(bh); } else hpfs_free_sectors(s, f, 1); @@ -421,22 +421,22 @@ void hpfs_truncate_btree(struct super_block *s, secno f, int fno, unsigned secs) while (btree->internal) { nodes = btree->n_used_nodes + btree->n_free_nodes; for (i = 0; i < btree->n_used_nodes; i++) - if (btree->u.internal[i].file_secno >= secs) goto f; + if (le32_to_cpu(btree->u.internal[i].file_secno) >= secs) goto f; brelse(bh); hpfs_error(s, "internal btree %08x doesn't end with -1", node); return; f: for (j = i + 1; j < btree->n_used_nodes; j++) - hpfs_ea_remove(s, btree->u.internal[j].down, 1, 0); + hpfs_ea_remove(s, le32_to_cpu(btree->u.internal[j].down), 1, 0); btree->n_used_nodes = i + 1; btree->n_free_nodes = nodes - btree->n_used_nodes; - btree->first_free = 8 + 8 * btree->n_used_nodes; + btree->first_free = cpu_to_le16(8 + 8 * btree->n_used_nodes); mark_buffer_dirty(bh); - if (btree->u.internal[i].file_secno == secs) { + if (btree->u.internal[i].file_secno == cpu_to_le32(secs)) { brelse(bh); return; } - node = btree->u.internal[i].down; + node = le32_to_cpu(btree->u.internal[i].down); brelse(bh); if (hpfs_sb(s)->sb_chk) if (hpfs_stop_cycles(s, node, &c1, &c2, "hpfs_truncate_btree")) @@ -446,25 +446,25 @@ void hpfs_truncate_btree(struct super_block *s, secno f, int fno, unsigned secs) } nodes = btree->n_used_nodes + btree->n_free_nodes; for (i = 0; i < btree->n_used_nodes; i++) - if (btree->u.external[i].file_secno + btree->u.external[i].length >= secs) goto ff; + if (le32_to_cpu(btree->u.external[i].file_secno) + le32_to_cpu(btree->u.external[i].length) >= secs) goto ff; brelse(bh); return; ff: - if (secs <= btree->u.external[i].file_secno) { + if (secs <= le32_to_cpu(btree->u.external[i].file_secno)) { hpfs_error(s, "there is an allocation error in file %08x, sector %08x", f, secs); if (i) i--; } - else if (btree->u.external[i].file_secno + btree->u.external[i].length > secs) { - hpfs_free_sectors(s, btree->u.external[i].disk_secno + secs - - btree->u.external[i].file_secno, btree->u.external[i].length - - secs + btree->u.external[i].file_secno); /* I hope gcc optimizes this :-) */ - btree->u.external[i].length = secs - btree->u.external[i].file_secno; + else if (le32_to_cpu(btree->u.external[i].file_secno) + le32_to_cpu(btree->u.external[i].length) > secs) { + hpfs_free_sectors(s, le32_to_cpu(btree->u.external[i].disk_secno) + secs - + le32_to_cpu(btree->u.external[i].file_secno), le32_to_cpu(btree->u.external[i].length) + - secs + le32_to_cpu(btree->u.external[i].file_secno)); /* I hope gcc optimizes this :-) */ + btree->u.external[i].length = cpu_to_le32(secs - le32_to_cpu(btree->u.external[i].file_secno)); } for (j = i + 1; j < btree->n_used_nodes; j++) - hpfs_free_sectors(s, btree->u.external[j].disk_secno, btree->u.external[j].length); + hpfs_free_sectors(s, le32_to_cpu(btree->u.external[j].disk_secno), le32_to_cpu(btree->u.external[j].length)); btree->n_used_nodes = i + 1; btree->n_free_nodes = nodes - btree->n_used_nodes; - btree->first_free = 8 + 12 * btree->n_used_nodes; + btree->first_free = cpu_to_le16(8 + 12 * btree->n_used_nodes); mark_buffer_dirty(bh); brelse(bh); } @@ -480,12 +480,12 @@ void hpfs_remove_fnode(struct super_block *s, fnode_secno fno) struct extended_attribute *ea_end; if (!(fnode = hpfs_map_fnode(s, fno, &bh))) return; if (!fnode->dirflag) hpfs_remove_btree(s, &fnode->btree); - else hpfs_remove_dtree(s, fnode->u.external[0].disk_secno); + else hpfs_remove_dtree(s, le32_to_cpu(fnode->u.external[0].disk_secno)); ea_end = fnode_end_ea(fnode); for (ea = fnode_ea(fnode); ea < ea_end; ea = next_ea(ea)) if (ea->indirect) hpfs_ea_remove(s, ea_sec(ea), ea->anode, ea_len(ea)); - hpfs_ea_ext_remove(s, fnode->ea_secno, fnode->ea_anode, fnode->ea_size_l); + hpfs_ea_ext_remove(s, le32_to_cpu(fnode->ea_secno), fnode->ea_anode, le32_to_cpu(fnode->ea_size_l)); brelse(bh); hpfs_free_sectors(s, fno, 1); } diff --git a/fs/hpfs/buffer.c b/fs/hpfs/buffer.c index 793cb9d943d2..9ecde27d1e29 100644 --- a/fs/hpfs/buffer.c +++ b/fs/hpfs/buffer.c @@ -9,22 +9,6 @@ #include <linux/slab.h> #include "hpfs_fn.h" -void hpfs_lock_creation(struct super_block *s) -{ -#ifdef DEBUG_LOCKS - printk("lock creation\n"); -#endif - mutex_lock(&hpfs_sb(s)->hpfs_creation_de); -} - -void hpfs_unlock_creation(struct super_block *s) -{ -#ifdef DEBUG_LOCKS - printk("unlock creation\n"); -#endif - mutex_unlock(&hpfs_sb(s)->hpfs_creation_de); -} - /* Map a sector into a buffer and return pointers to it and to the buffer. */ void *hpfs_map_sector(struct super_block *s, unsigned secno, struct buffer_head **bhp, @@ -32,6 +16,8 @@ void *hpfs_map_sector(struct super_block *s, unsigned secno, struct buffer_head { struct buffer_head *bh; + hpfs_lock_assert(s); + cond_resched(); *bhp = bh = sb_bread(s, secno); @@ -50,6 +36,8 @@ void *hpfs_get_sector(struct super_block *s, unsigned secno, struct buffer_head struct buffer_head *bh; /*return hpfs_map_sector(s, secno, bhp, 0);*/ + hpfs_lock_assert(s); + cond_resched(); if ((*bhp = bh = sb_getblk(s, secno)) != NULL) { @@ -70,6 +58,8 @@ void *hpfs_map_4sectors(struct super_block *s, unsigned secno, struct quad_buffe struct buffer_head *bh; char *data; + hpfs_lock_assert(s); + cond_resched(); if (secno & 3) { @@ -125,6 +115,8 @@ void *hpfs_get_4sectors(struct super_block *s, unsigned secno, { cond_resched(); + hpfs_lock_assert(s); + if (secno & 3) { printk("HPFS: hpfs_get_4sectors: unaligned read\n"); return NULL; diff --git a/fs/hpfs/dir.c b/fs/hpfs/dir.c index b3d7c0ddb609..f46ae025bfb5 100644 --- a/fs/hpfs/dir.c +++ b/fs/hpfs/dir.c @@ -88,9 +88,9 @@ static int hpfs_readdir(struct file *filp, void *dirent, filldir_t filldir) hpfs_error(inode->i_sb, "not a directory, fnode %08lx", (unsigned long)inode->i_ino); } - if (hpfs_inode->i_dno != fno->u.external[0].disk_secno) { + if (hpfs_inode->i_dno != le32_to_cpu(fno->u.external[0].disk_secno)) { e = 1; - hpfs_error(inode->i_sb, "corrupted inode: i_dno == %08x, fnode -> dnode == %08x", hpfs_inode->i_dno, fno->u.external[0].disk_secno); + hpfs_error(inode->i_sb, "corrupted inode: i_dno == %08x, fnode -> dnode == %08x", hpfs_inode->i_dno, le32_to_cpu(fno->u.external[0].disk_secno)); } brelse(bh); if (e) { @@ -156,7 +156,7 @@ static int hpfs_readdir(struct file *filp, void *dirent, filldir_t filldir) goto again; } tempname = hpfs_translate_name(inode->i_sb, de->name, de->namelen, lc, de->not_8x3); - if (filldir(dirent, tempname, de->namelen, old_pos, de->fnode, DT_UNKNOWN) < 0) { + if (filldir(dirent, tempname, de->namelen, old_pos, le32_to_cpu(de->fnode), DT_UNKNOWN) < 0) { filp->f_pos = old_pos; if (tempname != de->name) kfree(tempname); hpfs_brelse4(&qbh); @@ -221,7 +221,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name * Get inode number, what we're after. */ - ino = de->fnode; + ino = le32_to_cpu(de->fnode); /* * Go find or make an inode. @@ -236,7 +236,7 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name hpfs_init_inode(result); if (de->directory) hpfs_read_inode(result); - else if (de->ea_size && hpfs_sb(dir->i_sb)->sb_eas) + else if (le32_to_cpu(de->ea_size) && hpfs_sb(dir->i_sb)->sb_eas) hpfs_read_inode(result); else { result->i_mode |= S_IFREG; @@ -250,8 +250,6 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name hpfs_result = hpfs_i(result); if (!de->directory) hpfs_result->i_parent_dir = dir->i_ino; - hpfs_decide_conv(result, name, len); - if (de->has_acl || de->has_xtd_perm) if (!(dir->i_sb->s_flags & MS_RDONLY)) { hpfs_error(result->i_sb, "ACLs or XPERM found. This is probably HPFS386. This driver doesn't support it now. Send me some info on these structures"); goto bail1; @@ -263,19 +261,19 @@ struct dentry *hpfs_lookup(struct inode *dir, struct dentry *dentry, struct name */ if (!result->i_ctime.tv_sec) { - if (!(result->i_ctime.tv_sec = local_to_gmt(dir->i_sb, de->creation_date))) + if (!(result->i_ctime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(de->creation_date)))) result->i_ctime.tv_sec = 1; result->i_ctime.tv_nsec = 0; - result->i_mtime.tv_sec = local_to_gmt(dir->i_sb, de->write_date); + result->i_mtime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(de->write_date)); result->i_mtime.tv_nsec = 0; - result->i_atime.tv_sec = local_to_gmt(dir->i_sb, de->read_date); + result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(de->read_date)); result->i_atime.tv_nsec = 0; - hpfs_result->i_ea_size = de->ea_size; + hpfs_result->i_ea_size = le32_to_cpu(de->ea_size); if (!hpfs_result->i_ea_mode && de->read_only) result->i_mode &= ~0222; if (!de->directory) { if (result->i_size == -1) { - result->i_size = de->file_size; + result->i_size = le32_to_cpu(de->file_size); result->i_data.a_ops = &hpfs_aops; hpfs_i(result)->mmu_private = result->i_size; /* diff --git a/fs/hpfs/dnode.c b/fs/hpfs/dnode.c index 9b2ffadfc8c4..1e0e2ac30fd3 100644 --- a/fs/hpfs/dnode.c +++ b/fs/hpfs/dnode.c @@ -14,11 +14,11 @@ static loff_t get_pos(struct dnode *d, struct hpfs_dirent *fde) struct hpfs_dirent *de_end = dnode_end_de(d); int i = 1; for (de = dnode_first_de(d); de < de_end; de = de_next_de(de)) { - if (de == fde) return ((loff_t) d->self << 4) | (loff_t)i; + if (de == fde) return ((loff_t) le32_to_cpu(d->self) << 4) | (loff_t)i; i++; } printk("HPFS: get_pos: not_found\n"); - return ((loff_t)d->self << 4) | (loff_t)1; + return ((loff_t)le32_to_cpu(d->self) << 4) | (loff_t)1; } void hpfs_add_pos(struct inode *inode, loff_t *pos) @@ -130,29 +130,30 @@ static void set_last_pointer(struct super_block *s, struct dnode *d, dnode_secno { struct hpfs_dirent *de; if (!(de = dnode_last_de(d))) { - hpfs_error(s, "set_last_pointer: empty dnode %08x", d->self); + hpfs_error(s, "set_last_pointer: empty dnode %08x", le32_to_cpu(d->self)); return; } if (hpfs_sb(s)->sb_chk) { if (de->down) { hpfs_error(s, "set_last_pointer: dnode %08x has already last pointer %08x", - d->self, de_down_pointer(de)); + le32_to_cpu(d->self), de_down_pointer(de)); return; } - if (de->length != 32) { - hpfs_error(s, "set_last_pointer: bad last dirent in dnode %08x", d->self); + if (le16_to_cpu(de->length) != 32) { + hpfs_error(s, "set_last_pointer: bad last dirent in dnode %08x", le32_to_cpu(d->self)); return; } } if (ptr) { - if ((d->first_free += 4) > 2048) { - hpfs_error(s,"set_last_pointer: too long dnode %08x", d->self); - d->first_free -= 4; + d->first_free = cpu_to_le32(le32_to_cpu(d->first_free) + 4); + if (le32_to_cpu(d->first_free) > 2048) { + hpfs_error(s, "set_last_pointer: too long dnode %08x", le32_to_cpu(d->self)); + d->first_free = cpu_to_le32(le32_to_cpu(d->first_free) - 4); return; } - de->length = 36; + de->length = cpu_to_le16(36); de->down = 1; - *(dnode_secno *)((char *)de + 32) = ptr; + *(dnode_secno *)((char *)de + 32) = cpu_to_le32(ptr); } } @@ -168,7 +169,7 @@ struct hpfs_dirent *hpfs_add_de(struct super_block *s, struct dnode *d, for (de = dnode_first_de(d); de < de_end; de = de_next_de(de)) { int c = hpfs_compare_names(s, name, namelen, de->name, de->namelen, de->last); if (!c) { - hpfs_error(s, "name (%c,%d) already exists in dnode %08x", *name, namelen, d->self); + hpfs_error(s, "name (%c,%d) already exists in dnode %08x", *name, namelen, le32_to_cpu(d->self)); return NULL; } if (c < 0) break; @@ -176,15 +177,14 @@ struct hpfs_dirent *hpfs_add_de(struct super_block *s, struct dnode *d, memmove((char *)de + d_size, de, (char *)de_end - (char *)de); memset(de, 0, d_size); if (down_ptr) { - *(int *)((char *)de + d_size - 4) = down_ptr; + *(dnode_secno *)((char *)de + d_size - 4) = cpu_to_le32(down_ptr); de->down = 1; } - de->length = d_size; - if (down_ptr) de->down = 1; + de->length = cpu_to_le16(d_size); de->not_8x3 = hpfs_is_name_long(name, namelen); de->namelen = namelen; memcpy(de->name, name, namelen); - d->first_free += d_size; + d->first_free = cpu_to_le32(le32_to_cpu(d->first_free) + d_size); return de; } @@ -194,25 +194,25 @@ static void hpfs_delete_de(struct super_block *s, struct dnode *d, struct hpfs_dirent *de) { if (de->last) { - hpfs_error(s, "attempt to delete last dirent in dnode %08x", d->self); + hpfs_error(s, "attempt to delete last dirent in dnode %08x", le32_to_cpu(d->self)); return; } - d->first_free -= de->length; - memmove(de, de_next_de(de), d->first_free + (char *)d - (char *)de); + d->first_free = cpu_to_le32(le32_to_cpu(d->first_free) - le16_to_cpu(de->length)); + memmove(de, de_next_de(de), le32_to_cpu(d->first_free) + (char *)d - (char *)de); } static void fix_up_ptrs(struct super_block *s, struct dnode *d) { struct hpfs_dirent *de; struct hpfs_dirent *de_end = dnode_end_de(d); - dnode_secno dno = d->self; + dnode_secno dno = le32_to_cpu(d->self); for (de = dnode_first_de(d); de < de_end; de = de_next_de(de)) if (de->down) { struct quad_buffer_head qbh; struct dnode *dd; if ((dd = hpfs_map_dnode(s, de_down_pointer(de), &qbh))) { - if (dd->up != dno || dd->root_dnode) { - dd->up = dno; + if (le32_to_cpu(dd->up) != dno || dd->root_dnode) { + dd->up = cpu_to_le32(dno); dd->root_dnode = 0; hpfs_mark_4buffers_dirty(&qbh); } @@ -262,7 +262,7 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno, kfree(nname); return 1; } - if (d->first_free + de_size(namelen, down_ptr) <= 2048) { + if (le32_to_cpu(d->first_free) + de_size(namelen, down_ptr) <= 2048) { loff_t t; copy_de(de=hpfs_add_de(i->i_sb, d, name, namelen, down_ptr), new_de); t = get_pos(d, de); @@ -286,11 +286,11 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno, kfree(nname); return 1; } - memcpy(nd, d, d->first_free); + memcpy(nd, d, le32_to_cpu(d->first_free)); copy_de(de = hpfs_add_de(i->i_sb, nd, name, namelen, down_ptr), new_de); for_all_poss(i, hpfs_pos_ins, get_pos(nd, de), 1); h = ((char *)dnode_last_de(nd) - (char *)nd) / 2 + 10; - if (!(ad = hpfs_alloc_dnode(i->i_sb, d->up, &adno, &qbh1, 0))) { + if (!(ad = hpfs_alloc_dnode(i->i_sb, le32_to_cpu(d->up), &adno, &qbh1))) { hpfs_error(i->i_sb, "unable to alloc dnode - dnode tree will be corrupted"); hpfs_brelse4(&qbh); kfree(nd); @@ -313,20 +313,21 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno, down_ptr = adno; set_last_pointer(i->i_sb, ad, de->down ? de_down_pointer(de) : 0); de = de_next_de(de); - memmove((char *)nd + 20, de, nd->first_free + (char *)nd - (char *)de); - nd->first_free -= (char *)de - (char *)nd - 20; - memcpy(d, nd, nd->first_free); + memmove((char *)nd + 20, de, le32_to_cpu(nd->first_free) + (char *)nd - (char *)de); + nd->first_free = cpu_to_le32(le32_to_cpu(nd->first_free) - ((char *)de - (char *)nd - 20)); + memcpy(d, nd, le32_to_cpu(nd->first_free)); for_all_poss(i, hpfs_pos_del, (loff_t)dno << 4, pos); fix_up_ptrs(i->i_sb, ad); if (!d->root_dnode) { - dno = ad->up = d->up; + ad->up = d->up; + dno = le32_to_cpu(ad->up); hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); hpfs_mark_4buffers_dirty(&qbh1); hpfs_brelse4(&qbh1); goto go_up; } - if (!(rd = hpfs_alloc_dnode(i->i_sb, d->up, &rdno, &qbh2, 0))) { + if (!(rd = hpfs_alloc_dnode(i->i_sb, le32_to_cpu(d->up), &rdno, &qbh2))) { hpfs_error(i->i_sb, "unable to alloc dnode - dnode tree will be corrupted"); hpfs_brelse4(&qbh); hpfs_brelse4(&qbh1); @@ -338,7 +339,7 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno, i->i_blocks += 4; rd->root_dnode = 1; rd->up = d->up; - if (!(fnode = hpfs_map_fnode(i->i_sb, d->up, &bh))) { + if (!(fnode = hpfs_map_fnode(i->i_sb, le32_to_cpu(d->up), &bh))) { hpfs_free_dnode(i->i_sb, rdno); hpfs_brelse4(&qbh); hpfs_brelse4(&qbh1); @@ -347,10 +348,11 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno, kfree(nname); return 1; } - fnode->u.external[0].disk_secno = rdno; + fnode->u.external[0].disk_secno = cpu_to_le32(rdno); mark_buffer_dirty(bh); brelse(bh); - d->up = ad->up = hpfs_i(i)->i_dno = rdno; + hpfs_i(i)->i_dno = rdno; + d->up = ad->up = cpu_to_le32(rdno); d->root_dnode = ad->root_dnode = 0; hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); @@ -373,7 +375,7 @@ static int hpfs_add_to_dnode(struct inode *i, dnode_secno dno, int hpfs_add_dirent(struct inode *i, const unsigned char *name, unsigned namelen, - struct hpfs_dirent *new_de, int cdepth) + struct hpfs_dirent *new_de) { struct hpfs_inode_info *hpfs_inode = hpfs_i(i); struct dnode *d; @@ -403,7 +405,6 @@ int hpfs_add_dirent(struct inode *i, } } hpfs_brelse4(&qbh); - if (!cdepth) hpfs_lock_creation(i->i_sb); if (hpfs_check_free_dnodes(i->i_sb, FREE_DNODES_ADD)) { c = 1; goto ret; @@ -411,7 +412,6 @@ int hpfs_add_dirent(struct inode *i, i->i_version++; c = hpfs_add_to_dnode(i, dno, name, namelen, new_de, 0); ret: - if (!cdepth) hpfs_unlock_creation(i->i_sb); return c; } @@ -437,9 +437,9 @@ static secno move_to_top(struct inode *i, dnode_secno from, dnode_secno to) return 0; if (!(dnode = hpfs_map_dnode(i->i_sb, dno, &qbh))) return 0; if (hpfs_sb(i->i_sb)->sb_chk) { - if (dnode->up != chk_up) { + if (le32_to_cpu(dnode->up) != chk_up) { hpfs_error(i->i_sb, "move_to_top: up pointer from %08x should be %08x, is %08x", - dno, chk_up, dnode->up); + dno, chk_up, le32_to_cpu(dnode->up)); hpfs_brelse4(&qbh); return 0; } @@ -455,7 +455,7 @@ static secno move_to_top(struct inode *i, dnode_secno from, dnode_secno to) hpfs_brelse4(&qbh); } while (!(de = dnode_pre_last_de(dnode))) { - dnode_secno up = dnode->up; + dnode_secno up = le32_to_cpu(dnode->up); hpfs_brelse4(&qbh); hpfs_free_dnode(i->i_sb, dno); i->i_size -= 2048; @@ -474,8 +474,8 @@ static secno move_to_top(struct inode *i, dnode_secno from, dnode_secno to) hpfs_brelse4(&qbh); return 0; } - dnode->first_free -= 4; - de->length -= 4; + dnode->first_free = cpu_to_le32(le32_to_cpu(dnode->first_free) - 4); + de->length = cpu_to_le16(le16_to_cpu(de->length) - 4); de->down = 0; hpfs_mark_4buffers_dirty(&qbh); dno = up; @@ -483,12 +483,12 @@ static secno move_to_top(struct inode *i, dnode_secno from, dnode_secno to) t = get_pos(dnode, de); for_all_poss(i, hpfs_pos_subst, t, 4); for_all_poss(i, hpfs_pos_subst, t + 1, 5); - if (!(nde = kmalloc(de->length, GFP_NOFS))) { + if (!(nde = kmalloc(le16_to_cpu(de->length), GFP_NOFS))) { hpfs_error(i->i_sb, "out of memory for dirent - directory will be corrupted"); hpfs_brelse4(&qbh); return 0; } - memcpy(nde, de, de->length); + memcpy(nde, de, le16_to_cpu(de->length)); ddno = de->down ? de_down_pointer(de) : 0; hpfs_delete_de(i->i_sb, dnode, de); set_last_pointer(i->i_sb, dnode, ddno); @@ -517,11 +517,11 @@ static void delete_empty_dnode(struct inode *i, dnode_secno dno) try_it_again: if (hpfs_stop_cycles(i->i_sb, dno, &c1, &c2, "delete_empty_dnode")) return; if (!(dnode = hpfs_map_dnode(i->i_sb, dno, &qbh))) return; - if (dnode->first_free > 56) goto end; - if (dnode->first_free == 52 || dnode->first_free == 56) { + if (le32_to_cpu(dnode->first_free) > 56) goto end; + if (le32_to_cpu(dnode->first_free) == 52 || le32_to_cpu(dnode->first_free) == 56) { struct hpfs_dirent *de_end; int root = dnode->root_dnode; - up = dnode->up; + up = le32_to_cpu(dnode->up); de = dnode_first_de(dnode); down = de->down ? de_down_pointer(de) : 0; if (hpfs_sb(i->i_sb)->sb_chk) if (root && !down) { @@ -545,13 +545,13 @@ static void delete_empty_dnode(struct inode *i, dnode_secno dno) return; } if ((d1 = hpfs_map_dnode(i->i_sb, down, &qbh1))) { - d1->up = up; + d1->up = cpu_to_le32(up); d1->root_dnode = 1; hpfs_mark_4buffers_dirty(&qbh1); hpfs_brelse4(&qbh1); } if ((fnode = hpfs_map_fnode(i->i_sb, up, &bh))) { - fnode->u.external[0].disk_secno = down; + fnode->u.external[0].disk_secno = cpu_to_le32(down); mark_buffer_dirty(bh); brelse(bh); } @@ -570,22 +570,22 @@ static void delete_empty_dnode(struct inode *i, dnode_secno dno) for_all_poss(i, hpfs_pos_subst, ((loff_t)dno << 4) | 1, ((loff_t)up << 4) | p); if (!down) { de->down = 0; - de->length -= 4; - dnode->first_free -= 4; + de->length = cpu_to_le16(le16_to_cpu(de->length) - 4); + dnode->first_free = cpu_to_le32(le32_to_cpu(dnode->first_free) - 4); memmove(de_next_de(de), (char *)de_next_de(de) + 4, - (char *)dnode + dnode->first_free - (char *)de_next_de(de)); + (char *)dnode + le32_to_cpu(dnode->first_free) - (char *)de_next_de(de)); } else { struct dnode *d1; struct quad_buffer_head qbh1; - *(dnode_secno *) ((void *) de + de->length - 4) = down; + *(dnode_secno *) ((void *) de + le16_to_cpu(de->length) - 4) = down; if ((d1 = hpfs_map_dnode(i->i_sb, down, &qbh1))) { - d1->up = up; + d1->up = cpu_to_le32(up); hpfs_mark_4buffers_dirty(&qbh1); hpfs_brelse4(&qbh1); } } } else { - hpfs_error(i->i_sb, "delete_empty_dnode: dnode %08x, first_free == %03x", dno, dnode->first_free); + hpfs_error(i->i_sb, "delete_empty_dnode: dnode %08x, first_free == %03x", dno, le32_to_cpu(dnode->first_free)); goto end; } @@ -596,18 +596,18 @@ static void delete_empty_dnode(struct inode *i, dnode_secno dno) struct quad_buffer_head qbh1; if (!de_next->down) goto endm; ndown = de_down_pointer(de_next); - if (!(de_cp = kmalloc(de->length, GFP_NOFS))) { + if (!(de_cp = kmalloc(le16_to_cpu(de->length), GFP_NOFS))) { printk("HPFS: out of memory for dtree balancing\n"); goto endm; } - memcpy(de_cp, de, de->length); + memcpy(de_cp, de, le16_to_cpu(de->length)); hpfs_delete_de(i->i_sb, dnode, de); hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); for_all_poss(i, hpfs_pos_subst, ((loff_t)up << 4) | p, 4); for_all_poss(i, hpfs_pos_del, ((loff_t)up << 4) | p, 1); if (de_cp->down) if ((d1 = hpfs_map_dnode(i->i_sb, de_down_pointer(de_cp), &qbh1))) { - d1->up = ndown; + d1->up = cpu_to_le32(ndown); hpfs_mark_4buffers_dirty(&qbh1); hpfs_brelse4(&qbh1); } @@ -635,7 +635,7 @@ static void delete_empty_dnode(struct inode *i, dnode_secno dno) struct hpfs_dirent *del = dnode_last_de(d1); dlp = del->down ? de_down_pointer(del) : 0; if (!dlp && down) { - if (d1->first_free > 2044) { + if (le32_to_cpu(d1->first_free) > 2044) { if (hpfs_sb(i->i_sb)->sb_chk >= 2) { printk("HPFS: warning: unbalanced dnode tree, see hpfs.txt 4 more info\n"); printk("HPFS: warning: terminating balancing operation\n"); @@ -647,38 +647,38 @@ static void delete_empty_dnode(struct inode *i, dnode_secno dno) printk("HPFS: warning: unbalanced dnode tree, see hpfs.txt 4 more info\n"); printk("HPFS: warning: goin'on\n"); } - del->length += 4; + del->length = cpu_to_le16(le16_to_cpu(del->length) + 4); del->down = 1; - d1->first_free += 4; + d1->first_free = cpu_to_le32(le32_to_cpu(d1->first_free) + 4); } if (dlp && !down) { - del->length -= 4; + del->length = cpu_to_le16(le16_to_cpu(del->length) - 4); del->down = 0; - d1->first_free -= 4; + d1->first_free = cpu_to_le32(le32_to_cpu(d1->first_free) - 4); } else if (down) - *(dnode_secno *) ((void *) del + del->length - 4) = down; + *(dnode_secno *) ((void *) del + le16_to_cpu(del->length) - 4) = cpu_to_le32(down); } else goto endm; - if (!(de_cp = kmalloc(de_prev->length, GFP_NOFS))) { + if (!(de_cp = kmalloc(le16_to_cpu(de_prev->length), GFP_NOFS))) { printk("HPFS: out of memory for dtree balancing\n"); hpfs_brelse4(&qbh1); goto endm; } hpfs_mark_4buffers_dirty(&qbh1); hpfs_brelse4(&qbh1); - memcpy(de_cp, de_prev, de_prev->length); + memcpy(de_cp, de_prev, le16_to_cpu(de_prev->length)); hpfs_delete_de(i->i_sb, dnode, de_prev); if (!de_prev->down) { - de_prev->length += 4; + de_prev->length = cpu_to_le16(le16_to_cpu(de_prev->length) + 4); de_prev->down = 1; - dnode->first_free += 4; + dnode->first_free = cpu_to_le32(le32_to_cpu(dnode->first_free) + 4); } - *(dnode_secno *) ((void *) de_prev + de_prev->length - 4) = ndown; + *(dnode_secno *) ((void *) de_prev + le16_to_cpu(de_prev->length) - 4) = cpu_to_le32(ndown); hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); for_all_poss(i, hpfs_pos_subst, ((loff_t)up << 4) | (p - 1), 4); for_all_poss(i, hpfs_pos_subst, ((loff_t)up << 4) | p, ((loff_t)up << 4) | (p - 1)); if (down) if ((d1 = hpfs_map_dnode(i->i_sb, de_down_pointer(de), &qbh1))) { - d1->up = ndown; + d1->up = cpu_to_le32(ndown); hpfs_mark_4buffers_dirty(&qbh1); hpfs_brelse4(&qbh1); } @@ -701,7 +701,6 @@ int hpfs_remove_dirent(struct inode *i, dnode_secno dno, struct hpfs_dirent *de, { struct dnode *dnode = qbh->data; dnode_secno down = 0; - int lock = 0; loff_t t; if (de->first || de->last) { hpfs_error(i->i_sb, "hpfs_remove_dirent: attempt to delete first or last dirent in dnode %08x", dno); @@ -710,11 +709,8 @@ int hpfs_remove_dirent(struct inode *i, dnode_secno dno, struct hpfs_dirent *de, } if (de->down) down = de_down_pointer(de); if (depth && (de->down || (de == dnode_first_de(dnode) && de_next_de(de)->last))) { - lock = 1; - hpfs_lock_creation(i->i_sb); if (hpfs_check_free_dnodes(i->i_sb, FREE_DNODES_DEL)) { hpfs_brelse4(qbh); - hpfs_unlock_creation(i->i_sb); return 2; } } @@ -727,11 +723,9 @@ int hpfs_remove_dirent(struct inode *i, dnode_secno dno, struct hpfs_dirent *de, dnode_secno a = move_to_top(i, down, dno); for_all_poss(i, hpfs_pos_subst, 5, t); if (a) delete_empty_dnode(i, a); - if (lock) hpfs_unlock_creation(i->i_sb); return !a; } delete_empty_dnode(i, dno); - if (lock) hpfs_unlock_creation(i->i_sb); return 0; } @@ -751,8 +745,8 @@ void hpfs_count_dnodes(struct super_block *s, dnode_secno dno, int *n_dnodes, ptr = 0; go_up: if (!(dnode = hpfs_map_dnode(s, dno, &qbh))) return; - if (hpfs_sb(s)->sb_chk) if (odno && odno != -1 && dnode->up != odno) - hpfs_error(s, "hpfs_count_dnodes: bad up pointer; dnode %08x, down %08x points to %08x", odno, dno, dnode->up); + if (hpfs_sb(s)->sb_chk) if (odno && odno != -1 && le32_to_cpu(dnode->up) != odno) + hpfs_error(s, "hpfs_count_dnodes: bad up pointer; dnode %08x, down %08x points to %08x", odno, dno, le32_to_cpu(dnode->up)); de = dnode_first_de(dnode); if (ptr) while(1) { if (de->down) if (de_down_pointer(de) == ptr) goto process_de; @@ -776,7 +770,7 @@ void hpfs_count_dnodes(struct super_block *s, dnode_secno dno, int *n_dnodes, if (!de->first && !de->last && n_items) (*n_items)++; if ((de = de_next_de(de)) < dnode_end_de(dnode)) goto next_de; ptr = dno; - dno = dnode->up; + dno = le32_to_cpu(dnode->up); if (dnode->root_dnode) { hpfs_brelse4(&qbh); return; @@ -824,8 +818,8 @@ dnode_secno hpfs_de_as_down_as_possible(struct super_block *s, dnode_secno dno) return d; if (!(de = map_nth_dirent(s, d, 1, &qbh, NULL))) return dno; if (hpfs_sb(s)->sb_chk) - if (up && ((struct dnode *)qbh.data)->up != up) - hpfs_error(s, "hpfs_de_as_down_as_possible: bad up pointer; dnode %08x, down %08x points to %08x", up, d, ((struct dnode *)qbh.data)->up); + if (up && le32_to_cpu(((struct dnode *)qbh.data)->up) != up) + hpfs_error(s, "hpfs_de_as_down_as_possible: bad up pointer; dnode %08x, down %08x points to %08x", up, d, le32_to_cpu(((struct dnode *)qbh.data)->up)); if (!de->down) { hpfs_brelse4(&qbh); return d; @@ -874,7 +868,7 @@ struct hpfs_dirent *map_pos_dirent(struct inode *inode, loff_t *posp, /* Going up */ if (dnode->root_dnode) goto bail; - if (!(up_dnode = hpfs_map_dnode(inode->i_sb, dnode->up, &qbh0))) + if (!(up_dnode = hpfs_map_dnode(inode->i_sb, le32_to_cpu(dnode->up), &qbh0))) goto bail; end_up_de = dnode_end_de(up_dnode); @@ -882,16 +876,16 @@ struct hpfs_dirent *map_pos_dirent(struct inode *inode, loff_t *posp, for (up_de = dnode_first_de(up_dnode); up_de < end_up_de; up_de = de_next_de(up_de)) { if (!(++c & 077)) hpfs_error(inode->i_sb, - "map_pos_dirent: pos crossed dnode boundary; dnode = %08x", dnode->up); + "map_pos_dirent: pos crossed dnode boundary; dnode = %08x", le32_to_cpu(dnode->up)); if (up_de->down && de_down_pointer(up_de) == dno) { - *posp = ((loff_t) dnode->up << 4) + c; + *posp = ((loff_t) le32_to_cpu(dnode->up) << 4) + c; hpfs_brelse4(&qbh0); return de; } } hpfs_error(inode->i_sb, "map_pos_dirent: pointer to dnode %08x not found in parent dnode %08x", - dno, dnode->up); + dno, le32_to_cpu(dnode->up)); hpfs_brelse4(&qbh0); bail: @@ -1017,17 +1011,17 @@ struct hpfs_dirent *map_fnode_dirent(struct super_block *s, fnode_secno fno, /*name2[15] = 0xff;*/ name1len = 15; name2len = 256; } - if (!(upf = hpfs_map_fnode(s, f->up, &bh))) { + if (!(upf = hpfs_map_fnode(s, le32_to_cpu(f->up), &bh))) { kfree(name2); return NULL; } if (!upf->dirflag) { brelse(bh); - hpfs_error(s, "fnode %08x has non-directory parent %08x", fno, f->up); + hpfs_error(s, "fnode %08x has non-directory parent %08x", fno, le32_to_cpu(f->up)); kfree(name2); return NULL; } - dno = upf->u.external[0].disk_secno; + dno = le32_to_cpu(upf->u.external[0].disk_secno); brelse(bh); go_down: downd = 0; @@ -1049,7 +1043,7 @@ struct hpfs_dirent *map_fnode_dirent(struct super_block *s, fnode_secno fno, return NULL; } next_de: - if (de->fnode == fno) { + if (le32_to_cpu(de->fnode) == fno) { kfree(name2); return de; } @@ -1065,7 +1059,7 @@ struct hpfs_dirent *map_fnode_dirent(struct super_block *s, fnode_secno fno, goto go_down; } f: - if (de->fnode == fno) { + if (le32_to_cpu(de->fnode) == fno) { kfree(name2); return de; } @@ -1074,7 +1068,7 @@ struct hpfs_dirent *map_fnode_dirent(struct super_block *s, fnode_secno fno, if ((de = de_next_de(de)) < de_end) goto next_de; if (d->root_dnode) goto not_found; downd = dno; - dno = d->up; + dno = le32_to_cpu(d->up); hpfs_brelse4(qbh); if (hpfs_sb(s)->sb_chk) if (hpfs_stop_cycles(s, downd, &d1, &d2, "map_fnode_dirent #2")) { diff --git a/fs/hpfs/ea.c b/fs/hpfs/ea.c index 45e53d972b42..d8b84d113c89 100644 --- a/fs/hpfs/ea.c +++ b/fs/hpfs/ea.c @@ -24,7 +24,7 @@ void hpfs_ea_ext_remove(struct super_block *s, secno a, int ano, unsigned len) } if (hpfs_ea_read(s, a, ano, pos, 4, ex)) return; if (ea->indirect) { - if (ea->valuelen != 8) { + if (ea_valuelen(ea) != 8) { hpfs_error(s, "ea->indirect set while ea->valuelen!=8, %s %08x, pos %08x", ano ? "anode" : "sectors", a, pos); return; @@ -33,7 +33,7 @@ void hpfs_ea_ext_remove(struct super_block *s, secno a, int ano, unsigned len) return; hpfs_ea_remove(s, ea_sec(ea), ea->anode, ea_len(ea)); } - pos += ea->namelen + ea->valuelen + 5; + pos += ea->namelen + ea_valuelen(ea) + 5; } if (!ano) hpfs_free_sectors(s, a, (len+511) >> 9); else { @@ -76,24 +76,24 @@ int hpfs_read_ea(struct super_block *s, struct fnode *fnode, char *key, unsigned pos; int ano, len; secno a; + char ex[4 + 255 + 1 + 8]; struct extended_attribute *ea; struct extended_attribute *ea_end = fnode_end_ea(fnode); for (ea = fnode_ea(fnode); ea < ea_end; ea = next_ea(ea)) if (!strcmp(ea->name, key)) { if (ea->indirect) goto indirect; - if (ea->valuelen >= size) + if (ea_valuelen(ea) >= size) return -EINVAL; - memcpy(buf, ea_data(ea), ea->valuelen); - buf[ea->valuelen] = 0; + memcpy(buf, ea_data(ea), ea_valuelen(ea)); + buf[ea_valuelen(ea)] = 0; return 0; } - a = fnode->ea_secno; - len = fnode->ea_size_l; + a = le32_to_cpu(fnode->ea_secno); + len = le32_to_cpu(fnode->ea_size_l); ano = fnode->ea_anode; pos = 0; while (pos < len) { - char ex[4 + 255 + 1 + 8]; ea = (struct extended_attribute *)ex; if (pos + 4 > len) { hpfs_error(s, "EAs don't end correctly, %s %08x, len %08x", @@ -106,14 +106,14 @@ int hpfs_read_ea(struct super_block *s, struct fnode *fnode, char *key, if (!strcmp(ea->name, key)) { if (ea->indirect) goto indirect; - if (ea->valuelen >= size) + if (ea_valuelen(ea) >= size) return -EINVAL; - if (hpfs_ea_read(s, a, ano, pos + 4 + ea->namelen + 1, ea->valuelen, buf)) + if (hpfs_ea_read(s, a, ano, pos + 4 + ea->namelen + 1, ea_valuelen(ea), buf)) return -EIO; - buf[ea->valuelen] = 0; + buf[ea_valuelen(ea)] = 0; return 0; } - pos += ea->namelen + ea->valuelen + 5; + pos += ea->namelen + ea_valuelen(ea) + 5; } return -ENOENT; indirect: @@ -138,16 +138,16 @@ char *hpfs_get_ea(struct super_block *s, struct fnode *fnode, char *key, int *si if (!strcmp(ea->name, key)) { if (ea->indirect) return get_indirect_ea(s, ea->anode, ea_sec(ea), *size = ea_len(ea)); - if (!(ret = kmalloc((*size = ea->valuelen) + 1, GFP_NOFS))) { + if (!(ret = kmalloc((*size = ea_valuelen(ea)) + 1, GFP_NOFS))) { printk("HPFS: out of memory for EA\n"); return NULL; } - memcpy(ret, ea_data(ea), ea->valuelen); - ret[ea->valuelen] = 0; + memcpy(ret, ea_data(ea), ea_valuelen(ea)); + ret[ea_valuelen(ea)] = 0; return ret; } - a = fnode->ea_secno; - len = fnode->ea_size_l; + a = le32_to_cpu(fnode->ea_secno); + len = le32_to_cpu(fnode->ea_size_l); ano = fnode->ea_anode; pos = 0; while (pos < len) { @@ -164,18 +164,18 @@ char *hpfs_get_ea(struct super_block *s, struct fnode *fnode, char *key, int *si if (!strcmp(ea->name, key)) { if (ea->indirect) return get_indirect_ea(s, ea->anode, ea_sec(ea), *size = ea_len(ea)); - if (!(ret = kmalloc((*size = ea->valuelen) + 1, GFP_NOFS))) { + if (!(ret = kmalloc((*size = ea_valuelen(ea)) + 1, GFP_NOFS))) { printk("HPFS: out of memory for EA\n"); return NULL; } - if (hpfs_ea_read(s, a, ano, pos + 4 + ea->namelen + 1, ea->valuelen, ret)) { + if (hpfs_ea_read(s, a, ano, pos + 4 + ea->namelen + 1, ea_valuelen(ea), ret)) { kfree(ret); return NULL; } - ret[ea->valuelen] = 0; + ret[ea_valuelen(ea)] = 0; return ret; } - pos += ea->namelen + ea->valuelen + 5; + pos += ea->namelen + ea_valuelen(ea) + 5; } return NULL; } @@ -202,13 +202,13 @@ void hpfs_set_ea(struct inode *inode, struct fnode *fnode, const char *key, if (ea->indirect) { if (ea_len(ea) == size) set_indirect_ea(s, ea->anode, ea_sec(ea), data, size); - } else if (ea->valuelen == size) { + } else if (ea_valuelen(ea) == size) { memcpy(ea_data(ea), data, size); } return; } - a = fnode->ea_secno; - len = fnode->ea_size_l; + a = le32_to_cpu(fnode->ea_secno); + len = le32_to_cpu(fnode->ea_size_l); ano = fnode->ea_anode; pos = 0; while (pos < len) { @@ -228,68 +228,70 @@ void hpfs_set_ea(struct inode *inode, struct fnode *fnode, const char *key, set_indirect_ea(s, ea->anode, ea_sec(ea), data, size); } else { - if (ea->valuelen == size) + if (ea_valuelen(ea) == size) hpfs_ea_write(s, a, ano, pos + 4 + ea->namelen + 1, size, data); } return; } - pos += ea->namelen + ea->valuelen + 5; + pos += ea->namelen + ea_valuelen(ea) + 5; } - if (!fnode->ea_offs) { - /*if (fnode->ea_size_s) { + if (!le16_to_cpu(fnode->ea_offs)) { + /*if (le16_to_cpu(fnode->ea_size_s)) { hpfs_error(s, "fnode %08x: ea_size_s == %03x, ea_offs == 0", - inode->i_ino, fnode->ea_size_s); + inode->i_ino, le16_to_cpu(fnode->ea_size_s)); return; }*/ - fnode->ea_offs = 0xc4; + fnode->ea_offs = cpu_to_le16(0xc4); } - if (fnode->ea_offs < 0xc4 || fnode->ea_offs + fnode->acl_size_s + fnode->ea_size_s > 0x200) { + if (le16_to_cpu(fnode->ea_offs) < 0xc4 || le16_to_cpu(fnode->ea_offs) + le16_to_cpu(fnode->acl_size_s) + le16_to_cpu(fnode->ea_size_s) > 0x200) { hpfs_error(s, "fnode %08lx: ea_offs == %03x, ea_size_s == %03x", (unsigned long)inode->i_ino, - fnode->ea_offs, fnode->ea_size_s); + le32_to_cpu(fnode->ea_offs), le16_to_cpu(fnode->ea_size_s)); return; } - if ((fnode->ea_size_s || !fnode->ea_size_l) && - fnode->ea_offs + fnode->acl_size_s + fnode->ea_size_s + strlen(key) + size + 5 <= 0x200) { + if ((le16_to_cpu(fnode->ea_size_s) || !le32_to_cpu(fnode->ea_size_l)) && + le16_to_cpu(fnode->ea_offs) + le16_to_cpu(fnode->acl_size_s) + le16_to_cpu(fnode->ea_size_s) + strlen(key) + size + 5 <= 0x200) { ea = fnode_end_ea(fnode); *(char *)ea = 0; ea->namelen = strlen(key); - ea->valuelen = size; + ea->valuelen_lo = size; + ea->valuelen_hi = size >> 8; strcpy(ea->name, key); memcpy(ea_data(ea), data, size); - fnode->ea_size_s += strlen(key) + size + 5; + fnode->ea_size_s = cpu_to_le16(le16_to_cpu(fnode->ea_size_s) + strlen(key) + size + 5); goto ret; } /* Most the code here is 99.9993422% unused. I hope there are no bugs. But what .. HPFS.IFS has also bugs in ea management. */ - if (fnode->ea_size_s && !fnode->ea_size_l) { + if (le16_to_cpu(fnode->ea_size_s) && !le32_to_cpu(fnode->ea_size_l)) { secno n; struct buffer_head *bh; char *data; - if (!(n = hpfs_alloc_sector(s, fno, 1, 0, 1))) return; + if (!(n = hpfs_alloc_sector(s, fno, 1, 0))) return; if (!(data = hpfs_get_sector(s, n, &bh))) { hpfs_free_sectors(s, n, 1); return; } - memcpy(data, fnode_ea(fnode), fnode->ea_size_s); - fnode->ea_size_l = fnode->ea_size_s; - fnode->ea_size_s = 0; - fnode->ea_secno = n; - fnode->ea_anode = 0; + memcpy(data, fnode_ea(fnode), le16_to_cpu(fnode->ea_size_s)); + fnode->ea_size_l = cpu_to_le32(le16_to_cpu(fnode->ea_size_s)); + fnode->ea_size_s = cpu_to_le16(0); + fnode->ea_secno = cpu_to_le32(n); + fnode->ea_anode = cpu_to_le32(0); mark_buffer_dirty(bh); brelse(bh); } - pos = fnode->ea_size_l + 5 + strlen(key) + size; - len = (fnode->ea_size_l + 511) >> 9; + pos = le32_to_cpu(fnode->ea_size_l) + 5 + strlen(key) + size; + len = (le32_to_cpu(fnode->ea_size_l) + 511) >> 9; if (pos >= 30000) goto bail; while (((pos + 511) >> 9) > len) { if (!len) { - if (!(fnode->ea_secno = hpfs_alloc_sector(s, fno, 1, 0, 1))) - goto bail; + secno q = hpfs_alloc_sector(s, fno, 1, 0); + if (!q) goto bail; + fnode->ea_secno = cpu_to_le32(q); fnode->ea_anode = 0; len++; } else if (!fnode->ea_anode) { - if (hpfs_alloc_if_possible(s, fnode->ea_secno + len)) { + if (hpfs_alloc_if_possible(s, le32_to_cpu(fnode->ea_secno) + len)) { len++; } else { /* Aargh... don't know how to create ea anodes :-( */ @@ -298,26 +300,26 @@ void hpfs_set_ea(struct inode *inode, struct fnode *fnode, const char *key, anode_secno a_s; if (!(anode = hpfs_alloc_anode(s, fno, &a_s, &bh))) goto bail; - anode->up = fno; + anode->up = cpu_to_le32(fno); anode->btree.fnode_parent = 1; anode->btree.n_free_nodes--; anode->btree.n_used_nodes++; - anode->btree.first_free += 12; - anode->u.external[0].disk_secno = fnode->ea_secno; - anode->u.external[0].file_secno = 0; - anode->u.external[0].length = len; + anode->btree.first_free = cpu_to_le16(le16_to_cpu(anode->btree.first_free) + 12); + anode->u.external[0].disk_secno = cpu_to_le32(le32_to_cpu(fnode->ea_secno)); + anode->u.external[0].file_secno = cpu_to_le32(0); + anode->u.external[0].length = cpu_to_le32(len); mark_buffer_dirty(bh); brelse(bh); fnode->ea_anode = 1; - fnode->ea_secno = a_s;*/ + fnode->ea_secno = cpu_to_le32(a_s);*/ secno new_sec; int i; - if (!(new_sec = hpfs_alloc_sector(s, fno, 1, 1 - ((pos + 511) >> 9), 1))) + if (!(new_sec = hpfs_alloc_sector(s, fno, 1, 1 - ((pos + 511) >> 9)))) goto bail; for (i = 0; i < len; i++) { struct buffer_head *bh1, *bh2; void *b1, *b2; - if (!(b1 = hpfs_map_sector(s, fnode->ea_secno + i, &bh1, len - i - 1))) { + if (!(b1 = hpfs_map_sector(s, le32_to_cpu(fnode->ea_secno) + i, &bh1, len - i - 1))) { hpfs_free_sectors(s, new_sec, (pos + 511) >> 9); goto bail; } @@ -331,13 +333,13 @@ void hpfs_set_ea(struct inode *inode, struct fnode *fnode, const char *key, mark_buffer_dirty(bh2); brelse(bh2); } - hpfs_free_sectors(s, fnode->ea_secno, len); - fnode->ea_secno = new_sec; + hpfs_free_sectors(s, le32_to_cpu(fnode->ea_secno), len); + fnode->ea_secno = cpu_to_le32(new_sec); len = (pos + 511) >> 9; } } if (fnode->ea_anode) { - if (hpfs_add_sector_to_btree(s, fnode->ea_secno, + if (hpfs_add_sector_to_btree(s, le32_to_cpu(fnode->ea_secno), 0, len) != -1) { len++; } else { @@ -349,17 +351,17 @@ void hpfs_set_ea(struct inode *inode, struct fnode *fnode, const char *key, h[1] = strlen(key); h[2] = size & 0xff; h[3] = size >> 8; - if (hpfs_ea_write(s, fnode->ea_secno, fnode->ea_anode, fnode->ea_size_l, 4, h)) goto bail; - if (hpfs_ea_write(s, fnode->ea_secno, fnode->ea_anode, fnode->ea_size_l + 4, h[1] + 1, key)) goto bail; - if (hpfs_ea_write(s, fnode->ea_secno, fnode->ea_anode, fnode->ea_size_l + 5 + h[1], size, data)) goto bail; - fnode->ea_size_l = pos; + if (hpfs_ea_write(s, le32_to_cpu(fnode->ea_secno), fnode->ea_anode, le32_to_cpu(fnode->ea_size_l), 4, h)) goto bail; + if (hpfs_ea_write(s, le32_to_cpu(fnode->ea_secno), fnode->ea_anode, le32_to_cpu(fnode->ea_size_l) + 4, h[1] + 1, key)) goto bail; + if (hpfs_ea_write(s, le32_to_cpu(fnode->ea_secno), fnode->ea_anode, le32_to_cpu(fnode->ea_size_l) + 5 + h[1], size, data)) goto bail; + fnode->ea_size_l = cpu_to_le32(pos); ret: hpfs_i(inode)->i_ea_size += 5 + strlen(key) + size; return; bail: - if (fnode->ea_secno) - if (fnode->ea_anode) hpfs_truncate_btree(s, fnode->ea_secno, 1, (fnode->ea_size_l + 511) >> 9); - else hpfs_free_sectors(s, fnode->ea_secno + ((fnode->ea_size_l + 511) >> 9), len - ((fnode->ea_size_l + 511) >> 9)); - else fnode->ea_secno = fnode->ea_size_l = 0; + if (le32_to_cpu(fnode->ea_secno)) + if (fnode->ea_anode) hpfs_truncate_btree(s, le32_to_cpu(fnode->ea_secno), 1, (le32_to_cpu(fnode->ea_size_l) + 511) >> 9); + else hpfs_free_sectors(s, le32_to_cpu(fnode->ea_secno) + ((le32_to_cpu(fnode->ea_size_l) + 511) >> 9), len - ((le32_to_cpu(fnode->ea_size_l) + 511) >> 9)); + else fnode->ea_secno = fnode->ea_size_l = cpu_to_le32(0); } diff --git a/fs/hpfs/file.c b/fs/hpfs/file.c index 9b9eb6933e43..89c500ee5213 100644 --- a/fs/hpfs/file.c +++ b/fs/hpfs/file.c @@ -20,8 +20,8 @@ static int hpfs_file_release(struct inode *inode, struct file *file) int hpfs_file_fsync(struct file *file, int datasync) { - /*return file_fsync(file, datasync);*/ - return 0; /* Don't fsync :-) */ + struct inode *inode = file->f_mapping->host; + return sync_blockdev(inode->i_sb->s_bdev); } /* @@ -48,38 +48,46 @@ static secno hpfs_bmap(struct inode *inode, unsigned file_secno) static void hpfs_truncate(struct inode *i) { if (IS_IMMUTABLE(i)) return /*-EPERM*/; - hpfs_lock(i->i_sb); + hpfs_lock_assert(i->i_sb); + hpfs_i(i)->i_n_secs = 0; i->i_blocks = 1 + ((i->i_size + 511) >> 9); hpfs_i(i)->mmu_private = i->i_size; hpfs_truncate_btree(i->i_sb, i->i_ino, 1, ((i->i_size + 511) >> 9)); hpfs_write_inode(i); hpfs_i(i)->i_n_secs = 0; - hpfs_unlock(i->i_sb); } static int hpfs_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) { + int r; secno s; + hpfs_lock(inode->i_sb); s = hpfs_bmap(inode, iblock); if (s) { map_bh(bh_result, inode->i_sb, s); - return 0; + goto ret_0; } - if (!create) return 0; + if (!create) goto ret_0; if (iblock<<9 != hpfs_i(inode)->mmu_private) { BUG(); - return -EIO; + r = -EIO; + goto ret_r; } if ((s = hpfs_add_sector_to_btree(inode->i_sb, inode->i_ino, 1, inode->i_blocks - 1)) == -1) { hpfs_truncate_btree(inode->i_sb, inode->i_ino, 1, inode->i_blocks - 1); - return -ENOSPC; + r = -ENOSPC; + goto ret_r; } inode->i_blocks++; hpfs_i(inode)->mmu_private += 512; set_buffer_new(bh_result); map_bh(bh_result, inode->i_sb, s); - return 0; + ret_0: + r = 0; + ret_r: + hpfs_unlock(inode->i_sb); + return r; } static int hpfs_writepage(struct page *page, struct writeback_control *wbc) @@ -130,8 +138,11 @@ static ssize_t hpfs_file_write(struct file *file, const char __user *buf, ssize_t retval; retval = do_sync_write(file, buf, count, ppos); - if (retval > 0) + if (retval > 0) { + hpfs_lock(file->f_path.dentry->d_sb); hpfs_i(file->f_path.dentry->d_inode)->i_dirty = 1; + hpfs_unlock(file->f_path.dentry->d_sb); + } return retval; } diff --git a/fs/hpfs/hpfs.h b/fs/hpfs/hpfs.h index 0e84c73cd9c4..8b0650aae328 100644 --- a/fs/hpfs/hpfs.h +++ b/fs/hpfs/hpfs.h @@ -19,9 +19,13 @@ For definitive information on HPFS, ask somebody else -- this is guesswork. There are certain to be many mistakes. */ +#if !defined(__LITTLE_ENDIAN) && !defined(__BIG_ENDIAN) +#error unknown endian +#endif + /* Notation */ -typedef unsigned secno; /* sector number, partition relative */ +typedef u32 secno; /* sector number, partition relative */ typedef secno dnode_secno; /* sector number of a dnode */ typedef secno fnode_secno; /* sector number of an fnode */ @@ -38,28 +42,28 @@ typedef u32 time32_t; /* 32-bit time_t type */ struct hpfs_boot_block { - unsigned char jmp[3]; - unsigned char oem_id[8]; - unsigned char bytes_per_sector[2]; /* 512 */ - unsigned char sectors_per_cluster; - unsigned char n_reserved_sectors[2]; - unsigned char n_fats; - unsigned char n_rootdir_entries[2]; - unsigned char n_sectors_s[2]; - unsigned char media_byte; - unsigned short sectors_per_fat; - unsigned short sectors_per_track; - unsigned short heads_per_cyl; - unsigned int n_hidden_sectors; - unsigned int n_sectors_l; /* size of partition */ - unsigned char drive_number; - unsigned char mbz; - unsigned char sig_28h; /* 28h */ - unsigned char vol_serno[4]; - unsigned char vol_label[11]; - unsigned char sig_hpfs[8]; /* "HPFS " */ - unsigned char pad[448]; - unsigned short magic; /* aa55 */ + u8 jmp[3]; + u8 oem_id[8]; + u8 bytes_per_sector[2]; /* 512 */ + u8 sectors_per_cluster; + u8 n_reserved_sectors[2]; + u8 n_fats; + u8 n_rootdir_entries[2]; + u8 n_sectors_s[2]; + u8 media_byte; + u16 sectors_per_fat; + u16 sectors_per_track; + u16 heads_per_cyl; + u32 n_hidden_sectors; + u32 n_sectors_l; /* size of partition */ + u8 drive_number; + u8 mbz; + u8 sig_28h; /* 28h */ + u8 vol_serno[4]; + u8 vol_label[11]; + u8 sig_hpfs[8]; /* "HPFS " */ + u8 pad[448]; + u16 magic; /* aa55 */ }; @@ -71,31 +75,29 @@ struct hpfs_boot_block struct hpfs_super_block { - unsigned magic; /* f995 e849 */ - unsigned magic1; /* fa53 e9c5, more magic? */ - /*unsigned huh202;*/ /* ?? 202 = N. of B. in 1.00390625 S.*/ - char version; /* version of a filesystem usually 2 */ - char funcversion; /* functional version - oldest version + u32 magic; /* f995 e849 */ + u32 magic1; /* fa53 e9c5, more magic? */ + u8 version; /* version of a filesystem usually 2 */ + u8 funcversion; /* functional version - oldest version of filesystem that can understand this disk */ - unsigned short int zero; /* 0 */ + u16 zero; /* 0 */ fnode_secno root; /* fnode of root directory */ secno n_sectors; /* size of filesystem */ - unsigned n_badblocks; /* number of bad blocks */ + u32 n_badblocks; /* number of bad blocks */ secno bitmaps; /* pointers to free space bit maps */ - unsigned zero1; /* 0 */ + u32 zero1; /* 0 */ secno badblocks; /* bad block list */ - unsigned zero3; /* 0 */ + u32 zero3; /* 0 */ time32_t last_chkdsk; /* date last checked, 0 if never */ - /*unsigned zero4;*/ /* 0 */ - time32_t last_optimize; /* date last optimized, 0 if never */ + time32_t last_optimize; /* date last optimized, 0 if never */ secno n_dir_band; /* number of sectors in dir band */ secno dir_band_start; /* first sector in dir band */ secno dir_band_end; /* last sector in dir band */ secno dir_band_bitmap; /* free space map, 1 dnode per bit */ - char volume_name[32]; /* not used */ + u8 volume_name[32]; /* not used */ secno user_id_table; /* 8 preallocated sectors - user id */ - unsigned zero6[103]; /* 0 */ + u32 zero6[103]; /* 0 */ }; @@ -107,44 +109,65 @@ struct hpfs_super_block struct hpfs_spare_block { - unsigned magic; /* f991 1849 */ - unsigned magic1; /* fa52 29c5, more magic? */ - - unsigned dirty: 1; /* 0 clean, 1 "improperly stopped" */ - /*unsigned flag1234: 4;*/ /* unknown flags */ - unsigned sparedir_used: 1; /* spare dirblks used */ - unsigned hotfixes_used: 1; /* hotfixes used */ - unsigned bad_sector: 1; /* bad sector, corrupted disk (???) */ - unsigned bad_bitmap: 1; /* bad bitmap */ - unsigned fast: 1; /* partition was fast formatted */ - unsigned old_wrote: 1; /* old version wrote to partion */ - unsigned old_wrote_1: 1; /* old version wrote to partion (?) */ - unsigned install_dasd_limits: 1; /* HPFS386 flags */ - unsigned resynch_dasd_limits: 1; - unsigned dasd_limits_operational: 1; - unsigned multimedia_active: 1; - unsigned dce_acls_active: 1; - unsigned dasd_limits_dirty: 1; - unsigned flag67: 2; - unsigned char mm_contlgulty; - unsigned char unused; + u32 magic; /* f991 1849 */ + u32 magic1; /* fa52 29c5, more magic? */ + +#ifdef __LITTLE_ENDIAN + u8 dirty: 1; /* 0 clean, 1 "improperly stopped" */ + u8 sparedir_used: 1; /* spare dirblks used */ + u8 hotfixes_used: 1; /* hotfixes used */ + u8 bad_sector: 1; /* bad sector, corrupted disk (???) */ + u8 bad_bitmap: 1; /* bad bitmap */ + u8 fast: 1; /* partition was fast formatted */ + u8 old_wrote: 1; /* old version wrote to partion */ + u8 old_wrote_1: 1; /* old version wrote to partion (?) */ +#else + u8 old_wrote_1: 1; /* old version wrote to partion (?) */ + u8 old_wrote: 1; /* old version wrote to partion */ + u8 fast: 1; /* partition was fast formatted */ + u8 bad_bitmap: 1; /* bad bitmap */ + u8 bad_sector: 1; /* bad sector, corrupted disk (???) */ + u8 hotfixes_used: 1; /* hotfixes used */ + u8 sparedir_used: 1; /* spare dirblks used */ + u8 dirty: 1; /* 0 clean, 1 "improperly stopped" */ +#endif + +#ifdef __LITTLE_ENDIAN + u8 install_dasd_limits: 1; /* HPFS386 flags */ + u8 resynch_dasd_limits: 1; + u8 dasd_limits_operational: 1; + u8 multimedia_active: 1; + u8 dce_acls_active: 1; + u8 dasd_limits_dirty: 1; + u8 flag67: 2; +#else + u8 flag67: 2; + u8 dasd_limits_dirty: 1; + u8 dce_acls_active: 1; + u8 multimedia_active: 1; + u8 dasd_limits_operational: 1; + u8 resynch_dasd_limits: 1; + u8 install_dasd_limits: 1; /* HPFS386 flags */ +#endif + + u8 mm_contlgulty; + u8 unused; secno hotfix_map; /* info about remapped bad sectors */ - unsigned n_spares_used; /* number of hotfixes */ - unsigned n_spares; /* number of spares in hotfix map */ - unsigned n_dnode_spares_free; /* spare dnodes unused */ - unsigned n_dnode_spares; /* length of spare_dnodes[] list, + u32 n_spares_used; /* number of hotfixes */ + u32 n_spares; /* number of spares in hotfix map */ + u32 n_dnode_spares_free; /* spare dnodes unused */ + u32 n_dnode_spares; /* length of spare_dnodes[] list, follows in this block*/ secno code_page_dir; /* code page directory block */ - unsigned n_code_pages; /* number of code pages */ - /*unsigned large_numbers[2];*/ /* ?? */ - unsigned super_crc; /* on HPFS386 and LAN Server this is + u32 n_code_pages; /* number of code pages */ + u32 super_crc; /* on HPFS386 and LAN Server this is checksum of superblock, on normal OS/2 unused */ - unsigned spare_crc; /* on HPFS386 checksum of spareblock */ - unsigned zero1[15]; /* unused */ + u32 spare_crc; /* on HPFS386 checksum of spareblock */ + u32 zero1[15]; /* unused */ dnode_secno spare_dnodes[100]; /* emergency free dnode list */ - unsigned zero2[1]; /* room for more? */ + u32 zero2[1]; /* room for more? */ }; /* The bad block list is 4 sectors long. The first word must be zero, @@ -179,18 +202,18 @@ struct hpfs_spare_block struct code_page_directory { - unsigned magic; /* 4945 21f7 */ - unsigned n_code_pages; /* number of pointers following */ - unsigned zero1[2]; + u32 magic; /* 4945 21f7 */ + u32 n_code_pages; /* number of pointers following */ + u32 zero1[2]; struct { - unsigned short ix; /* index */ - unsigned short code_page_number; /* code page number */ - unsigned bounds; /* matches corresponding word + u16 ix; /* index */ + u16 code_page_number; /* code page number */ + u32 bounds; /* matches corresponding word in data block */ secno code_page_data; /* sector number of a code_page_data containing c.p. array */ - unsigned short index; /* index in c.p. array in that sector*/ - unsigned short unknown; /* some unknown value; usually 0; + u16 index; /* index in c.p. array in that sector*/ + u16 unknown; /* some unknown value; usually 0; 2 in Japanese version */ } array[31]; /* unknown length */ }; @@ -201,21 +224,21 @@ struct code_page_directory struct code_page_data { - unsigned magic; /* 8945 21f7 */ - unsigned n_used; /* # elements used in c_p_data[] */ - unsigned bounds[3]; /* looks a bit like + u32 magic; /* 8945 21f7 */ + u32 n_used; /* # elements used in c_p_data[] */ + u32 bounds[3]; /* looks a bit like (beg1,end1), (beg2,end2) one byte each */ - unsigned short offs[3]; /* offsets from start of sector + u16 offs[3]; /* offsets from start of sector to start of c_p_data[ix] */ struct { - unsigned short ix; /* index */ - unsigned short code_page_number; /* code page number */ - unsigned short unknown; /* the same as in cp directory */ - unsigned char map[128]; /* upcase table for chars 80..ff */ - unsigned short zero2; + u16 ix; /* index */ + u16 code_page_number; /* code page number */ + u16 unknown; /* the same as in cp directory */ + u8 map[128]; /* upcase table for chars 80..ff */ + u16 zero2; } code_page[3]; - unsigned char incognita[78]; + u8 incognita[78]; }; @@ -255,50 +278,84 @@ struct code_page_data #define DNODE_MAGIC 0x77e40aae struct dnode { - unsigned magic; /* 77e4 0aae */ - unsigned first_free; /* offset from start of dnode to + u32 magic; /* 77e4 0aae */ + u32 first_free; /* offset from start of dnode to first free dir entry */ - unsigned root_dnode:1; /* Is it root dnode? */ - unsigned increment_me:31; /* some kind of activity counter? - Neither HPFS.IFS nor CHKDSK cares +#ifdef __LITTLE_ENDIAN + u8 root_dnode: 1; /* Is it root dnode? */ + u8 increment_me: 7; /* some kind of activity counter? */ + /* Neither HPFS.IFS nor CHKDSK cares + if you change this word */ +#else + u8 increment_me: 7; /* some kind of activity counter? */ + /* Neither HPFS.IFS nor CHKDSK cares if you change this word */ + u8 root_dnode: 1; /* Is it root dnode? */ +#endif + u8 increment_me2[3]; secno up; /* (root dnode) directory's fnode (nonroot) parent dnode */ dnode_secno self; /* pointer to this dnode */ - unsigned char dirent[2028]; /* one or more dirents */ + u8 dirent[2028]; /* one or more dirents */ }; struct hpfs_dirent { - unsigned short length; /* offset to next dirent */ - unsigned first: 1; /* set on phony ^A^A (".") entry */ - unsigned has_acl: 1; - unsigned down: 1; /* down pointer present (after name) */ - unsigned last: 1; /* set on phony \377 entry */ - unsigned has_ea: 1; /* entry has EA */ - unsigned has_xtd_perm: 1; /* has extended perm list (???) */ - unsigned has_explicit_acl: 1; - unsigned has_needea: 1; /* ?? some EA has NEEDEA set + u16 length; /* offset to next dirent */ + +#ifdef __LITTLE_ENDIAN + u8 first: 1; /* set on phony ^A^A (".") entry */ + u8 has_acl: 1; + u8 down: 1; /* down pointer present (after name) */ + u8 last: 1; /* set on phony \377 entry */ + u8 has_ea: 1; /* entry has EA */ + u8 has_xtd_perm: 1; /* has extended perm list (???) */ + u8 has_explicit_acl: 1; + u8 has_needea: 1; /* ?? some EA has NEEDEA set + I have no idea why this is + interesting in a dir entry */ +#else + u8 has_needea: 1; /* ?? some EA has NEEDEA set I have no idea why this is interesting in a dir entry */ - unsigned read_only: 1; /* dos attrib */ - unsigned hidden: 1; /* dos attrib */ - unsigned system: 1; /* dos attrib */ - unsigned flag11: 1; /* would be volume label dos attrib */ - unsigned directory: 1; /* dos attrib */ - unsigned archive: 1; /* dos attrib */ - unsigned not_8x3: 1; /* name is not 8.3 */ - unsigned flag15: 1; + u8 has_explicit_acl: 1; + u8 has_xtd_perm: 1; /* has extended perm list (???) */ + u8 has_ea: 1; /* entry has EA */ + u8 last: 1; /* set on phony \377 entry */ + u8 down: 1; /* down pointer present (after name) */ + u8 has_acl: 1; + u8 first: 1; /* set on phony ^A^A (".") entry */ +#endif + +#ifdef __LITTLE_ENDIAN + u8 read_only: 1; /* dos attrib */ + u8 hidden: 1; /* dos attrib */ + u8 system: 1; /* dos attrib */ + u8 flag11: 1; /* would be volume label dos attrib */ + u8 directory: 1; /* dos attrib */ + u8 archive: 1; /* dos attrib */ + u8 not_8x3: 1; /* name is not 8.3 */ + u8 flag15: 1; +#else + u8 flag15: 1; + u8 not_8x3: 1; /* name is not 8.3 */ + u8 archive: 1; /* dos attrib */ + u8 directory: 1; /* dos attrib */ + u8 flag11: 1; /* would be volume label dos attrib */ + u8 system: 1; /* dos attrib */ + u8 hidden: 1; /* dos attrib */ + u8 read_only: 1; /* dos attrib */ +#endif + fnode_secno fnode; /* fnode giving allocation info */ time32_t write_date; /* mtime */ - unsigned file_size; /* file length, bytes */ + u32 file_size; /* file length, bytes */ time32_t read_date; /* atime */ time32_t creation_date; /* ctime */ - unsigned ea_size; /* total EA length, bytes */ - unsigned char no_of_acls : 3; /* number of ACL's */ - unsigned char reserver : 5; - unsigned char ix; /* code page index (of filename), see + u32 ea_size; /* total EA length, bytes */ + u8 no_of_acls; /* number of ACL's (low 3 bits) */ + u8 ix; /* code page index (of filename), see struct code_page_data */ - unsigned char namelen, name[1]; /* file name */ + u8 namelen, name[1]; /* file name */ /* dnode_secno down; btree down pointer, if present, follows name on next word boundary, or maybe it precedes next dirent, which is on a word boundary. */ @@ -318,38 +375,50 @@ struct hpfs_dirent { struct bplus_leaf_node { - unsigned file_secno; /* first file sector in extent */ - unsigned length; /* length, sectors */ + u32 file_secno; /* first file sector in extent */ + u32 length; /* length, sectors */ secno disk_secno; /* first corresponding disk sector */ }; struct bplus_internal_node { - unsigned file_secno; /* subtree maps sectors < this */ + u32 file_secno; /* subtree maps sectors < this */ anode_secno down; /* pointer to subtree */ }; struct bplus_header { - unsigned hbff: 1; /* high bit of first free entry offset */ - unsigned flag1: 1; - unsigned flag2: 1; - unsigned flag3: 1; - unsigned flag4: 1; - unsigned fnode_parent: 1; /* ? we're pointed to by an fnode, +#ifdef __LITTLE_ENDIAN + u8 hbff: 1; /* high bit of first free entry offset */ + u8 flag1234: 4; + u8 fnode_parent: 1; /* ? we're pointed to by an fnode, the data btree or some ea or the main ea bootage pointer ea_secno */ /* also can get set in fnodes, which may be a chkdsk glitch or may mean this bit is irrelevant in fnodes, or this interpretation is all wet */ - unsigned binary_search: 1; /* suggest binary search (unused) */ - unsigned internal: 1; /* 1 -> (internal) tree of anodes + u8 binary_search: 1; /* suggest binary search (unused) */ + u8 internal: 1; /* 1 -> (internal) tree of anodes + 0 -> (leaf) list of extents */ +#else + u8 internal: 1; /* 1 -> (internal) tree of anodes 0 -> (leaf) list of extents */ - unsigned char fill[3]; - unsigned char n_free_nodes; /* free nodes in following array */ - unsigned char n_used_nodes; /* used nodes in following array */ - unsigned short first_free; /* offset from start of header to + u8 binary_search: 1; /* suggest binary search (unused) */ + u8 fnode_parent: 1; /* ? we're pointed to by an fnode, + the data btree or some ea or the + main ea bootage pointer ea_secno */ + /* also can get set in fnodes, which + may be a chkdsk glitch or may mean + this bit is irrelevant in fnodes, + or this interpretation is all wet */ + u8 flag1234: 4; + u8 hbff: 1; /* high bit of first free entry offset */ +#endif + u8 fill[3]; + u8 n_free_nodes; /* free nodes in following array */ + u8 n_used_nodes; /* used nodes in following array */ + u16 first_free; /* offset from start of header to first free node in array */ union { struct bplus_internal_node internal[0]; /* (internal) 2-word entries giving @@ -369,37 +438,38 @@ struct bplus_header struct fnode { - unsigned magic; /* f7e4 0aae */ - unsigned zero1[2]; /* read history */ - unsigned char len, name[15]; /* true length, truncated name */ + u32 magic; /* f7e4 0aae */ + u32 zero1[2]; /* read history */ + u8 len, name[15]; /* true length, truncated name */ fnode_secno up; /* pointer to file's directory fnode */ - /*unsigned zero2[3];*/ secno acl_size_l; secno acl_secno; - unsigned short acl_size_s; - char acl_anode; - char zero2; /* history bit count */ - unsigned ea_size_l; /* length of disk-resident ea's */ + u16 acl_size_s; + u8 acl_anode; + u8 zero2; /* history bit count */ + u32 ea_size_l; /* length of disk-resident ea's */ secno ea_secno; /* first sector of disk-resident ea's*/ - unsigned short ea_size_s; /* length of fnode-resident ea's */ - - unsigned flag0: 1; - unsigned ea_anode: 1; /* 1 -> ea_secno is an anode */ - unsigned flag2: 1; - unsigned flag3: 1; - unsigned flag4: 1; - unsigned flag5: 1; - unsigned flag6: 1; - unsigned flag7: 1; - unsigned dirflag: 1; /* 1 -> directory. first & only extent + u16 ea_size_s; /* length of fnode-resident ea's */ + +#ifdef __LITTLE_ENDIAN + u8 flag0: 1; + u8 ea_anode: 1; /* 1 -> ea_secno is an anode */ + u8 flag234567: 6; +#else + u8 flag234567: 6; + u8 ea_anode: 1; /* 1 -> ea_secno is an anode */ + u8 flag0: 1; +#endif + +#ifdef __LITTLE_ENDIAN + u8 dirflag: 1; /* 1 -> directory. first & only extent points to dnode. */ - unsigned flag9: 1; - unsigned flag10: 1; - unsigned flag11: 1; - unsigned flag12: 1; - unsigned flag13: 1; - unsigned flag14: 1; - unsigned flag15: 1; + u8 flag9012345: 7; +#else + u8 flag9012345: 7; + u8 dirflag: 1; /* 1 -> directory. first & only extent + points to dnode. */ +#endif struct bplus_header btree; /* b+ tree, 8 extents or 12 subtrees */ union { @@ -407,17 +477,16 @@ struct fnode struct bplus_internal_node internal[12]; } u; - unsigned file_size; /* file length, bytes */ - unsigned n_needea; /* number of EA's with NEEDEA set */ - char user_id[16]; /* unused */ - unsigned short ea_offs; /* offset from start of fnode + u32 file_size; /* file length, bytes */ + u32 n_needea; /* number of EA's with NEEDEA set */ + u8 user_id[16]; /* unused */ + u16 ea_offs; /* offset from start of fnode to first fnode-resident ea */ - char dasd_limit_treshhold; - char dasd_limit_delta; - unsigned dasd_limit; - unsigned dasd_usage; - /*unsigned zero5[2];*/ - unsigned char ea[316]; /* zero or more EA's, packed together + u8 dasd_limit_treshhold; + u8 dasd_limit_delta; + u32 dasd_limit; + u32 dasd_usage; + u8 ea[316]; /* zero or more EA's, packed together with no alignment padding. (Do not use this name, get here via fnode + ea_offs. I think.) */ @@ -430,7 +499,7 @@ struct fnode struct anode { - unsigned magic; /* 37e4 0aae */ + u32 magic; /* 37e4 0aae */ anode_secno self; /* pointer to this anode */ secno up; /* parent anode or fnode */ @@ -440,7 +509,7 @@ struct anode struct bplus_internal_node internal[60]; } u; - unsigned fill[3]; /* unused */ + u32 fill[3]; /* unused */ }; @@ -461,25 +530,31 @@ struct anode struct extended_attribute { - unsigned indirect: 1; /* 1 -> value gives sector number +#ifdef __LITTLE_ENDIAN + u8 indirect: 1; /* 1 -> value gives sector number where real value starts */ - unsigned anode: 1; /* 1 -> sector is an anode + u8 anode: 1; /* 1 -> sector is an anode + that points to fragmented value */ + u8 flag23456: 5; + u8 needea: 1; /* required ea */ +#else + u8 needea: 1; /* required ea */ + u8 flag23456: 5; + u8 anode: 1; /* 1 -> sector is an anode that points to fragmented value */ - unsigned flag2: 1; - unsigned flag3: 1; - unsigned flag4: 1; - unsigned flag5: 1; - unsigned flag6: 1; - unsigned needea: 1; /* required ea */ - unsigned char namelen; /* length of name, bytes */ - unsigned short valuelen; /* length of value, bytes */ - unsigned char name[0]; + u8 indirect: 1; /* 1 -> value gives sector number + where real value starts */ +#endif + u8 namelen; /* length of name, bytes */ + u8 valuelen_lo; /* length of value, bytes */ + u8 valuelen_hi; /* length of value, bytes */ + u8 name[0]; /* - unsigned char name[namelen]; ascii attrib name - unsigned char nul; terminating '\0', not counted - unsigned char value[valuelen]; value, arbitrary + u8 name[namelen]; ascii attrib name + u8 nul; terminating '\0', not counted + u8 value[valuelen]; value, arbitrary if this.indirect, valuelen is 8 and the value is - unsigned length; real length of value, bytes + u32 length; real length of value, bytes secno secno; sector address where it starts if this.anode, the above sector number is the root of an anode tree which points to the value. diff --git a/fs/hpfs/hpfs_fn.h b/fs/hpfs/hpfs_fn.h index c15adbca07ff..dd552f862c8f 100644 --- a/fs/hpfs/hpfs_fn.h +++ b/fs/hpfs/hpfs_fn.h @@ -13,6 +13,7 @@ #include <linux/pagemap.h> #include <linux/buffer_head.h> #include <linux/slab.h> +#include <asm/unaligned.h> #include "hpfs.h" @@ -51,18 +52,16 @@ struct hpfs_inode_info { unsigned i_disk_sec; /* (files) minimalist cache of alloc info */ unsigned i_n_secs; /* (files) minimalist cache of alloc info */ unsigned i_ea_size; /* size of extended attributes */ - unsigned i_conv : 2; /* (files) crlf->newline hackery */ unsigned i_ea_mode : 1; /* file's permission is stored in ea */ unsigned i_ea_uid : 1; /* file's uid is stored in ea */ unsigned i_ea_gid : 1; /* file's gid is stored in ea */ unsigned i_dirty : 1; - struct mutex i_mutex; - struct mutex i_parent_mutex; loff_t **i_rddir_off; struct inode vfs_inode; }; struct hpfs_sb_info { + struct mutex hpfs_mutex; /* global hpfs lock */ ino_t sb_root; /* inode number of root dir */ unsigned sb_fs_size; /* file system size, sectors */ unsigned sb_bitmaps; /* sector number of bitmap list */ @@ -74,7 +73,6 @@ struct hpfs_sb_info { uid_t sb_uid; /* uid from mount options */ gid_t sb_gid; /* gid from mount options */ umode_t sb_mode; /* mode from mount options */ - unsigned sb_conv : 2; /* crlf->newline hackery */ unsigned sb_eas : 2; /* eas: 0-ignore, 1-ro, 2-rw */ unsigned sb_err : 2; /* on errs: 0-cont, 1-ro, 2-panic */ unsigned sb_chk : 2; /* checks: 0-no, 1-normal, 2-strict */ @@ -87,20 +85,9 @@ struct hpfs_sb_info { unsigned *sb_bmp_dir; /* main bitmap directory */ unsigned sb_c_bitmap; /* current bitmap */ unsigned sb_max_fwd_alloc; /* max forwad allocation */ - struct mutex hpfs_creation_de; /* when creating dirents, nobody else - can alloc blocks */ - /*unsigned sb_mounting : 1;*/ int sb_timeshift; }; -/* - * conv= options - */ - -#define CONV_BINARY 0 /* no conversion */ -#define CONV_TEXT 1 /* crlf->newline */ -#define CONV_AUTO 2 /* decide based on file contents */ - /* Four 512-byte buffers and the 2k block obtained by concatenating them */ struct quad_buffer_head { @@ -113,7 +100,7 @@ struct quad_buffer_head { static inline dnode_secno de_down_pointer (struct hpfs_dirent *de) { CHKCOND(de->down,("HPFS: de_down_pointer: !de->down\n")); - return *(dnode_secno *) ((void *) de + de->length - 4); + return le32_to_cpu(*(dnode_secno *) ((void *) de + le16_to_cpu(de->length) - 4)); } /* The first dir entry in a dnode */ @@ -127,41 +114,46 @@ static inline struct hpfs_dirent *dnode_first_de (struct dnode *dnode) static inline struct hpfs_dirent *dnode_end_de (struct dnode *dnode) { - CHKCOND(dnode->first_free>=0x14 && dnode->first_free<=0xa00,("HPFS: dnode_end_de: dnode->first_free = %d\n",(int)dnode->first_free)); - return (void *) dnode + dnode->first_free; + CHKCOND(le32_to_cpu(dnode->first_free)>=0x14 && le32_to_cpu(dnode->first_free)<=0xa00,("HPFS: dnode_end_de: dnode->first_free = %x\n",(unsigned)le32_to_cpu(dnode->first_free))); + return (void *) dnode + le32_to_cpu(dnode->first_free); } /* The dir entry after dir entry de */ static inline struct hpfs_dirent *de_next_de (struct hpfs_dirent *de) { - CHKCOND(de->length>=0x20 && de->length<0x800,("HPFS: de_next_de: de->length = %d\n",(int)de->length)); - return (void *) de + de->length; + CHKCOND(le16_to_cpu(de->length)>=0x20 && le16_to_cpu(de->length)<0x800,("HPFS: de_next_de: de->length = %x\n",(unsigned)le16_to_cpu(de->length))); + return (void *) de + le16_to_cpu(de->length); } static inline struct extended_attribute *fnode_ea(struct fnode *fnode) { - return (struct extended_attribute *)((char *)fnode + fnode->ea_offs + fnode->acl_size_s); + return (struct extended_attribute *)((char *)fnode + le16_to_cpu(fnode->ea_offs) + le16_to_cpu(fnode->acl_size_s)); } static inline struct extended_attribute *fnode_end_ea(struct fnode *fnode) { - return (struct extended_attribute *)((char *)fnode + fnode->ea_offs + fnode->acl_size_s + fnode->ea_size_s); + return (struct extended_attribute *)((char *)fnode + le16_to_cpu(fnode->ea_offs) + le16_to_cpu(fnode->acl_size_s) + le16_to_cpu(fnode->ea_size_s)); +} + +static unsigned ea_valuelen(struct extended_attribute *ea) +{ + return ea->valuelen_lo + 256 * ea->valuelen_hi; } static inline struct extended_attribute *next_ea(struct extended_attribute *ea) { - return (struct extended_attribute *)((char *)ea + 5 + ea->namelen + ea->valuelen); + return (struct extended_attribute *)((char *)ea + 5 + ea->namelen + ea_valuelen(ea)); } static inline secno ea_sec(struct extended_attribute *ea) { - return *(secno *)((char *)ea + 9 + ea->namelen); + return le32_to_cpu(get_unaligned((secno *)((char *)ea + 9 + ea->namelen))); } static inline secno ea_len(struct extended_attribute *ea) { - return *(secno *)((char *)ea + 5 + ea->namelen); + return le32_to_cpu(get_unaligned((secno *)((char *)ea + 5 + ea->namelen))); } static inline char *ea_data(struct extended_attribute *ea) @@ -186,13 +178,13 @@ static inline void copy_de(struct hpfs_dirent *dst, struct hpfs_dirent *src) dst->not_8x3 = n; } -static inline unsigned tstbits(unsigned *bmp, unsigned b, unsigned n) +static inline unsigned tstbits(u32 *bmp, unsigned b, unsigned n) { int i; if ((b >= 0x4000) || (b + n - 1 >= 0x4000)) return n; - if (!((bmp[(b & 0x3fff) >> 5] >> (b & 0x1f)) & 1)) return 1; + if (!((le32_to_cpu(bmp[(b & 0x3fff) >> 5]) >> (b & 0x1f)) & 1)) return 1; for (i = 1; i < n; i++) - if (/*b+i < 0x4000 &&*/ !((bmp[((b+i) & 0x3fff) >> 5] >> ((b+i) & 0x1f)) & 1)) + if (!((le32_to_cpu(bmp[((b+i) & 0x3fff) >> 5]) >> ((b+i) & 0x1f)) & 1)) return i + 1; return 0; } @@ -200,12 +192,12 @@ static inline unsigned tstbits(unsigned *bmp, unsigned b, unsigned n) /* alloc.c */ int hpfs_chk_sectors(struct super_block *, secno, int, char *); -secno hpfs_alloc_sector(struct super_block *, secno, unsigned, int, int); +secno hpfs_alloc_sector(struct super_block *, secno, unsigned, int); int hpfs_alloc_if_possible(struct super_block *, secno); void hpfs_free_sectors(struct super_block *, secno, unsigned); int hpfs_check_free_dnodes(struct super_block *, int); void hpfs_free_dnode(struct super_block *, secno); -struct dnode *hpfs_alloc_dnode(struct super_block *, secno, dnode_secno *, struct quad_buffer_head *, int); +struct dnode *hpfs_alloc_dnode(struct super_block *, secno, dnode_secno *, struct quad_buffer_head *); struct fnode *hpfs_alloc_fnode(struct super_block *, secno, fnode_secno *, struct buffer_head **); struct anode *hpfs_alloc_anode(struct super_block *, secno, anode_secno *, struct buffer_head **); @@ -222,8 +214,6 @@ void hpfs_remove_fnode(struct super_block *, fnode_secno fno); /* buffer.c */ -void hpfs_lock_creation(struct super_block *); -void hpfs_unlock_creation(struct super_block *); void *hpfs_map_sector(struct super_block *, unsigned, struct buffer_head **, int); void *hpfs_get_sector(struct super_block *, unsigned, struct buffer_head **); void *hpfs_map_4sectors(struct super_block *, unsigned, struct quad_buffer_head *, int); @@ -247,7 +237,7 @@ void hpfs_del_pos(struct inode *, loff_t *); struct hpfs_dirent *hpfs_add_de(struct super_block *, struct dnode *, const unsigned char *, unsigned, secno); int hpfs_add_dirent(struct inode *, const unsigned char *, unsigned, - struct hpfs_dirent *, int); + struct hpfs_dirent *); int hpfs_remove_dirent(struct inode *, dnode_secno, struct hpfs_dirent *, struct quad_buffer_head *, int); void hpfs_count_dnodes(struct super_block *, dnode_secno, int *, int *, int *); dnode_secno hpfs_de_as_down_as_possible(struct super_block *, dnode_secno dno); @@ -303,7 +293,6 @@ int hpfs_compare_names(struct super_block *, const unsigned char *, unsigned, const unsigned char *, unsigned, int); int hpfs_is_name_long(const unsigned char *, unsigned); void hpfs_adjust_length(const unsigned char *, unsigned *); -void hpfs_decide_conv(struct inode *, const unsigned char *, unsigned); /* namei.c */ @@ -346,21 +335,26 @@ static inline time32_t gmt_to_local(struct super_block *s, time_t t) /* * Locking: * - * hpfs_lock() is a leftover from the big kernel lock. - * Right now, these functions are empty and only left - * for documentation purposes. The file system no longer - * works on SMP systems, so the lock is not needed - * any more. + * hpfs_lock() locks the whole filesystem. It must be taken + * on any method called by the VFS. * - * If someone is interested in making it work again, this - * would be the place to start by adding a per-superblock - * mutex and fixing all the bugs and performance issues - * caused by that. + * We don't do any per-file locking anymore, it is hard to + * review and HPFS is not performance-sensitive anyway. */ static inline void hpfs_lock(struct super_block *s) { + struct hpfs_sb_info *sbi = hpfs_sb(s); + mutex_lock(&sbi->hpfs_mutex); } static inline void hpfs_unlock(struct super_block *s) { + struct hpfs_sb_info *sbi = hpfs_sb(s); + mutex_unlock(&sbi->hpfs_mutex); +} + +static inline void hpfs_lock_assert(struct super_block *s) +{ + struct hpfs_sb_info *sbi = hpfs_sb(s); + WARN_ON(!mutex_is_locked(&sbi->hpfs_mutex)); } diff --git a/fs/hpfs/inode.c b/fs/hpfs/inode.c index 87f1f787e767..338cd8368451 100644 --- a/fs/hpfs/inode.c +++ b/fs/hpfs/inode.c @@ -17,7 +17,6 @@ void hpfs_init_inode(struct inode *i) i->i_uid = hpfs_sb(sb)->sb_uid; i->i_gid = hpfs_sb(sb)->sb_gid; i->i_mode = hpfs_sb(sb)->sb_mode; - hpfs_inode->i_conv = hpfs_sb(sb)->sb_conv; i->i_size = -1; i->i_blocks = -1; @@ -116,8 +115,8 @@ void hpfs_read_inode(struct inode *i) i->i_mode |= S_IFDIR; i->i_op = &hpfs_dir_iops; i->i_fop = &hpfs_dir_ops; - hpfs_inode->i_parent_dir = fnode->up; - hpfs_inode->i_dno = fnode->u.external[0].disk_secno; + hpfs_inode->i_parent_dir = le32_to_cpu(fnode->up); + hpfs_inode->i_dno = le32_to_cpu(fnode->u.external[0].disk_secno); if (hpfs_sb(sb)->sb_chk >= 2) { struct buffer_head *bh0; if (hpfs_map_fnode(sb, hpfs_inode->i_parent_dir, &bh0)) brelse(bh0); @@ -133,7 +132,7 @@ void hpfs_read_inode(struct inode *i) i->i_op = &hpfs_file_iops; i->i_fop = &hpfs_file_ops; i->i_nlink = 1; - i->i_size = fnode->file_size; + i->i_size = le32_to_cpu(fnode->file_size); i->i_blocks = ((i->i_size + 511) >> 9) + 1; i->i_data.a_ops = &hpfs_aops; hpfs_i(i)->mmu_private = i->i_size; @@ -144,7 +143,7 @@ void hpfs_read_inode(struct inode *i) static void hpfs_write_inode_ea(struct inode *i, struct fnode *fnode) { struct hpfs_inode_info *hpfs_inode = hpfs_i(i); - /*if (fnode->acl_size_l || fnode->acl_size_s) { + /*if (le32_to_cpu(fnode->acl_size_l) || le16_to_cpu(fnode->acl_size_s)) { Some unknown structures like ACL may be in fnode, we'd better not overwrite them hpfs_error(i->i_sb, "fnode %08x has some unknown HPFS386 stuctures", i->i_ino); @@ -187,9 +186,7 @@ void hpfs_write_inode(struct inode *i) kfree(hpfs_inode->i_rddir_off); hpfs_inode->i_rddir_off = NULL; } - mutex_lock(&hpfs_inode->i_parent_mutex); if (!i->i_nlink) { - mutex_unlock(&hpfs_inode->i_parent_mutex); return; } parent = iget_locked(i->i_sb, hpfs_inode->i_parent_dir); @@ -200,14 +197,9 @@ void hpfs_write_inode(struct inode *i) hpfs_read_inode(parent); unlock_new_inode(parent); } - mutex_lock(&hpfs_inode->i_mutex); hpfs_write_inode_nolock(i); - mutex_unlock(&hpfs_inode->i_mutex); iput(parent); - } else { - mark_inode_dirty(i); } - mutex_unlock(&hpfs_inode->i_parent_mutex); } void hpfs_write_inode_nolock(struct inode *i) @@ -226,30 +218,30 @@ void hpfs_write_inode_nolock(struct inode *i) } } else de = NULL; if (S_ISREG(i->i_mode)) { - fnode->file_size = i->i_size; - if (de) de->file_size = i->i_size; + fnode->file_size = cpu_to_le32(i->i_size); + if (de) de->file_size = cpu_to_le32(i->i_size); } else if (S_ISDIR(i->i_mode)) { - fnode->file_size = 0; - if (de) de->file_size = 0; + fnode->file_size = cpu_to_le32(0); + if (de) de->file_size = cpu_to_le32(0); } hpfs_write_inode_ea(i, fnode); if (de) { - de->write_date = gmt_to_local(i->i_sb, i->i_mtime.tv_sec); - de->read_date = gmt_to_local(i->i_sb, i->i_atime.tv_sec); - de->creation_date = gmt_to_local(i->i_sb, i->i_ctime.tv_sec); + de->write_date = cpu_to_le32(gmt_to_local(i->i_sb, i->i_mtime.tv_sec)); + de->read_date = cpu_to_le32(gmt_to_local(i->i_sb, i->i_atime.tv_sec)); + de->creation_date = cpu_to_le32(gmt_to_local(i->i_sb, i->i_ctime.tv_sec)); de->read_only = !(i->i_mode & 0222); - de->ea_size = hpfs_inode->i_ea_size; + de->ea_size = cpu_to_le32(hpfs_inode->i_ea_size); hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); } if (S_ISDIR(i->i_mode)) { if ((de = map_dirent(i, hpfs_inode->i_dno, "\001\001", 2, NULL, &qbh))) { - de->write_date = gmt_to_local(i->i_sb, i->i_mtime.tv_sec); - de->read_date = gmt_to_local(i->i_sb, i->i_atime.tv_sec); - de->creation_date = gmt_to_local(i->i_sb, i->i_ctime.tv_sec); + de->write_date = cpu_to_le32(gmt_to_local(i->i_sb, i->i_mtime.tv_sec)); + de->read_date = cpu_to_le32(gmt_to_local(i->i_sb, i->i_atime.tv_sec)); + de->creation_date = cpu_to_le32(gmt_to_local(i->i_sb, i->i_ctime.tv_sec)); de->read_only = !(i->i_mode & 0222); - de->ea_size = /*hpfs_inode->i_ea_size*/0; - de->file_size = 0; + de->ea_size = cpu_to_le32(/*hpfs_inode->i_ea_size*/0); + de->file_size = cpu_to_le32(0); hpfs_mark_4buffers_dirty(&qbh); hpfs_brelse4(&qbh); } else @@ -269,6 +261,10 @@ int hpfs_setattr(struct dentry *dentry, struct iattr *attr) hpfs_lock(inode->i_sb); if (inode->i_ino == hpfs_sb(inode->i_sb)->sb_root) goto out_unlock; + if ((attr->ia_valid & ATTR_UID) && attr->ia_uid >= 0x10000) + goto out_unlock; + if ((attr->ia_valid & ATTR_GID) && attr->ia_gid >= 0x10000) + goto out_unlock; if ((attr->ia_valid & ATTR_SIZE) && attr->ia_size > inode->i_size) goto out_unlock; @@ -284,7 +280,6 @@ int hpfs_setattr(struct dentry *dentry, struct iattr *attr) } setattr_copy(inode, attr); - mark_inode_dirty(inode); hpfs_write_inode(inode); diff --git a/fs/hpfs/map.c b/fs/hpfs/map.c index 840d033ecee8..a790821366a7 100644 --- a/fs/hpfs/map.c +++ b/fs/hpfs/map.c @@ -21,7 +21,7 @@ unsigned int *hpfs_map_bitmap(struct super_block *s, unsigned bmp_block, hpfs_error(s, "hpfs_map_bitmap called with bad parameter: %08x at %s", bmp_block, id); return NULL; } - sec = hpfs_sb(s)->sb_bmp_dir[bmp_block]; + sec = le32_to_cpu(hpfs_sb(s)->sb_bmp_dir[bmp_block]); if (!sec || sec > hpfs_sb(s)->sb_fs_size-4) { hpfs_error(s, "invalid bitmap block pointer %08x -> %08x at %s", bmp_block, sec, id); return NULL; @@ -46,18 +46,18 @@ unsigned char *hpfs_load_code_page(struct super_block *s, secno cps) struct code_page_data *cpd; struct code_page_directory *cp = hpfs_map_sector(s, cps, &bh, 0); if (!cp) return NULL; - if (cp->magic != CP_DIR_MAGIC) { - printk("HPFS: Code page directory magic doesn't match (magic = %08x)\n", cp->magic); + if (le32_to_cpu(cp->magic) != CP_DIR_MAGIC) { + printk("HPFS: Code page directory magic doesn't match (magic = %08x)\n", le32_to_cpu(cp->magic)); brelse(bh); return NULL; } - if (!cp->n_code_pages) { + if (!le32_to_cpu(cp->n_code_pages)) { printk("HPFS: n_code_pages == 0\n"); brelse(bh); return NULL; } - cpds = cp->array[0].code_page_data; - cpi = cp->array[0].index; + cpds = le32_to_cpu(cp->array[0].code_page_data); + cpi = le16_to_cpu(cp->array[0].index); brelse(bh); if (cpi >= 3) { @@ -66,12 +66,12 @@ unsigned char *hpfs_load_code_page(struct super_block *s, secno cps) } if (!(cpd = hpfs_map_sector(s, cpds, &bh, 0))) return NULL; - if ((unsigned)cpd->offs[cpi] > 0x178) { + if (le16_to_cpu(cpd->offs[cpi]) > 0x178) { printk("HPFS: Code page index out of sector\n"); brelse(bh); return NULL; } - ptr = (unsigned char *)cpd + cpd->offs[cpi] + 6; + ptr = (unsigned char *)cpd + le16_to_cpu(cpd->offs[cpi]) + 6; if (!(cp_table = kmalloc(256, GFP_KERNEL))) { printk("HPFS: out of memory for code page table\n"); brelse(bh); @@ -125,7 +125,7 @@ struct fnode *hpfs_map_fnode(struct super_block *s, ino_t ino, struct buffer_hea if (hpfs_sb(s)->sb_chk) { struct extended_attribute *ea; struct extended_attribute *ea_end; - if (fnode->magic != FNODE_MAGIC) { + if (le32_to_cpu(fnode->magic) != FNODE_MAGIC) { hpfs_error(s, "bad magic on fnode %08lx", (unsigned long)ino); goto bail; @@ -138,7 +138,7 @@ struct fnode *hpfs_map_fnode(struct super_block *s, ino_t ino, struct buffer_hea (unsigned long)ino); goto bail; } - if (fnode->btree.first_free != + if (le16_to_cpu(fnode->btree.first_free) != 8 + fnode->btree.n_used_nodes * (fnode->btree.internal ? 8 : 12)) { hpfs_error(s, "bad first_free pointer in fnode %08lx", @@ -146,12 +146,12 @@ struct fnode *hpfs_map_fnode(struct super_block *s, ino_t ino, struct buffer_hea goto bail; } } - if (fnode->ea_size_s && ((signed int)fnode->ea_offs < 0xc4 || - (signed int)fnode->ea_offs + fnode->acl_size_s + fnode->ea_size_s > 0x200)) { + if (le16_to_cpu(fnode->ea_size_s) && (le16_to_cpu(fnode->ea_offs) < 0xc4 || + le16_to_cpu(fnode->ea_offs) + le16_to_cpu(fnode->acl_size_s) + le16_to_cpu(fnode->ea_size_s) > 0x200)) { hpfs_error(s, "bad EA info in fnode %08lx: ea_offs == %04x ea_size_s == %04x", (unsigned long)ino, - fnode->ea_offs, fnode->ea_size_s); + le16_to_cpu(fnode->ea_offs), le16_to_cpu(fnode->ea_size_s)); goto bail; } ea = fnode_ea(fnode); @@ -178,16 +178,20 @@ struct anode *hpfs_map_anode(struct super_block *s, anode_secno ano, struct buff if (hpfs_sb(s)->sb_chk) if (hpfs_chk_sectors(s, ano, 1, "anode")) return NULL; if ((anode = hpfs_map_sector(s, ano, bhp, ANODE_RD_AHEAD))) if (hpfs_sb(s)->sb_chk) { - if (anode->magic != ANODE_MAGIC || anode->self != ano) { + if (le32_to_cpu(anode->magic) != ANODE_MAGIC) { hpfs_error(s, "bad magic on anode %08x", ano); goto bail; } + if (le32_to_cpu(anode->self) != ano) { + hpfs_error(s, "self pointer invalid on anode %08x", ano); + goto bail; + } if ((unsigned)anode->btree.n_used_nodes + (unsigned)anode->btree.n_free_nodes != (anode->btree.internal ? 60 : 40)) { hpfs_error(s, "bad number of nodes in anode %08x", ano); goto bail; } - if (anode->btree.first_free != + if (le16_to_cpu(anode->btree.first_free) != 8 + anode->btree.n_used_nodes * (anode->btree.internal ? 8 : 12)) { hpfs_error(s, "bad first_free pointer in anode %08x", ano); goto bail; @@ -219,26 +223,26 @@ struct dnode *hpfs_map_dnode(struct super_block *s, unsigned secno, unsigned p, pp = 0; unsigned char *d = (unsigned char *)dnode; int b = 0; - if (dnode->magic != DNODE_MAGIC) { + if (le32_to_cpu(dnode->magic) != DNODE_MAGIC) { hpfs_error(s, "bad magic on dnode %08x", secno); goto bail; } - if (dnode->self != secno) - hpfs_error(s, "bad self pointer on dnode %08x self = %08x", secno, dnode->self); + if (le32_to_cpu(dnode->self) != secno) + hpfs_error(s, "bad self pointer on dnode %08x self = %08x", secno, le32_to_cpu(dnode->self)); /* Check dirents - bad dirents would cause infinite loops or shooting to memory */ - if (dnode->first_free > 2048/* || dnode->first_free < 84*/) { - hpfs_error(s, "dnode %08x has first_free == %08x", secno, dnode->first_free); + if (le32_to_cpu(dnode->first_free) > 2048) { + hpfs_error(s, "dnode %08x has first_free == %08x", secno, le32_to_cpu(dnode->first_free)); goto bail; } - for (p = 20; p < dnode->first_free; p += d[p] + (d[p+1] << 8)) { + for (p = 20; p < le32_to_cpu(dnode->first_free); p += d[p] + (d[p+1] << 8)) { struct hpfs_dirent *de = (struct hpfs_dirent *)((char *)dnode + p); - if (de->length > 292 || (de->length < 32) || (de->length & 3) || p + de->length > 2048) { + if (le16_to_cpu(de->length) > 292 || (le16_to_cpu(de->length) < 32) || (le16_to_cpu(de->length) & 3) || p + le16_to_cpu(de->length) > 2048) { hpfs_error(s, "bad dirent size in dnode %08x, dirent %03x, last %03x", secno, p, pp); goto bail; } - if (((31 + de->namelen + de->down*4 + 3) & ~3) != de->length) { - if (((31 + de->namelen + de->down*4 + 3) & ~3) < de->length && s->s_flags & MS_RDONLY) goto ok; + if (((31 + de->namelen + de->down*4 + 3) & ~3) != le16_to_cpu(de->length)) { + if (((31 + de->namelen + de->down*4 + 3) & ~3) < le16_to_cpu(de->length) && s->s_flags & MS_RDONLY) goto ok; hpfs_error(s, "namelen does not match dirent size in dnode %08x, dirent %03x, last %03x", secno, p, pp); goto bail; } @@ -251,7 +255,7 @@ struct dnode *hpfs_map_dnode(struct super_block *s, unsigned secno, pp = p; } - if (p != dnode->first_free) { + if (p != le32_to_cpu(dnode->first_free)) { hpfs_error(s, "size on last dirent does not match first_free; dnode %08x", secno); goto bail; } @@ -277,7 +281,7 @@ dnode_secno hpfs_fnode_dno(struct super_block *s, ino_t ino) if (!fnode) return 0; - dno = fnode->u.external[0].disk_secno; + dno = le32_to_cpu(fnode->u.external[0].disk_secno); brelse(bh); return dno; } diff --git a/fs/hpfs/name.c b/fs/hpfs/name.c index f24736d7a439..9acdf338def0 100644 --- a/fs/hpfs/name.c +++ b/fs/hpfs/name.c @@ -8,39 +8,6 @@ #include "hpfs_fn.h" -static const char *text_postfix[]={ -".ASM", ".BAS", ".BAT", ".C", ".CC", ".CFG", ".CMD", ".CON", ".CPP", ".DEF", -".DOC", ".DPR", ".ERX", ".H", ".HPP", ".HTM", ".HTML", ".JAVA", ".LOG", ".PAS", -".RC", ".TEX", ".TXT", ".Y", ""}; - -static const char *text_prefix[]={ -"AUTOEXEC.", "CHANGES", "COPYING", "CONFIG.", "CREDITS", "FAQ", "FILE_ID.DIZ", -"MAKEFILE", "READ.ME", "README", "TERMCAP", ""}; - -void hpfs_decide_conv(struct inode *inode, const unsigned char *name, unsigned len) -{ - struct hpfs_inode_info *hpfs_inode = hpfs_i(inode); - int i; - if (hpfs_inode->i_conv != CONV_AUTO) return; - for (i = 0; *text_postfix[i]; i++) { - int l = strlen(text_postfix[i]); - if (l <= len) - if (!hpfs_compare_names(inode->i_sb, text_postfix[i], l, name + len - l, l, 0)) - goto text; - } - for (i = 0; *text_prefix[i]; i++) { - int l = strlen(text_prefix[i]); - if (l <= len) - if (!hpfs_compare_names(inode->i_sb, text_prefix[i], l, name, l, 0)) - goto text; - } - hpfs_inode->i_conv = CONV_BINARY; - return; - text: - hpfs_inode->i_conv = CONV_TEXT; - return; -} - static inline int not_allowed_char(unsigned char c) { return c<' ' || c=='"' || c=='*' || c=='/' || c==':' || c=='<' || diff --git a/fs/hpfs/namei.c b/fs/hpfs/namei.c index d5f8c8a19023..acf95dab2aac 100644 --- a/fs/hpfs/namei.c +++ b/fs/hpfs/namei.c @@ -29,7 +29,7 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) fnode = hpfs_alloc_fnode(dir->i_sb, hpfs_i(dir)->i_dno, &fno, &bh); if (!fnode) goto bail; - dnode = hpfs_alloc_dnode(dir->i_sb, fno, &dno, &qbh0, 1); + dnode = hpfs_alloc_dnode(dir->i_sb, fno, &dno, &qbh0); if (!dnode) goto bail1; memset(&dee, 0, sizeof dee); @@ -37,8 +37,8 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) if (!(mode & 0222)) dee.read_only = 1; /*dee.archive = 0;*/ dee.hidden = name[0] == '.'; - dee.fnode = fno; - dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds()); + dee.fnode = cpu_to_le32(fno); + dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds())); result = new_inode(dir->i_sb); if (!result) goto bail2; @@ -46,7 +46,7 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) result->i_ino = fno; hpfs_i(result)->i_parent_dir = dir->i_ino; hpfs_i(result)->i_dno = dno; - result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date); + result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date)); result->i_ctime.tv_nsec = 0; result->i_mtime.tv_nsec = 0; result->i_atime.tv_nsec = 0; @@ -60,8 +60,7 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) if (dee.read_only) result->i_mode &= ~0222; - mutex_lock(&hpfs_i(dir)->i_mutex); - r = hpfs_add_dirent(dir, name, len, &dee, 0); + r = hpfs_add_dirent(dir, name, len, &dee); if (r == 1) goto bail3; if (r == -1) { @@ -70,21 +69,21 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) } fnode->len = len; memcpy(fnode->name, name, len > 15 ? 15 : len); - fnode->up = dir->i_ino; + fnode->up = cpu_to_le32(dir->i_ino); fnode->dirflag = 1; fnode->btree.n_free_nodes = 7; fnode->btree.n_used_nodes = 1; - fnode->btree.first_free = 0x14; - fnode->u.external[0].disk_secno = dno; - fnode->u.external[0].file_secno = -1; + fnode->btree.first_free = cpu_to_le16(0x14); + fnode->u.external[0].disk_secno = cpu_to_le32(dno); + fnode->u.external[0].file_secno = cpu_to_le32(-1); dnode->root_dnode = 1; - dnode->up = fno; + dnode->up = cpu_to_le32(fno); de = hpfs_add_de(dir->i_sb, dnode, "\001\001", 2, 0); - de->creation_date = de->write_date = de->read_date = gmt_to_local(dir->i_sb, get_seconds()); + de->creation_date = de->write_date = de->read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds())); if (!(mode & 0222)) de->read_only = 1; de->first = de->directory = 1; /*de->hidden = de->system = 0;*/ - de->fnode = fno; + de->fnode = cpu_to_le32(fno); mark_buffer_dirty(bh); brelse(bh); hpfs_mark_4buffers_dirty(&qbh0); @@ -101,11 +100,9 @@ static int hpfs_mkdir(struct inode *dir, struct dentry *dentry, int mode) hpfs_write_inode_nolock(result); } d_instantiate(dentry, result); - mutex_unlock(&hpfs_i(dir)->i_mutex); hpfs_unlock(dir->i_sb); return 0; bail3: - mutex_unlock(&hpfs_i(dir)->i_mutex); iput(result); bail2: hpfs_brelse4(&qbh0); @@ -140,8 +137,8 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc if (!(mode & 0222)) dee.read_only = 1; dee.archive = 1; dee.hidden = name[0] == '.'; - dee.fnode = fno; - dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds()); + dee.fnode = cpu_to_le32(fno); + dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds())); result = new_inode(dir->i_sb); if (!result) @@ -154,9 +151,8 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc result->i_op = &hpfs_file_iops; result->i_fop = &hpfs_file_ops; result->i_nlink = 1; - hpfs_decide_conv(result, name, len); hpfs_i(result)->i_parent_dir = dir->i_ino; - result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date); + result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date)); result->i_ctime.tv_nsec = 0; result->i_mtime.tv_nsec = 0; result->i_atime.tv_nsec = 0; @@ -168,8 +164,7 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc result->i_data.a_ops = &hpfs_aops; hpfs_i(result)->mmu_private = 0; - mutex_lock(&hpfs_i(dir)->i_mutex); - r = hpfs_add_dirent(dir, name, len, &dee, 0); + r = hpfs_add_dirent(dir, name, len, &dee); if (r == 1) goto bail2; if (r == -1) { @@ -178,7 +173,7 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc } fnode->len = len; memcpy(fnode->name, name, len > 15 ? 15 : len); - fnode->up = dir->i_ino; + fnode->up = cpu_to_le32(dir->i_ino); mark_buffer_dirty(bh); brelse(bh); @@ -193,12 +188,10 @@ static int hpfs_create(struct inode *dir, struct dentry *dentry, int mode, struc hpfs_write_inode_nolock(result); } d_instantiate(dentry, result); - mutex_unlock(&hpfs_i(dir)->i_mutex); hpfs_unlock(dir->i_sb); return 0; bail2: - mutex_unlock(&hpfs_i(dir)->i_mutex); iput(result); bail1: brelse(bh); @@ -232,8 +225,8 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t if (!(mode & 0222)) dee.read_only = 1; dee.archive = 1; dee.hidden = name[0] == '.'; - dee.fnode = fno; - dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds()); + dee.fnode = cpu_to_le32(fno); + dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds())); result = new_inode(dir->i_sb); if (!result) @@ -242,7 +235,7 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t hpfs_init_inode(result); result->i_ino = fno; hpfs_i(result)->i_parent_dir = dir->i_ino; - result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date); + result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date)); result->i_ctime.tv_nsec = 0; result->i_mtime.tv_nsec = 0; result->i_atime.tv_nsec = 0; @@ -254,8 +247,7 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t result->i_blocks = 1; init_special_inode(result, mode, rdev); - mutex_lock(&hpfs_i(dir)->i_mutex); - r = hpfs_add_dirent(dir, name, len, &dee, 0); + r = hpfs_add_dirent(dir, name, len, &dee); if (r == 1) goto bail2; if (r == -1) { @@ -264,19 +256,17 @@ static int hpfs_mknod(struct inode *dir, struct dentry *dentry, int mode, dev_t } fnode->len = len; memcpy(fnode->name, name, len > 15 ? 15 : len); - fnode->up = dir->i_ino; + fnode->up = cpu_to_le32(dir->i_ino); mark_buffer_dirty(bh); insert_inode_hash(result); hpfs_write_inode_nolock(result); d_instantiate(dentry, result); - mutex_unlock(&hpfs_i(dir)->i_mutex); brelse(bh); hpfs_unlock(dir->i_sb); return 0; bail2: - mutex_unlock(&hpfs_i(dir)->i_mutex); iput(result); bail1: brelse(bh); @@ -310,8 +300,8 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy memset(&dee, 0, sizeof dee); dee.archive = 1; dee.hidden = name[0] == '.'; - dee.fnode = fno; - dee.creation_date = dee.write_date = dee.read_date = gmt_to_local(dir->i_sb, get_seconds()); + dee.fnode = cpu_to_le32(fno); + dee.creation_date = dee.write_date = dee.read_date = cpu_to_le32(gmt_to_local(dir->i_sb, get_seconds())); result = new_inode(dir->i_sb); if (!result) @@ -319,7 +309,7 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy result->i_ino = fno; hpfs_init_inode(result); hpfs_i(result)->i_parent_dir = dir->i_ino; - result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, dee.creation_date); + result->i_ctime.tv_sec = result->i_mtime.tv_sec = result->i_atime.tv_sec = local_to_gmt(dir->i_sb, le32_to_cpu(dee.creation_date)); result->i_ctime.tv_nsec = 0; result->i_mtime.tv_nsec = 0; result->i_atime.tv_nsec = 0; @@ -333,8 +323,7 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy result->i_op = &page_symlink_inode_operations; result->i_data.a_ops = &hpfs_symlink_aops; - mutex_lock(&hpfs_i(dir)->i_mutex); - r = hpfs_add_dirent(dir, name, len, &dee, 0); + r = hpfs_add_dirent(dir, name, len, &dee); if (r == 1) goto bail2; if (r == -1) { @@ -343,7 +332,7 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy } fnode->len = len; memcpy(fnode->name, name, len > 15 ? 15 : len); - fnode->up = dir->i_ino; + fnode->up = cpu_to_le32(dir->i_ino); hpfs_set_ea(result, fnode, "SYMLINK", symlink, strlen(symlink)); mark_buffer_dirty(bh); brelse(bh); @@ -352,11 +341,9 @@ static int hpfs_symlink(struct inode *dir, struct dentry *dentry, const char *sy hpfs_write_inode_nolock(result); d_instantiate(dentry, result); - mutex_unlock(&hpfs_i(dir)->i_mutex); hpfs_unlock(dir->i_sb); return 0; bail2: - mutex_unlock(&hpfs_i(dir)->i_mutex); iput(result); bail1: brelse(bh); @@ -374,7 +361,6 @@ static int hpfs_unlink(struct inode *dir, struct dentry *dentry) struct hpfs_dirent *de; struct inode *inode = dentry->d_inode; dnode_secno dno; - fnode_secno fno; int r; int rep = 0; int err; @@ -382,8 +368,6 @@ static int hpfs_unlink(struct inode *dir, struct dentry *dentry) hpfs_lock(dir->i_sb); hpfs_adjust_length(name, &len); again: - mutex_lock(&hpfs_i(inode)->i_parent_mutex); - mutex_lock(&hpfs_i(dir)->i_mutex); err = -ENOENT; de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, &dno, &qbh); if (!de) @@ -397,7 +381,6 @@ again: if (de->directory) goto out1; - fno = de->fnode; r = hpfs_remove_dirent(dir, dno, de, &qbh, 1); switch (r) { case 1: @@ -410,11 +393,8 @@ again: if (rep++) break; - mutex_unlock(&hpfs_i(dir)->i_mutex); - mutex_unlock(&hpfs_i(inode)->i_parent_mutex); dentry_unhash(dentry); if (!d_unhashed(dentry)) { - dput(dentry); hpfs_unlock(dir->i_sb); return -ENOSPC; } @@ -422,7 +402,6 @@ again: !S_ISREG(inode->i_mode) || get_write_access(inode)) { d_rehash(dentry); - dput(dentry); } else { struct iattr newattrs; /*printk("HPFS: truncating file before delete.\n");*/ @@ -430,7 +409,6 @@ again: newattrs.ia_valid = ATTR_SIZE | ATTR_CTIME; err = notify_change(dentry, &newattrs); put_write_access(inode); - dput(dentry); if (!err) goto again; } @@ -445,8 +423,6 @@ again: out1: hpfs_brelse4(&qbh); out: - mutex_unlock(&hpfs_i(dir)->i_mutex); - mutex_unlock(&hpfs_i(inode)->i_parent_mutex); hpfs_unlock(dir->i_sb); return err; } @@ -459,15 +435,12 @@ static int hpfs_rmdir(struct inode *dir, struct dentry *dentry) struct hpfs_dirent *de; struct inode *inode = dentry->d_inode; dnode_secno dno; - fnode_secno fno; int n_items = 0; int err; int r; hpfs_adjust_length(name, &len); hpfs_lock(dir->i_sb); - mutex_lock(&hpfs_i(inode)->i_parent_mutex); - mutex_lock(&hpfs_i(dir)->i_mutex); err = -ENOENT; de = map_dirent(dir, hpfs_i(dir)->i_dno, name, len, &dno, &qbh); if (!de) @@ -486,7 +459,6 @@ static int hpfs_rmdir(struct inode *dir, struct dentry *dentry) if (n_items) goto out1; - fno = de->fnode; r = hpfs_remove_dirent(dir, dno, de, &qbh, 1); switch (r) { case 1: @@ -505,8 +477,6 @@ static int hpfs_rmdir(struct inode *dir, struct dentry *dentry) out1: hpfs_brelse4(&qbh); out: - mutex_unlock(&hpfs_i(dir)->i_mutex); - mutex_unlock(&hpfs_i(inode)->i_parent_mutex); hpfs_unlock(dir->i_sb); return err; } @@ -562,18 +532,13 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry, struct buffer_head *bh; struct fnode *fnode; int err; + if ((err = hpfs_chk_name(new_name, &new_len))) return err; err = 0; hpfs_adjust_length(old_name, &old_len); hpfs_lock(i->i_sb); /* order doesn't matter, due to VFS exclusion */ - mutex_lock(&hpfs_i(i)->i_parent_mutex); - if (new_inode) - mutex_lock(&hpfs_i(new_inode)->i_parent_mutex); - mutex_lock(&hpfs_i(old_dir)->i_mutex); - if (new_dir != old_dir) - mutex_lock(&hpfs_i(new_dir)->i_mutex); /* Erm? Moving over the empty non-busy directory is perfectly legal */ if (new_inode && S_ISDIR(new_inode->i_mode)) { @@ -610,9 +575,7 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (new_dir == old_dir) hpfs_brelse4(&qbh); - hpfs_lock_creation(i->i_sb); - if ((r = hpfs_add_dirent(new_dir, new_name, new_len, &de, 1))) { - hpfs_unlock_creation(i->i_sb); + if ((r = hpfs_add_dirent(new_dir, new_name, new_len, &de))) { if (r == -1) hpfs_error(new_dir->i_sb, "hpfs_rename: dirent already exists!"); err = r == 1 ? -ENOSPC : -EFSERROR; if (new_dir != old_dir) hpfs_brelse4(&qbh); @@ -621,20 +584,17 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry, if (new_dir == old_dir) if (!(dep = map_dirent(old_dir, hpfs_i(old_dir)->i_dno, old_name, old_len, &dno, &qbh))) { - hpfs_unlock_creation(i->i_sb); hpfs_error(i->i_sb, "lookup succeeded but map dirent failed at #2"); err = -ENOENT; goto end1; } if ((r = hpfs_remove_dirent(old_dir, dno, dep, &qbh, 0))) { - hpfs_unlock_creation(i->i_sb); hpfs_error(i->i_sb, "hpfs_rename: could not remove dirent"); err = r == 2 ? -ENOSPC : -EFSERROR; goto end1; } - hpfs_unlock_creation(i->i_sb); - + end: hpfs_i(i)->i_parent_dir = new_dir->i_ino; if (S_ISDIR(i->i_mode)) { @@ -642,22 +602,14 @@ static int hpfs_rename(struct inode *old_dir, struct dentry *old_dentry, drop_nlink(old_dir); } if ((fnode = hpfs_map_fnode(i->i_sb, i->i_ino, &bh))) { - fnode->up = new_dir->i_ino; + fnode->up = cpu_to_le32(new_dir->i_ino); fnode->len = new_len; memcpy(fnode->name, new_name, new_len>15?15:new_len); if (new_len < 15) memset(&fnode->name[new_len], 0, 15 - new_len); mark_buffer_dirty(bh); brelse(bh); } - hpfs_i(i)->i_conv = hpfs_sb(i->i_sb)->sb_conv; - hpfs_decide_conv(i, new_name, new_len); end1: - if (old_dir != new_dir) - mutex_unlock(&hpfs_i(new_dir)->i_mutex); - mutex_unlock(&hpfs_i(old_dir)->i_mutex); - mutex_unlock(&hpfs_i(i)->i_parent_mutex); - if (new_inode) - mutex_unlock(&hpfs_i(new_inode)->i_parent_mutex); hpfs_unlock(i->i_sb); return err; } diff --git a/fs/hpfs/super.c b/fs/hpfs/super.c index c89b40808587..98580a3b5005 100644 --- a/fs/hpfs/super.c +++ b/fs/hpfs/super.c @@ -18,15 +18,16 @@ /* Mark the filesystem dirty, so that chkdsk checks it when os/2 booted */ -static void mark_dirty(struct super_block *s) +static void mark_dirty(struct super_block *s, int remount) { - if (hpfs_sb(s)->sb_chkdsk && !(s->s_flags & MS_RDONLY)) { + if (hpfs_sb(s)->sb_chkdsk && (remount || !(s->s_flags & MS_RDONLY))) { struct buffer_head *bh; struct hpfs_spare_block *sb; if ((sb = hpfs_map_sector(s, 17, &bh, 0))) { sb->dirty = 1; sb->old_wrote = 0; mark_buffer_dirty(bh); + sync_dirty_buffer(bh); brelse(bh); } } @@ -40,10 +41,12 @@ static void unmark_dirty(struct super_block *s) struct buffer_head *bh; struct hpfs_spare_block *sb; if (s->s_flags & MS_RDONLY) return; + sync_blockdev(s->s_bdev); if ((sb = hpfs_map_sector(s, 17, &bh, 0))) { sb->dirty = hpfs_sb(s)->sb_chkdsk > 1 - hpfs_sb(s)->sb_was_error; sb->old_wrote = hpfs_sb(s)->sb_chkdsk >= 2 && !hpfs_sb(s)->sb_was_error; mark_buffer_dirty(bh); + sync_dirty_buffer(bh); brelse(bh); } } @@ -63,13 +66,13 @@ void hpfs_error(struct super_block *s, const char *fmt, ...) if (!hpfs_sb(s)->sb_was_error) { if (hpfs_sb(s)->sb_err == 2) { printk("; crashing the system because you wanted it\n"); - mark_dirty(s); + mark_dirty(s, 0); panic("HPFS panic"); } else if (hpfs_sb(s)->sb_err == 1) { if (s->s_flags & MS_RDONLY) printk("; already mounted read-only\n"); else { printk("; remounting read-only\n"); - mark_dirty(s); + mark_dirty(s, 0); s->s_flags |= MS_RDONLY; } } else if (s->s_flags & MS_RDONLY) printk("; going on - but anything won't be destroyed because it's read-only\n"); @@ -102,9 +105,12 @@ static void hpfs_put_super(struct super_block *s) { struct hpfs_sb_info *sbi = hpfs_sb(s); + hpfs_lock(s); + unmark_dirty(s); + hpfs_unlock(s); + kfree(sbi->sb_cp_table); kfree(sbi->sb_bmp_dir); - unmark_dirty(s); s->s_fs_info = NULL; kfree(sbi); } @@ -129,7 +135,7 @@ static unsigned count_bitmaps(struct super_block *s) n_bands = (hpfs_sb(s)->sb_fs_size + 0x3fff) >> 14; count = 0; for (n = 0; n < n_bands; n++) - count += hpfs_count_one_bitmap(s, hpfs_sb(s)->sb_bmp_dir[n]); + count += hpfs_count_one_bitmap(s, le32_to_cpu(hpfs_sb(s)->sb_bmp_dir[n])); return count; } @@ -188,8 +194,6 @@ static void init_once(void *foo) { struct hpfs_inode_info *ei = (struct hpfs_inode_info *) foo; - mutex_init(&ei->i_mutex); - mutex_init(&ei->i_parent_mutex); inode_init_once(&ei->vfs_inode); } @@ -218,7 +222,6 @@ static void destroy_inodecache(void) enum { Opt_help, Opt_uid, Opt_gid, Opt_umask, Opt_case_lower, Opt_case_asis, - Opt_conv_binary, Opt_conv_text, Opt_conv_auto, Opt_check_none, Opt_check_normal, Opt_check_strict, Opt_err_cont, Opt_err_ro, Opt_err_panic, Opt_eas_no, Opt_eas_ro, Opt_eas_rw, @@ -233,9 +236,6 @@ static const match_table_t tokens = { {Opt_umask, "umask=%o"}, {Opt_case_lower, "case=lower"}, {Opt_case_asis, "case=asis"}, - {Opt_conv_binary, "conv=binary"}, - {Opt_conv_text, "conv=text"}, - {Opt_conv_auto, "conv=auto"}, {Opt_check_none, "check=none"}, {Opt_check_normal, "check=normal"}, {Opt_check_strict, "check=strict"}, @@ -253,7 +253,7 @@ static const match_table_t tokens = { }; static int parse_opts(char *opts, uid_t *uid, gid_t *gid, umode_t *umask, - int *lowercase, int *conv, int *eas, int *chk, int *errs, + int *lowercase, int *eas, int *chk, int *errs, int *chkdsk, int *timeshift) { char *p; @@ -295,15 +295,6 @@ static int parse_opts(char *opts, uid_t *uid, gid_t *gid, umode_t *umask, case Opt_case_asis: *lowercase = 0; break; - case Opt_conv_binary: - *conv = CONV_BINARY; - break; - case Opt_conv_text: - *conv = CONV_TEXT; - break; - case Opt_conv_auto: - *conv = CONV_AUTO; - break; case Opt_check_none: *chk = 0; break; @@ -370,9 +361,6 @@ HPFS filesystem options:\n\ umask=xxx set mode of files that don't have mode specified in eas\n\ case=lower lowercase all files\n\ case=asis do not lowercase files (default)\n\ - conv=binary do not convert CR/LF -> LF (default)\n\ - conv=auto convert only files with known text extensions\n\ - conv=text convert all files\n\ check=none no fs checks - kernel may crash on corrupted filesystem\n\ check=normal do some checks - it should not crash (default)\n\ check=strict do extra time-consuming checks, used for debugging\n\ @@ -394,7 +382,7 @@ static int hpfs_remount_fs(struct super_block *s, int *flags, char *data) uid_t uid; gid_t gid; umode_t umask; - int lowercase, conv, eas, chk, errs, chkdsk, timeshift; + int lowercase, eas, chk, errs, chkdsk, timeshift; int o; struct hpfs_sb_info *sbi = hpfs_sb(s); char *new_opts = kstrdup(data, GFP_KERNEL); @@ -405,11 +393,11 @@ static int hpfs_remount_fs(struct super_block *s, int *flags, char *data) lock_super(s); uid = sbi->sb_uid; gid = sbi->sb_gid; umask = 0777 & ~sbi->sb_mode; - lowercase = sbi->sb_lowercase; conv = sbi->sb_conv; + lowercase = sbi->sb_lowercase; eas = sbi->sb_eas; chk = sbi->sb_chk; chkdsk = sbi->sb_chkdsk; errs = sbi->sb_err; timeshift = sbi->sb_timeshift; - if (!(o = parse_opts(data, &uid, &gid, &umask, &lowercase, &conv, + if (!(o = parse_opts(data, &uid, &gid, &umask, &lowercase, &eas, &chk, &errs, &chkdsk, ×hift))) { printk("HPFS: bad mount options.\n"); goto out_err; @@ -427,11 +415,11 @@ static int hpfs_remount_fs(struct super_block *s, int *flags, char *data) sbi->sb_uid = uid; sbi->sb_gid = gid; sbi->sb_mode = 0777 & ~umask; - sbi->sb_lowercase = lowercase; sbi->sb_conv = conv; + sbi->sb_lowercase = lowercase; sbi->sb_eas = eas; sbi->sb_chk = chk; sbi->sb_chkdsk = chkdsk; sbi->sb_err = errs; sbi->sb_timeshift = timeshift; - if (!(*flags & MS_RDONLY)) mark_dirty(s); + if (!(*flags & MS_RDONLY)) mark_dirty(s, 1); replace_mount_options(s, new_opts); @@ -471,7 +459,7 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) uid_t uid; gid_t gid; umode_t umask; - int lowercase, conv, eas, chk, errs, chkdsk, timeshift; + int lowercase, eas, chk, errs, chkdsk, timeshift; dnode_secno root_dno; struct hpfs_dirent *de = NULL; @@ -479,11 +467,6 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) int o; - if (num_possible_cpus() > 1) { - printk(KERN_ERR "HPFS is not SMP safe\n"); - return -EINVAL; - } - save_mount_options(s, options); sbi = kzalloc(sizeof(*sbi), GFP_KERNEL); @@ -495,20 +478,20 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) sbi->sb_bmp_dir = NULL; sbi->sb_cp_table = NULL; - mutex_init(&sbi->hpfs_creation_de); + mutex_init(&sbi->hpfs_mutex); + hpfs_lock(s); uid = current_uid(); gid = current_gid(); umask = current_umask(); lowercase = 0; - conv = CONV_BINARY; eas = 2; chk = 1; errs = 1; chkdsk = 1; timeshift = 0; - if (!(o = parse_opts(options, &uid, &gid, &umask, &lowercase, &conv, + if (!(o = parse_opts(options, &uid, &gid, &umask, &lowercase, &eas, &chk, &errs, &chkdsk, ×hift))) { printk("HPFS: bad mount options.\n"); goto bail0; @@ -526,9 +509,9 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) if (!(spareblock = hpfs_map_sector(s, 17, &bh2, 0))) goto bail3; /* Check magics */ - if (/*bootblock->magic != BB_MAGIC - ||*/ superblock->magic != SB_MAGIC - || spareblock->magic != SP_MAGIC) { + if (/*le16_to_cpu(bootblock->magic) != BB_MAGIC + ||*/ le32_to_cpu(superblock->magic) != SB_MAGIC + || le32_to_cpu(spareblock->magic) != SP_MAGIC) { if (!silent) printk("HPFS: Bad magic ... probably not HPFS\n"); goto bail4; } @@ -549,19 +532,18 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) s->s_op = &hpfs_sops; s->s_d_op = &hpfs_dentry_operations; - sbi->sb_root = superblock->root; - sbi->sb_fs_size = superblock->n_sectors; - sbi->sb_bitmaps = superblock->bitmaps; - sbi->sb_dirband_start = superblock->dir_band_start; - sbi->sb_dirband_size = superblock->n_dir_band; - sbi->sb_dmap = superblock->dir_band_bitmap; + sbi->sb_root = le32_to_cpu(superblock->root); + sbi->sb_fs_size = le32_to_cpu(superblock->n_sectors); + sbi->sb_bitmaps = le32_to_cpu(superblock->bitmaps); + sbi->sb_dirband_start = le32_to_cpu(superblock->dir_band_start); + sbi->sb_dirband_size = le32_to_cpu(superblock->n_dir_band); + sbi->sb_dmap = le32_to_cpu(superblock->dir_band_bitmap); sbi->sb_uid = uid; sbi->sb_gid = gid; sbi->sb_mode = 0777 & ~umask; sbi->sb_n_free = -1; sbi->sb_n_free_dnodes = -1; sbi->sb_lowercase = lowercase; - sbi->sb_conv = conv; sbi->sb_eas = eas; sbi->sb_chk = chk; sbi->sb_chkdsk = chkdsk; @@ -573,7 +555,7 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) sbi->sb_max_fwd_alloc = 0xffffff; /* Load bitmap directory */ - if (!(sbi->sb_bmp_dir = hpfs_load_bitmap_directory(s, superblock->bitmaps))) + if (!(sbi->sb_bmp_dir = hpfs_load_bitmap_directory(s, le32_to_cpu(superblock->bitmaps)))) goto bail4; /* Check for general fs errors*/ @@ -591,20 +573,20 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) mark_buffer_dirty(bh2); } - if (spareblock->hotfixes_used || spareblock->n_spares_used) { + if (le32_to_cpu(spareblock->hotfixes_used) || le32_to_cpu(spareblock->n_spares_used)) { if (errs >= 2) { printk("HPFS: Hotfixes not supported here, try chkdsk\n"); - mark_dirty(s); + mark_dirty(s, 0); goto bail4; } hpfs_error(s, "hotfixes not supported here, try chkdsk"); if (errs == 0) printk("HPFS: Proceeding, but your filesystem will be probably corrupted by this driver...\n"); else printk("HPFS: This driver may read bad files or crash when operating on disk with hotfixes.\n"); } - if (spareblock->n_dnode_spares != spareblock->n_dnode_spares_free) { + if (le32_to_cpu(spareblock->n_dnode_spares) != le32_to_cpu(spareblock->n_dnode_spares_free)) { if (errs >= 2) { printk("HPFS: Spare dnodes used, try chkdsk\n"); - mark_dirty(s); + mark_dirty(s, 0); goto bail4; } hpfs_error(s, "warning: spare dnodes used, try chkdsk"); @@ -612,26 +594,26 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) } if (chk) { unsigned a; - if (superblock->dir_band_end - superblock->dir_band_start + 1 != superblock->n_dir_band || - superblock->dir_band_end < superblock->dir_band_start || superblock->n_dir_band > 0x4000) { + if (le32_to_cpu(superblock->dir_band_end) - le32_to_cpu(superblock->dir_band_start) + 1 != le32_to_cpu(superblock->n_dir_band) || + le32_to_cpu(superblock->dir_band_end) < le32_to_cpu(superblock->dir_band_start) || le32_to_cpu(superblock->n_dir_band) > 0x4000) { hpfs_error(s, "dir band size mismatch: dir_band_start==%08x, dir_band_end==%08x, n_dir_band==%08x", - superblock->dir_band_start, superblock->dir_band_end, superblock->n_dir_band); + le32_to_cpu(superblock->dir_band_start), le32_to_cpu(superblock->dir_band_end), le32_to_cpu(superblock->n_dir_band)); goto bail4; } a = sbi->sb_dirband_size; sbi->sb_dirband_size = 0; - if (hpfs_chk_sectors(s, superblock->dir_band_start, superblock->n_dir_band, "dir_band") || - hpfs_chk_sectors(s, superblock->dir_band_bitmap, 4, "dir_band_bitmap") || - hpfs_chk_sectors(s, superblock->bitmaps, 4, "bitmaps")) { - mark_dirty(s); + if (hpfs_chk_sectors(s, le32_to_cpu(superblock->dir_band_start), le32_to_cpu(superblock->n_dir_band), "dir_band") || + hpfs_chk_sectors(s, le32_to_cpu(superblock->dir_band_bitmap), 4, "dir_band_bitmap") || + hpfs_chk_sectors(s, le32_to_cpu(superblock->bitmaps), 4, "bitmaps")) { + mark_dirty(s, 0); goto bail4; } sbi->sb_dirband_size = a; } else printk("HPFS: You really don't want any checks? You are crazy...\n"); /* Load code page table */ - if (spareblock->n_code_pages) - if (!(sbi->sb_cp_table = hpfs_load_code_page(s, spareblock->code_page_dir))) + if (le32_to_cpu(spareblock->n_code_pages)) + if (!(sbi->sb_cp_table = hpfs_load_code_page(s, le32_to_cpu(spareblock->code_page_dir)))) printk("HPFS: Warning: code page support is disabled\n"); brelse(bh2); @@ -660,13 +642,13 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) if (!de) hpfs_error(s, "unable to find root dir"); else { - root->i_atime.tv_sec = local_to_gmt(s, de->read_date); + root->i_atime.tv_sec = local_to_gmt(s, le32_to_cpu(de->read_date)); root->i_atime.tv_nsec = 0; - root->i_mtime.tv_sec = local_to_gmt(s, de->write_date); + root->i_mtime.tv_sec = local_to_gmt(s, le32_to_cpu(de->write_date)); root->i_mtime.tv_nsec = 0; - root->i_ctime.tv_sec = local_to_gmt(s, de->creation_date); + root->i_ctime.tv_sec = local_to_gmt(s, le32_to_cpu(de->creation_date)); root->i_ctime.tv_nsec = 0; - hpfs_i(root)->i_ea_size = de->ea_size; + hpfs_i(root)->i_ea_size = le16_to_cpu(de->ea_size); hpfs_i(root)->i_parent_dir = root->i_ino; if (root->i_size == -1) root->i_size = 2048; @@ -674,6 +656,7 @@ static int hpfs_fill_super(struct super_block *s, void *options, int silent) root->i_blocks = 5; hpfs_brelse4(&qbh); } + hpfs_unlock(s); return 0; bail4: brelse(bh2); @@ -681,6 +664,7 @@ bail3: brelse(bh1); bail2: brelse(bh0); bail1: bail0: + hpfs_unlock(s); kfree(sbi->sb_bmp_dir); kfree(sbi->sb_cp_table); s->s_fs_info = NULL; diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index b9eeb1cd03ff..7aafeb8fa300 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -412,10 +412,10 @@ static int hugetlb_vmtruncate(struct inode *inode, loff_t offset) pgoff = offset >> PAGE_SHIFT; i_size_write(inode, offset); - spin_lock(&mapping->i_mmap_lock); + mutex_lock(&mapping->i_mmap_mutex); if (!prio_tree_empty(&mapping->i_mmap)) hugetlb_vmtruncate_list(&mapping->i_mmap, pgoff); - spin_unlock(&mapping->i_mmap_lock); + mutex_unlock(&mapping->i_mmap_mutex); truncate_hugepages(inode, offset); return 0; } @@ -921,7 +921,8 @@ static int can_do_hugetlb_shm(void) return capable(CAP_IPC_LOCK) || in_group_p(sysctl_hugetlb_shm_group); } -struct file *hugetlb_file_setup(const char *name, size_t size, int acctflag, +struct file *hugetlb_file_setup(const char *name, size_t size, + vm_flags_t acctflag, struct user_struct **user, int creat_flags) { int error = -ENOMEM; diff --git a/fs/inode.c b/fs/inode.c index 33c963d08ab4..0f7e88a7803f 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -1,9 +1,7 @@ /* - * linux/fs/inode.c - * * (C) 1997 Linus Torvalds + * (C) 1999 Andrea Arcangeli <andrea@suse.de> (dynamic inode allocation) */ - #include <linux/fs.h> #include <linux/mm.h> #include <linux/dcache.h> @@ -24,12 +22,14 @@ #include <linux/mount.h> #include <linux/async.h> #include <linux/posix_acl.h> +#include <linux/prefetch.h> #include <linux/ima.h> #include <linux/cred.h> +#include <linux/buffer_head.h> /* for inode_has_buffers */ #include "internal.h" /* - * inode locking rules. + * Inode locking rules: * * inode->i_lock protects: * inode->i_state, inode->i_hash, __iget() @@ -59,54 +59,11 @@ * inode_hash_lock */ -/* - * This is needed for the following functions: - * - inode_has_buffers - * - invalidate_bdev - * - * FIXME: remove all knowledge of the buffer layer from this file - */ -#include <linux/buffer_head.h> - -/* - * New inode.c implementation. - * - * This implementation has the basic premise of trying - * to be extremely low-overhead and SMP-safe, yet be - * simple enough to be "obviously correct". - * - * Famous last words. - */ - -/* inode dynamic allocation 1999, Andrea Arcangeli <andrea@suse.de> */ - -/* #define INODE_PARANOIA 1 */ -/* #define INODE_DEBUG 1 */ - -/* - * Inode lookup is no longer as critical as it used to be: - * most of the lookups are going to be through the dcache. - */ -#define I_HASHBITS i_hash_shift -#define I_HASHMASK i_hash_mask - static unsigned int i_hash_mask __read_mostly; static unsigned int i_hash_shift __read_mostly; static struct hlist_head *inode_hashtable __read_mostly; static __cacheline_aligned_in_smp DEFINE_SPINLOCK(inode_hash_lock); -/* - * Each inode can be on two separate lists. One is - * the hash list of the inode, used for lookups. The - * other linked list is the "type" list: - * "in_use" - valid inode, i_count > 0, i_nlink > 0 - * "dirty" - as "in_use" but also dirty - * "unused" - valid inode, i_count = 0 - * - * A "dirty" list is maintained for each super block, - * allowing for low-overhead inode sync() operations. - */ - static LIST_HEAD(inode_lru); static DEFINE_SPINLOCK(inode_lru_lock); @@ -325,12 +282,11 @@ void address_space_init_once(struct address_space *mapping) memset(mapping, 0, sizeof(*mapping)); INIT_RADIX_TREE(&mapping->page_tree, GFP_ATOMIC); spin_lock_init(&mapping->tree_lock); - spin_lock_init(&mapping->i_mmap_lock); + mutex_init(&mapping->i_mmap_mutex); INIT_LIST_HEAD(&mapping->private_list); spin_lock_init(&mapping->private_lock); INIT_RAW_PRIO_TREE_ROOT(&mapping->i_mmap); INIT_LIST_HEAD(&mapping->i_mmap_nonlinear); - mutex_init(&mapping->unmap_mutex); } EXPORT_SYMBOL(address_space_init_once); @@ -424,8 +380,8 @@ static unsigned long hash(struct super_block *sb, unsigned long hashval) tmp = (hashval * (unsigned long)sb) ^ (GOLDEN_RATIO_PRIME + hashval) / L1_CACHE_BYTES; - tmp = tmp ^ ((tmp ^ GOLDEN_RATIO_PRIME) >> I_HASHBITS); - return tmp & I_HASHMASK; + tmp = tmp ^ ((tmp ^ GOLDEN_RATIO_PRIME) >> i_hash_shift); + return tmp & i_hash_mask; } /** @@ -751,8 +707,12 @@ static void prune_icache(int nr_to_scan) * This function is passed the number of inodes to scan, and it returns the * total number of remaining possibly-reclaimable inodes. */ -static int shrink_icache_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) +static int shrink_icache_memory(struct shrinker *shrink, + struct shrink_control *sc) { + int nr = sc->nr_to_scan; + gfp_t gfp_mask = sc->gfp_mask; + if (nr) { /* * Nasty deadlock avoidance. We may hold various FS locks, diff --git a/fs/jbd/commit.c b/fs/jbd/commit.c index da871ee084d3..72ffa974b0b8 100644 --- a/fs/jbd/commit.c +++ b/fs/jbd/commit.c @@ -302,12 +302,6 @@ void journal_commit_transaction(journal_t *journal) * all outstanding updates to complete. */ -#ifdef COMMIT_STATS - spin_lock(&journal->j_list_lock); - summarise_journal_usage(journal); - spin_unlock(&journal->j_list_lock); -#endif - /* Do we need to erase the effects of a prior journal_flush? */ if (journal->j_flags & JFS_FLUSHED) { jbd_debug(3, "super block updated\n"); @@ -362,7 +356,7 @@ void journal_commit_transaction(journal_t *journal) * we do not require it to remember exactly which old buffers it * has reserved. This is consistent with the existing behaviour * that multiple journal_get_write_access() calls to the same - * buffer are perfectly permissable. + * buffer are perfectly permissible. */ while (commit_transaction->t_reserved_list) { jh = commit_transaction->t_reserved_list; @@ -722,8 +716,13 @@ wait_for_iobuf: required. */ JBUFFER_TRACE(jh, "file as BJ_Forget"); journal_file_buffer(jh, commit_transaction, BJ_Forget); - /* Wake up any transactions which were waiting for this - IO to complete */ + /* + * Wake up any transactions which were waiting for this + * IO to complete. The barrier must be here so that changes + * by journal_file_buffer() take effect before wake_up_bit() + * does the waitqueue check. + */ + smp_mb(); wake_up_bit(&bh->b_state, BH_Unshadow); JBUFFER_TRACE(jh, "brelse shadowed buffer"); __brelse(bh); diff --git a/fs/jbd/journal.c b/fs/jbd/journal.c index eb11601f2e00..e2d4285fbe90 100644 --- a/fs/jbd/journal.c +++ b/fs/jbd/journal.c @@ -437,9 +437,12 @@ int __log_space_left(journal_t *journal) int __log_start_commit(journal_t *journal, tid_t target) { /* - * Are we already doing a recent enough commit? + * The only transaction we can possibly wait upon is the + * currently running transaction (if it exists). Otherwise, + * the target tid must be an old one. */ - if (!tid_geq(journal->j_commit_request, target)) { + if (journal->j_running_transaction && + journal->j_running_transaction->t_tid == target) { /* * We want a new commit: OK, mark the request and wakeup the * commit thread. We do _not_ do the commit ourselves. @@ -451,7 +454,14 @@ int __log_start_commit(journal_t *journal, tid_t target) journal->j_commit_sequence); wake_up(&journal->j_wait_commit); return 1; - } + } else if (!tid_geq(journal->j_commit_request, target)) + /* This should never happen, but if it does, preserve + the evidence before kjournald goes into a loop and + increments j_commit_sequence beyond all recognition. */ + WARN_ONCE(1, "jbd: bad log_start_commit: %u %u %u %u\n", + journal->j_commit_request, journal->j_commit_sequence, + target, journal->j_running_transaction ? + journal->j_running_transaction->t_tid : 0); return 0; } @@ -770,7 +780,7 @@ journal_t * journal_init_dev(struct block_device *bdev, journal->j_wbufsize = n; journal->j_wbuf = kmalloc(n * sizeof(struct buffer_head*), GFP_KERNEL); if (!journal->j_wbuf) { - printk(KERN_ERR "%s: Cant allocate bhs for commit thread\n", + printk(KERN_ERR "%s: Can't allocate bhs for commit thread\n", __func__); goto out_err; } @@ -831,7 +841,7 @@ journal_t * journal_init_inode (struct inode *inode) journal->j_wbufsize = n; journal->j_wbuf = kmalloc(n * sizeof(struct buffer_head*), GFP_KERNEL); if (!journal->j_wbuf) { - printk(KERN_ERR "%s: Cant allocate bhs for commit thread\n", + printk(KERN_ERR "%s: Can't allocate bhs for commit thread\n", __func__); goto out_err; } diff --git a/fs/jbd/revoke.c b/fs/jbd/revoke.c index d29018307e2e..305a90763154 100644 --- a/fs/jbd/revoke.c +++ b/fs/jbd/revoke.c @@ -71,7 +71,7 @@ * switching hash tables under them. For operations on the lists of entries in * the hash table j_revoke_lock is used. * - * Finally, also replay code uses the hash tables but at this moment noone else + * Finally, also replay code uses the hash tables but at this moment no one else * can touch them (filesystem isn't mounted yet) and hence no locking is * needed. */ diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index 5b2e4c30a2a1..f7ee81a065da 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -266,7 +266,8 @@ static handle_t *new_handle(int nblocks) * This function is visible to journal users (like ext3fs), so is not * called with the journal already locked. * - * Return a pointer to a newly allocated handle, or NULL on failure + * Return a pointer to a newly allocated handle, or an ERR_PTR() value + * on failure. */ handle_t *journal_start(journal_t *journal, int nblocks) { @@ -1392,7 +1393,7 @@ int journal_stop(handle_t *handle) * by 30x or more... * * We try and optimize the sleep time against what the underlying disk - * can do, instead of having a static sleep time. This is usefull for + * can do, instead of having a static sleep time. This is useful for * the case where our storage is so fast that it is more optimal to go * ahead and force a flush and wait for the transaction to be committed * than it is to wait for an arbitrary amount of time for new writers to diff --git a/fs/jbd2/commit.c b/fs/jbd2/commit.c index fa36d7662b21..7f21cf3aaf92 100644 --- a/fs/jbd2/commit.c +++ b/fs/jbd2/commit.c @@ -105,6 +105,8 @@ static int journal_submit_commit_record(journal_t *journal, int ret; struct timespec now = current_kernel_time(); + *cbh = NULL; + if (is_journal_aborted(journal)) return 0; @@ -217,7 +219,6 @@ static int journal_submit_data_buffers(journal_t *journal, ret = err; spin_lock(&journal->j_list_lock); J_ASSERT(jinode->i_transaction == commit_transaction); - commit_transaction->t_flushed_data_blocks = 1; clear_bit(__JI_COMMIT_RUNNING, &jinode->i_flags); smp_mb__after_clear_bit(); wake_up_bit(&jinode->i_flags, __JI_COMMIT_RUNNING); @@ -336,12 +337,6 @@ void jbd2_journal_commit_transaction(journal_t *journal) * all outstanding updates to complete. */ -#ifdef COMMIT_STATS - spin_lock(&journal->j_list_lock); - summarise_journal_usage(journal); - spin_unlock(&journal->j_list_lock); -#endif - /* Do we need to erase the effects of a prior jbd2_journal_flush? */ if (journal->j_flags & JBD2_FLUSHED) { jbd_debug(3, "super block updated\n"); @@ -403,7 +398,7 @@ void jbd2_journal_commit_transaction(journal_t *journal) * we do not require it to remember exactly which old buffers it * has reserved. This is consistent with the existing behaviour * that multiple jbd2_journal_get_write_access() calls to the same - * buffer are perfectly permissable. + * buffer are perfectly permissible. */ while (commit_transaction->t_reserved_list) { jh = commit_transaction->t_reserved_list; @@ -676,12 +671,16 @@ start_journal_io: err = 0; } + write_lock(&journal->j_state_lock); + J_ASSERT(commit_transaction->t_state == T_COMMIT); + commit_transaction->t_state = T_COMMIT_DFLUSH; + write_unlock(&journal->j_state_lock); /* * If the journal is not located on the file system device, * then we must flush the file system device before we issue * the commit record */ - if (commit_transaction->t_flushed_data_blocks && + if (commit_transaction->t_need_data_flush && (journal->j_fs_dev != journal->j_dev) && (journal->j_flags & JBD2_BARRIER)) blkdev_issue_flush(journal->j_fs_dev, GFP_KERNEL, NULL); @@ -758,8 +757,13 @@ wait_for_iobuf: required. */ JBUFFER_TRACE(jh, "file as BJ_Forget"); jbd2_journal_file_buffer(jh, commit_transaction, BJ_Forget); - /* Wake up any transactions which were waiting for this - IO to complete */ + /* + * Wake up any transactions which were waiting for this IO to + * complete. The barrier must be here so that changes by + * jbd2_journal_file_buffer() take effect before wake_up_bit() + * does the waitqueue check. + */ + smp_mb(); wake_up_bit(&bh->b_state, BH_Unshadow); JBUFFER_TRACE(jh, "brelse shadowed buffer"); __brelse(bh); @@ -798,6 +802,10 @@ wait_for_iobuf: jbd2_journal_abort(journal, err); jbd_debug(3, "JBD: commit phase 5\n"); + write_lock(&journal->j_state_lock); + J_ASSERT(commit_transaction->t_state == T_COMMIT_DFLUSH); + commit_transaction->t_state = T_COMMIT_JFLUSH; + write_unlock(&journal->j_state_lock); if (!JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT)) { @@ -806,7 +814,7 @@ wait_for_iobuf: if (err) __jbd2_journal_abort_hard(journal); } - if (!err && !is_journal_aborted(journal)) + if (cbh) err = journal_wait_on_commit_record(journal, cbh); if (JBD2_HAS_INCOMPAT_FEATURE(journal, JBD2_FEATURE_INCOMPAT_ASYNC_COMMIT) && @@ -953,7 +961,7 @@ restart_loop: jbd_debug(3, "JBD: commit phase 7\n"); - J_ASSERT(commit_transaction->t_state == T_COMMIT); + J_ASSERT(commit_transaction->t_state == T_COMMIT_JFLUSH); commit_transaction->t_start = jiffies; stats.run.rs_logging = jbd2_time_diff(stats.run.rs_logging, diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 90407b8fece7..9a7826990304 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -479,9 +479,12 @@ int __jbd2_log_space_left(journal_t *journal) int __jbd2_log_start_commit(journal_t *journal, tid_t target) { /* - * Are we already doing a recent enough commit? + * The only transaction we can possibly wait upon is the + * currently running transaction (if it exists). Otherwise, + * the target tid must be an old one. */ - if (!tid_geq(journal->j_commit_request, target)) { + if (journal->j_running_transaction && + journal->j_running_transaction->t_tid == target) { /* * We want a new commit: OK, mark the request and wakeup the * commit thread. We do _not_ do the commit ourselves. @@ -493,7 +496,15 @@ int __jbd2_log_start_commit(journal_t *journal, tid_t target) journal->j_commit_sequence); wake_up(&journal->j_wait_commit); return 1; - } + } else if (!tid_geq(journal->j_commit_request, target)) + /* This should never happen, but if it does, preserve + the evidence before kjournald goes into a loop and + increments j_commit_sequence beyond all recognition. */ + WARN_ONCE(1, "jbd: bad log_start_commit: %u %u %u %u\n", + journal->j_commit_request, + journal->j_commit_sequence, + target, journal->j_running_transaction ? + journal->j_running_transaction->t_tid : 0); return 0; } @@ -577,6 +588,47 @@ int jbd2_journal_start_commit(journal_t *journal, tid_t *ptid) } /* + * Return 1 if a given transaction has not yet sent barrier request + * connected with a transaction commit. If 0 is returned, transaction + * may or may not have sent the barrier. Used to avoid sending barrier + * twice in common cases. + */ +int jbd2_trans_will_send_data_barrier(journal_t *journal, tid_t tid) +{ + int ret = 0; + transaction_t *commit_trans; + + if (!(journal->j_flags & JBD2_BARRIER)) + return 0; + read_lock(&journal->j_state_lock); + /* Transaction already committed? */ + if (tid_geq(journal->j_commit_sequence, tid)) + goto out; + commit_trans = journal->j_committing_transaction; + if (!commit_trans || commit_trans->t_tid != tid) { + ret = 1; + goto out; + } + /* + * Transaction is being committed and we already proceeded to + * submitting a flush to fs partition? + */ + if (journal->j_fs_dev != journal->j_dev) { + if (!commit_trans->t_need_data_flush || + commit_trans->t_state >= T_COMMIT_DFLUSH) + goto out; + } else { + if (commit_trans->t_state >= T_COMMIT_JFLUSH) + goto out; + } + ret = 1; +out: + read_unlock(&journal->j_state_lock); + return ret; +} +EXPORT_SYMBOL(jbd2_trans_will_send_data_barrier); + +/* * Wait for a specified commit to complete. * The caller may not hold the journal lock. */ @@ -917,7 +969,7 @@ journal_t * jbd2_journal_init_dev(struct block_device *bdev, journal->j_wbufsize = n; journal->j_wbuf = kmalloc(n * sizeof(struct buffer_head*), GFP_KERNEL); if (!journal->j_wbuf) { - printk(KERN_ERR "%s: Cant allocate bhs for commit thread\n", + printk(KERN_ERR "%s: Can't allocate bhs for commit thread\n", __func__); goto out_err; } @@ -983,7 +1035,7 @@ journal_t * jbd2_journal_init_inode (struct inode *inode) journal->j_wbufsize = n; journal->j_wbuf = kmalloc(n * sizeof(struct buffer_head*), GFP_KERNEL); if (!journal->j_wbuf) { - printk(KERN_ERR "%s: Cant allocate bhs for commit thread\n", + printk(KERN_ERR "%s: Can't allocate bhs for commit thread\n", __func__); goto out_err; } @@ -2413,10 +2465,12 @@ const char *jbd2_dev_to_name(dev_t device) new_dev = kmalloc(sizeof(struct devname_cache), GFP_KERNEL); if (!new_dev) return "NODEV-ALLOCFAILURE"; /* Something non-NULL */ + bd = bdget(device); spin_lock(&devname_cache_lock); if (devcache[i]) { if (devcache[i]->device == device) { kfree(new_dev); + bdput(bd); ret = devcache[i]->devname; spin_unlock(&devname_cache_lock); return ret; @@ -2425,7 +2479,6 @@ const char *jbd2_dev_to_name(dev_t device) } devcache[i] = new_dev; devcache[i]->device = device; - bd = bdget(device); if (bd) { bdevname(bd, devcache[i]->devname); bdput(bd); diff --git a/fs/jbd2/revoke.c b/fs/jbd2/revoke.c index 9ad321fd63fd..69fd93588118 100644 --- a/fs/jbd2/revoke.c +++ b/fs/jbd2/revoke.c @@ -71,7 +71,7 @@ * switching hash tables under them. For operations on the lists of entries in * the hash table j_revoke_lock is used. * - * Finally, also replay code uses the hash tables but at this moment noone else + * Finally, also replay code uses the hash tables but at this moment no one else * can touch them (filesystem isn't mounted yet) and hence no locking is * needed. */ diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index 1d1191050f99..3eec82d32fd4 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -82,7 +82,7 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction) */ /* - * Update transiaction's maximum wait time, if debugging is enabled. + * Update transaction's maximum wait time, if debugging is enabled. * * In order for t_max_wait to be reliable, it must be protected by a * lock. But doing so will mean that start_this_handle() can not be @@ -91,11 +91,10 @@ jbd2_get_transaction(journal_t *journal, transaction_t *transaction) * means that maximum wait time reported by the jbd2_run_stats * tracepoint will always be zero. */ -static inline void update_t_max_wait(transaction_t *transaction) +static inline void update_t_max_wait(transaction_t *transaction, + unsigned long ts) { #ifdef CONFIG_JBD2_DEBUG - unsigned long ts = jiffies; - if (jbd2_journal_enable_debug && time_after(transaction->t_start, ts)) { ts = jbd2_time_diff(ts, transaction->t_start); @@ -121,6 +120,7 @@ static int start_this_handle(journal_t *journal, handle_t *handle, tid_t tid; int needed, need_to_start; int nblocks = handle->h_buffer_credits; + unsigned long ts = jiffies; if (nblocks > journal->j_max_transaction_buffers) { printk(KERN_ERR "JBD: %s wants too many credits (%d > %d)\n", @@ -271,7 +271,7 @@ repeat: /* OK, account for the buffers that this operation expects to * use and add the handle to the running transaction. */ - update_t_max_wait(transaction); + update_t_max_wait(transaction, ts); handle->h_transaction = transaction; atomic_inc(&transaction->t_updates); atomic_inc(&transaction->t_handle_count); @@ -316,7 +316,8 @@ static handle_t *new_handle(int nblocks) * This function is visible to journal users (like ext3fs), so is not * called with the journal already locked. * - * Return a pointer to a newly allocated handle, or NULL on failure + * Return a pointer to a newly allocated handle, or an ERR_PTR() value + * on failure. */ handle_t *jbd2__journal_start(journal_t *journal, int nblocks, int gfp_mask) { @@ -921,8 +922,8 @@ int jbd2_journal_get_create_access(handle_t *handle, struct buffer_head *bh) */ JBUFFER_TRACE(jh, "cancelling revoke"); jbd2_journal_cancel_revoke(handle, jh); - jbd2_journal_put_journal_head(jh); out: + jbd2_journal_put_journal_head(jh); return err; } @@ -1403,7 +1404,7 @@ int jbd2_journal_stop(handle_t *handle) /* * Once we drop t_updates, if it goes to zero the transaction - * could start commiting on us and eventually disappear. So + * could start committing on us and eventually disappear. So * once we do this, we must not dereference transaction * pointer again. */ @@ -2147,6 +2148,13 @@ int jbd2_journal_file_inode(handle_t *handle, struct jbd2_inode *jinode) jinode->i_next_transaction == transaction) goto done; + /* + * We only ever set this variable to 1 so the test is safe. Since + * t_need_data_flush is likely to be set, we do the test to save some + * cacheline bouncing + */ + if (!transaction->t_need_data_flush) + transaction->t_need_data_flush = 1; /* On some different transaction's list - should be * the committing one */ if (jinode->i_transaction) { diff --git a/fs/jffs2/TODO b/fs/jffs2/TODO index 5d3ea4070f01..ca28964abd4b 100644 --- a/fs/jffs2/TODO +++ b/fs/jffs2/TODO @@ -11,7 +11,7 @@ - checkpointing (do we need this? scan is quite fast) - make the scan code populate real inodes so read_inode just after mount doesn't have to read the flash twice for large files. - Make this a per-inode option, changable with chattr, so you can + Make this a per-inode option, changeable with chattr, so you can decide which inodes should be in-core immediately after mount. - test, test, test diff --git a/fs/jffs2/dir.c b/fs/jffs2/dir.c index 82faddd1f321..4bca6a2e5c07 100644 --- a/fs/jffs2/dir.c +++ b/fs/jffs2/dir.c @@ -75,7 +75,6 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target, struct nameidata *nd) { struct jffs2_inode_info *dir_f; - struct jffs2_sb_info *c; struct jffs2_full_dirent *fd = NULL, *fd_list; uint32_t ino = 0; struct inode *inode = NULL; @@ -86,7 +85,6 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target, return ERR_PTR(-ENAMETOOLONG); dir_f = JFFS2_INODE_INFO(dir_i); - c = JFFS2_SB_INFO(dir_i->i_sb); mutex_lock(&dir_f->sem); @@ -119,7 +117,6 @@ static struct dentry *jffs2_lookup(struct inode *dir_i, struct dentry *target, static int jffs2_readdir(struct file *filp, void *dirent, filldir_t filldir) { struct jffs2_inode_info *f; - struct jffs2_sb_info *c; struct inode *inode = filp->f_path.dentry->d_inode; struct jffs2_full_dirent *fd; unsigned long offset, curofs; @@ -127,7 +124,6 @@ static int jffs2_readdir(struct file *filp, void *dirent, filldir_t filldir) D1(printk(KERN_DEBUG "jffs2_readdir() for dir_i #%lu\n", filp->f_path.dentry->d_inode->i_ino)); f = JFFS2_INODE_INFO(inode); - c = JFFS2_SB_INFO(inode->i_sb); offset = filp->f_pos; diff --git a/fs/jffs2/fs.c b/fs/jffs2/fs.c index e896e67767eb..46ad619b6124 100644 --- a/fs/jffs2/fs.c +++ b/fs/jffs2/fs.c @@ -357,7 +357,7 @@ error: return ERR_PTR(ret); } -void jffs2_dirty_inode(struct inode *inode) +void jffs2_dirty_inode(struct inode *inode, int flags) { struct iattr iattr; diff --git a/fs/jffs2/os-linux.h b/fs/jffs2/os-linux.h index 00bae7cc2e48..65c6c43ca482 100644 --- a/fs/jffs2/os-linux.h +++ b/fs/jffs2/os-linux.h @@ -172,7 +172,7 @@ int jffs2_setattr (struct dentry *, struct iattr *); int jffs2_do_setattr (struct inode *, struct iattr *); struct inode *jffs2_iget(struct super_block *, unsigned long); void jffs2_evict_inode (struct inode *); -void jffs2_dirty_inode(struct inode *inode); +void jffs2_dirty_inode(struct inode *inode, int flags); struct inode *jffs2_new_inode (struct inode *dir_i, int mode, struct jffs2_raw_inode *ri); int jffs2_statfs (struct dentry *, struct kstatfs *); diff --git a/fs/jffs2/readinode.c b/fs/jffs2/readinode.c index d32ee9412cb9..2ab1a0d91210 100644 --- a/fs/jffs2/readinode.c +++ b/fs/jffs2/readinode.c @@ -24,7 +24,7 @@ * * Returns: 0 if the data CRC is correct; * 1 - if incorrect; - * error code if an error occured. + * error code if an error occurred. */ static int check_node_data(struct jffs2_sb_info *c, struct jffs2_tmp_dnode_info *tn) { diff --git a/fs/jffs2/scan.c b/fs/jffs2/scan.c index b632dddcb482..8d8cd3419d02 100644 --- a/fs/jffs2/scan.c +++ b/fs/jffs2/scan.c @@ -94,7 +94,7 @@ int jffs2_scan_medium(struct jffs2_sb_info *c) uint32_t buf_size = 0; struct jffs2_summary *s = NULL; /* summary info collected by the scan process */ #ifndef __ECOS - size_t pointlen; + size_t pointlen, try_size; if (c->mtd->point) { ret = c->mtd->point(c->mtd, 0, c->mtd->size, &pointlen, @@ -113,18 +113,21 @@ int jffs2_scan_medium(struct jffs2_sb_info *c) /* For NAND it's quicker to read a whole eraseblock at a time, apparently */ if (jffs2_cleanmarker_oob(c)) - buf_size = c->sector_size; + try_size = c->sector_size; else - buf_size = PAGE_SIZE; + try_size = PAGE_SIZE; - /* Respect kmalloc limitations */ - if (buf_size > 128*1024) - buf_size = 128*1024; + D1(printk(KERN_DEBUG "Trying to allocate readbuf of %zu " + "bytes\n", try_size)); - D1(printk(KERN_DEBUG "Allocating readbuf of %d bytes\n", buf_size)); - flashbuf = kmalloc(buf_size, GFP_KERNEL); + flashbuf = mtd_kmalloc_up_to(c->mtd, &try_size); if (!flashbuf) return -ENOMEM; + + D1(printk(KERN_DEBUG "Allocated readbuf of %zu bytes\n", + try_size)); + + buf_size = (uint32_t)try_size; } if (jffs2_sum_active()) { diff --git a/fs/jffs2/summary.c b/fs/jffs2/summary.c index 800171dca53b..e537fb0e0184 100644 --- a/fs/jffs2/summary.c +++ b/fs/jffs2/summary.c @@ -121,7 +121,7 @@ int jffs2_sum_add_inode_mem(struct jffs2_summary *s, struct jffs2_raw_inode *ri, temp->nodetype = ri->nodetype; temp->inode = ri->ino; temp->version = ri->version; - temp->offset = cpu_to_je32(ofs); /* relative offset from the begining of the jeb */ + temp->offset = cpu_to_je32(ofs); /* relative offset from the beginning of the jeb */ temp->totlen = ri->totlen; temp->next = NULL; @@ -139,7 +139,7 @@ int jffs2_sum_add_dirent_mem(struct jffs2_summary *s, struct jffs2_raw_dirent *r temp->nodetype = rd->nodetype; temp->totlen = rd->totlen; - temp->offset = cpu_to_je32(ofs); /* relative from the begining of the jeb */ + temp->offset = cpu_to_je32(ofs); /* relative from the beginning of the jeb */ temp->pino = rd->pino; temp->version = rd->version; temp->ino = rd->ino; diff --git a/fs/jffs2/wbuf.c b/fs/jffs2/wbuf.c index 07ee1546b2fa..4515bea0268f 100644 --- a/fs/jffs2/wbuf.c +++ b/fs/jffs2/wbuf.c @@ -1116,7 +1116,7 @@ int jffs2_write_nand_cleanmarker(struct jffs2_sb_info *c, /* * On NAND we try to mark this block bad. If the block was erased more - * than MAX_ERASE_FAILURES we mark it finaly bad. + * than MAX_ERASE_FAILURES we mark it finally bad. * Don't care about failures. This block remains on the erase-pending * or badblock list as long as nobody manipulates the flash with * a bootloader or something like that. diff --git a/fs/jfs/inode.c b/fs/jfs/inode.c index eddbb373209e..109655904bbc 100644 --- a/fs/jfs/inode.c +++ b/fs/jfs/inode.c @@ -173,7 +173,7 @@ void jfs_evict_inode(struct inode *inode) dquot_drop(inode); } -void jfs_dirty_inode(struct inode *inode) +void jfs_dirty_inode(struct inode *inode, int flags) { static int noisy = 5; diff --git a/fs/jfs/jfs_dmap.c b/fs/jfs/jfs_dmap.c index c92ea3b3ea5e..4496872cf4e7 100644 --- a/fs/jfs/jfs_dmap.c +++ b/fs/jfs/jfs_dmap.c @@ -1649,7 +1649,7 @@ static int dbFindCtl(struct bmap * bmp, int l2nb, int level, s64 * blkno) } /* search the tree within the dmap control page for - * sufficent free space. if sufficient free space is found, + * sufficient free space. if sufficient free space is found, * dbFindLeaf() returns the index of the leaf at which * free space was found. */ @@ -2744,7 +2744,7 @@ static int dbJoin(dmtree_t * tp, int leafno, int newval) /* check which (leafno or buddy) is the left buddy. * the left buddy gets to claim the blocks resulting * from the join while the right gets to claim none. - * the left buddy is also eligable to participate in + * the left buddy is also eligible to participate in * a join at the next higher level while the right * is not. * diff --git a/fs/jfs/jfs_extent.c b/fs/jfs/jfs_extent.c index 5d3bbd10f8db..e5fe8506ed16 100644 --- a/fs/jfs/jfs_extent.c +++ b/fs/jfs/jfs_extent.c @@ -126,7 +126,7 @@ extAlloc(struct inode *ip, s64 xlen, s64 pno, xad_t * xp, bool abnr) /* allocate the disk blocks for the extent. initially, extBalloc() * will try to allocate disk blocks for the requested size (xlen). - * if this fails (xlen contiguous free blocks not avaliable), it'll + * if this fails (xlen contiguous free blocks not available), it'll * try to allocate a smaller number of blocks (producing a smaller * extent), with this smaller number of blocks consisting of the * requested number of blocks rounded down to the next smaller @@ -481,7 +481,7 @@ int extFill(struct inode *ip, xad_t * xp) * * initially, we will try to allocate disk blocks for the * requested size (nblocks). if this fails (nblocks - * contiguous free blocks not avaliable), we'll try to allocate + * contiguous free blocks not available), we'll try to allocate * a smaller number of blocks (producing a smaller extent), with * this smaller number of blocks consisting of the requested * number of blocks rounded down to the next smaller power of 2 @@ -575,7 +575,7 @@ extBalloc(struct inode *ip, s64 hint, s64 * nblocks, s64 * blkno) * to a new set of blocks. If moving the extent, we initially * will try to allocate disk blocks for the requested size * (newnblks). if this fails (new contiguous free blocks not - * avaliable), we'll try to allocate a smaller number of + * available), we'll try to allocate a smaller number of * blocks (producing a smaller extent), with this smaller * number of blocks consisting of the requested number of * blocks rounded down to the next smaller power of 2 diff --git a/fs/jfs/jfs_imap.c b/fs/jfs/jfs_imap.c index 3a09423b6c22..ed53a4740168 100644 --- a/fs/jfs/jfs_imap.c +++ b/fs/jfs/jfs_imap.c @@ -1069,7 +1069,7 @@ int diFree(struct inode *ip) */ if (iagp->nfreeexts == cpu_to_le32(EXTSPERIAG - 1)) { /* in preparation for removing the iag from the - * ag extent free list, read the iags preceeding + * ag extent free list, read the iags preceding * and following the iag on the ag extent free * list. */ @@ -1095,7 +1095,7 @@ int diFree(struct inode *ip) int inofreefwd = le32_to_cpu(iagp->inofreefwd); /* in preparation for removing the iag from the - * ag inode free list, read the iags preceeding + * ag inode free list, read the iags preceding * and following the iag on the ag inode free * list. before reading these iags, we must make * sure that we already don't have them in hand @@ -1681,7 +1681,7 @@ diAllocAG(struct inomap * imap, int agno, bool dir, struct inode *ip) * try to allocate a new extent of free inodes. */ if (addext) { - /* if free space is not avaliable for this new extent, try + /* if free space is not available for this new extent, try * below to allocate a free and existing (already backed) * inode from the ag. */ @@ -2036,7 +2036,7 @@ static int diAllocBit(struct inomap * imap, struct iag * iagp, int ino) /* check if this is the last free inode within the iag. * if so, it will have to be removed from the ag free - * inode list, so get the iags preceeding and following + * inode list, so get the iags preceding and following * it on the list. */ if (iagp->nfreeinos == cpu_to_le32(1)) { @@ -2208,7 +2208,7 @@ static int diNewExt(struct inomap * imap, struct iag * iagp, int extno) /* check if this is the last free extent within the * iag. if so, the iag must be removed from the ag - * free extent list, so get the iags preceeding and + * free extent list, so get the iags preceding and * following the iag on this list. */ if (iagp->nfreeexts == cpu_to_le32(1)) { @@ -2504,7 +2504,7 @@ diNewIAG(struct inomap * imap, int *iagnop, int agno, struct metapage ** mpp) } - /* get the next avaliable iag number */ + /* get the next available iag number */ iagno = imap->im_nextiag; /* make sure that we have not exceeded the maximum inode @@ -2615,7 +2615,7 @@ diNewIAG(struct inomap * imap, int *iagnop, int agno, struct metapage ** mpp) duplicateIXtree(sb, blkno, xlen, &xaddr); - /* update the next avaliable iag number */ + /* update the next available iag number */ imap->im_nextiag += 1; /* Add the iag to the iag free list so we don't lose the iag diff --git a/fs/jfs/jfs_inode.h b/fs/jfs/jfs_inode.h index 155e91eff07d..ec2fb8b945fc 100644 --- a/fs/jfs/jfs_inode.h +++ b/fs/jfs/jfs_inode.h @@ -28,7 +28,7 @@ extern struct inode *jfs_iget(struct super_block *, unsigned long); extern int jfs_commit_inode(struct inode *, int); extern int jfs_write_inode(struct inode *, struct writeback_control *); extern void jfs_evict_inode(struct inode *); -extern void jfs_dirty_inode(struct inode *); +extern void jfs_dirty_inode(struct inode *, int); extern void jfs_truncate(struct inode *); extern void jfs_truncate_nolock(struct inode *, loff_t); extern void jfs_free_zero_link(struct inode *); diff --git a/fs/jfs/jfs_logmgr.h b/fs/jfs/jfs_logmgr.h index 9236bc49ae7f..e38c21598850 100644 --- a/fs/jfs/jfs_logmgr.h +++ b/fs/jfs/jfs_logmgr.h @@ -288,7 +288,7 @@ struct lrd { /* * SYNCPT: log sync point * - * replay log upto syncpt address specified; + * replay log up to syncpt address specified; */ struct { __le32 sync; /* 4: syncpt address (0 = here) */ diff --git a/fs/jfs/jfs_metapage.h b/fs/jfs/jfs_metapage.h index d94f8d9e87d7..a78beda85f68 100644 --- a/fs/jfs/jfs_metapage.h +++ b/fs/jfs/jfs_metapage.h @@ -75,7 +75,7 @@ extern void grab_metapage(struct metapage *); extern void force_metapage(struct metapage *); /* - * hold_metapage and put_metapage are used in conjuction. The page lock + * hold_metapage and put_metapage are used in conjunction. The page lock * is not dropped between the two, so no other threads can get or release * the metapage */ diff --git a/fs/jfs/jfs_txnmgr.c b/fs/jfs/jfs_txnmgr.c index 9466957ec841..f6cc0c09ec63 100644 --- a/fs/jfs/jfs_txnmgr.c +++ b/fs/jfs/jfs_txnmgr.c @@ -636,7 +636,7 @@ struct tlock *txLock(tid_t tid, struct inode *ip, struct metapage * mp, * the inode of the page and available to all anonymous * transactions until txCommit() time at which point * they are transferred to the transaction tlock list of - * the commiting transaction of the inode) + * the committing transaction of the inode) */ if (xtid == 0) { tlck->tid = tid; diff --git a/fs/jfs/resize.c b/fs/jfs/resize.c index 1aba0039f1c9..8ea5efb5a34e 100644 --- a/fs/jfs/resize.c +++ b/fs/jfs/resize.c @@ -57,7 +57,7 @@ * 2. compute new FSCKSize from new LVSize; * 3. set new FSSize as MIN(FSSize, LVSize-(LogSize+FSCKSize)) where * assert(new FSSize >= old FSSize), - * i.e., file system must not be shrinked; + * i.e., file system must not be shrunk; */ int jfs_extendfs(struct super_block *sb, s64 newLVSize, int newLogSize) { @@ -182,7 +182,7 @@ int jfs_extendfs(struct super_block *sb, s64 newLVSize, int newLogSize) */ newFSSize = newLVSize - newLogSize - newFSCKSize; - /* file system cannot be shrinked */ + /* file system cannot be shrunk */ if (newFSSize < bmp->db_mapsize) { rc = -EINVAL; goto out; diff --git a/fs/jfs/super.c b/fs/jfs/super.c index eeca48a031ab..06c8a67cbe76 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -644,7 +644,7 @@ static int jfs_show_options(struct seq_file *seq, struct vfsmount *vfs) /* Read data from quotafile - avoid pagecache and such because we cannot afford * acquiring the locks... As quota files are never truncated and quota code - * itself serializes the operations (and noone else should touch the files) + * itself serializes the operations (and no one else should touch the files) * we don't have to be afraid of races */ static ssize_t jfs_quota_read(struct super_block *sb, int type, char *data, size_t len, loff_t off) diff --git a/fs/logfs/dev_bdev.c b/fs/logfs/dev_bdev.c index 1adc8d455f0e..df0de27c2733 100644 --- a/fs/logfs/dev_bdev.c +++ b/fs/logfs/dev_bdev.c @@ -10,6 +10,7 @@ #include <linux/blkdev.h> #include <linux/buffer_head.h> #include <linux/gfp.h> +#include <linux/prefetch.h> #define PAGE_OFS(ofs) ((ofs) & (PAGE_SIZE-1)) diff --git a/fs/logfs/dev_mtd.c b/fs/logfs/dev_mtd.c index 7466e9dcc8c5..339e17e9133d 100644 --- a/fs/logfs/dev_mtd.c +++ b/fs/logfs/dev_mtd.c @@ -60,7 +60,7 @@ static int mtd_write(struct super_block *sb, loff_t ofs, size_t len, void *buf) * asynchronous properties. So just to prevent the first implementor of such * a thing from breaking logfs in 2350, we do the usual pointless dance to * declare a completion variable and wait for completion before returning - * from mtd_erase(). What an excercise in futility! + * from mtd_erase(). What an exercise in futility! */ static void logfs_erase_callback(struct erase_info *ei) { diff --git a/fs/logfs/dir.c b/fs/logfs/dir.c index f9ddf0c388c8..9ed89d1663f8 100644 --- a/fs/logfs/dir.c +++ b/fs/logfs/dir.c @@ -92,7 +92,7 @@ static int beyond_eof(struct inode *inode, loff_t bix) * so short names (len <= 9) don't even occupy the complete 32bit name * space. A prime >256 ensures short names quickly spread the 32bit * name space. Add about 26 for the estimated amount of information - * of each character and pick a prime nearby, preferrably a bit-sparse + * of each character and pick a prime nearby, preferably a bit-sparse * one. */ static u32 hash_32(const char *s, int len, u32 seed) diff --git a/fs/logfs/readwrite.c b/fs/logfs/readwrite.c index ee99a9f5dfd3..d8d09380c7de 100644 --- a/fs/logfs/readwrite.c +++ b/fs/logfs/readwrite.c @@ -481,7 +481,7 @@ static int inode_write_alias(struct super_block *sb, val = inode_val0(inode); break; case INODE_USED_OFS: - val = cpu_to_be64(li->li_used_bytes);; + val = cpu_to_be64(li->li_used_bytes); break; case INODE_SIZE_OFS: val = cpu_to_be64(i_size_read(inode)); @@ -1616,7 +1616,7 @@ int logfs_rewrite_block(struct inode *inode, u64 bix, u64 ofs, err = logfs_write_buf(inode, page, flags); if (!err && shrink_level(gc_level) == 0) { /* Rewrite cannot mark the inode dirty but has to - * write it immediatly. + * write it immediately. * Q: Can't we just create an alias for the inode * instead? And if not, why not? */ diff --git a/fs/logfs/super.c b/fs/logfs/super.c index 33435e4b14d2..ce03a182c771 100644 --- a/fs/logfs/super.c +++ b/fs/logfs/super.c @@ -480,10 +480,6 @@ static int logfs_read_sb(struct super_block *sb, int read_only) !read_only) return -EIO; - mutex_init(&super->s_dirop_mutex); - mutex_init(&super->s_object_alias_mutex); - INIT_LIST_HEAD(&super->s_freeing_list); - ret = logfs_init_rw(sb); if (ret) return ret; @@ -601,6 +597,10 @@ static struct dentry *logfs_mount(struct file_system_type *type, int flags, if (!super) return ERR_PTR(-ENOMEM); + mutex_init(&super->s_dirop_mutex); + mutex_init(&super->s_object_alias_mutex); + INIT_LIST_HEAD(&super->s_freeing_list); + if (!devname) err = logfs_get_sb_bdev(super, type, devname); else if (strncmp(devname, "mtd", 3)) diff --git a/fs/mbcache.c b/fs/mbcache.c index a25444ab2baf..8c32ef3ba88e 100644 --- a/fs/mbcache.c +++ b/fs/mbcache.c @@ -90,7 +90,8 @@ static DEFINE_SPINLOCK(mb_cache_spinlock); * What the mbcache registers as to get shrunk dynamically. */ -static int mb_cache_shrink_fn(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask); +static int mb_cache_shrink_fn(struct shrinker *shrink, + struct shrink_control *sc); static struct shrinker mb_cache_shrinker = { .shrink = mb_cache_shrink_fn, @@ -156,18 +157,19 @@ forget: * gets low. * * @shrink: (ignored) - * @nr_to_scan: Number of objects to scan - * @gfp_mask: (ignored) + * @sc: shrink_control passed from reclaim * * Returns the number of objects which are present in the cache. */ static int -mb_cache_shrink_fn(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask) +mb_cache_shrink_fn(struct shrinker *shrink, struct shrink_control *sc) { LIST_HEAD(free_list); struct mb_cache *cache; struct mb_cache_entry *entry, *tmp; int count = 0; + int nr_to_scan = sc->nr_to_scan; + gfp_t gfp_mask = sc->gfp_mask; mb_debug("trying to free %d entries", nr_to_scan); spin_lock(&mb_cache_spinlock); @@ -542,7 +544,7 @@ __mb_cache_entry_find(struct list_head *l, struct list_head *head, * mb_cache_entry_find_first() * * Find the first cache entry on a given device with a certain key in - * an additional index. Additonal matches can be found with + * an additional index. Additional matches can be found with * mb_cache_entry_find_next(). Returns NULL if no match was found. The * returned cache entry is locked for shared access ("multiple readers"). * diff --git a/fs/mpage.c b/fs/mpage.c index 0afc809e46e0..fdfae9fa98cd 100644 --- a/fs/mpage.c +++ b/fs/mpage.c @@ -27,6 +27,7 @@ #include <linux/writeback.h> #include <linux/backing-dev.h> #include <linux/pagevec.h> +#include <linux/cleancache.h> /* * I/O completion handler for multipage BIOs. @@ -271,6 +272,12 @@ do_mpage_readpage(struct bio *bio, struct page *page, unsigned nr_pages, SetPageMappedToDisk(page); } + if (fully_mapped && blocks_per_page == 1 && !PageUptodate(page) && + cleancache_get_page(page) == 0) { + SetPageUptodate(page); + goto confused; + } + /* * This page will go to BIO. Do we need to send this BIO off first? */ diff --git a/fs/namei.c b/fs/namei.c index 3cb616d38d9c..1ab641f2e78e 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -70,7 +70,7 @@ * name indicated by the symlink. The old code always complained that the * name already exists, due to not following the symlink even if its target * is nonexistent. The new semantics affects also mknod() and link() when - * the name is a symlink pointing to a non-existant name. + * the name is a symlink pointing to a non-existent name. * * I don't know which semantics is the right one, since I have no access * to standards. But I found by trial that HP-UX 9.0 has the full "new" @@ -179,7 +179,7 @@ EXPORT_SYMBOL(putname); static int acl_permission_check(struct inode *inode, int mask, unsigned int flags, int (*check_acl)(struct inode *inode, int mask, unsigned int flags)) { - umode_t mode = inode->i_mode; + unsigned int mode = inode->i_mode; mask &= MAY_READ | MAY_WRITE | MAY_EXEC; @@ -391,79 +391,28 @@ void path_put(struct path *path) } EXPORT_SYMBOL(path_put); -/** - * nameidata_drop_rcu - drop this nameidata out of rcu-walk - * @nd: nameidata pathwalk data to drop - * Returns: 0 on success, -ECHILD on failure - * +/* * Path walking has 2 modes, rcu-walk and ref-walk (see - * Documentation/filesystems/path-lookup.txt). __drop_rcu* functions attempt - * to drop out of rcu-walk mode and take normal reference counts on dentries - * and vfsmounts to transition to rcu-walk mode. __drop_rcu* functions take - * refcounts at the last known good point before rcu-walk got stuck, so - * ref-walk may continue from there. If this is not successful (eg. a seqcount - * has changed), then failure is returned and path walk restarts from the - * beginning in ref-walk mode. - * - * nameidata_drop_rcu attempts to drop the current nd->path and nd->root into - * ref-walk. Must be called from rcu-walk context. + * Documentation/filesystems/path-lookup.txt). In situations when we can't + * continue in RCU mode, we attempt to drop out of rcu-walk mode and grab + * normal reference counts on dentries and vfsmounts to transition to rcu-walk + * mode. Refcounts are grabbed at the last known good point before rcu-walk + * got stuck, so ref-walk may continue from there. If this is not successful + * (eg. a seqcount has changed), then failure is returned and it's up to caller + * to restart the path walk from the beginning in ref-walk mode. */ -static int nameidata_drop_rcu(struct nameidata *nd) -{ - struct fs_struct *fs = current->fs; - struct dentry *dentry = nd->path.dentry; - int want_root = 0; - - BUG_ON(!(nd->flags & LOOKUP_RCU)); - if (nd->root.mnt && !(nd->flags & LOOKUP_ROOT)) { - want_root = 1; - spin_lock(&fs->lock); - if (nd->root.mnt != fs->root.mnt || - nd->root.dentry != fs->root.dentry) - goto err_root; - } - spin_lock(&dentry->d_lock); - if (!__d_rcu_to_refcount(dentry, nd->seq)) - goto err; - BUG_ON(nd->inode != dentry->d_inode); - spin_unlock(&dentry->d_lock); - if (want_root) { - path_get(&nd->root); - spin_unlock(&fs->lock); - } - mntget(nd->path.mnt); - - rcu_read_unlock(); - br_read_unlock(vfsmount_lock); - nd->flags &= ~LOOKUP_RCU; - return 0; -err: - spin_unlock(&dentry->d_lock); -err_root: - if (want_root) - spin_unlock(&fs->lock); - return -ECHILD; -} - -/* Try to drop out of rcu-walk mode if we were in it, otherwise do nothing. */ -static inline int nameidata_drop_rcu_maybe(struct nameidata *nd) -{ - if (nd->flags & LOOKUP_RCU) - return nameidata_drop_rcu(nd); - return 0; -} /** - * nameidata_dentry_drop_rcu - drop nameidata and dentry out of rcu-walk - * @nd: nameidata pathwalk data to drop - * @dentry: dentry to drop + * unlazy_walk - try to switch to ref-walk mode. + * @nd: nameidata pathwalk data + * @dentry: child of nd->path.dentry or NULL * Returns: 0 on success, -ECHILD on failure * - * nameidata_dentry_drop_rcu attempts to drop the current nd->path and nd->root, - * and dentry into ref-walk. @dentry must be a path found by a do_lookup call on - * @nd. Must be called from rcu-walk context. + * unlazy_walk attempts to legitimize the current nd->path, nd->root and dentry + * for ref-walk mode. @dentry must be a path found by a do_lookup call on + * @nd or NULL. Must be called from rcu-walk context. */ -static int nameidata_dentry_drop_rcu(struct nameidata *nd, struct dentry *dentry) +static int unlazy_walk(struct nameidata *nd, struct dentry *dentry) { struct fs_struct *fs = current->fs; struct dentry *parent = nd->path.dentry; @@ -478,18 +427,25 @@ static int nameidata_dentry_drop_rcu(struct nameidata *nd, struct dentry *dentry goto err_root; } spin_lock(&parent->d_lock); - spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); - if (!__d_rcu_to_refcount(dentry, nd->seq)) - goto err; - /* - * If the sequence check on the child dentry passed, then the child has - * not been removed from its parent. This means the parent dentry must - * be valid and able to take a reference at this point. - */ - BUG_ON(!IS_ROOT(dentry) && dentry->d_parent != parent); - BUG_ON(!parent->d_count); - parent->d_count++; - spin_unlock(&dentry->d_lock); + if (!dentry) { + if (!__d_rcu_to_refcount(parent, nd->seq)) + goto err_parent; + BUG_ON(nd->inode != parent->d_inode); + } else { + spin_lock_nested(&dentry->d_lock, DENTRY_D_LOCK_NESTED); + if (!__d_rcu_to_refcount(dentry, nd->seq)) + goto err_child; + /* + * If the sequence check on the child dentry passed, then + * the child has not been removed from its parent. This + * means the parent dentry must be valid and able to take + * a reference at this point. + */ + BUG_ON(!IS_ROOT(dentry) && dentry->d_parent != parent); + BUG_ON(!parent->d_count); + parent->d_count++; + spin_unlock(&dentry->d_lock); + } spin_unlock(&parent->d_lock); if (want_root) { path_get(&nd->root); @@ -501,8 +457,10 @@ static int nameidata_dentry_drop_rcu(struct nameidata *nd, struct dentry *dentry br_read_unlock(vfsmount_lock); nd->flags &= ~LOOKUP_RCU; return 0; -err: + +err_child: spin_unlock(&dentry->d_lock); +err_parent: spin_unlock(&parent->d_lock); err_root: if (want_root) @@ -510,59 +468,6 @@ err_root: return -ECHILD; } -/* Try to drop out of rcu-walk mode if we were in it, otherwise do nothing. */ -static inline int nameidata_dentry_drop_rcu_maybe(struct nameidata *nd, struct dentry *dentry) -{ - if (nd->flags & LOOKUP_RCU) { - if (unlikely(nameidata_dentry_drop_rcu(nd, dentry))) { - nd->flags &= ~LOOKUP_RCU; - if (!(nd->flags & LOOKUP_ROOT)) - nd->root.mnt = NULL; - rcu_read_unlock(); - br_read_unlock(vfsmount_lock); - return -ECHILD; - } - } - return 0; -} - -/** - * nameidata_drop_rcu_last - drop nameidata ending path walk out of rcu-walk - * @nd: nameidata pathwalk data to drop - * Returns: 0 on success, -ECHILD on failure - * - * nameidata_drop_rcu_last attempts to drop the current nd->path into ref-walk. - * nd->path should be the final element of the lookup, so nd->root is discarded. - * Must be called from rcu-walk context. - */ -static int nameidata_drop_rcu_last(struct nameidata *nd) -{ - struct dentry *dentry = nd->path.dentry; - - BUG_ON(!(nd->flags & LOOKUP_RCU)); - nd->flags &= ~LOOKUP_RCU; - if (!(nd->flags & LOOKUP_ROOT)) - nd->root.mnt = NULL; - spin_lock(&dentry->d_lock); - if (!__d_rcu_to_refcount(dentry, nd->seq)) - goto err_unlock; - BUG_ON(nd->inode != dentry->d_inode); - spin_unlock(&dentry->d_lock); - - mntget(nd->path.mnt); - - rcu_read_unlock(); - br_read_unlock(vfsmount_lock); - - return 0; - -err_unlock: - spin_unlock(&dentry->d_lock); - rcu_read_unlock(); - br_read_unlock(vfsmount_lock); - return -ECHILD; -} - /** * release_open_intent - free up open intent resources * @nd: pointer to nameidata @@ -606,26 +511,39 @@ do_revalidate(struct dentry *dentry, struct nameidata *nd) return dentry; } -/* - * handle_reval_path - force revalidation of a dentry - * - * In some situations the path walking code will trust dentries without - * revalidating them. This causes problems for filesystems that depend on - * d_revalidate to handle file opens (e.g. NFSv4). When FS_REVAL_DOT is set - * (which indicates that it's possible for the dentry to go stale), force - * a d_revalidate call before proceeding. +/** + * complete_walk - successful completion of path walk + * @nd: pointer nameidata * - * Returns 0 if the revalidation was successful. If the revalidation fails, - * either return the error returned by d_revalidate or -ESTALE if the - * revalidation it just returned 0. If d_revalidate returns 0, we attempt to - * invalidate the dentry. It's up to the caller to handle putting references - * to the path if necessary. + * If we had been in RCU mode, drop out of it and legitimize nd->path. + * Revalidate the final result, unless we'd already done that during + * the path walk or the filesystem doesn't ask for it. Return 0 on + * success, -error on failure. In case of failure caller does not + * need to drop nd->path. */ -static inline int handle_reval_path(struct nameidata *nd) +static int complete_walk(struct nameidata *nd) { struct dentry *dentry = nd->path.dentry; int status; + if (nd->flags & LOOKUP_RCU) { + nd->flags &= ~LOOKUP_RCU; + if (!(nd->flags & LOOKUP_ROOT)) + nd->root.mnt = NULL; + spin_lock(&dentry->d_lock); + if (unlikely(!__d_rcu_to_refcount(dentry, nd->seq))) { + spin_unlock(&dentry->d_lock); + rcu_read_unlock(); + br_read_unlock(vfsmount_lock); + return -ECHILD; + } + BUG_ON(nd->inode != dentry->d_inode); + spin_unlock(&dentry->d_lock); + mntget(nd->path.mnt); + rcu_read_unlock(); + br_read_unlock(vfsmount_lock); + } + if (likely(!(nd->flags & LOOKUP_JUMPED))) return 0; @@ -643,6 +561,7 @@ static inline int handle_reval_path(struct nameidata *nd) if (!status) status = -ESTALE; + path_put(&nd->path); return status; } @@ -697,6 +616,7 @@ static __always_inline void set_root_rcu(struct nameidata *nd) do { seq = read_seqcount_begin(&fs->seq); nd->root = fs->root; + nd->seq = __read_seqcount_begin(&nd->root.dentry->d_seq); } while (read_seqcount_retry(&fs->seq, seq)); } } @@ -999,12 +919,11 @@ static inline bool managed_dentry_might_block(struct dentry *dentry) } /* - * Skip to top of mountpoint pile in rcuwalk mode. We abort the rcu-walk if we - * meet a managed dentry and we're not walking to "..". True is returned to - * continue, false to abort. + * Try to skip to top of mountpoint pile in rcuwalk mode. Fail if + * we meet a managed dentry that would need blocking. */ static bool __follow_mount_rcu(struct nameidata *nd, struct path *path, - struct inode **inode, bool reverse_transit) + struct inode **inode) { for (;;) { struct vfsmount *mounted; @@ -1013,8 +932,7 @@ static bool __follow_mount_rcu(struct nameidata *nd, struct path *path, * that wants to block transit. */ *inode = path->dentry->d_inode; - if (!reverse_transit && - unlikely(managed_dentry_might_block(path->dentry))) + if (unlikely(managed_dentry_might_block(path->dentry))) return false; if (!d_mountpoint(path->dentry)) @@ -1027,16 +945,24 @@ static bool __follow_mount_rcu(struct nameidata *nd, struct path *path, path->dentry = mounted->mnt_root; nd->seq = read_seqcount_begin(&path->dentry->d_seq); } - - if (unlikely(path->dentry->d_flags & DCACHE_NEED_AUTOMOUNT)) - return reverse_transit; return true; } -static int follow_dotdot_rcu(struct nameidata *nd) +static void follow_mount_rcu(struct nameidata *nd) { - struct inode *inode = nd->inode; + while (d_mountpoint(nd->path.dentry)) { + struct vfsmount *mounted; + mounted = __lookup_mnt(nd->path.mnt, nd->path.dentry, 1); + if (!mounted) + break; + nd->path.mnt = mounted; + nd->path.dentry = mounted->mnt_root; + nd->seq = read_seqcount_begin(&nd->path.dentry->d_seq); + } +} +static int follow_dotdot_rcu(struct nameidata *nd) +{ set_root_rcu(nd); while (1) { @@ -1052,7 +978,6 @@ static int follow_dotdot_rcu(struct nameidata *nd) seq = read_seqcount_begin(&parent->d_seq); if (read_seqcount_retry(&old->d_seq, nd->seq)) goto failed; - inode = parent->d_inode; nd->path.dentry = parent; nd->seq = seq; break; @@ -1060,10 +985,9 @@ static int follow_dotdot_rcu(struct nameidata *nd) if (!follow_up_rcu(&nd->path)) break; nd->seq = read_seqcount_begin(&nd->path.dentry->d_seq); - inode = nd->path.dentry->d_inode; } - __follow_mount_rcu(nd, &nd->path, &inode, true); - nd->inode = inode; + follow_mount_rcu(nd); + nd->inode = nd->path.dentry->d_inode; return 0; failed: @@ -1237,16 +1161,14 @@ static int do_lookup(struct nameidata *nd, struct qstr *name, } path->mnt = mnt; path->dentry = dentry; - if (likely(__follow_mount_rcu(nd, path, inode, false))) - return 0; + if (unlikely(!__follow_mount_rcu(nd, path, inode))) + goto unlazy; + if (unlikely(path->dentry->d_flags & DCACHE_NEED_AUTOMOUNT)) + goto unlazy; + return 0; unlazy: - if (dentry) { - if (nameidata_dentry_drop_rcu(nd, dentry)) - return -ECHILD; - } else { - if (nameidata_drop_rcu(nd)) - return -ECHILD; - } + if (unlazy_walk(nd, dentry)) + return -ECHILD; } else { dentry = __d_lookup(parent, name); } @@ -1302,7 +1224,7 @@ static inline int may_lookup(struct nameidata *nd) int err = exec_permission(nd->inode, IPERM_FLAG_RCU); if (err != -ECHILD) return err; - if (nameidata_drop_rcu(nd)) + if (unlazy_walk(nd, NULL)) return -ECHILD; } return exec_permission(nd->inode, 0); @@ -1356,8 +1278,12 @@ static inline int walk_component(struct nameidata *nd, struct path *path, return -ENOENT; } if (unlikely(inode->i_op->follow_link) && follow) { - if (nameidata_dentry_drop_rcu_maybe(nd, path->dentry)) - return -ECHILD; + if (nd->flags & LOOKUP_RCU) { + if (unlikely(unlazy_walk(nd, path->dentry))) { + terminate_walk(nd); + return -ECHILD; + } + } BUG_ON(inode != path->dentry->d_inode); return 1; } @@ -1377,12 +1303,12 @@ static inline int nested_symlink(struct path *path, struct nameidata *nd) { int res; - BUG_ON(nd->depth >= MAX_NESTED_LINKS); if (unlikely(current->link_count >= MAX_NESTED_LINKS)) { path_put_conditional(path, nd); path_put(&nd->path); return -ELOOP; } + BUG_ON(nd->depth >= MAX_NESTED_LINKS); nd->depth++; current->link_count++; @@ -1656,18 +1582,8 @@ static int path_lookupat(int dfd, const char *name, } } - if (nd->flags & LOOKUP_RCU) { - /* went all way through without dropping RCU */ - BUG_ON(err); - if (nameidata_drop_rcu_last(nd)) - err = -ECHILD; - } - - if (!err) { - err = handle_reval_path(nd); - if (err) - path_put(&nd->path); - } + if (!err) + err = complete_walk(nd); if (!err && nd->flags & LOOKUP_DIRECTORY) { if (!nd->inode->i_op->lookup) { @@ -2133,13 +2049,9 @@ static struct file *do_last(struct nameidata *nd, struct path *path, return ERR_PTR(error); /* fallthrough */ case LAST_ROOT: - if (nd->flags & LOOKUP_RCU) { - if (nameidata_drop_rcu_last(nd)) - return ERR_PTR(-ECHILD); - } - error = handle_reval_path(nd); + error = complete_walk(nd); if (error) - goto exit; + return ERR_PTR(error); audit_inode(pathname, nd->path.dentry); if (open_flag & O_CREAT) { error = -EISDIR; @@ -2147,10 +2059,9 @@ static struct file *do_last(struct nameidata *nd, struct path *path, } goto ok; case LAST_BIND: - /* can't be RCU mode here */ - error = handle_reval_path(nd); + error = complete_walk(nd); if (error) - goto exit; + return ERR_PTR(error); audit_inode(pathname, dir); goto ok; } @@ -2169,10 +2080,9 @@ static struct file *do_last(struct nameidata *nd, struct path *path, if (error) /* symlink */ return NULL; /* sayonara */ - if (nd->flags & LOOKUP_RCU) { - if (nameidata_drop_rcu_last(nd)) - return ERR_PTR(-ECHILD); - } + error = complete_walk(nd); + if (error) + return ERR_PTR(-ECHILD); error = -ENOTDIR; if (nd->flags & LOOKUP_DIRECTORY) { @@ -2184,11 +2094,9 @@ static struct file *do_last(struct nameidata *nd, struct path *path, } /* create side of things */ - - if (nd->flags & LOOKUP_RCU) { - if (nameidata_drop_rcu_last(nd)) - return ERR_PTR(-ECHILD); - } + error = complete_walk(nd); + if (error) + return ERR_PTR(error); audit_inode(pathname, dir); error = -EISDIR; @@ -2628,10 +2536,10 @@ SYSCALL_DEFINE2(mkdir, const char __user *, pathname, int, mode) } /* - * We try to drop the dentry early: we should have - * a usage count of 2 if we're the only user of this - * dentry, and if that is true (possibly after pruning - * the dcache), then we drop the dentry now. + * The dentry_unhash() helper will try to drop the dentry early: we + * should have a usage count of 2 if we're the only user of this + * dentry, and if that is true (possibly after pruning the dcache), + * then we drop the dentry now. * * A low-level filesystem can, if it choses, legally * do a @@ -2644,10 +2552,9 @@ SYSCALL_DEFINE2(mkdir, const char __user *, pathname, int, mode) */ void dentry_unhash(struct dentry *dentry) { - dget(dentry); shrink_dcache_parent(dentry); spin_lock(&dentry->d_lock); - if (dentry->d_count == 2) + if (dentry->d_count == 1) __d_drop(dentry); spin_unlock(&dentry->d_lock); } @@ -2663,25 +2570,26 @@ int vfs_rmdir(struct inode *dir, struct dentry *dentry) return -EPERM; mutex_lock(&dentry->d_inode->i_mutex); - dentry_unhash(dentry); + + error = -EBUSY; if (d_mountpoint(dentry)) - error = -EBUSY; - else { - error = security_inode_rmdir(dir, dentry); - if (!error) { - error = dir->i_op->rmdir(dir, dentry); - if (!error) { - dentry->d_inode->i_flags |= S_DEAD; - dont_mount(dentry); - } - } - } + goto out; + + error = security_inode_rmdir(dir, dentry); + if (error) + goto out; + + error = dir->i_op->rmdir(dir, dentry); + if (error) + goto out; + + dentry->d_inode->i_flags |= S_DEAD; + dont_mount(dentry); + +out: mutex_unlock(&dentry->d_inode->i_mutex); - if (!error) { + if (!error) d_delete(dentry); - } - dput(dentry); - return error; } @@ -3052,12 +2960,7 @@ SYSCALL_DEFINE2(link, const char __user *, oldname, const char __user *, newname * HOWEVER, it relies on the assumption that any object with ->lookup() * has no more than 1 dentry. If "hybrid" objects will ever appear, * we'd better make sure that there's no link(2) for them. - * d) some filesystems don't support opened-but-unlinked directories, - * either because of layout or because they are not ready to deal with - * all cases correctly. The latter will be fixed (taking this sort of - * stuff into VFS), but the former is not going away. Solution: the same - * trick as in rmdir(). - * e) conversion from fhandle to dentry may come in the wrong moment - when + * d) conversion from fhandle to dentry may come in the wrong moment - when * we are removing the target. Solution: we will have to grab ->i_mutex * in the fhandle_to_dentry code. [FIXME - current nfsfh.c relies on * ->i_mutex on parents, which works but leads to some truly excessive @@ -3067,7 +2970,7 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { int error = 0; - struct inode *target; + struct inode *target = new_dentry->d_inode; /* * If we are going to change the parent - check write permissions, @@ -3083,26 +2986,24 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, if (error) return error; - target = new_dentry->d_inode; if (target) mutex_lock(&target->i_mutex); - if (d_mountpoint(old_dentry)||d_mountpoint(new_dentry)) - error = -EBUSY; - else { - if (target) - dentry_unhash(new_dentry); - error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry); - } + + error = -EBUSY; + if (d_mountpoint(old_dentry) || d_mountpoint(new_dentry)) + goto out; + + error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry); + if (error) + goto out; + if (target) { - if (!error) { - target->i_flags |= S_DEAD; - dont_mount(new_dentry); - } - mutex_unlock(&target->i_mutex); - if (d_unhashed(new_dentry)) - d_rehash(new_dentry); - dput(new_dentry); + target->i_flags |= S_DEAD; + dont_mount(new_dentry); } +out: + if (target) + mutex_unlock(&target->i_mutex); if (!error) if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE)) d_move(old_dentry,new_dentry); @@ -3112,7 +3013,7 @@ static int vfs_rename_dir(struct inode *old_dir, struct dentry *old_dentry, static int vfs_rename_other(struct inode *old_dir, struct dentry *old_dentry, struct inode *new_dir, struct dentry *new_dentry) { - struct inode *target; + struct inode *target = new_dentry->d_inode; int error; error = security_inode_rename(old_dir, old_dentry, new_dir, new_dentry); @@ -3120,19 +3021,22 @@ static int vfs_rename_other(struct inode *old_dir, struct dentry *old_dentry, return error; dget(new_dentry); - target = new_dentry->d_inode; if (target) mutex_lock(&target->i_mutex); + + error = -EBUSY; if (d_mountpoint(old_dentry)||d_mountpoint(new_dentry)) - error = -EBUSY; - else - error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry); - if (!error) { - if (target) - dont_mount(new_dentry); - if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE)) - d_move(old_dentry, new_dentry); - } + goto out; + + error = old_dir->i_op->rename(old_dir, old_dentry, new_dir, new_dentry); + if (error) + goto out; + + if (target) + dont_mount(new_dentry); + if (!(old_dir->i_sb->s_type->fs_flags & FS_RENAME_DOES_D_MOVE)) + d_move(old_dentry, new_dentry); +out: if (target) mutex_unlock(&target->i_mutex); dput(new_dentry); diff --git a/fs/namespace.c b/fs/namespace.c index 7dba2ed03429..fe59bd145d21 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1030,18 +1030,6 @@ const struct seq_operations mounts_op = { .show = show_vfsmnt }; -static int uuid_is_nil(u8 *uuid) -{ - int i; - u8 *cp = (u8 *)uuid; - - for (i = 0; i < 16; i++) { - if (*cp++) - return 0; - } - return 1; -} - static int show_mountinfo(struct seq_file *m, void *v) { struct proc_mounts *p = m->private; @@ -1085,10 +1073,6 @@ static int show_mountinfo(struct seq_file *m, void *v) if (IS_MNT_UNBINDABLE(mnt)) seq_puts(m, " unbindable"); - if (!uuid_is_nil(mnt->mnt_sb->s_uuid)) - /* print the uuid */ - seq_printf(m, " uuid:%pU", mnt->mnt_sb->s_uuid); - /* Filesystem specific data */ seq_puts(m, " - "); show_type(m, sb); @@ -1711,7 +1695,7 @@ static int graft_tree(struct vfsmount *mnt, struct path *path) static int flags_to_propagation_type(int flags) { - int type = flags & ~MS_REC; + int type = flags & ~(MS_REC | MS_SILENT); /* Fail if any non-propagation flags are set */ if (type & ~(MS_SHARED | MS_PRIVATE | MS_SLAVE | MS_UNBINDABLE)) diff --git a/fs/ncpfs/dir.c b/fs/ncpfs/dir.c index f6946bb5cb55..9c51f621e901 100644 --- a/fs/ncpfs/dir.c +++ b/fs/ncpfs/dir.c @@ -1033,6 +1033,11 @@ static int ncp_rmdir(struct inode *dir, struct dentry *dentry) DPRINTK("ncp_rmdir: removing %s/%s\n", dentry->d_parent->d_name.name, dentry->d_name.name); + /* + * fail with EBUSY if there are still references to this + * directory. + */ + dentry_unhash(dentry); error = -EBUSY; if (!d_unhashed(dentry)) goto out; @@ -1139,6 +1144,17 @@ static int ncp_rename(struct inode *old_dir, struct dentry *old_dentry, old_dentry->d_parent->d_name.name, old_dentry->d_name.name, new_dentry->d_parent->d_name.name, new_dentry->d_name.name); + if (new_dentry->d_inode && S_ISDIR(new_dentry->d_inode->i_mode)) { + /* + * fail with EBUSY if there are still references to this + * directory. + */ + dentry_unhash(new_dentry); + error = -EBUSY; + if (!d_unhashed(new_dentry)) + goto out; + } + ncp_age_dentry(server, old_dentry); ncp_age_dentry(server, new_dentry); diff --git a/fs/ncpfs/inode.c b/fs/ncpfs/inode.c index 00a1d1c3d3a4..202f370526a7 100644 --- a/fs/ncpfs/inode.c +++ b/fs/ncpfs/inode.c @@ -461,7 +461,7 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) #endif struct ncp_entry_info finfo; - data.wdog_pid = NULL; + memset(&data, 0, sizeof(data)); server = kzalloc(sizeof(struct ncp_server), GFP_KERNEL); if (!server) return -ENOMEM; @@ -496,7 +496,6 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) struct ncp_mount_data_v4* md = (struct ncp_mount_data_v4*)raw_data; data.flags = md->flags; - data.int_flags = 0; data.mounted_uid = md->mounted_uid; data.wdog_pid = find_get_pid(md->wdog_pid); data.ncp_fd = md->ncp_fd; @@ -507,7 +506,6 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) data.file_mode = md->file_mode; data.dir_mode = md->dir_mode; data.info_fd = -1; - data.mounted_vol[0] = 0; } break; default: @@ -596,7 +594,7 @@ static int ncp_fill_super(struct super_block *sb, void *raw_data, int silent) /* server->priv.data = NULL; */ server->m = data; - /* Althought anything producing this is buggy, it happens + /* Although anything producing this is buggy, it happens now because of PATH_MAX changes.. */ if (server->m.time_out < 1) { server->m.time_out = 10; diff --git a/fs/ncpfs/mmap.c b/fs/ncpfs/mmap.c index a7c07b44b100..e5d71b27a5b0 100644 --- a/fs/ncpfs/mmap.c +++ b/fs/ncpfs/mmap.c @@ -16,6 +16,7 @@ #include <linux/mman.h> #include <linux/string.h> #include <linux/fcntl.h> +#include <linux/memcontrol.h> #include <asm/uaccess.h> #include <asm/system.h> @@ -92,6 +93,7 @@ static int ncp_file_mmap_fault(struct vm_area_struct *area, * -- wli */ count_vm_event(PGMAJFAULT); + mem_cgroup_count_vm_event(area->vm_mm, PGMAJFAULT); return VM_FAULT_MAJOR; } diff --git a/fs/nfs/callback_xdr.c b/fs/nfs/callback_xdr.c index 14e0f9371d14..00ecf62ce7c1 100644 --- a/fs/nfs/callback_xdr.c +++ b/fs/nfs/callback_xdr.c @@ -241,7 +241,7 @@ static __be32 decode_layoutrecall_args(struct svc_rqst *rqstp, args->cbl_layout_type = ntohl(*p++); /* Depite the spec's xdr, iomode really belongs in the FILE switch, - * as it is unuseable and ignored with the other types. + * as it is unusable and ignored with the other types. */ iomode = ntohl(*p++); args->cbl_layoutchanged = ntohl(*p++); diff --git a/fs/nfs/delegation.c b/fs/nfs/delegation.c index bbbc6bf5cb2e..dd25c2aec375 100644 --- a/fs/nfs/delegation.c +++ b/fs/nfs/delegation.c @@ -21,25 +21,13 @@ #include "delegation.h" #include "internal.h" -static void nfs_do_free_delegation(struct nfs_delegation *delegation) -{ - kfree(delegation); -} - -static void nfs_free_delegation_callback(struct rcu_head *head) -{ - struct nfs_delegation *delegation = container_of(head, struct nfs_delegation, rcu); - - nfs_do_free_delegation(delegation); -} - static void nfs_free_delegation(struct nfs_delegation *delegation) { if (delegation->cred) { put_rpccred(delegation->cred); delegation->cred = NULL; } - call_rcu(&delegation->rcu, nfs_free_delegation_callback); + kfree_rcu(delegation, rcu); } /** diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index 7237672216c8..424e47773a84 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -2042,11 +2042,14 @@ static void nfs_access_free_list(struct list_head *head) } } -int nfs_access_cache_shrinker(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask) +int nfs_access_cache_shrinker(struct shrinker *shrink, + struct shrink_control *sc) { LIST_HEAD(head); struct nfs_inode *nfsi, *next; struct nfs_access_entry *cache; + int nr_to_scan = sc->nr_to_scan; + gfp_t gfp_mask = sc->gfp_mask; if ((gfp_mask & GFP_KERNEL) != GFP_KERNEL) return (nr_to_scan == 0) ? 0 : -1; diff --git a/fs/nfs/file.c b/fs/nfs/file.c index 3ac5bd695e5e..2f093ed16980 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -301,7 +301,7 @@ nfs_file_mmap(struct file * file, struct vm_area_struct * vma) * disk, but it retrieves and clears ctx->error after synching, despite * the two being set at the same time in nfs_context_set_write_error(). * This is because the former is used to notify the _next_ call to - * nfs_file_write() that a write error occured, and hence cause it to + * nfs_file_write() that a write error occurred, and hence cause it to * fall back to doing a synchronous write. */ static int diff --git a/fs/nfs/inode.c b/fs/nfs/inode.c index 57bb31ad7a5e..873c6fa8bc3b 100644 --- a/fs/nfs/inode.c +++ b/fs/nfs/inode.c @@ -1298,8 +1298,12 @@ static int nfs_update_inode(struct inode *inode, struct nfs_fattr *fattr) i_size_write(inode, new_isize); invalid |= NFS_INO_INVALID_ATTR|NFS_INO_INVALID_DATA; } - dprintk("NFS: isize change on server for file %s/%ld\n", - inode->i_sb->s_id, inode->i_ino); + dprintk("NFS: isize change on server for file %s/%ld " + "(%Ld to %Ld)\n", + inode->i_sb->s_id, + inode->i_ino, + (long long)cur_isize, + (long long)new_isize); } } else invalid |= save_cache_validity & (NFS_INO_INVALID_ATTR diff --git a/fs/nfs/internal.h b/fs/nfs/internal.h index ce118ce885dd..2df6ca7b5898 100644 --- a/fs/nfs/internal.h +++ b/fs/nfs/internal.h @@ -234,7 +234,7 @@ extern int nfs_init_client(struct nfs_client *clp, /* dir.c */ extern int nfs_access_cache_shrinker(struct shrinker *shrink, - int nr_to_scan, gfp_t gfp_mask); + struct shrink_control *sc); /* inode.c */ extern struct workqueue_struct *nfsiod_workqueue; diff --git a/fs/nfs/namespace.c b/fs/nfs/namespace.c index 9166fcb66da2..1f063bacd285 100644 --- a/fs/nfs/namespace.c +++ b/fs/nfs/namespace.c @@ -119,7 +119,7 @@ Elong: } #ifdef CONFIG_NFS_V4 -static rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors, struct inode *inode) +static rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors) { struct gss_api_mech *mech; struct xdr_netobj oid; @@ -148,67 +148,64 @@ static rpc_authflavor_t nfs_find_best_sec(struct nfs4_secinfo_flavors *flavors, return pseudoflavor; } -static rpc_authflavor_t nfs_negotiate_security(const struct dentry *parent, const struct dentry *dentry) +static int nfs_negotiate_security(const struct dentry *parent, + const struct dentry *dentry, + rpc_authflavor_t *flavor) { - int status = 0; struct page *page; struct nfs4_secinfo_flavors *flavors; int (*secinfo)(struct inode *, const struct qstr *, struct nfs4_secinfo_flavors *); - rpc_authflavor_t flavor = RPC_AUTH_UNIX; + int ret = -EPERM; secinfo = NFS_PROTO(parent->d_inode)->secinfo; if (secinfo != NULL) { page = alloc_page(GFP_KERNEL); if (!page) { - status = -ENOMEM; + ret = -ENOMEM; goto out; } flavors = page_address(page); - status = secinfo(parent->d_inode, &dentry->d_name, flavors); - flavor = nfs_find_best_sec(flavors, dentry->d_inode); + ret = secinfo(parent->d_inode, &dentry->d_name, flavors); + *flavor = nfs_find_best_sec(flavors); put_page(page); } - return flavor; - out: - status = -ENOMEM; - return status; + return ret; } -static rpc_authflavor_t nfs_lookup_with_sec(struct nfs_server *server, struct dentry *parent, - struct dentry *dentry, struct path *path, - struct nfs_fh *fh, struct nfs_fattr *fattr) +static int nfs_lookup_with_sec(struct nfs_server *server, struct dentry *parent, + struct dentry *dentry, struct path *path, + struct nfs_fh *fh, struct nfs_fattr *fattr, + rpc_authflavor_t *flavor) { - rpc_authflavor_t flavor; struct rpc_clnt *clone; struct rpc_auth *auth; int err; - flavor = nfs_negotiate_security(parent, path->dentry); - if (flavor < 0) + err = nfs_negotiate_security(parent, path->dentry, flavor); + if (err < 0) goto out; clone = rpc_clone_client(server->client); - auth = rpcauth_create(flavor, clone); + auth = rpcauth_create(*flavor, clone); if (!auth) { - flavor = -EIO; + err = -EIO; goto out_shutdown; } err = server->nfs_client->rpc_ops->lookup(clone, parent->d_inode, &path->dentry->d_name, fh, fattr); - if (err < 0) - flavor = err; out_shutdown: rpc_shutdown_client(clone); out: - return flavor; + return err; } #else /* CONFIG_NFS_V4 */ -static inline rpc_authflavor_t nfs_lookup_with_sec(struct nfs_server *server, - struct dentry *parent, struct dentry *dentry, - struct path *path, struct nfs_fh *fh, - struct nfs_fattr *fattr) +static inline int nfs_lookup_with_sec(struct nfs_server *server, + struct dentry *parent, struct dentry *dentry, + struct path *path, struct nfs_fh *fh, + struct nfs_fattr *fattr, + rpc_authflavor_t *flavor) { return -EPERM; } @@ -234,7 +231,7 @@ struct vfsmount *nfs_d_automount(struct path *path) struct nfs_fh *fh = NULL; struct nfs_fattr *fattr = NULL; int err; - rpc_authflavor_t flavor = 1; + rpc_authflavor_t flavor = RPC_AUTH_UNIX; dprintk("--> nfs_d_automount()\n"); @@ -255,13 +252,8 @@ struct vfsmount *nfs_d_automount(struct path *path) err = server->nfs_client->rpc_ops->lookup(server->client, parent->d_inode, &path->dentry->d_name, fh, fattr); - if (err == -EPERM) { - flavor = nfs_lookup_with_sec(server, parent, path->dentry, path, fh, fattr); - if (flavor < 0) - err = flavor; - else - err = 0; - } + if (err == -EPERM && NFS_PROTO(parent->d_inode)->secinfo != NULL) + err = nfs_lookup_with_sec(server, parent, path->dentry, path, fh, fattr, &flavor); dput(parent); if (err != 0) { mnt = ERR_PTR(err); diff --git a/fs/nfs/nfs4_fs.h b/fs/nfs/nfs4_fs.h index e1c261ddd65d..c4a69833dd0d 100644 --- a/fs/nfs/nfs4_fs.h +++ b/fs/nfs/nfs4_fs.h @@ -47,6 +47,7 @@ enum nfs4_client_state { NFS4CLNT_LAYOUTRECALL, NFS4CLNT_SESSION_RESET, NFS4CLNT_RECALL_SLOT, + NFS4CLNT_LEASE_CONFIRM, }; enum nfs4_session_state { diff --git a/fs/nfs/nfs4filelayout.c b/fs/nfs/nfs4filelayout.c index 6f8192f4cfc7..be79dc9f386d 100644 --- a/fs/nfs/nfs4filelayout.c +++ b/fs/nfs/nfs4filelayout.c @@ -117,6 +117,8 @@ static int filelayout_async_handle_error(struct rpc_task *task, case -EKEYEXPIRED: rpc_delay(task, FILELAYOUT_POLL_RETRY_MAX); break; + case -NFS4ERR_RETRY_UNCACHED_REP: + break; default: dprintk("%s DS error. Retry through MDS %d\n", __func__, task->tk_status); @@ -416,7 +418,8 @@ static int filelayout_check_layout(struct pnfs_layout_hdr *lo, struct nfs4_filelayout_segment *fl, struct nfs4_layoutget_res *lgr, - struct nfs4_deviceid *id) + struct nfs4_deviceid *id, + gfp_t gfp_flags) { struct nfs4_file_layout_dsaddr *dsaddr; int status = -EINVAL; @@ -439,7 +442,7 @@ filelayout_check_layout(struct pnfs_layout_hdr *lo, /* find and reference the deviceid */ dsaddr = nfs4_fl_find_get_deviceid(id); if (dsaddr == NULL) { - dsaddr = get_device_info(lo->plh_inode, id); + dsaddr = get_device_info(lo->plh_inode, id, gfp_flags); if (dsaddr == NULL) goto out; } @@ -500,7 +503,8 @@ static int filelayout_decode_layout(struct pnfs_layout_hdr *flo, struct nfs4_filelayout_segment *fl, struct nfs4_layoutget_res *lgr, - struct nfs4_deviceid *id) + struct nfs4_deviceid *id, + gfp_t gfp_flags) { struct xdr_stream stream; struct xdr_buf buf = { @@ -516,7 +520,7 @@ filelayout_decode_layout(struct pnfs_layout_hdr *flo, dprintk("%s: set_layout_map Begin\n", __func__); - scratch = alloc_page(GFP_KERNEL); + scratch = alloc_page(gfp_flags); if (!scratch) return -ENOMEM; @@ -554,13 +558,13 @@ filelayout_decode_layout(struct pnfs_layout_hdr *flo, goto out_err; fl->fh_array = kzalloc(fl->num_fh * sizeof(struct nfs_fh *), - GFP_KERNEL); + gfp_flags); if (!fl->fh_array) goto out_err; for (i = 0; i < fl->num_fh; i++) { /* Do we want to use a mempool here? */ - fl->fh_array[i] = kmalloc(sizeof(struct nfs_fh), GFP_KERNEL); + fl->fh_array[i] = kmalloc(sizeof(struct nfs_fh), gfp_flags); if (!fl->fh_array[i]) goto out_err_free; @@ -605,19 +609,20 @@ filelayout_free_lseg(struct pnfs_layout_segment *lseg) static struct pnfs_layout_segment * filelayout_alloc_lseg(struct pnfs_layout_hdr *layoutid, - struct nfs4_layoutget_res *lgr) + struct nfs4_layoutget_res *lgr, + gfp_t gfp_flags) { struct nfs4_filelayout_segment *fl; int rc; struct nfs4_deviceid id; dprintk("--> %s\n", __func__); - fl = kzalloc(sizeof(*fl), GFP_KERNEL); + fl = kzalloc(sizeof(*fl), gfp_flags); if (!fl) return NULL; - rc = filelayout_decode_layout(layoutid, fl, lgr, &id); - if (rc != 0 || filelayout_check_layout(layoutid, fl, lgr, &id)) { + rc = filelayout_decode_layout(layoutid, fl, lgr, &id, gfp_flags); + if (rc != 0 || filelayout_check_layout(layoutid, fl, lgr, &id, gfp_flags)) { _filelayout_free_lseg(fl); return NULL; } @@ -633,7 +638,7 @@ filelayout_alloc_lseg(struct pnfs_layout_hdr *layoutid, int size = (fl->stripe_type == STRIPE_SPARSE) ? fl->dsaddr->ds_num : fl->dsaddr->stripe_count; - fl->commit_buckets = kcalloc(size, sizeof(struct list_head), GFP_KERNEL); + fl->commit_buckets = kcalloc(size, sizeof(struct list_head), gfp_flags); if (!fl->commit_buckets) { filelayout_free_lseg(&fl->generic_hdr); return NULL; diff --git a/fs/nfs/nfs4filelayout.h b/fs/nfs/nfs4filelayout.h index 085a354e0f08..2b461d77b43a 100644 --- a/fs/nfs/nfs4filelayout.h +++ b/fs/nfs/nfs4filelayout.h @@ -33,7 +33,7 @@ #include "pnfs.h" /* - * Field testing shows we need to support upto 4096 stripe indices. + * Field testing shows we need to support up to 4096 stripe indices. * We store each index as a u8 (u32 on the wire) to keep the memory footprint * reasonable. This in turn means we support a maximum of 256 * RFC 5661 multipath_list4 structures. @@ -104,6 +104,6 @@ extern struct nfs4_file_layout_dsaddr * nfs4_fl_find_get_deviceid(struct nfs4_deviceid *dev_id); extern void nfs4_fl_put_deviceid(struct nfs4_file_layout_dsaddr *dsaddr); struct nfs4_file_layout_dsaddr * -get_device_info(struct inode *inode, struct nfs4_deviceid *dev_id); +get_device_info(struct inode *inode, struct nfs4_deviceid *dev_id, gfp_t gfp_flags); #endif /* FS_NFS_NFS4FILELAYOUT_H */ diff --git a/fs/nfs/nfs4filelayoutdev.c b/fs/nfs/nfs4filelayoutdev.c index de5350f2b249..db07c7af1395 100644 --- a/fs/nfs/nfs4filelayoutdev.c +++ b/fs/nfs/nfs4filelayoutdev.c @@ -225,11 +225,11 @@ nfs4_fl_free_deviceid(struct nfs4_file_layout_dsaddr *dsaddr) } static struct nfs4_pnfs_ds * -nfs4_pnfs_ds_add(struct inode *inode, u32 ip_addr, u32 port) +nfs4_pnfs_ds_add(struct inode *inode, u32 ip_addr, u32 port, gfp_t gfp_flags) { struct nfs4_pnfs_ds *tmp_ds, *ds; - ds = kzalloc(sizeof(*tmp_ds), GFP_KERNEL); + ds = kzalloc(sizeof(*tmp_ds), gfp_flags); if (!ds) goto out; @@ -261,7 +261,7 @@ out: * Currently only support ipv4, and one multi-path address. */ static struct nfs4_pnfs_ds * -decode_and_add_ds(struct xdr_stream *streamp, struct inode *inode) +decode_and_add_ds(struct xdr_stream *streamp, struct inode *inode, gfp_t gfp_flags) { struct nfs4_pnfs_ds *ds = NULL; char *buf; @@ -303,7 +303,7 @@ decode_and_add_ds(struct xdr_stream *streamp, struct inode *inode) rlen); goto out_err; } - buf = kmalloc(rlen + 1, GFP_KERNEL); + buf = kmalloc(rlen + 1, gfp_flags); if (!buf) { dprintk("%s: Not enough memory\n", __func__); goto out_err; @@ -333,7 +333,7 @@ decode_and_add_ds(struct xdr_stream *streamp, struct inode *inode) sscanf(pstr, "-%d-%d", &tmp[0], &tmp[1]); port = htons((tmp[0] << 8) | (tmp[1])); - ds = nfs4_pnfs_ds_add(inode, ip_addr, port); + ds = nfs4_pnfs_ds_add(inode, ip_addr, port, gfp_flags); dprintk("%s: Decoded address and port %s\n", __func__, buf); out_free: kfree(buf); @@ -343,7 +343,7 @@ out_err: /* Decode opaque device data and return the result */ static struct nfs4_file_layout_dsaddr* -decode_device(struct inode *ino, struct pnfs_device *pdev) +decode_device(struct inode *ino, struct pnfs_device *pdev, gfp_t gfp_flags) { int i; u32 cnt, num; @@ -362,7 +362,7 @@ decode_device(struct inode *ino, struct pnfs_device *pdev) struct page *scratch; /* set up xdr stream */ - scratch = alloc_page(GFP_KERNEL); + scratch = alloc_page(gfp_flags); if (!scratch) goto out_err; @@ -384,7 +384,7 @@ decode_device(struct inode *ino, struct pnfs_device *pdev) } /* read stripe indices */ - stripe_indices = kcalloc(cnt, sizeof(u8), GFP_KERNEL); + stripe_indices = kcalloc(cnt, sizeof(u8), gfp_flags); if (!stripe_indices) goto out_err_free_scratch; @@ -423,7 +423,7 @@ decode_device(struct inode *ino, struct pnfs_device *pdev) dsaddr = kzalloc(sizeof(*dsaddr) + (sizeof(struct nfs4_pnfs_ds *) * (num - 1)), - GFP_KERNEL); + gfp_flags); if (!dsaddr) goto out_err_free_stripe_indices; @@ -452,7 +452,7 @@ decode_device(struct inode *ino, struct pnfs_device *pdev) for (j = 0; j < mp_count; j++) { if (j == 0) { dsaddr->ds_list[i] = decode_and_add_ds(&stream, - ino); + ino, gfp_flags); if (dsaddr->ds_list[i] == NULL) goto out_err_free_deviceid; } else { @@ -503,12 +503,12 @@ out_err: * available devices. */ static struct nfs4_file_layout_dsaddr * -decode_and_add_device(struct inode *inode, struct pnfs_device *dev) +decode_and_add_device(struct inode *inode, struct pnfs_device *dev, gfp_t gfp_flags) { struct nfs4_file_layout_dsaddr *d, *new; long hash; - new = decode_device(inode, dev); + new = decode_device(inode, dev, gfp_flags); if (!new) { printk(KERN_WARNING "%s: Could not decode or add device\n", __func__); @@ -537,7 +537,7 @@ decode_and_add_device(struct inode *inode, struct pnfs_device *dev) * of available devices, and return it. */ struct nfs4_file_layout_dsaddr * -get_device_info(struct inode *inode, struct nfs4_deviceid *dev_id) +get_device_info(struct inode *inode, struct nfs4_deviceid *dev_id, gfp_t gfp_flags) { struct pnfs_device *pdev = NULL; u32 max_resp_sz; @@ -556,17 +556,17 @@ get_device_info(struct inode *inode, struct nfs4_deviceid *dev_id) dprintk("%s inode %p max_resp_sz %u max_pages %d\n", __func__, inode, max_resp_sz, max_pages); - pdev = kzalloc(sizeof(struct pnfs_device), GFP_KERNEL); + pdev = kzalloc(sizeof(struct pnfs_device), gfp_flags); if (pdev == NULL) return NULL; - pages = kzalloc(max_pages * sizeof(struct page *), GFP_KERNEL); + pages = kzalloc(max_pages * sizeof(struct page *), gfp_flags); if (pages == NULL) { kfree(pdev); return NULL; } for (i = 0; i < max_pages; i++) { - pages[i] = alloc_page(GFP_KERNEL); + pages[i] = alloc_page(gfp_flags); if (!pages[i]) goto out_free; } @@ -587,7 +587,7 @@ get_device_info(struct inode *inode, struct nfs4_deviceid *dev_id) * Found new device, need to decode it and then add it to the * list of known devices for this mountpoint. */ - dsaddr = decode_and_add_device(inode, pdev); + dsaddr = decode_and_add_device(inode, pdev, gfp_flags); out_free: for (i = 0; i < max_pages; i++) __free_page(pages[i]); diff --git a/fs/nfs/nfs4proc.c b/fs/nfs/nfs4proc.c index dfd1e6d7e6c3..d0e15dba7a5a 100644 --- a/fs/nfs/nfs4proc.c +++ b/fs/nfs/nfs4proc.c @@ -46,6 +46,7 @@ #include <linux/nfs4.h> #include <linux/nfs_fs.h> #include <linux/nfs_page.h> +#include <linux/nfs_mount.h> #include <linux/namei.h> #include <linux/mount.h> #include <linux/module.h> @@ -266,9 +267,11 @@ static int nfs4_handle_exception(struct nfs_server *server, int errorcode, struc break; nfs4_schedule_stateid_recovery(server, state); goto wait_on_recovery; + case -NFS4ERR_EXPIRED: + if (state != NULL) + nfs4_schedule_stateid_recovery(server, state); case -NFS4ERR_STALE_STATEID: case -NFS4ERR_STALE_CLIENTID: - case -NFS4ERR_EXPIRED: nfs4_schedule_lease_recovery(clp); goto wait_on_recovery; #if defined(CONFIG_NFS_V4_1) @@ -299,6 +302,7 @@ static int nfs4_handle_exception(struct nfs_server *server, int errorcode, struc ret = nfs4_delay(server->client, &exception->timeout); if (ret != 0) break; + case -NFS4ERR_RETRY_UNCACHED_REP: case -NFS4ERR_OLD_STATEID: exception->retry = 1; break; @@ -443,8 +447,8 @@ static int nfs41_sequence_done(struct rpc_task *task, struct nfs4_sequence_res * if (res->sr_status == 1) res->sr_status = NFS_OK; - /* -ERESTARTSYS can result in skipping nfs41_sequence_setup */ - if (!res->sr_slot) + /* don't increment the sequence number if the task wasn't sent */ + if (!RPC_WAS_SENT(task)) goto out; /* Check the SEQUENCE operation status */ @@ -2185,9 +2189,14 @@ static int nfs4_lookup_root(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs4_exception exception = { }; int err; do { - err = nfs4_handle_exception(server, - _nfs4_lookup_root(server, fhandle, info), - &exception); + err = _nfs4_lookup_root(server, fhandle, info); + switch (err) { + case 0: + case -NFS4ERR_WRONGSEC: + break; + default: + err = nfs4_handle_exception(server, err, &exception); + } } while (exception.retry); return err; } @@ -2204,31 +2213,51 @@ static int nfs4_lookup_root_sec(struct nfs_server *server, struct nfs_fh *fhandl goto out; } ret = nfs4_lookup_root(server, fhandle, info); - if (ret < 0) - ret = -EAGAIN; out: return ret; } -/* - * get the file handle for the "/" directory on the server - */ -static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, +static int nfs4_find_root_sec(struct nfs_server *server, struct nfs_fh *fhandle, struct nfs_fsinfo *info) { int i, len, status = 0; - rpc_authflavor_t flav_array[NFS_MAX_SECFLAVORS + 2]; + rpc_authflavor_t flav_array[NFS_MAX_SECFLAVORS]; - flav_array[0] = RPC_AUTH_UNIX; - len = gss_mech_list_pseudoflavors(&flav_array[1]); - flav_array[1+len] = RPC_AUTH_NULL; - len += 2; + len = gss_mech_list_pseudoflavors(&flav_array[0]); + flav_array[len] = RPC_AUTH_NULL; + len += 1; for (i = 0; i < len; i++) { status = nfs4_lookup_root_sec(server, fhandle, info, flav_array[i]); - if (status == 0) - break; + if (status == -NFS4ERR_WRONGSEC || status == -EACCES) + continue; + break; } + /* + * -EACCESS could mean that the user doesn't have correct permissions + * to access the mount. It could also mean that we tried to mount + * with a gss auth flavor, but rpc.gssd isn't running. Either way, + * existing mount programs don't handle -EACCES very well so it should + * be mapped to -EPERM instead. + */ + if (status == -EACCES) + status = -EPERM; + return status; +} + +/* + * get the file handle for the "/" directory on the server + */ +static int nfs4_proc_get_root(struct nfs_server *server, struct nfs_fh *fhandle, + struct nfs_fsinfo *info) +{ + int status = nfs4_lookup_root(server, fhandle, info); + if ((status == -NFS4ERR_WRONGSEC) && !(server->flags & NFS_MOUNT_SECFLAVOUR)) + /* + * A status of -NFS4ERR_WRONGSEC will be mapped to -EPERM + * by nfs4_map_errors() as this function exits. + */ + status = nfs4_find_root_sec(server, fhandle, info); if (status == 0) status = nfs4_server_capabilities(server, fhandle); if (status == 0) @@ -3643,9 +3672,11 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, break; nfs4_schedule_stateid_recovery(server, state); goto wait_on_recovery; + case -NFS4ERR_EXPIRED: + if (state != NULL) + nfs4_schedule_stateid_recovery(server, state); case -NFS4ERR_STALE_STATEID: case -NFS4ERR_STALE_CLIENTID: - case -NFS4ERR_EXPIRED: nfs4_schedule_lease_recovery(clp); goto wait_on_recovery; #if defined(CONFIG_NFS_V4_1) @@ -3669,6 +3700,7 @@ nfs4_async_handle_error(struct rpc_task *task, const struct nfs_server *server, rpc_delay(task, NFS4_POLL_RETRY_MAX); task->tk_status = 0; return -EAGAIN; + case -NFS4ERR_RETRY_UNCACHED_REP: case -NFS4ERR_OLD_STATEID: task->tk_status = 0; return -EAGAIN; @@ -3725,21 +3757,20 @@ int nfs4_proc_setclientid(struct nfs_client *clp, u32 program, sizeof(setclientid.sc_uaddr), "%s.%u.%u", clp->cl_ipaddr, port >> 8, port & 255); - status = rpc_call_sync(clp->cl_rpcclient, &msg, 0); + status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); if (status != -NFS4ERR_CLID_INUSE) break; - if (signalled()) + if (loop != 0) { + ++clp->cl_id_uniquifier; break; - if (loop++ & 1) - ssleep(clp->cl_lease_time / HZ + 1); - else - if (++clp->cl_id_uniquifier == 0) - break; + } + ++loop; + ssleep(clp->cl_lease_time / HZ + 1); } return status; } -static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, +int nfs4_proc_setclientid_confirm(struct nfs_client *clp, struct nfs4_setclientid_res *arg, struct rpc_cred *cred) { @@ -3754,7 +3785,7 @@ static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, int status; now = jiffies; - status = rpc_call_sync(clp->cl_rpcclient, &msg, 0); + status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); if (status == 0) { spin_lock(&clp->cl_lock); clp->cl_lease_time = fsinfo.lease_time * HZ; @@ -3764,26 +3795,6 @@ static int _nfs4_proc_setclientid_confirm(struct nfs_client *clp, return status; } -int nfs4_proc_setclientid_confirm(struct nfs_client *clp, - struct nfs4_setclientid_res *arg, - struct rpc_cred *cred) -{ - long timeout = 0; - int err; - do { - err = _nfs4_proc_setclientid_confirm(clp, arg, cred); - switch (err) { - case 0: - return err; - case -NFS4ERR_RESOURCE: - /* The IBM lawyers misread another document! */ - case -NFS4ERR_DELAY: - err = nfs4_delay(clp->cl_rpcclient, &timeout); - } - } while (err == 0); - return err; -} - struct nfs4_delegreturndata { struct nfs4_delegreturnargs args; struct nfs4_delegreturnres res; @@ -4536,6 +4547,7 @@ int nfs4_lock_delegation_recall(struct nfs4_state *state, struct file_lock *fl) case -ESTALE: goto out; case -NFS4ERR_EXPIRED: + nfs4_schedule_stateid_recovery(server, state); case -NFS4ERR_STALE_CLIENTID: case -NFS4ERR_STALE_STATEID: nfs4_schedule_lease_recovery(server->nfs_client); @@ -4788,7 +4800,7 @@ int nfs4_proc_exchange_id(struct nfs_client *clp, struct rpc_cred *cred) init_utsname()->domainname, clp->cl_rpcclient->cl_auth->au_flavor); - status = rpc_call_sync(clp->cl_rpcclient, &msg, 0); + status = rpc_call_sync(clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); if (!status) status = nfs4_check_cl_exchange_flags(clp->cl_exchange_flags); dprintk("<-- %s status= %d\n", __func__, status); @@ -4839,6 +4851,8 @@ static void nfs4_get_lease_time_done(struct rpc_task *task, void *calldata) dprintk("%s Retry: tk_status %d\n", __func__, task->tk_status); rpc_delay(task, NFS4_POLL_RETRY_MIN); task->tk_status = 0; + /* fall through */ + case -NFS4ERR_RETRY_UNCACHED_REP: nfs_restart_rpc(task, data->clp); return; } @@ -4871,7 +4885,8 @@ int nfs4_proc_get_lease_time(struct nfs_client *clp, struct nfs_fsinfo *fsinfo) .rpc_client = clp->cl_rpcclient, .rpc_message = &msg, .callback_ops = &nfs4_get_lease_time_ops, - .callback_data = &data + .callback_data = &data, + .flags = RPC_TASK_TIMEOUT, }; int status; @@ -5173,7 +5188,7 @@ static int _nfs4_proc_create_session(struct nfs_client *clp) nfs4_init_channel_attrs(&args); args.flags = (SESSION4_PERSIST | SESSION4_BACK_CHAN); - status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0); + status = rpc_call_sync(session->clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); if (!status) /* Verify the session's negotiated channel_attrs values */ @@ -5196,20 +5211,10 @@ int nfs4_proc_create_session(struct nfs_client *clp) int status; unsigned *ptr; struct nfs4_session *session = clp->cl_session; - long timeout = 0; - int err; dprintk("--> %s clp=%p session=%p\n", __func__, clp, session); - do { - status = _nfs4_proc_create_session(clp); - if (status == -NFS4ERR_DELAY) { - err = nfs4_delay(clp->cl_rpcclient, &timeout); - if (err) - status = err; - } - } while (status == -NFS4ERR_DELAY); - + status = _nfs4_proc_create_session(clp); if (status) goto out; @@ -5250,7 +5255,7 @@ int nfs4_proc_destroy_session(struct nfs4_session *session) msg.rpc_argp = session; msg.rpc_resp = NULL; msg.rpc_cred = NULL; - status = rpc_call_sync(session->clp->cl_rpcclient, &msg, 0); + status = rpc_call_sync(session->clp->cl_rpcclient, &msg, RPC_TASK_TIMEOUT); if (status) printk(KERN_WARNING @@ -5483,6 +5488,8 @@ static int nfs41_reclaim_complete_handle_errors(struct rpc_task *task, struct nf break; case -NFS4ERR_DELAY: rpc_delay(task, NFS4_POLL_RETRY_MAX); + /* fall through */ + case -NFS4ERR_RETRY_UNCACHED_REP: return -EAGAIN; default: nfs4_schedule_lease_recovery(clp); diff --git a/fs/nfs/nfs4state.c b/fs/nfs/nfs4state.c index a6804f704d9d..e97dd219f84f 100644 --- a/fs/nfs/nfs4state.c +++ b/fs/nfs/nfs4state.c @@ -64,10 +64,15 @@ static LIST_HEAD(nfs4_clientid_list); int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred) { - struct nfs4_setclientid_res clid; + struct nfs4_setclientid_res clid = { + .clientid = clp->cl_clientid, + .confirm = clp->cl_confirm, + }; unsigned short port; int status; + if (test_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state)) + goto do_confirm; port = nfs_callback_tcpport; if (clp->cl_addr.ss_family == AF_INET6) port = nfs_callback_tcpport6; @@ -75,10 +80,14 @@ int nfs4_init_clientid(struct nfs_client *clp, struct rpc_cred *cred) status = nfs4_proc_setclientid(clp, NFS4_CALLBACK, port, cred, &clid); if (status != 0) goto out; + clp->cl_clientid = clid.clientid; + clp->cl_confirm = clid.confirm; + set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); +do_confirm: status = nfs4_proc_setclientid_confirm(clp, &clid, cred); if (status != 0) goto out; - clp->cl_clientid = clid.clientid; + clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); nfs4_schedule_state_renewal(clp); out: return status; @@ -230,13 +239,18 @@ int nfs41_init_clientid(struct nfs_client *clp, struct rpc_cred *cred) { int status; + if (test_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state)) + goto do_confirm; nfs4_begin_drain_session(clp); status = nfs4_proc_exchange_id(clp, cred); if (status != 0) goto out; + set_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); +do_confirm: status = nfs4_proc_create_session(clp); if (status != 0) goto out; + clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); nfs41_setup_state_renewal(clp); nfs_mark_client_ready(clp, NFS_CS_READY); out: @@ -1452,7 +1466,10 @@ static int nfs4_reclaim_lease(struct nfs_client *clp) #ifdef CONFIG_NFS_V4_1 void nfs4_schedule_session_recovery(struct nfs4_session *session) { - nfs4_schedule_lease_recovery(session->clp); + struct nfs_client *clp = session->clp; + + set_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state); + nfs4_schedule_lease_recovery(clp); } EXPORT_SYMBOL_GPL(nfs4_schedule_session_recovery); @@ -1535,6 +1552,7 @@ static int nfs4_reset_session(struct nfs_client *clp) status = nfs4_recovery_handle_error(clp, status); goto out; } + clear_bit(NFS4CLNT_SESSION_RESET, &clp->cl_state); /* create_session negotiated new slot table */ clear_bit(NFS4CLNT_RECALL_SLOT, &clp->cl_state); @@ -1584,20 +1602,23 @@ static int nfs4_recall_slot(struct nfs_client *clp) { return 0; } */ static void nfs4_set_lease_expired(struct nfs_client *clp, int status) { - if (nfs4_has_session(clp)) { - switch (status) { - case -NFS4ERR_DELAY: - case -NFS4ERR_CLID_INUSE: - case -EAGAIN: - break; + switch (status) { + case -NFS4ERR_CLID_INUSE: + case -NFS4ERR_STALE_CLIENTID: + clear_bit(NFS4CLNT_LEASE_CONFIRM, &clp->cl_state); + break; + case -NFS4ERR_DELAY: + case -ETIMEDOUT: + case -EAGAIN: + ssleep(1); + break; - case -EKEYEXPIRED: - nfs4_warn_keyexpired(clp->cl_hostname); - case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery - * in nfs4_exchange_id */ - default: - return; - } + case -EKEYEXPIRED: + nfs4_warn_keyexpired(clp->cl_hostname); + case -NFS4ERR_NOT_SAME: /* FixMe: implement recovery + * in nfs4_exchange_id */ + default: + return; } set_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state); } @@ -1607,7 +1628,7 @@ static void nfs4_state_manager(struct nfs_client *clp) int status = 0; /* Ensure exclusive access to NFSv4 state */ - for(;;) { + do { if (test_and_clear_bit(NFS4CLNT_LEASE_EXPIRED, &clp->cl_state)) { /* We're going to have to re-establish a clientid */ status = nfs4_reclaim_lease(clp); @@ -1691,7 +1712,7 @@ static void nfs4_state_manager(struct nfs_client *clp) break; if (test_and_set_bit(NFS4CLNT_MANAGER_RUNNING, &clp->cl_state) != 0) break; - } + } while (atomic_read(&clp->cl_count) > 1); return; out_error: printk(KERN_WARNING "Error: state manager failed on NFSv4 server %s" diff --git a/fs/nfs/nfs4xdr.c b/fs/nfs/nfs4xdr.c index dddfb5795d7b..c3ccd2c46834 100644 --- a/fs/nfs/nfs4xdr.c +++ b/fs/nfs/nfs4xdr.c @@ -1452,26 +1452,25 @@ static void encode_read(struct xdr_stream *xdr, const struct nfs_readargs *args, static void encode_readdir(struct xdr_stream *xdr, const struct nfs4_readdir_arg *readdir, struct rpc_rqst *req, struct compound_hdr *hdr) { - uint32_t attrs[2] = {0, 0}; + uint32_t attrs[2] = { + FATTR4_WORD0_RDATTR_ERROR, + FATTR4_WORD1_MOUNTED_ON_FILEID, + }; uint32_t dircount = readdir->count >> 1; __be32 *p; if (readdir->plus) { attrs[0] |= FATTR4_WORD0_TYPE|FATTR4_WORD0_CHANGE|FATTR4_WORD0_SIZE| - FATTR4_WORD0_FSID|FATTR4_WORD0_FILEHANDLE; + FATTR4_WORD0_FSID|FATTR4_WORD0_FILEHANDLE|FATTR4_WORD0_FILEID; attrs[1] |= FATTR4_WORD1_MODE|FATTR4_WORD1_NUMLINKS|FATTR4_WORD1_OWNER| FATTR4_WORD1_OWNER_GROUP|FATTR4_WORD1_RAWDEV| FATTR4_WORD1_SPACE_USED|FATTR4_WORD1_TIME_ACCESS| FATTR4_WORD1_TIME_METADATA|FATTR4_WORD1_TIME_MODIFY; dircount >>= 1; } - attrs[0] |= FATTR4_WORD0_RDATTR_ERROR|FATTR4_WORD0_FILEID; - attrs[1] |= FATTR4_WORD1_MOUNTED_ON_FILEID; - /* Switch to mounted_on_fileid if the server supports it */ - if (readdir->bitmask[1] & FATTR4_WORD1_MOUNTED_ON_FILEID) - attrs[0] &= ~FATTR4_WORD0_FILEID; - else - attrs[1] &= ~FATTR4_WORD1_MOUNTED_ON_FILEID; + /* Use mounted_on_fileid only if the server supports it */ + if (!(readdir->bitmask[1] & FATTR4_WORD1_MOUNTED_ON_FILEID)) + attrs[0] |= FATTR4_WORD0_FILEID; p = reserve_space(xdr, 12+NFS4_VERIFIER_SIZE+20); *p++ = cpu_to_be32(OP_READDIR); @@ -3140,7 +3139,7 @@ static int decode_attr_mounted_on_fileid(struct xdr_stream *xdr, uint32_t *bitma goto out_overflow; xdr_decode_hyper(p, fileid); bitmap[1] &= ~FATTR4_WORD1_MOUNTED_ON_FILEID; - ret = NFS_ATTR_FATTR_FILEID; + ret = NFS_ATTR_FATTR_MOUNTED_ON_FILEID; } dprintk("%s: fileid=%Lu\n", __func__, (unsigned long long)*fileid); return ret; @@ -4002,7 +4001,6 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap, { int status; umode_t fmode = 0; - uint64_t fileid; uint32_t type; status = decode_attr_type(xdr, bitmap, &type); @@ -4101,13 +4099,10 @@ static int decode_getfattr_attrs(struct xdr_stream *xdr, uint32_t *bitmap, goto xdr_error; fattr->valid |= status; - status = decode_attr_mounted_on_fileid(xdr, bitmap, &fileid); + status = decode_attr_mounted_on_fileid(xdr, bitmap, &fattr->mounted_on_fileid); if (status < 0) goto xdr_error; - if (status != 0 && !(fattr->valid & status)) { - fattr->fileid = fileid; - fattr->valid |= status; - } + fattr->valid |= status; xdr_error: dprintk("%s: xdr returned %d\n", __func__, -status); @@ -4838,17 +4833,21 @@ static int decode_secinfo(struct xdr_stream *xdr, struct nfs4_secinfo_res *res) struct nfs4_secinfo_flavor *sec_flavor; int status; __be32 *p; - int i; + int i, num_flavors; status = decode_op_hdr(xdr, OP_SECINFO); + if (status) + goto out; p = xdr_inline_decode(xdr, 4); if (unlikely(!p)) goto out_overflow; - res->flavors->num_flavors = be32_to_cpup(p); - for (i = 0; i < res->flavors->num_flavors; i++) { + res->flavors->num_flavors = 0; + num_flavors = be32_to_cpup(p); + + for (i = 0; i < num_flavors; i++) { sec_flavor = &res->flavors->flavors[i]; - if ((char *)&sec_flavor[1] - (char *)res > PAGE_SIZE) + if ((char *)&sec_flavor[1] - (char *)res->flavors > PAGE_SIZE) break; p = xdr_inline_decode(xdr, 4); @@ -4857,13 +4856,15 @@ static int decode_secinfo(struct xdr_stream *xdr, struct nfs4_secinfo_res *res) sec_flavor->flavor = be32_to_cpup(p); if (sec_flavor->flavor == RPC_AUTH_GSS) { - if (decode_secinfo_gss(xdr, sec_flavor)) - break; + status = decode_secinfo_gss(xdr, sec_flavor); + if (status) + goto out; } + res->flavors->num_flavors++; } - return 0; - +out: + return status; out_overflow: print_overflow_msg(__func__, xdr); return -EIO; @@ -6408,7 +6409,9 @@ int nfs4_decode_dirent(struct xdr_stream *xdr, struct nfs_entry *entry, if (decode_getfattr_attrs(xdr, bitmap, entry->fattr, entry->fh, entry->server, 1) < 0) goto out_overflow; - if (entry->fattr->valid & NFS_ATTR_FATTR_FILEID) + if (entry->fattr->valid & NFS_ATTR_FATTR_MOUNTED_ON_FILEID) + entry->ino = entry->fattr->mounted_on_fileid; + else if (entry->fattr->valid & NFS_ATTR_FATTR_FILEID) entry->ino = entry->fattr->fileid; entry->d_type = DT_UNKNOWN; diff --git a/fs/nfs/nfsroot.c b/fs/nfs/nfsroot.c index c541093a5bf2..c4744e1d513c 100644 --- a/fs/nfs/nfsroot.c +++ b/fs/nfs/nfsroot.c @@ -87,7 +87,7 @@ #define NFS_ROOT "/tftpboot/%s" /* Default NFSROOT mount options. */ -#define NFS_DEF_OPTIONS "udp" +#define NFS_DEF_OPTIONS "vers=2,udp,rsize=4096,wsize=4096" /* Parameters passed from the kernel command line */ static char nfs_root_parms[256] __initdata = ""; diff --git a/fs/nfs/pnfs.c b/fs/nfs/pnfs.c index d9ab97269ce6..101c85a3644e 100644 --- a/fs/nfs/pnfs.c +++ b/fs/nfs/pnfs.c @@ -383,6 +383,7 @@ pnfs_destroy_all_layouts(struct nfs_client *clp) plh_layouts); dprintk("%s freeing layout for inode %lu\n", __func__, lo->plh_inode->i_ino); + list_del_init(&lo->plh_layouts); pnfs_destroy_layout(NFS_I(lo->plh_inode)); } } @@ -466,7 +467,8 @@ pnfs_choose_layoutget_stateid(nfs4_stateid *dst, struct pnfs_layout_hdr *lo, static struct pnfs_layout_segment * send_layoutget(struct pnfs_layout_hdr *lo, struct nfs_open_context *ctx, - u32 iomode) + u32 iomode, + gfp_t gfp_flags) { struct inode *ino = lo->plh_inode; struct nfs_server *server = NFS_SERVER(ino); @@ -479,7 +481,7 @@ send_layoutget(struct pnfs_layout_hdr *lo, dprintk("--> %s\n", __func__); BUG_ON(ctx == NULL); - lgp = kzalloc(sizeof(*lgp), GFP_KERNEL); + lgp = kzalloc(sizeof(*lgp), gfp_flags); if (lgp == NULL) return NULL; @@ -487,12 +489,12 @@ send_layoutget(struct pnfs_layout_hdr *lo, max_resp_sz = server->nfs_client->cl_session->fc_attrs.max_resp_sz; max_pages = max_resp_sz >> PAGE_SHIFT; - pages = kzalloc(max_pages * sizeof(struct page *), GFP_KERNEL); + pages = kzalloc(max_pages * sizeof(struct page *), gfp_flags); if (!pages) goto out_err_free; for (i = 0; i < max_pages; i++) { - pages[i] = alloc_page(GFP_KERNEL); + pages[i] = alloc_page(gfp_flags); if (!pages[i]) goto out_err_free; } @@ -508,6 +510,7 @@ send_layoutget(struct pnfs_layout_hdr *lo, lgp->args.layout.pages = pages; lgp->args.layout.pglen = max_pages * PAGE_SIZE; lgp->lsegpp = &lseg; + lgp->gfp_flags = gfp_flags; /* Synchronously retrieve layout information from server and * store in lseg. @@ -665,11 +668,11 @@ pnfs_insert_layout(struct pnfs_layout_hdr *lo, } static struct pnfs_layout_hdr * -alloc_init_layout_hdr(struct inode *ino) +alloc_init_layout_hdr(struct inode *ino, gfp_t gfp_flags) { struct pnfs_layout_hdr *lo; - lo = kzalloc(sizeof(struct pnfs_layout_hdr), GFP_KERNEL); + lo = kzalloc(sizeof(struct pnfs_layout_hdr), gfp_flags); if (!lo) return NULL; atomic_set(&lo->plh_refcount, 1); @@ -681,7 +684,7 @@ alloc_init_layout_hdr(struct inode *ino) } static struct pnfs_layout_hdr * -pnfs_find_alloc_layout(struct inode *ino) +pnfs_find_alloc_layout(struct inode *ino, gfp_t gfp_flags) { struct nfs_inode *nfsi = NFS_I(ino); struct pnfs_layout_hdr *new = NULL; @@ -696,7 +699,7 @@ pnfs_find_alloc_layout(struct inode *ino) return nfsi->layout; } spin_unlock(&ino->i_lock); - new = alloc_init_layout_hdr(ino); + new = alloc_init_layout_hdr(ino, gfp_flags); spin_lock(&ino->i_lock); if (likely(nfsi->layout == NULL)) /* Won the race? */ @@ -756,7 +759,8 @@ pnfs_find_lseg(struct pnfs_layout_hdr *lo, u32 iomode) struct pnfs_layout_segment * pnfs_update_layout(struct inode *ino, struct nfs_open_context *ctx, - enum pnfs_iomode iomode) + enum pnfs_iomode iomode, + gfp_t gfp_flags) { struct nfs_inode *nfsi = NFS_I(ino); struct nfs_client *clp = NFS_SERVER(ino)->nfs_client; @@ -767,7 +771,7 @@ pnfs_update_layout(struct inode *ino, if (!pnfs_enabled_sb(NFS_SERVER(ino))) return NULL; spin_lock(&ino->i_lock); - lo = pnfs_find_alloc_layout(ino); + lo = pnfs_find_alloc_layout(ino, gfp_flags); if (lo == NULL) { dprintk("%s ERROR: can't get pnfs_layout_hdr\n", __func__); goto out_unlock; @@ -807,7 +811,7 @@ pnfs_update_layout(struct inode *ino, spin_unlock(&clp->cl_lock); } - lseg = send_layoutget(lo, ctx, iomode); + lseg = send_layoutget(lo, ctx, iomode, gfp_flags); if (!lseg && first) { spin_lock(&clp->cl_lock); list_del_init(&lo->plh_layouts); @@ -846,7 +850,7 @@ pnfs_layout_process(struct nfs4_layoutget *lgp) goto out; } /* Inject layout blob into I/O device driver */ - lseg = NFS_SERVER(ino)->pnfs_curr_ld->alloc_lseg(lo, res); + lseg = NFS_SERVER(ino)->pnfs_curr_ld->alloc_lseg(lo, res, lgp->gfp_flags); if (!lseg || IS_ERR(lseg)) { if (!lseg) status = -ENOMEM; @@ -899,7 +903,8 @@ static int pnfs_read_pg_test(struct nfs_pageio_descriptor *pgio, /* This is first coelesce call for a series of nfs_pages */ pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode, prev->wb_context, - IOMODE_READ); + IOMODE_READ, + GFP_KERNEL); } return NFS_SERVER(pgio->pg_inode)->pnfs_curr_ld->pg_test(pgio, prev, req); } @@ -921,7 +926,8 @@ static int pnfs_write_pg_test(struct nfs_pageio_descriptor *pgio, /* This is first coelesce call for a series of nfs_pages */ pgio->pg_lseg = pnfs_update_layout(pgio->pg_inode, prev->wb_context, - IOMODE_RW); + IOMODE_RW, + GFP_NOFS); } return NFS_SERVER(pgio->pg_inode)->pnfs_curr_ld->pg_test(pgio, prev, req); } @@ -1003,7 +1009,8 @@ void pnfs_set_layoutcommit(struct nfs_write_data *wdata) { struct nfs_inode *nfsi = NFS_I(wdata->inode); - loff_t end_pos = wdata->args.offset + wdata->res.count; + loff_t end_pos = wdata->mds_offset + wdata->res.count; + bool mark_as_dirty = false; spin_lock(&nfsi->vfs_inode.i_lock); if (!test_and_set_bit(NFS_INO_LAYOUTCOMMIT, &nfsi->flags)) { @@ -1011,13 +1018,18 @@ pnfs_set_layoutcommit(struct nfs_write_data *wdata) get_lseg(wdata->lseg); wdata->lseg->pls_lc_cred = get_rpccred(wdata->args.context->state->owner->so_cred); - mark_inode_dirty_sync(wdata->inode); + mark_as_dirty = true; dprintk("%s: Set layoutcommit for inode %lu ", __func__, wdata->inode->i_ino); } if (end_pos > wdata->lseg->pls_end_pos) wdata->lseg->pls_end_pos = end_pos; spin_unlock(&nfsi->vfs_inode.i_lock); + + /* if pnfs_layoutcommit_inode() runs between inode locks, the next one + * will be a noop because NFS_INO_LAYOUTCOMMIT will not be set */ + if (mark_as_dirty) + mark_inode_dirty_sync(wdata->inode); } EXPORT_SYMBOL_GPL(pnfs_set_layoutcommit); diff --git a/fs/nfs/pnfs.h b/fs/nfs/pnfs.h index bc4827202e7a..0c015bad9e7a 100644 --- a/fs/nfs/pnfs.h +++ b/fs/nfs/pnfs.h @@ -70,7 +70,7 @@ struct pnfs_layoutdriver_type { const u32 id; const char *name; struct module *owner; - struct pnfs_layout_segment * (*alloc_lseg) (struct pnfs_layout_hdr *layoutid, struct nfs4_layoutget_res *lgr); + struct pnfs_layout_segment * (*alloc_lseg) (struct pnfs_layout_hdr *layoutid, struct nfs4_layoutget_res *lgr, gfp_t gfp_flags); void (*free_lseg) (struct pnfs_layout_segment *lseg); /* test for nfs page cache coalescing */ @@ -126,7 +126,7 @@ void get_layout_hdr(struct pnfs_layout_hdr *lo); void put_lseg(struct pnfs_layout_segment *lseg); struct pnfs_layout_segment * pnfs_update_layout(struct inode *ino, struct nfs_open_context *ctx, - enum pnfs_iomode access_type); + enum pnfs_iomode access_type, gfp_t gfp_flags); void set_pnfs_layoutdriver(struct nfs_server *, u32 id); void unset_pnfs_layoutdriver(struct nfs_server *); enum pnfs_try_status pnfs_try_to_write_data(struct nfs_write_data *, @@ -245,7 +245,7 @@ static inline void put_lseg(struct pnfs_layout_segment *lseg) static inline struct pnfs_layout_segment * pnfs_update_layout(struct inode *ino, struct nfs_open_context *ctx, - enum pnfs_iomode access_type) + enum pnfs_iomode access_type, gfp_t gfp_flags) { return NULL; } diff --git a/fs/nfs/read.c b/fs/nfs/read.c index 7cded2b12a05..2bcf0dc306a1 100644 --- a/fs/nfs/read.c +++ b/fs/nfs/read.c @@ -288,7 +288,7 @@ static int nfs_pagein_multi(struct nfs_pageio_descriptor *desc) atomic_set(&req->wb_complete, requests); BUG_ON(desc->pg_lseg != NULL); - lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_READ); + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_READ, GFP_KERNEL); ClearPageError(page); offset = 0; nbytes = desc->pg_count; @@ -351,7 +351,7 @@ static int nfs_pagein_one(struct nfs_pageio_descriptor *desc) } req = nfs_list_entry(data->pages.next); if ((!lseg) && list_is_singular(&data->pages)) - lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_READ); + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_READ, GFP_KERNEL); ret = nfs_read_rpcsetup(req, data, &nfs_read_full_ops, desc->pg_count, 0, lseg); diff --git a/fs/nfs/super.c b/fs/nfs/super.c index 2b8e9a5e366a..e288f06d3fa7 100644 --- a/fs/nfs/super.c +++ b/fs/nfs/super.c @@ -1004,6 +1004,7 @@ static int nfs_parse_security_flavors(char *value, return 0; } + mnt->flags |= NFS_MOUNT_SECFLAVOUR; mnt->auth_flavor_len = 1; return 1; } @@ -1976,6 +1977,15 @@ nfs_remount(struct super_block *sb, int *flags, char *raw_data) if (error < 0) goto out; + /* + * noac is a special case. It implies -o sync, but that's not + * necessarily reflected in the mtab options. do_remount_sb + * will clear MS_SYNCHRONOUS if -o sync wasn't specified in the + * remount options, so we have to explicitly reset it. + */ + if (data->flags & NFS_MOUNT_NOAC) + *flags |= MS_SYNCHRONOUS; + /* compare new mount options with old ones */ error = nfs_compare_remount_data(nfss, data); out: @@ -2235,8 +2245,7 @@ static struct dentry *nfs_fs_mount(struct file_system_type *fs_type, if (!s->s_root) { /* initial superblock/root creation */ nfs_fill_super(s, data); - nfs_fscache_get_super_cookie( - s, data ? data->fscache_uniq : NULL, NULL); + nfs_fscache_get_super_cookie(s, data->fscache_uniq, NULL); } mntroot = nfs_get_root(s, mntfh, dev_name); diff --git a/fs/nfs/write.c b/fs/nfs/write.c index af0c6279a4a7..49c715b4ac92 100644 --- a/fs/nfs/write.c +++ b/fs/nfs/write.c @@ -542,11 +542,15 @@ nfs_scan_commit(struct inode *inode, struct list_head *dst, pgoff_t idx_start, u if (!nfs_need_commit(nfsi)) return 0; + spin_lock(&inode->i_lock); ret = nfs_scan_list(nfsi, dst, idx_start, npages, NFS_PAGE_TAG_COMMIT); if (ret > 0) nfsi->ncommit -= ret; + spin_unlock(&inode->i_lock); + if (nfs_need_commit(NFS_I(inode))) __mark_inode_dirty(inode, I_DIRTY_DATASYNC); + return ret; } #else @@ -676,7 +680,6 @@ static int nfs_writepage_setup(struct nfs_open_context *ctx, struct page *page, req = nfs_setup_write_request(ctx, page, offset, count); if (IS_ERR(req)) return PTR_ERR(req); - nfs_mark_request_dirty(req); /* Update file length */ nfs_grow_file(page, offset, count); nfs_mark_uptodate(page, req->wb_pgbase, req->wb_bytes); @@ -936,7 +939,7 @@ static int nfs_flush_multi(struct nfs_pageio_descriptor *desc) atomic_set(&req->wb_complete, requests); BUG_ON(desc->pg_lseg); - lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_RW); + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_RW, GFP_NOFS); ClearPageError(page); offset = 0; nbytes = desc->pg_count; @@ -1010,7 +1013,7 @@ static int nfs_flush_one(struct nfs_pageio_descriptor *desc) } req = nfs_list_entry(data->pages.next); if ((!lseg) && list_is_singular(&data->pages)) - lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_RW); + lseg = pnfs_update_layout(desc->pg_inode, req->wb_context, IOMODE_RW, GFP_NOFS); if ((desc->pg_ioflags & FLUSH_COND_STABLE) && (desc->pg_moreio || NFS_I(desc->pg_inode)->ncommit)) @@ -1414,8 +1417,7 @@ static void nfs_commit_done(struct rpc_task *task, void *calldata) task->tk_pid, task->tk_status); /* Call the NFS version-specific code */ - if (NFS_PROTO(data->inode)->commit_done(task, data) != 0) - return; + NFS_PROTO(data->inode)->commit_done(task, data); } void nfs_commit_release_pages(struct nfs_write_data *data) @@ -1483,9 +1485,7 @@ int nfs_commit_inode(struct inode *inode, int how) res = nfs_commit_set_lock(NFS_I(inode), may_wait); if (res <= 0) goto out_mark_dirty; - spin_lock(&inode->i_lock); res = nfs_scan_commit(inode, &head, 0, 0); - spin_unlock(&inode->i_lock); if (res) { int error; diff --git a/fs/nfs_common/nfsacl.c b/fs/nfs_common/nfsacl.c index ec0f277be7f5..6940439bd609 100644 --- a/fs/nfs_common/nfsacl.c +++ b/fs/nfs_common/nfsacl.c @@ -173,7 +173,7 @@ xdr_nfsace_decode(struct xdr_array2_desc *desc, void *elem) return -EINVAL; break; case ACL_MASK: - /* Solaris sometimes sets additonal bits in the mask */ + /* Solaris sometimes sets additional bits in the mask */ entry->e_perm &= S_IRWXO; break; default: diff --git a/fs/nfsd/nfs3xdr.c b/fs/nfsd/nfs3xdr.c index 291f2d2551b3..08c6e36ab2eb 100644 --- a/fs/nfsd/nfs3xdr.c +++ b/fs/nfsd/nfs3xdr.c @@ -702,7 +702,7 @@ nfs3svc_encode_readres(struct svc_rqst *rqstp, __be32 *p, *p++ = htonl(resp->eof); *p++ = htonl(resp->count); /* xdr opaque count */ xdr_ressize_check(rqstp, p); - /* now update rqstp->rq_res to reflect data aswell */ + /* now update rqstp->rq_res to reflect data as well */ rqstp->rq_res.page_len = resp->count; if (resp->count & 3) { /* need to pad the tail */ diff --git a/fs/nfsd/nfs4state.c b/fs/nfsd/nfs4state.c index a8e4e91587ab..e98f3c2e9492 100644 --- a/fs/nfsd/nfs4state.c +++ b/fs/nfsd/nfs4state.c @@ -258,6 +258,7 @@ static void nfs4_put_deleg_lease(struct nfs4_file *fp) if (atomic_dec_and_test(&fp->fi_delegees)) { vfs_setlease(fp->fi_deleg_file, F_UNLCK, &fp->fi_lease); fp->fi_lease = NULL; + fput(fp->fi_deleg_file); fp->fi_deleg_file = NULL; } } @@ -397,9 +398,12 @@ static void unhash_generic_stateid(struct nfs4_stateid *stp) static void free_generic_stateid(struct nfs4_stateid *stp) { - int oflag = nfs4_access_bmap_to_omode(stp); + int oflag; - nfs4_file_put_access(stp->st_file, oflag); + if (stp->st_access_bmap) { + oflag = nfs4_access_bmap_to_omode(stp); + nfs4_file_put_access(stp->st_file, oflag); + } put_nfs4_file(stp->st_file); kmem_cache_free(stateid_slab, stp); } @@ -3079,7 +3083,7 @@ check_special_stateids(svc_fh *current_fh, stateid_t *stateid, int flags) if (ONE_STATEID(stateid) && (flags & RD_STATE)) return nfs_ok; else if (locks_in_grace()) { - /* Answer in remaining cases depends on existance of + /* Answer in remaining cases depends on existence of * conflicting state; so we must wait out the grace period. */ return nfserr_grace; } else if (flags & WR_STATE) @@ -3699,7 +3703,7 @@ find_lockstateowner_str(struct inode *inode, clientid_t *clid, /* * Alloc a lock owner structure. * Called in nfsd4_lock - therefore, OPEN and OPEN_CONFIRM (if needed) has - * occured. + * occurred. * * strhashval = lock_ownerstr_hashval */ diff --git a/fs/nfsd/nfsxdr.c b/fs/nfsd/nfsxdr.c index 4ce005dbf3e6..65ec595e2226 100644 --- a/fs/nfsd/nfsxdr.c +++ b/fs/nfsd/nfsxdr.c @@ -451,7 +451,7 @@ nfssvc_encode_readres(struct svc_rqst *rqstp, __be32 *p, *p++ = htonl(resp->count); xdr_ressize_check(rqstp, p); - /* now update rqstp->rq_res to reflect data aswell */ + /* now update rqstp->rq_res to reflect data as well */ rqstp->rq_res.page_len = resp->count; if (resp->count & 3) { /* need to pad the tail */ diff --git a/fs/nfsd/stats.c b/fs/nfsd/stats.c index 5232d3e8fb2f..a2e2402b2afb 100644 --- a/fs/nfsd/stats.c +++ b/fs/nfsd/stats.c @@ -8,7 +8,7 @@ * Statistsics for the reply cache * fh <stale> <total-lookups> <anonlookups> <dir-not-in-dcache> <nondir-not-in-dcache> * statistics for filehandle lookup - * io <bytes-read> <bytes-writtten> + * io <bytes-read> <bytes-written> * statistics for IO throughput * th <threads> <fullcnt> <10%-20%> <20%-30%> ... <90%-100%> <100%> * time (seconds) when nfsd thread usage above thresholds diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index f4e056ae53e4..d5718273bb32 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -1365,7 +1365,7 @@ do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, goto out; if (!(iap->ia_valid & ATTR_MODE)) iap->ia_mode = 0; - err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE); + err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_EXEC); if (err) goto out; @@ -1387,6 +1387,13 @@ do_nfsd_create(struct svc_rqst *rqstp, struct svc_fh *fhp, if (IS_ERR(dchild)) goto out_nfserr; + /* If file doesn't exist, check for permissions to create one */ + if (!dchild->d_inode) { + err = fh_verify(rqstp, fhp, S_IFDIR, NFSD_MAY_CREATE); + if (err) + goto out; + } + err = fh_compose(resfhp, fhp->fh_export, dchild, fhp); if (err) goto out; diff --git a/fs/nilfs2/alloc.c b/fs/nilfs2/alloc.c index 0a0a66d98cce..eed4d7b26249 100644 --- a/fs/nilfs2/alloc.c +++ b/fs/nilfs2/alloc.c @@ -489,8 +489,8 @@ int nilfs_palloc_prepare_alloc_entry(struct inode *inode, void nilfs_palloc_commit_alloc_entry(struct inode *inode, struct nilfs_palloc_req *req) { - nilfs_mdt_mark_buffer_dirty(req->pr_bitmap_bh); - nilfs_mdt_mark_buffer_dirty(req->pr_desc_bh); + mark_buffer_dirty(req->pr_bitmap_bh); + mark_buffer_dirty(req->pr_desc_bh); nilfs_mdt_mark_dirty(inode); brelse(req->pr_bitmap_bh); @@ -527,8 +527,8 @@ void nilfs_palloc_commit_free_entry(struct inode *inode, kunmap(req->pr_bitmap_bh->b_page); kunmap(req->pr_desc_bh->b_page); - nilfs_mdt_mark_buffer_dirty(req->pr_desc_bh); - nilfs_mdt_mark_buffer_dirty(req->pr_bitmap_bh); + mark_buffer_dirty(req->pr_desc_bh); + mark_buffer_dirty(req->pr_bitmap_bh); nilfs_mdt_mark_dirty(inode); brelse(req->pr_bitmap_bh); @@ -646,7 +646,7 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) unsigned long group, group_offset; int i, j, n, ret; - for (i = 0; i < nitems; i += n) { + for (i = 0; i < nitems; i = j) { group = nilfs_palloc_group(inode, entry_nrs[i], &group_offset); ret = nilfs_palloc_get_desc_block(inode, group, 0, &desc_bh); if (ret < 0) @@ -683,8 +683,8 @@ int nilfs_palloc_freev(struct inode *inode, __u64 *entry_nrs, size_t nitems) kunmap(bitmap_bh->b_page); kunmap(desc_bh->b_page); - nilfs_mdt_mark_buffer_dirty(desc_bh); - nilfs_mdt_mark_buffer_dirty(bitmap_bh); + mark_buffer_dirty(desc_bh); + mark_buffer_dirty(bitmap_bh); nilfs_mdt_mark_dirty(inode); brelse(bitmap_bh); diff --git a/fs/nilfs2/bmap.c b/fs/nilfs2/bmap.c index 4723f04e9b12..aadbd0b5e3e8 100644 --- a/fs/nilfs2/bmap.c +++ b/fs/nilfs2/bmap.c @@ -34,7 +34,9 @@ struct inode *nilfs_bmap_get_dat(const struct nilfs_bmap *bmap) { - return NILFS_I_NILFS(bmap->b_inode)->ns_dat; + struct the_nilfs *nilfs = bmap->b_inode->i_sb->s_fs_info; + + return nilfs->ns_dat; } static int nilfs_bmap_convert_error(struct nilfs_bmap *bmap, diff --git a/fs/nilfs2/btnode.c b/fs/nilfs2/btnode.c index 609cd223eea8..a35ae35e6932 100644 --- a/fs/nilfs2/btnode.c +++ b/fs/nilfs2/btnode.c @@ -34,12 +34,6 @@ #include "page.h" #include "btnode.h" -void nilfs_btnode_cache_init(struct address_space *btnc, - struct backing_dev_info *bdi) -{ - nilfs_mapping_init(btnc, bdi); -} - void nilfs_btnode_cache_clear(struct address_space *btnc) { invalidate_mapping_pages(btnc, 0, -1); @@ -62,7 +56,7 @@ nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr) BUG(); } memset(bh->b_data, 0, 1 << inode->i_blkbits); - bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; + bh->b_bdev = inode->i_sb->s_bdev; bh->b_blocknr = blocknr; set_buffer_mapped(bh); set_buffer_uptodate(bh); @@ -94,10 +88,11 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, if (pblocknr == 0) { pblocknr = blocknr; if (inode->i_ino != NILFS_DAT_INO) { - struct inode *dat = NILFS_I_NILFS(inode)->ns_dat; + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; /* blocknr is a virtual block number */ - err = nilfs_dat_translate(dat, blocknr, &pblocknr); + err = nilfs_dat_translate(nilfs->ns_dat, blocknr, + &pblocknr); if (unlikely(err)) { brelse(bh); goto out_locked; @@ -120,7 +115,7 @@ int nilfs_btnode_submit_block(struct address_space *btnc, __u64 blocknr, goto found; } set_buffer_mapped(bh); - bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; + bh->b_bdev = inode->i_sb->s_bdev; bh->b_blocknr = pblocknr; /* set block address for read */ bh->b_end_io = end_buffer_read_sync; get_bh(bh); @@ -259,7 +254,7 @@ void nilfs_btnode_commit_change_key(struct address_space *btnc, "invalid oldkey %lld (newkey=%lld)", (unsigned long long)oldkey, (unsigned long long)newkey); - nilfs_btnode_mark_dirty(obh); + mark_buffer_dirty(obh); spin_lock_irq(&btnc->tree_lock); radix_tree_delete(&btnc->page_tree, oldkey); @@ -271,7 +266,7 @@ void nilfs_btnode_commit_change_key(struct address_space *btnc, unlock_page(opage); } else { nilfs_copy_buffer(nbh, obh); - nilfs_btnode_mark_dirty(nbh); + mark_buffer_dirty(nbh); nbh->b_blocknr = newkey; ctxt->bh = nbh; diff --git a/fs/nilfs2/btnode.h b/fs/nilfs2/btnode.h index 1b8ebd888c28..3a4dd2d8d3fc 100644 --- a/fs/nilfs2/btnode.h +++ b/fs/nilfs2/btnode.h @@ -37,7 +37,6 @@ struct nilfs_btnode_chkey_ctxt { struct buffer_head *newbh; }; -void nilfs_btnode_cache_init(struct address_space *, struct backing_dev_info *); void nilfs_btnode_cache_clear(struct address_space *); struct buffer_head *nilfs_btnode_create_block(struct address_space *btnc, __u64 blocknr); @@ -51,7 +50,4 @@ void nilfs_btnode_commit_change_key(struct address_space *, void nilfs_btnode_abort_change_key(struct address_space *, struct nilfs_btnode_chkey_ctxt *); -#define nilfs_btnode_mark_dirty(bh) nilfs_mark_buffer_dirty(bh) - - #endif /* _NILFS_BTNODE_H */ diff --git a/fs/nilfs2/btree.c b/fs/nilfs2/btree.c index d451ae0e0bf3..7eafe468a29c 100644 --- a/fs/nilfs2/btree.c +++ b/fs/nilfs2/btree.c @@ -714,7 +714,7 @@ static void nilfs_btree_promote_key(struct nilfs_bmap *btree, nilfs_btree_get_nonroot_node(path, level), path[level].bp_index, key); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); } while ((path[level].bp_index == 0) && (++level < nilfs_btree_height(btree) - 1)); } @@ -739,7 +739,7 @@ static void nilfs_btree_do_insert(struct nilfs_bmap *btree, nilfs_btree_node_insert(node, path[level].bp_index, *keyp, *ptrp, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (path[level].bp_index == 0) nilfs_btree_promote_key(btree, path, level + 1, @@ -777,9 +777,9 @@ static void nilfs_btree_carry_left(struct nilfs_bmap *btree, nilfs_btree_node_move_left(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -823,9 +823,9 @@ static void nilfs_btree_carry_right(struct nilfs_bmap *btree, nilfs_btree_node_move_right(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); path[level + 1].bp_index++; nilfs_btree_promote_key(btree, path, level + 1, @@ -870,9 +870,9 @@ static void nilfs_btree_split(struct nilfs_bmap *btree, nilfs_btree_node_move_right(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); newkey = nilfs_btree_node_get_key(right, 0); newptr = path[level].bp_newreq.bpr_ptr; @@ -919,7 +919,7 @@ static void nilfs_btree_grow(struct nilfs_bmap *btree, nilfs_btree_node_set_level(root, level + 1); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); path[level].bp_bh = path[level].bp_sib_bh; path[level].bp_sib_bh = NULL; @@ -1194,7 +1194,7 @@ static void nilfs_btree_do_delete(struct nilfs_bmap *btree, nilfs_btree_node_delete(node, path[level].bp_index, keyp, ptrp, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (path[level].bp_index == 0) nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -1226,9 +1226,9 @@ static void nilfs_btree_borrow_left(struct nilfs_bmap *btree, nilfs_btree_node_move_right(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); nilfs_btree_promote_key(btree, path, level + 1, nilfs_btree_node_get_key(node, 0)); @@ -1258,9 +1258,9 @@ static void nilfs_btree_borrow_right(struct nilfs_bmap *btree, nilfs_btree_node_move_left(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); path[level + 1].bp_index++; nilfs_btree_promote_key(btree, path, level + 1, @@ -1289,7 +1289,7 @@ static void nilfs_btree_concat_left(struct nilfs_bmap *btree, nilfs_btree_node_move_left(left, node, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_sib_bh)) - nilfs_btnode_mark_dirty(path[level].bp_sib_bh); + mark_buffer_dirty(path[level].bp_sib_bh); nilfs_btnode_delete(path[level].bp_bh); path[level].bp_bh = path[level].bp_sib_bh; @@ -1315,7 +1315,7 @@ static void nilfs_btree_concat_right(struct nilfs_bmap *btree, nilfs_btree_node_move_left(node, right, n, ncblk, ncblk); if (!buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); nilfs_btnode_delete(path[level].bp_sib_bh); path[level].bp_sib_bh = NULL; @@ -1709,7 +1709,7 @@ nilfs_btree_commit_convert_and_insert(struct nilfs_bmap *btree, nilfs_btree_node_init(node, 0, 1, n, ncblk, keys, ptrs); nilfs_btree_node_insert(node, n, key, dreq->bpr_ptr, ncblk); if (!buffer_dirty(bh)) - nilfs_btnode_mark_dirty(bh); + mark_buffer_dirty(bh); if (!nilfs_bmap_dirty(btree)) nilfs_bmap_set_dirty(btree); @@ -1787,7 +1787,7 @@ static int nilfs_btree_propagate_p(struct nilfs_bmap *btree, { while ((++level < nilfs_btree_height(btree) - 1) && !buffer_dirty(path[level].bp_bh)) - nilfs_btnode_mark_dirty(path[level].bp_bh); + mark_buffer_dirty(path[level].bp_bh); return 0; } @@ -2229,7 +2229,7 @@ static int nilfs_btree_mark(struct nilfs_bmap *btree, __u64 key, int level) } if (!buffer_dirty(bh)) - nilfs_btnode_mark_dirty(bh); + mark_buffer_dirty(bh); brelse(bh); if (!nilfs_bmap_dirty(btree)) nilfs_bmap_set_dirty(btree); diff --git a/fs/nilfs2/cpfile.c b/fs/nilfs2/cpfile.c index 5ff15a8a1024..c9b342c8b503 100644 --- a/fs/nilfs2/cpfile.c +++ b/fs/nilfs2/cpfile.c @@ -216,14 +216,14 @@ int nilfs_cpfile_get_checkpoint(struct inode *cpfile, if (!nilfs_cpfile_is_in_first(cpfile, cno)) nilfs_cpfile_block_add_valid_checkpoints(cpfile, cp_bh, kaddr, 1); - nilfs_mdt_mark_buffer_dirty(cp_bh); + mark_buffer_dirty(cp_bh); kaddr = kmap_atomic(header_bh->b_page, KM_USER0); header = nilfs_cpfile_block_get_header(cpfile, header_bh, kaddr); le64_add_cpu(&header->ch_ncheckpoints, 1); kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); } @@ -326,7 +326,7 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, } if (nicps > 0) { tnicps += nicps; - nilfs_mdt_mark_buffer_dirty(cp_bh); + mark_buffer_dirty(cp_bh); nilfs_mdt_mark_dirty(cpfile); if (!nilfs_cpfile_is_in_first(cpfile, cno)) { count = @@ -358,7 +358,7 @@ int nilfs_cpfile_delete_checkpoints(struct inode *cpfile, header = nilfs_cpfile_block_get_header(cpfile, header_bh, kaddr); le64_add_cpu(&header->ch_ncheckpoints, -(u64)tnicps); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); kunmap_atomic(kaddr, KM_USER0); } @@ -671,10 +671,10 @@ static int nilfs_cpfile_set_snapshot(struct inode *cpfile, __u64 cno) le64_add_cpu(&header->ch_nsnapshots, 1); kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(prev_bh); - nilfs_mdt_mark_buffer_dirty(curr_bh); - nilfs_mdt_mark_buffer_dirty(cp_bh); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(prev_bh); + mark_buffer_dirty(curr_bh); + mark_buffer_dirty(cp_bh); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); brelse(prev_bh); @@ -774,10 +774,10 @@ static int nilfs_cpfile_clear_snapshot(struct inode *cpfile, __u64 cno) le64_add_cpu(&header->ch_nsnapshots, -1); kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(next_bh); - nilfs_mdt_mark_buffer_dirty(prev_bh); - nilfs_mdt_mark_buffer_dirty(cp_bh); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(next_bh); + mark_buffer_dirty(prev_bh); + mark_buffer_dirty(cp_bh); + mark_buffer_dirty(header_bh); nilfs_mdt_mark_dirty(cpfile); brelse(prev_bh); diff --git a/fs/nilfs2/dat.c b/fs/nilfs2/dat.c index 59e5fe742f7b..fcc2f869af16 100644 --- a/fs/nilfs2/dat.c +++ b/fs/nilfs2/dat.c @@ -54,7 +54,7 @@ static int nilfs_dat_prepare_entry(struct inode *dat, static void nilfs_dat_commit_entry(struct inode *dat, struct nilfs_palloc_req *req) { - nilfs_mdt_mark_buffer_dirty(req->pr_entry_bh); + mark_buffer_dirty(req->pr_entry_bh); nilfs_mdt_mark_dirty(dat); brelse(req->pr_entry_bh); } @@ -361,7 +361,7 @@ int nilfs_dat_move(struct inode *dat, __u64 vblocknr, sector_t blocknr) entry->de_blocknr = cpu_to_le64(blocknr); kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(entry_bh); + mark_buffer_dirty(entry_bh); nilfs_mdt_mark_dirty(dat); brelse(entry_bh); diff --git a/fs/nilfs2/file.c b/fs/nilfs2/file.c index 397e73258631..d7eeca62febd 100644 --- a/fs/nilfs2/file.c +++ b/fs/nilfs2/file.c @@ -111,7 +111,6 @@ static int nilfs_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) nilfs_transaction_commit(inode->i_sb); mapped: - SetPageChecked(page); wait_on_page_writeback(page); return VM_FAULT_LOCKED; } diff --git a/fs/nilfs2/gcinode.c b/fs/nilfs2/gcinode.c index 1c2a3e23f8b2..08a07a218d26 100644 --- a/fs/nilfs2/gcinode.c +++ b/fs/nilfs2/gcinode.c @@ -48,9 +48,6 @@ #include "dat.h" #include "ifile.h" -static const struct address_space_operations def_gcinode_aops = { -}; - /* * nilfs_gccache_submit_read_data() - add data buffer and submit read request * @inode - gc inode @@ -87,9 +84,9 @@ int nilfs_gccache_submit_read_data(struct inode *inode, sector_t blkoff, goto out; if (pbn == 0) { - struct inode *dat_inode = NILFS_I_NILFS(inode)->ns_dat; - /* use original dat, not gc dat. */ - err = nilfs_dat_translate(dat_inode, vbn, &pbn); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + + err = nilfs_dat_translate(nilfs->ns_dat, vbn, &pbn); if (unlikely(err)) { /* -EIO, -ENOMEM, -ENOENT */ brelse(bh); goto failed; @@ -103,7 +100,7 @@ int nilfs_gccache_submit_read_data(struct inode *inode, sector_t blkoff, } if (!buffer_mapped(bh)) { - bh->b_bdev = NILFS_I_NILFS(inode)->ns_bdev; + bh->b_bdev = inode->i_sb->s_bdev; set_buffer_mapped(bh); } bh->b_blocknr = pbn; @@ -160,15 +157,11 @@ int nilfs_gccache_wait_and_mark_dirty(struct buffer_head *bh) if (buffer_dirty(bh)) return -EEXIST; - if (buffer_nilfs_node(bh)) { - if (nilfs_btree_broken_node_block(bh)) { - clear_buffer_uptodate(bh); - return -EIO; - } - nilfs_btnode_mark_dirty(bh); - } else { - nilfs_mark_buffer_dirty(bh); + if (buffer_nilfs_node(bh) && nilfs_btree_broken_node_block(bh)) { + clear_buffer_uptodate(bh); + return -EIO; } + mark_buffer_dirty(bh); return 0; } @@ -178,7 +171,7 @@ int nilfs_init_gcinode(struct inode *inode) inode->i_mode = S_IFREG; mapping_set_gfp_mask(inode->i_mapping, GFP_NOFS); - inode->i_mapping->a_ops = &def_gcinode_aops; + inode->i_mapping->a_ops = &empty_aops; inode->i_mapping->backing_dev_info = inode->i_sb->s_bdi; ii->i_flags = 0; diff --git a/fs/nilfs2/ifile.c b/fs/nilfs2/ifile.c index bfc73d3a30ed..684d76300a80 100644 --- a/fs/nilfs2/ifile.c +++ b/fs/nilfs2/ifile.c @@ -80,7 +80,7 @@ int nilfs_ifile_create_inode(struct inode *ifile, ino_t *out_ino, return ret; } nilfs_palloc_commit_alloc_entry(ifile, &req); - nilfs_mdt_mark_buffer_dirty(req.pr_entry_bh); + mark_buffer_dirty(req.pr_entry_bh); nilfs_mdt_mark_dirty(ifile); *out_ino = (ino_t)req.pr_entry_nr; *out_bh = req.pr_entry_bh; @@ -128,7 +128,7 @@ int nilfs_ifile_delete_inode(struct inode *ifile, ino_t ino) raw_inode->i_flags = 0; kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(req.pr_entry_bh); + mark_buffer_dirty(req.pr_entry_bh); brelse(req.pr_entry_bh); nilfs_palloc_commit_free_entry(ifile, &req); diff --git a/fs/nilfs2/inode.c b/fs/nilfs2/inode.c index c0aa27490c02..b954878ad6ce 100644 --- a/fs/nilfs2/inode.c +++ b/fs/nilfs2/inode.c @@ -74,14 +74,14 @@ int nilfs_get_block(struct inode *inode, sector_t blkoff, struct buffer_head *bh_result, int create) { struct nilfs_inode_info *ii = NILFS_I(inode); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; __u64 blknum = 0; int err = 0, ret; - struct inode *dat = NILFS_I_NILFS(inode)->ns_dat; unsigned maxblocks = bh_result->b_size >> inode->i_blkbits; - down_read(&NILFS_MDT(dat)->mi_sem); + down_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem); ret = nilfs_bmap_lookup_contig(ii->i_bmap, blkoff, &blknum, maxblocks); - up_read(&NILFS_MDT(dat)->mi_sem); + up_read(&NILFS_MDT(nilfs->ns_dat)->mi_sem); if (ret >= 0) { /* found */ map_bh(bh_result, inode->i_sb, blknum); if (ret > 0) @@ -596,6 +596,16 @@ void nilfs_write_inode_common(struct inode *inode, raw_inode->i_flags = cpu_to_le32(ii->i_flags); raw_inode->i_generation = cpu_to_le32(inode->i_generation); + if (NILFS_ROOT_METADATA_FILE(inode->i_ino)) { + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + + /* zero-fill unused portion in the case of super root block */ + raw_inode->i_xattr = 0; + raw_inode->i_pad = 0; + memset((void *)raw_inode + sizeof(*raw_inode), 0, + nilfs->ns_inode_size - sizeof(*raw_inode)); + } + if (has_bmap) nilfs_bmap_write(ii->i_bmap, raw_inode); else if (S_ISCHR(inode->i_mode) || S_ISBLK(inode->i_mode)) @@ -872,8 +882,7 @@ int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty) return -EINVAL; /* NILFS_I_DIRTY may remain for freeing inode */ } - list_del(&ii->i_dirty); - list_add_tail(&ii->i_dirty, &nilfs->ns_dirty_files); + list_move_tail(&ii->i_dirty, &nilfs->ns_dirty_files); set_bit(NILFS_I_QUEUED, &ii->i_state); } spin_unlock(&nilfs->ns_inode_lock); @@ -892,7 +901,7 @@ int nilfs_mark_inode_dirty(struct inode *inode) return err; } nilfs_update_inode(inode, ibh); - nilfs_mdt_mark_buffer_dirty(ibh); + mark_buffer_dirty(ibh); nilfs_mdt_mark_dirty(NILFS_I(inode)->i_root->ifile); brelse(ibh); return 0; @@ -908,7 +917,7 @@ int nilfs_mark_inode_dirty(struct inode *inode) * construction. This function can be called both as a single operation * and as a part of indivisible file operations. */ -void nilfs_dirty_inode(struct inode *inode) +void nilfs_dirty_inode(struct inode *inode, int flags) { struct nilfs_transaction_info ti; struct nilfs_mdt_info *mdi = NILFS_MDT(inode); @@ -931,7 +940,7 @@ void nilfs_dirty_inode(struct inode *inode) int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len) { - struct the_nilfs *nilfs = NILFS_I_NILFS(inode); + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; __u64 logical = 0, phys = 0, size = 0; __u32 flags = 0; loff_t isize; diff --git a/fs/nilfs2/ioctl.c b/fs/nilfs2/ioctl.c index f2469ba6246b..41d6743d303c 100644 --- a/fs/nilfs2/ioctl.c +++ b/fs/nilfs2/ioctl.c @@ -698,6 +698,63 @@ static int nilfs_ioctl_sync(struct inode *inode, struct file *filp, return 0; } +static int nilfs_ioctl_resize(struct inode *inode, struct file *filp, + void __user *argp) +{ + __u64 newsize; + int ret = -EPERM; + + if (!capable(CAP_SYS_ADMIN)) + goto out; + + ret = mnt_want_write(filp->f_path.mnt); + if (ret) + goto out; + + ret = -EFAULT; + if (copy_from_user(&newsize, argp, sizeof(newsize))) + goto out_drop_write; + + ret = nilfs_resize_fs(inode->i_sb, newsize); + +out_drop_write: + mnt_drop_write(filp->f_path.mnt); +out: + return ret; +} + +static int nilfs_ioctl_set_alloc_range(struct inode *inode, void __user *argp) +{ + struct the_nilfs *nilfs = inode->i_sb->s_fs_info; + __u64 range[2]; + __u64 minseg, maxseg; + unsigned long segbytes; + int ret = -EPERM; + + if (!capable(CAP_SYS_ADMIN)) + goto out; + + ret = -EFAULT; + if (copy_from_user(range, argp, sizeof(__u64[2]))) + goto out; + + ret = -ERANGE; + if (range[1] > i_size_read(inode->i_sb->s_bdev->bd_inode)) + goto out; + + segbytes = nilfs->ns_blocks_per_segment * nilfs->ns_blocksize; + + minseg = range[0] + segbytes - 1; + do_div(minseg, segbytes); + maxseg = NILFS_SB2_OFFSET_BYTES(range[1]); + do_div(maxseg, segbytes); + maxseg--; + + ret = nilfs_sufile_set_alloc_range(nilfs->ns_sufile, minseg, maxseg); +out: + return ret; +} + static int nilfs_ioctl_get_info(struct inode *inode, struct file *filp, unsigned int cmd, void __user *argp, size_t membsz, @@ -763,6 +820,10 @@ long nilfs_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return nilfs_ioctl_clean_segments(inode, filp, cmd, argp); case NILFS_IOCTL_SYNC: return nilfs_ioctl_sync(inode, filp, cmd, argp); + case NILFS_IOCTL_RESIZE: + return nilfs_ioctl_resize(inode, filp, argp); + case NILFS_IOCTL_SET_ALLOC_RANGE: + return nilfs_ioctl_set_alloc_range(inode, argp); default: return -ENOTTY; } diff --git a/fs/nilfs2/mdt.c b/fs/nilfs2/mdt.c index a649b05f7069..800e8d78a83b 100644 --- a/fs/nilfs2/mdt.c +++ b/fs/nilfs2/mdt.c @@ -66,7 +66,7 @@ nilfs_mdt_insert_new_block(struct inode *inode, unsigned long block, kunmap_atomic(kaddr, KM_USER0); set_buffer_uptodate(bh); - nilfs_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(inode); return 0; } @@ -355,7 +355,7 @@ int nilfs_mdt_mark_block_dirty(struct inode *inode, unsigned long block) err = nilfs_mdt_read_block(inode, block, 0, &bh); if (unlikely(err)) return err; - nilfs_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(inode); brelse(bh); return 0; @@ -450,9 +450,9 @@ int nilfs_mdt_setup_shadow_map(struct inode *inode, INIT_LIST_HEAD(&shadow->frozen_buffers); address_space_init_once(&shadow->frozen_data); - nilfs_mapping_init(&shadow->frozen_data, bdi); + nilfs_mapping_init(&shadow->frozen_data, inode, bdi); address_space_init_once(&shadow->frozen_btnodes); - nilfs_mapping_init(&shadow->frozen_btnodes, bdi); + nilfs_mapping_init(&shadow->frozen_btnodes, inode, bdi); mi->mi_shadow = shadow; return 0; } diff --git a/fs/nilfs2/mdt.h b/fs/nilfs2/mdt.h index ed68563ec708..ab20a4baa50f 100644 --- a/fs/nilfs2/mdt.h +++ b/fs/nilfs2/mdt.h @@ -64,11 +64,6 @@ static inline struct nilfs_mdt_info *NILFS_MDT(const struct inode *inode) return inode->i_private; } -static inline struct the_nilfs *NILFS_I_NILFS(struct inode *inode) -{ - return inode->i_sb->s_fs_info; -} - /* Default GFP flags using highmem */ #define NILFS_MDT_GFP (__GFP_WAIT | __GFP_IO | __GFP_HIGHMEM) @@ -93,8 +88,6 @@ int nilfs_mdt_freeze_buffer(struct inode *inode, struct buffer_head *bh); struct buffer_head *nilfs_mdt_get_frozen_buffer(struct inode *inode, struct buffer_head *bh); -#define nilfs_mdt_mark_buffer_dirty(bh) nilfs_mark_buffer_dirty(bh) - static inline void nilfs_mdt_mark_dirty(struct inode *inode) { if (!test_bit(NILFS_I_DIRTY, &NILFS_I(inode)->i_state)) @@ -108,7 +101,7 @@ static inline void nilfs_mdt_clear_dirty(struct inode *inode) static inline __u64 nilfs_mdt_cno(struct inode *inode) { - return NILFS_I_NILFS(inode)->ns_cno; + return ((struct the_nilfs *)inode->i_sb->s_fs_info)->ns_cno; } #define nilfs_mdt_bgl_lock(inode, bg) \ diff --git a/fs/nilfs2/nilfs.h b/fs/nilfs2/nilfs.h index a8dd344303cb..f02b9ad43a21 100644 --- a/fs/nilfs2/nilfs.h +++ b/fs/nilfs2/nilfs.h @@ -80,12 +80,6 @@ static inline struct inode *NILFS_BTNC_I(struct address_space *btnc) return &ii->vfs_inode; } -static inline struct inode *NILFS_AS_I(struct address_space *mapping) -{ - return (mapping->host) ? : - container_of(mapping, struct inode, i_data); -} - /* * Dynamic state flags of NILFS on-memory inode (i_state) */ @@ -275,7 +269,7 @@ int nilfs_load_inode_block(struct inode *inode, struct buffer_head **pbh); extern int nilfs_inode_dirty(struct inode *); int nilfs_set_file_dirty(struct inode *inode, unsigned nr_dirty); extern int nilfs_mark_inode_dirty(struct inode *); -extern void nilfs_dirty_inode(struct inode *); +extern void nilfs_dirty_inode(struct inode *, int flags); int nilfs_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); @@ -298,6 +292,7 @@ struct nilfs_super_block **nilfs_prepare_super(struct super_block *sb, int flip); int nilfs_commit_super(struct super_block *sb, int flag); int nilfs_cleanup_super(struct super_block *sb); +int nilfs_resize_fs(struct super_block *sb, __u64 newsize); int nilfs_attach_checkpoint(struct super_block *sb, __u64 cno, int curr_mnt, struct nilfs_root **root); int nilfs_checkpoint_is_mounted(struct super_block *sb, __u64 cno); diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c index 1168059c7efd..65221a04c6f0 100644 --- a/fs/nilfs2/page.c +++ b/fs/nilfs2/page.c @@ -37,8 +37,7 @@ #define NILFS_BUFFER_INHERENT_BITS \ ((1UL << BH_Uptodate) | (1UL << BH_Mapped) | (1UL << BH_NILFS_Node) | \ - (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Allocated) | \ - (1UL << BH_NILFS_Checked)) + (1UL << BH_NILFS_Volatile) | (1UL << BH_NILFS_Checked)) static struct buffer_head * __nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index, @@ -59,19 +58,6 @@ __nilfs_get_page_block(struct page *page, unsigned long block, pgoff_t index, return bh; } -/* - * Since the page cache of B-tree node pages or data page cache of pseudo - * inodes does not have a valid mapping->host pointer, calling - * mark_buffer_dirty() for their buffers causes a NULL pointer dereference; - * it calls __mark_inode_dirty(NULL) through __set_page_dirty(). - * To avoid this problem, the old style mark_buffer_dirty() is used instead. - */ -void nilfs_mark_buffer_dirty(struct buffer_head *bh) -{ - if (!buffer_dirty(bh) && !test_set_buffer_dirty(bh)) - __set_page_dirty_nobuffers(bh->b_page); -} - struct buffer_head *nilfs_grab_buffer(struct inode *inode, struct address_space *mapping, unsigned long blkoff, @@ -183,7 +169,7 @@ int nilfs_page_buffers_clean(struct page *page) void nilfs_page_bug(struct page *page) { struct address_space *m; - unsigned long ino = 0; + unsigned long ino; if (unlikely(!page)) { printk(KERN_CRIT "NILFS_PAGE_BUG(NULL)\n"); @@ -191,11 +177,8 @@ void nilfs_page_bug(struct page *page) } m = page->mapping; - if (m) { - struct inode *inode = NILFS_AS_I(m); - if (inode != NULL) - ino = inode->i_ino; - } + ino = m ? m->host->i_ino : 0; + printk(KERN_CRIT "NILFS_PAGE_BUG(%p): cnt=%d index#=%llu flags=0x%lx " "mapping=%p ino=%lu\n", page, atomic_read(&page->_count), @@ -217,56 +200,6 @@ void nilfs_page_bug(struct page *page) } /** - * nilfs_alloc_private_page - allocate a private page with buffer heads - * - * Return Value: On success, a pointer to the allocated page is returned. - * On error, NULL is returned. - */ -struct page *nilfs_alloc_private_page(struct block_device *bdev, int size, - unsigned long state) -{ - struct buffer_head *bh, *head, *tail; - struct page *page; - - page = alloc_page(GFP_NOFS); /* page_count of the returned page is 1 */ - if (unlikely(!page)) - return NULL; - - lock_page(page); - head = alloc_page_buffers(page, size, 0); - if (unlikely(!head)) { - unlock_page(page); - __free_page(page); - return NULL; - } - - bh = head; - do { - bh->b_state = (1UL << BH_NILFS_Allocated) | state; - tail = bh; - bh->b_bdev = bdev; - bh = bh->b_this_page; - } while (bh); - - tail->b_this_page = head; - attach_page_buffers(page, head); - - return page; -} - -void nilfs_free_private_page(struct page *page) -{ - BUG_ON(!PageLocked(page)); - BUG_ON(page->mapping); - - if (page_has_buffers(page) && !try_to_free_buffers(page)) - NILFS_PAGE_BUG(page, "failed to free page"); - - unlock_page(page); - __free_page(page); -} - -/** * nilfs_copy_page -- copy the page with buffers * @dst: destination page * @src: source page @@ -492,10 +425,10 @@ unsigned nilfs_page_count_clean_buffers(struct page *page, return nc; } -void nilfs_mapping_init(struct address_space *mapping, +void nilfs_mapping_init(struct address_space *mapping, struct inode *inode, struct backing_dev_info *bdi) { - mapping->host = NULL; + mapping->host = inode; mapping->flags = 0; mapping_set_gfp_mask(mapping, GFP_NOFS); mapping->assoc_mapping = NULL; diff --git a/fs/nilfs2/page.h b/fs/nilfs2/page.h index f06b79ad7493..fb7de71605a0 100644 --- a/fs/nilfs2/page.h +++ b/fs/nilfs2/page.h @@ -38,14 +38,12 @@ enum { BH_NILFS_Redirected, }; -BUFFER_FNS(NILFS_Allocated, nilfs_allocated) /* nilfs private buffers */ BUFFER_FNS(NILFS_Node, nilfs_node) /* nilfs node buffers */ BUFFER_FNS(NILFS_Volatile, nilfs_volatile) BUFFER_FNS(NILFS_Checked, nilfs_checked) /* buffer is verified */ BUFFER_FNS(NILFS_Redirected, nilfs_redirected) /* redirected to a copy */ -void nilfs_mark_buffer_dirty(struct buffer_head *bh); int __nilfs_clear_page_dirty(struct page *); struct buffer_head *nilfs_grab_buffer(struct inode *, struct address_space *, @@ -54,14 +52,11 @@ void nilfs_forget_buffer(struct buffer_head *); void nilfs_copy_buffer(struct buffer_head *, struct buffer_head *); int nilfs_page_buffers_clean(struct page *); void nilfs_page_bug(struct page *); -struct page *nilfs_alloc_private_page(struct block_device *, int, - unsigned long); -void nilfs_free_private_page(struct page *); int nilfs_copy_dirty_pages(struct address_space *, struct address_space *); void nilfs_copy_back_pages(struct address_space *, struct address_space *); void nilfs_clear_dirty_pages(struct address_space *); -void nilfs_mapping_init(struct address_space *mapping, +void nilfs_mapping_init(struct address_space *mapping, struct inode *inode, struct backing_dev_info *bdi); unsigned nilfs_page_count_clean_buffers(struct page *, unsigned, unsigned); unsigned long nilfs_find_uncommitted_extent(struct inode *inode, diff --git a/fs/nilfs2/recovery.c b/fs/nilfs2/recovery.c index ba4a64518f38..a604ac0331b2 100644 --- a/fs/nilfs2/recovery.c +++ b/fs/nilfs2/recovery.c @@ -387,9 +387,9 @@ static int nilfs_scan_dsync_log(struct the_nilfs *nilfs, sector_t start_blocknr, static void dispose_recovery_list(struct list_head *head) { while (!list_empty(head)) { - struct nilfs_recovery_block *rb - = list_entry(head->next, - struct nilfs_recovery_block, list); + struct nilfs_recovery_block *rb; + + rb = list_first_entry(head, struct nilfs_recovery_block, list); list_del(&rb->list); kfree(rb); } @@ -416,9 +416,9 @@ static int nilfs_segment_list_add(struct list_head *head, __u64 segnum) void nilfs_dispose_segment_list(struct list_head *head) { while (!list_empty(head)) { - struct nilfs_segment_entry *ent - = list_entry(head->next, - struct nilfs_segment_entry, list); + struct nilfs_segment_entry *ent; + + ent = list_first_entry(head, struct nilfs_segment_entry, list); list_del(&ent->list); kfree(ent); } diff --git a/fs/nilfs2/segbuf.c b/fs/nilfs2/segbuf.c index 2853ff20f85a..850a7c0228fb 100644 --- a/fs/nilfs2/segbuf.c +++ b/fs/nilfs2/segbuf.c @@ -239,12 +239,15 @@ nilfs_segbuf_fill_in_super_root_crc(struct nilfs_segment_buffer *segbuf, u32 seed) { struct nilfs_super_root *raw_sr; + struct the_nilfs *nilfs = segbuf->sb_super->s_fs_info; + unsigned srsize; u32 crc; raw_sr = (struct nilfs_super_root *)segbuf->sb_super_root->b_data; + srsize = NILFS_SR_BYTES(nilfs->ns_inode_size); crc = crc32_le(seed, (unsigned char *)raw_sr + sizeof(raw_sr->sr_sum), - NILFS_SR_BYTES - sizeof(raw_sr->sr_sum)); + srsize - sizeof(raw_sr->sr_sum)); raw_sr->sr_sum = cpu_to_le32(crc); } @@ -254,18 +257,6 @@ static void nilfs_release_buffers(struct list_head *list) list_for_each_entry_safe(bh, n, list, b_assoc_buffers) { list_del_init(&bh->b_assoc_buffers); - if (buffer_nilfs_allocated(bh)) { - struct page *clone_page = bh->b_page; - - /* remove clone page */ - brelse(bh); - page_cache_release(clone_page); /* for each bh */ - if (page_count(clone_page) <= 2) { - lock_page(clone_page); - nilfs_free_private_page(clone_page); - } - continue; - } brelse(bh); } } diff --git a/fs/nilfs2/segment.c b/fs/nilfs2/segment.c index afe4f2183454..141646e88fb5 100644 --- a/fs/nilfs2/segment.c +++ b/fs/nilfs2/segment.c @@ -655,13 +655,10 @@ static size_t nilfs_lookup_dirty_data_buffers(struct inode *inode, if (unlikely(page->index > last)) break; - if (mapping->host) { - lock_page(page); - if (!page_has_buffers(page)) - create_empty_buffers(page, - 1 << inode->i_blkbits, 0); - unlock_page(page); - } + lock_page(page); + if (!page_has_buffers(page)) + create_empty_buffers(page, 1 << inode->i_blkbits, 0); + unlock_page(page); bh = head = page_buffers(page); do { @@ -809,7 +806,7 @@ static int nilfs_segctor_create_checkpoint(struct nilfs_sc_info *sci) /* The following code is duplicated with cpfile. But, it is needed to collect the checkpoint even if it was not newly created */ - nilfs_mdt_mark_buffer_dirty(bh_cp); + mark_buffer_dirty(bh_cp); nilfs_mdt_mark_dirty(nilfs->ns_cpfile); nilfs_cpfile_put_checkpoint( nilfs->ns_cpfile, nilfs->ns_cno, bh_cp); @@ -889,12 +886,14 @@ static void nilfs_segctor_fill_in_super_root(struct nilfs_sc_info *sci, { struct buffer_head *bh_sr; struct nilfs_super_root *raw_sr; - unsigned isz = nilfs->ns_inode_size; + unsigned isz, srsz; bh_sr = NILFS_LAST_SEGBUF(&sci->sc_segbufs)->sb_super_root; raw_sr = (struct nilfs_super_root *)bh_sr->b_data; + isz = nilfs->ns_inode_size; + srsz = NILFS_SR_BYTES(isz); - raw_sr->sr_bytes = cpu_to_le16(NILFS_SR_BYTES); + raw_sr->sr_bytes = cpu_to_le16(srsz); raw_sr->sr_nongc_ctime = cpu_to_le64(nilfs_doing_gc() ? nilfs->ns_nongc_ctime : sci->sc_seg_ctime); @@ -906,6 +905,7 @@ static void nilfs_segctor_fill_in_super_root(struct nilfs_sc_info *sci, NILFS_SR_CPFILE_OFFSET(isz), 1); nilfs_write_inode_common(nilfs->ns_sufile, (void *)raw_sr + NILFS_SR_SUFILE_OFFSET(isz), 1); + memset((void *)raw_sr + srsz, 0, nilfs->ns_blocksize - srsz); } static void nilfs_redirty_inodes(struct list_head *head) @@ -954,8 +954,8 @@ static int nilfs_segctor_apply_buffers(struct nilfs_sc_info *sci, dispose_buffers: while (!list_empty(listp)) { - bh = list_entry(listp->next, struct buffer_head, - b_assoc_buffers); + bh = list_first_entry(listp, struct buffer_head, + b_assoc_buffers); list_del_init(&bh->b_assoc_buffers); brelse(bh); } @@ -1500,10 +1500,7 @@ nilfs_segctor_update_payload_blocknr(struct nilfs_sc_info *sci, nblocks = le32_to_cpu(finfo->fi_nblocks); ndatablk = le32_to_cpu(finfo->fi_ndatablk); - if (buffer_nilfs_node(bh)) - inode = NILFS_BTNC_I(bh->b_page->mapping); - else - inode = NILFS_AS_I(bh->b_page->mapping); + inode = bh->b_page->mapping->host; if (mode == SC_LSEG_DSYNC) sc_op = &nilfs_sc_dsync_ops; @@ -1556,83 +1553,24 @@ static int nilfs_segctor_assign(struct nilfs_sc_info *sci, int mode) return 0; } -static int -nilfs_copy_replace_page_buffers(struct page *page, struct list_head *out) -{ - struct page *clone_page; - struct buffer_head *bh, *head, *bh2; - void *kaddr; - - bh = head = page_buffers(page); - - clone_page = nilfs_alloc_private_page(bh->b_bdev, bh->b_size, 0); - if (unlikely(!clone_page)) - return -ENOMEM; - - bh2 = page_buffers(clone_page); - kaddr = kmap_atomic(page, KM_USER0); - do { - if (list_empty(&bh->b_assoc_buffers)) - continue; - get_bh(bh2); - page_cache_get(clone_page); /* for each bh */ - memcpy(bh2->b_data, kaddr + bh_offset(bh), bh2->b_size); - bh2->b_blocknr = bh->b_blocknr; - list_replace(&bh->b_assoc_buffers, &bh2->b_assoc_buffers); - list_add_tail(&bh->b_assoc_buffers, out); - } while (bh = bh->b_this_page, bh2 = bh2->b_this_page, bh != head); - kunmap_atomic(kaddr, KM_USER0); - - if (!TestSetPageWriteback(clone_page)) - account_page_writeback(clone_page); - unlock_page(clone_page); - - return 0; -} - -static int nilfs_test_page_to_be_frozen(struct page *page) -{ - struct address_space *mapping = page->mapping; - - if (!mapping || !mapping->host || S_ISDIR(mapping->host->i_mode)) - return 0; - - if (page_mapped(page)) { - ClearPageChecked(page); - return 1; - } - return PageChecked(page); -} - -static int nilfs_begin_page_io(struct page *page, struct list_head *out) +static void nilfs_begin_page_io(struct page *page) { if (!page || PageWriteback(page)) /* For split b-tree node pages, this function may be called twice. We ignore the 2nd or later calls by this check. */ - return 0; + return; lock_page(page); clear_page_dirty_for_io(page); set_page_writeback(page); unlock_page(page); - - if (nilfs_test_page_to_be_frozen(page)) { - int err = nilfs_copy_replace_page_buffers(page, out); - if (unlikely(err)) - return err; - } - return 0; } -static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, - struct page **failed_page) +static void nilfs_segctor_prepare_write(struct nilfs_sc_info *sci) { struct nilfs_segment_buffer *segbuf; struct page *bd_page = NULL, *fs_page = NULL; - struct list_head *list = &sci->sc_copied_buffers; - int err; - *failed_page = NULL; list_for_each_entry(segbuf, &sci->sc_segbufs, sb_list) { struct buffer_head *bh; @@ -1662,11 +1600,7 @@ static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, break; } if (bh->b_page != fs_page) { - err = nilfs_begin_page_io(fs_page, list); - if (unlikely(err)) { - *failed_page = fs_page; - goto out; - } + nilfs_begin_page_io(fs_page); fs_page = bh->b_page; } } @@ -1677,11 +1611,7 @@ static int nilfs_segctor_prepare_write(struct nilfs_sc_info *sci, set_page_writeback(bd_page); unlock_page(bd_page); } - err = nilfs_begin_page_io(fs_page, list); - if (unlikely(err)) - *failed_page = fs_page; - out: - return err; + nilfs_begin_page_io(fs_page); } static int nilfs_segctor_write(struct nilfs_sc_info *sci, @@ -1694,24 +1624,6 @@ static int nilfs_segctor_write(struct nilfs_sc_info *sci, return ret; } -static void __nilfs_end_page_io(struct page *page, int err) -{ - if (!err) { - if (!nilfs_page_buffers_clean(page)) - __set_page_dirty_nobuffers(page); - ClearPageError(page); - } else { - __set_page_dirty_nobuffers(page); - SetPageError(page); - } - - if (buffer_nilfs_allocated(page_buffers(page))) { - if (TestClearPageWriteback(page)) - dec_zone_page_state(page, NR_WRITEBACK); - } else - end_page_writeback(page); -} - static void nilfs_end_page_io(struct page *page, int err) { if (!page) @@ -1738,40 +1650,19 @@ static void nilfs_end_page_io(struct page *page, int err) return; } - __nilfs_end_page_io(page, err); -} - -static void nilfs_clear_copied_buffers(struct list_head *list, int err) -{ - struct buffer_head *bh, *head; - struct page *page; - - while (!list_empty(list)) { - bh = list_entry(list->next, struct buffer_head, - b_assoc_buffers); - page = bh->b_page; - page_cache_get(page); - head = bh = page_buffers(page); - do { - if (!list_empty(&bh->b_assoc_buffers)) { - list_del_init(&bh->b_assoc_buffers); - if (!err) { - set_buffer_uptodate(bh); - clear_buffer_dirty(bh); - clear_buffer_delay(bh); - clear_buffer_nilfs_volatile(bh); - } - brelse(bh); /* for b_assoc_buffers */ - } - } while ((bh = bh->b_this_page) != head); - - __nilfs_end_page_io(page, err); - page_cache_release(page); + if (!err) { + if (!nilfs_page_buffers_clean(page)) + __set_page_dirty_nobuffers(page); + ClearPageError(page); + } else { + __set_page_dirty_nobuffers(page); + SetPageError(page); } + + end_page_writeback(page); } -static void nilfs_abort_logs(struct list_head *logs, struct page *failed_page, - int err) +static void nilfs_abort_logs(struct list_head *logs, int err) { struct nilfs_segment_buffer *segbuf; struct page *bd_page = NULL, *fs_page = NULL; @@ -1801,8 +1692,6 @@ static void nilfs_abort_logs(struct list_head *logs, struct page *failed_page, } if (bh->b_page != fs_page) { nilfs_end_page_io(fs_page, err); - if (fs_page && fs_page == failed_page) - return; fs_page = bh->b_page; } } @@ -1821,12 +1710,11 @@ static void nilfs_segctor_abort_construction(struct nilfs_sc_info *sci, list_splice_tail_init(&sci->sc_write_logs, &logs); ret = nilfs_wait_on_logs(&logs); - nilfs_abort_logs(&logs, NULL, ret ? : err); + nilfs_abort_logs(&logs, ret ? : err); list_splice_tail_init(&sci->sc_segbufs, &logs); nilfs_cancel_segusage(&logs, nilfs->ns_sufile); nilfs_free_incomplete_logs(&logs, nilfs); - nilfs_clear_copied_buffers(&sci->sc_copied_buffers, err); if (sci->sc_stage.flags & NILFS_CF_SUFREED) { ret = nilfs_sufile_cancel_freev(nilfs->ns_sufile, @@ -1920,8 +1808,6 @@ static void nilfs_segctor_complete_write(struct nilfs_sc_info *sci) nilfs_end_page_io(fs_page, 0); - nilfs_clear_copied_buffers(&sci->sc_copied_buffers, 0); - nilfs_drop_collected_inodes(&sci->sc_dirty_files); if (nilfs_doing_gc()) @@ -1979,7 +1865,7 @@ static int nilfs_segctor_collect_dirty_files(struct nilfs_sc_info *sci, "failed to get inode block.\n"); return err; } - nilfs_mdt_mark_buffer_dirty(ibh); + mark_buffer_dirty(ibh); nilfs_mdt_mark_dirty(ifile); spin_lock(&nilfs->ns_inode_lock); if (likely(!ii->i_bh)) @@ -1991,8 +1877,7 @@ static int nilfs_segctor_collect_dirty_files(struct nilfs_sc_info *sci, clear_bit(NILFS_I_QUEUED, &ii->i_state); set_bit(NILFS_I_BUSY, &ii->i_state); - list_del(&ii->i_dirty); - list_add_tail(&ii->i_dirty, &sci->sc_dirty_files); + list_move_tail(&ii->i_dirty, &sci->sc_dirty_files); } spin_unlock(&nilfs->ns_inode_lock); @@ -2014,8 +1899,7 @@ static void nilfs_segctor_drop_written_files(struct nilfs_sc_info *sci, clear_bit(NILFS_I_BUSY, &ii->i_state); brelse(ii->i_bh); ii->i_bh = NULL; - list_del(&ii->i_dirty); - list_add_tail(&ii->i_dirty, &ti->ti_garbage); + list_move_tail(&ii->i_dirty, &ti->ti_garbage); } spin_unlock(&nilfs->ns_inode_lock); } @@ -2026,7 +1910,6 @@ static void nilfs_segctor_drop_written_files(struct nilfs_sc_info *sci, static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) { struct the_nilfs *nilfs = sci->sc_super->s_fs_info; - struct page *failed_page; int err; sci->sc_stage.scnt = NILFS_ST_INIT; @@ -2081,11 +1964,7 @@ static int nilfs_segctor_do_construct(struct nilfs_sc_info *sci, int mode) nilfs_segctor_update_segusage(sci, nilfs->ns_sufile); /* Write partial segments */ - err = nilfs_segctor_prepare_write(sci, &failed_page); - if (err) { - nilfs_abort_logs(&sci->sc_segbufs, failed_page, err); - goto failed_to_write; - } + nilfs_segctor_prepare_write(sci); nilfs_add_checksums_on_logs(&sci->sc_segbufs, nilfs->ns_crc_seed); @@ -2687,7 +2566,6 @@ static struct nilfs_sc_info *nilfs_segctor_new(struct super_block *sb, INIT_LIST_HEAD(&sci->sc_segbufs); INIT_LIST_HEAD(&sci->sc_write_logs); INIT_LIST_HEAD(&sci->sc_gc_inodes); - INIT_LIST_HEAD(&sci->sc_copied_buffers); init_timer(&sci->sc_timer); sci->sc_interval = HZ * NILFS_SC_DEFAULT_TIMEOUT; @@ -2741,8 +2619,6 @@ static void nilfs_segctor_destroy(struct nilfs_sc_info *sci) if (flag || !nilfs_segctor_confirm(sci)) nilfs_segctor_write_out(sci); - WARN_ON(!list_empty(&sci->sc_copied_buffers)); - if (!list_empty(&sci->sc_dirty_files)) { nilfs_warning(sci->sc_super, __func__, "dirty file(s) after the final construction\n"); diff --git a/fs/nilfs2/segment.h b/fs/nilfs2/segment.h index 6c02a86745fb..38a1d0013314 100644 --- a/fs/nilfs2/segment.h +++ b/fs/nilfs2/segment.h @@ -92,7 +92,6 @@ struct nilfs_segsum_pointer { * @sc_nblk_inc: Block count of current generation * @sc_dirty_files: List of files to be written * @sc_gc_inodes: List of GC inodes having blocks to be written - * @sc_copied_buffers: List of copied buffers (buffer heads) to freeze data * @sc_freesegs: array of segment numbers to be freed * @sc_nfreesegs: number of segments on @sc_freesegs * @sc_dsync_inode: inode whose data pages are written for a sync operation @@ -136,7 +135,6 @@ struct nilfs_sc_info { struct list_head sc_dirty_files; struct list_head sc_gc_inodes; - struct list_head sc_copied_buffers; __u64 *sc_freesegs; size_t sc_nfreesegs; diff --git a/fs/nilfs2/sufile.c b/fs/nilfs2/sufile.c index 1d6f488ccae8..0a0aba617d8a 100644 --- a/fs/nilfs2/sufile.c +++ b/fs/nilfs2/sufile.c @@ -33,7 +33,9 @@ struct nilfs_sufile_info { struct nilfs_mdt_info mi; - unsigned long ncleansegs; + unsigned long ncleansegs;/* number of clean segments */ + __u64 allocmin; /* lower limit of allocatable segment range */ + __u64 allocmax; /* upper limit of allocatable segment range */ }; static inline struct nilfs_sufile_info *NILFS_SUI(struct inode *sufile) @@ -96,6 +98,13 @@ nilfs_sufile_get_segment_usage_block(struct inode *sufile, __u64 segnum, create, NULL, bhp); } +static int nilfs_sufile_delete_segment_usage_block(struct inode *sufile, + __u64 segnum) +{ + return nilfs_mdt_delete_block(sufile, + nilfs_sufile_get_blkoff(sufile, segnum)); +} + static void nilfs_sufile_mod_counter(struct buffer_head *header_bh, u64 ncleanadd, u64 ndirtyadd) { @@ -108,7 +117,7 @@ static void nilfs_sufile_mod_counter(struct buffer_head *header_bh, le64_add_cpu(&header->sh_ndirtysegs, ndirtyadd); kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(header_bh); + mark_buffer_dirty(header_bh); } /** @@ -248,6 +257,35 @@ int nilfs_sufile_update(struct inode *sufile, __u64 segnum, int create, } /** + * nilfs_sufile_set_alloc_range - limit range of segment to be allocated + * @sufile: inode of segment usage file + * @start: minimum segment number of allocatable region (inclusive) + * @end: maximum segment number of allocatable region (inclusive) + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-ERANGE - invalid segment region + */ +int nilfs_sufile_set_alloc_range(struct inode *sufile, __u64 start, __u64 end) +{ + struct nilfs_sufile_info *sui = NILFS_SUI(sufile); + __u64 nsegs; + int ret = -ERANGE; + + down_write(&NILFS_MDT(sufile)->mi_sem); + nsegs = nilfs_sufile_get_nsegments(sufile); + + if (start <= end && end < nsegs) { + sui->allocmin = start; + sui->allocmax = end; + ret = 0; + } + up_write(&NILFS_MDT(sufile)->mi_sem); + return ret; +} + +/** * nilfs_sufile_alloc - allocate a segment * @sufile: inode of segment usage file * @segnump: pointer to segment number @@ -269,11 +307,12 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) struct buffer_head *header_bh, *su_bh; struct nilfs_sufile_header *header; struct nilfs_segment_usage *su; + struct nilfs_sufile_info *sui = NILFS_SUI(sufile); size_t susz = NILFS_MDT(sufile)->mi_entry_size; __u64 segnum, maxsegnum, last_alloc; void *kaddr; - unsigned long nsegments, ncleansegs, nsus; - int ret, i, j; + unsigned long nsegments, ncleansegs, nsus, cnt; + int ret, j; down_write(&NILFS_MDT(sufile)->mi_sem); @@ -287,13 +326,31 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) kunmap_atomic(kaddr, KM_USER0); nsegments = nilfs_sufile_get_nsegments(sufile); + maxsegnum = sui->allocmax; segnum = last_alloc + 1; - maxsegnum = nsegments - 1; - for (i = 0; i < nsegments; i += nsus) { - if (segnum >= nsegments) { - /* wrap around */ - segnum = 0; - maxsegnum = last_alloc; + if (segnum < sui->allocmin || segnum > sui->allocmax) + segnum = sui->allocmin; + + for (cnt = 0; cnt < nsegments; cnt += nsus) { + if (segnum > maxsegnum) { + if (cnt < sui->allocmax - sui->allocmin + 1) { + /* + * wrap around in the limited region. + * if allocation started from + * sui->allocmin, this never happens. + */ + segnum = sui->allocmin; + maxsegnum = last_alloc; + } else if (segnum > sui->allocmin && + sui->allocmax + 1 < nsegments) { + segnum = sui->allocmax + 1; + maxsegnum = nsegments - 1; + } else if (sui->allocmin > 0) { + segnum = 0; + maxsegnum = sui->allocmin - 1; + } else { + break; /* never happens */ + } } ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 1, &su_bh); @@ -319,9 +376,9 @@ int nilfs_sufile_alloc(struct inode *sufile, __u64 *segnump) header->sh_last_alloc = cpu_to_le64(segnum); kunmap_atomic(kaddr, KM_USER0); - NILFS_SUI(sufile)->ncleansegs--; - nilfs_mdt_mark_buffer_dirty(header_bh); - nilfs_mdt_mark_buffer_dirty(su_bh); + sui->ncleansegs--; + mark_buffer_dirty(header_bh); + mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); brelse(su_bh); *segnump = segnum; @@ -364,7 +421,7 @@ void nilfs_sufile_do_cancel_free(struct inode *sufile, __u64 segnum, nilfs_sufile_mod_counter(header_bh, -1, 1); NILFS_SUI(sufile)->ncleansegs--; - nilfs_mdt_mark_buffer_dirty(su_bh); + mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } @@ -395,7 +452,7 @@ void nilfs_sufile_do_scrap(struct inode *sufile, __u64 segnum, nilfs_sufile_mod_counter(header_bh, clean ? (u64)-1 : 0, dirty ? 0 : 1); NILFS_SUI(sufile)->ncleansegs -= clean; - nilfs_mdt_mark_buffer_dirty(su_bh); + mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } @@ -421,7 +478,7 @@ void nilfs_sufile_do_free(struct inode *sufile, __u64 segnum, sudirty = nilfs_segment_usage_dirty(su); nilfs_segment_usage_set_clean(su); kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(su_bh); + mark_buffer_dirty(su_bh); nilfs_sufile_mod_counter(header_bh, 1, sudirty ? (u64)-1 : 0); NILFS_SUI(sufile)->ncleansegs++; @@ -441,7 +498,7 @@ int nilfs_sufile_mark_dirty(struct inode *sufile, __u64 segnum) ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, &bh); if (!ret) { - nilfs_mdt_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(sufile); brelse(bh); } @@ -476,7 +533,7 @@ int nilfs_sufile_set_segment_usage(struct inode *sufile, __u64 segnum, su->su_nblocks = cpu_to_le32(nblocks); kunmap_atomic(kaddr, KM_USER0); - nilfs_mdt_mark_buffer_dirty(bh); + mark_buffer_dirty(bh); nilfs_mdt_mark_dirty(sufile); brelse(bh); @@ -505,7 +562,7 @@ int nilfs_sufile_get_stat(struct inode *sufile, struct nilfs_sustat *sustat) { struct buffer_head *header_bh; struct nilfs_sufile_header *header; - struct the_nilfs *nilfs = NILFS_I_NILFS(sufile); + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; void *kaddr; int ret; @@ -555,11 +612,183 @@ void nilfs_sufile_do_set_error(struct inode *sufile, __u64 segnum, nilfs_sufile_mod_counter(header_bh, -1, 0); NILFS_SUI(sufile)->ncleansegs--; } - nilfs_mdt_mark_buffer_dirty(su_bh); + mark_buffer_dirty(su_bh); nilfs_mdt_mark_dirty(sufile); } /** + * nilfs_sufile_truncate_range - truncate range of segment array + * @sufile: inode of segment usage file + * @start: start segment number (inclusive) + * @end: end segment number (inclusive) + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-EINVAL - Invalid number of segments specified + * + * %-EBUSY - Dirty or active segments are present in the range + */ +static int nilfs_sufile_truncate_range(struct inode *sufile, + __u64 start, __u64 end) +{ + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; + struct buffer_head *header_bh; + struct buffer_head *su_bh; + struct nilfs_segment_usage *su, *su2; + size_t susz = NILFS_MDT(sufile)->mi_entry_size; + unsigned long segusages_per_block; + unsigned long nsegs, ncleaned; + __u64 segnum; + void *kaddr; + ssize_t n, nc; + int ret; + int j; + + nsegs = nilfs_sufile_get_nsegments(sufile); + + ret = -EINVAL; + if (start > end || start >= nsegs) + goto out; + + ret = nilfs_sufile_get_header_block(sufile, &header_bh); + if (ret < 0) + goto out; + + segusages_per_block = nilfs_sufile_segment_usages_per_block(sufile); + ncleaned = 0; + + for (segnum = start; segnum <= end; segnum += n) { + n = min_t(unsigned long, + segusages_per_block - + nilfs_sufile_get_offset(sufile, segnum), + end - segnum + 1); + ret = nilfs_sufile_get_segment_usage_block(sufile, segnum, 0, + &su_bh); + if (ret < 0) { + if (ret != -ENOENT) + goto out_header; + /* hole */ + continue; + } + kaddr = kmap_atomic(su_bh->b_page, KM_USER0); + su = nilfs_sufile_block_get_segment_usage( + sufile, segnum, su_bh, kaddr); + su2 = su; + for (j = 0; j < n; j++, su = (void *)su + susz) { + if ((le32_to_cpu(su->su_flags) & + ~(1UL << NILFS_SEGMENT_USAGE_ERROR)) || + nilfs_segment_is_active(nilfs, segnum + j)) { + ret = -EBUSY; + kunmap_atomic(kaddr, KM_USER0); + brelse(su_bh); + goto out_header; + } + } + nc = 0; + for (su = su2, j = 0; j < n; j++, su = (void *)su + susz) { + if (nilfs_segment_usage_error(su)) { + nilfs_segment_usage_set_clean(su); + nc++; + } + } + kunmap_atomic(kaddr, KM_USER0); + if (nc > 0) { + mark_buffer_dirty(su_bh); + ncleaned += nc; + } + brelse(su_bh); + + if (n == segusages_per_block) { + /* make hole */ + nilfs_sufile_delete_segment_usage_block(sufile, segnum); + } + } + ret = 0; + +out_header: + if (ncleaned > 0) { + NILFS_SUI(sufile)->ncleansegs += ncleaned; + nilfs_sufile_mod_counter(header_bh, ncleaned, 0); + nilfs_mdt_mark_dirty(sufile); + } + brelse(header_bh); +out: + return ret; +} + +/** + * nilfs_sufile_resize - resize segment array + * @sufile: inode of segment usage file + * @newnsegs: new number of segments + * + * Return Value: On success, 0 is returned. On error, one of the + * following negative error codes is returned. + * + * %-EIO - I/O error. + * + * %-ENOMEM - Insufficient amount of memory available. + * + * %-ENOSPC - Enough free space is not left for shrinking + * + * %-EBUSY - Dirty or active segments exist in the region to be truncated + */ +int nilfs_sufile_resize(struct inode *sufile, __u64 newnsegs) +{ + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; + struct buffer_head *header_bh; + struct nilfs_sufile_header *header; + struct nilfs_sufile_info *sui = NILFS_SUI(sufile); + void *kaddr; + unsigned long nsegs, nrsvsegs; + int ret = 0; + + down_write(&NILFS_MDT(sufile)->mi_sem); + + nsegs = nilfs_sufile_get_nsegments(sufile); + if (nsegs == newnsegs) + goto out; + + ret = -ENOSPC; + nrsvsegs = nilfs_nrsvsegs(nilfs, newnsegs); + if (newnsegs < nsegs && nsegs - newnsegs + nrsvsegs > sui->ncleansegs) + goto out; + + ret = nilfs_sufile_get_header_block(sufile, &header_bh); + if (ret < 0) + goto out; + + if (newnsegs > nsegs) { + sui->ncleansegs += newnsegs - nsegs; + } else /* newnsegs < nsegs */ { + ret = nilfs_sufile_truncate_range(sufile, newnsegs, nsegs - 1); + if (ret < 0) + goto out_header; + + sui->ncleansegs -= nsegs - newnsegs; + } + + kaddr = kmap_atomic(header_bh->b_page, KM_USER0); + header = kaddr + bh_offset(header_bh); + header->sh_ncleansegs = cpu_to_le64(sui->ncleansegs); + kunmap_atomic(kaddr, KM_USER0); + + mark_buffer_dirty(header_bh); + nilfs_mdt_mark_dirty(sufile); + nilfs_set_nsegments(nilfs, newnsegs); + +out_header: + brelse(header_bh); +out: + up_write(&NILFS_MDT(sufile)->mi_sem); + return ret; +} + +/** * nilfs_sufile_get_suinfo - * @sufile: inode of segment usage file * @segnum: segment number to start looking @@ -583,7 +812,7 @@ ssize_t nilfs_sufile_get_suinfo(struct inode *sufile, __u64 segnum, void *buf, struct nilfs_segment_usage *su; struct nilfs_suinfo *si = buf; size_t susz = NILFS_MDT(sufile)->mi_entry_size; - struct the_nilfs *nilfs = NILFS_I_NILFS(sufile); + struct the_nilfs *nilfs = sufile->i_sb->s_fs_info; void *kaddr; unsigned long nsegs, segusages_per_block; ssize_t n; @@ -679,6 +908,9 @@ int nilfs_sufile_read(struct super_block *sb, size_t susize, kunmap_atomic(kaddr, KM_USER0); brelse(header_bh); + sui->allocmax = nilfs_sufile_get_nsegments(sufile) - 1; + sui->allocmin = 0; + unlock_new_inode(sufile); out: *inodep = sufile; diff --git a/fs/nilfs2/sufile.h b/fs/nilfs2/sufile.h index a943fbacb45b..e84bc5b51fc1 100644 --- a/fs/nilfs2/sufile.h +++ b/fs/nilfs2/sufile.h @@ -31,11 +31,12 @@ static inline unsigned long nilfs_sufile_get_nsegments(struct inode *sufile) { - return NILFS_I_NILFS(sufile)->ns_nsegments; + return ((struct the_nilfs *)sufile->i_sb->s_fs_info)->ns_nsegments; } unsigned long nilfs_sufile_get_ncleansegs(struct inode *sufile); +int nilfs_sufile_set_alloc_range(struct inode *sufile, __u64 start, __u64 end); int nilfs_sufile_alloc(struct inode *, __u64 *); int nilfs_sufile_mark_dirty(struct inode *sufile, __u64 segnum); int nilfs_sufile_set_segment_usage(struct inode *sufile, __u64 segnum, @@ -61,6 +62,7 @@ void nilfs_sufile_do_cancel_free(struct inode *, __u64, struct buffer_head *, void nilfs_sufile_do_set_error(struct inode *, __u64, struct buffer_head *, struct buffer_head *); +int nilfs_sufile_resize(struct inode *sufile, __u64 newnsegs); int nilfs_sufile_read(struct super_block *sb, size_t susize, struct nilfs_inode *raw_inode, struct inode **inodep); diff --git a/fs/nilfs2/super.c b/fs/nilfs2/super.c index 062cca065195..8351c44a7320 100644 --- a/fs/nilfs2/super.c +++ b/fs/nilfs2/super.c @@ -56,6 +56,7 @@ #include "btnode.h" #include "page.h" #include "cpfile.h" +#include "sufile.h" /* nilfs_sufile_resize(), nilfs_sufile_set_alloc_range() */ #include "ifile.h" #include "dat.h" #include "segment.h" @@ -165,7 +166,7 @@ struct inode *nilfs_alloc_inode(struct super_block *sb) ii->i_state = 0; ii->i_cno = 0; ii->vfs_inode.i_version = 1; - nilfs_btnode_cache_init(&ii->i_btnode_cache, sb->s_bdi); + nilfs_mapping_init(&ii->i_btnode_cache, &ii->vfs_inode, sb->s_bdi); return &ii->vfs_inode; } @@ -347,6 +348,134 @@ int nilfs_cleanup_super(struct super_block *sb) return ret; } +/** + * nilfs_move_2nd_super - relocate secondary super block + * @sb: super block instance + * @sb2off: new offset of the secondary super block (in bytes) + */ +static int nilfs_move_2nd_super(struct super_block *sb, loff_t sb2off) +{ + struct the_nilfs *nilfs = sb->s_fs_info; + struct buffer_head *nsbh; + struct nilfs_super_block *nsbp; + sector_t blocknr, newblocknr; + unsigned long offset; + int sb2i = -1; /* array index of the secondary superblock */ + int ret = 0; + + /* nilfs->ns_sem must be locked by the caller. */ + if (nilfs->ns_sbh[1] && + nilfs->ns_sbh[1]->b_blocknr > nilfs->ns_first_data_block) { + sb2i = 1; + blocknr = nilfs->ns_sbh[1]->b_blocknr; + } else if (nilfs->ns_sbh[0]->b_blocknr > nilfs->ns_first_data_block) { + sb2i = 0; + blocknr = nilfs->ns_sbh[0]->b_blocknr; + } + if (sb2i >= 0 && (u64)blocknr << nilfs->ns_blocksize_bits == sb2off) + goto out; /* super block location is unchanged */ + + /* Get new super block buffer */ + newblocknr = sb2off >> nilfs->ns_blocksize_bits; + offset = sb2off & (nilfs->ns_blocksize - 1); + nsbh = sb_getblk(sb, newblocknr); + if (!nsbh) { + printk(KERN_WARNING + "NILFS warning: unable to move secondary superblock " + "to block %llu\n", (unsigned long long)newblocknr); + ret = -EIO; + goto out; + } + nsbp = (void *)nsbh->b_data + offset; + memset(nsbp, 0, nilfs->ns_blocksize); + + if (sb2i >= 0) { + memcpy(nsbp, nilfs->ns_sbp[sb2i], nilfs->ns_sbsize); + brelse(nilfs->ns_sbh[sb2i]); + nilfs->ns_sbh[sb2i] = nsbh; + nilfs->ns_sbp[sb2i] = nsbp; + } else if (nilfs->ns_sbh[0]->b_blocknr < nilfs->ns_first_data_block) { + /* secondary super block will be restored to index 1 */ + nilfs->ns_sbh[1] = nsbh; + nilfs->ns_sbp[1] = nsbp; + } else { + brelse(nsbh); + } +out: + return ret; +} + +/** + * nilfs_resize_fs - resize the filesystem + * @sb: super block instance + * @newsize: new size of the filesystem (in bytes) + */ +int nilfs_resize_fs(struct super_block *sb, __u64 newsize) +{ + struct the_nilfs *nilfs = sb->s_fs_info; + struct nilfs_super_block **sbp; + __u64 devsize, newnsegs; + loff_t sb2off; + int ret; + + ret = -ERANGE; + devsize = i_size_read(sb->s_bdev->bd_inode); + if (newsize > devsize) + goto out; + + /* + * Write lock is required to protect some functions depending + * on the number of segments, the number of reserved segments, + * and so forth. + */ + down_write(&nilfs->ns_segctor_sem); + + sb2off = NILFS_SB2_OFFSET_BYTES(newsize); + newnsegs = sb2off >> nilfs->ns_blocksize_bits; + do_div(newnsegs, nilfs->ns_blocks_per_segment); + + ret = nilfs_sufile_resize(nilfs->ns_sufile, newnsegs); + up_write(&nilfs->ns_segctor_sem); + if (ret < 0) + goto out; + + ret = nilfs_construct_segment(sb); + if (ret < 0) + goto out; + + down_write(&nilfs->ns_sem); + nilfs_move_2nd_super(sb, sb2off); + ret = -EIO; + sbp = nilfs_prepare_super(sb, 0); + if (likely(sbp)) { + nilfs_set_log_cursor(sbp[0], nilfs); + /* + * Drop NILFS_RESIZE_FS flag for compatibility with + * mount-time resize which may be implemented in a + * future release. + */ + sbp[0]->s_state = cpu_to_le16(le16_to_cpu(sbp[0]->s_state) & + ~NILFS_RESIZE_FS); + sbp[0]->s_dev_size = cpu_to_le64(newsize); + sbp[0]->s_nsegments = cpu_to_le64(nilfs->ns_nsegments); + if (sbp[1]) + memcpy(sbp[1], sbp[0], nilfs->ns_sbsize); + ret = nilfs_commit_super(sb, NILFS_SB_COMMIT_ALL); + } + up_write(&nilfs->ns_sem); + + /* + * Reset the range of allocatable segments last. This order + * is important in the case of expansion because the secondary + * superblock must be protected from log write until migration + * completes. + */ + if (!ret) + nilfs_sufile_set_alloc_range(nilfs->ns_sufile, 0, newnsegs - 1); +out: + return ret; +} + static void nilfs_put_super(struct super_block *sb) { struct the_nilfs *nilfs = sb->s_fs_info; diff --git a/fs/nilfs2/the_nilfs.c b/fs/nilfs2/the_nilfs.c index d2acd1a651f3..d32714094375 100644 --- a/fs/nilfs2/the_nilfs.c +++ b/fs/nilfs2/the_nilfs.c @@ -363,6 +363,24 @@ static unsigned long long nilfs_max_size(unsigned int blkbits) return res; } +/** + * nilfs_nrsvsegs - calculate the number of reserved segments + * @nilfs: nilfs object + * @nsegs: total number of segments + */ +unsigned long nilfs_nrsvsegs(struct the_nilfs *nilfs, unsigned long nsegs) +{ + return max_t(unsigned long, NILFS_MIN_NRSVSEGS, + DIV_ROUND_UP(nsegs * nilfs->ns_r_segments_percentage, + 100)); +} + +void nilfs_set_nsegments(struct the_nilfs *nilfs, unsigned long nsegs) +{ + nilfs->ns_nsegments = nsegs; + nilfs->ns_nrsvsegs = nilfs_nrsvsegs(nilfs, nsegs); +} + static int nilfs_store_disk_layout(struct the_nilfs *nilfs, struct nilfs_super_block *sbp) { @@ -389,13 +407,9 @@ static int nilfs_store_disk_layout(struct the_nilfs *nilfs, } nilfs->ns_first_data_block = le64_to_cpu(sbp->s_first_data_block); - nilfs->ns_nsegments = le64_to_cpu(sbp->s_nsegments); nilfs->ns_r_segments_percentage = le32_to_cpu(sbp->s_r_segments_percentage); - nilfs->ns_nrsvsegs = - max_t(unsigned long, NILFS_MIN_NRSVSEGS, - DIV_ROUND_UP(nilfs->ns_nsegments * - nilfs->ns_r_segments_percentage, 100)); + nilfs_set_nsegments(nilfs, le64_to_cpu(sbp->s_nsegments)); nilfs->ns_crc_seed = le32_to_cpu(sbp->s_crc_seed); return 0; } diff --git a/fs/nilfs2/the_nilfs.h b/fs/nilfs2/the_nilfs.h index f4968145c2a3..9992b11312ff 100644 --- a/fs/nilfs2/the_nilfs.h +++ b/fs/nilfs2/the_nilfs.h @@ -268,6 +268,8 @@ struct the_nilfs *alloc_nilfs(struct block_device *bdev); void destroy_nilfs(struct the_nilfs *nilfs); int init_nilfs(struct the_nilfs *nilfs, struct super_block *sb, char *data); int load_nilfs(struct the_nilfs *nilfs, struct super_block *sb); +unsigned long nilfs_nrsvsegs(struct the_nilfs *nilfs, unsigned long nsegs); +void nilfs_set_nsegments(struct the_nilfs *nilfs, unsigned long nsegs); int nilfs_discard_segments(struct the_nilfs *, __u64 *, size_t); int nilfs_count_free_blocks(struct the_nilfs *, sector_t *); struct nilfs_root *nilfs_lookup_root(struct the_nilfs *nilfs, __u64 cno); diff --git a/fs/notify/fanotify/fanotify_user.c b/fs/notify/fanotify/fanotify_user.c index 6b1305dc26c0..9fde1c00a296 100644 --- a/fs/notify/fanotify/fanotify_user.c +++ b/fs/notify/fanotify/fanotify_user.c @@ -164,7 +164,7 @@ static int process_access_response(struct fsnotify_group *group, fd, response); /* * make sure the response is valid, if invalid we do nothing and either - * userspace can send a valid responce or we will clean it up after the + * userspace can send a valid response or we will clean it up after the * timeout */ switch (response) { diff --git a/fs/notify/inotify/inotify_fsnotify.c b/fs/notify/inotify/inotify_fsnotify.c index 0348d0c8f65e..e3cbd746f64a 100644 --- a/fs/notify/inotify/inotify_fsnotify.c +++ b/fs/notify/inotify/inotify_fsnotify.c @@ -194,7 +194,7 @@ static int idr_callback(int id, void *p, void *data) static void inotify_free_group_priv(struct fsnotify_group *group) { - /* ideally the idr is empty and we won't hit the BUG in teh callback */ + /* ideally the idr is empty and we won't hit the BUG in the callback */ idr_for_each(&group->inotify_data.idr, idr_callback, group); idr_remove_all(&group->inotify_data.idr); idr_destroy(&group->inotify_data.idr); diff --git a/fs/notify/mark.c b/fs/notify/mark.c index 50c00856f730..252ab1f6452b 100644 --- a/fs/notify/mark.c +++ b/fs/notify/mark.c @@ -24,7 +24,7 @@ * referencing this object. The object typically will live inside the kernel * with a refcnt of 2, one for each list it is on (i_list, g_list). Any task * which can find this object holding the appropriete locks, can take a reference - * and the object itself is guarenteed to survive until the reference is dropped. + * and the object itself is guaranteed to survive until the reference is dropped. * * LOCKING: * There are 3 spinlocks involved with fsnotify inode marks and they MUST diff --git a/fs/ntfs/attrib.c b/fs/ntfs/attrib.c index f5094ee224c1..f14fde2b03d6 100644 --- a/fs/ntfs/attrib.c +++ b/fs/ntfs/attrib.c @@ -197,7 +197,7 @@ err_out: } else if (ctx_needs_reset) { /* * If there is no attribute list, restoring the search context - * is acomplished simply by copying the saved context back over + * is accomplished simply by copying the saved context back over * the caller supplied context. If there is an attribute list, * things are more complicated as we need to deal with mapping * of mft records and resulting potential changes in pointers. @@ -1181,7 +1181,7 @@ not_found: * for, i.e. if one wants to add the attribute to the mft record this is the * correct place to insert its attribute list entry into. * - * When -errno != -ENOENT, an error occured during the lookup. @ctx->attr is + * When -errno != -ENOENT, an error occurred during the lookup. @ctx->attr is * then undefined and in particular you should not rely on it not changing. */ int ntfs_attr_lookup(const ATTR_TYPE type, const ntfschar *name, diff --git a/fs/ntfs/compress.c b/fs/ntfs/compress.c index ef9ed854255c..ee4144ce5d7c 100644 --- a/fs/ntfs/compress.c +++ b/fs/ntfs/compress.c @@ -501,7 +501,7 @@ int ntfs_read_compressed_block(struct page *page) VCN start_vcn = (((s64)index << PAGE_CACHE_SHIFT) & ~cb_size_mask) >> vol->cluster_size_bits; /* - * The first vcn after the last wanted vcn (minumum alignment is again + * The first vcn after the last wanted vcn (minimum alignment is again * PAGE_CACHE_SIZE. */ VCN end_vcn = ((((s64)(index + 1UL) << PAGE_CACHE_SHIFT) + cb_size - 1) diff --git a/fs/ntfs/inode.c b/fs/ntfs/inode.c index 0b56c6b7ec01..c05d6dcf77a4 100644 --- a/fs/ntfs/inode.c +++ b/fs/ntfs/inode.c @@ -622,7 +622,7 @@ static int ntfs_read_locked_inode(struct inode *vi) */ /* Everyone gets all permissions. */ vi->i_mode |= S_IRWXUGO; - /* If read-only, noone gets write permissions. */ + /* If read-only, no one gets write permissions. */ if (IS_RDONLY(vi)) vi->i_mode &= ~S_IWUGO; if (m->flags & MFT_RECORD_IS_DIRECTORY) { @@ -2529,7 +2529,7 @@ retry_truncate: * specifies that the behaviour is unspecified thus we do not * have to do anything. This means that in our implementation * in the rare case that the file is mmap()ped and a write - * occured into the mmap()ped region just beyond the file size + * occurred into the mmap()ped region just beyond the file size * and writepage has not yet been called to write out the page * (which would clear the area beyond the file size) and we now * extend the file size to incorporate this dirty region diff --git a/fs/ntfs/layout.h b/fs/ntfs/layout.h index 8b2549f672bf..faece7190866 100644 --- a/fs/ntfs/layout.h +++ b/fs/ntfs/layout.h @@ -286,7 +286,7 @@ typedef le16 MFT_RECORD_FLAGS; * fragmented. Volume free space includes the empty part of the mft zone and * when the volume's free 88% are used up, the mft zone is shrunk by a factor * of 2, thus making more space available for more files/data. This process is - * repeated everytime there is no more free space except for the mft zone until + * repeated every time there is no more free space except for the mft zone until * there really is no more free space. */ @@ -1657,13 +1657,13 @@ typedef enum { * pointed to by the Owner field was provided by a defaulting mechanism * rather than explicitly provided by the original provider of the * security descriptor. This may affect the treatment of the SID with - * respect to inheritence of an owner. + * respect to inheritance of an owner. * * SE_GROUP_DEFAULTED - This boolean flag, when set, indicates that the SID in * the Group field was provided by a defaulting mechanism rather than * explicitly provided by the original provider of the security * descriptor. This may affect the treatment of the SID with respect to - * inheritence of a primary group. + * inheritance of a primary group. * * SE_DACL_PRESENT - This boolean flag, when set, indicates that the security * descriptor contains a discretionary ACL. If this flag is set and the @@ -1674,7 +1674,7 @@ typedef enum { * pointed to by the Dacl field was provided by a defaulting mechanism * rather than explicitly provided by the original provider of the * security descriptor. This may affect the treatment of the ACL with - * respect to inheritence of an ACL. This flag is ignored if the + * respect to inheritance of an ACL. This flag is ignored if the * DaclPresent flag is not set. * * SE_SACL_PRESENT - This boolean flag, when set, indicates that the security @@ -1686,7 +1686,7 @@ typedef enum { * pointed to by the Sacl field was provided by a defaulting mechanism * rather than explicitly provided by the original provider of the * security descriptor. This may affect the treatment of the ACL with - * respect to inheritence of an ACL. This flag is ignored if the + * respect to inheritance of an ACL. This flag is ignored if the * SaclPresent flag is not set. * * SE_SELF_RELATIVE - This boolean flag, when set, indicates that the security @@ -2283,7 +2283,7 @@ typedef struct { // the key_length is zero, then the vcn immediately // follows the INDEX_ENTRY_HEADER. Regardless of // key_length, the address of the 8-byte boundary - // alligned vcn of INDEX_ENTRY{_HEADER} *ie is given by + // aligned vcn of INDEX_ENTRY{_HEADER} *ie is given by // (char*)ie + le16_to_cpu(ie*)->length) - sizeof(VCN), // where sizeof(VCN) can be hardcoded as 8 if wanted. */ } __attribute__ ((__packed__)) INDEX_ENTRY; diff --git a/fs/ntfs/logfile.c b/fs/ntfs/logfile.c index 4dadcdf3d451..c71de292c5ad 100644 --- a/fs/ntfs/logfile.c +++ b/fs/ntfs/logfile.c @@ -669,7 +669,7 @@ err_out: * of cases where we think that a volume is dirty when in fact it is clean. * This should only affect volumes that have not been shutdown cleanly but did * not have any pending, non-check-pointed i/o, i.e. they were completely idle - * at least for the five seconds preceeding the unclean shutdown. + * at least for the five seconds preceding the unclean shutdown. * * This function assumes that the $LogFile journal has already been consistency * checked by a call to ntfs_check_logfile() and in particular if the $LogFile diff --git a/fs/ntfs/logfile.h b/fs/ntfs/logfile.h index b5a6f08bd35c..aa2b6ac3f0a4 100644 --- a/fs/ntfs/logfile.h +++ b/fs/ntfs/logfile.h @@ -222,7 +222,7 @@ typedef struct { /* 24*/ sle64 file_size; /* Usable byte size of the log file. If the restart_area_offset + the offset of the file_size are > 510 then corruption has - occured. This is the very first check when + occurred. This is the very first check when starting with the restart_area as if it fails it means that some of the above values will be corrupted by the multi sector diff --git a/fs/ntfs/mft.c b/fs/ntfs/mft.c index 326e7475a22a..382857f9c7db 100644 --- a/fs/ntfs/mft.c +++ b/fs/ntfs/mft.c @@ -73,7 +73,7 @@ static inline MFT_RECORD *map_mft_record_page(ntfs_inode *ni) if (index > end_index || (i_size & ~PAGE_CACHE_MASK) < ofs + vol->mft_record_size) { page = ERR_PTR(-ENOENT); - ntfs_error(vol->sb, "Attemt to read mft record 0x%lx, " + ntfs_error(vol->sb, "Attempt to read mft record 0x%lx, " "which is beyond the end of the mft. " "This is probably a bug in the ntfs " "driver.", ni->mft_no); @@ -1442,7 +1442,7 @@ static int ntfs_mft_bitmap_extend_allocation_nolock(ntfs_volume *vol) // Note: It will need to be a special mft record and if none of // those are available it gets rather complicated... ntfs_error(vol->sb, "Not enough space in this mft record to " - "accomodate extended mft bitmap attribute " + "accommodate extended mft bitmap attribute " "extent. Cannot handle this yet."); ret = -EOPNOTSUPP; goto undo_alloc; @@ -1879,7 +1879,7 @@ static int ntfs_mft_data_extend_allocation_nolock(ntfs_volume *vol) // and we would then need to update all references to this mft // record appropriately. This is rather complicated... ntfs_error(vol->sb, "Not enough space in this mft record to " - "accomodate extended mft data attribute " + "accommodate extended mft data attribute " "extent. Cannot handle this yet."); ret = -EOPNOTSUPP; goto undo_alloc; @@ -2357,7 +2357,7 @@ ntfs_inode *ntfs_mft_record_alloc(ntfs_volume *vol, const int mode, } #ifdef DEBUG read_lock_irqsave(&mftbmp_ni->size_lock, flags); - ntfs_debug("Status of mftbmp after initialized extention: " + ntfs_debug("Status of mftbmp after initialized extension: " "allocated_size 0x%llx, data_size 0x%llx, " "initialized_size 0x%llx.", (long long)mftbmp_ni->allocated_size, diff --git a/fs/ntfs/runlist.c b/fs/ntfs/runlist.c index 56a9a6d25a2a..eac7d6788a10 100644 --- a/fs/ntfs/runlist.c +++ b/fs/ntfs/runlist.c @@ -1243,7 +1243,7 @@ err_out: * write. * * This is used when building the mapping pairs array of a runlist to compress - * a given logical cluster number (lcn) or a specific run length to the minumum + * a given logical cluster number (lcn) or a specific run length to the minimum * size possible. * * Return the number of bytes written on success. On error, i.e. the diff --git a/fs/ntfs/super.c b/fs/ntfs/super.c index 29099a07b9fe..b52706da4645 100644 --- a/fs/ntfs/super.c +++ b/fs/ntfs/super.c @@ -458,7 +458,7 @@ static int ntfs_remount(struct super_block *sb, int *flags, char *opt) * the volume on boot and updates them. * * When remounting read-only, mark the volume clean if no volume errors - * have occured. + * have occurred. */ if ((sb->s_flags & MS_RDONLY) && !(*flags & MS_RDONLY)) { static const char *es = ". Cannot remount read-write."; @@ -1269,7 +1269,7 @@ static int check_windows_hibernation_status(ntfs_volume *vol) "hibernated on the volume."); return 0; } - /* A real error occured. */ + /* A real error occurred. */ ntfs_error(vol->sb, "Failed to find inode number for " "hiberfil.sys."); return ret; @@ -1370,7 +1370,7 @@ static bool load_and_init_quota(ntfs_volume *vol) NVolSetQuotaOutOfDate(vol); return true; } - /* A real error occured. */ + /* A real error occurred. */ ntfs_error(vol->sb, "Failed to find inode number for $Quota."); return false; } @@ -1454,7 +1454,7 @@ not_enabled: NVolSetUsnJrnlStamped(vol); return true; } - /* A real error occured. */ + /* A real error occurred. */ ntfs_error(vol->sb, "Failed to find inode number for " "$UsnJrnl."); return false; @@ -2292,7 +2292,7 @@ static void ntfs_put_super(struct super_block *sb) ntfs_commit_inode(vol->mft_ino); /* - * If a read-write mount and no volume errors have occured, mark the + * If a read-write mount and no volume errors have occurred, mark the * volume clean. Also, re-commit all affected inodes. */ if (!(sb->s_flags & MS_RDONLY)) { @@ -2496,7 +2496,7 @@ static s64 get_nr_free_clusters(ntfs_volume *vol) if (vol->nr_clusters & 63) nr_free += 64 - (vol->nr_clusters & 63); up_read(&vol->lcnbmp_lock); - /* If errors occured we may well have gone below zero, fix this. */ + /* If errors occurred we may well have gone below zero, fix this. */ if (nr_free < 0) nr_free = 0; ntfs_debug("Exiting."); @@ -2561,7 +2561,7 @@ static unsigned long __get_nr_free_mft_records(ntfs_volume *vol, } ntfs_debug("Finished reading $MFT/$BITMAP, last index = 0x%lx.", index - 1); - /* If errors occured we may well have gone below zero, fix this. */ + /* If errors occurred we may well have gone below zero, fix this. */ if (nr_free < 0) nr_free = 0; ntfs_debug("Exiting."); diff --git a/fs/ocfs2/Makefile b/fs/ocfs2/Makefile index d8a0313e99e6..f17e58b32989 100644 --- a/fs/ocfs2/Makefile +++ b/fs/ocfs2/Makefile @@ -30,6 +30,7 @@ ocfs2-objs := \ namei.o \ refcounttree.o \ reservations.o \ + move_extents.o \ resize.o \ slot_map.o \ suballoc.o \ diff --git a/fs/ocfs2/alloc.c b/fs/ocfs2/alloc.c index b27a0d86f8c5..ed553c60de82 100644 --- a/fs/ocfs2/alloc.c +++ b/fs/ocfs2/alloc.c @@ -29,6 +29,7 @@ #include <linux/highmem.h> #include <linux/swap.h> #include <linux/quotaops.h> +#include <linux/blkdev.h> #include <cluster/masklog.h> @@ -4519,7 +4520,7 @@ set_tail_append: } /* - * Helper function called at the begining of an insert. + * Helper function called at the beginning of an insert. * * This computes a few things that are commonly used in the process of * inserting into the btree: @@ -7184,3 +7185,168 @@ out_commit: out: return ret; } + +static int ocfs2_trim_extent(struct super_block *sb, + struct ocfs2_group_desc *gd, + u32 start, u32 count) +{ + u64 discard, bcount; + + bcount = ocfs2_clusters_to_blocks(sb, count); + discard = le64_to_cpu(gd->bg_blkno) + + ocfs2_clusters_to_blocks(sb, start); + + trace_ocfs2_trim_extent(sb, (unsigned long long)discard, bcount); + + return sb_issue_discard(sb, discard, bcount, GFP_NOFS, 0); +} + +static int ocfs2_trim_group(struct super_block *sb, + struct ocfs2_group_desc *gd, + u32 start, u32 max, u32 minbits) +{ + int ret = 0, count = 0, next; + void *bitmap = gd->bg_bitmap; + + if (le16_to_cpu(gd->bg_free_bits_count) < minbits) + return 0; + + trace_ocfs2_trim_group((unsigned long long)le64_to_cpu(gd->bg_blkno), + start, max, minbits); + + while (start < max) { + start = ocfs2_find_next_zero_bit(bitmap, max, start); + if (start >= max) + break; + next = ocfs2_find_next_bit(bitmap, max, start); + + if ((next - start) >= minbits) { + ret = ocfs2_trim_extent(sb, gd, + start, next - start); + if (ret < 0) { + mlog_errno(ret); + break; + } + count += next - start; + } + start = next + 1; + + if (fatal_signal_pending(current)) { + count = -ERESTARTSYS; + break; + } + + if ((le16_to_cpu(gd->bg_free_bits_count) - count) < minbits) + break; + } + + if (ret < 0) + count = ret; + + return count; +} + +int ocfs2_trim_fs(struct super_block *sb, struct fstrim_range *range) +{ + struct ocfs2_super *osb = OCFS2_SB(sb); + u64 start, len, trimmed, first_group, last_group, group; + int ret, cnt; + u32 first_bit, last_bit, minlen; + struct buffer_head *main_bm_bh = NULL; + struct inode *main_bm_inode = NULL; + struct buffer_head *gd_bh = NULL; + struct ocfs2_dinode *main_bm; + struct ocfs2_group_desc *gd = NULL; + + start = range->start >> osb->s_clustersize_bits; + len = range->len >> osb->s_clustersize_bits; + minlen = range->minlen >> osb->s_clustersize_bits; + trimmed = 0; + + if (!len) { + range->len = 0; + return 0; + } + + if (minlen >= osb->bitmap_cpg) + return -EINVAL; + + main_bm_inode = ocfs2_get_system_file_inode(osb, + GLOBAL_BITMAP_SYSTEM_INODE, + OCFS2_INVALID_SLOT); + if (!main_bm_inode) { + ret = -EIO; + mlog_errno(ret); + goto out; + } + + mutex_lock(&main_bm_inode->i_mutex); + + ret = ocfs2_inode_lock(main_bm_inode, &main_bm_bh, 0); + if (ret < 0) { + mlog_errno(ret); + goto out_mutex; + } + main_bm = (struct ocfs2_dinode *)main_bm_bh->b_data; + + if (start >= le32_to_cpu(main_bm->i_clusters)) { + ret = -EINVAL; + goto out_unlock; + } + + if (start + len > le32_to_cpu(main_bm->i_clusters)) + len = le32_to_cpu(main_bm->i_clusters) - start; + + trace_ocfs2_trim_fs(start, len, minlen); + + /* Determine first and last group to examine based on start and len */ + first_group = ocfs2_which_cluster_group(main_bm_inode, start); + if (first_group == osb->first_cluster_group_blkno) + first_bit = start; + else + first_bit = start - ocfs2_blocks_to_clusters(sb, first_group); + last_group = ocfs2_which_cluster_group(main_bm_inode, start + len - 1); + last_bit = osb->bitmap_cpg; + + for (group = first_group; group <= last_group;) { + if (first_bit + len >= osb->bitmap_cpg) + last_bit = osb->bitmap_cpg; + else + last_bit = first_bit + len; + + ret = ocfs2_read_group_descriptor(main_bm_inode, + main_bm, group, + &gd_bh); + if (ret < 0) { + mlog_errno(ret); + break; + } + + gd = (struct ocfs2_group_desc *)gd_bh->b_data; + cnt = ocfs2_trim_group(sb, gd, first_bit, last_bit, minlen); + brelse(gd_bh); + gd_bh = NULL; + if (cnt < 0) { + ret = cnt; + mlog_errno(ret); + break; + } + + trimmed += cnt; + len -= osb->bitmap_cpg - first_bit; + first_bit = 0; + if (group == osb->first_cluster_group_blkno) + group = ocfs2_clusters_to_blocks(sb, osb->bitmap_cpg); + else + group += ocfs2_clusters_to_blocks(sb, osb->bitmap_cpg); + } + range->len = trimmed * sb->s_blocksize; +out_unlock: + ocfs2_inode_unlock(main_bm_inode, 0); + brelse(main_bm_bh); +out_mutex: + mutex_unlock(&main_bm_inode->i_mutex); + iput(main_bm_inode); +out: + return ret; +} diff --git a/fs/ocfs2/alloc.h b/fs/ocfs2/alloc.h index 3bd08a03251c..ca381c584127 100644 --- a/fs/ocfs2/alloc.h +++ b/fs/ocfs2/alloc.h @@ -239,6 +239,7 @@ int ocfs2_find_leaf(struct ocfs2_caching_info *ci, struct buffer_head **leaf_bh); int ocfs2_search_extent_list(struct ocfs2_extent_list *el, u32 v_cluster); +int ocfs2_trim_fs(struct super_block *sb, struct fstrim_range *range); /* * Helper function to look at the # of clusters in an extent record. */ diff --git a/fs/ocfs2/aops.h b/fs/ocfs2/aops.h index eceb456037c1..75cf3ad987a6 100644 --- a/fs/ocfs2/aops.h +++ b/fs/ocfs2/aops.h @@ -71,7 +71,7 @@ static inline void ocfs2_iocb_set_rw_locked(struct kiocb *iocb, int level) /* * Using a named enum representing lock types in terms of #N bit stored in - * iocb->private, which is going to be used for communication bewteen + * iocb->private, which is going to be used for communication between * ocfs2_dio_end_io() and ocfs2_file_aio_write/read(). */ enum ocfs2_iocb_lock_bits { diff --git a/fs/ocfs2/cluster/heartbeat.c b/fs/ocfs2/cluster/heartbeat.c index 2461eb3272ed..9a3e6bbff27b 100644 --- a/fs/ocfs2/cluster/heartbeat.c +++ b/fs/ocfs2/cluster/heartbeat.c @@ -539,25 +539,41 @@ static int o2hb_verify_crc(struct o2hb_region *reg, /* We want to make sure that nobody is heartbeating on top of us -- * this will help detect an invalid configuration. */ -static int o2hb_check_last_timestamp(struct o2hb_region *reg) +static void o2hb_check_last_timestamp(struct o2hb_region *reg) { - int node_num, ret; struct o2hb_disk_slot *slot; struct o2hb_disk_heartbeat_block *hb_block; + char *errstr; - node_num = o2nm_this_node(); - - ret = 1; - slot = ®->hr_slots[node_num]; + slot = ®->hr_slots[o2nm_this_node()]; /* Don't check on our 1st timestamp */ - if (slot->ds_last_time) { - hb_block = slot->ds_raw_block; + if (!slot->ds_last_time) + return; - if (le64_to_cpu(hb_block->hb_seq) != slot->ds_last_time) - ret = 0; - } + hb_block = slot->ds_raw_block; + if (le64_to_cpu(hb_block->hb_seq) == slot->ds_last_time && + le64_to_cpu(hb_block->hb_generation) == slot->ds_last_generation && + hb_block->hb_node == slot->ds_node_num) + return; - return ret; +#define ERRSTR1 "Another node is heartbeating on device" +#define ERRSTR2 "Heartbeat generation mismatch on device" +#define ERRSTR3 "Heartbeat sequence mismatch on device" + + if (hb_block->hb_node != slot->ds_node_num) + errstr = ERRSTR1; + else if (le64_to_cpu(hb_block->hb_generation) != + slot->ds_last_generation) + errstr = ERRSTR2; + else + errstr = ERRSTR3; + + mlog(ML_ERROR, "%s (%s): expected(%u:0x%llx, 0x%llx), " + "ondisk(%u:0x%llx, 0x%llx)\n", errstr, reg->hr_dev_name, + slot->ds_node_num, (unsigned long long)slot->ds_last_generation, + (unsigned long long)slot->ds_last_time, hb_block->hb_node, + (unsigned long long)le64_to_cpu(hb_block->hb_generation), + (unsigned long long)le64_to_cpu(hb_block->hb_seq)); } static inline void o2hb_prepare_block(struct o2hb_region *reg, @@ -983,9 +999,7 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg) /* With an up to date view of the slots, we can check that no * other node has been improperly configured to heartbeat in * our slot. */ - if (!o2hb_check_last_timestamp(reg)) - mlog(ML_ERROR, "Device \"%s\": another node is heartbeating " - "in our slot!\n", reg->hr_dev_name); + o2hb_check_last_timestamp(reg); /* fill in the proper info for our next heartbeat */ o2hb_prepare_block(reg, reg->hr_generation); @@ -999,8 +1013,8 @@ static int o2hb_do_disk_heartbeat(struct o2hb_region *reg) } i = -1; - while((i = find_next_bit(configured_nodes, O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES) { - + while((i = find_next_bit(configured_nodes, + O2NM_MAX_NODES, i + 1)) < O2NM_MAX_NODES) { change |= o2hb_check_slot(reg, ®->hr_slots[i]); } @@ -1690,6 +1704,7 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg, struct file *filp = NULL; struct inode *inode = NULL; ssize_t ret = -EINVAL; + int live_threshold; if (reg->hr_bdev) goto out; @@ -1766,8 +1781,18 @@ static ssize_t o2hb_region_dev_write(struct o2hb_region *reg, * A node is considered live after it has beat LIVE_THRESHOLD * times. We're not steady until we've given them a chance * _after_ our first read. + * The default threshold is bare minimum so as to limit the delay + * during mounts. For global heartbeat, the threshold doubled for the + * first region. */ - atomic_set(®->hr_steady_iterations, O2HB_LIVE_THRESHOLD + 1); + live_threshold = O2HB_LIVE_THRESHOLD; + if (o2hb_global_heartbeat_active()) { + spin_lock(&o2hb_live_lock); + if (o2hb_pop_count(&o2hb_region_bitmap, O2NM_MAX_REGIONS) == 1) + live_threshold <<= 1; + spin_unlock(&o2hb_live_lock); + } + atomic_set(®->hr_steady_iterations, live_threshold + 1); hb_task = kthread_run(o2hb_thread, reg, "o2hb-%s", reg->hr_item.ci_name); @@ -2275,7 +2300,7 @@ void o2hb_free_hb_set(struct config_group *group) kfree(hs); } -/* hb callback registration and issueing */ +/* hb callback registration and issuing */ static struct o2hb_callback *hbcall_from_type(enum o2hb_callback_type type) { diff --git a/fs/ocfs2/cluster/quorum.c b/fs/ocfs2/cluster/quorum.c index a87366750f23..8f9cea1597af 100644 --- a/fs/ocfs2/cluster/quorum.c +++ b/fs/ocfs2/cluster/quorum.c @@ -89,7 +89,7 @@ static void o2quo_fence_self(void) }; } -/* Indicate that a timeout occured on a hearbeat region write. The +/* Indicate that a timeout occurred on a hearbeat region write. The * other nodes in the cluster may consider us dead at that time so we * want to "fence" ourselves so that we don't scribble on the disk * after they think they've recovered us. This can't solve all @@ -261,7 +261,7 @@ void o2quo_hb_still_up(u8 node) spin_unlock(&qs->qs_lock); } -/* This is analagous to hb_up. as a node's connection comes up we delay the +/* This is analogous to hb_up. as a node's connection comes up we delay the * quorum decision until we see it heartbeating. the hold will be droped in * hb_up or hb_down. it might be perpetuated by con_err until hb_down. if * it's already heartbeating we we might be dropping a hold that conn_up got. diff --git a/fs/ocfs2/cluster/sys.c b/fs/ocfs2/cluster/sys.c index bc702dab5d1f..a4b07730b2e1 100644 --- a/fs/ocfs2/cluster/sys.c +++ b/fs/ocfs2/cluster/sys.c @@ -57,7 +57,6 @@ static struct kset *o2cb_kset; void o2cb_sys_shutdown(void) { mlog_sys_shutdown(); - sysfs_remove_link(NULL, "o2cb"); kset_unregister(o2cb_kset); } @@ -69,14 +68,6 @@ int o2cb_sys_init(void) if (!o2cb_kset) return -ENOMEM; - /* - * Create this symlink for backwards compatibility with old - * versions of ocfs2-tools which look for things in /sys/o2cb. - */ - ret = sysfs_create_link(NULL, &o2cb_kset->kobj, "o2cb"); - if (ret) - goto error; - ret = sysfs_create_group(&o2cb_kset->kobj, &o2cb_attr_group); if (ret) goto error; diff --git a/fs/ocfs2/cluster/tcp.c b/fs/ocfs2/cluster/tcp.c index ee04ff5ee603..db5ee4b4f47a 100644 --- a/fs/ocfs2/cluster/tcp.c +++ b/fs/ocfs2/cluster/tcp.c @@ -565,7 +565,7 @@ static void o2net_set_nn_state(struct o2net_node *nn, * the work queue actually being up. */ if (!valid && o2net_wq) { unsigned long delay; - /* delay if we're withing a RECONNECT_DELAY of the + /* delay if we're within a RECONNECT_DELAY of the * last attempt */ delay = (nn->nn_last_connect_attempt + msecs_to_jiffies(o2net_reconnect_delay())) diff --git a/fs/ocfs2/dir.c b/fs/ocfs2/dir.c index 9fe5b8fd658f..8582e3f4f120 100644 --- a/fs/ocfs2/dir.c +++ b/fs/ocfs2/dir.c @@ -2868,7 +2868,7 @@ static int ocfs2_expand_inline_dir(struct inode *dir, struct buffer_head *di_bh, bytes = blocks_wanted << sb->s_blocksize_bits; struct ocfs2_super *osb = OCFS2_SB(dir->i_sb); struct ocfs2_inode_info *oi = OCFS2_I(dir); - struct ocfs2_alloc_context *data_ac; + struct ocfs2_alloc_context *data_ac = NULL; struct ocfs2_alloc_context *meta_ac = NULL; struct buffer_head *dirdata_bh = NULL; struct buffer_head *dx_root_bh = NULL; diff --git a/fs/ocfs2/dlm/dlmcommon.h b/fs/ocfs2/dlm/dlmcommon.h index 4bdf7baee344..d602abb51b61 100644 --- a/fs/ocfs2/dlm/dlmcommon.h +++ b/fs/ocfs2/dlm/dlmcommon.h @@ -144,6 +144,7 @@ struct dlm_ctxt wait_queue_head_t dlm_join_events; unsigned long live_nodes_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; unsigned long domain_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; + unsigned long exit_domain_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; unsigned long recovery_map[BITS_TO_LONGS(O2NM_MAX_NODES)]; struct dlm_recovery_ctxt reco; spinlock_t master_lock; @@ -401,6 +402,18 @@ static inline int dlm_lvb_is_empty(char *lvb) return 1; } +static inline char *dlm_list_in_text(enum dlm_lockres_list idx) +{ + if (idx == DLM_GRANTED_LIST) + return "granted"; + else if (idx == DLM_CONVERTING_LIST) + return "converting"; + else if (idx == DLM_BLOCKED_LIST) + return "blocked"; + else + return "unknown"; +} + static inline struct list_head * dlm_list_idx_to_ptr(struct dlm_lock_resource *res, enum dlm_lockres_list idx) { @@ -448,6 +461,7 @@ enum { DLM_FINALIZE_RECO_MSG = 518, DLM_QUERY_REGION = 519, DLM_QUERY_NODEINFO = 520, + DLM_BEGIN_EXIT_DOMAIN_MSG = 521, }; struct dlm_reco_node_data diff --git a/fs/ocfs2/dlm/dlmdebug.c b/fs/ocfs2/dlm/dlmdebug.c index 04a32be0aeb9..56f82cb912e3 100644 --- a/fs/ocfs2/dlm/dlmdebug.c +++ b/fs/ocfs2/dlm/dlmdebug.c @@ -756,6 +756,12 @@ static int debug_state_print(struct dlm_ctxt *dlm, char *buf, int len) buf + out, len - out); out += snprintf(buf + out, len - out, "\n"); + /* Exit Domain Map: xx xx xx */ + out += snprintf(buf + out, len - out, "Exit Domain Map: "); + out += stringify_nodemap(dlm->exit_domain_map, O2NM_MAX_NODES, + buf + out, len - out); + out += snprintf(buf + out, len - out, "\n"); + /* Live Map: xx xx xx */ out += snprintf(buf + out, len - out, "Live Map: "); out += stringify_nodemap(dlm->live_nodes_map, O2NM_MAX_NODES, diff --git a/fs/ocfs2/dlm/dlmdomain.c b/fs/ocfs2/dlm/dlmdomain.c index 7540a492eaba..6ed6b95dcf93 100644 --- a/fs/ocfs2/dlm/dlmdomain.c +++ b/fs/ocfs2/dlm/dlmdomain.c @@ -132,10 +132,12 @@ static DECLARE_WAIT_QUEUE_HEAD(dlm_domain_events); * New in version 1.1: * - Message DLM_QUERY_REGION added to support global heartbeat * - Message DLM_QUERY_NODEINFO added to allow online node removes + * New in version 1.2: + * - Message DLM_BEGIN_EXIT_DOMAIN_MSG added to mark start of exit domain */ static const struct dlm_protocol_version dlm_protocol = { .pv_major = 1, - .pv_minor = 1, + .pv_minor = 2, }; #define DLM_DOMAIN_BACKOFF_MS 200 @@ -449,14 +451,18 @@ redo_bucket: dropped = dlm_empty_lockres(dlm, res); spin_lock(&res->spinlock); - __dlm_lockres_calc_usage(dlm, res); - iter = res->hash_node.next; + if (dropped) + __dlm_lockres_calc_usage(dlm, res); + else + iter = res->hash_node.next; spin_unlock(&res->spinlock); dlm_lockres_put(res); - if (dropped) + if (dropped) { + cond_resched_lock(&dlm->spinlock); goto redo_bucket; + } } cond_resched_lock(&dlm->spinlock); num += n; @@ -486,6 +492,28 @@ static int dlm_no_joining_node(struct dlm_ctxt *dlm) return ret; } +static int dlm_begin_exit_domain_handler(struct o2net_msg *msg, u32 len, + void *data, void **ret_data) +{ + struct dlm_ctxt *dlm = data; + unsigned int node; + struct dlm_exit_domain *exit_msg = (struct dlm_exit_domain *) msg->buf; + + if (!dlm_grab(dlm)) + return 0; + + node = exit_msg->node_idx; + mlog(0, "%s: Node %u sent a begin exit domain message\n", dlm->name, node); + + spin_lock(&dlm->spinlock); + set_bit(node, dlm->exit_domain_map); + spin_unlock(&dlm->spinlock); + + dlm_put(dlm); + + return 0; +} + static void dlm_mark_domain_leaving(struct dlm_ctxt *dlm) { /* Yikes, a double spinlock! I need domain_lock for the dlm @@ -542,6 +570,7 @@ static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data, spin_lock(&dlm->spinlock); clear_bit(node, dlm->domain_map); + clear_bit(node, dlm->exit_domain_map); __dlm_print_nodes(dlm); /* notify anything attached to the heartbeat events */ @@ -554,29 +583,56 @@ static int dlm_exit_domain_handler(struct o2net_msg *msg, u32 len, void *data, return 0; } -static int dlm_send_one_domain_exit(struct dlm_ctxt *dlm, +static int dlm_send_one_domain_exit(struct dlm_ctxt *dlm, u32 msg_type, unsigned int node) { int status; struct dlm_exit_domain leave_msg; - mlog(0, "Asking node %u if we can leave the domain %s me = %u\n", - node, dlm->name, dlm->node_num); + mlog(0, "%s: Sending domain exit message %u to node %u\n", dlm->name, + msg_type, node); memset(&leave_msg, 0, sizeof(leave_msg)); leave_msg.node_idx = dlm->node_num; - status = o2net_send_message(DLM_EXIT_DOMAIN_MSG, dlm->key, - &leave_msg, sizeof(leave_msg), node, - NULL); + status = o2net_send_message(msg_type, dlm->key, &leave_msg, + sizeof(leave_msg), node, NULL); if (status < 0) - mlog(ML_ERROR, "Error %d when sending message %u (key 0x%x) to " - "node %u\n", status, DLM_EXIT_DOMAIN_MSG, dlm->key, node); - mlog(0, "status return %d from o2net_send_message\n", status); + mlog(ML_ERROR, "Error %d sending domain exit message %u " + "to node %u on domain %s\n", status, msg_type, node, + dlm->name); return status; } +static void dlm_begin_exit_domain(struct dlm_ctxt *dlm) +{ + int node = -1; + + /* Support for begin exit domain was added in 1.2 */ + if (dlm->dlm_locking_proto.pv_major == 1 && + dlm->dlm_locking_proto.pv_minor < 2) + return; + + /* + * Unlike DLM_EXIT_DOMAIN_MSG, DLM_BEGIN_EXIT_DOMAIN_MSG is purely + * informational. Meaning if a node does not receive the message, + * so be it. + */ + spin_lock(&dlm->spinlock); + while (1) { + node = find_next_bit(dlm->domain_map, O2NM_MAX_NODES, node + 1); + if (node >= O2NM_MAX_NODES) + break; + if (node == dlm->node_num) + continue; + + spin_unlock(&dlm->spinlock); + dlm_send_one_domain_exit(dlm, DLM_BEGIN_EXIT_DOMAIN_MSG, node); + spin_lock(&dlm->spinlock); + } + spin_unlock(&dlm->spinlock); +} static void dlm_leave_domain(struct dlm_ctxt *dlm) { @@ -602,7 +658,8 @@ static void dlm_leave_domain(struct dlm_ctxt *dlm) clear_node = 1; - status = dlm_send_one_domain_exit(dlm, node); + status = dlm_send_one_domain_exit(dlm, DLM_EXIT_DOMAIN_MSG, + node); if (status < 0 && status != -ENOPROTOOPT && status != -ENOTCONN) { @@ -677,6 +734,7 @@ void dlm_unregister_domain(struct dlm_ctxt *dlm) if (leave) { mlog(0, "shutting down domain %s\n", dlm->name); + dlm_begin_exit_domain(dlm); /* We changed dlm state, notify the thread */ dlm_kick_thread(dlm, NULL); @@ -909,6 +967,7 @@ static int dlm_assert_joined_handler(struct o2net_msg *msg, u32 len, void *data, * leftover join state. */ BUG_ON(dlm->joining_node != assert->node_idx); set_bit(assert->node_idx, dlm->domain_map); + clear_bit(assert->node_idx, dlm->exit_domain_map); __dlm_set_joining_node(dlm, DLM_LOCK_RES_OWNER_UNKNOWN); printk(KERN_NOTICE "o2dlm: Node %u joins domain %s\n", @@ -1614,7 +1673,8 @@ static int dlm_try_to_join_domain(struct dlm_ctxt *dlm) spin_unlock(&dlm->spinlock); /* Support for global heartbeat and node info was added in 1.1 */ - if (dlm_protocol.pv_major > 1 || dlm_protocol.pv_minor > 0) { + if (dlm->dlm_locking_proto.pv_major > 1 || + dlm->dlm_locking_proto.pv_minor > 0) { status = dlm_send_nodeinfo(dlm, ctxt->yes_resp_map); if (status) { mlog_errno(status); @@ -1792,6 +1852,13 @@ static int dlm_register_domain_handlers(struct dlm_ctxt *dlm) if (status) goto bail; + status = o2net_register_handler(DLM_BEGIN_EXIT_DOMAIN_MSG, dlm->key, + sizeof(struct dlm_exit_domain), + dlm_begin_exit_domain_handler, + dlm, NULL, &dlm->dlm_domain_handlers); + if (status) + goto bail; + bail: if (status) dlm_unregister_domain_handlers(dlm); diff --git a/fs/ocfs2/dlm/dlmmaster.c b/fs/ocfs2/dlm/dlmmaster.c index 9d67610dfc74..11eefb8c12e9 100644 --- a/fs/ocfs2/dlm/dlmmaster.c +++ b/fs/ocfs2/dlm/dlmmaster.c @@ -808,7 +808,7 @@ lookup: dlm_mle_detach_hb_events(dlm, mle); dlm_put_mle(mle); mle = NULL; - /* this is lame, but we cant wait on either + /* this is lame, but we can't wait on either * the mle or lockres waitqueue here */ if (mig) msleep(100); @@ -843,7 +843,7 @@ lookup: /* finally add the lockres to its hash bucket */ __dlm_insert_lockres(dlm, res); - /* since this lockres is new it doesnt not require the spinlock */ + /* since this lockres is new it doesn't not require the spinlock */ dlm_lockres_grab_inflight_ref_new(dlm, res); /* if this node does not become the master make sure to drop @@ -2339,65 +2339,55 @@ static void dlm_deref_lockres_worker(struct dlm_work_item *item, void *data) dlm_lockres_put(res); } -/* Checks whether the lockres can be migrated. Returns 0 if yes, < 0 - * if not. If 0, numlocks is set to the number of locks in the lockres. +/* + * A migrateable resource is one that is : + * 1. locally mastered, and, + * 2. zero local locks, and, + * 3. one or more non-local locks, or, one or more references + * Returns 1 if yes, 0 if not. */ static int dlm_is_lockres_migrateable(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, - int *numlocks, - int *hasrefs) + struct dlm_lock_resource *res) { - int ret; - int i; - int count = 0; + enum dlm_lockres_list idx; + int nonlocal = 0, node_ref; struct list_head *queue; struct dlm_lock *lock; + u64 cookie; assert_spin_locked(&res->spinlock); - *numlocks = 0; - *hasrefs = 0; - - ret = -EINVAL; - if (res->owner == DLM_LOCK_RES_OWNER_UNKNOWN) { - mlog(0, "cannot migrate lockres with unknown owner!\n"); - goto leave; - } - - if (res->owner != dlm->node_num) { - mlog(0, "cannot migrate lockres this node doesn't own!\n"); - goto leave; - } + if (res->owner != dlm->node_num) + return 0; - ret = 0; - queue = &res->granted; - for (i = 0; i < 3; i++) { + for (idx = DLM_GRANTED_LIST; idx <= DLM_BLOCKED_LIST; idx++) { + queue = dlm_list_idx_to_ptr(res, idx); list_for_each_entry(lock, queue, list) { - ++count; - if (lock->ml.node == dlm->node_num) { - mlog(0, "found a lock owned by this node still " - "on the %s queue! will not migrate this " - "lockres\n", (i == 0 ? "granted" : - (i == 1 ? "converting" : - "blocked"))); - ret = -ENOTEMPTY; - goto leave; + if (lock->ml.node != dlm->node_num) { + nonlocal++; + continue; } + cookie = be64_to_cpu(lock->ml.cookie); + mlog(0, "%s: Not migrateable res %.*s, lock %u:%llu on " + "%s list\n", dlm->name, res->lockname.len, + res->lockname.name, + dlm_get_lock_cookie_node(cookie), + dlm_get_lock_cookie_seq(cookie), + dlm_list_in_text(idx)); + return 0; } - queue++; } - *numlocks = count; - - count = find_next_bit(res->refmap, O2NM_MAX_NODES, 0); - if (count < O2NM_MAX_NODES) - *hasrefs = 1; + if (!nonlocal) { + node_ref = find_next_bit(res->refmap, O2NM_MAX_NODES, 0); + if (node_ref >= O2NM_MAX_NODES) + return 0; + } - mlog(0, "%s: res %.*s, Migrateable, locks %d, refs %d\n", dlm->name, - res->lockname.len, res->lockname.name, *numlocks, *hasrefs); + mlog(0, "%s: res %.*s, Migrateable\n", dlm->name, res->lockname.len, + res->lockname.name); -leave: - return ret; + return 1; } /* @@ -2406,8 +2396,7 @@ leave: static int dlm_migrate_lockres(struct dlm_ctxt *dlm, - struct dlm_lock_resource *res, - u8 target) + struct dlm_lock_resource *res, u8 target) { struct dlm_master_list_entry *mle = NULL; struct dlm_master_list_entry *oldmle = NULL; @@ -2416,37 +2405,20 @@ static int dlm_migrate_lockres(struct dlm_ctxt *dlm, const char *name; unsigned int namelen; int mle_added = 0; - int numlocks, hasrefs; int wake = 0; if (!dlm_grab(dlm)) return -EINVAL; + BUG_ON(target == O2NM_MAX_NODES); + name = res->lockname.name; namelen = res->lockname.len; - mlog(0, "%s: Migrating %.*s to %u\n", dlm->name, namelen, name, target); - - /* - * ensure this lockres is a proper candidate for migration - */ - spin_lock(&res->spinlock); - ret = dlm_is_lockres_migrateable(dlm, res, &numlocks, &hasrefs); - if (ret < 0) { - spin_unlock(&res->spinlock); - goto leave; - } - spin_unlock(&res->spinlock); - - /* no work to do */ - if (numlocks == 0 && !hasrefs) - goto leave; - - /* - * preallocate up front - * if this fails, abort - */ + mlog(0, "%s: Migrating %.*s to node %u\n", dlm->name, namelen, name, + target); + /* preallocate up front. if this fails, abort */ ret = -ENOMEM; mres = (struct dlm_migratable_lockres *) __get_free_page(GFP_NOFS); if (!mres) { @@ -2462,35 +2434,10 @@ static int dlm_migrate_lockres(struct dlm_ctxt *dlm, ret = 0; /* - * find a node to migrate the lockres to - */ - - spin_lock(&dlm->spinlock); - /* pick a new node */ - if (!test_bit(target, dlm->domain_map) || - target >= O2NM_MAX_NODES) { - target = dlm_pick_migration_target(dlm, res); - } - mlog(0, "%s: res %.*s, Node %u chosen for migration\n", dlm->name, - namelen, name, target); - - if (target >= O2NM_MAX_NODES || - !test_bit(target, dlm->domain_map)) { - /* target chosen is not alive */ - ret = -EINVAL; - } - - if (ret) { - spin_unlock(&dlm->spinlock); - goto fail; - } - - mlog(0, "continuing with target = %u\n", target); - - /* * clear any existing master requests and * add the migration mle to the list */ + spin_lock(&dlm->spinlock); spin_lock(&dlm->master_lock); ret = dlm_add_migration_mle(dlm, res, mle, &oldmle, name, namelen, target, dlm->node_num); @@ -2531,6 +2478,7 @@ fail: dlm_put_mle(mle); } else if (mle) { kmem_cache_free(dlm_mle_cache, mle); + mle = NULL; } goto leave; } @@ -2574,6 +2522,9 @@ fail: res->state &= ~DLM_LOCK_RES_MIGRATING; wake = 1; spin_unlock(&res->spinlock); + if (dlm_is_host_down(ret)) + dlm_wait_for_node_death(dlm, target, + DLM_NODE_DEATH_WAIT_MAX); goto leave; } @@ -2649,69 +2600,52 @@ leave: if (wake) wake_up(&res->wq); - /* TODO: cleanup */ if (mres) free_page((unsigned long)mres); dlm_put(dlm); - mlog(0, "returning %d\n", ret); + mlog(0, "%s: Migrating %.*s to %u, returns %d\n", dlm->name, namelen, + name, target, ret); return ret; } #define DLM_MIGRATION_RETRY_MS 100 -/* Should be called only after beginning the domain leave process. +/* + * Should be called only after beginning the domain leave process. * There should not be any remaining locks on nonlocal lock resources, * and there should be no local locks left on locally mastered resources. * * Called with the dlm spinlock held, may drop it to do migration, but * will re-acquire before exit. * - * Returns: 1 if dlm->spinlock was dropped/retaken, 0 if never dropped */ + * Returns: 1 if dlm->spinlock was dropped/retaken, 0 if never dropped + */ int dlm_empty_lockres(struct dlm_ctxt *dlm, struct dlm_lock_resource *res) { int ret; int lock_dropped = 0; - int numlocks, hasrefs; + u8 target = O2NM_MAX_NODES; + + assert_spin_locked(&dlm->spinlock); spin_lock(&res->spinlock); - if (res->owner != dlm->node_num) { - if (!__dlm_lockres_unused(res)) { - mlog(ML_ERROR, "%s:%.*s: this node is not master, " - "trying to free this but locks remain\n", - dlm->name, res->lockname.len, res->lockname.name); - } - spin_unlock(&res->spinlock); - goto leave; - } + if (dlm_is_lockres_migrateable(dlm, res)) + target = dlm_pick_migration_target(dlm, res); + spin_unlock(&res->spinlock); - /* No need to migrate a lockres having no locks */ - ret = dlm_is_lockres_migrateable(dlm, res, &numlocks, &hasrefs); - if (ret >= 0 && numlocks == 0 && !hasrefs) { - spin_unlock(&res->spinlock); + if (target == O2NM_MAX_NODES) goto leave; - } - spin_unlock(&res->spinlock); /* Wheee! Migrate lockres here! Will sleep so drop spinlock. */ spin_unlock(&dlm->spinlock); lock_dropped = 1; - while (1) { - ret = dlm_migrate_lockres(dlm, res, O2NM_MAX_NODES); - if (ret >= 0) - break; - if (ret == -ENOTEMPTY) { - mlog(ML_ERROR, "lockres %.*s still has local locks!\n", - res->lockname.len, res->lockname.name); - BUG(); - } - - mlog(0, "lockres %.*s: migrate failed, " - "retrying\n", res->lockname.len, - res->lockname.name); - msleep(DLM_MIGRATION_RETRY_MS); - } + ret = dlm_migrate_lockres(dlm, res, target); + if (ret) + mlog(0, "%s: res %.*s, Migrate to node %u failed with %d\n", + dlm->name, res->lockname.len, res->lockname.name, + target, ret); spin_lock(&dlm->spinlock); leave: return lock_dropped; @@ -2895,61 +2829,55 @@ static void dlm_remove_nonlocal_locks(struct dlm_ctxt *dlm, } } -/* for now this is not too intelligent. we will - * need stats to make this do the right thing. - * this just finds the first lock on one of the - * queues and uses that node as the target. */ +/* + * Pick a node to migrate the lock resource to. This function selects a + * potential target based first on the locks and then on refmap. It skips + * nodes that are in the process of exiting the domain. + */ static u8 dlm_pick_migration_target(struct dlm_ctxt *dlm, struct dlm_lock_resource *res) { - int i; + enum dlm_lockres_list idx; struct list_head *queue = &res->granted; struct dlm_lock *lock; - int nodenum; + int noderef; + u8 nodenum = O2NM_MAX_NODES; assert_spin_locked(&dlm->spinlock); + assert_spin_locked(&res->spinlock); - spin_lock(&res->spinlock); - for (i=0; i<3; i++) { + /* Go through all the locks */ + for (idx = DLM_GRANTED_LIST; idx <= DLM_BLOCKED_LIST; idx++) { + queue = dlm_list_idx_to_ptr(res, idx); list_for_each_entry(lock, queue, list) { - /* up to the caller to make sure this node - * is alive */ - if (lock->ml.node != dlm->node_num) { - spin_unlock(&res->spinlock); - return lock->ml.node; - } + if (lock->ml.node == dlm->node_num) + continue; + if (test_bit(lock->ml.node, dlm->exit_domain_map)) + continue; + nodenum = lock->ml.node; + goto bail; } - queue++; - } - - nodenum = find_next_bit(res->refmap, O2NM_MAX_NODES, 0); - if (nodenum < O2NM_MAX_NODES) { - spin_unlock(&res->spinlock); - return nodenum; } - spin_unlock(&res->spinlock); - mlog(0, "have not found a suitable target yet! checking domain map\n"); - /* ok now we're getting desperate. pick anyone alive. */ - nodenum = -1; + /* Go thru the refmap */ + noderef = -1; while (1) { - nodenum = find_next_bit(dlm->domain_map, - O2NM_MAX_NODES, nodenum+1); - mlog(0, "found %d in domain map\n", nodenum); - if (nodenum >= O2NM_MAX_NODES) + noderef = find_next_bit(res->refmap, O2NM_MAX_NODES, + noderef + 1); + if (noderef >= O2NM_MAX_NODES) break; - if (nodenum != dlm->node_num) { - mlog(0, "picking %d\n", nodenum); - return nodenum; - } + if (noderef == dlm->node_num) + continue; + if (test_bit(noderef, dlm->exit_domain_map)) + continue; + nodenum = noderef; + goto bail; } - mlog(0, "giving up. no master to migrate to\n"); - return DLM_LOCK_RES_OWNER_UNKNOWN; +bail: + return nodenum; } - - /* this is called by the new master once all lockres * data has been received */ static int dlm_do_migrate_request(struct dlm_ctxt *dlm, diff --git a/fs/ocfs2/dlm/dlmrecovery.c b/fs/ocfs2/dlm/dlmrecovery.c index f1beb6fc254d..7efab6d28a21 100644 --- a/fs/ocfs2/dlm/dlmrecovery.c +++ b/fs/ocfs2/dlm/dlmrecovery.c @@ -2393,6 +2393,7 @@ static void __dlm_hb_node_down(struct dlm_ctxt *dlm, int idx) mlog(0, "node %u being removed from domain map!\n", idx); clear_bit(idx, dlm->domain_map); + clear_bit(idx, dlm->exit_domain_map); /* wake up migration waiters if a node goes down. * perhaps later we can genericize this for other waiters. */ wake_up(&dlm->migration_wq); diff --git a/fs/ocfs2/dlmfs/dlmfs.c b/fs/ocfs2/dlmfs/dlmfs.c index 8c5c0eddc365..b42076797049 100644 --- a/fs/ocfs2/dlmfs/dlmfs.c +++ b/fs/ocfs2/dlmfs/dlmfs.c @@ -88,7 +88,7 @@ struct workqueue_struct *user_dlm_worker; * signifies a bast fired on the lock. */ #define DLMFS_CAPABILITIES "bast stackglue" -extern int param_set_dlmfs_capabilities(const char *val, +static int param_set_dlmfs_capabilities(const char *val, struct kernel_param *kp) { printk(KERN_ERR "%s: readonly parameter\n", kp->name); diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 41565ae52856..b1e35a392ca5 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -1607,6 +1607,9 @@ static void ocfs2_calc_trunc_pos(struct inode *inode, range = le32_to_cpu(rec->e_cpos) + ocfs2_rec_clusters(el, rec); if (le32_to_cpu(rec->e_cpos) >= trunc_start) { + /* + * remove an entire extent record. + */ *trunc_cpos = le32_to_cpu(rec->e_cpos); /* * Skip holes if any. @@ -1617,7 +1620,16 @@ static void ocfs2_calc_trunc_pos(struct inode *inode, *blkno = le64_to_cpu(rec->e_blkno); *trunc_end = le32_to_cpu(rec->e_cpos); } else if (range > trunc_start) { + /* + * remove a partial extent record, which means we're + * removing the last extent record. + */ *trunc_cpos = trunc_start; + /* + * skip hole if any. + */ + if (range < *trunc_end) + *trunc_end = range; *trunc_len = *trunc_end - trunc_start; coff = trunc_start - le32_to_cpu(rec->e_cpos); *blkno = le64_to_cpu(rec->e_blkno) + @@ -2658,6 +2670,7 @@ const struct file_operations ocfs2_fops_no_plocks = { .flock = ocfs2_flock, .splice_read = ocfs2_file_splice_read, .splice_write = ocfs2_file_splice_write, + .fallocate = ocfs2_fallocate, }; const struct file_operations ocfs2_dops_no_plocks = { diff --git a/fs/ocfs2/inode.c b/fs/ocfs2/inode.c index 177d3a6c2a5f..b4c8bb6b8d28 100644 --- a/fs/ocfs2/inode.c +++ b/fs/ocfs2/inode.c @@ -835,7 +835,7 @@ static int ocfs2_inode_is_valid_to_delete(struct inode *inode) /* If we have allowd wipe of this inode for another node, it * will be marked here so we can safely skip it. Recovery will - * cleanup any inodes we might inadvertantly skip here. */ + * cleanup any inodes we might inadvertently skip here. */ if (oi->ip_flags & OCFS2_INODE_SKIP_DELETE) goto bail_unlock; @@ -917,7 +917,7 @@ static int ocfs2_query_inode_wipe(struct inode *inode, * the inode open lock in ocfs2_read_locked_inode(). When we * get to ->delete_inode(), each node tries to convert it's * lock to an exclusive. Trylocks are serialized by the inode - * meta data lock. If the upconvert suceeds, we know the inode + * meta data lock. If the upconvert succeeds, we know the inode * is no longer live and can be deleted. * * Though we call this with the meta data lock held, the diff --git a/fs/ocfs2/ioctl.c b/fs/ocfs2/ioctl.c index 8f13c5989eae..bc91072b7219 100644 --- a/fs/ocfs2/ioctl.c +++ b/fs/ocfs2/ioctl.c @@ -22,6 +22,11 @@ #include "ioctl.h" #include "resize.h" #include "refcounttree.h" +#include "sysfile.h" +#include "dir.h" +#include "buffer_head_io.h" +#include "suballoc.h" +#include "move_extents.h" #include <linux/ext2_fs.h> @@ -35,31 +40,27 @@ * be -EFAULT. The error will be returned from the ioctl(2) call. It's * just a best-effort to tell userspace that this request caused the error. */ -static inline void __o2info_set_request_error(struct ocfs2_info_request *kreq, +static inline void o2info_set_request_error(struct ocfs2_info_request *kreq, struct ocfs2_info_request __user *req) { kreq->ir_flags |= OCFS2_INFO_FL_ERROR; (void)put_user(kreq->ir_flags, (__u32 __user *)&(req->ir_flags)); } -#define o2info_set_request_error(a, b) \ - __o2info_set_request_error((struct ocfs2_info_request *)&(a), b) - -static inline void __o2info_set_request_filled(struct ocfs2_info_request *req) +static inline void o2info_set_request_filled(struct ocfs2_info_request *req) { req->ir_flags |= OCFS2_INFO_FL_FILLED; } -#define o2info_set_request_filled(a) \ - __o2info_set_request_filled((struct ocfs2_info_request *)&(a)) - -static inline void __o2info_clear_request_filled(struct ocfs2_info_request *req) +static inline void o2info_clear_request_filled(struct ocfs2_info_request *req) { req->ir_flags &= ~OCFS2_INFO_FL_FILLED; } -#define o2info_clear_request_filled(a) \ - __o2info_clear_request_filled((struct ocfs2_info_request *)&(a)) +static inline int o2info_coherent(struct ocfs2_info_request *req) +{ + return (!(req->ir_flags & OCFS2_INFO_FL_NON_COHERENT)); +} static int ocfs2_get_inode_attr(struct inode *inode, unsigned *flags) { @@ -153,7 +154,7 @@ int ocfs2_info_handle_blocksize(struct inode *inode, oib.ib_blocksize = inode->i_sb->s_blocksize; - o2info_set_request_filled(oib); + o2info_set_request_filled(&oib.ib_req); if (o2info_to_user(oib, req)) goto bail; @@ -161,7 +162,7 @@ int ocfs2_info_handle_blocksize(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oib, req); + o2info_set_request_error(&oib.ib_req, req); return status; } @@ -178,7 +179,7 @@ int ocfs2_info_handle_clustersize(struct inode *inode, oic.ic_clustersize = osb->s_clustersize; - o2info_set_request_filled(oic); + o2info_set_request_filled(&oic.ic_req); if (o2info_to_user(oic, req)) goto bail; @@ -186,7 +187,7 @@ int ocfs2_info_handle_clustersize(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oic, req); + o2info_set_request_error(&oic.ic_req, req); return status; } @@ -203,7 +204,7 @@ int ocfs2_info_handle_maxslots(struct inode *inode, oim.im_max_slots = osb->max_slots; - o2info_set_request_filled(oim); + o2info_set_request_filled(&oim.im_req); if (o2info_to_user(oim, req)) goto bail; @@ -211,7 +212,7 @@ int ocfs2_info_handle_maxslots(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oim, req); + o2info_set_request_error(&oim.im_req, req); return status; } @@ -228,7 +229,7 @@ int ocfs2_info_handle_label(struct inode *inode, memcpy(oil.il_label, osb->vol_label, OCFS2_MAX_VOL_LABEL_LEN); - o2info_set_request_filled(oil); + o2info_set_request_filled(&oil.il_req); if (o2info_to_user(oil, req)) goto bail; @@ -236,7 +237,7 @@ int ocfs2_info_handle_label(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oil, req); + o2info_set_request_error(&oil.il_req, req); return status; } @@ -253,7 +254,7 @@ int ocfs2_info_handle_uuid(struct inode *inode, memcpy(oiu.iu_uuid_str, osb->uuid_str, OCFS2_TEXT_UUID_LEN + 1); - o2info_set_request_filled(oiu); + o2info_set_request_filled(&oiu.iu_req); if (o2info_to_user(oiu, req)) goto bail; @@ -261,7 +262,7 @@ int ocfs2_info_handle_uuid(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oiu, req); + o2info_set_request_error(&oiu.iu_req, req); return status; } @@ -280,7 +281,7 @@ int ocfs2_info_handle_fs_features(struct inode *inode, oif.if_incompat_features = osb->s_feature_incompat; oif.if_ro_compat_features = osb->s_feature_ro_compat; - o2info_set_request_filled(oif); + o2info_set_request_filled(&oif.if_req); if (o2info_to_user(oif, req)) goto bail; @@ -288,7 +289,7 @@ int ocfs2_info_handle_fs_features(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oif, req); + o2info_set_request_error(&oif.if_req, req); return status; } @@ -305,7 +306,7 @@ int ocfs2_info_handle_journal_size(struct inode *inode, oij.ij_journal_size = osb->journal->j_inode->i_size; - o2info_set_request_filled(oij); + o2info_set_request_filled(&oij.ij_req); if (o2info_to_user(oij, req)) goto bail; @@ -313,7 +314,408 @@ int ocfs2_info_handle_journal_size(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oij, req); + o2info_set_request_error(&oij.ij_req, req); + + return status; +} + +int ocfs2_info_scan_inode_alloc(struct ocfs2_super *osb, + struct inode *inode_alloc, u64 blkno, + struct ocfs2_info_freeinode *fi, u32 slot) +{ + int status = 0, unlock = 0; + + struct buffer_head *bh = NULL; + struct ocfs2_dinode *dinode_alloc = NULL; + + if (inode_alloc) + mutex_lock(&inode_alloc->i_mutex); + + if (o2info_coherent(&fi->ifi_req)) { + status = ocfs2_inode_lock(inode_alloc, &bh, 0); + if (status < 0) { + mlog_errno(status); + goto bail; + } + unlock = 1; + } else { + status = ocfs2_read_blocks_sync(osb, blkno, 1, &bh); + if (status < 0) { + mlog_errno(status); + goto bail; + } + } + + dinode_alloc = (struct ocfs2_dinode *)bh->b_data; + + fi->ifi_stat[slot].lfi_total = + le32_to_cpu(dinode_alloc->id1.bitmap1.i_total); + fi->ifi_stat[slot].lfi_free = + le32_to_cpu(dinode_alloc->id1.bitmap1.i_total) - + le32_to_cpu(dinode_alloc->id1.bitmap1.i_used); + +bail: + if (unlock) + ocfs2_inode_unlock(inode_alloc, 0); + + if (inode_alloc) + mutex_unlock(&inode_alloc->i_mutex); + + brelse(bh); + + return status; +} + +int ocfs2_info_handle_freeinode(struct inode *inode, + struct ocfs2_info_request __user *req) +{ + u32 i; + u64 blkno = -1; + char namebuf[40]; + int status = -EFAULT, type = INODE_ALLOC_SYSTEM_INODE; + struct ocfs2_info_freeinode *oifi = NULL; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct inode *inode_alloc = NULL; + + oifi = kzalloc(sizeof(struct ocfs2_info_freeinode), GFP_KERNEL); + if (!oifi) { + status = -ENOMEM; + mlog_errno(status); + goto bail; + } + + if (o2info_from_user(*oifi, req)) + goto bail; + + oifi->ifi_slotnum = osb->max_slots; + + for (i = 0; i < oifi->ifi_slotnum; i++) { + if (o2info_coherent(&oifi->ifi_req)) { + inode_alloc = ocfs2_get_system_file_inode(osb, type, i); + if (!inode_alloc) { + mlog(ML_ERROR, "unable to get alloc inode in " + "slot %u\n", i); + status = -EIO; + goto bail; + } + } else { + ocfs2_sprintf_system_inode_name(namebuf, + sizeof(namebuf), + type, i); + status = ocfs2_lookup_ino_from_name(osb->sys_root_inode, + namebuf, + strlen(namebuf), + &blkno); + if (status < 0) { + status = -ENOENT; + goto bail; + } + } + + status = ocfs2_info_scan_inode_alloc(osb, inode_alloc, blkno, oifi, i); + if (status < 0) + goto bail; + + iput(inode_alloc); + inode_alloc = NULL; + } + + o2info_set_request_filled(&oifi->ifi_req); + + if (o2info_to_user(*oifi, req)) + goto bail; + + status = 0; +bail: + if (status) + o2info_set_request_error(&oifi->ifi_req, req); + + kfree(oifi); + + return status; +} + +static void o2ffg_update_histogram(struct ocfs2_info_free_chunk_list *hist, + unsigned int chunksize) +{ + int index; + + index = __ilog2_u32(chunksize); + if (index >= OCFS2_INFO_MAX_HIST) + index = OCFS2_INFO_MAX_HIST - 1; + + hist->fc_chunks[index]++; + hist->fc_clusters[index] += chunksize; +} + +static void o2ffg_update_stats(struct ocfs2_info_freefrag_stats *stats, + unsigned int chunksize) +{ + if (chunksize > stats->ffs_max) + stats->ffs_max = chunksize; + + if (chunksize < stats->ffs_min) + stats->ffs_min = chunksize; + + stats->ffs_avg += chunksize; + stats->ffs_free_chunks_real++; +} + +void ocfs2_info_update_ffg(struct ocfs2_info_freefrag *ffg, + unsigned int chunksize) +{ + o2ffg_update_histogram(&(ffg->iff_ffs.ffs_fc_hist), chunksize); + o2ffg_update_stats(&(ffg->iff_ffs), chunksize); +} + +int ocfs2_info_freefrag_scan_chain(struct ocfs2_super *osb, + struct inode *gb_inode, + struct ocfs2_dinode *gb_dinode, + struct ocfs2_chain_rec *rec, + struct ocfs2_info_freefrag *ffg, + u32 chunks_in_group) +{ + int status = 0, used; + u64 blkno; + + struct buffer_head *bh = NULL; + struct ocfs2_group_desc *bg = NULL; + + unsigned int max_bits, num_clusters; + unsigned int offset = 0, cluster, chunk; + unsigned int chunk_free, last_chunksize = 0; + + if (!le32_to_cpu(rec->c_free)) + goto bail; + + do { + if (!bg) + blkno = le64_to_cpu(rec->c_blkno); + else + blkno = le64_to_cpu(bg->bg_next_group); + + if (bh) { + brelse(bh); + bh = NULL; + } + + if (o2info_coherent(&ffg->iff_req)) + status = ocfs2_read_group_descriptor(gb_inode, + gb_dinode, + blkno, &bh); + else + status = ocfs2_read_blocks_sync(osb, blkno, 1, &bh); + + if (status < 0) { + mlog(ML_ERROR, "Can't read the group descriptor # " + "%llu from device.", (unsigned long long)blkno); + status = -EIO; + goto bail; + } + + bg = (struct ocfs2_group_desc *)bh->b_data; + + if (!le16_to_cpu(bg->bg_free_bits_count)) + continue; + + max_bits = le16_to_cpu(bg->bg_bits); + offset = 0; + + for (chunk = 0; chunk < chunks_in_group; chunk++) { + /* + * last chunk may be not an entire one. + */ + if ((offset + ffg->iff_chunksize) > max_bits) + num_clusters = max_bits - offset; + else + num_clusters = ffg->iff_chunksize; + + chunk_free = 0; + for (cluster = 0; cluster < num_clusters; cluster++) { + used = ocfs2_test_bit(offset, + (unsigned long *)bg->bg_bitmap); + /* + * - chunk_free counts free clusters in #N chunk. + * - last_chunksize records the size(in) clusters + * for the last real free chunk being counted. + */ + if (!used) { + last_chunksize++; + chunk_free++; + } + + if (used && last_chunksize) { + ocfs2_info_update_ffg(ffg, + last_chunksize); + last_chunksize = 0; + } + + offset++; + } + + if (chunk_free == ffg->iff_chunksize) + ffg->iff_ffs.ffs_free_chunks++; + } + + /* + * need to update the info for last free chunk. + */ + if (last_chunksize) + ocfs2_info_update_ffg(ffg, last_chunksize); + + } while (le64_to_cpu(bg->bg_next_group)); + +bail: + brelse(bh); + + return status; +} + +int ocfs2_info_freefrag_scan_bitmap(struct ocfs2_super *osb, + struct inode *gb_inode, u64 blkno, + struct ocfs2_info_freefrag *ffg) +{ + u32 chunks_in_group; + int status = 0, unlock = 0, i; + + struct buffer_head *bh = NULL; + struct ocfs2_chain_list *cl = NULL; + struct ocfs2_chain_rec *rec = NULL; + struct ocfs2_dinode *gb_dinode = NULL; + + if (gb_inode) + mutex_lock(&gb_inode->i_mutex); + + if (o2info_coherent(&ffg->iff_req)) { + status = ocfs2_inode_lock(gb_inode, &bh, 0); + if (status < 0) { + mlog_errno(status); + goto bail; + } + unlock = 1; + } else { + status = ocfs2_read_blocks_sync(osb, blkno, 1, &bh); + if (status < 0) { + mlog_errno(status); + goto bail; + } + } + + gb_dinode = (struct ocfs2_dinode *)bh->b_data; + cl = &(gb_dinode->id2.i_chain); + + /* + * Chunksize(in) clusters from userspace should be + * less than clusters in a group. + */ + if (ffg->iff_chunksize > le16_to_cpu(cl->cl_cpg)) { + status = -EINVAL; + goto bail; + } + + memset(&ffg->iff_ffs, 0, sizeof(struct ocfs2_info_freefrag_stats)); + + ffg->iff_ffs.ffs_min = ~0U; + ffg->iff_ffs.ffs_clusters = + le32_to_cpu(gb_dinode->id1.bitmap1.i_total); + ffg->iff_ffs.ffs_free_clusters = ffg->iff_ffs.ffs_clusters - + le32_to_cpu(gb_dinode->id1.bitmap1.i_used); + + chunks_in_group = le16_to_cpu(cl->cl_cpg) / ffg->iff_chunksize + 1; + + for (i = 0; i < le16_to_cpu(cl->cl_next_free_rec); i++) { + rec = &(cl->cl_recs[i]); + status = ocfs2_info_freefrag_scan_chain(osb, gb_inode, + gb_dinode, + rec, ffg, + chunks_in_group); + if (status) + goto bail; + } + + if (ffg->iff_ffs.ffs_free_chunks_real) + ffg->iff_ffs.ffs_avg = (ffg->iff_ffs.ffs_avg / + ffg->iff_ffs.ffs_free_chunks_real); +bail: + if (unlock) + ocfs2_inode_unlock(gb_inode, 0); + + if (gb_inode) + mutex_unlock(&gb_inode->i_mutex); + + if (gb_inode) + iput(gb_inode); + + brelse(bh); + + return status; +} + +int ocfs2_info_handle_freefrag(struct inode *inode, + struct ocfs2_info_request __user *req) +{ + u64 blkno = -1; + char namebuf[40]; + int status = -EFAULT, type = GLOBAL_BITMAP_SYSTEM_INODE; + + struct ocfs2_info_freefrag *oiff; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct inode *gb_inode = NULL; + + oiff = kzalloc(sizeof(struct ocfs2_info_freefrag), GFP_KERNEL); + if (!oiff) { + status = -ENOMEM; + mlog_errno(status); + goto bail; + } + + if (o2info_from_user(*oiff, req)) + goto bail; + /* + * chunksize from userspace should be power of 2. + */ + if ((oiff->iff_chunksize & (oiff->iff_chunksize - 1)) || + (!oiff->iff_chunksize)) { + status = -EINVAL; + goto bail; + } + + if (o2info_coherent(&oiff->iff_req)) { + gb_inode = ocfs2_get_system_file_inode(osb, type, + OCFS2_INVALID_SLOT); + if (!gb_inode) { + mlog(ML_ERROR, "unable to get global_bitmap inode\n"); + status = -EIO; + goto bail; + } + } else { + ocfs2_sprintf_system_inode_name(namebuf, sizeof(namebuf), type, + OCFS2_INVALID_SLOT); + status = ocfs2_lookup_ino_from_name(osb->sys_root_inode, + namebuf, + strlen(namebuf), + &blkno); + if (status < 0) { + status = -ENOENT; + goto bail; + } + } + + status = ocfs2_info_freefrag_scan_bitmap(osb, gb_inode, blkno, oiff); + if (status < 0) + goto bail; + + o2info_set_request_filled(&oiff->iff_req); + + if (o2info_to_user(*oiff, req)) + goto bail; + + status = 0; +bail: + if (status) + o2info_set_request_error(&oiff->iff_req, req); + + kfree(oiff); return status; } @@ -327,7 +729,7 @@ int ocfs2_info_handle_unknown(struct inode *inode, if (o2info_from_user(oir, req)) goto bail; - o2info_clear_request_filled(oir); + o2info_clear_request_filled(&oir); if (o2info_to_user(oir, req)) goto bail; @@ -335,7 +737,7 @@ int ocfs2_info_handle_unknown(struct inode *inode, status = 0; bail: if (status) - o2info_set_request_error(oir, req); + o2info_set_request_error(&oir, req); return status; } @@ -389,6 +791,14 @@ int ocfs2_info_handle_request(struct inode *inode, if (oir.ir_size == sizeof(struct ocfs2_info_journal_size)) status = ocfs2_info_handle_journal_size(inode, req); break; + case OCFS2_INFO_FREEINODE: + if (oir.ir_size == sizeof(struct ocfs2_info_freeinode)) + status = ocfs2_info_handle_freeinode(inode, req); + break; + case OCFS2_INFO_FREEFRAG: + if (oir.ir_size == sizeof(struct ocfs2_info_freefrag)) + status = ocfs2_info_handle_freefrag(inode, req); + break; default: status = ocfs2_info_handle_unknown(inode, req); break; @@ -542,6 +952,31 @@ long ocfs2_ioctl(struct file *filp, unsigned int cmd, unsigned long arg) return -EFAULT; return ocfs2_info_handle(inode, &info, 0); + case FITRIM: + { + struct super_block *sb = inode->i_sb; + struct fstrim_range range; + int ret = 0; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + if (copy_from_user(&range, (struct fstrim_range *)arg, + sizeof(range))) + return -EFAULT; + + ret = ocfs2_trim_fs(sb, &range); + if (ret < 0) + return ret; + + if (copy_to_user((struct fstrim_range *)arg, &range, + sizeof(range))) + return -EFAULT; + + return 0; + } + case OCFS2_IOC_MOVE_EXT: + return ocfs2_ioctl_move_extents(filp, (void __user *)arg); default: return -ENOTTY; } @@ -569,6 +1004,7 @@ long ocfs2_compat_ioctl(struct file *file, unsigned cmd, unsigned long arg) case OCFS2_IOC_GROUP_EXTEND: case OCFS2_IOC_GROUP_ADD: case OCFS2_IOC_GROUP_ADD64: + case FITRIM: break; case OCFS2_IOC_REFLINK: if (copy_from_user(&args, (struct reflink_arguments *)arg, @@ -584,6 +1020,8 @@ long ocfs2_compat_ioctl(struct file *file, unsigned cmd, unsigned long arg) return -EFAULT; return ocfs2_info_handle(inode, &info, 1); + case OCFS2_IOC_MOVE_EXT: + break; default: return -ENOIOCTLCMD; } diff --git a/fs/ocfs2/journal.c b/fs/ocfs2/journal.c index dcc2d9327150..295d56454e8b 100644 --- a/fs/ocfs2/journal.c +++ b/fs/ocfs2/journal.c @@ -1260,6 +1260,9 @@ void ocfs2_complete_mount_recovery(struct ocfs2_super *osb) { struct ocfs2_journal *journal = osb->journal; + if (ocfs2_is_hard_readonly(osb)) + return; + /* No need to queue up our truncate_log as regular cleanup will catch * that */ ocfs2_queue_recovery_completion(journal, osb->slot_num, @@ -1368,7 +1371,7 @@ skip_recovery: mlog_errno(status); /* Now it is right time to recover quotas... We have to do this under - * superblock lock so that noone can start using the slot (and crash) + * superblock lock so that no one can start using the slot (and crash) * before we recover it */ for (i = 0; i < rm_quota_used; i++) { qrec = ocfs2_begin_quota_recovery(osb, rm_quota[i]); diff --git a/fs/ocfs2/journal.h b/fs/ocfs2/journal.h index 6180da1e37e6..68cf2f6d3c6a 100644 --- a/fs/ocfs2/journal.h +++ b/fs/ocfs2/journal.h @@ -215,7 +215,7 @@ static inline void ocfs2_checkpoint_inode(struct inode *inode) /* WARNING: This only kicks off a single * checkpoint. If someone races you and adds more * metadata to the journal, you won't know, and will - * wind up waiting *alot* longer than necessary. Right + * wind up waiting *a lot* longer than necessary. Right * now we only use this in clear_inode so that's * OK. */ ocfs2_start_checkpoint(osb); diff --git a/fs/ocfs2/move_extents.c b/fs/ocfs2/move_extents.c new file mode 100644 index 000000000000..cd9427023d2e --- /dev/null +++ b/fs/ocfs2/move_extents.c @@ -0,0 +1,1152 @@ +/* -*- mode: c; c-basic-offset: 8; -*- + * vim: noexpandtab sw=8 ts=8 sts=0: + * + * move_extents.c + * + * Copyright (C) 2011 Oracle. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + */ +#include <linux/fs.h> +#include <linux/types.h> +#include <linux/mount.h> +#include <linux/swap.h> + +#include <cluster/masklog.h> + +#include "ocfs2.h" +#include "ocfs2_ioctl.h" + +#include "alloc.h" +#include "aops.h" +#include "dlmglue.h" +#include "extent_map.h" +#include "inode.h" +#include "journal.h" +#include "suballoc.h" +#include "uptodate.h" +#include "super.h" +#include "dir.h" +#include "buffer_head_io.h" +#include "sysfile.h" +#include "suballoc.h" +#include "refcounttree.h" +#include "move_extents.h" + +struct ocfs2_move_extents_context { + struct inode *inode; + struct file *file; + int auto_defrag; + int partial; + int credits; + u32 new_phys_cpos; + u32 clusters_moved; + u64 refcount_loc; + struct ocfs2_move_extents *range; + struct ocfs2_extent_tree et; + struct ocfs2_alloc_context *meta_ac; + struct ocfs2_alloc_context *data_ac; + struct ocfs2_cached_dealloc_ctxt dealloc; +}; + +static int __ocfs2_move_extent(handle_t *handle, + struct ocfs2_move_extents_context *context, + u32 cpos, u32 len, u32 p_cpos, u32 new_p_cpos, + int ext_flags) +{ + int ret = 0, index; + struct inode *inode = context->inode; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct ocfs2_extent_rec *rec, replace_rec; + struct ocfs2_path *path = NULL; + struct ocfs2_extent_list *el; + u64 ino = ocfs2_metadata_cache_owner(context->et.et_ci); + u64 old_blkno = ocfs2_clusters_to_blocks(inode->i_sb, p_cpos); + + ret = ocfs2_duplicate_clusters_by_page(handle, context->file, cpos, + p_cpos, new_p_cpos, len); + if (ret) { + mlog_errno(ret); + goto out; + } + + memset(&replace_rec, 0, sizeof(replace_rec)); + replace_rec.e_cpos = cpu_to_le32(cpos); + replace_rec.e_leaf_clusters = cpu_to_le16(len); + replace_rec.e_blkno = cpu_to_le64(ocfs2_clusters_to_blocks(inode->i_sb, + new_p_cpos)); + + path = ocfs2_new_path_from_et(&context->et); + if (!path) { + ret = -ENOMEM; + mlog_errno(ret); + goto out; + } + + ret = ocfs2_find_path(INODE_CACHE(inode), path, cpos); + if (ret) { + mlog_errno(ret); + goto out; + } + + el = path_leaf_el(path); + + index = ocfs2_search_extent_list(el, cpos); + if (index == -1 || index >= le16_to_cpu(el->l_next_free_rec)) { + ocfs2_error(inode->i_sb, + "Inode %llu has an extent at cpos %u which can no " + "longer be found.\n", + (unsigned long long)ino, cpos); + ret = -EROFS; + goto out; + } + + rec = &el->l_recs[index]; + + BUG_ON(ext_flags != rec->e_flags); + /* + * after moving/defraging to new location, the extent is not going + * to be refcounted anymore. + */ + replace_rec.e_flags = ext_flags & ~OCFS2_EXT_REFCOUNTED; + + ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), + context->et.et_root_bh, + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret) { + mlog_errno(ret); + goto out; + } + + ret = ocfs2_split_extent(handle, &context->et, path, index, + &replace_rec, context->meta_ac, + &context->dealloc); + if (ret) { + mlog_errno(ret); + goto out; + } + + ocfs2_journal_dirty(handle, context->et.et_root_bh); + + context->new_phys_cpos = new_p_cpos; + + /* + * need I to append truncate log for old clusters? + */ + if (old_blkno) { + if (ext_flags & OCFS2_EXT_REFCOUNTED) + ret = ocfs2_decrease_refcount(inode, handle, + ocfs2_blocks_to_clusters(osb->sb, + old_blkno), + len, context->meta_ac, + &context->dealloc, 1); + else + ret = ocfs2_truncate_log_append(osb, handle, + old_blkno, len); + } + +out: + return ret; +} + +/* + * lock allocators, and reserving appropriate number of bits for + * meta blocks and data clusters. + * + * in some cases, we don't need to reserve clusters, just let data_ac + * be NULL. + */ +static int ocfs2_lock_allocators_move_extents(struct inode *inode, + struct ocfs2_extent_tree *et, + u32 clusters_to_move, + u32 extents_to_split, + struct ocfs2_alloc_context **meta_ac, + struct ocfs2_alloc_context **data_ac, + int extra_blocks, + int *credits) +{ + int ret, num_free_extents; + unsigned int max_recs_needed = 2 * extents_to_split + clusters_to_move; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + + num_free_extents = ocfs2_num_free_extents(osb, et); + if (num_free_extents < 0) { + ret = num_free_extents; + mlog_errno(ret); + goto out; + } + + if (!num_free_extents || + (ocfs2_sparse_alloc(osb) && num_free_extents < max_recs_needed)) + extra_blocks += ocfs2_extend_meta_needed(et->et_root_el); + + ret = ocfs2_reserve_new_metadata_blocks(osb, extra_blocks, meta_ac); + if (ret) { + mlog_errno(ret); + goto out; + } + + if (data_ac) { + ret = ocfs2_reserve_clusters(osb, clusters_to_move, data_ac); + if (ret) { + mlog_errno(ret); + goto out; + } + } + + *credits += ocfs2_calc_extend_credits(osb->sb, et->et_root_el, + clusters_to_move + 2); + + mlog(0, "reserve metadata_blocks: %d, data_clusters: %u, credits: %d\n", + extra_blocks, clusters_to_move, *credits); +out: + if (ret) { + if (*meta_ac) { + ocfs2_free_alloc_context(*meta_ac); + *meta_ac = NULL; + } + } + + return ret; +} + +/* + * Using one journal handle to guarantee the data consistency in case + * crash happens anywhere. + * + * XXX: defrag can end up with finishing partial extent as requested, + * due to not enough contiguous clusters can be found in allocator. + */ +static int ocfs2_defrag_extent(struct ocfs2_move_extents_context *context, + u32 cpos, u32 phys_cpos, u32 *len, int ext_flags) +{ + int ret, credits = 0, extra_blocks = 0, partial = context->partial; + handle_t *handle; + struct inode *inode = context->inode; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct inode *tl_inode = osb->osb_tl_inode; + struct ocfs2_refcount_tree *ref_tree = NULL; + u32 new_phys_cpos, new_len; + u64 phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos); + + if ((ext_flags & OCFS2_EXT_REFCOUNTED) && *len) { + + BUG_ON(!(OCFS2_I(inode)->ip_dyn_features & + OCFS2_HAS_REFCOUNT_FL)); + + BUG_ON(!context->refcount_loc); + + ret = ocfs2_lock_refcount_tree(osb, context->refcount_loc, 1, + &ref_tree, NULL); + if (ret) { + mlog_errno(ret); + return ret; + } + + ret = ocfs2_prepare_refcount_change_for_del(inode, + context->refcount_loc, + phys_blkno, + *len, + &credits, + &extra_blocks); + if (ret) { + mlog_errno(ret); + goto out; + } + } + + ret = ocfs2_lock_allocators_move_extents(inode, &context->et, *len, 1, + &context->meta_ac, + &context->data_ac, + extra_blocks, &credits); + if (ret) { + mlog_errno(ret); + goto out; + } + + /* + * should be using allocation reservation strategy there? + * + * if (context->data_ac) + * context->data_ac->ac_resv = &OCFS2_I(inode)->ip_la_data_resv; + */ + + mutex_lock(&tl_inode->i_mutex); + + if (ocfs2_truncate_log_needs_flush(osb)) { + ret = __ocfs2_flush_truncate_log(osb); + if (ret < 0) { + mlog_errno(ret); + goto out_unlock_mutex; + } + } + + handle = ocfs2_start_trans(osb, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + mlog_errno(ret); + goto out_unlock_mutex; + } + + ret = __ocfs2_claim_clusters(handle, context->data_ac, 1, *len, + &new_phys_cpos, &new_len); + if (ret) { + mlog_errno(ret); + goto out_commit; + } + + /* + * allowing partial extent moving is kind of 'pros and cons', it makes + * whole defragmentation less likely to fail, on the contrary, the bad + * thing is it may make the fs even more fragmented after moving, let + * userspace make a good decision here. + */ + if (new_len != *len) { + mlog(0, "len_claimed: %u, len: %u\n", new_len, *len); + if (!partial) { + context->range->me_flags &= ~OCFS2_MOVE_EXT_FL_COMPLETE; + ret = -ENOSPC; + goto out_commit; + } + } + + mlog(0, "cpos: %u, phys_cpos: %u, new_phys_cpos: %u\n", cpos, + phys_cpos, new_phys_cpos); + + ret = __ocfs2_move_extent(handle, context, cpos, new_len, phys_cpos, + new_phys_cpos, ext_flags); + if (ret) + mlog_errno(ret); + + if (partial && (new_len != *len)) + *len = new_len; + + /* + * Here we should write the new page out first if we are + * in write-back mode. + */ + ret = ocfs2_cow_sync_writeback(inode->i_sb, context->inode, cpos, *len); + if (ret) + mlog_errno(ret); + +out_commit: + ocfs2_commit_trans(osb, handle); + +out_unlock_mutex: + mutex_unlock(&tl_inode->i_mutex); + + if (context->data_ac) { + ocfs2_free_alloc_context(context->data_ac); + context->data_ac = NULL; + } + + if (context->meta_ac) { + ocfs2_free_alloc_context(context->meta_ac); + context->meta_ac = NULL; + } + +out: + if (ref_tree) + ocfs2_unlock_refcount_tree(osb, ref_tree, 1); + + return ret; +} + +/* + * find the victim alloc group, where #blkno fits. + */ +static int ocfs2_find_victim_alloc_group(struct inode *inode, + u64 vict_blkno, + int type, int slot, + int *vict_bit, + struct buffer_head **ret_bh) +{ + int ret, i, bits_per_unit = 0; + u64 blkno; + char namebuf[40]; + + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct buffer_head *ac_bh = NULL, *gd_bh = NULL; + struct ocfs2_chain_list *cl; + struct ocfs2_chain_rec *rec; + struct ocfs2_dinode *ac_dinode; + struct ocfs2_group_desc *bg; + + ocfs2_sprintf_system_inode_name(namebuf, sizeof(namebuf), type, slot); + ret = ocfs2_lookup_ino_from_name(osb->sys_root_inode, namebuf, + strlen(namebuf), &blkno); + if (ret) { + ret = -ENOENT; + goto out; + } + + ret = ocfs2_read_blocks_sync(osb, blkno, 1, &ac_bh); + if (ret) { + mlog_errno(ret); + goto out; + } + + ac_dinode = (struct ocfs2_dinode *)ac_bh->b_data; + cl = &(ac_dinode->id2.i_chain); + rec = &(cl->cl_recs[0]); + + if (type == GLOBAL_BITMAP_SYSTEM_INODE) + bits_per_unit = osb->s_clustersize_bits - + inode->i_sb->s_blocksize_bits; + /* + * 'vict_blkno' was out of the valid range. + */ + if ((vict_blkno < le64_to_cpu(rec->c_blkno)) || + (vict_blkno >= (le32_to_cpu(ac_dinode->id1.bitmap1.i_total) << + bits_per_unit))) { + ret = -EINVAL; + goto out; + } + + for (i = 0; i < le16_to_cpu(cl->cl_next_free_rec); i++) { + + rec = &(cl->cl_recs[i]); + if (!rec) + continue; + + bg = NULL; + + do { + if (!bg) + blkno = le64_to_cpu(rec->c_blkno); + else + blkno = le64_to_cpu(bg->bg_next_group); + + if (gd_bh) { + brelse(gd_bh); + gd_bh = NULL; + } + + ret = ocfs2_read_blocks_sync(osb, blkno, 1, &gd_bh); + if (ret) { + mlog_errno(ret); + goto out; + } + + bg = (struct ocfs2_group_desc *)gd_bh->b_data; + + if (vict_blkno < (le64_to_cpu(bg->bg_blkno) + + le16_to_cpu(bg->bg_bits))) { + + *ret_bh = gd_bh; + *vict_bit = (vict_blkno - blkno) >> + bits_per_unit; + mlog(0, "find the victim group: #%llu, " + "total_bits: %u, vict_bit: %u\n", + blkno, le16_to_cpu(bg->bg_bits), + *vict_bit); + goto out; + } + + } while (le64_to_cpu(bg->bg_next_group)); + } + + ret = -EINVAL; +out: + brelse(ac_bh); + + /* + * caller has to release the gd_bh properly. + */ + return ret; +} + +/* + * XXX: helper to validate and adjust moving goal. + */ +static int ocfs2_validate_and_adjust_move_goal(struct inode *inode, + struct ocfs2_move_extents *range) +{ + int ret, goal_bit = 0; + + struct buffer_head *gd_bh = NULL; + struct ocfs2_group_desc *bg = NULL; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + int c_to_b = 1 << (osb->s_clustersize_bits - + inode->i_sb->s_blocksize_bits); + + /* + * make goal become cluster aligned. + */ + range->me_goal = ocfs2_block_to_cluster_start(inode->i_sb, + range->me_goal); + /* + * moving goal is not allowd to start with a group desc blok(#0 blk) + * let's compromise to the latter cluster. + */ + if (range->me_goal == le64_to_cpu(bg->bg_blkno)) + range->me_goal += c_to_b; + + /* + * validate goal sits within global_bitmap, and return the victim + * group desc + */ + ret = ocfs2_find_victim_alloc_group(inode, range->me_goal, + GLOBAL_BITMAP_SYSTEM_INODE, + OCFS2_INVALID_SLOT, + &goal_bit, &gd_bh); + if (ret) + goto out; + + bg = (struct ocfs2_group_desc *)gd_bh->b_data; + + /* + * movement is not gonna cross two groups. + */ + if ((le16_to_cpu(bg->bg_bits) - goal_bit) * osb->s_clustersize < + range->me_len) { + ret = -EINVAL; + goto out; + } + /* + * more exact validations/adjustments will be performed later during + * moving operation for each extent range. + */ + mlog(0, "extents get ready to be moved to #%llu block\n", + range->me_goal); + +out: + brelse(gd_bh); + + return ret; +} + +static void ocfs2_probe_alloc_group(struct inode *inode, struct buffer_head *bh, + int *goal_bit, u32 move_len, u32 max_hop, + u32 *phys_cpos) +{ + int i, used, last_free_bits = 0, base_bit = *goal_bit; + struct ocfs2_group_desc *gd = (struct ocfs2_group_desc *)bh->b_data; + u32 base_cpos = ocfs2_blocks_to_clusters(inode->i_sb, + le64_to_cpu(gd->bg_blkno)); + + for (i = base_bit; i < le16_to_cpu(gd->bg_bits); i++) { + + used = ocfs2_test_bit(i, (unsigned long *)gd->bg_bitmap); + if (used) { + /* + * we even tried searching the free chunk by jumping + * a 'max_hop' distance, but still failed. + */ + if ((i - base_bit) > max_hop) { + *phys_cpos = 0; + break; + } + + if (last_free_bits) + last_free_bits = 0; + + continue; + } else + last_free_bits++; + + if (last_free_bits == move_len) { + *goal_bit = i; + *phys_cpos = base_cpos + i; + break; + } + } + + mlog(0, "found phys_cpos: %u to fit the wanted moving.\n", *phys_cpos); +} + +static int ocfs2_alloc_dinode_update_counts(struct inode *inode, + handle_t *handle, + struct buffer_head *di_bh, + u32 num_bits, + u16 chain) +{ + int ret; + u32 tmp_used; + struct ocfs2_dinode *di = (struct ocfs2_dinode *) di_bh->b_data; + struct ocfs2_chain_list *cl = + (struct ocfs2_chain_list *) &di->id2.i_chain; + + ret = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh, + OCFS2_JOURNAL_ACCESS_WRITE); + if (ret < 0) { + mlog_errno(ret); + goto out; + } + + tmp_used = le32_to_cpu(di->id1.bitmap1.i_used); + di->id1.bitmap1.i_used = cpu_to_le32(num_bits + tmp_used); + le32_add_cpu(&cl->cl_recs[chain].c_free, -num_bits); + ocfs2_journal_dirty(handle, di_bh); + +out: + return ret; +} + +static inline int ocfs2_block_group_set_bits(handle_t *handle, + struct inode *alloc_inode, + struct ocfs2_group_desc *bg, + struct buffer_head *group_bh, + unsigned int bit_off, + unsigned int num_bits) +{ + int status; + void *bitmap = bg->bg_bitmap; + int journal_type = OCFS2_JOURNAL_ACCESS_WRITE; + + /* All callers get the descriptor via + * ocfs2_read_group_descriptor(). Any corruption is a code bug. */ + BUG_ON(!OCFS2_IS_VALID_GROUP_DESC(bg)); + BUG_ON(le16_to_cpu(bg->bg_free_bits_count) < num_bits); + + mlog(0, "block_group_set_bits: off = %u, num = %u\n", bit_off, + num_bits); + + if (ocfs2_is_cluster_bitmap(alloc_inode)) + journal_type = OCFS2_JOURNAL_ACCESS_UNDO; + + status = ocfs2_journal_access_gd(handle, + INODE_CACHE(alloc_inode), + group_bh, + journal_type); + if (status < 0) { + mlog_errno(status); + goto bail; + } + + le16_add_cpu(&bg->bg_free_bits_count, -num_bits); + if (le16_to_cpu(bg->bg_free_bits_count) > le16_to_cpu(bg->bg_bits)) { + ocfs2_error(alloc_inode->i_sb, "Group descriptor # %llu has bit" + " count %u but claims %u are freed. num_bits %d", + (unsigned long long)le64_to_cpu(bg->bg_blkno), + le16_to_cpu(bg->bg_bits), + le16_to_cpu(bg->bg_free_bits_count), num_bits); + return -EROFS; + } + while (num_bits--) + ocfs2_set_bit(bit_off++, bitmap); + + ocfs2_journal_dirty(handle, group_bh); + +bail: + return status; +} + +static int ocfs2_move_extent(struct ocfs2_move_extents_context *context, + u32 cpos, u32 phys_cpos, u32 *new_phys_cpos, + u32 len, int ext_flags) +{ + int ret, credits = 0, extra_blocks = 0, goal_bit = 0; + handle_t *handle; + struct inode *inode = context->inode; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + struct inode *tl_inode = osb->osb_tl_inode; + struct inode *gb_inode = NULL; + struct buffer_head *gb_bh = NULL; + struct buffer_head *gd_bh = NULL; + struct ocfs2_group_desc *gd; + struct ocfs2_refcount_tree *ref_tree = NULL; + u32 move_max_hop = ocfs2_blocks_to_clusters(inode->i_sb, + context->range->me_threshold); + u64 phys_blkno, new_phys_blkno; + + phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, phys_cpos); + + if ((ext_flags & OCFS2_EXT_REFCOUNTED) && len) { + + BUG_ON(!(OCFS2_I(inode)->ip_dyn_features & + OCFS2_HAS_REFCOUNT_FL)); + + BUG_ON(!context->refcount_loc); + + ret = ocfs2_lock_refcount_tree(osb, context->refcount_loc, 1, + &ref_tree, NULL); + if (ret) { + mlog_errno(ret); + return ret; + } + + ret = ocfs2_prepare_refcount_change_for_del(inode, + context->refcount_loc, + phys_blkno, + len, + &credits, + &extra_blocks); + if (ret) { + mlog_errno(ret); + goto out; + } + } + + ret = ocfs2_lock_allocators_move_extents(inode, &context->et, len, 1, + &context->meta_ac, + NULL, extra_blocks, &credits); + if (ret) { + mlog_errno(ret); + goto out; + } + + /* + * need to count 2 extra credits for global_bitmap inode and + * group descriptor. + */ + credits += OCFS2_INODE_UPDATE_CREDITS + 1; + + /* + * ocfs2_move_extent() didn't reserve any clusters in lock_allocators() + * logic, while we still need to lock the global_bitmap. + */ + gb_inode = ocfs2_get_system_file_inode(osb, GLOBAL_BITMAP_SYSTEM_INODE, + OCFS2_INVALID_SLOT); + if (!gb_inode) { + mlog(ML_ERROR, "unable to get global_bitmap inode\n"); + ret = -EIO; + goto out; + } + + mutex_lock(&gb_inode->i_mutex); + + ret = ocfs2_inode_lock(gb_inode, &gb_bh, 1); + if (ret) { + mlog_errno(ret); + goto out_unlock_gb_mutex; + } + + mutex_lock(&tl_inode->i_mutex); + + handle = ocfs2_start_trans(osb, credits); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + mlog_errno(ret); + goto out_unlock_tl_inode; + } + + new_phys_blkno = ocfs2_clusters_to_blocks(inode->i_sb, *new_phys_cpos); + ret = ocfs2_find_victim_alloc_group(inode, new_phys_blkno, + GLOBAL_BITMAP_SYSTEM_INODE, + OCFS2_INVALID_SLOT, + &goal_bit, &gd_bh); + if (ret) { + mlog_errno(ret); + goto out_commit; + } + + /* + * probe the victim cluster group to find a proper + * region to fit wanted movement, it even will perfrom + * a best-effort attempt by compromising to a threshold + * around the goal. + */ + ocfs2_probe_alloc_group(inode, gd_bh, &goal_bit, len, move_max_hop, + new_phys_cpos); + if (!new_phys_cpos) { + ret = -ENOSPC; + goto out_commit; + } + + ret = __ocfs2_move_extent(handle, context, cpos, len, phys_cpos, + *new_phys_cpos, ext_flags); + if (ret) { + mlog_errno(ret); + goto out_commit; + } + + gd = (struct ocfs2_group_desc *)gd_bh->b_data; + ret = ocfs2_alloc_dinode_update_counts(gb_inode, handle, gb_bh, len, + le16_to_cpu(gd->bg_chain)); + if (ret) { + mlog_errno(ret); + goto out_commit; + } + + ret = ocfs2_block_group_set_bits(handle, gb_inode, gd, gd_bh, + goal_bit, len); + if (ret) + mlog_errno(ret); + + /* + * Here we should write the new page out first if we are + * in write-back mode. + */ + ret = ocfs2_cow_sync_writeback(inode->i_sb, context->inode, cpos, len); + if (ret) + mlog_errno(ret); + +out_commit: + ocfs2_commit_trans(osb, handle); + brelse(gd_bh); + +out_unlock_tl_inode: + mutex_unlock(&tl_inode->i_mutex); + + ocfs2_inode_unlock(gb_inode, 1); +out_unlock_gb_mutex: + mutex_unlock(&gb_inode->i_mutex); + brelse(gb_bh); + iput(gb_inode); + +out: + if (context->meta_ac) { + ocfs2_free_alloc_context(context->meta_ac); + context->meta_ac = NULL; + } + + if (ref_tree) + ocfs2_unlock_refcount_tree(osb, ref_tree, 1); + + return ret; +} + +/* + * Helper to calculate the defraging length in one run according to threshold. + */ +static void ocfs2_calc_extent_defrag_len(u32 *alloc_size, u32 *len_defraged, + u32 threshold, int *skip) +{ + if ((*alloc_size + *len_defraged) < threshold) { + /* + * proceed defragmentation until we meet the thresh + */ + *len_defraged += *alloc_size; + } else if (*len_defraged == 0) { + /* + * XXX: skip a large extent. + */ + *skip = 1; + } else { + /* + * split this extent to coalesce with former pieces as + * to reach the threshold. + * + * we're done here with one cycle of defragmentation + * in a size of 'thresh', resetting 'len_defraged' + * forces a new defragmentation. + */ + *alloc_size = threshold - *len_defraged; + *len_defraged = 0; + } +} + +static int __ocfs2_move_extents_range(struct buffer_head *di_bh, + struct ocfs2_move_extents_context *context) +{ + int ret = 0, flags, do_defrag, skip = 0; + u32 cpos, phys_cpos, move_start, len_to_move, alloc_size; + u32 len_defraged = 0, defrag_thresh = 0, new_phys_cpos = 0; + + struct inode *inode = context->inode; + struct ocfs2_dinode *di = (struct ocfs2_dinode *)di_bh->b_data; + struct ocfs2_move_extents *range = context->range; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + + if ((inode->i_size == 0) || (range->me_len == 0)) + return 0; + + if (OCFS2_I(inode)->ip_dyn_features & OCFS2_INLINE_DATA_FL) + return 0; + + context->refcount_loc = le64_to_cpu(di->i_refcount_loc); + + ocfs2_init_dinode_extent_tree(&context->et, INODE_CACHE(inode), di_bh); + ocfs2_init_dealloc_ctxt(&context->dealloc); + + /* + * TO-DO XXX: + * + * - xattr extents. + */ + + do_defrag = context->auto_defrag; + + /* + * extents moving happens in unit of clusters, for the sake + * of simplicity, we may ignore two clusters where 'byte_start' + * and 'byte_start + len' were within. + */ + move_start = ocfs2_clusters_for_bytes(osb->sb, range->me_start); + len_to_move = (range->me_start + range->me_len) >> + osb->s_clustersize_bits; + if (len_to_move >= move_start) + len_to_move -= move_start; + else + len_to_move = 0; + + if (do_defrag) { + defrag_thresh = range->me_threshold >> osb->s_clustersize_bits; + if (defrag_thresh <= 1) + goto done; + } else + new_phys_cpos = ocfs2_blocks_to_clusters(inode->i_sb, + range->me_goal); + + mlog(0, "Inode: %llu, start: %llu, len: %llu, cstart: %u, clen: %u, " + "thresh: %u\n", + (unsigned long long)OCFS2_I(inode)->ip_blkno, + (unsigned long long)range->me_start, + (unsigned long long)range->me_len, + move_start, len_to_move, defrag_thresh); + + cpos = move_start; + while (len_to_move) { + ret = ocfs2_get_clusters(inode, cpos, &phys_cpos, &alloc_size, + &flags); + if (ret) { + mlog_errno(ret); + goto out; + } + + if (alloc_size > len_to_move) + alloc_size = len_to_move; + + /* + * XXX: how to deal with a hole: + * + * - skip the hole of course + * - force a new defragmentation + */ + if (!phys_cpos) { + if (do_defrag) + len_defraged = 0; + + goto next; + } + + if (do_defrag) { + ocfs2_calc_extent_defrag_len(&alloc_size, &len_defraged, + defrag_thresh, &skip); + /* + * skip large extents + */ + if (skip) { + skip = 0; + goto next; + } + + mlog(0, "#Defrag: cpos: %u, phys_cpos: %u, " + "alloc_size: %u, len_defraged: %u\n", + cpos, phys_cpos, alloc_size, len_defraged); + + ret = ocfs2_defrag_extent(context, cpos, phys_cpos, + &alloc_size, flags); + } else { + ret = ocfs2_move_extent(context, cpos, phys_cpos, + &new_phys_cpos, alloc_size, + flags); + + new_phys_cpos += alloc_size; + } + + if (ret < 0) { + mlog_errno(ret); + goto out; + } + + context->clusters_moved += alloc_size; +next: + cpos += alloc_size; + len_to_move -= alloc_size; + } + +done: + range->me_flags |= OCFS2_MOVE_EXT_FL_COMPLETE; + +out: + range->me_moved_len = ocfs2_clusters_to_bytes(osb->sb, + context->clusters_moved); + range->me_new_offset = ocfs2_clusters_to_bytes(osb->sb, + context->new_phys_cpos); + + ocfs2_schedule_truncate_log_flush(osb, 1); + ocfs2_run_deallocs(osb, &context->dealloc); + + return ret; +} + +static int ocfs2_move_extents(struct ocfs2_move_extents_context *context) +{ + int status; + handle_t *handle; + struct inode *inode = context->inode; + struct ocfs2_dinode *di; + struct buffer_head *di_bh = NULL; + struct ocfs2_super *osb = OCFS2_SB(inode->i_sb); + + if (!inode) + return -ENOENT; + + if (ocfs2_is_hard_readonly(osb) || ocfs2_is_soft_readonly(osb)) + return -EROFS; + + mutex_lock(&inode->i_mutex); + + /* + * This prevents concurrent writes from other nodes + */ + status = ocfs2_rw_lock(inode, 1); + if (status) { + mlog_errno(status); + goto out; + } + + status = ocfs2_inode_lock(inode, &di_bh, 1); + if (status) { + mlog_errno(status); + goto out_rw_unlock; + } + + /* + * rememer ip_xattr_sem also needs to be held if necessary + */ + down_write(&OCFS2_I(inode)->ip_alloc_sem); + + status = __ocfs2_move_extents_range(di_bh, context); + + up_write(&OCFS2_I(inode)->ip_alloc_sem); + if (status) { + mlog_errno(status); + goto out_inode_unlock; + } + + /* + * We update ctime for these changes + */ + handle = ocfs2_start_trans(osb, OCFS2_INODE_UPDATE_CREDITS); + if (IS_ERR(handle)) { + status = PTR_ERR(handle); + mlog_errno(status); + goto out_inode_unlock; + } + + status = ocfs2_journal_access_di(handle, INODE_CACHE(inode), di_bh, + OCFS2_JOURNAL_ACCESS_WRITE); + if (status) { + mlog_errno(status); + goto out_commit; + } + + di = (struct ocfs2_dinode *)di_bh->b_data; + inode->i_ctime = CURRENT_TIME; + di->i_ctime = cpu_to_le64(inode->i_ctime.tv_sec); + di->i_ctime_nsec = cpu_to_le32(inode->i_ctime.tv_nsec); + + ocfs2_journal_dirty(handle, di_bh); + +out_commit: + ocfs2_commit_trans(osb, handle); + +out_inode_unlock: + brelse(di_bh); + ocfs2_inode_unlock(inode, 1); +out_rw_unlock: + ocfs2_rw_unlock(inode, 1); +out: + mutex_unlock(&inode->i_mutex); + + return status; +} + +int ocfs2_ioctl_move_extents(struct file *filp, void __user *argp) +{ + int status; + + struct inode *inode = filp->f_path.dentry->d_inode; + struct ocfs2_move_extents range; + struct ocfs2_move_extents_context *context = NULL; + + status = mnt_want_write(filp->f_path.mnt); + if (status) + return status; + + if ((!S_ISREG(inode->i_mode)) || !(filp->f_mode & FMODE_WRITE)) + goto out; + + if (inode->i_flags & (S_IMMUTABLE|S_APPEND)) { + status = -EPERM; + goto out; + } + + context = kzalloc(sizeof(struct ocfs2_move_extents_context), GFP_NOFS); + if (!context) { + status = -ENOMEM; + mlog_errno(status); + goto out; + } + + context->inode = inode; + context->file = filp; + + if (argp) { + if (copy_from_user(&range, (struct ocfs2_move_extents *)argp, + sizeof(range))) { + status = -EFAULT; + goto out; + } + } else { + status = -EINVAL; + goto out; + } + + if (range.me_start > i_size_read(inode)) + goto out; + + if (range.me_start + range.me_len > i_size_read(inode)) + range.me_len = i_size_read(inode) - range.me_start; + + context->range = ⦥ + + if (range.me_flags & OCFS2_MOVE_EXT_FL_AUTO_DEFRAG) { + context->auto_defrag = 1; + /* + * ok, the default theshold for the defragmentation + * is 1M, since our maximum clustersize was 1M also. + * any thought? + */ + if (!range.me_threshold) + range.me_threshold = 1024 * 1024; + + if (range.me_threshold > i_size_read(inode)) + range.me_threshold = i_size_read(inode); + + if (range.me_flags & OCFS2_MOVE_EXT_FL_PART_DEFRAG) + context->partial = 1; + } else { + /* + * first best-effort attempt to validate and adjust the goal + * (physical address in block), while it can't guarantee later + * operation can succeed all the time since global_bitmap may + * change a bit over time. + */ + + status = ocfs2_validate_and_adjust_move_goal(inode, &range); + if (status) + goto out; + } + + status = ocfs2_move_extents(context); + if (status) + mlog_errno(status); +out: + /* + * movement/defragmentation may end up being partially completed, + * that's the reason why we need to return userspace the finished + * length and new_offset even if failure happens somewhere. + */ + if (argp) { + if (copy_to_user((struct ocfs2_move_extents *)argp, &range, + sizeof(range))) + status = -EFAULT; + } + + kfree(context); + + mnt_drop_write(filp->f_path.mnt); + + return status; +} diff --git a/fs/ocfs2/move_extents.h b/fs/ocfs2/move_extents.h new file mode 100644 index 000000000000..4e143e811441 --- /dev/null +++ b/fs/ocfs2/move_extents.h @@ -0,0 +1,22 @@ +/* -*- mode: c; c-basic-offset: 8; -*- + * vim: noexpandtab sw=8 ts=8 sts=0: + * + * move_extents.h + * + * Copyright (C) 2011 Oracle. All rights reserved. + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public + * License version 2 as published by the Free Software Foundation. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + * General Public License for more details. + */ +#ifndef OCFS2_MOVE_EXTENTS_H +#define OCFS2_MOVE_EXTENTS_H + +int ocfs2_ioctl_move_extents(struct file *filp, void __user *argp); + +#endif /* OCFS2_MOVE_EXTENTS_H */ diff --git a/fs/ocfs2/namei.c b/fs/ocfs2/namei.c index 28f2cc1080d8..e5d738cd9cc0 100644 --- a/fs/ocfs2/namei.c +++ b/fs/ocfs2/namei.c @@ -2128,7 +2128,7 @@ leave: } /** - * ocfs2_prep_new_orphaned_file() - Prepare the orphan dir to recieve a newly + * ocfs2_prep_new_orphaned_file() - Prepare the orphan dir to receive a newly * allocated file. This is different from the typical 'add to orphan dir' * operation in that the inode does not yet exist. This is a problem because * the orphan dir stringifies the inode block number to come up with it's diff --git a/fs/ocfs2/ocfs2_fs.h b/fs/ocfs2/ocfs2_fs.h index bf2e7764920e..938387a10d5d 100644 --- a/fs/ocfs2/ocfs2_fs.h +++ b/fs/ocfs2/ocfs2_fs.h @@ -441,7 +441,7 @@ static unsigned char ocfs2_type_by_mode[S_IFMT >> S_SHIFT] = { struct ocfs2_block_check { /*00*/ __le32 bc_crc32e; /* 802.3 Ethernet II CRC32 */ __le16 bc_ecc; /* Single-error-correction parity vector. - This is a simple Hamming code dependant + This is a simple Hamming code dependent on the blocksize. OCFS2's maximum blocksize, 4K, requires 16 parity bits, so we fit in __le16. */ @@ -750,7 +750,7 @@ struct ocfs2_dinode { after an unclean shutdown */ } journal1; - } id1; /* Inode type dependant 1 */ + } id1; /* Inode type dependent 1 */ /*C0*/ union { struct ocfs2_super_block i_super; struct ocfs2_local_alloc i_lab; @@ -1019,7 +1019,7 @@ struct ocfs2_xattr_entry { __le16 xe_name_offset; /* byte offset from the 1st entry in the local xattr storage(inode, xattr block or xattr bucket). */ - __u8 xe_name_len; /* xattr name len, does't include prefix. */ + __u8 xe_name_len; /* xattr name len, doesn't include prefix. */ __u8 xe_type; /* the low 7 bits indicate the name prefix * type and the highest bit indicates whether * the EA is stored in the local storage. */ diff --git a/fs/ocfs2/ocfs2_ioctl.h b/fs/ocfs2/ocfs2_ioctl.h index b46f39bf7438..5b27ff1fa577 100644 --- a/fs/ocfs2/ocfs2_ioctl.h +++ b/fs/ocfs2/ocfs2_ioctl.h @@ -142,6 +142,38 @@ struct ocfs2_info_journal_size { __u64 ij_journal_size; }; +struct ocfs2_info_freeinode { + struct ocfs2_info_request ifi_req; + struct ocfs2_info_local_freeinode { + __u64 lfi_total; + __u64 lfi_free; + } ifi_stat[OCFS2_MAX_SLOTS]; + __u32 ifi_slotnum; /* out */ + __u32 ifi_pad; +}; + +#define OCFS2_INFO_MAX_HIST (32) + +struct ocfs2_info_freefrag { + struct ocfs2_info_request iff_req; + struct ocfs2_info_freefrag_stats { /* (out) */ + struct ocfs2_info_free_chunk_list { + __u32 fc_chunks[OCFS2_INFO_MAX_HIST]; + __u32 fc_clusters[OCFS2_INFO_MAX_HIST]; + } ffs_fc_hist; + __u32 ffs_clusters; + __u32 ffs_free_clusters; + __u32 ffs_free_chunks; + __u32 ffs_free_chunks_real; + __u32 ffs_min; /* Minimum free chunksize in clusters */ + __u32 ffs_max; + __u32 ffs_avg; + __u32 ffs_pad; + } iff_ffs; + __u32 iff_chunksize; /* chunksize in clusters(in) */ + __u32 iff_pad; +}; + /* Codes for ocfs2_info_request */ enum ocfs2_info_type { OCFS2_INFO_CLUSTERSIZE = 1, @@ -151,6 +183,8 @@ enum ocfs2_info_type { OCFS2_INFO_UUID, OCFS2_INFO_FS_FEATURES, OCFS2_INFO_JOURNAL_SIZE, + OCFS2_INFO_FREEINODE, + OCFS2_INFO_FREEFRAG, OCFS2_INFO_NUM_TYPES }; @@ -171,4 +205,38 @@ enum ocfs2_info_type { #define OCFS2_IOC_INFO _IOR('o', 5, struct ocfs2_info) +struct ocfs2_move_extents { +/* All values are in bytes */ + /* in */ + __u64 me_start; /* Virtual start in the file to move */ + __u64 me_len; /* Length of the extents to be moved */ + __u64 me_goal; /* Physical offset of the goal, + it's in block unit */ + __u64 me_threshold; /* Maximum distance from goal or threshold + for auto defragmentation */ + __u64 me_flags; /* Flags for the operation: + * - auto defragmentation. + * - refcount,xattr cases. + */ + /* out */ + __u64 me_moved_len; /* Moved/defraged length */ + __u64 me_new_offset; /* Resulting physical location */ + __u32 me_reserved[2]; /* Reserved for futhure */ +}; + +#define OCFS2_MOVE_EXT_FL_AUTO_DEFRAG (0x00000001) /* Kernel manages to + claim new clusters + as the goal place + for extents moving */ +#define OCFS2_MOVE_EXT_FL_PART_DEFRAG (0x00000002) /* Allow partial extent + moving, is to make + movement less likely + to fail, may make fs + even more fragmented */ +#define OCFS2_MOVE_EXT_FL_COMPLETE (0x00000004) /* Move or defragmenation + completely gets done. + */ + +#define OCFS2_IOC_MOVE_EXT _IOW('o', 6, struct ocfs2_move_extents) + #endif /* OCFS2_IOCTL_H */ diff --git a/fs/ocfs2/ocfs2_trace.h b/fs/ocfs2/ocfs2_trace.h index a1dae5bb54ac..3b481f490633 100644 --- a/fs/ocfs2/ocfs2_trace.h +++ b/fs/ocfs2/ocfs2_trace.h @@ -688,6 +688,31 @@ TRACE_EVENT(ocfs2_cache_block_dealloc, __entry->blkno, __entry->bit) ); +TRACE_EVENT(ocfs2_trim_extent, + TP_PROTO(struct super_block *sb, unsigned long long blk, + unsigned long long count), + TP_ARGS(sb, blk, count), + TP_STRUCT__entry( + __field(int, dev_major) + __field(int, dev_minor) + __field(unsigned long long, blk) + __field(__u64, count) + ), + TP_fast_assign( + __entry->dev_major = MAJOR(sb->s_dev); + __entry->dev_minor = MINOR(sb->s_dev); + __entry->blk = blk; + __entry->count = count; + ), + TP_printk("%d %d %llu %llu", + __entry->dev_major, __entry->dev_minor, + __entry->blk, __entry->count) +); + +DEFINE_OCFS2_ULL_UINT_UINT_UINT_EVENT(ocfs2_trim_group); + +DEFINE_OCFS2_ULL_ULL_ULL_EVENT(ocfs2_trim_fs); + /* End of trace events for fs/ocfs2/alloc.c. */ /* Trace events for fs/ocfs2/localalloc.c. */ diff --git a/fs/ocfs2/quota_global.c b/fs/ocfs2/quota_global.c index 279aef68025b..92fcd575775a 100644 --- a/fs/ocfs2/quota_global.c +++ b/fs/ocfs2/quota_global.c @@ -556,7 +556,7 @@ int __ocfs2_sync_dquot(struct dquot *dquot, int freeing) spin_unlock(&dq_data_lock); err = ocfs2_qinfo_lock(info, freeing); if (err < 0) { - mlog(ML_ERROR, "Failed to lock quota info, loosing quota write" + mlog(ML_ERROR, "Failed to lock quota info, losing quota write" " (type=%d, id=%u)\n", dquot->dq_type, (unsigned)dquot->dq_id); goto out; diff --git a/fs/ocfs2/refcounttree.c b/fs/ocfs2/refcounttree.c index 5d32749c896d..ebfd3825f12a 100644 --- a/fs/ocfs2/refcounttree.c +++ b/fs/ocfs2/refcounttree.c @@ -66,7 +66,7 @@ struct ocfs2_cow_context { u32 *num_clusters, unsigned int *extent_flags); int (*cow_duplicate_clusters)(handle_t *handle, - struct ocfs2_cow_context *context, + struct file *file, u32 cpos, u32 old_cluster, u32 new_cluster, u32 new_len); }; @@ -2921,20 +2921,21 @@ static int ocfs2_clear_cow_buffer(handle_t *handle, struct buffer_head *bh) return 0; } -static int ocfs2_duplicate_clusters_by_page(handle_t *handle, - struct ocfs2_cow_context *context, - u32 cpos, u32 old_cluster, - u32 new_cluster, u32 new_len) +int ocfs2_duplicate_clusters_by_page(handle_t *handle, + struct file *file, + u32 cpos, u32 old_cluster, + u32 new_cluster, u32 new_len) { int ret = 0, partial; - struct ocfs2_caching_info *ci = context->data_et.et_ci; + struct inode *inode = file->f_path.dentry->d_inode; + struct ocfs2_caching_info *ci = INODE_CACHE(inode); struct super_block *sb = ocfs2_metadata_cache_get_super(ci); u64 new_block = ocfs2_clusters_to_blocks(sb, new_cluster); struct page *page; pgoff_t page_index; unsigned int from, to, readahead_pages; loff_t offset, end, map_end; - struct address_space *mapping = context->inode->i_mapping; + struct address_space *mapping = inode->i_mapping; trace_ocfs2_duplicate_clusters_by_page(cpos, old_cluster, new_cluster, new_len); @@ -2948,8 +2949,8 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle, * We only duplicate pages until we reach the page contains i_size - 1. * So trim 'end' to i_size. */ - if (end > i_size_read(context->inode)) - end = i_size_read(context->inode); + if (end > i_size_read(inode)) + end = i_size_read(inode); while (offset < end) { page_index = offset >> PAGE_CACHE_SHIFT; @@ -2972,10 +2973,9 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle, if (PAGE_CACHE_SIZE <= OCFS2_SB(sb)->s_clustersize) BUG_ON(PageDirty(page)); - if (PageReadahead(page) && context->file) { + if (PageReadahead(page)) { page_cache_async_readahead(mapping, - &context->file->f_ra, - context->file, + &file->f_ra, file, page, page_index, readahead_pages); } @@ -2999,8 +2999,7 @@ static int ocfs2_duplicate_clusters_by_page(handle_t *handle, } } - ocfs2_map_and_dirty_page(context->inode, - handle, from, to, + ocfs2_map_and_dirty_page(inode, handle, from, to, page, 0, &new_block); mark_page_accessed(page); unlock: @@ -3015,14 +3014,15 @@ unlock: return ret; } -static int ocfs2_duplicate_clusters_by_jbd(handle_t *handle, - struct ocfs2_cow_context *context, - u32 cpos, u32 old_cluster, - u32 new_cluster, u32 new_len) +int ocfs2_duplicate_clusters_by_jbd(handle_t *handle, + struct file *file, + u32 cpos, u32 old_cluster, + u32 new_cluster, u32 new_len) { int ret = 0; - struct super_block *sb = context->inode->i_sb; - struct ocfs2_caching_info *ci = context->data_et.et_ci; + struct inode *inode = file->f_path.dentry->d_inode; + struct super_block *sb = inode->i_sb; + struct ocfs2_caching_info *ci = INODE_CACHE(inode); int i, blocks = ocfs2_clusters_to_blocks(sb, new_len); u64 old_block = ocfs2_clusters_to_blocks(sb, old_cluster); u64 new_block = ocfs2_clusters_to_blocks(sb, new_cluster); @@ -3145,8 +3145,8 @@ static int ocfs2_replace_clusters(handle_t *handle, /*If the old clusters is unwritten, no need to duplicate. */ if (!(ext_flags & OCFS2_EXT_UNWRITTEN)) { - ret = context->cow_duplicate_clusters(handle, context, cpos, - old, new, len); + ret = context->cow_duplicate_clusters(handle, context->file, + cpos, old, new, len); if (ret) { mlog_errno(ret); goto out; @@ -3162,22 +3162,22 @@ out: return ret; } -static int ocfs2_cow_sync_writeback(struct super_block *sb, - struct ocfs2_cow_context *context, - u32 cpos, u32 num_clusters) +int ocfs2_cow_sync_writeback(struct super_block *sb, + struct inode *inode, + u32 cpos, u32 num_clusters) { int ret = 0; loff_t offset, end, map_end; pgoff_t page_index; struct page *page; - if (ocfs2_should_order_data(context->inode)) + if (ocfs2_should_order_data(inode)) return 0; offset = ((loff_t)cpos) << OCFS2_SB(sb)->s_clustersize_bits; end = offset + (num_clusters << OCFS2_SB(sb)->s_clustersize_bits); - ret = filemap_fdatawrite_range(context->inode->i_mapping, + ret = filemap_fdatawrite_range(inode->i_mapping, offset, end - 1); if (ret < 0) { mlog_errno(ret); @@ -3190,7 +3190,7 @@ static int ocfs2_cow_sync_writeback(struct super_block *sb, if (map_end > end) map_end = end; - page = find_or_create_page(context->inode->i_mapping, + page = find_or_create_page(inode->i_mapping, page_index, GFP_NOFS); BUG_ON(!page); @@ -3349,7 +3349,7 @@ static int ocfs2_make_clusters_writable(struct super_block *sb, * in write-back mode. */ if (context->get_clusters == ocfs2_di_get_clusters) { - ret = ocfs2_cow_sync_writeback(sb, context, cpos, + ret = ocfs2_cow_sync_writeback(sb, context->inode, cpos, orig_num_clusters); if (ret) mlog_errno(ret); @@ -3706,7 +3706,7 @@ int ocfs2_refcount_cow_xattr(struct inode *inode, context->cow_start = cow_start; context->cow_len = cow_len; context->ref_tree = ref_tree; - context->ref_root_bh = ref_root_bh;; + context->ref_root_bh = ref_root_bh; context->cow_object = xv; context->cow_duplicate_clusters = ocfs2_duplicate_clusters_by_jbd; diff --git a/fs/ocfs2/refcounttree.h b/fs/ocfs2/refcounttree.h index c8ce46f7d8e3..7754608c83a4 100644 --- a/fs/ocfs2/refcounttree.h +++ b/fs/ocfs2/refcounttree.h @@ -84,6 +84,17 @@ int ocfs2_refcount_cow_xattr(struct inode *inode, struct buffer_head *ref_root_bh, u32 cpos, u32 write_len, struct ocfs2_post_refcount *post); +int ocfs2_duplicate_clusters_by_page(handle_t *handle, + struct file *file, + u32 cpos, u32 old_cluster, + u32 new_cluster, u32 new_len); +int ocfs2_duplicate_clusters_by_jbd(handle_t *handle, + struct file *file, + u32 cpos, u32 old_cluster, + u32 new_cluster, u32 new_len); +int ocfs2_cow_sync_writeback(struct super_block *sb, + struct inode *inode, + u32 cpos, u32 num_clusters); int ocfs2_add_refcount_flag(struct inode *inode, struct ocfs2_extent_tree *data_et, struct ocfs2_caching_info *ref_ci, diff --git a/fs/ocfs2/reservations.h b/fs/ocfs2/reservations.h index 1e49cc29d06c..42c2b804f3fd 100644 --- a/fs/ocfs2/reservations.h +++ b/fs/ocfs2/reservations.h @@ -29,7 +29,7 @@ struct ocfs2_alloc_reservation { struct rb_node r_node; - unsigned int r_start; /* Begining of current window */ + unsigned int r_start; /* Beginning of current window */ unsigned int r_len; /* Length of the window */ unsigned int r_last_len; /* Length of most recent alloc */ diff --git a/fs/ocfs2/stackglue.h b/fs/ocfs2/stackglue.h index 8ce7398ae1d2..1ec56fdb8d0d 100644 --- a/fs/ocfs2/stackglue.h +++ b/fs/ocfs2/stackglue.h @@ -126,7 +126,7 @@ struct ocfs2_stack_operations { * * ->connect() must not return until it is guaranteed that * - * - Node down notifications for the filesystem will be recieved + * - Node down notifications for the filesystem will be received * and passed to conn->cc_recovery_handler(). * - Locking requests for the filesystem will be processed. */ diff --git a/fs/ocfs2/suballoc.c b/fs/ocfs2/suballoc.c index ab6e2061074f..ba5d97e4a73e 100644 --- a/fs/ocfs2/suballoc.c +++ b/fs/ocfs2/suballoc.c @@ -1511,7 +1511,7 @@ static int ocfs2_cluster_group_search(struct inode *inode, max_bits = le16_to_cpu(gd->bg_bits); /* Tail groups in cluster bitmaps which aren't cpg - * aligned are prone to partial extention by a failed + * aligned are prone to partial extension by a failed * fs resize. If the file system resize never got to * update the dinode cluster count, then we don't want * to trust any clusters past it, regardless of what @@ -2459,7 +2459,7 @@ static int _ocfs2_free_suballoc_bits(handle_t *handle, /* The alloc_bh comes from ocfs2_free_dinode() or * ocfs2_free_clusters(). The callers have all locked the * allocator and gotten alloc_bh from the lock call. This - * validates the dinode buffer. Any corruption that has happended + * validates the dinode buffer. Any corruption that has happened * is a code bug. */ BUG_ON(!OCFS2_IS_VALID_DINODE(fe)); BUG_ON((count + start_bit) > ocfs2_bits_per_group(cl)); diff --git a/fs/ocfs2/super.c b/fs/ocfs2/super.c index 69fa11b35aa4..cdbaf5e97308 100644 --- a/fs/ocfs2/super.c +++ b/fs/ocfs2/super.c @@ -41,6 +41,7 @@ #include <linux/mount.h> #include <linux/seq_file.h> #include <linux/quotaops.h> +#include <linux/cleancache.h> #define CREATE_TRACE_POINTS #include "ocfs2_trace.h" @@ -78,7 +79,7 @@ static struct kmem_cache *ocfs2_inode_cachep = NULL; struct kmem_cache *ocfs2_dquot_cachep; struct kmem_cache *ocfs2_qf_chunk_cachep; -/* OCFS2 needs to schedule several differnt types of work which +/* OCFS2 needs to schedule several different types of work which * require cluster locking, disk I/O, recovery waits, etc. Since these * types of work tend to be heavy we avoid using the kernel events * workqueue and schedule on our own. */ @@ -1566,7 +1567,7 @@ static int ocfs2_show_options(struct seq_file *s, struct vfsmount *mnt) if (osb->preferred_slot != OCFS2_INVALID_SLOT) seq_printf(s, ",preferred_slot=%d", osb->preferred_slot); - if (osb->s_atime_quantum != OCFS2_DEFAULT_ATIME_QUANTUM) + if (!(mnt->mnt_flags & MNT_NOATIME) && !(mnt->mnt_flags & MNT_RELATIME)) seq_printf(s, ",atime_quantum=%u", osb->s_atime_quantum); if (osb->osb_commit_interval) @@ -2352,6 +2353,7 @@ static int ocfs2_initialize_super(struct super_block *sb, mlog_errno(status); goto bail; } + cleancache_init_shared_fs((char *)&uuid_net_key, sb); bail: return status; diff --git a/fs/ocfs2/xattr.c b/fs/ocfs2/xattr.c index 57a215dc2d9b..81ecf9c0bf0a 100644 --- a/fs/ocfs2/xattr.c +++ b/fs/ocfs2/xattr.c @@ -3554,7 +3554,7 @@ int ocfs2_xattr_set(struct inode *inode, down_write(&OCFS2_I(inode)->ip_xattr_sem); /* * Scan inode and external block to find the same name - * extended attribute and collect search infomation. + * extended attribute and collect search information. */ ret = ocfs2_xattr_ibody_find(inode, name_index, name, &xis); if (ret) @@ -3578,7 +3578,7 @@ int ocfs2_xattr_set(struct inode *inode, goto cleanup; } - /* Check whether the value is refcounted and do some prepartion. */ + /* Check whether the value is refcounted and do some preparation. */ if (OCFS2_I(inode)->ip_dyn_features & OCFS2_HAS_REFCOUNT_FL && (!xis.not_found || !xbs.not_found)) { ret = ocfs2_prepare_refcount_xattr(inode, di, &xi, diff --git a/fs/omfs/dir.c b/fs/omfs/dir.c index de4ff29f1e05..3b8d3979e03b 100644 --- a/fs/omfs/dir.c +++ b/fs/omfs/dir.c @@ -240,7 +240,9 @@ static int omfs_remove(struct inode *dir, struct dentry *dentry) struct inode *inode = dentry->d_inode; int ret; - if (S_ISDIR(inode->i_mode) && !omfs_dir_is_empty(inode)) + + if (S_ISDIR(inode->i_mode) && + !omfs_dir_is_empty(inode)) return -ENOTEMPTY; ret = omfs_delete_entry(dentry); diff --git a/fs/partitions/check.c b/fs/partitions/check.c index ac546975031f..f82e762eeca2 100644 --- a/fs/partitions/check.c +++ b/fs/partitions/check.c @@ -255,7 +255,13 @@ ssize_t part_discard_alignment_show(struct device *dev, struct device_attribute *attr, char *buf) { struct hd_struct *p = dev_to_part(dev); - return sprintf(buf, "%u\n", p->discard_alignment); + struct gendisk *disk = dev_to_disk(dev); + unsigned int alignment = 0; + + if (disk->queue) + alignment = queue_limit_discard_alignment(&disk->queue->limits, + p->start_sect); + return sprintf(buf, "%u\n", alignment); } ssize_t part_stat_show(struct device *dev, @@ -449,8 +455,6 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno, p->start_sect = start; p->alignment_offset = queue_limit_alignment_offset(&disk->queue->limits, start); - p->discard_alignment = - queue_limit_discard_alignment(&disk->queue->limits, start); p->nr_sects = len; p->partno = partno; p->policy = get_disk_ro(disk); @@ -500,7 +504,7 @@ struct hd_struct *add_partition(struct gendisk *disk, int partno, /* everything is up and running, commence */ rcu_assign_pointer(ptbl->part[partno], p); - /* suppress uevent if the disk supresses it */ + /* suppress uevent if the disk suppresses it */ if (!dev_get_uevent_suppress(ddev)) kobject_uevent(&pdev->kobj, KOBJ_ADD); @@ -585,7 +589,7 @@ rescan: /* * If any partition code tried to read beyond EOD, try * unlocking native capacity even if partition table is - * sucessfully read as we could be missing some partitions. + * successfully read as we could be missing some partitions. */ if (state->access_beyond_eod) { printk(KERN_WARNING diff --git a/fs/partitions/efi.c b/fs/partitions/efi.c index ac0ccb5026a2..6296b403c67a 100644 --- a/fs/partitions/efi.c +++ b/fs/partitions/efi.c @@ -310,6 +310,15 @@ static int is_gpt_valid(struct parsed_partitions *state, u64 lba, goto fail; } + /* Check the GUID Partition Table header size */ + if (le32_to_cpu((*gpt)->header_size) > + bdev_logical_block_size(state->bdev)) { + pr_debug("GUID Partition Table Header size is wrong: %u > %u\n", + le32_to_cpu((*gpt)->header_size), + bdev_logical_block_size(state->bdev)); + goto fail; + } + /* Check the GUID Partition Table CRC */ origcrc = le32_to_cpu((*gpt)->header_crc32); (*gpt)->header_crc32 = 0; @@ -348,6 +357,12 @@ static int is_gpt_valid(struct parsed_partitions *state, u64 lba, goto fail; } + /* Check that sizeof_partition_entry has the correct value */ + if (le32_to_cpu((*gpt)->sizeof_partition_entry) != sizeof(gpt_entry)) { + pr_debug("GUID Partitition Entry Size check failed.\n"); + goto fail; + } + if (!(*ptes = alloc_read_gpt_entries(state, *gpt))) goto fail; diff --git a/fs/partitions/ldm.c b/fs/partitions/ldm.c index b10e3540d5b7..af9fdf046769 100644 --- a/fs/partitions/ldm.c +++ b/fs/partitions/ldm.c @@ -565,7 +565,7 @@ static bool ldm_validate_partition_table(struct parsed_partitions *state) data = read_part_sector(state, 0, §); if (!data) { - ldm_crit ("Disk read failed."); + ldm_info ("Disk read failed."); return false; } @@ -1299,6 +1299,11 @@ static bool ldm_frag_add (const u8 *data, int size, struct list_head *frags) BUG_ON (!data || !frags); + if (size < 2 * VBLK_SIZE_HEAD) { + ldm_error("Value of size is to small."); + return false; + } + group = get_unaligned_be32(data + 0x08); rec = get_unaligned_be16(data + 0x0C); num = get_unaligned_be16(data + 0x0E); @@ -1306,6 +1311,10 @@ static bool ldm_frag_add (const u8 *data, int size, struct list_head *frags) ldm_error ("A VBLK claims to have %d parts.", num); return false; } + if (rec >= num) { + ldm_error("REC value (%d) exceeds NUM value (%d)", rec, num); + return false; + } list_for_each (item, frags) { f = list_entry (item, struct frag, list); @@ -1326,6 +1335,11 @@ static bool ldm_frag_add (const u8 *data, int size, struct list_head *frags) list_add_tail (&f->list, frags); found: + if (rec >= f->num) { + ldm_error("REC value (%d) exceeds NUM value (%d)", rec, f->num); + return false; + } + if (f->map & (1 << rec)) { ldm_error ("Duplicate VBLK, part %d.", rec); f->map &= 0x7F; /* Mark the group as broken */ @@ -1334,10 +1348,9 @@ found: f->map |= (1 << rec); - if (num > 0) { - data += VBLK_SIZE_HEAD; - size -= VBLK_SIZE_HEAD; - } + data += VBLK_SIZE_HEAD; + size -= VBLK_SIZE_HEAD; + memcpy (f->data+rec*(size-VBLK_SIZE_HEAD)+VBLK_SIZE_HEAD, data, size); return true; diff --git a/fs/proc/Makefile b/fs/proc/Makefile index df434c5f28fb..c1c729335924 100644 --- a/fs/proc/Makefile +++ b/fs/proc/Makefile @@ -20,6 +20,7 @@ proc-y += stat.o proc-y += uptime.o proc-y += version.o proc-y += softirqs.o +proc-y += namespaces.o proc-$(CONFIG_PROC_SYSCTL) += proc_sysctl.o proc-$(CONFIG_NET) += proc_net.o proc-$(CONFIG_PROC_KCORE) += kcore.o diff --git a/fs/proc/array.c b/fs/proc/array.c index 5e4f776b0917..9b45ee84fbcc 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -131,7 +131,7 @@ static inline void task_name(struct seq_file *m, struct task_struct *p) * you can test for combinations of others with * simple bit tests. */ -static const char *task_state_array[] = { +static const char * const task_state_array[] = { "R (running)", /* 0 */ "S (sleeping)", /* 1 */ "D (disk sleep)", /* 2 */ @@ -147,7 +147,7 @@ static const char *task_state_array[] = { static inline const char *get_task_state(struct task_struct *tsk) { unsigned int state = (tsk->state & TASK_REPORT) | tsk->exit_state; - const char **p = &task_state_array[0]; + const char * const *p = &task_state_array[0]; BUILD_BUG_ON(1 + ilog2(TASK_STATE_MAX) != ARRAY_SIZE(task_state_array)); diff --git a/fs/proc/base.c b/fs/proc/base.c index 5a670c11aeac..4ede550517a6 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -220,7 +220,7 @@ static struct mm_struct *__check_mem_permission(struct task_struct *task) } /* - * Noone else is allowed. + * No one else is allowed. */ mmput(mm); return ERR_PTR(-EPERM); @@ -600,7 +600,7 @@ static int proc_fd_access_allowed(struct inode *inode) return allowed; } -static int proc_setattr(struct dentry *dentry, struct iattr *attr) +int proc_setattr(struct dentry *dentry, struct iattr *attr) { int error; struct inode *inode = dentry->d_inode; @@ -894,20 +894,20 @@ static ssize_t mem_write(struct file * file, const char __user *buf, if (!task) goto out_no_task; + copied = -ENOMEM; + page = (char *)__get_free_page(GFP_TEMPORARY); + if (!page) + goto out_task; + mm = check_mem_permission(task); copied = PTR_ERR(mm); if (IS_ERR(mm)) - goto out_task; + goto out_free; copied = -EIO; if (file->private_data != (void *)((long)current->self_exec_id)) goto out_mm; - copied = -ENOMEM; - page = (char *)__get_free_page(GFP_TEMPORARY); - if (!page) - goto out_mm; - copied = 0; while (count > 0) { int this_len, retval; @@ -929,9 +929,11 @@ static ssize_t mem_write(struct file * file, const char __user *buf, count -= retval; } *ppos = dst; - free_page((unsigned long) page); + out_mm: mmput(mm); +out_free: + free_page((unsigned long) page); out_task: put_task_struct(task); out_no_task: @@ -1059,7 +1061,7 @@ static ssize_t oom_adjust_write(struct file *file, const char __user *buf, { struct task_struct *task; char buffer[PROC_NUMBUF]; - long oom_adjust; + int oom_adjust; unsigned long flags; int err; @@ -1071,7 +1073,7 @@ static ssize_t oom_adjust_write(struct file *file, const char __user *buf, goto out; } - err = strict_strtol(strstrip(buffer), 0, &oom_adjust); + err = kstrtoint(strstrip(buffer), 0, &oom_adjust); if (err) goto out; if ((oom_adjust < OOM_ADJUST_MIN || oom_adjust > OOM_ADJUST_MAX) && @@ -1168,7 +1170,7 @@ static ssize_t oom_score_adj_write(struct file *file, const char __user *buf, struct task_struct *task; char buffer[PROC_NUMBUF]; unsigned long flags; - long oom_score_adj; + int oom_score_adj; int err; memset(buffer, 0, sizeof(buffer)); @@ -1179,7 +1181,7 @@ static ssize_t oom_score_adj_write(struct file *file, const char __user *buf, goto out; } - err = strict_strtol(strstrip(buffer), 0, &oom_score_adj); + err = kstrtoint(strstrip(buffer), 0, &oom_score_adj); if (err) goto out; if (oom_score_adj < OOM_SCORE_ADJ_MIN || @@ -1468,7 +1470,7 @@ sched_autogroup_write(struct file *file, const char __user *buf, struct inode *inode = file->f_path.dentry->d_inode; struct task_struct *p; char buffer[PROC_NUMBUF]; - long nice; + int nice; int err; memset(buffer, 0, sizeof(buffer)); @@ -1477,9 +1479,9 @@ sched_autogroup_write(struct file *file, const char __user *buf, if (copy_from_user(buffer, buf, count)) return -EFAULT; - err = strict_strtol(strstrip(buffer), 0, &nice); - if (err) - return -EINVAL; + err = kstrtoint(strstrip(buffer), 0, &nice); + if (err < 0) + return err; p = get_proc_task(inode); if (!p) @@ -1576,57 +1578,6 @@ static const struct file_operations proc_pid_set_comm_operations = { .release = single_release, }; -/* - * We added or removed a vma mapping the executable. The vmas are only mapped - * during exec and are not mapped with the mmap system call. - * Callers must hold down_write() on the mm's mmap_sem for these - */ -void added_exe_file_vma(struct mm_struct *mm) -{ - mm->num_exe_file_vmas++; -} - -void removed_exe_file_vma(struct mm_struct *mm) -{ - mm->num_exe_file_vmas--; - if ((mm->num_exe_file_vmas == 0) && mm->exe_file){ - fput(mm->exe_file); - mm->exe_file = NULL; - } - -} - -void set_mm_exe_file(struct mm_struct *mm, struct file *new_exe_file) -{ - if (new_exe_file) - get_file(new_exe_file); - if (mm->exe_file) - fput(mm->exe_file); - mm->exe_file = new_exe_file; - mm->num_exe_file_vmas = 0; -} - -struct file *get_mm_exe_file(struct mm_struct *mm) -{ - struct file *exe_file; - - /* We need mmap_sem to protect against races with removal of - * VM_EXECUTABLE vmas */ - down_read(&mm->mmap_sem); - exe_file = mm->exe_file; - if (exe_file) - get_file(exe_file); - up_read(&mm->mmap_sem); - return exe_file; -} - -void dup_mm_exe_file(struct mm_struct *oldmm, struct mm_struct *newmm) -{ - /* It's safe to write the exe_file pointer without exe_file_lock because - * this is called during fork when the task is not yet in /proc */ - newmm->exe_file = get_mm_exe_file(oldmm); -} - static int proc_exe_link(struct inode *inode, struct path *exe_path) { struct task_struct *task; @@ -1736,8 +1687,7 @@ static int task_dumpable(struct task_struct *task) return 0; } - -static struct inode *proc_pid_make_inode(struct super_block * sb, struct task_struct *task) +struct inode *proc_pid_make_inode(struct super_block * sb, struct task_struct *task) { struct inode * inode; struct proc_inode *ei; @@ -1779,7 +1729,7 @@ out_unlock: return NULL; } -static int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) +int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat) { struct inode *inode = dentry->d_inode; struct task_struct *task; @@ -1820,7 +1770,7 @@ static int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat * made this apply to all per process world readable and executable * directories. */ -static int pid_revalidate(struct dentry *dentry, struct nameidata *nd) +int pid_revalidate(struct dentry *dentry, struct nameidata *nd) { struct inode *inode; struct task_struct *task; @@ -1862,7 +1812,7 @@ static int pid_delete_dentry(const struct dentry * dentry) return !proc_pid(dentry->d_inode)->tasks[PIDTYPE_PID].first; } -static const struct dentry_operations pid_dentry_operations = +const struct dentry_operations pid_dentry_operations = { .d_revalidate = pid_revalidate, .d_delete = pid_delete_dentry, @@ -1870,9 +1820,6 @@ static const struct dentry_operations pid_dentry_operations = /* Lookups */ -typedef struct dentry *instantiate_t(struct inode *, struct dentry *, - struct task_struct *, const void *); - /* * Fill a directory entry. * @@ -1885,8 +1832,8 @@ typedef struct dentry *instantiate_t(struct inode *, struct dentry *, * reported by readdir in sync with the inode numbers reported * by stat. */ -static int proc_fill_cache(struct file *filp, void *dirent, filldir_t filldir, - char *name, int len, +int proc_fill_cache(struct file *filp, void *dirent, filldir_t filldir, + const char *name, int len, instantiate_t instantiate, struct task_struct *task, const void *ptr) { struct dentry *child, *dir = filp->f_path.dentry; @@ -2820,6 +2767,7 @@ static const struct pid_entry tgid_base_stuff[] = { DIR("task", S_IRUGO|S_IXUGO, proc_task_inode_operations, proc_task_operations), DIR("fd", S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations), DIR("fdinfo", S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations), + DIR("ns", S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations), #ifdef CONFIG_NET DIR("net", S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations), #endif @@ -3124,11 +3072,16 @@ static int proc_pid_fill_cache(struct file *filp, void *dirent, filldir_t filldi /* for the /proc/ directory itself, after non-process stuff has been done */ int proc_pid_readdir(struct file * filp, void * dirent, filldir_t filldir) { - unsigned int nr = filp->f_pos - FIRST_PROCESS_ENTRY; - struct task_struct *reaper = get_proc_task(filp->f_path.dentry->d_inode); + unsigned int nr; + struct task_struct *reaper; struct tgid_iter iter; struct pid_namespace *ns; + if (filp->f_pos >= PID_MAX_LIMIT + TGID_OFFSET) + goto out_no_task; + nr = filp->f_pos - FIRST_PROCESS_ENTRY; + + reaper = get_proc_task(filp->f_path.dentry->d_inode); if (!reaper) goto out_no_task; @@ -3163,6 +3116,7 @@ out_no_task: static const struct pid_entry tid_base_stuff[] = { DIR("fd", S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations), DIR("fdinfo", S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations), + DIR("ns", S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations), REG("environ", S_IRUSR, proc_environ_operations), INF("auxv", S_IRUSR, proc_pid_auxv), ONE("status", S_IRUGO, proc_pid_status), diff --git a/fs/proc/generic.c b/fs/proc/generic.c index f1281339b6fa..f1637f17c37c 100644 --- a/fs/proc/generic.c +++ b/fs/proc/generic.c @@ -674,6 +674,7 @@ struct proc_dir_entry *proc_mkdir_mode(const char *name, mode_t mode, } return ent; } +EXPORT_SYMBOL(proc_mkdir_mode); struct proc_dir_entry *proc_net_mkdir(struct net *net, const char *name, struct proc_dir_entry *parent) diff --git a/fs/proc/inode.c b/fs/proc/inode.c index d15aa1b1cc8f..74b48cfa1bb2 100644 --- a/fs/proc/inode.c +++ b/fs/proc/inode.c @@ -28,6 +28,7 @@ static void proc_evict_inode(struct inode *inode) { struct proc_dir_entry *de; struct ctl_table_header *head; + const struct proc_ns_operations *ns_ops; truncate_inode_pages(&inode->i_data, 0); end_writeback(inode); @@ -44,6 +45,10 @@ static void proc_evict_inode(struct inode *inode) rcu_assign_pointer(PROC_I(inode)->sysctl, NULL); sysctl_head_put(head); } + /* Release any associated namespace */ + ns_ops = PROC_I(inode)->ns_ops; + if (ns_ops && ns_ops->put) + ns_ops->put(PROC_I(inode)->ns); } static struct kmem_cache * proc_inode_cachep; @@ -62,6 +67,8 @@ static struct inode *proc_alloc_inode(struct super_block *sb) ei->pde = NULL; ei->sysctl = NULL; ei->sysctl_entry = NULL; + ei->ns = NULL; + ei->ns_ops = NULL; inode = &ei->vfs_inode; inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; return inode; diff --git a/fs/proc/internal.h b/fs/proc/internal.h index c03e8d3a3a5b..7838e5cfec14 100644 --- a/fs/proc/internal.h +++ b/fs/proc/internal.h @@ -61,6 +61,14 @@ extern const struct file_operations proc_pagemap_operations; extern const struct file_operations proc_net_operations; extern const struct inode_operations proc_net_inode_operations; +struct proc_maps_private { + struct pid *pid; + struct task_struct *task; +#ifdef CONFIG_MMU + struct vm_area_struct *tail_vma; +#endif +}; + void proc_init_inodecache(void); static inline struct pid *proc_pid(struct inode *inode) @@ -119,3 +127,21 @@ struct inode *proc_get_inode(struct super_block *, struct proc_dir_entry *); */ int proc_readdir(struct file *, void *, filldir_t); struct dentry *proc_lookup(struct inode *, struct dentry *, struct nameidata *); + + + +/* Lookups */ +typedef struct dentry *instantiate_t(struct inode *, struct dentry *, + struct task_struct *, const void *); +int proc_fill_cache(struct file *filp, void *dirent, filldir_t filldir, + const char *name, int len, + instantiate_t instantiate, struct task_struct *task, const void *ptr); +int pid_revalidate(struct dentry *dentry, struct nameidata *nd); +struct inode *proc_pid_make_inode(struct super_block * sb, struct task_struct *task); +extern const struct dentry_operations pid_dentry_operations; +int pid_getattr(struct vfsmount *mnt, struct dentry *dentry, struct kstat *stat); +int proc_setattr(struct dentry *dentry, struct iattr *attr); + +extern const struct inode_operations proc_ns_dir_inode_operations; +extern const struct file_operations proc_ns_dir_operations; + diff --git a/fs/proc/namespaces.c b/fs/proc/namespaces.c new file mode 100644 index 000000000000..781dec5bd682 --- /dev/null +++ b/fs/proc/namespaces.c @@ -0,0 +1,198 @@ +#include <linux/proc_fs.h> +#include <linux/nsproxy.h> +#include <linux/sched.h> +#include <linux/ptrace.h> +#include <linux/fs_struct.h> +#include <linux/mount.h> +#include <linux/path.h> +#include <linux/namei.h> +#include <linux/file.h> +#include <linux/utsname.h> +#include <net/net_namespace.h> +#include <linux/mnt_namespace.h> +#include <linux/ipc_namespace.h> +#include <linux/pid_namespace.h> +#include "internal.h" + + +static const struct proc_ns_operations *ns_entries[] = { +#ifdef CONFIG_NET_NS + &netns_operations, +#endif +#ifdef CONFIG_UTS_NS + &utsns_operations, +#endif +#ifdef CONFIG_IPC_NS + &ipcns_operations, +#endif +}; + +static const struct file_operations ns_file_operations = { + .llseek = no_llseek, +}; + +static struct dentry *proc_ns_instantiate(struct inode *dir, + struct dentry *dentry, struct task_struct *task, const void *ptr) +{ + const struct proc_ns_operations *ns_ops = ptr; + struct inode *inode; + struct proc_inode *ei; + struct dentry *error = ERR_PTR(-ENOENT); + + inode = proc_pid_make_inode(dir->i_sb, task); + if (!inode) + goto out; + + ei = PROC_I(inode); + inode->i_mode = S_IFREG|S_IRUSR; + inode->i_fop = &ns_file_operations; + ei->ns_ops = ns_ops; + ei->ns = ns_ops->get(task); + if (!ei->ns) + goto out_iput; + + dentry->d_op = &pid_dentry_operations; + d_add(dentry, inode); + /* Close the race of the process dying before we return the dentry */ + if (pid_revalidate(dentry, NULL)) + error = NULL; +out: + return error; +out_iput: + iput(inode); + goto out; +} + +static int proc_ns_fill_cache(struct file *filp, void *dirent, + filldir_t filldir, struct task_struct *task, + const struct proc_ns_operations *ops) +{ + return proc_fill_cache(filp, dirent, filldir, + ops->name, strlen(ops->name), + proc_ns_instantiate, task, ops); +} + +static int proc_ns_dir_readdir(struct file *filp, void *dirent, + filldir_t filldir) +{ + int i; + struct dentry *dentry = filp->f_path.dentry; + struct inode *inode = dentry->d_inode; + struct task_struct *task = get_proc_task(inode); + const struct proc_ns_operations **entry, **last; + ino_t ino; + int ret; + + ret = -ENOENT; + if (!task) + goto out_no_task; + + ret = -EPERM; + if (!ptrace_may_access(task, PTRACE_MODE_READ)) + goto out; + + ret = 0; + i = filp->f_pos; + switch (i) { + case 0: + ino = inode->i_ino; + if (filldir(dirent, ".", 1, i, ino, DT_DIR) < 0) + goto out; + i++; + filp->f_pos++; + /* fall through */ + case 1: + ino = parent_ino(dentry); + if (filldir(dirent, "..", 2, i, ino, DT_DIR) < 0) + goto out; + i++; + filp->f_pos++; + /* fall through */ + default: + i -= 2; + if (i >= ARRAY_SIZE(ns_entries)) { + ret = 1; + goto out; + } + entry = ns_entries + i; + last = &ns_entries[ARRAY_SIZE(ns_entries) - 1]; + while (entry <= last) { + if (proc_ns_fill_cache(filp, dirent, filldir, + task, *entry) < 0) + goto out; + filp->f_pos++; + entry++; + } + } + + ret = 1; +out: + put_task_struct(task); +out_no_task: + return ret; +} + +const struct file_operations proc_ns_dir_operations = { + .read = generic_read_dir, + .readdir = proc_ns_dir_readdir, +}; + +static struct dentry *proc_ns_dir_lookup(struct inode *dir, + struct dentry *dentry, struct nameidata *nd) +{ + struct dentry *error; + struct task_struct *task = get_proc_task(dir); + const struct proc_ns_operations **entry, **last; + unsigned int len = dentry->d_name.len; + + error = ERR_PTR(-ENOENT); + + if (!task) + goto out_no_task; + + error = ERR_PTR(-EPERM); + if (!ptrace_may_access(task, PTRACE_MODE_READ)) + goto out; + + last = &ns_entries[ARRAY_SIZE(ns_entries) - 1]; + for (entry = ns_entries; entry <= last; entry++) { + if (strlen((*entry)->name) != len) + continue; + if (!memcmp(dentry->d_name.name, (*entry)->name, len)) + break; + } + error = ERR_PTR(-ENOENT); + if (entry > last) + goto out; + + error = proc_ns_instantiate(dir, dentry, task, *entry); +out: + put_task_struct(task); +out_no_task: + return error; +} + +const struct inode_operations proc_ns_dir_inode_operations = { + .lookup = proc_ns_dir_lookup, + .getattr = pid_getattr, + .setattr = proc_setattr, +}; + +struct file *proc_ns_fget(int fd) +{ + struct file *file; + + file = fget(fd); + if (!file) + return ERR_PTR(-EBADF); + + if (file->f_op != &ns_file_operations) + goto out_invalid; + + return file; + +out_invalid: + fput(file); + return ERR_PTR(-EINVAL); +} + diff --git a/fs/proc/stat.c b/fs/proc/stat.c index 1cffa2b8a2fc..9758b654a1bc 100644 --- a/fs/proc/stat.c +++ b/fs/proc/stat.c @@ -138,9 +138,9 @@ static int stat_open(struct inode *inode, struct file *file) struct seq_file *m; int res; - /* don't ask for more than the kmalloc() max size, currently 128 KB */ - if (size > 128 * 1024) - size = 128 * 1024; + /* don't ask for more than the kmalloc() max size */ + if (size > KMALLOC_MAX_SIZE) + size = KMALLOC_MAX_SIZE; buf = kmalloc(size, GFP_KERNEL); if (!buf) return -ENOMEM; diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 2e7addfd9803..25b6a887adb9 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -211,10 +211,10 @@ static void show_map_vma(struct seq_file *m, struct vm_area_struct *vma) { struct mm_struct *mm = vma->vm_mm; struct file *file = vma->vm_file; - int flags = vma->vm_flags; + vm_flags_t flags = vma->vm_flags; unsigned long ino = 0; unsigned long long pgoff = 0; - unsigned long start; + unsigned long start, end; dev_t dev = 0; int len; @@ -227,13 +227,15 @@ static void show_map_vma(struct seq_file *m, struct vm_area_struct *vma) /* We don't show the stack guard page in /proc/maps */ start = vma->vm_start; - if (vma->vm_flags & VM_GROWSDOWN) - if (!vma_stack_continue(vma->vm_prev, vma->vm_start)) - start += PAGE_SIZE; + if (stack_guard_page_start(vma, start)) + start += PAGE_SIZE; + end = vma->vm_end; + if (stack_guard_page_end(vma, end)) + end -= PAGE_SIZE; seq_printf(m, "%08lx-%08lx %c%c%c%c %08llx %02x:%02x %lu %n", start, - vma->vm_end, + end, flags & VM_READ ? 'r' : '-', flags & VM_WRITE ? 'w' : '-', flags & VM_EXEC ? 'x' : '-', @@ -534,15 +536,17 @@ static ssize_t clear_refs_write(struct file *file, const char __user *buf, char buffer[PROC_NUMBUF]; struct mm_struct *mm; struct vm_area_struct *vma; - long type; + int type; + int rv; memset(buffer, 0, sizeof(buffer)); if (count > sizeof(buffer) - 1) count = sizeof(buffer) - 1; if (copy_from_user(buffer, buf, count)) return -EFAULT; - if (strict_strtol(strstrip(buffer), 10, &type)) - return -EINVAL; + rv = kstrtoint(strstrip(buffer), 10, &type); + if (rv < 0) + return rv; if (type < CLEAR_REFS_ALL || type > CLEAR_REFS_MAPPED) return -EINVAL; task = get_proc_task(file->f_path.dentry->d_inode); @@ -767,18 +771,12 @@ static ssize_t pagemap_read(struct file *file, char __user *buf, if (!task) goto out; - mm = mm_for_maps(task); - ret = PTR_ERR(mm); - if (!mm || IS_ERR(mm)) - goto out_task; - ret = -EINVAL; /* file position must be aligned */ if ((*ppos % PM_ENTRY_BYTES) || (count % PM_ENTRY_BYTES)) goto out_task; ret = 0; - if (!count) goto out_task; @@ -786,7 +784,12 @@ static ssize_t pagemap_read(struct file *file, char __user *buf, pm.buffer = kmalloc(pm.len, GFP_TEMPORARY); ret = -ENOMEM; if (!pm.buffer) - goto out_mm; + goto out_task; + + mm = mm_for_maps(task); + ret = PTR_ERR(mm); + if (!mm || IS_ERR(mm)) + goto out_free; pagemap_walk.pmd_entry = pagemap_pte_range; pagemap_walk.pte_hole = pagemap_pte_hole; @@ -829,7 +832,7 @@ static ssize_t pagemap_read(struct file *file, char __user *buf, len = min(count, PM_ENTRY_BYTES * pm.pos); if (copy_to_user(buf, pm.buffer, len)) { ret = -EFAULT; - goto out_free; + goto out_mm; } copied += len; buf += len; @@ -839,10 +842,10 @@ static ssize_t pagemap_read(struct file *file, char __user *buf, if (!ret || ret == PM_END_OF_BUFFER) ret = copied; -out_free: - kfree(pm.buffer); out_mm: mmput(mm); +out_free: + kfree(pm.buffer); out_task: put_task_struct(task); out: @@ -856,7 +859,192 @@ const struct file_operations proc_pagemap_operations = { #endif /* CONFIG_PROC_PAGE_MONITOR */ #ifdef CONFIG_NUMA -extern int show_numa_map(struct seq_file *m, void *v); + +struct numa_maps { + struct vm_area_struct *vma; + unsigned long pages; + unsigned long anon; + unsigned long active; + unsigned long writeback; + unsigned long mapcount_max; + unsigned long dirty; + unsigned long swapcache; + unsigned long node[MAX_NUMNODES]; +}; + +struct numa_maps_private { + struct proc_maps_private proc_maps; + struct numa_maps md; +}; + +static void gather_stats(struct page *page, struct numa_maps *md, int pte_dirty) +{ + int count = page_mapcount(page); + + md->pages++; + if (pte_dirty || PageDirty(page)) + md->dirty++; + + if (PageSwapCache(page)) + md->swapcache++; + + if (PageActive(page) || PageUnevictable(page)) + md->active++; + + if (PageWriteback(page)) + md->writeback++; + + if (PageAnon(page)) + md->anon++; + + if (count > md->mapcount_max) + md->mapcount_max = count; + + md->node[page_to_nid(page)]++; +} + +static int gather_pte_stats(pmd_t *pmd, unsigned long addr, + unsigned long end, struct mm_walk *walk) +{ + struct numa_maps *md; + spinlock_t *ptl; + pte_t *orig_pte; + pte_t *pte; + + md = walk->private; + orig_pte = pte = pte_offset_map_lock(walk->mm, pmd, addr, &ptl); + do { + struct page *page; + int nid; + + if (!pte_present(*pte)) + continue; + + page = vm_normal_page(md->vma, addr, *pte); + if (!page) + continue; + + if (PageReserved(page)) + continue; + + nid = page_to_nid(page); + if (!node_isset(nid, node_states[N_HIGH_MEMORY])) + continue; + + gather_stats(page, md, pte_dirty(*pte)); + + } while (pte++, addr += PAGE_SIZE, addr != end); + pte_unmap_unlock(orig_pte, ptl); + return 0; +} +#ifdef CONFIG_HUGETLB_PAGE +static int gather_hugetbl_stats(pte_t *pte, unsigned long hmask, + unsigned long addr, unsigned long end, struct mm_walk *walk) +{ + struct numa_maps *md; + struct page *page; + + if (pte_none(*pte)) + return 0; + + page = pte_page(*pte); + if (!page) + return 0; + + md = walk->private; + gather_stats(page, md, pte_dirty(*pte)); + return 0; +} + +#else +static int gather_hugetbl_stats(pte_t *pte, unsigned long hmask, + unsigned long addr, unsigned long end, struct mm_walk *walk) +{ + return 0; +} +#endif + +/* + * Display pages allocated per node and memory policy via /proc. + */ +static int show_numa_map(struct seq_file *m, void *v) +{ + struct numa_maps_private *numa_priv = m->private; + struct proc_maps_private *proc_priv = &numa_priv->proc_maps; + struct vm_area_struct *vma = v; + struct numa_maps *md = &numa_priv->md; + struct file *file = vma->vm_file; + struct mm_struct *mm = vma->vm_mm; + struct mm_walk walk = {}; + struct mempolicy *pol; + int n; + char buffer[50]; + + if (!mm) + return 0; + + /* Ensure we start with an empty set of numa_maps statistics. */ + memset(md, 0, sizeof(*md)); + + md->vma = vma; + + walk.hugetlb_entry = gather_hugetbl_stats; + walk.pmd_entry = gather_pte_stats; + walk.private = md; + walk.mm = mm; + + pol = get_vma_policy(proc_priv->task, vma, vma->vm_start); + mpol_to_str(buffer, sizeof(buffer), pol, 0); + mpol_cond_put(pol); + + seq_printf(m, "%08lx %s", vma->vm_start, buffer); + + if (file) { + seq_printf(m, " file="); + seq_path(m, &file->f_path, "\n\t= "); + } else if (vma->vm_start <= mm->brk && vma->vm_end >= mm->start_brk) { + seq_printf(m, " heap"); + } else if (vma->vm_start <= mm->start_stack && + vma->vm_end >= mm->start_stack) { + seq_printf(m, " stack"); + } + + walk_page_range(vma->vm_start, vma->vm_end, &walk); + + if (!md->pages) + goto out; + + if (md->anon) + seq_printf(m, " anon=%lu", md->anon); + + if (md->dirty) + seq_printf(m, " dirty=%lu", md->dirty); + + if (md->pages != md->anon && md->pages != md->dirty) + seq_printf(m, " mapped=%lu", md->pages); + + if (md->mapcount_max > 1) + seq_printf(m, " mapmax=%lu", md->mapcount_max); + + if (md->swapcache) + seq_printf(m, " swapcache=%lu", md->swapcache); + + if (md->active < md->pages && !is_vm_hugetlb_page(vma)) + seq_printf(m, " active=%lu", md->active); + + if (md->writeback) + seq_printf(m, " writeback=%lu", md->writeback); + + for_each_node_state(n, N_HIGH_MEMORY) + if (md->node[n]) + seq_printf(m, " N%d=%lu", n, md->node[n]); +out: + seq_putc(m, '\n'); + + if (m->count < m->size) + m->version = (vma != proc_priv->tail_vma) ? vma->vm_start : 0; + return 0; +} static const struct seq_operations proc_pid_numa_maps_op = { .start = m_start, @@ -867,7 +1055,20 @@ static const struct seq_operations proc_pid_numa_maps_op = { static int numa_maps_open(struct inode *inode, struct file *file) { - return do_maps_open(inode, file, &proc_pid_numa_maps_op); + struct numa_maps_private *priv; + int ret = -ENOMEM; + priv = kzalloc(sizeof(*priv), GFP_KERNEL); + if (priv) { + priv->proc_maps.pid = proc_pid(inode); + ret = seq_open(file, &proc_pid_numa_maps_op); + if (!ret) { + struct seq_file *m = file->private_data; + m->private = priv; + } else { + kfree(priv); + } + } + return ret; } const struct file_operations proc_numa_maps_operations = { @@ -876,4 +1077,4 @@ const struct file_operations proc_numa_maps_operations = { .llseek = seq_lseek, .release = seq_release_private, }; -#endif +#endif /* CONFIG_NUMA */ diff --git a/fs/proc/vmcore.c b/fs/proc/vmcore.c index 74802bc5ded9..cd99bf557650 100644 --- a/fs/proc/vmcore.c +++ b/fs/proc/vmcore.c @@ -35,6 +35,46 @@ static u64 vmcore_size; static struct proc_dir_entry *proc_vmcore = NULL; +/* + * Returns > 0 for RAM pages, 0 for non-RAM pages, < 0 on error + * The called function has to take care of module refcounting. + */ +static int (*oldmem_pfn_is_ram)(unsigned long pfn); + +int register_oldmem_pfn_is_ram(int (*fn)(unsigned long pfn)) +{ + if (oldmem_pfn_is_ram) + return -EBUSY; + oldmem_pfn_is_ram = fn; + return 0; +} +EXPORT_SYMBOL_GPL(register_oldmem_pfn_is_ram); + +void unregister_oldmem_pfn_is_ram(void) +{ + oldmem_pfn_is_ram = NULL; + wmb(); +} +EXPORT_SYMBOL_GPL(unregister_oldmem_pfn_is_ram); + +static int pfn_is_ram(unsigned long pfn) +{ + int (*fn)(unsigned long pfn); + /* pfn is ram unless fn() checks pagetype */ + int ret = 1; + + /* + * Ask hypervisor if the pfn is really ram. + * A ballooned page contains no data and reading from such a page + * will cause high load in the hypervisor. + */ + fn = oldmem_pfn_is_ram; + if (fn) + ret = fn(pfn); + + return ret; +} + /* Reads a page from the oldmem device from given offset. */ static ssize_t read_from_oldmem(char *buf, size_t count, u64 *ppos, int userbuf) @@ -55,9 +95,15 @@ static ssize_t read_from_oldmem(char *buf, size_t count, else nr_bytes = count; - tmp = copy_oldmem_page(pfn, buf, nr_bytes, offset, userbuf); - if (tmp < 0) - return tmp; + /* If pfn is not ram, return zeros for sparse dump files */ + if (pfn_is_ram(pfn) == 0) + memset(buf, 0, nr_bytes); + else { + tmp = copy_oldmem_page(pfn, buf, nr_bytes, + offset, userbuf); + if (tmp < 0) + return tmp; + } *ppos += nr_bytes; count -= nr_bytes; buf += nr_bytes; diff --git a/fs/pstore/Kconfig b/fs/pstore/Kconfig index 867d0ac026ce..8007ae7c0d8c 100644 --- a/fs/pstore/Kconfig +++ b/fs/pstore/Kconfig @@ -1,5 +1,5 @@ config PSTORE - bool "Persistant store support" + bool "Persistent store support" default n help This option enables generic access to platform level diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c index f835a25625ff..f2c3ff20ea68 100644 --- a/fs/pstore/platform.c +++ b/fs/pstore/platform.c @@ -152,21 +152,27 @@ EXPORT_SYMBOL_GPL(pstore_register); void pstore_get_records(void) { struct pstore_info *psi = psinfo; - size_t size; + ssize_t size; u64 id; enum pstore_type_id type; struct timespec time; - int failed = 0; + int failed = 0, rc; if (!psi) return; mutex_lock(&psinfo->buf_mutex); + rc = psi->open(psi); + if (rc) + goto out; + while ((size = psi->read(&id, &type, &time)) > 0) { - if (pstore_mkfile(type, psi->name, id, psi->buf, size, + if (pstore_mkfile(type, psi->name, id, psi->buf, (size_t)size, time, psi->erase)) failed++; } + psi->close(psi); +out: mutex_unlock(&psinfo->buf_mutex); if (failed) diff --git a/fs/quota/dquot.c b/fs/quota/dquot.c index fcc8ae75d874..5b572c89e6c4 100644 --- a/fs/quota/dquot.c +++ b/fs/quota/dquot.c @@ -442,7 +442,7 @@ EXPORT_SYMBOL(dquot_acquire); */ int dquot_commit(struct dquot *dquot) { - int ret = 0, ret2 = 0; + int ret = 0; struct quota_info *dqopt = sb_dqopt(dquot->dq_sb); mutex_lock(&dqopt->dqio_mutex); @@ -454,15 +454,10 @@ int dquot_commit(struct dquot *dquot) spin_unlock(&dq_list_lock); /* Inactive dquot can be only if there was error during read/init * => we have better not writing it */ - if (test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) { + if (test_bit(DQ_ACTIVE_B, &dquot->dq_flags)) ret = dqopt->ops[dquot->dq_type]->commit_dqblk(dquot); - if (info_dirty(&dqopt->info[dquot->dq_type])) { - ret2 = dqopt->ops[dquot->dq_type]->write_file_info( - dquot->dq_sb, dquot->dq_type); - } - if (ret >= 0) - ret = ret2; - } + else + ret = -EIO; out_sem: mutex_unlock(&dqopt->dqio_mutex); return ret; @@ -696,8 +691,11 @@ static void prune_dqcache(int count) * This is called from kswapd when we think we need some * more memory */ -static int shrink_dqcache_memory(struct shrinker *shrink, int nr, gfp_t gfp_mask) +static int shrink_dqcache_memory(struct shrinker *shrink, + struct shrink_control *sc) { + int nr = sc->nr_to_scan; + if (nr) { spin_lock(&dq_list_lock); prune_dqcache(nr); @@ -956,7 +954,7 @@ static inline int dqput_blocks(struct dquot *dquot) /* * Remove references to dquots from inode and add dquot to list for freeing - * if we have the last referece to dquot + * if we have the last reference to dquot * We can't race with anybody because we hold dqptr_sem for writing... */ static int remove_inode_dquot_ref(struct inode *inode, int type, diff --git a/fs/ramfs/file-nommu.c b/fs/ramfs/file-nommu.c index 9eead2c796b7..fbb0b478a346 100644 --- a/fs/ramfs/file-nommu.c +++ b/fs/ramfs/file-nommu.c @@ -112,6 +112,7 @@ int ramfs_nommu_expand_for_mapping(struct inode *inode, size_t newsize) SetPageDirty(page); unlock_page(page); + put_page(page); } return 0; diff --git a/fs/reiserfs/journal.c b/fs/reiserfs/journal.c index c77514bd5776..c5e82ece7c6c 100644 --- a/fs/reiserfs/journal.c +++ b/fs/reiserfs/journal.c @@ -1,7 +1,7 @@ /* ** Write ahead logging implementation copyright Chris Mason 2000 ** -** The background commits make this code very interelated, and +** The background commits make this code very interrelated, and ** overly complex. I need to rethink things a bit....The major players: ** ** journal_begin -- call with the number of blocks you expect to log. @@ -2725,7 +2725,7 @@ int journal_init(struct super_block *sb, const char *j_dev_name, REISERFS_DISK_OFFSET_IN_BYTES / sb->s_blocksize + 2); - /* Sanity check to see is the standard journal fitting withing first bitmap + /* Sanity check to see is the standard journal fitting within first bitmap (actual for small blocksizes) */ if (!SB_ONDISK_JOURNAL_DEVICE(sb) && (SB_JOURNAL_1st_RESERVED_BLOCK(sb) + diff --git a/fs/reiserfs/lock.c b/fs/reiserfs/lock.c index b87aa2c1afc1..7df1ce48203a 100644 --- a/fs/reiserfs/lock.c +++ b/fs/reiserfs/lock.c @@ -15,7 +15,7 @@ * for this mutex, no need for a system wide mutex facility. * * Also this lock is often released before a call that could block because - * reiserfs performances were partialy based on the release while schedule() + * reiserfs performances were partially based on the release while schedule() * property of the Bkl. */ void reiserfs_write_lock(struct super_block *s) diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index 0aab04f46827..aa91089162cb 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -393,7 +393,7 @@ void add_save_link(struct reiserfs_transaction_handle *th, /* body of "save" link */ link = INODE_PKEY(inode)->k_dir_id; - /* put "save" link inot tree, don't charge quota to anyone */ + /* put "save" link into tree, don't charge quota to anyone */ retval = reiserfs_insert_item(th, &path, &key, &ih, NULL, (char *)&link); if (retval) { @@ -568,7 +568,7 @@ static void destroy_inodecache(void) } /* we don't mark inodes dirty, we just log them */ -static void reiserfs_dirty_inode(struct inode *inode) +static void reiserfs_dirty_inode(struct inode *inode, int flags) { struct reiserfs_transaction_handle th; @@ -2104,7 +2104,7 @@ out: /* Read data from quotafile - avoid pagecache and such because we cannot afford * acquiring the locks... As quota files are never truncated and quota code - * itself serializes the operations (and noone else should touch the files) + * itself serializes the operations (and no one else should touch the files) * we don't have to be afraid of races */ static ssize_t reiserfs_quota_read(struct super_block *sb, int type, char *data, size_t len, loff_t off) diff --git a/fs/reiserfs/xattr.c b/fs/reiserfs/xattr.c index 5c11ca82b782..e8a62f41b458 100644 --- a/fs/reiserfs/xattr.c +++ b/fs/reiserfs/xattr.c @@ -98,14 +98,12 @@ static int xattr_rmdir(struct inode *dir, struct dentry *dentry) reiserfs_mutex_lock_nested_safe(&dentry->d_inode->i_mutex, I_MUTEX_CHILD, dir->i_sb); - dentry_unhash(dentry); error = dir->i_op->rmdir(dir, dentry); if (!error) dentry->d_inode->i_flags |= S_DEAD; mutex_unlock(&dentry->d_inode->i_mutex); if (!error) d_delete(dentry); - dput(dentry); return error; } @@ -396,7 +394,7 @@ static struct page *reiserfs_get_page(struct inode *dir, size_t n) struct address_space *mapping = dir->i_mapping; struct page *page; /* We can deadlock if we try to free dentries, - and an unlink/rmdir has just occured - GFP_NOFS avoids this */ + and an unlink/rmdir has just occurred - GFP_NOFS avoids this */ mapping_set_gfp_mask(mapping, GFP_NOFS); page = read_mapping_page(mapping, n >> PAGE_CACHE_SHIFT, NULL); if (!IS_ERR(page)) { diff --git a/fs/splice.c b/fs/splice.c index 50a5d978da16..aa866d309695 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -162,6 +162,14 @@ static const struct pipe_buf_operations user_page_pipe_buf_ops = { .get = generic_pipe_buf_get, }; +static void wakeup_pipe_readers(struct pipe_inode_info *pipe) +{ + smp_mb(); + if (waitqueue_active(&pipe->wait)) + wake_up_interruptible(&pipe->wait); + kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); +} + /** * splice_to_pipe - fill passed data into a pipe * @pipe: pipe to fill @@ -247,12 +255,8 @@ ssize_t splice_to_pipe(struct pipe_inode_info *pipe, pipe_unlock(pipe); - if (do_wakeup) { - smp_mb(); - if (waitqueue_active(&pipe->wait)) - wake_up_interruptible(&pipe->wait); - kill_fasync(&pipe->fasync_readers, SIGIO, POLL_IN); - } + if (do_wakeup) + wakeup_pipe_readers(pipe); while (page_nr < spd_pages) spd->spd_release(spd, page_nr++); @@ -1892,12 +1896,9 @@ retry: /* * If we put data in the output pipe, wakeup any potential readers. */ - if (ret > 0) { - smp_mb(); - if (waitqueue_active(&opipe->wait)) - wake_up_interruptible(&opipe->wait); - kill_fasync(&opipe->fasync_readers, SIGIO, POLL_IN); - } + if (ret > 0) + wakeup_pipe_readers(opipe); + if (input_wakeup) wakeup_pipe_writers(ipipe); @@ -1976,12 +1977,8 @@ static int link_pipe(struct pipe_inode_info *ipipe, /* * If we put data in the output pipe, wakeup any potential readers. */ - if (ret > 0) { - smp_mb(); - if (waitqueue_active(&opipe->wait)) - wake_up_interruptible(&opipe->wait); - kill_fasync(&opipe->fasync_readers, SIGIO, POLL_IN); - } + if (ret > 0) + wakeup_pipe_readers(opipe); return ret; } diff --git a/fs/squashfs/Kconfig b/fs/squashfs/Kconfig index efc309fa3035..7797218d0b30 100644 --- a/fs/squashfs/Kconfig +++ b/fs/squashfs/Kconfig @@ -42,7 +42,7 @@ config SQUASHFS_LZO select LZO_DECOMPRESS help Saying Y here includes support for reading Squashfs file systems - compressed with LZO compresssion. LZO compression is mainly + compressed with LZO compression. LZO compression is mainly aimed at embedded systems with slower CPUs where the overheads of zlib are too high. @@ -57,7 +57,7 @@ config SQUASHFS_XZ select XZ_DEC help Saying Y here includes support for reading Squashfs file systems - compressed with XZ compresssion. XZ gives better compression than + compressed with XZ compression. XZ gives better compression than the default zlib compression, at the expense of greater CPU and memory overhead. diff --git a/fs/squashfs/block.c b/fs/squashfs/block.c index 8ab48bc2fa7d..ed0eb2a921f4 100644 --- a/fs/squashfs/block.c +++ b/fs/squashfs/block.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/cache.c b/fs/squashfs/cache.c index 26b15ae34d6f..f744be98cd5a 100644 --- a/fs/squashfs/cache.c +++ b/fs/squashfs/cache.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -29,7 +29,7 @@ * plus functions layered ontop of the generic cache implementation to * access the metadata and fragment caches. * - * To avoid out of memory and fragmentation isssues with vmalloc the cache + * To avoid out of memory and fragmentation issues with vmalloc the cache * uses sequences of kmalloced PAGE_CACHE_SIZE buffers. * * It should be noted that the cache is not used for file datablocks, these @@ -104,7 +104,7 @@ struct squashfs_cache_entry *squashfs_cache_get(struct super_block *sb, entry = &cache->entry[i]; /* - * Initialise choosen cache entry, and fill it in from + * Initialise chosen cache entry, and fill it in from * disk. */ cache->unused--; @@ -286,7 +286,7 @@ cleanup: /* - * Copy upto length bytes from cache entry to buffer starting at offset bytes + * Copy up to length bytes from cache entry to buffer starting at offset bytes * into the cache entry. If there's not length bytes then copy the number of * bytes available. In all cases return the number of bytes copied. */ @@ -393,19 +393,36 @@ struct squashfs_cache_entry *squashfs_get_datablock(struct super_block *sb, /* * Read a filesystem table (uncompressed sequence of bytes) from disk */ -int squashfs_read_table(struct super_block *sb, void *buffer, u64 block, - int length) +void *squashfs_read_table(struct super_block *sb, u64 block, int length) { int pages = (length + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; int i, res; - void **data = kcalloc(pages, sizeof(void *), GFP_KERNEL); - if (data == NULL) - return -ENOMEM; + void *table, *buffer, **data; + + table = buffer = kmalloc(length, GFP_KERNEL); + if (table == NULL) + return ERR_PTR(-ENOMEM); + + data = kcalloc(pages, sizeof(void *), GFP_KERNEL); + if (data == NULL) { + res = -ENOMEM; + goto failed; + } for (i = 0; i < pages; i++, buffer += PAGE_CACHE_SIZE) data[i] = buffer; + res = squashfs_read_data(sb, data, block, length | SQUASHFS_COMPRESSED_BIT_BLOCK, NULL, length, pages); + kfree(data); - return res; + + if (res < 0) + goto failed; + + return table; + +failed: + kfree(table); + return ERR_PTR(res); } diff --git a/fs/squashfs/decompressor.c b/fs/squashfs/decompressor.c index e921bd213738..9f1b0bb96f13 100644 --- a/fs/squashfs/decompressor.c +++ b/fs/squashfs/decompressor.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/decompressor.h b/fs/squashfs/decompressor.h index 099745ad5691..8ba70cff09a6 100644 --- a/fs/squashfs/decompressor.h +++ b/fs/squashfs/decompressor.h @@ -4,7 +4,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/dir.c b/fs/squashfs/dir.c index 3f79cd1d0c19..9dfe2ce0fb70 100644 --- a/fs/squashfs/dir.c +++ b/fs/squashfs/dir.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/export.c b/fs/squashfs/export.c index 7f93d5a9ee05..5e1101ff276f 100644 --- a/fs/squashfs/export.c +++ b/fs/squashfs/export.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -121,30 +121,38 @@ static struct dentry *squashfs_get_parent(struct dentry *child) * Read uncompressed inode lookup table indexes off disk into memory */ __le64 *squashfs_read_inode_lookup_table(struct super_block *sb, - u64 lookup_table_start, unsigned int inodes) + u64 lookup_table_start, u64 next_table, unsigned int inodes) { unsigned int length = SQUASHFS_LOOKUP_BLOCK_BYTES(inodes); - __le64 *inode_lookup_table; - int err; + __le64 *table; TRACE("In read_inode_lookup_table, length %d\n", length); - /* Allocate inode lookup table indexes */ - inode_lookup_table = kmalloc(length, GFP_KERNEL); - if (inode_lookup_table == NULL) { - ERROR("Failed to allocate inode lookup table\n"); - return ERR_PTR(-ENOMEM); - } + /* Sanity check values */ + + /* there should always be at least one inode */ + if (inodes == 0) + return ERR_PTR(-EINVAL); + + /* length bytes should not extend into the next table - this check + * also traps instances where lookup_table_start is incorrectly larger + * than the next table start + */ + if (lookup_table_start + length > next_table) + return ERR_PTR(-EINVAL); + + table = squashfs_read_table(sb, lookup_table_start, length); - err = squashfs_read_table(sb, inode_lookup_table, lookup_table_start, - length); - if (err < 0) { - ERROR("unable to read inode lookup table\n"); - kfree(inode_lookup_table); - return ERR_PTR(err); + /* + * table[0] points to the first inode lookup table metadata block, + * this should be less than lookup_table_start + */ + if (!IS_ERR(table) && le64_to_cpu(table[0]) >= lookup_table_start) { + kfree(table); + return ERR_PTR(-EINVAL); } - return inode_lookup_table; + return table; } diff --git a/fs/squashfs/file.c b/fs/squashfs/file.c index a25c5060bdcb..38bb1c640559 100644 --- a/fs/squashfs/file.c +++ b/fs/squashfs/file.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/fragment.c b/fs/squashfs/fragment.c index 7eef571443c6..0ed6edbc5c71 100644 --- a/fs/squashfs/fragment.c +++ b/fs/squashfs/fragment.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -71,26 +71,29 @@ int squashfs_frag_lookup(struct super_block *sb, unsigned int fragment, * Read the uncompressed fragment lookup table indexes off disk into memory */ __le64 *squashfs_read_fragment_index_table(struct super_block *sb, - u64 fragment_table_start, unsigned int fragments) + u64 fragment_table_start, u64 next_table, unsigned int fragments) { unsigned int length = SQUASHFS_FRAGMENT_INDEX_BYTES(fragments); - __le64 *fragment_index; - int err; + __le64 *table; - /* Allocate fragment lookup table indexes */ - fragment_index = kmalloc(length, GFP_KERNEL); - if (fragment_index == NULL) { - ERROR("Failed to allocate fragment index table\n"); - return ERR_PTR(-ENOMEM); - } + /* + * Sanity check, length bytes should not extend into the next table - + * this check also traps instances where fragment_table_start is + * incorrectly larger than the next table start + */ + if (fragment_table_start + length > next_table) + return ERR_PTR(-EINVAL); + + table = squashfs_read_table(sb, fragment_table_start, length); - err = squashfs_read_table(sb, fragment_index, fragment_table_start, - length); - if (err < 0) { - ERROR("unable to read fragment index table\n"); - kfree(fragment_index); - return ERR_PTR(err); + /* + * table[0] points to the first fragment table metadata block, this + * should be less than fragment_table_start + */ + if (!IS_ERR(table) && le64_to_cpu(table[0]) >= fragment_table_start) { + kfree(table); + return ERR_PTR(-EINVAL); } - return fragment_index; + return table; } diff --git a/fs/squashfs/id.c b/fs/squashfs/id.c index d8f32452638e..d38ea3dab951 100644 --- a/fs/squashfs/id.c +++ b/fs/squashfs/id.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -66,27 +66,37 @@ int squashfs_get_id(struct super_block *sb, unsigned int index, * Read uncompressed id lookup table indexes from disk into memory */ __le64 *squashfs_read_id_index_table(struct super_block *sb, - u64 id_table_start, unsigned short no_ids) + u64 id_table_start, u64 next_table, unsigned short no_ids) { unsigned int length = SQUASHFS_ID_BLOCK_BYTES(no_ids); - __le64 *id_table; - int err; + __le64 *table; TRACE("In read_id_index_table, length %d\n", length); - /* Allocate id lookup table indexes */ - id_table = kmalloc(length, GFP_KERNEL); - if (id_table == NULL) { - ERROR("Failed to allocate id index table\n"); - return ERR_PTR(-ENOMEM); - } + /* Sanity check values */ + + /* there should always be at least one id */ + if (no_ids == 0) + return ERR_PTR(-EINVAL); + + /* + * length bytes should not extend into the next table - this check + * also traps instances where id_table_start is incorrectly larger + * than the next table start + */ + if (id_table_start + length > next_table) + return ERR_PTR(-EINVAL); + + table = squashfs_read_table(sb, id_table_start, length); - err = squashfs_read_table(sb, id_table, id_table_start, length); - if (err < 0) { - ERROR("unable to read id index table\n"); - kfree(id_table); - return ERR_PTR(err); + /* + * table[0] points to the first id lookup table metadata block, this + * should be less than id_table_start + */ + if (!IS_ERR(table) && le64_to_cpu(table[0]) >= id_table_start) { + kfree(table); + return ERR_PTR(-EINVAL); } - return id_table; + return table; } diff --git a/fs/squashfs/inode.c b/fs/squashfs/inode.c index 62e63ad25075..04bebcaa2373 100644 --- a/fs/squashfs/inode.c +++ b/fs/squashfs/inode.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/namei.c b/fs/squashfs/namei.c index 5d922a6701ab..4bc63ac64bc0 100644 --- a/fs/squashfs/namei.c +++ b/fs/squashfs/namei.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/squashfs.h b/fs/squashfs/squashfs.h index 1f2e608b8785..e3be6a71cfa7 100644 --- a/fs/squashfs/squashfs.h +++ b/fs/squashfs/squashfs.h @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -44,24 +44,24 @@ extern struct squashfs_cache_entry *squashfs_get_fragment(struct super_block *, u64, int); extern struct squashfs_cache_entry *squashfs_get_datablock(struct super_block *, u64, int); -extern int squashfs_read_table(struct super_block *, void *, u64, int); +extern void *squashfs_read_table(struct super_block *, u64, int); /* decompressor.c */ extern const struct squashfs_decompressor *squashfs_lookup_decompressor(int); extern void *squashfs_decompressor_init(struct super_block *, unsigned short); /* export.c */ -extern __le64 *squashfs_read_inode_lookup_table(struct super_block *, u64, +extern __le64 *squashfs_read_inode_lookup_table(struct super_block *, u64, u64, unsigned int); /* fragment.c */ extern int squashfs_frag_lookup(struct super_block *, unsigned int, u64 *); extern __le64 *squashfs_read_fragment_index_table(struct super_block *, - u64, unsigned int); + u64, u64, unsigned int); /* id.c */ extern int squashfs_get_id(struct super_block *, unsigned int, unsigned int *); -extern __le64 *squashfs_read_id_index_table(struct super_block *, u64, +extern __le64 *squashfs_read_id_index_table(struct super_block *, u64, u64, unsigned short); /* inode.c */ diff --git a/fs/squashfs/squashfs_fs.h b/fs/squashfs/squashfs_fs.h index 4582c568ef4d..b4a4e539a08c 100644 --- a/fs/squashfs/squashfs_fs.h +++ b/fs/squashfs/squashfs_fs.h @@ -4,7 +4,7 @@ * Squashfs * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/squashfs_fs_i.h b/fs/squashfs/squashfs_fs_i.h index 359baefc01fc..73588e7700ed 100644 --- a/fs/squashfs/squashfs_fs_i.h +++ b/fs/squashfs/squashfs_fs_i.h @@ -4,7 +4,7 @@ * Squashfs * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/squashfs_fs_sb.h b/fs/squashfs/squashfs_fs_sb.h index d9037a5215f0..651f0b31d296 100644 --- a/fs/squashfs/squashfs_fs_sb.h +++ b/fs/squashfs/squashfs_fs_sb.h @@ -4,7 +4,7 @@ * Squashfs * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/super.c b/fs/squashfs/super.c index 5c8184c061a4..7438850c62d0 100644 --- a/fs/squashfs/super.c +++ b/fs/squashfs/super.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -83,7 +83,7 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent) long long root_inode; unsigned short flags; unsigned int fragments; - u64 lookup_table_start, xattr_id_table_start; + u64 lookup_table_start, xattr_id_table_start, next_table; int err; TRACE("Entered squashfs_fill_superblock\n"); @@ -95,12 +95,6 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent) } msblk = sb->s_fs_info; - sblk = kzalloc(sizeof(*sblk), GFP_KERNEL); - if (sblk == NULL) { - ERROR("Failed to allocate squashfs_super_block\n"); - goto failure; - } - msblk->devblksize = sb_min_blocksize(sb, BLOCK_SIZE); msblk->devblksize_log2 = ffz(~msblk->devblksize); @@ -114,10 +108,12 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent) * of bytes_used) we need to set it to an initial sensible dummy value */ msblk->bytes_used = sizeof(*sblk); - err = squashfs_read_table(sb, sblk, SQUASHFS_START, sizeof(*sblk)); + sblk = squashfs_read_table(sb, SQUASHFS_START, sizeof(*sblk)); - if (err < 0) { + if (IS_ERR(sblk)) { ERROR("unable to read squashfs_super_block\n"); + err = PTR_ERR(sblk); + sblk = NULL; goto failed_mount; } @@ -218,18 +214,61 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent) goto failed_mount; } + /* Handle xattrs */ + sb->s_xattr = squashfs_xattr_handlers; + xattr_id_table_start = le64_to_cpu(sblk->xattr_id_table_start); + if (xattr_id_table_start == SQUASHFS_INVALID_BLK) { + next_table = msblk->bytes_used; + goto allocate_id_index_table; + } + + /* Allocate and read xattr id lookup table */ + msblk->xattr_id_table = squashfs_read_xattr_id_table(sb, + xattr_id_table_start, &msblk->xattr_table, &msblk->xattr_ids); + if (IS_ERR(msblk->xattr_id_table)) { + ERROR("unable to read xattr id index table\n"); + err = PTR_ERR(msblk->xattr_id_table); + msblk->xattr_id_table = NULL; + if (err != -ENOTSUPP) + goto failed_mount; + } + next_table = msblk->xattr_table; + +allocate_id_index_table: /* Allocate and read id index table */ msblk->id_table = squashfs_read_id_index_table(sb, - le64_to_cpu(sblk->id_table_start), le16_to_cpu(sblk->no_ids)); + le64_to_cpu(sblk->id_table_start), next_table, + le16_to_cpu(sblk->no_ids)); if (IS_ERR(msblk->id_table)) { + ERROR("unable to read id index table\n"); err = PTR_ERR(msblk->id_table); msblk->id_table = NULL; goto failed_mount; } + next_table = le64_to_cpu(msblk->id_table[0]); + + /* Handle inode lookup table */ + lookup_table_start = le64_to_cpu(sblk->lookup_table_start); + if (lookup_table_start == SQUASHFS_INVALID_BLK) + goto handle_fragments; + + /* Allocate and read inode lookup table */ + msblk->inode_lookup_table = squashfs_read_inode_lookup_table(sb, + lookup_table_start, next_table, msblk->inodes); + if (IS_ERR(msblk->inode_lookup_table)) { + ERROR("unable to read inode lookup table\n"); + err = PTR_ERR(msblk->inode_lookup_table); + msblk->inode_lookup_table = NULL; + goto failed_mount; + } + next_table = le64_to_cpu(msblk->inode_lookup_table[0]); + sb->s_export_op = &squashfs_export_ops; + +handle_fragments: fragments = le32_to_cpu(sblk->fragments); if (fragments == 0) - goto allocate_lookup_table; + goto check_directory_table; msblk->fragment_cache = squashfs_cache_init("fragment", SQUASHFS_CACHED_FRAGMENTS, msblk->block_size); @@ -240,45 +279,29 @@ static int squashfs_fill_super(struct super_block *sb, void *data, int silent) /* Allocate and read fragment index table */ msblk->fragment_index = squashfs_read_fragment_index_table(sb, - le64_to_cpu(sblk->fragment_table_start), fragments); + le64_to_cpu(sblk->fragment_table_start), next_table, fragments); if (IS_ERR(msblk->fragment_index)) { + ERROR("unable to read fragment index table\n"); err = PTR_ERR(msblk->fragment_index); msblk->fragment_index = NULL; goto failed_mount; } + next_table = le64_to_cpu(msblk->fragment_index[0]); -allocate_lookup_table: - lookup_table_start = le64_to_cpu(sblk->lookup_table_start); - if (lookup_table_start == SQUASHFS_INVALID_BLK) - goto allocate_xattr_table; - - /* Allocate and read inode lookup table */ - msblk->inode_lookup_table = squashfs_read_inode_lookup_table(sb, - lookup_table_start, msblk->inodes); - if (IS_ERR(msblk->inode_lookup_table)) { - err = PTR_ERR(msblk->inode_lookup_table); - msblk->inode_lookup_table = NULL; +check_directory_table: + /* Sanity check directory_table */ + if (msblk->directory_table >= next_table) { + err = -EINVAL; goto failed_mount; } - sb->s_export_op = &squashfs_export_ops; - -allocate_xattr_table: - sb->s_xattr = squashfs_xattr_handlers; - xattr_id_table_start = le64_to_cpu(sblk->xattr_id_table_start); - if (xattr_id_table_start == SQUASHFS_INVALID_BLK) - goto allocate_root; - - /* Allocate and read xattr id lookup table */ - msblk->xattr_id_table = squashfs_read_xattr_id_table(sb, - xattr_id_table_start, &msblk->xattr_table, &msblk->xattr_ids); - if (IS_ERR(msblk->xattr_id_table)) { - err = PTR_ERR(msblk->xattr_id_table); - msblk->xattr_id_table = NULL; - if (err != -ENOTSUPP) - goto failed_mount; + /* Sanity check inode_table */ + if (msblk->inode_table >= msblk->directory_table) { + err = -EINVAL; + goto failed_mount; } -allocate_root: + + /* allocate root */ root = new_inode(sb); if (!root) { err = -ENOMEM; @@ -318,11 +341,6 @@ failed_mount: sb->s_fs_info = NULL; kfree(sblk); return err; - -failure: - kfree(sb->s_fs_info); - sb->s_fs_info = NULL; - return -ENOMEM; } @@ -475,5 +493,5 @@ static const struct super_operations squashfs_super_ops = { module_init(init_squashfs_fs); module_exit(exit_squashfs_fs); MODULE_DESCRIPTION("squashfs 4.0, a compressed read-only filesystem"); -MODULE_AUTHOR("Phillip Lougher <phillip@lougher.demon.co.uk>"); +MODULE_AUTHOR("Phillip Lougher <phillip@squashfs.org.uk>"); MODULE_LICENSE("GPL"); diff --git a/fs/squashfs/symlink.c b/fs/squashfs/symlink.c index ec86434921e1..1191817264cc 100644 --- a/fs/squashfs/symlink.c +++ b/fs/squashfs/symlink.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/xattr.c b/fs/squashfs/xattr.c index 3876c36699a1..92fcde7b4d61 100644 --- a/fs/squashfs/xattr.c +++ b/fs/squashfs/xattr.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2010 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/xattr.h b/fs/squashfs/xattr.h index b634efce4bde..c83f5d9ec125 100644 --- a/fs/squashfs/xattr.h +++ b/fs/squashfs/xattr.h @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2010 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -31,6 +31,7 @@ static inline __le64 *squashfs_read_xattr_id_table(struct super_block *sb, u64 start, u64 *xattr_table_start, int *xattr_ids) { ERROR("Xattrs in filesystem, these will be ignored\n"); + *xattr_table_start = start; return ERR_PTR(-ENOTSUPP); } diff --git a/fs/squashfs/xattr_id.c b/fs/squashfs/xattr_id.c index 05385dbe1465..c89607d690c4 100644 --- a/fs/squashfs/xattr_id.c +++ b/fs/squashfs/xattr_id.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2010 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License @@ -67,34 +67,29 @@ __le64 *squashfs_read_xattr_id_table(struct super_block *sb, u64 start, u64 *xattr_table_start, int *xattr_ids) { unsigned int len; - __le64 *xid_table; - struct squashfs_xattr_id_table id_table; - int err; + struct squashfs_xattr_id_table *id_table; + + id_table = squashfs_read_table(sb, start, sizeof(*id_table)); + if (IS_ERR(id_table)) + return (__le64 *) id_table; + + *xattr_table_start = le64_to_cpu(id_table->xattr_table_start); + *xattr_ids = le32_to_cpu(id_table->xattr_ids); + kfree(id_table); + + /* Sanity check values */ + + /* there is always at least one xattr id */ + if (*xattr_ids == 0) + return ERR_PTR(-EINVAL); + + /* xattr_table should be less than start */ + if (*xattr_table_start >= start) + return ERR_PTR(-EINVAL); - err = squashfs_read_table(sb, &id_table, start, sizeof(id_table)); - if (err < 0) { - ERROR("unable to read xattr id table\n"); - return ERR_PTR(err); - } - *xattr_table_start = le64_to_cpu(id_table.xattr_table_start); - *xattr_ids = le32_to_cpu(id_table.xattr_ids); len = SQUASHFS_XATTR_BLOCK_BYTES(*xattr_ids); TRACE("In read_xattr_index_table, length %d\n", len); - /* Allocate xattr id lookup table indexes */ - xid_table = kmalloc(len, GFP_KERNEL); - if (xid_table == NULL) { - ERROR("Failed to allocate xattr id index table\n"); - return ERR_PTR(-ENOMEM); - } - - err = squashfs_read_table(sb, xid_table, start + sizeof(id_table), len); - if (err < 0) { - ERROR("unable to read xattr id index table\n"); - kfree(xid_table); - return ERR_PTR(err); - } - - return xid_table; + return squashfs_read_table(sb, start + sizeof(*id_table), len); } diff --git a/fs/squashfs/xz_wrapper.c b/fs/squashfs/xz_wrapper.c index aa47a286d1f8..1760b7d108f6 100644 --- a/fs/squashfs/xz_wrapper.c +++ b/fs/squashfs/xz_wrapper.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/squashfs/zlib_wrapper.c b/fs/squashfs/zlib_wrapper.c index 517688b32ffa..55d918fd2d86 100644 --- a/fs/squashfs/zlib_wrapper.c +++ b/fs/squashfs/zlib_wrapper.c @@ -2,7 +2,7 @@ * Squashfs - a compressed read only filesystem for Linux * * Copyright (c) 2002, 2003, 2004, 2005, 2006, 2007, 2008, 2009 - * Phillip Lougher <phillip@lougher.demon.co.uk> + * Phillip Lougher <phillip@squashfs.org.uk> * * This program is free software; you can redistribute it and/or * modify it under the terms of the GNU General Public License diff --git a/fs/super.c b/fs/super.c index 8a06881b1920..c75593953c52 100644 --- a/fs/super.c +++ b/fs/super.c @@ -31,6 +31,7 @@ #include <linux/mutex.h> #include <linux/backing-dev.h> #include <linux/rculist_bl.h> +#include <linux/cleancache.h> #include "internal.h" @@ -112,6 +113,7 @@ static struct super_block *alloc_super(struct file_system_type *type) s->s_maxbytes = MAX_NON_LFS; s->s_op = &default_op; s->s_time_gran = 1000000000; + s->cleancache_poolid = -1; } out: return s; @@ -177,6 +179,7 @@ void deactivate_locked_super(struct super_block *s) { struct file_system_type *fs = s->s_type; if (atomic_dec_and_test(&s->s_active)) { + cleancache_flush_fs(s); fs->kill_sb(s); /* * We need to call rcu_barrier so all the delayed rcu free @@ -948,8 +951,7 @@ mount_fs(struct file_system_type *type, int flags, const char *name, void *data) * filesystems should never set s_maxbytes larger than MAX_LFS_FILESIZE * but s_maxbytes was an unsigned long long for many releases. Throw * this warning for a little while to try and catch filesystems that - * violate this rule. This warning should be either removed or - * converted to a BUG() in 2.6.34. + * violate this rule. */ WARN((sb->s_maxbytes < 0), "%s set sb->s_maxbytes to " "negative value (%lld)\n", type->name, sb->s_maxbytes); diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index da3fefe91a8f..1ad8c93c1b85 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -24,13 +24,6 @@ #include "sysfs.h" -/* used in crash dumps to help with debugging */ -static char last_sysfs_file[PATH_MAX]; -void sysfs_printk_last_file(void) -{ - printk(KERN_EMERG "last sysfs file: %s\n", last_sysfs_file); -} - /* * There's one sysfs_buffer for each open file and one * sysfs_open_dirent for each sysfs_dirent with one or more open @@ -337,11 +330,6 @@ static int sysfs_open_file(struct inode *inode, struct file *file) struct sysfs_buffer *buffer; const struct sysfs_ops *ops; int error = -EACCES; - char *p; - - p = d_path(&file->f_path, last_sysfs_file, sizeof(last_sysfs_file)); - if (!IS_ERR(p)) - memmove(last_sysfs_file, p, strlen(p) + 1); /* need attr_sd for attr and ops, its parent for kobj */ if (!sysfs_get_active(attr_sd)) diff --git a/fs/sysfs/group.c b/fs/sysfs/group.c index c8769dc222d8..194414f8298c 100644 --- a/fs/sysfs/group.c +++ b/fs/sysfs/group.c @@ -101,9 +101,9 @@ int sysfs_create_group(struct kobject *kobj, } /** - * sysfs_update_group - given a directory kobject, create an attribute group - * @kobj: The kobject to create the group on - * @grp: The attribute group to create + * sysfs_update_group - given a directory kobject, update an attribute group + * @kobj: The kobject to update the group on + * @grp: The attribute group to update * * This function updates an attribute group. Unlike * sysfs_create_group(), it will explicitly not warn or error if any diff --git a/fs/timerfd.c b/fs/timerfd.c index 8c4fc1425b3e..f67acbdda5e8 100644 --- a/fs/timerfd.c +++ b/fs/timerfd.c @@ -22,16 +22,24 @@ #include <linux/anon_inodes.h> #include <linux/timerfd.h> #include <linux/syscalls.h> +#include <linux/rcupdate.h> struct timerfd_ctx { struct hrtimer tmr; ktime_t tintv; + ktime_t moffs; wait_queue_head_t wqh; u64 ticks; int expired; int clockid; + struct rcu_head rcu; + struct list_head clist; + bool might_cancel; }; +static LIST_HEAD(cancel_list); +static DEFINE_SPINLOCK(cancel_lock); + /* * This gets called when the timer event triggers. We set the "expired" * flag, but we do not re-arm the timer (in case it's necessary, @@ -51,6 +59,63 @@ static enum hrtimer_restart timerfd_tmrproc(struct hrtimer *htmr) return HRTIMER_NORESTART; } +/* + * Called when the clock was set to cancel the timers in the cancel + * list. + */ +void timerfd_clock_was_set(void) +{ + ktime_t moffs = ktime_get_monotonic_offset(); + struct timerfd_ctx *ctx; + unsigned long flags; + + rcu_read_lock(); + list_for_each_entry_rcu(ctx, &cancel_list, clist) { + if (!ctx->might_cancel) + continue; + spin_lock_irqsave(&ctx->wqh.lock, flags); + if (ctx->moffs.tv64 != moffs.tv64) { + ctx->moffs.tv64 = KTIME_MAX; + wake_up_locked(&ctx->wqh); + } + spin_unlock_irqrestore(&ctx->wqh.lock, flags); + } + rcu_read_unlock(); +} + +static void timerfd_remove_cancel(struct timerfd_ctx *ctx) +{ + if (ctx->might_cancel) { + ctx->might_cancel = false; + spin_lock(&cancel_lock); + list_del_rcu(&ctx->clist); + spin_unlock(&cancel_lock); + } +} + +static bool timerfd_canceled(struct timerfd_ctx *ctx) +{ + if (!ctx->might_cancel || ctx->moffs.tv64 != KTIME_MAX) + return false; + ctx->moffs = ktime_get_monotonic_offset(); + return true; +} + +static void timerfd_setup_cancel(struct timerfd_ctx *ctx, int flags) +{ + if (ctx->clockid == CLOCK_REALTIME && (flags & TFD_TIMER_ABSTIME) && + (flags & TFD_TIMER_CANCEL_ON_SET)) { + if (!ctx->might_cancel) { + ctx->might_cancel = true; + spin_lock(&cancel_lock); + list_add_rcu(&ctx->clist, &cancel_list); + spin_unlock(&cancel_lock); + } + } else if (ctx->might_cancel) { + timerfd_remove_cancel(ctx); + } +} + static ktime_t timerfd_get_remaining(struct timerfd_ctx *ctx) { ktime_t remaining; @@ -59,11 +124,12 @@ static ktime_t timerfd_get_remaining(struct timerfd_ctx *ctx) return remaining.tv64 < 0 ? ktime_set(0, 0): remaining; } -static void timerfd_setup(struct timerfd_ctx *ctx, int flags, - const struct itimerspec *ktmr) +static int timerfd_setup(struct timerfd_ctx *ctx, int flags, + const struct itimerspec *ktmr) { enum hrtimer_mode htmode; ktime_t texp; + int clockid = ctx->clockid; htmode = (flags & TFD_TIMER_ABSTIME) ? HRTIMER_MODE_ABS: HRTIMER_MODE_REL; @@ -72,19 +138,24 @@ static void timerfd_setup(struct timerfd_ctx *ctx, int flags, ctx->expired = 0; ctx->ticks = 0; ctx->tintv = timespec_to_ktime(ktmr->it_interval); - hrtimer_init(&ctx->tmr, ctx->clockid, htmode); + hrtimer_init(&ctx->tmr, clockid, htmode); hrtimer_set_expires(&ctx->tmr, texp); ctx->tmr.function = timerfd_tmrproc; - if (texp.tv64 != 0) + if (texp.tv64 != 0) { hrtimer_start(&ctx->tmr, texp, htmode); + if (timerfd_canceled(ctx)) + return -ECANCELED; + } + return 0; } static int timerfd_release(struct inode *inode, struct file *file) { struct timerfd_ctx *ctx = file->private_data; + timerfd_remove_cancel(ctx); hrtimer_cancel(&ctx->tmr); - kfree(ctx); + kfree_rcu(ctx, rcu); return 0; } @@ -118,8 +189,21 @@ static ssize_t timerfd_read(struct file *file, char __user *buf, size_t count, res = -EAGAIN; else res = wait_event_interruptible_locked_irq(ctx->wqh, ctx->ticks); + + /* + * If clock has changed, we do not care about the + * ticks and we do not rearm the timer. Userspace must + * reevaluate anyway. + */ + if (timerfd_canceled(ctx)) { + ctx->ticks = 0; + ctx->expired = 0; + res = -ECANCELED; + } + if (ctx->ticks) { ticks = ctx->ticks; + if (ctx->expired && ctx->tintv.tv64) { /* * If tintv.tv64 != 0, this is a periodic timer that @@ -183,6 +267,7 @@ SYSCALL_DEFINE2(timerfd_create, int, clockid, int, flags) init_waitqueue_head(&ctx->wqh); ctx->clockid = clockid; hrtimer_init(&ctx->tmr, clockid, HRTIMER_MODE_ABS); + ctx->moffs = ktime_get_monotonic_offset(); ufd = anon_inode_getfd("[timerfd]", &timerfd_fops, ctx, O_RDWR | (flags & TFD_SHARED_FCNTL_FLAGS)); @@ -199,6 +284,7 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, struct file *file; struct timerfd_ctx *ctx; struct itimerspec ktmr, kotmr; + int ret; if (copy_from_user(&ktmr, utmr, sizeof(ktmr))) return -EFAULT; @@ -213,6 +299,8 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, return PTR_ERR(file); ctx = file->private_data; + timerfd_setup_cancel(ctx, flags); + /* * We need to stop the existing timer before reprogramming * it to the new values. @@ -240,14 +328,14 @@ SYSCALL_DEFINE4(timerfd_settime, int, ufd, int, flags, /* * Re-program the timer to the new value ... */ - timerfd_setup(ctx, flags, &ktmr); + ret = timerfd_setup(ctx, flags, &ktmr); spin_unlock_irq(&ctx->wqh.lock); fput(file); if (otmr && copy_to_user(otmr, &kotmr, sizeof(kotmr))) return -EFAULT; - return 0; + return ret; } SYSCALL_DEFINE2(timerfd_gettime, int, ufd, struct itimerspec __user *, otmr) diff --git a/fs/ubifs/Kconfig b/fs/ubifs/Kconfig index d7440904be17..f8b0160da2da 100644 --- a/fs/ubifs/Kconfig +++ b/fs/ubifs/Kconfig @@ -47,7 +47,7 @@ config UBIFS_FS_DEBUG bool "Enable debugging support" depends on UBIFS_FS select DEBUG_FS - select KALLSYMS_ALL + select KALLSYMS help This option enables UBIFS debugging support. It makes sure various assertions, self-checks, debugging messages and test modes are compiled diff --git a/fs/ubifs/budget.c b/fs/ubifs/budget.c index c8ff0d1ae5d3..315de66e52b2 100644 --- a/fs/ubifs/budget.c +++ b/fs/ubifs/budget.c @@ -106,7 +106,7 @@ static long long get_liability(struct ubifs_info *c) long long liab; spin_lock(&c->space_lock); - liab = c->budg_idx_growth + c->budg_data_growth + c->budg_dd_growth; + liab = c->bi.idx_growth + c->bi.data_growth + c->bi.dd_growth; spin_unlock(&c->space_lock); return liab; } @@ -147,7 +147,7 @@ static int make_free_space(struct ubifs_info *c) if (liab2 < liab1) return -EAGAIN; - dbg_budg("new liability %lld (not shrinked)", liab2); + dbg_budg("new liability %lld (not shrunk)", liab2); /* Liability did not shrink again, try GC */ dbg_budg("Run GC"); @@ -180,7 +180,7 @@ int ubifs_calc_min_idx_lebs(struct ubifs_info *c) int idx_lebs; long long idx_size; - idx_size = c->old_idx_sz + c->budg_idx_growth + c->budg_uncommitted_idx; + idx_size = c->bi.old_idx_sz + c->bi.idx_growth + c->bi.uncommitted_idx; /* And make sure we have thrice the index size of space reserved */ idx_size += idx_size << 1; /* @@ -292,13 +292,13 @@ static int can_use_rp(struct ubifs_info *c) * budgeted index space to the size of the current index, multiplies this by 3, * and makes sure this does not exceed the amount of free LEBs. * - * Notes about @c->min_idx_lebs and @c->lst.idx_lebs variables: + * Notes about @c->bi.min_idx_lebs and @c->lst.idx_lebs variables: * o @c->lst.idx_lebs is the number of LEBs the index currently uses. It might * be large, because UBIFS does not do any index consolidation as long as * there is free space. IOW, the index may take a lot of LEBs, but the LEBs * will contain a lot of dirt. - * o @c->min_idx_lebs is the number of LEBS the index presumably takes. IOW, - * the index may be consolidated to take up to @c->min_idx_lebs LEBs. + * o @c->bi.min_idx_lebs is the number of LEBS the index presumably takes. IOW, + * the index may be consolidated to take up to @c->bi.min_idx_lebs LEBs. * * This function returns zero in case of success, and %-ENOSPC in case of * failure. @@ -343,13 +343,13 @@ static int do_budget_space(struct ubifs_info *c) c->lst.taken_empty_lebs; if (unlikely(rsvd_idx_lebs > lebs)) { dbg_budg("out of indexing space: min_idx_lebs %d (old %d), " - "rsvd_idx_lebs %d", min_idx_lebs, c->min_idx_lebs, + "rsvd_idx_lebs %d", min_idx_lebs, c->bi.min_idx_lebs, rsvd_idx_lebs); return -ENOSPC; } available = ubifs_calc_available(c, min_idx_lebs); - outstanding = c->budg_data_growth + c->budg_dd_growth; + outstanding = c->bi.data_growth + c->bi.dd_growth; if (unlikely(available < outstanding)) { dbg_budg("out of data space: available %lld, outstanding %lld", @@ -360,7 +360,7 @@ static int do_budget_space(struct ubifs_info *c) if (available - outstanding <= c->rp_size && !can_use_rp(c)) return -ENOSPC; - c->min_idx_lebs = min_idx_lebs; + c->bi.min_idx_lebs = min_idx_lebs; return 0; } @@ -393,11 +393,11 @@ static int calc_data_growth(const struct ubifs_info *c, { int data_growth; - data_growth = req->new_ino ? c->inode_budget : 0; + data_growth = req->new_ino ? c->bi.inode_budget : 0; if (req->new_page) - data_growth += c->page_budget; + data_growth += c->bi.page_budget; if (req->new_dent) - data_growth += c->dent_budget; + data_growth += c->bi.dent_budget; data_growth += req->new_ino_d; return data_growth; } @@ -413,12 +413,12 @@ static int calc_dd_growth(const struct ubifs_info *c, { int dd_growth; - dd_growth = req->dirtied_page ? c->page_budget : 0; + dd_growth = req->dirtied_page ? c->bi.page_budget : 0; if (req->dirtied_ino) - dd_growth += c->inode_budget << (req->dirtied_ino - 1); + dd_growth += c->bi.inode_budget << (req->dirtied_ino - 1); if (req->mod_dent) - dd_growth += c->dent_budget; + dd_growth += c->bi.dent_budget; dd_growth += req->dirtied_ino_d; return dd_growth; } @@ -460,19 +460,19 @@ int ubifs_budget_space(struct ubifs_info *c, struct ubifs_budget_req *req) again: spin_lock(&c->space_lock); - ubifs_assert(c->budg_idx_growth >= 0); - ubifs_assert(c->budg_data_growth >= 0); - ubifs_assert(c->budg_dd_growth >= 0); + ubifs_assert(c->bi.idx_growth >= 0); + ubifs_assert(c->bi.data_growth >= 0); + ubifs_assert(c->bi.dd_growth >= 0); - if (unlikely(c->nospace) && (c->nospace_rp || !can_use_rp(c))) { + if (unlikely(c->bi.nospace) && (c->bi.nospace_rp || !can_use_rp(c))) { dbg_budg("no space"); spin_unlock(&c->space_lock); return -ENOSPC; } - c->budg_idx_growth += idx_growth; - c->budg_data_growth += data_growth; - c->budg_dd_growth += dd_growth; + c->bi.idx_growth += idx_growth; + c->bi.data_growth += data_growth; + c->bi.dd_growth += dd_growth; err = do_budget_space(c); if (likely(!err)) { @@ -484,9 +484,9 @@ again: } /* Restore the old values */ - c->budg_idx_growth -= idx_growth; - c->budg_data_growth -= data_growth; - c->budg_dd_growth -= dd_growth; + c->bi.idx_growth -= idx_growth; + c->bi.data_growth -= data_growth; + c->bi.dd_growth -= dd_growth; spin_unlock(&c->space_lock); if (req->fast) { @@ -506,9 +506,9 @@ again: goto again; } dbg_budg("FS is full, -ENOSPC"); - c->nospace = 1; + c->bi.nospace = 1; if (can_use_rp(c) || c->rp_size == 0) - c->nospace_rp = 1; + c->bi.nospace_rp = 1; smp_wmb(); } else ubifs_err("cannot budget space, error %d", err); @@ -523,8 +523,8 @@ again: * This function releases the space budgeted by 'ubifs_budget_space()'. Note, * since the index changes (which were budgeted for in @req->idx_growth) will * only be written to the media on commit, this function moves the index budget - * from @c->budg_idx_growth to @c->budg_uncommitted_idx. The latter will be - * zeroed by the commit operation. + * from @c->bi.idx_growth to @c->bi.uncommitted_idx. The latter will be zeroed + * by the commit operation. */ void ubifs_release_budget(struct ubifs_info *c, struct ubifs_budget_req *req) { @@ -553,23 +553,23 @@ void ubifs_release_budget(struct ubifs_info *c, struct ubifs_budget_req *req) if (!req->data_growth && !req->dd_growth) return; - c->nospace = c->nospace_rp = 0; + c->bi.nospace = c->bi.nospace_rp = 0; smp_wmb(); spin_lock(&c->space_lock); - c->budg_idx_growth -= req->idx_growth; - c->budg_uncommitted_idx += req->idx_growth; - c->budg_data_growth -= req->data_growth; - c->budg_dd_growth -= req->dd_growth; - c->min_idx_lebs = ubifs_calc_min_idx_lebs(c); - - ubifs_assert(c->budg_idx_growth >= 0); - ubifs_assert(c->budg_data_growth >= 0); - ubifs_assert(c->budg_dd_growth >= 0); - ubifs_assert(c->min_idx_lebs < c->main_lebs); - ubifs_assert(!(c->budg_idx_growth & 7)); - ubifs_assert(!(c->budg_data_growth & 7)); - ubifs_assert(!(c->budg_dd_growth & 7)); + c->bi.idx_growth -= req->idx_growth; + c->bi.uncommitted_idx += req->idx_growth; + c->bi.data_growth -= req->data_growth; + c->bi.dd_growth -= req->dd_growth; + c->bi.min_idx_lebs = ubifs_calc_min_idx_lebs(c); + + ubifs_assert(c->bi.idx_growth >= 0); + ubifs_assert(c->bi.data_growth >= 0); + ubifs_assert(c->bi.dd_growth >= 0); + ubifs_assert(c->bi.min_idx_lebs < c->main_lebs); + ubifs_assert(!(c->bi.idx_growth & 7)); + ubifs_assert(!(c->bi.data_growth & 7)); + ubifs_assert(!(c->bi.dd_growth & 7)); spin_unlock(&c->space_lock); } @@ -586,13 +586,13 @@ void ubifs_convert_page_budget(struct ubifs_info *c) { spin_lock(&c->space_lock); /* Release the index growth reservation */ - c->budg_idx_growth -= c->max_idx_node_sz << UBIFS_BLOCKS_PER_PAGE_SHIFT; + c->bi.idx_growth -= c->max_idx_node_sz << UBIFS_BLOCKS_PER_PAGE_SHIFT; /* Release the data growth reservation */ - c->budg_data_growth -= c->page_budget; + c->bi.data_growth -= c->bi.page_budget; /* Increase the dirty data growth reservation instead */ - c->budg_dd_growth += c->page_budget; + c->bi.dd_growth += c->bi.page_budget; /* And re-calculate the indexing space reservation */ - c->min_idx_lebs = ubifs_calc_min_idx_lebs(c); + c->bi.min_idx_lebs = ubifs_calc_min_idx_lebs(c); spin_unlock(&c->space_lock); } @@ -612,7 +612,7 @@ void ubifs_release_dirty_inode_budget(struct ubifs_info *c, memset(&req, 0, sizeof(struct ubifs_budget_req)); /* The "no space" flags will be cleared because dd_growth is > 0 */ - req.dd_growth = c->inode_budget + ALIGN(ui->data_len, 8); + req.dd_growth = c->bi.inode_budget + ALIGN(ui->data_len, 8); ubifs_release_budget(c, &req); } @@ -682,9 +682,9 @@ long long ubifs_get_free_space_nolock(struct ubifs_info *c) int rsvd_idx_lebs, lebs; long long available, outstanding, free; - ubifs_assert(c->min_idx_lebs == ubifs_calc_min_idx_lebs(c)); - outstanding = c->budg_data_growth + c->budg_dd_growth; - available = ubifs_calc_available(c, c->min_idx_lebs); + ubifs_assert(c->bi.min_idx_lebs == ubifs_calc_min_idx_lebs(c)); + outstanding = c->bi.data_growth + c->bi.dd_growth; + available = ubifs_calc_available(c, c->bi.min_idx_lebs); /* * When reporting free space to user-space, UBIFS guarantees that it is @@ -697,8 +697,8 @@ long long ubifs_get_free_space_nolock(struct ubifs_info *c) * Note, the calculations below are similar to what we have in * 'do_budget_space()', so refer there for comments. */ - if (c->min_idx_lebs > c->lst.idx_lebs) - rsvd_idx_lebs = c->min_idx_lebs - c->lst.idx_lebs; + if (c->bi.min_idx_lebs > c->lst.idx_lebs) + rsvd_idx_lebs = c->bi.min_idx_lebs - c->lst.idx_lebs; else rsvd_idx_lebs = 0; lebs = c->lst.empty_lebs + c->freeable_cnt + c->idx_gc_cnt - diff --git a/fs/ubifs/commit.c b/fs/ubifs/commit.c index b148fbc80f8d..87cd0ead8633 100644 --- a/fs/ubifs/commit.c +++ b/fs/ubifs/commit.c @@ -182,7 +182,7 @@ static int do_commit(struct ubifs_info *c) c->mst_node->root_len = cpu_to_le32(zroot.len); c->mst_node->ihead_lnum = cpu_to_le32(c->ihead_lnum); c->mst_node->ihead_offs = cpu_to_le32(c->ihead_offs); - c->mst_node->index_size = cpu_to_le64(c->old_idx_sz); + c->mst_node->index_size = cpu_to_le64(c->bi.old_idx_sz); c->mst_node->lpt_lnum = cpu_to_le32(c->lpt_lnum); c->mst_node->lpt_offs = cpu_to_le32(c->lpt_offs); c->mst_node->nhead_lnum = cpu_to_le32(c->nhead_lnum); @@ -577,7 +577,7 @@ int dbg_check_old_index(struct ubifs_info *c, struct ubifs_zbranch *zroot) size_t sz; if (!(ubifs_chk_flags & UBIFS_CHK_OLD_IDX)) - goto out; + return 0; INIT_LIST_HEAD(&list); diff --git a/fs/ubifs/debug.c b/fs/ubifs/debug.c index f25a7339f800..0bb2bcef0de9 100644 --- a/fs/ubifs/debug.c +++ b/fs/ubifs/debug.c @@ -34,7 +34,6 @@ #include <linux/moduleparam.h> #include <linux/debugfs.h> #include <linux/math64.h> -#include <linux/slab.h> #ifdef CONFIG_UBIFS_FS_DEBUG @@ -43,15 +42,12 @@ DEFINE_SPINLOCK(dbg_lock); static char dbg_key_buf0[128]; static char dbg_key_buf1[128]; -unsigned int ubifs_msg_flags; unsigned int ubifs_chk_flags; unsigned int ubifs_tst_flags; -module_param_named(debug_msgs, ubifs_msg_flags, uint, S_IRUGO | S_IWUSR); module_param_named(debug_chks, ubifs_chk_flags, uint, S_IRUGO | S_IWUSR); module_param_named(debug_tsts, ubifs_tst_flags, uint, S_IRUGO | S_IWUSR); -MODULE_PARM_DESC(debug_msgs, "Debug message type flags"); MODULE_PARM_DESC(debug_chks, "Debug check flags"); MODULE_PARM_DESC(debug_tsts, "Debug special test flags"); @@ -317,6 +313,8 @@ void dbg_dump_node(const struct ubifs_info *c, const void *node) printk(KERN_DEBUG "\tflags %#x\n", sup_flags); printk(KERN_DEBUG "\t big_lpt %u\n", !!(sup_flags & UBIFS_FLG_BIGLPT)); + printk(KERN_DEBUG "\t space_fixup %u\n", + !!(sup_flags & UBIFS_FLG_SPACE_FIXUP)); printk(KERN_DEBUG "\tmin_io_size %u\n", le32_to_cpu(sup->min_io_size)); printk(KERN_DEBUG "\tleb_size %u\n", @@ -602,7 +600,7 @@ void dbg_dump_lstats(const struct ubifs_lp_stats *lst) spin_unlock(&dbg_lock); } -void dbg_dump_budg(struct ubifs_info *c) +void dbg_dump_budg(struct ubifs_info *c, const struct ubifs_budg_info *bi) { int i; struct rb_node *rb; @@ -610,26 +608,42 @@ void dbg_dump_budg(struct ubifs_info *c) struct ubifs_gced_idx_leb *idx_gc; long long available, outstanding, free; - ubifs_assert(spin_is_locked(&c->space_lock)); + spin_lock(&c->space_lock); spin_lock(&dbg_lock); - printk(KERN_DEBUG "(pid %d) Budgeting info: budg_data_growth %lld, " - "budg_dd_growth %lld, budg_idx_growth %lld\n", current->pid, - c->budg_data_growth, c->budg_dd_growth, c->budg_idx_growth); - printk(KERN_DEBUG "\tdata budget sum %lld, total budget sum %lld, " - "freeable_cnt %d\n", c->budg_data_growth + c->budg_dd_growth, - c->budg_data_growth + c->budg_dd_growth + c->budg_idx_growth, - c->freeable_cnt); - printk(KERN_DEBUG "\tmin_idx_lebs %d, old_idx_sz %lld, " - "calc_idx_sz %lld, idx_gc_cnt %d\n", c->min_idx_lebs, - c->old_idx_sz, c->calc_idx_sz, c->idx_gc_cnt); + printk(KERN_DEBUG "(pid %d) Budgeting info: data budget sum %lld, " + "total budget sum %lld\n", current->pid, + bi->data_growth + bi->dd_growth, + bi->data_growth + bi->dd_growth + bi->idx_growth); + printk(KERN_DEBUG "\tbudg_data_growth %lld, budg_dd_growth %lld, " + "budg_idx_growth %lld\n", bi->data_growth, bi->dd_growth, + bi->idx_growth); + printk(KERN_DEBUG "\tmin_idx_lebs %d, old_idx_sz %llu, " + "uncommitted_idx %lld\n", bi->min_idx_lebs, bi->old_idx_sz, + bi->uncommitted_idx); + printk(KERN_DEBUG "\tpage_budget %d, inode_budget %d, dent_budget %d\n", + bi->page_budget, bi->inode_budget, bi->dent_budget); + printk(KERN_DEBUG "\tnospace %u, nospace_rp %u\n", + bi->nospace, bi->nospace_rp); + printk(KERN_DEBUG "\tdark_wm %d, dead_wm %d, max_idx_node_sz %d\n", + c->dark_wm, c->dead_wm, c->max_idx_node_sz); + + if (bi != &c->bi) + /* + * If we are dumping saved budgeting data, do not print + * additional information which is about the current state, not + * the old one which corresponded to the saved budgeting data. + */ + goto out_unlock; + + printk(KERN_DEBUG "\tfreeable_cnt %d, calc_idx_sz %lld, idx_gc_cnt %d\n", + c->freeable_cnt, c->calc_idx_sz, c->idx_gc_cnt); printk(KERN_DEBUG "\tdirty_pg_cnt %ld, dirty_zn_cnt %ld, " "clean_zn_cnt %ld\n", atomic_long_read(&c->dirty_pg_cnt), atomic_long_read(&c->dirty_zn_cnt), atomic_long_read(&c->clean_zn_cnt)); - printk(KERN_DEBUG "\tdark_wm %d, dead_wm %d, max_idx_node_sz %d\n", - c->dark_wm, c->dead_wm, c->max_idx_node_sz); printk(KERN_DEBUG "\tgc_lnum %d, ihead_lnum %d\n", c->gc_lnum, c->ihead_lnum); + /* If we are in R/O mode, journal heads do not exist */ if (c->jheads) for (i = 0; i < c->jhead_cnt; i++) @@ -648,13 +662,15 @@ void dbg_dump_budg(struct ubifs_info *c) printk(KERN_DEBUG "\tcommit state %d\n", c->cmt_state); /* Print budgeting predictions */ - available = ubifs_calc_available(c, c->min_idx_lebs); - outstanding = c->budg_data_growth + c->budg_dd_growth; + available = ubifs_calc_available(c, c->bi.min_idx_lebs); + outstanding = c->bi.data_growth + c->bi.dd_growth; free = ubifs_get_free_space_nolock(c); printk(KERN_DEBUG "Budgeting predictions:\n"); printk(KERN_DEBUG "\tavailable: %lld, outstanding %lld, free %lld\n", available, outstanding, free); +out_unlock: spin_unlock(&dbg_lock); + spin_unlock(&c->space_lock); } void dbg_dump_lprop(const struct ubifs_info *c, const struct ubifs_lprops *lp) @@ -729,7 +745,13 @@ void dbg_dump_lprop(const struct ubifs_info *c, const struct ubifs_lprops *lp) if (bud->lnum == lp->lnum) { int head = 0; for (i = 0; i < c->jhead_cnt; i++) { - if (lp->lnum == c->jheads[i].wbuf.lnum) { + /* + * Note, if we are in R/O mode or in the middle + * of mounting/re-mounting, the write-buffers do + * not exist. + */ + if (c->jheads && + lp->lnum == c->jheads[i].wbuf.lnum) { printk(KERN_CONT ", jhead %s", dbg_jhead(i)); head = 1; @@ -972,11 +994,41 @@ void dbg_dump_index(struct ubifs_info *c) void dbg_save_space_info(struct ubifs_info *c) { struct ubifs_debug_info *d = c->dbg; - - ubifs_get_lp_stats(c, &d->saved_lst); + int freeable_cnt; spin_lock(&c->space_lock); + memcpy(&d->saved_lst, &c->lst, sizeof(struct ubifs_lp_stats)); + memcpy(&d->saved_bi, &c->bi, sizeof(struct ubifs_budg_info)); + d->saved_idx_gc_cnt = c->idx_gc_cnt; + + /* + * We use a dirty hack here and zero out @c->freeable_cnt, because it + * affects the free space calculations, and UBIFS might not know about + * all freeable eraseblocks. Indeed, we know about freeable eraseblocks + * only when we read their lprops, and we do this only lazily, upon the + * need. So at any given point of time @c->freeable_cnt might be not + * exactly accurate. + * + * Just one example about the issue we hit when we did not zero + * @c->freeable_cnt. + * 1. The file-system is mounted R/O, c->freeable_cnt is %0. We save the + * amount of free space in @d->saved_free + * 2. We re-mount R/W, which makes UBIFS to read the "lsave" + * information from flash, where we cache LEBs from various + * categories ('ubifs_remount_fs()' -> 'ubifs_lpt_init()' + * -> 'lpt_init_wr()' -> 'read_lsave()' -> 'ubifs_lpt_lookup()' + * -> 'ubifs_get_pnode()' -> 'update_cats()' + * -> 'ubifs_add_to_cat()'). + * 3. Lsave contains a freeable eraseblock, and @c->freeable_cnt + * becomes %1. + * 4. We calculate the amount of free space when the re-mount is + * finished in 'dbg_check_space_info()' and it does not match + * @d->saved_free. + */ + freeable_cnt = c->freeable_cnt; + c->freeable_cnt = 0; d->saved_free = ubifs_get_free_space_nolock(c); + c->freeable_cnt = freeable_cnt; spin_unlock(&c->space_lock); } @@ -993,12 +1045,15 @@ int dbg_check_space_info(struct ubifs_info *c) { struct ubifs_debug_info *d = c->dbg; struct ubifs_lp_stats lst; - long long avail, free; + long long free; + int freeable_cnt; spin_lock(&c->space_lock); - avail = ubifs_calc_available(c, c->min_idx_lebs); + freeable_cnt = c->freeable_cnt; + c->freeable_cnt = 0; + free = ubifs_get_free_space_nolock(c); + c->freeable_cnt = freeable_cnt; spin_unlock(&c->space_lock); - free = ubifs_get_free_space(c); if (free != d->saved_free) { ubifs_err("free space changed from %lld to %lld", @@ -1011,14 +1066,14 @@ int dbg_check_space_info(struct ubifs_info *c) out: ubifs_msg("saved lprops statistics dump"); dbg_dump_lstats(&d->saved_lst); - ubifs_get_lp_stats(c, &lst); - + ubifs_msg("saved budgeting info dump"); + dbg_dump_budg(c, &d->saved_bi); + ubifs_msg("saved idx_gc_cnt %d", d->saved_idx_gc_cnt); ubifs_msg("current lprops statistics dump"); + ubifs_get_lp_stats(c, &lst); dbg_dump_lstats(&lst); - - spin_lock(&c->space_lock); - dbg_dump_budg(c); - spin_unlock(&c->space_lock); + ubifs_msg("current budgeting info dump"); + dbg_dump_budg(c, &c->bi); dump_stack(); return -EINVAL; } @@ -1762,6 +1817,8 @@ static struct fsck_inode *add_inode(struct ubifs_info *c, struct rb_node **p, *parent = NULL; struct fsck_inode *fscki; ino_t inum = key_inum_flash(c, &ino->key); + struct inode *inode; + struct ubifs_inode *ui; p = &fsckd->inodes.rb_node; while (*p) { @@ -1785,19 +1842,46 @@ static struct fsck_inode *add_inode(struct ubifs_info *c, if (!fscki) return ERR_PTR(-ENOMEM); + inode = ilookup(c->vfs_sb, inum); + fscki->inum = inum; - fscki->nlink = le32_to_cpu(ino->nlink); - fscki->size = le64_to_cpu(ino->size); - fscki->xattr_cnt = le32_to_cpu(ino->xattr_cnt); - fscki->xattr_sz = le32_to_cpu(ino->xattr_size); - fscki->xattr_nms = le32_to_cpu(ino->xattr_names); - fscki->mode = le32_to_cpu(ino->mode); + /* + * If the inode is present in the VFS inode cache, use it instead of + * the on-flash inode which might be out-of-date. E.g., the size might + * be out-of-date. If we do not do this, the following may happen, for + * example: + * 1. A power cut happens + * 2. We mount the file-system R/O, the replay process fixes up the + * inode size in the VFS cache, but on on-flash. + * 3. 'check_leaf()' fails because it hits a data node beyond inode + * size. + */ + if (!inode) { + fscki->nlink = le32_to_cpu(ino->nlink); + fscki->size = le64_to_cpu(ino->size); + fscki->xattr_cnt = le32_to_cpu(ino->xattr_cnt); + fscki->xattr_sz = le32_to_cpu(ino->xattr_size); + fscki->xattr_nms = le32_to_cpu(ino->xattr_names); + fscki->mode = le32_to_cpu(ino->mode); + } else { + ui = ubifs_inode(inode); + fscki->nlink = inode->i_nlink; + fscki->size = inode->i_size; + fscki->xattr_cnt = ui->xattr_cnt; + fscki->xattr_sz = ui->xattr_size; + fscki->xattr_nms = ui->xattr_names; + fscki->mode = inode->i_mode; + iput(inode); + } + if (S_ISDIR(fscki->mode)) { fscki->calc_sz = UBIFS_INO_NODE_SZ; fscki->calc_cnt = 2; } + rb_link_node(&fscki->rb, parent, p); rb_insert_color(&fscki->rb, &fsckd->inodes); + return fscki; } @@ -2390,7 +2474,8 @@ int dbg_check_nondata_nodes_order(struct ubifs_info *c, struct list_head *head) hashb = key_block(c, &sb->key); if (hasha > hashb) { - ubifs_err("larger hash %u goes before %u", hasha, hashb); + ubifs_err("larger hash %u goes before %u", + hasha, hashb); goto error_dump; } } @@ -2406,14 +2491,12 @@ error_dump: return 0; } -static int invocation_cnt; - int dbg_force_in_the_gaps(void) { - if (!dbg_force_in_the_gaps_enabled) + if (!(ubifs_chk_flags & UBIFS_CHK_GEN)) return 0; - /* Force in-the-gaps every 8th commit */ - return !((invocation_cnt++) & 0x7); + + return !(random32() & 7); } /* Failure mode for recovery testing */ @@ -2601,7 +2684,7 @@ int dbg_leb_read(struct ubi_volume_desc *desc, int lnum, char *buf, int offset, int len, int check) { if (in_failure_mode(desc)) - return -EIO; + return -EROFS; return ubi_leb_read(desc, lnum, buf, offset, len, check); } @@ -2611,7 +2694,7 @@ int dbg_leb_write(struct ubi_volume_desc *desc, int lnum, const void *buf, int err, failing; if (in_failure_mode(desc)) - return -EIO; + return -EROFS; failing = do_fail(desc, lnum, 1); if (failing) cut_data(buf, len); @@ -2619,7 +2702,7 @@ int dbg_leb_write(struct ubi_volume_desc *desc, int lnum, const void *buf, if (err) return err; if (failing) - return -EIO; + return -EROFS; return 0; } @@ -2629,12 +2712,12 @@ int dbg_leb_change(struct ubi_volume_desc *desc, int lnum, const void *buf, int err; if (do_fail(desc, lnum, 1)) - return -EIO; + return -EROFS; err = ubi_leb_change(desc, lnum, buf, len, dtype); if (err) return err; if (do_fail(desc, lnum, 1)) - return -EIO; + return -EROFS; return 0; } @@ -2643,12 +2726,12 @@ int dbg_leb_erase(struct ubi_volume_desc *desc, int lnum) int err; if (do_fail(desc, lnum, 0)) - return -EIO; + return -EROFS; err = ubi_leb_erase(desc, lnum); if (err) return err; if (do_fail(desc, lnum, 0)) - return -EIO; + return -EROFS; return 0; } @@ -2657,19 +2740,19 @@ int dbg_leb_unmap(struct ubi_volume_desc *desc, int lnum) int err; if (do_fail(desc, lnum, 0)) - return -EIO; + return -EROFS; err = ubi_leb_unmap(desc, lnum); if (err) return err; if (do_fail(desc, lnum, 0)) - return -EIO; + return -EROFS; return 0; } int dbg_is_mapped(struct ubi_volume_desc *desc, int lnum) { if (in_failure_mode(desc)) - return -EIO; + return -EROFS; return ubi_is_mapped(desc, lnum); } @@ -2678,12 +2761,12 @@ int dbg_leb_map(struct ubi_volume_desc *desc, int lnum, int dtype) int err; if (do_fail(desc, lnum, 0)) - return -EIO; + return -EROFS; err = ubi_leb_map(desc, lnum, dtype); if (err) return err; if (do_fail(desc, lnum, 0)) - return -EIO; + return -EROFS; return 0; } @@ -2753,7 +2836,7 @@ void dbg_debugfs_exit(void) static int open_debugfs_file(struct inode *inode, struct file *file) { file->private_data = inode->i_private; - return 0; + return nonseekable_open(inode, file); } static ssize_t write_debugfs_file(struct file *file, const char __user *buf, @@ -2764,18 +2847,15 @@ static ssize_t write_debugfs_file(struct file *file, const char __user *buf, if (file->f_path.dentry == d->dfs_dump_lprops) dbg_dump_lprops(c); - else if (file->f_path.dentry == d->dfs_dump_budg) { - spin_lock(&c->space_lock); - dbg_dump_budg(c); - spin_unlock(&c->space_lock); - } else if (file->f_path.dentry == d->dfs_dump_tnc) { + else if (file->f_path.dentry == d->dfs_dump_budg) + dbg_dump_budg(c, &c->bi); + else if (file->f_path.dentry == d->dfs_dump_tnc) { mutex_lock(&c->tnc_mutex); dbg_dump_tnc(c); mutex_unlock(&c->tnc_mutex); } else return -EINVAL; - *ppos += count; return count; } @@ -2783,7 +2863,7 @@ static const struct file_operations dfs_fops = { .open = open_debugfs_file, .write = write_debugfs_file, .owner = THIS_MODULE, - .llseek = default_llseek, + .llseek = no_llseek, }; /** @@ -2806,40 +2886,38 @@ int dbg_debugfs_init_fs(struct ubifs_info *c) struct ubifs_debug_info *d = c->dbg; sprintf(d->dfs_dir_name, "ubi%d_%d", c->vi.ubi_num, c->vi.vol_id); - d->dfs_dir = debugfs_create_dir(d->dfs_dir_name, dfs_rootdir); - if (IS_ERR(d->dfs_dir)) { - err = PTR_ERR(d->dfs_dir); - ubifs_err("cannot create \"%s\" debugfs directory, error %d\n", - d->dfs_dir_name, err); + fname = d->dfs_dir_name; + dent = debugfs_create_dir(fname, dfs_rootdir); + if (IS_ERR_OR_NULL(dent)) goto out; - } + d->dfs_dir = dent; fname = "dump_lprops"; dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops); - if (IS_ERR(dent)) + if (IS_ERR_OR_NULL(dent)) goto out_remove; d->dfs_dump_lprops = dent; fname = "dump_budg"; dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops); - if (IS_ERR(dent)) + if (IS_ERR_OR_NULL(dent)) goto out_remove; d->dfs_dump_budg = dent; fname = "dump_tnc"; dent = debugfs_create_file(fname, S_IWUSR, d->dfs_dir, c, &dfs_fops); - if (IS_ERR(dent)) + if (IS_ERR_OR_NULL(dent)) goto out_remove; d->dfs_dump_tnc = dent; return 0; out_remove: - err = PTR_ERR(dent); - ubifs_err("cannot create \"%s\" debugfs directory, error %d\n", - fname, err); debugfs_remove_recursive(d->dfs_dir); out: + err = dent ? PTR_ERR(dent) : -ENODEV; + ubifs_err("cannot create \"%s\" debugfs directory, error %d\n", + fname, err); return err; } diff --git a/fs/ubifs/debug.h b/fs/ubifs/debug.h index 919f0de29d8f..a811ac4a26bb 100644 --- a/fs/ubifs/debug.h +++ b/fs/ubifs/debug.h @@ -23,8 +23,16 @@ #ifndef __UBIFS_DEBUG_H__ #define __UBIFS_DEBUG_H__ +/* Checking helper functions */ +typedef int (*dbg_leaf_callback)(struct ubifs_info *c, + struct ubifs_zbranch *zbr, void *priv); +typedef int (*dbg_znode_callback)(struct ubifs_info *c, + struct ubifs_znode *znode, void *priv); + #ifdef CONFIG_UBIFS_FS_DEBUG +#include <linux/random.h> + /** * ubifs_debug_info - per-FS debugging information. * @old_zroot: old index root - used by 'dbg_check_old_index()' @@ -44,13 +52,15 @@ * @new_ihead_offs: used by debugging to check @c->ihead_offs * * @saved_lst: saved lprops statistics (used by 'dbg_save_space_info()') - * @saved_free: saved free space (used by 'dbg_save_space_info()') + * @saved_bi: saved budgeting information + * @saved_free: saved amount of free space + * @saved_idx_gc_cnt: saved value of @c->idx_gc_cnt * - * dfs_dir_name: name of debugfs directory containing this file-system's files - * dfs_dir: direntry object of the file-system debugfs directory - * dfs_dump_lprops: "dump lprops" debugfs knob - * dfs_dump_budg: "dump budgeting information" debugfs knob - * dfs_dump_tnc: "dump TNC" debugfs knob + * @dfs_dir_name: name of debugfs directory containing this file-system's files + * @dfs_dir: direntry object of the file-system debugfs directory + * @dfs_dump_lprops: "dump lprops" debugfs knob + * @dfs_dump_budg: "dump budgeting information" debugfs knob + * @dfs_dump_tnc: "dump TNC" debugfs knob */ struct ubifs_debug_info { struct ubifs_zbranch old_zroot; @@ -70,7 +80,9 @@ struct ubifs_debug_info { int new_ihead_offs; struct ubifs_lp_stats saved_lst; + struct ubifs_budg_info saved_bi; long long saved_free; + int saved_idx_gc_cnt; char dfs_dir_name[100]; struct dentry *dfs_dir; @@ -95,23 +107,7 @@ struct ubifs_debug_info { } \ } while (0) -#define dbg_dump_stack() do { \ - if (!dbg_failure_mode) \ - dump_stack(); \ -} while (0) - -/* Generic debugging messages */ -#define dbg_msg(fmt, ...) do { \ - spin_lock(&dbg_lock); \ - printk(KERN_DEBUG "UBIFS DBG (pid %d): %s: " fmt "\n", current->pid, \ - __func__, ##__VA_ARGS__); \ - spin_unlock(&dbg_lock); \ -} while (0) - -#define dbg_do_msg(typ, fmt, ...) do { \ - if (ubifs_msg_flags & typ) \ - dbg_msg(fmt, ##__VA_ARGS__); \ -} while (0) +#define dbg_dump_stack() dump_stack() #define dbg_err(fmt, ...) do { \ spin_lock(&dbg_lock); \ @@ -131,77 +127,40 @@ const char *dbg_key_str1(const struct ubifs_info *c, #define DBGKEY(key) dbg_key_str0(c, (key)) #define DBGKEY1(key) dbg_key_str1(c, (key)) -/* General messages */ -#define dbg_gen(fmt, ...) dbg_do_msg(UBIFS_MSG_GEN, fmt, ##__VA_ARGS__) +#define ubifs_dbg_msg(type, fmt, ...) do { \ + spin_lock(&dbg_lock); \ + pr_debug("UBIFS DBG " type ": " fmt "\n", ##__VA_ARGS__); \ + spin_unlock(&dbg_lock); \ +} while (0) +/* Just a debugging messages not related to any specific UBIFS subsystem */ +#define dbg_msg(fmt, ...) ubifs_dbg_msg("msg", fmt, ##__VA_ARGS__) +/* General messages */ +#define dbg_gen(fmt, ...) ubifs_dbg_msg("gen", fmt, ##__VA_ARGS__) /* Additional journal messages */ -#define dbg_jnl(fmt, ...) dbg_do_msg(UBIFS_MSG_JNL, fmt, ##__VA_ARGS__) - +#define dbg_jnl(fmt, ...) ubifs_dbg_msg("jnl", fmt, ##__VA_ARGS__) /* Additional TNC messages */ -#define dbg_tnc(fmt, ...) dbg_do_msg(UBIFS_MSG_TNC, fmt, ##__VA_ARGS__) - +#define dbg_tnc(fmt, ...) ubifs_dbg_msg("tnc", fmt, ##__VA_ARGS__) /* Additional lprops messages */ -#define dbg_lp(fmt, ...) dbg_do_msg(UBIFS_MSG_LP, fmt, ##__VA_ARGS__) - +#define dbg_lp(fmt, ...) ubifs_dbg_msg("lp", fmt, ##__VA_ARGS__) /* Additional LEB find messages */ -#define dbg_find(fmt, ...) dbg_do_msg(UBIFS_MSG_FIND, fmt, ##__VA_ARGS__) - +#define dbg_find(fmt, ...) ubifs_dbg_msg("find", fmt, ##__VA_ARGS__) /* Additional mount messages */ -#define dbg_mnt(fmt, ...) dbg_do_msg(UBIFS_MSG_MNT, fmt, ##__VA_ARGS__) - +#define dbg_mnt(fmt, ...) ubifs_dbg_msg("mnt", fmt, ##__VA_ARGS__) /* Additional I/O messages */ -#define dbg_io(fmt, ...) dbg_do_msg(UBIFS_MSG_IO, fmt, ##__VA_ARGS__) - +#define dbg_io(fmt, ...) ubifs_dbg_msg("io", fmt, ##__VA_ARGS__) /* Additional commit messages */ -#define dbg_cmt(fmt, ...) dbg_do_msg(UBIFS_MSG_CMT, fmt, ##__VA_ARGS__) - +#define dbg_cmt(fmt, ...) ubifs_dbg_msg("cmt", fmt, ##__VA_ARGS__) /* Additional budgeting messages */ -#define dbg_budg(fmt, ...) dbg_do_msg(UBIFS_MSG_BUDG, fmt, ##__VA_ARGS__) - +#define dbg_budg(fmt, ...) ubifs_dbg_msg("budg", fmt, ##__VA_ARGS__) /* Additional log messages */ -#define dbg_log(fmt, ...) dbg_do_msg(UBIFS_MSG_LOG, fmt, ##__VA_ARGS__) - +#define dbg_log(fmt, ...) ubifs_dbg_msg("log", fmt, ##__VA_ARGS__) /* Additional gc messages */ -#define dbg_gc(fmt, ...) dbg_do_msg(UBIFS_MSG_GC, fmt, ##__VA_ARGS__) - +#define dbg_gc(fmt, ...) ubifs_dbg_msg("gc", fmt, ##__VA_ARGS__) /* Additional scan messages */ -#define dbg_scan(fmt, ...) dbg_do_msg(UBIFS_MSG_SCAN, fmt, ##__VA_ARGS__) - +#define dbg_scan(fmt, ...) ubifs_dbg_msg("scan", fmt, ##__VA_ARGS__) /* Additional recovery messages */ -#define dbg_rcvry(fmt, ...) dbg_do_msg(UBIFS_MSG_RCVRY, fmt, ##__VA_ARGS__) - -/* - * Debugging message type flags. - * - * UBIFS_MSG_GEN: general messages - * UBIFS_MSG_JNL: journal messages - * UBIFS_MSG_MNT: mount messages - * UBIFS_MSG_CMT: commit messages - * UBIFS_MSG_FIND: LEB find messages - * UBIFS_MSG_BUDG: budgeting messages - * UBIFS_MSG_GC: garbage collection messages - * UBIFS_MSG_TNC: TNC messages - * UBIFS_MSG_LP: lprops messages - * UBIFS_MSG_IO: I/O messages - * UBIFS_MSG_LOG: log messages - * UBIFS_MSG_SCAN: scan messages - * UBIFS_MSG_RCVRY: recovery messages - */ -enum { - UBIFS_MSG_GEN = 0x1, - UBIFS_MSG_JNL = 0x2, - UBIFS_MSG_MNT = 0x4, - UBIFS_MSG_CMT = 0x8, - UBIFS_MSG_FIND = 0x10, - UBIFS_MSG_BUDG = 0x20, - UBIFS_MSG_GC = 0x40, - UBIFS_MSG_TNC = 0x80, - UBIFS_MSG_LP = 0x100, - UBIFS_MSG_IO = 0x200, - UBIFS_MSG_LOG = 0x400, - UBIFS_MSG_SCAN = 0x800, - UBIFS_MSG_RCVRY = 0x1000, -}; +#define dbg_rcvry(fmt, ...) ubifs_dbg_msg("rcvry", fmt, ##__VA_ARGS__) /* * Debugging check flags. @@ -227,11 +186,9 @@ enum { /* * Special testing flags. * - * UBIFS_TST_FORCE_IN_THE_GAPS: force the use of in-the-gaps method * UBIFS_TST_RCVRY: failure mode for recovery testing */ enum { - UBIFS_TST_FORCE_IN_THE_GAPS = 0x2, UBIFS_TST_RCVRY = 0x4, }; @@ -256,7 +213,7 @@ void dbg_dump_lpt_node(const struct ubifs_info *c, void *node, int lnum, int offs); void dbg_dump_budget_req(const struct ubifs_budget_req *req); void dbg_dump_lstats(const struct ubifs_lp_stats *lst); -void dbg_dump_budg(struct ubifs_info *c); +void dbg_dump_budg(struct ubifs_info *c, const struct ubifs_budg_info *bi); void dbg_dump_lprop(const struct ubifs_info *c, const struct ubifs_lprops *lp); void dbg_dump_lprops(struct ubifs_info *c); void dbg_dump_lpt_info(struct ubifs_info *c); @@ -270,11 +227,6 @@ void dbg_dump_tnc(struct ubifs_info *c); void dbg_dump_index(struct ubifs_info *c); void dbg_dump_lpt_lebs(const struct ubifs_info *c); -/* Checking helper functions */ -typedef int (*dbg_leaf_callback)(struct ubifs_info *c, - struct ubifs_zbranch *zbr, void *priv); -typedef int (*dbg_znode_callback)(struct ubifs_info *c, - struct ubifs_znode *znode, void *priv); int dbg_walk_index(struct ubifs_info *c, dbg_leaf_callback leaf_cb, dbg_znode_callback znode_cb, void *priv); @@ -295,7 +247,6 @@ int dbg_check_idx_size(struct ubifs_info *c, long long idx_size); int dbg_check_filesystem(struct ubifs_info *c); void dbg_check_heap(struct ubifs_info *c, struct ubifs_lpt_heap *heap, int cat, int add_pos); -int dbg_check_lprops(struct ubifs_info *c); int dbg_check_lpt_nodes(struct ubifs_info *c, struct ubifs_cnode *cnode, int row, int col); int dbg_check_inode_size(struct ubifs_info *c, const struct inode *inode, @@ -304,18 +255,16 @@ int dbg_check_data_nodes_order(struct ubifs_info *c, struct list_head *head); int dbg_check_nondata_nodes_order(struct ubifs_info *c, struct list_head *head); /* Force the use of in-the-gaps method for testing */ - -#define dbg_force_in_the_gaps_enabled \ - (ubifs_tst_flags & UBIFS_TST_FORCE_IN_THE_GAPS) - +static inline int dbg_force_in_the_gaps_enabled(void) +{ + return ubifs_chk_flags & UBIFS_CHK_GEN; +} int dbg_force_in_the_gaps(void); /* Failure mode for recovery testing */ - #define dbg_failure_mode (ubifs_tst_flags & UBIFS_TST_RCVRY) #ifndef UBIFS_DBG_PRESERVE_UBI - #define ubi_leb_read dbg_leb_read #define ubi_leb_write dbg_leb_write #define ubi_leb_change dbg_leb_change @@ -323,7 +272,6 @@ int dbg_force_in_the_gaps(void); #define ubi_leb_unmap dbg_leb_unmap #define ubi_is_mapped dbg_is_mapped #define ubi_leb_map dbg_leb_map - #endif int dbg_leb_read(struct ubi_volume_desc *desc, int lnum, char *buf, int offset, @@ -370,89 +318,127 @@ void dbg_debugfs_exit_fs(struct ubifs_info *c); __func__, __LINE__, current->pid); \ } while (0) -#define dbg_err(fmt, ...) do { \ - if (0) \ - ubifs_err(fmt, ##__VA_ARGS__); \ +#define dbg_err(fmt, ...) do { \ + if (0) \ + ubifs_err(fmt, ##__VA_ARGS__); \ } while (0) -#define dbg_msg(fmt, ...) do { \ - if (0) \ - printk(KERN_DEBUG "UBIFS DBG (pid %d): %s: " fmt "\n", \ - current->pid, __func__, ##__VA_ARGS__); \ +#define ubifs_dbg_msg(fmt, ...) do { \ + if (0) \ + pr_debug(fmt "\n", ##__VA_ARGS__); \ } while (0) #define dbg_dump_stack() #define ubifs_assert_cmt_locked(c) -#define dbg_gen(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_jnl(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_tnc(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_lp(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_find(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_mnt(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_io(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_cmt(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_budg(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_log(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_gc(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_scan(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) -#define dbg_rcvry(fmt, ...) dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_msg(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_gen(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_jnl(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_tnc(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_lp(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_find(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_mnt(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_io(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_cmt(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_budg(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_log(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_gc(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_scan(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) +#define dbg_rcvry(fmt, ...) ubifs_dbg_msg(fmt, ##__VA_ARGS__) #define DBGKEY(key) ((char *)(key)) #define DBGKEY1(key) ((char *)(key)) -#define ubifs_debugging_init(c) 0 -#define ubifs_debugging_exit(c) ({}) - -#define dbg_ntype(type) "" -#define dbg_cstate(cmt_state) "" -#define dbg_jhead(jhead) "" -#define dbg_get_key_dump(c, key) ({}) -#define dbg_dump_inode(c, inode) ({}) -#define dbg_dump_node(c, node) ({}) -#define dbg_dump_lpt_node(c, node, lnum, offs) ({}) -#define dbg_dump_budget_req(req) ({}) -#define dbg_dump_lstats(lst) ({}) -#define dbg_dump_budg(c) ({}) -#define dbg_dump_lprop(c, lp) ({}) -#define dbg_dump_lprops(c) ({}) -#define dbg_dump_lpt_info(c) ({}) -#define dbg_dump_leb(c, lnum) ({}) -#define dbg_dump_znode(c, znode) ({}) -#define dbg_dump_heap(c, heap, cat) ({}) -#define dbg_dump_pnode(c, pnode, parent, iip) ({}) -#define dbg_dump_tnc(c) ({}) -#define dbg_dump_index(c) ({}) -#define dbg_dump_lpt_lebs(c) ({}) - -#define dbg_walk_index(c, leaf_cb, znode_cb, priv) 0 -#define dbg_old_index_check_init(c, zroot) 0 -#define dbg_save_space_info(c) ({}) -#define dbg_check_space_info(c) 0 -#define dbg_check_old_index(c, zroot) 0 -#define dbg_check_cats(c) 0 -#define dbg_check_ltab(c) 0 -#define dbg_chk_lpt_free_spc(c) 0 -#define dbg_chk_lpt_sz(c, action, len) 0 -#define dbg_check_synced_i_size(inode) 0 -#define dbg_check_dir_size(c, dir) 0 -#define dbg_check_tnc(c, x) 0 -#define dbg_check_idx_size(c, idx_size) 0 -#define dbg_check_filesystem(c) 0 -#define dbg_check_heap(c, heap, cat, add_pos) ({}) -#define dbg_check_lprops(c) 0 -#define dbg_check_lpt_nodes(c, cnode, row, col) 0 -#define dbg_check_inode_size(c, inode, size) 0 -#define dbg_check_data_nodes_order(c, head) 0 -#define dbg_check_nondata_nodes_order(c, head) 0 -#define dbg_force_in_the_gaps_enabled 0 -#define dbg_force_in_the_gaps() 0 -#define dbg_failure_mode 0 - -#define dbg_debugfs_init() 0 -#define dbg_debugfs_exit() -#define dbg_debugfs_init_fs(c) 0 -#define dbg_debugfs_exit_fs(c) 0 +static inline int ubifs_debugging_init(struct ubifs_info *c) { return 0; } +static inline void ubifs_debugging_exit(struct ubifs_info *c) { return; } +static inline const char *dbg_ntype(int type) { return ""; } +static inline const char *dbg_cstate(int cmt_state) { return ""; } +static inline const char *dbg_jhead(int jhead) { return ""; } +static inline const char * +dbg_get_key_dump(const struct ubifs_info *c, + const union ubifs_key *key) { return ""; } +static inline void dbg_dump_inode(const struct ubifs_info *c, + const struct inode *inode) { return; } +static inline void dbg_dump_node(const struct ubifs_info *c, + const void *node) { return; } +static inline void dbg_dump_lpt_node(const struct ubifs_info *c, + void *node, int lnum, + int offs) { return; } +static inline void +dbg_dump_budget_req(const struct ubifs_budget_req *req) { return; } +static inline void +dbg_dump_lstats(const struct ubifs_lp_stats *lst) { return; } +static inline void +dbg_dump_budg(struct ubifs_info *c, + const struct ubifs_budg_info *bi) { return; } +static inline void dbg_dump_lprop(const struct ubifs_info *c, + const struct ubifs_lprops *lp) { return; } +static inline void dbg_dump_lprops(struct ubifs_info *c) { return; } +static inline void dbg_dump_lpt_info(struct ubifs_info *c) { return; } +static inline void dbg_dump_leb(const struct ubifs_info *c, + int lnum) { return; } +static inline void +dbg_dump_znode(const struct ubifs_info *c, + const struct ubifs_znode *znode) { return; } +static inline void dbg_dump_heap(struct ubifs_info *c, + struct ubifs_lpt_heap *heap, + int cat) { return; } +static inline void dbg_dump_pnode(struct ubifs_info *c, + struct ubifs_pnode *pnode, + struct ubifs_nnode *parent, + int iip) { return; } +static inline void dbg_dump_tnc(struct ubifs_info *c) { return; } +static inline void dbg_dump_index(struct ubifs_info *c) { return; } +static inline void dbg_dump_lpt_lebs(const struct ubifs_info *c) { return; } + +static inline int dbg_walk_index(struct ubifs_info *c, + dbg_leaf_callback leaf_cb, + dbg_znode_callback znode_cb, + void *priv) { return 0; } +static inline void dbg_save_space_info(struct ubifs_info *c) { return; } +static inline int dbg_check_space_info(struct ubifs_info *c) { return 0; } +static inline int dbg_check_lprops(struct ubifs_info *c) { return 0; } +static inline int +dbg_old_index_check_init(struct ubifs_info *c, + struct ubifs_zbranch *zroot) { return 0; } +static inline int +dbg_check_old_index(struct ubifs_info *c, + struct ubifs_zbranch *zroot) { return 0; } +static inline int dbg_check_cats(struct ubifs_info *c) { return 0; } +static inline int dbg_check_ltab(struct ubifs_info *c) { return 0; } +static inline int dbg_chk_lpt_free_spc(struct ubifs_info *c) { return 0; } +static inline int dbg_chk_lpt_sz(struct ubifs_info *c, + int action, int len) { return 0; } +static inline int dbg_check_synced_i_size(struct inode *inode) { return 0; } +static inline int dbg_check_dir_size(struct ubifs_info *c, + const struct inode *dir) { return 0; } +static inline int dbg_check_tnc(struct ubifs_info *c, int extra) { return 0; } +static inline int dbg_check_idx_size(struct ubifs_info *c, + long long idx_size) { return 0; } +static inline int dbg_check_filesystem(struct ubifs_info *c) { return 0; } +static inline void dbg_check_heap(struct ubifs_info *c, + struct ubifs_lpt_heap *heap, + int cat, int add_pos) { return; } +static inline int dbg_check_lpt_nodes(struct ubifs_info *c, + struct ubifs_cnode *cnode, int row, int col) { return 0; } +static inline int dbg_check_inode_size(struct ubifs_info *c, + const struct inode *inode, + loff_t size) { return 0; } +static inline int +dbg_check_data_nodes_order(struct ubifs_info *c, + struct list_head *head) { return 0; } +static inline int +dbg_check_nondata_nodes_order(struct ubifs_info *c, + struct list_head *head) { return 0; } + +static inline int dbg_force_in_the_gaps(void) { return 0; } +#define dbg_force_in_the_gaps_enabled() 0 +#define dbg_failure_mode 0 + +static inline int dbg_debugfs_init(void) { return 0; } +static inline void dbg_debugfs_exit(void) { return; } +static inline int dbg_debugfs_init_fs(struct ubifs_info *c) { return 0; } +static inline int dbg_debugfs_exit_fs(struct ubifs_info *c) { return 0; } #endif /* !CONFIG_UBIFS_FS_DEBUG */ #endif /* !__UBIFS_DEBUG_H__ */ diff --git a/fs/ubifs/dir.c b/fs/ubifs/dir.c index 7217d67a80a6..ef5abd38f0bf 100644 --- a/fs/ubifs/dir.c +++ b/fs/ubifs/dir.c @@ -603,7 +603,7 @@ static int ubifs_unlink(struct inode *dir, struct dentry *dentry) ubifs_release_budget(c, &req); else { /* We've deleted something - clean the "no space" flags */ - c->nospace = c->nospace_rp = 0; + c->bi.nospace = c->bi.nospace_rp = 0; smp_wmb(); } return 0; @@ -693,7 +693,7 @@ static int ubifs_rmdir(struct inode *dir, struct dentry *dentry) ubifs_release_budget(c, &req); else { /* We've deleted something - clean the "no space" flags */ - c->nospace = c->nospace_rp = 0; + c->bi.nospace = c->bi.nospace_rp = 0; smp_wmb(); } return 0; diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index 28be1e6a65e8..5e7fccfc4b29 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -212,7 +212,7 @@ static void release_new_page_budget(struct ubifs_info *c) */ static void release_existing_page_budget(struct ubifs_info *c) { - struct ubifs_budget_req req = { .dd_growth = c->page_budget}; + struct ubifs_budget_req req = { .dd_growth = c->bi.page_budget}; ubifs_release_budget(c, &req); } @@ -971,11 +971,11 @@ static int do_writepage(struct page *page, int len) * the page locked, and it locks @ui_mutex. However, write-back does take inode * @i_mutex, which means other VFS operations may be run on this inode at the * same time. And the problematic one is truncation to smaller size, from where - * we have to call 'truncate_setsize()', which first changes @inode->i_size, then - * drops the truncated pages. And while dropping the pages, it takes the page - * lock. This means that 'do_truncation()' cannot call 'truncate_setsize()' with - * @ui_mutex locked, because it would deadlock with 'ubifs_writepage()'. This - * means that @inode->i_size is changed while @ui_mutex is unlocked. + * we have to call 'truncate_setsize()', which first changes @inode->i_size, + * then drops the truncated pages. And while dropping the pages, it takes the + * page lock. This means that 'do_truncation()' cannot call 'truncate_setsize()' + * with @ui_mutex locked, because it would deadlock with 'ubifs_writepage()'. + * This means that @inode->i_size is changed while @ui_mutex is unlocked. * * XXX(truncate): with the new truncate sequence this is not true anymore, * and the calls to truncate_setsize can be move around freely. They should @@ -1189,7 +1189,7 @@ out_budg: if (budgeted) ubifs_release_budget(c, &req); else { - c->nospace = c->nospace_rp = 0; + c->bi.nospace = c->bi.nospace_rp = 0; smp_wmb(); } return err; @@ -1312,6 +1312,13 @@ int ubifs_fsync(struct file *file, int datasync) dbg_gen("syncing inode %lu", inode->i_ino); + if (c->ro_mount) + /* + * For some really strange reasons VFS does not filter out + * 'fsync()' for R/O mounted file-systems as per 2.6.39. + */ + return 0; + /* * VFS has already synchronized dirty pages for this inode. Synchronize * the inode unless this is a 'datasync()' call. @@ -1429,10 +1436,11 @@ static int ubifs_releasepage(struct page *page, gfp_t unused_gfp_flags) } /* - * mmap()d file has taken write protection fault and is being made - * writable. UBIFS must ensure page is budgeted for. + * mmap()d file has taken write protection fault and is being made writable. + * UBIFS must ensure page is budgeted for. */ -static int ubifs_vm_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) +static int ubifs_vm_page_mkwrite(struct vm_area_struct *vma, + struct vm_fault *vmf) { struct page *page = vmf->page; struct inode *inode = vma->vm_file->f_path.dentry->d_inode; @@ -1533,7 +1541,6 @@ static int ubifs_file_mmap(struct file *file, struct vm_area_struct *vma) { int err; - /* 'generic_file_mmap()' takes care of NOMMU case */ err = generic_file_mmap(file, vma); if (err) return err; diff --git a/fs/ubifs/find.c b/fs/ubifs/find.c index 1d54383d1269..2559d174e004 100644 --- a/fs/ubifs/find.c +++ b/fs/ubifs/find.c @@ -252,8 +252,8 @@ int ubifs_find_dirty_leb(struct ubifs_info *c, struct ubifs_lprops *ret_lp, * But if the index takes fewer LEBs than it is reserved for it, * this function must avoid picking those reserved LEBs. */ - if (c->min_idx_lebs >= c->lst.idx_lebs) { - rsvd_idx_lebs = c->min_idx_lebs - c->lst.idx_lebs; + if (c->bi.min_idx_lebs >= c->lst.idx_lebs) { + rsvd_idx_lebs = c->bi.min_idx_lebs - c->lst.idx_lebs; exclude_index = 1; } spin_unlock(&c->space_lock); @@ -276,7 +276,7 @@ int ubifs_find_dirty_leb(struct ubifs_info *c, struct ubifs_lprops *ret_lp, pick_free = 0; } else { spin_lock(&c->space_lock); - exclude_index = (c->min_idx_lebs >= c->lst.idx_lebs); + exclude_index = (c->bi.min_idx_lebs >= c->lst.idx_lebs); spin_unlock(&c->space_lock); } @@ -501,8 +501,8 @@ int ubifs_find_free_space(struct ubifs_info *c, int min_space, int *offs, /* Check if there are enough empty LEBs for commit */ spin_lock(&c->space_lock); - if (c->min_idx_lebs > c->lst.idx_lebs) - rsvd_idx_lebs = c->min_idx_lebs - c->lst.idx_lebs; + if (c->bi.min_idx_lebs > c->lst.idx_lebs) + rsvd_idx_lebs = c->bi.min_idx_lebs - c->lst.idx_lebs; else rsvd_idx_lebs = 0; lebs = c->lst.empty_lebs + c->freeable_cnt + c->idx_gc_cnt - diff --git a/fs/ubifs/gc.c b/fs/ubifs/gc.c index 151f10882820..ded29f6224c2 100644 --- a/fs/ubifs/gc.c +++ b/fs/ubifs/gc.c @@ -100,6 +100,10 @@ static int switch_gc_head(struct ubifs_info *c) if (err) return err; + err = ubifs_wbuf_sync_nolock(wbuf); + if (err) + return err; + err = ubifs_add_bud_to_log(c, GCHD, gc_lnum, 0); if (err) return err; @@ -118,7 +122,7 @@ static int switch_gc_head(struct ubifs_info *c) * This function compares data nodes @a and @b. Returns %1 if @a has greater * inode or block number, and %-1 otherwise. */ -int data_nodes_cmp(void *priv, struct list_head *a, struct list_head *b) +static int data_nodes_cmp(void *priv, struct list_head *a, struct list_head *b) { ino_t inuma, inumb; struct ubifs_info *c = priv; @@ -161,7 +165,8 @@ int data_nodes_cmp(void *priv, struct list_head *a, struct list_head *b) * first and sorted by length in descending order. Directory entry nodes go * after inode nodes and are sorted in ascending hash valuer order. */ -int nondata_nodes_cmp(void *priv, struct list_head *a, struct list_head *b) +static int nondata_nodes_cmp(void *priv, struct list_head *a, + struct list_head *b) { ino_t inuma, inumb; struct ubifs_info *c = priv; @@ -473,6 +478,37 @@ int ubifs_garbage_collect_leb(struct ubifs_info *c, struct ubifs_lprops *lp) ubifs_assert(c->gc_lnum != lnum); ubifs_assert(wbuf->lnum != lnum); + if (lp->free + lp->dirty == c->leb_size) { + /* Special case - a free LEB */ + dbg_gc("LEB %d is free, return it", lp->lnum); + ubifs_assert(!(lp->flags & LPROPS_INDEX)); + + if (lp->free != c->leb_size) { + /* + * Write buffers must be sync'd before unmapping + * freeable LEBs, because one of them may contain data + * which obsoletes something in 'lp->pnum'. + */ + err = gc_sync_wbufs(c); + if (err) + return err; + err = ubifs_change_one_lp(c, lp->lnum, c->leb_size, + 0, 0, 0, 0); + if (err) + return err; + } + err = ubifs_leb_unmap(c, lp->lnum); + if (err) + return err; + + if (c->gc_lnum == -1) { + c->gc_lnum = lnum; + return LEB_RETAINED; + } + + return LEB_FREED; + } + /* * We scan the entire LEB even though we only really need to scan up to * (c->leb_size - lp->free). @@ -682,37 +718,6 @@ int ubifs_garbage_collect(struct ubifs_info *c, int anyway) "(min. space %d)", lp.lnum, lp.free, lp.dirty, lp.free + lp.dirty, min_space); - if (lp.free + lp.dirty == c->leb_size) { - /* An empty LEB was returned */ - dbg_gc("LEB %d is free, return it", lp.lnum); - /* - * ubifs_find_dirty_leb() doesn't return freeable index - * LEBs. - */ - ubifs_assert(!(lp.flags & LPROPS_INDEX)); - if (lp.free != c->leb_size) { - /* - * Write buffers must be sync'd before - * unmapping freeable LEBs, because one of them - * may contain data which obsoletes something - * in 'lp.pnum'. - */ - ret = gc_sync_wbufs(c); - if (ret) - goto out; - ret = ubifs_change_one_lp(c, lp.lnum, - c->leb_size, 0, 0, 0, - 0); - if (ret) - goto out; - } - ret = ubifs_leb_unmap(c, lp.lnum); - if (ret) - goto out; - ret = lp.lnum; - break; - } - space_before = c->leb_size - wbuf->offs - wbuf->used; if (wbuf->lnum == -1) space_before = 0; diff --git a/fs/ubifs/io.c b/fs/ubifs/io.c index dfd168b7807e..166951e0dcd3 100644 --- a/fs/ubifs/io.c +++ b/fs/ubifs/io.c @@ -393,7 +393,7 @@ int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf) ubifs_assert(wbuf->size % c->min_io_size == 0); ubifs_assert(!c->ro_media && !c->ro_mount); if (c->leb_size - wbuf->offs >= c->max_write_size) - ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size )); + ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size)); if (c->ro_error) return -EROFS; @@ -452,8 +452,8 @@ int ubifs_wbuf_sync_nolock(struct ubifs_wbuf *wbuf) * @dtype: data type * * This function targets the write-buffer to logical eraseblock @lnum:@offs. - * The write-buffer is synchronized if it is not empty. Returns zero in case of - * success and a negative error code in case of failure. + * The write-buffer has to be empty. Returns zero in case of success and a + * negative error code in case of failure. */ int ubifs_wbuf_seek_nolock(struct ubifs_wbuf *wbuf, int lnum, int offs, int dtype) @@ -465,13 +465,7 @@ int ubifs_wbuf_seek_nolock(struct ubifs_wbuf *wbuf, int lnum, int offs, ubifs_assert(offs >= 0 && offs <= c->leb_size); ubifs_assert(offs % c->min_io_size == 0 && !(offs & 7)); ubifs_assert(lnum != wbuf->lnum); - - if (wbuf->used > 0) { - int err = ubifs_wbuf_sync_nolock(wbuf); - - if (err) - return err; - } + ubifs_assert(wbuf->used == 0); spin_lock(&wbuf->lock); wbuf->lnum = lnum; @@ -573,7 +567,7 @@ out_timers: int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) { struct ubifs_info *c = wbuf->c; - int err, written, n, aligned_len = ALIGN(len, 8), offs; + int err, written, n, aligned_len = ALIGN(len, 8); dbg_io("%d bytes (%s) to jhead %s wbuf at LEB %d:%d", len, dbg_ntype(((struct ubifs_ch *)buf)->node_type), @@ -588,7 +582,7 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) ubifs_assert(mutex_is_locked(&wbuf->io_mutex)); ubifs_assert(!c->ro_media && !c->ro_mount); if (c->leb_size - wbuf->offs >= c->max_write_size) - ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size )); + ubifs_assert(!((wbuf->offs + wbuf->size) % c->max_write_size)); if (c->leb_size - wbuf->offs - wbuf->used < aligned_len) { err = -ENOSPC; @@ -636,7 +630,6 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) goto exit; } - offs = wbuf->offs; written = 0; if (wbuf->used) { @@ -653,7 +646,7 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) if (err) goto out; - offs += wbuf->size; + wbuf->offs += wbuf->size; len -= wbuf->avail; aligned_len -= wbuf->avail; written += wbuf->avail; @@ -672,7 +665,7 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) if (err) goto out; - offs += wbuf->size; + wbuf->offs += wbuf->size; len -= wbuf->size; aligned_len -= wbuf->size; written += wbuf->size; @@ -687,12 +680,13 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) n = aligned_len >> c->max_write_shift; if (n) { n <<= c->max_write_shift; - dbg_io("write %d bytes to LEB %d:%d", n, wbuf->lnum, offs); - err = ubi_leb_write(c->ubi, wbuf->lnum, buf + written, offs, n, - wbuf->dtype); + dbg_io("write %d bytes to LEB %d:%d", n, wbuf->lnum, + wbuf->offs); + err = ubi_leb_write(c->ubi, wbuf->lnum, buf + written, + wbuf->offs, n, wbuf->dtype); if (err) goto out; - offs += n; + wbuf->offs += n; aligned_len -= n; len -= n; written += n; @@ -707,7 +701,6 @@ int ubifs_wbuf_write_nolock(struct ubifs_wbuf *wbuf, void *buf, int len) */ memcpy(wbuf->buf, buf + written, len); - wbuf->offs = offs; if (c->leb_size - wbuf->offs >= c->max_write_size) wbuf->size = c->max_write_size; else diff --git a/fs/ubifs/journal.c b/fs/ubifs/journal.c index aed25e864227..34b1679e6e3a 100644 --- a/fs/ubifs/journal.c +++ b/fs/ubifs/journal.c @@ -141,14 +141,8 @@ again: * LEB with some empty space. */ lnum = ubifs_find_free_space(c, len, &offs, squeeze); - if (lnum >= 0) { - /* Found an LEB, add it to the journal head */ - err = ubifs_add_bud_to_log(c, jhead, lnum, offs); - if (err) - goto out_return; - /* A new bud was successfully allocated and added to the log */ + if (lnum >= 0) goto out; - } err = lnum; if (err != -ENOSPC) @@ -203,12 +197,23 @@ again: return 0; } - err = ubifs_add_bud_to_log(c, jhead, lnum, 0); - if (err) - goto out_return; offs = 0; out: + /* + * Make sure we synchronize the write-buffer before we add the new bud + * to the log. Otherwise we may have a power cut after the log + * reference node for the last bud (@lnum) is written but before the + * write-buffer data are written to the next-to-last bud + * (@wbuf->lnum). And the effect would be that the recovery would see + * that there is corruption in the next-to-last bud. + */ + err = ubifs_wbuf_sync_nolock(wbuf); + if (err) + goto out_return; + err = ubifs_add_bud_to_log(c, jhead, lnum, offs); + if (err) + goto out_return; err = ubifs_wbuf_seek_nolock(wbuf, lnum, offs, wbuf->dtype); if (err) goto out_unlock; @@ -380,10 +385,8 @@ out: if (err == -ENOSPC) { /* This are some budgeting problems, print useful information */ down_write(&c->commit_sem); - spin_lock(&c->space_lock); dbg_dump_stack(); - dbg_dump_budg(c); - spin_unlock(&c->space_lock); + dbg_dump_budg(c, &c->bi); dbg_dump_lprops(c); cmt_retries = dbg_check_lprops(c); up_write(&c->commit_sem); diff --git a/fs/ubifs/log.c b/fs/ubifs/log.c index 4d0cb1241460..affea9494ae2 100644 --- a/fs/ubifs/log.c +++ b/fs/ubifs/log.c @@ -100,20 +100,6 @@ struct ubifs_wbuf *ubifs_get_wbuf(struct ubifs_info *c, int lnum) } /** - * next_log_lnum - switch to the next log LEB. - * @c: UBIFS file-system description object - * @lnum: current log LEB - */ -static inline int next_log_lnum(const struct ubifs_info *c, int lnum) -{ - lnum += 1; - if (lnum > c->log_last) - lnum = UBIFS_LOG_LNUM; - - return lnum; -} - -/** * empty_log_bytes - calculate amount of empty space in the log. * @c: UBIFS file-system description object */ @@ -175,26 +161,6 @@ void ubifs_add_bud(struct ubifs_info *c, struct ubifs_bud *bud) } /** - * ubifs_create_buds_lists - create journal head buds lists for remount rw. - * @c: UBIFS file-system description object - */ -void ubifs_create_buds_lists(struct ubifs_info *c) -{ - struct rb_node *p; - - spin_lock(&c->buds_lock); - p = rb_first(&c->buds); - while (p) { - struct ubifs_bud *bud = rb_entry(p, struct ubifs_bud, rb); - struct ubifs_jhead *jhead = &c->jheads[bud->jhead]; - - list_add_tail(&bud->list, &jhead->buds_list); - p = rb_next(p); - } - spin_unlock(&c->buds_lock); -} - -/** * ubifs_add_bud_to_log - add a new bud to the log. * @c: UBIFS file-system description object * @jhead: journal head the bud belongs to @@ -277,7 +243,7 @@ int ubifs_add_bud_to_log(struct ubifs_info *c, int jhead, int lnum, int offs) ref->jhead = cpu_to_le32(jhead); if (c->lhead_offs > c->leb_size - c->ref_node_alsz) { - c->lhead_lnum = next_log_lnum(c, c->lhead_lnum); + c->lhead_lnum = ubifs_next_log_lnum(c, c->lhead_lnum); c->lhead_offs = 0; } @@ -445,7 +411,7 @@ int ubifs_log_start_commit(struct ubifs_info *c, int *ltail_lnum) /* Switch to the next log LEB */ if (c->lhead_offs) { - c->lhead_lnum = next_log_lnum(c, c->lhead_lnum); + c->lhead_lnum = ubifs_next_log_lnum(c, c->lhead_lnum); c->lhead_offs = 0; } @@ -466,7 +432,7 @@ int ubifs_log_start_commit(struct ubifs_info *c, int *ltail_lnum) c->lhead_offs += len; if (c->lhead_offs == c->leb_size) { - c->lhead_lnum = next_log_lnum(c, c->lhead_lnum); + c->lhead_lnum = ubifs_next_log_lnum(c, c->lhead_lnum); c->lhead_offs = 0; } @@ -553,7 +519,7 @@ int ubifs_log_post_commit(struct ubifs_info *c, int old_ltail_lnum) } mutex_lock(&c->log_mutex); for (lnum = old_ltail_lnum; lnum != c->ltail_lnum; - lnum = next_log_lnum(c, lnum)) { + lnum = ubifs_next_log_lnum(c, lnum)) { dbg_log("unmap log LEB %d", lnum); err = ubifs_leb_unmap(c, lnum); if (err) @@ -662,7 +628,7 @@ static int add_node(struct ubifs_info *c, void *buf, int *lnum, int *offs, err = ubifs_leb_change(c, *lnum, buf, sz, UBI_SHORTTERM); if (err) return err; - *lnum = next_log_lnum(c, *lnum); + *lnum = ubifs_next_log_lnum(c, *lnum); *offs = 0; } memcpy(buf + *offs, node, len); @@ -732,7 +698,7 @@ int ubifs_consolidate_log(struct ubifs_info *c) ubifs_scan_destroy(sleb); if (lnum == c->lhead_lnum) break; - lnum = next_log_lnum(c, lnum); + lnum = ubifs_next_log_lnum(c, lnum); } if (offs) { int sz = ALIGN(offs, c->min_io_size); @@ -752,7 +718,7 @@ int ubifs_consolidate_log(struct ubifs_info *c) /* Unmap remaining LEBs */ lnum = write_lnum; do { - lnum = next_log_lnum(c, lnum); + lnum = ubifs_next_log_lnum(c, lnum); err = ubifs_leb_unmap(c, lnum); if (err) return err; diff --git a/fs/ubifs/lprops.c b/fs/ubifs/lprops.c index 0ee0847f2421..667884f4a615 100644 --- a/fs/ubifs/lprops.c +++ b/fs/ubifs/lprops.c @@ -1007,21 +1007,11 @@ out: } /** - * struct scan_check_data - data provided to scan callback function. - * @lst: LEB properties statistics - * @err: error code - */ -struct scan_check_data { - struct ubifs_lp_stats lst; - int err; -}; - -/** * scan_check_cb - scan callback. * @c: the UBIFS file-system description object * @lp: LEB properties to scan * @in_tree: whether the LEB properties are in main memory - * @data: information passed to and from the caller of the scan + * @lst: lprops statistics to update * * This function returns a code that indicates whether the scan should continue * (%LPT_SCAN_CONTINUE), whether the LEB properties should be added to the tree @@ -1030,11 +1020,10 @@ struct scan_check_data { */ static int scan_check_cb(struct ubifs_info *c, const struct ubifs_lprops *lp, int in_tree, - struct scan_check_data *data) + struct ubifs_lp_stats *lst) { struct ubifs_scan_leb *sleb; struct ubifs_scan_node *snod; - struct ubifs_lp_stats *lst = &data->lst; int cat, lnum = lp->lnum, is_idx = 0, used = 0, free, dirty, ret; void *buf = NULL; @@ -1044,7 +1033,7 @@ static int scan_check_cb(struct ubifs_info *c, if (cat != (lp->flags & LPROPS_CAT_MASK)) { ubifs_err("bad LEB category %d expected %d", (lp->flags & LPROPS_CAT_MASK), cat); - goto out; + return -EINVAL; } } @@ -1078,7 +1067,7 @@ static int scan_check_cb(struct ubifs_info *c, } if (!found) { ubifs_err("bad LPT list (category %d)", cat); - goto out; + return -EINVAL; } } } @@ -1090,45 +1079,40 @@ static int scan_check_cb(struct ubifs_info *c, if ((lp->hpos != -1 && heap->arr[lp->hpos]->lnum != lnum) || lp != heap->arr[lp->hpos]) { ubifs_err("bad LPT heap (category %d)", cat); - goto out; + return -EINVAL; } } buf = __vmalloc(c->leb_size, GFP_NOFS, PAGE_KERNEL); - if (!buf) { - ubifs_err("cannot allocate memory to scan LEB %d", lnum); - goto out; + if (!buf) + return -ENOMEM; + + /* + * After an unclean unmount, empty and freeable LEBs + * may contain garbage - do not scan them. + */ + if (lp->free == c->leb_size) { + lst->empty_lebs += 1; + lst->total_free += c->leb_size; + lst->total_dark += ubifs_calc_dark(c, c->leb_size); + return LPT_SCAN_CONTINUE; + } + if (lp->free + lp->dirty == c->leb_size && + !(lp->flags & LPROPS_INDEX)) { + lst->total_free += lp->free; + lst->total_dirty += lp->dirty; + lst->total_dark += ubifs_calc_dark(c, c->leb_size); + return LPT_SCAN_CONTINUE; } sleb = ubifs_scan(c, lnum, 0, buf, 0); if (IS_ERR(sleb)) { - /* - * After an unclean unmount, empty and freeable LEBs - * may contain garbage. - */ - if (lp->free == c->leb_size) { - ubifs_err("scan errors were in empty LEB " - "- continuing checking"); - lst->empty_lebs += 1; - lst->total_free += c->leb_size; - lst->total_dark += ubifs_calc_dark(c, c->leb_size); - ret = LPT_SCAN_CONTINUE; - goto exit; - } - - if (lp->free + lp->dirty == c->leb_size && - !(lp->flags & LPROPS_INDEX)) { - ubifs_err("scan errors were in freeable LEB " - "- continuing checking"); - lst->total_free += lp->free; - lst->total_dirty += lp->dirty; - lst->total_dark += ubifs_calc_dark(c, c->leb_size); - ret = LPT_SCAN_CONTINUE; - goto exit; + ret = PTR_ERR(sleb); + if (ret == -EUCLEAN) { + dbg_dump_lprops(c); + dbg_dump_budg(c, &c->bi); } - data->err = PTR_ERR(sleb); - ret = LPT_SCAN_STOP; - goto exit; + goto out; } is_idx = -1; @@ -1246,10 +1230,8 @@ static int scan_check_cb(struct ubifs_info *c, } ubifs_scan_destroy(sleb); - ret = LPT_SCAN_CONTINUE; -exit: vfree(buf); - return ret; + return LPT_SCAN_CONTINUE; out_print: ubifs_err("bad accounting of LEB %d: free %d, dirty %d flags %#x, " @@ -1258,10 +1240,10 @@ out_print: dbg_dump_leb(c, lnum); out_destroy: ubifs_scan_destroy(sleb); + ret = -EINVAL; out: vfree(buf); - data->err = -EINVAL; - return LPT_SCAN_STOP; + return ret; } /** @@ -1278,8 +1260,7 @@ out: int dbg_check_lprops(struct ubifs_info *c) { int i, err; - struct scan_check_data data; - struct ubifs_lp_stats *lst = &data.lst; + struct ubifs_lp_stats lst; if (!(ubifs_chk_flags & UBIFS_CHK_LPROPS)) return 0; @@ -1294,29 +1275,23 @@ int dbg_check_lprops(struct ubifs_info *c) return err; } - memset(lst, 0, sizeof(struct ubifs_lp_stats)); - - data.err = 0; + memset(&lst, 0, sizeof(struct ubifs_lp_stats)); err = ubifs_lpt_scan_nolock(c, c->main_first, c->leb_cnt - 1, (ubifs_lpt_scan_callback)scan_check_cb, - &data); + &lst); if (err && err != -ENOSPC) goto out; - if (data.err) { - err = data.err; - goto out; - } - if (lst->empty_lebs != c->lst.empty_lebs || - lst->idx_lebs != c->lst.idx_lebs || - lst->total_free != c->lst.total_free || - lst->total_dirty != c->lst.total_dirty || - lst->total_used != c->lst.total_used) { + if (lst.empty_lebs != c->lst.empty_lebs || + lst.idx_lebs != c->lst.idx_lebs || + lst.total_free != c->lst.total_free || + lst.total_dirty != c->lst.total_dirty || + lst.total_used != c->lst.total_used) { ubifs_err("bad overall accounting"); ubifs_err("calculated: empty_lebs %d, idx_lebs %d, " "total_free %lld, total_dirty %lld, total_used %lld", - lst->empty_lebs, lst->idx_lebs, lst->total_free, - lst->total_dirty, lst->total_used); + lst.empty_lebs, lst.idx_lebs, lst.total_free, + lst.total_dirty, lst.total_used); ubifs_err("read from lprops: empty_lebs %d, idx_lebs %d, " "total_free %lld, total_dirty %lld, total_used %lld", c->lst.empty_lebs, c->lst.idx_lebs, c->lst.total_free, @@ -1325,11 +1300,11 @@ int dbg_check_lprops(struct ubifs_info *c) goto out; } - if (lst->total_dead != c->lst.total_dead || - lst->total_dark != c->lst.total_dark) { + if (lst.total_dead != c->lst.total_dead || + lst.total_dark != c->lst.total_dark) { ubifs_err("bad dead/dark space accounting"); ubifs_err("calculated: total_dead %lld, total_dark %lld", - lst->total_dead, lst->total_dark); + lst.total_dead, lst.total_dark); ubifs_err("read from lprops: total_dead %lld, total_dark %lld", c->lst.total_dead, c->lst.total_dark); err = -EINVAL; diff --git a/fs/ubifs/lpt.c b/fs/ubifs/lpt.c index 72775d35b99e..ef5155e109a2 100644 --- a/fs/ubifs/lpt.c +++ b/fs/ubifs/lpt.c @@ -1270,10 +1270,9 @@ static int read_pnode(struct ubifs_info *c, struct ubifs_nnode *parent, int iip) lnum = branch->lnum; offs = branch->offs; pnode = kzalloc(sizeof(struct ubifs_pnode), GFP_NOFS); - if (!pnode) { - err = -ENOMEM; - goto out; - } + if (!pnode) + return -ENOMEM; + if (lnum == 0) { /* * This pnode was not written which just means that the LEB diff --git a/fs/ubifs/lpt_commit.c b/fs/ubifs/lpt_commit.c index 0c9c69bd983a..dfcb5748a7dc 100644 --- a/fs/ubifs/lpt_commit.c +++ b/fs/ubifs/lpt_commit.c @@ -29,6 +29,12 @@ #include <linux/slab.h> #include "ubifs.h" +#ifdef CONFIG_UBIFS_FS_DEBUG +static int dbg_populate_lsave(struct ubifs_info *c); +#else +#define dbg_populate_lsave(c) 0 +#endif + /** * first_dirty_cnode - find first dirty cnode. * @c: UBIFS file-system description object @@ -586,7 +592,7 @@ static struct ubifs_pnode *next_pnode_to_dirty(struct ubifs_info *c, if (nnode->nbranch[iip].lnum) break; } - } while (iip >= UBIFS_LPT_FANOUT); + } while (iip >= UBIFS_LPT_FANOUT); /* Go right */ nnode = ubifs_get_nnode(c, nnode, iip); @@ -815,6 +821,10 @@ static void populate_lsave(struct ubifs_info *c) c->lpt_drty_flgs |= LSAVE_DIRTY; ubifs_add_lpt_dirt(c, c->lsave_lnum, c->lsave_sz); } + + if (dbg_populate_lsave(c)) + return; + list_for_each_entry(lprops, &c->empty_list, list) { c->lsave[cnt++] = lprops->lnum; if (cnt >= c->lsave_cnt) @@ -1994,4 +2004,47 @@ void dbg_dump_lpt_lebs(const struct ubifs_info *c) current->pid); } +/** + * dbg_populate_lsave - debugging version of 'populate_lsave()' + * @c: UBIFS file-system description object + * + * This is a debugging version for 'populate_lsave()' which populates lsave + * with random LEBs instead of useful LEBs, which is good for test coverage. + * Returns zero if lsave has not been populated (this debugging feature is + * disabled) an non-zero if lsave has been populated. + */ +static int dbg_populate_lsave(struct ubifs_info *c) +{ + struct ubifs_lprops *lprops; + struct ubifs_lpt_heap *heap; + int i; + + if (!(ubifs_chk_flags & UBIFS_CHK_GEN)) + return 0; + if (random32() & 3) + return 0; + + for (i = 0; i < c->lsave_cnt; i++) + c->lsave[i] = c->main_first; + + list_for_each_entry(lprops, &c->empty_list, list) + c->lsave[random32() % c->lsave_cnt] = lprops->lnum; + list_for_each_entry(lprops, &c->freeable_list, list) + c->lsave[random32() % c->lsave_cnt] = lprops->lnum; + list_for_each_entry(lprops, &c->frdi_idx_list, list) + c->lsave[random32() % c->lsave_cnt] = lprops->lnum; + + heap = &c->lpt_heap[LPROPS_DIRTY_IDX - 1]; + for (i = 0; i < heap->cnt; i++) + c->lsave[random32() % c->lsave_cnt] = heap->arr[i]->lnum; + heap = &c->lpt_heap[LPROPS_DIRTY - 1]; + for (i = 0; i < heap->cnt; i++) + c->lsave[random32() % c->lsave_cnt] = heap->arr[i]->lnum; + heap = &c->lpt_heap[LPROPS_FREE - 1]; + for (i = 0; i < heap->cnt; i++) + c->lsave[random32() % c->lsave_cnt] = heap->arr[i]->lnum; + + return 1; +} + #endif /* CONFIG_UBIFS_FS_DEBUG */ diff --git a/fs/ubifs/master.c b/fs/ubifs/master.c index 21f47afdacff..278c2382e8c2 100644 --- a/fs/ubifs/master.c +++ b/fs/ubifs/master.c @@ -148,7 +148,7 @@ static int validate_master(const struct ubifs_info *c) } main_sz = (long long)c->main_lebs * c->leb_size; - if (c->old_idx_sz & 7 || c->old_idx_sz >= main_sz) { + if (c->bi.old_idx_sz & 7 || c->bi.old_idx_sz >= main_sz) { err = 9; goto out; } @@ -218,7 +218,7 @@ static int validate_master(const struct ubifs_info *c) } if (c->lst.total_dead + c->lst.total_dark + - c->lst.total_used + c->old_idx_sz > main_sz) { + c->lst.total_used + c->bi.old_idx_sz > main_sz) { err = 21; goto out; } @@ -286,7 +286,7 @@ int ubifs_read_master(struct ubifs_info *c) c->gc_lnum = le32_to_cpu(c->mst_node->gc_lnum); c->ihead_lnum = le32_to_cpu(c->mst_node->ihead_lnum); c->ihead_offs = le32_to_cpu(c->mst_node->ihead_offs); - c->old_idx_sz = le64_to_cpu(c->mst_node->index_size); + c->bi.old_idx_sz = le64_to_cpu(c->mst_node->index_size); c->lpt_lnum = le32_to_cpu(c->mst_node->lpt_lnum); c->lpt_offs = le32_to_cpu(c->mst_node->lpt_offs); c->nhead_lnum = le32_to_cpu(c->mst_node->nhead_lnum); @@ -305,7 +305,7 @@ int ubifs_read_master(struct ubifs_info *c) c->lst.total_dead = le64_to_cpu(c->mst_node->total_dead); c->lst.total_dark = le64_to_cpu(c->mst_node->total_dark); - c->calc_idx_sz = c->old_idx_sz; + c->calc_idx_sz = c->bi.old_idx_sz; if (c->mst_node->flags & cpu_to_le32(UBIFS_MST_NO_ORPHS)) c->no_orphs = 1; diff --git a/fs/ubifs/misc.h b/fs/ubifs/misc.h index c3de04dc952a..0b5296a9a4c5 100644 --- a/fs/ubifs/misc.h +++ b/fs/ubifs/misc.h @@ -340,4 +340,21 @@ static inline void ubifs_release_lprops(struct ubifs_info *c) mutex_unlock(&c->lp_mutex); } +/** + * ubifs_next_log_lnum - switch to the next log LEB. + * @c: UBIFS file-system description object + * @lnum: current log LEB + * + * This helper function returns the log LEB number which goes next after LEB + * 'lnum'. + */ +static inline int ubifs_next_log_lnum(const struct ubifs_info *c, int lnum) +{ + lnum += 1; + if (lnum > c->log_last) + lnum = UBIFS_LOG_LNUM; + + return lnum; +} + #endif /* __UBIFS_MISC_H__ */ diff --git a/fs/ubifs/orphan.c b/fs/ubifs/orphan.c index 09df318e368f..bd644bf587a8 100644 --- a/fs/ubifs/orphan.c +++ b/fs/ubifs/orphan.c @@ -673,7 +673,8 @@ static int kill_orphans(struct ubifs_info *c) sleb = ubifs_scan(c, lnum, 0, c->sbuf, 1); if (IS_ERR(sleb)) { if (PTR_ERR(sleb) == -EUCLEAN) - sleb = ubifs_recover_leb(c, lnum, 0, c->sbuf, 0); + sleb = ubifs_recover_leb(c, lnum, 0, + c->sbuf, 0); if (IS_ERR(sleb)) { err = PTR_ERR(sleb); break; diff --git a/fs/ubifs/recovery.c b/fs/ubifs/recovery.c index 936f2cbfe6b6..731d9e2e7b50 100644 --- a/fs/ubifs/recovery.c +++ b/fs/ubifs/recovery.c @@ -317,6 +317,32 @@ int ubifs_recover_master_node(struct ubifs_info *c) goto out_free; } memcpy(c->rcvrd_mst_node, c->mst_node, UBIFS_MST_NODE_SZ); + + /* + * We had to recover the master node, which means there was an + * unclean reboot. However, it is possible that the master node + * is clean at this point, i.e., %UBIFS_MST_DIRTY is not set. + * E.g., consider the following chain of events: + * + * 1. UBIFS was cleanly unmounted, so the master node is clean + * 2. UBIFS is being mounted R/W and starts changing the master + * node in the first (%UBIFS_MST_LNUM). A power cut happens, + * so this LEB ends up with some amount of garbage at the + * end. + * 3. UBIFS is being mounted R/O. We reach this place and + * recover the master node from the second LEB + * (%UBIFS_MST_LNUM + 1). But we cannot update the media + * because we are being mounted R/O. We have to defer the + * operation. + * 4. However, this master node (@c->mst_node) is marked as + * clean (since the step 1). And if we just return, the + * mount code will be confused and won't recover the master + * node when it is re-mounter R/W later. + * + * Thus, to force the recovery by marking the master node as + * dirty. + */ + c->mst_node->flags |= cpu_to_le32(UBIFS_MST_DIRTY); } else { /* Write the recovered master node */ c->max_sqnum = le64_to_cpu(mst->ch.sqnum) - 1; @@ -538,13 +564,16 @@ static int fix_unclean_leb(struct ubifs_info *c, struct ubifs_scan_leb *sleb, } /** - * drop_incomplete_group - drop nodes from an incomplete group. + * drop_last_node - drop the last node or group of nodes. * @sleb: scanned LEB information * @offs: offset of dropped nodes is returned here + * @grouped: non-zero if whole group of nodes have to be dropped * - * This function returns %1 if nodes are dropped and %0 otherwise. + * This is a helper function for 'ubifs_recover_leb()' which drops the last + * node of the scanned LEB or the last group of nodes if @grouped is not zero. + * This function returns %1 if a node was dropped and %0 otherwise. */ -static int drop_incomplete_group(struct ubifs_scan_leb *sleb, int *offs) +static int drop_last_node(struct ubifs_scan_leb *sleb, int *offs, int grouped) { int dropped = 0; @@ -563,6 +592,8 @@ static int drop_incomplete_group(struct ubifs_scan_leb *sleb, int *offs) kfree(snod); sleb->nodes_cnt -= 1; dropped = 1; + if (!grouped) + break; } return dropped; } @@ -583,8 +614,7 @@ static int drop_incomplete_group(struct ubifs_scan_leb *sleb, int *offs) struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, int offs, void *sbuf, int grouped) { - int err, len = c->leb_size - offs, need_clean = 0, quiet = 1; - int empty_chkd = 0, start = offs; + int ret = 0, err, len = c->leb_size - offs, start = offs, min_io_unit; struct ubifs_scan_leb *sleb; void *buf = sbuf + offs; @@ -594,12 +624,8 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, if (IS_ERR(sleb)) return sleb; - if (sleb->ecc) - need_clean = 1; - + ubifs_assert(len >= 8); while (len >= 8) { - int ret; - dbg_scan("look at LEB %d:%d (%d bytes left)", lnum, offs, len); @@ -609,8 +635,7 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, * Scan quietly until there is an error from which we cannot * recover */ - ret = ubifs_scan_a_node(c, buf, len, lnum, offs, quiet); - + ret = ubifs_scan_a_node(c, buf, len, lnum, offs, 0); if (ret == SCANNED_A_NODE) { /* A valid node, and not a padding node */ struct ubifs_ch *ch = buf; @@ -623,70 +648,32 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, offs += node_len; buf += node_len; len -= node_len; - continue; - } - - if (ret > 0) { + } else if (ret > 0) { /* Padding bytes or a valid padding node */ offs += ret; buf += ret; len -= ret; - continue; - } - - if (ret == SCANNED_EMPTY_SPACE) { - if (!is_empty(buf, len)) { - if (!is_last_write(c, buf, offs)) - break; - clean_buf(c, &buf, lnum, &offs, &len); - need_clean = 1; - } - empty_chkd = 1; + } else if (ret == SCANNED_EMPTY_SPACE || + ret == SCANNED_GARBAGE || + ret == SCANNED_A_BAD_PAD_NODE || + ret == SCANNED_A_CORRUPT_NODE) { + dbg_rcvry("found corruption - %d", ret); break; - } - - if (ret == SCANNED_GARBAGE || ret == SCANNED_A_BAD_PAD_NODE) - if (is_last_write(c, buf, offs)) { - clean_buf(c, &buf, lnum, &offs, &len); - need_clean = 1; - empty_chkd = 1; - break; - } - - if (ret == SCANNED_A_CORRUPT_NODE) - if (no_more_nodes(c, buf, len, lnum, offs)) { - clean_buf(c, &buf, lnum, &offs, &len); - need_clean = 1; - empty_chkd = 1; - break; - } - - if (quiet) { - /* Redo the last scan but noisily */ - quiet = 0; - continue; - } - - switch (ret) { - case SCANNED_GARBAGE: - dbg_err("garbage"); - goto corrupted; - case SCANNED_A_CORRUPT_NODE: - case SCANNED_A_BAD_PAD_NODE: - dbg_err("bad node"); - goto corrupted; - default: - dbg_err("unknown"); + } else { + dbg_err("unexpected return value %d", ret); err = -EINVAL; goto error; } } - if (!empty_chkd && !is_empty(buf, len)) { - if (is_last_write(c, buf, offs)) { - clean_buf(c, &buf, lnum, &offs, &len); - need_clean = 1; - } else { + if (ret == SCANNED_GARBAGE || ret == SCANNED_A_BAD_PAD_NODE) { + if (!is_last_write(c, buf, offs)) + goto corrupted_rescan; + } else if (ret == SCANNED_A_CORRUPT_NODE) { + if (!no_more_nodes(c, buf, len, lnum, offs)) + goto corrupted_rescan; + } else if (!is_empty(buf, len)) { + if (!is_last_write(c, buf, offs)) { int corruption = first_non_ff(buf, len); /* @@ -702,29 +689,82 @@ struct ubifs_scan_leb *ubifs_recover_leb(struct ubifs_info *c, int lnum, } } - /* Drop nodes from incomplete group */ - if (grouped && drop_incomplete_group(sleb, &offs)) { - buf = sbuf + offs; - len = c->leb_size - offs; - clean_buf(c, &buf, lnum, &offs, &len); - need_clean = 1; - } + min_io_unit = round_down(offs, c->min_io_size); + if (grouped) + /* + * If nodes are grouped, always drop the incomplete group at + * the end. + */ + drop_last_node(sleb, &offs, 1); - if (offs % c->min_io_size) { - clean_buf(c, &buf, lnum, &offs, &len); - need_clean = 1; - } + /* + * While we are in the middle of the same min. I/O unit keep dropping + * nodes. So basically, what we want is to make sure that the last min. + * I/O unit where we saw the corruption is dropped completely with all + * the uncorrupted node which may possibly sit there. + * + * In other words, let's name the min. I/O unit where the corruption + * starts B, and the previous min. I/O unit A. The below code tries to + * deal with a situation when half of B contains valid nodes or the end + * of a valid node, and the second half of B contains corrupted data or + * garbage. This means that UBIFS had been writing to B just before the + * power cut happened. I do not know how realistic is this scenario + * that half of the min. I/O unit had been written successfully and the + * other half not, but this is possible in our 'failure mode emulation' + * infrastructure at least. + * + * So what is the problem, why we need to drop those nodes? Whey can't + * we just clean-up the second half of B by putting a padding node + * there? We can, and this works fine with one exception which was + * reproduced with power cut emulation testing and happens extremely + * rarely. The description follows, but it is worth noting that that is + * only about the GC head, so we could do this trick only if the bud + * belongs to the GC head, but it does not seem to be worth an + * additional "if" statement. + * + * So, imagine the file-system is full, we run GC which is moving valid + * nodes from LEB X to LEB Y (obviously, LEB Y is the current GC head + * LEB). The @c->gc_lnum is -1, which means that GC will retain LEB X + * and will try to continue. Imagine that LEB X is currently the + * dirtiest LEB, and the amount of used space in LEB Y is exactly the + * same as amount of free space in LEB X. + * + * And a power cut happens when nodes are moved from LEB X to LEB Y. We + * are here trying to recover LEB Y which is the GC head LEB. We find + * the min. I/O unit B as described above. Then we clean-up LEB Y by + * padding min. I/O unit. And later 'ubifs_rcvry_gc_commit()' function + * fails, because it cannot find a dirty LEB which could be GC'd into + * LEB Y! Even LEB X does not match because the amount of valid nodes + * there does not fit the free space in LEB Y any more! And this is + * because of the padding node which we added to LEB Y. The + * user-visible effect of this which I once observed and analysed is + * that we cannot mount the file-system with -ENOSPC error. + * + * So obviously, to make sure that situation does not happen we should + * free min. I/O unit B in LEB Y completely and the last used min. I/O + * unit in LEB Y should be A. This is basically what the below code + * tries to do. + */ + while (min_io_unit == round_down(offs, c->min_io_size) && + min_io_unit != offs && + drop_last_node(sleb, &offs, grouped)); + + buf = sbuf + offs; + len = c->leb_size - offs; + clean_buf(c, &buf, lnum, &offs, &len); ubifs_end_scan(c, sleb, lnum, offs); - if (need_clean) { - err = fix_unclean_leb(c, sleb, start); - if (err) - goto error; - } + err = fix_unclean_leb(c, sleb, start); + if (err) + goto error; return sleb; +corrupted_rescan: + /* Re-scan the corrupted data with verbose messages */ + dbg_err("corruptio %d", ret); + ubifs_scan_a_node(c, buf, len, lnum, offs, 1); corrupted: ubifs_scanned_corruption(c, lnum, offs, buf); err = -EUCLEAN; @@ -1044,6 +1084,53 @@ int ubifs_clean_lebs(const struct ubifs_info *c, void *sbuf) } /** + * grab_empty_leb - grab an empty LEB to use as GC LEB and run commit. + * @c: UBIFS file-system description object + * + * This is a helper function for 'ubifs_rcvry_gc_commit()' which grabs an empty + * LEB to be used as GC LEB (@c->gc_lnum), and then runs the commit. Returns + * zero in case of success and a negative error code in case of failure. + */ +static int grab_empty_leb(struct ubifs_info *c) +{ + int lnum, err; + + /* + * Note, it is very important to first search for an empty LEB and then + * run the commit, not vice-versa. The reason is that there might be + * only one empty LEB at the moment, the one which has been the + * @c->gc_lnum just before the power cut happened. During the regular + * UBIFS operation (not now) @c->gc_lnum is marked as "taken", so no + * one but GC can grab it. But at this moment this single empty LEB is + * not marked as taken, so if we run commit - what happens? Right, the + * commit will grab it and write the index there. Remember that the + * index always expands as long as there is free space, and it only + * starts consolidating when we run out of space. + * + * IOW, if we run commit now, we might not be able to find a free LEB + * after this. + */ + lnum = ubifs_find_free_leb_for_idx(c); + if (lnum < 0) { + dbg_err("could not find an empty LEB"); + dbg_dump_lprops(c); + dbg_dump_budg(c, &c->bi); + return lnum; + } + + /* Reset the index flag */ + err = ubifs_change_one_lp(c, lnum, LPROPS_NC, LPROPS_NC, 0, + LPROPS_INDEX, 0); + if (err) + return err; + + c->gc_lnum = lnum; + dbg_rcvry("found empty LEB %d, run commit", lnum); + + return ubifs_run_commit(c); +} + +/** * ubifs_rcvry_gc_commit - recover the GC LEB number and run the commit. * @c: UBIFS file-system description object * @@ -1065,71 +1152,26 @@ int ubifs_rcvry_gc_commit(struct ubifs_info *c) { struct ubifs_wbuf *wbuf = &c->jheads[GCHD].wbuf; struct ubifs_lprops lp; - int lnum, err; + int err; + + dbg_rcvry("GC head LEB %d, offs %d", wbuf->lnum, wbuf->offs); c->gc_lnum = -1; - if (wbuf->lnum == -1) { - dbg_rcvry("no GC head LEB"); - goto find_free; - } - /* - * See whether the used space in the dirtiest LEB fits in the GC head - * LEB. - */ - if (wbuf->offs == c->leb_size) { - dbg_rcvry("no room in GC head LEB"); - goto find_free; - } + if (wbuf->lnum == -1 || wbuf->offs == c->leb_size) + return grab_empty_leb(c); + err = ubifs_find_dirty_leb(c, &lp, wbuf->offs, 2); if (err) { - /* - * There are no dirty or empty LEBs subject to here being - * enough for the index. Try to use - * 'ubifs_find_free_leb_for_idx()', which will return any empty - * LEBs (ignoring index requirements). If the index then - * doesn't have enough LEBs the recovery commit will fail - - * which is the same result anyway i.e. recovery fails. So - * there is no problem ignoring index requirements and just - * grabbing a free LEB since we have already established there - * is not a dirty LEB we could have used instead. - */ - if (err == -ENOSPC) { - dbg_rcvry("could not find a dirty LEB"); - goto find_free; - } - return err; - } - ubifs_assert(!(lp.flags & LPROPS_INDEX)); - lnum = lp.lnum; - if (lp.free + lp.dirty == c->leb_size) { - /* An empty LEB was returned */ - if (lp.free != c->leb_size) { - err = ubifs_change_one_lp(c, lnum, c->leb_size, - 0, 0, 0, 0); - if (err) - return err; - } - err = ubifs_leb_unmap(c, lnum); - if (err) - return err; - c->gc_lnum = lnum; - dbg_rcvry("allocated LEB %d for GC", lnum); - /* Run the commit */ - dbg_rcvry("committing"); - return ubifs_run_commit(c); - } - /* - * There was no empty LEB so the used space in the dirtiest LEB must fit - * in the GC head LEB. - */ - if (lp.free + lp.dirty < wbuf->offs) { - dbg_rcvry("LEB %d doesn't fit in GC head LEB %d:%d", - lnum, wbuf->lnum, wbuf->offs); - err = ubifs_return_leb(c, lnum); - if (err) + if (err != -ENOSPC) return err; - goto find_free; + + dbg_rcvry("could not find a dirty LEB"); + return grab_empty_leb(c); } + + ubifs_assert(!(lp.flags & LPROPS_INDEX)); + ubifs_assert(lp.free + lp.dirty >= wbuf->offs); + /* * We run the commit before garbage collection otherwise subsequent * mounts will see the GC and orphan deletion in a different order. @@ -1138,11 +1180,8 @@ int ubifs_rcvry_gc_commit(struct ubifs_info *c) err = ubifs_run_commit(c); if (err) return err; - /* - * The data in the dirtiest LEB fits in the GC head LEB, so do the GC - * - use locking to keep 'ubifs_assert()' happy. - */ - dbg_rcvry("GC'ing LEB %d", lnum); + + dbg_rcvry("GC'ing LEB %d", lp.lnum); mutex_lock_nested(&wbuf->io_mutex, wbuf->jhead); err = ubifs_garbage_collect_leb(c, &lp); if (err >= 0) { @@ -1158,37 +1197,17 @@ int ubifs_rcvry_gc_commit(struct ubifs_info *c) err = -EINVAL; return err; } - if (err != LEB_RETAINED) { - dbg_err("GC returned %d", err); + + ubifs_assert(err == LEB_RETAINED); + if (err != LEB_RETAINED) return -EINVAL; - } + err = ubifs_leb_unmap(c, c->gc_lnum); if (err) return err; - dbg_rcvry("allocated LEB %d for GC", lnum); - return 0; -find_free: - /* - * There is no GC head LEB or the free space in the GC head LEB is too - * small, or there are not dirty LEBs. Allocate gc_lnum by calling - * 'ubifs_find_free_leb_for_idx()' so GC is not run. - */ - lnum = ubifs_find_free_leb_for_idx(c); - if (lnum < 0) { - dbg_err("could not find an empty LEB"); - return lnum; - } - /* And reset the index flag */ - err = ubifs_change_one_lp(c, lnum, LPROPS_NC, LPROPS_NC, 0, - LPROPS_INDEX, 0); - if (err) - return err; - c->gc_lnum = lnum; - dbg_rcvry("allocated LEB %d for GC", lnum); - /* Run the commit */ - dbg_rcvry("committing"); - return ubifs_run_commit(c); + dbg_rcvry("allocated LEB %d for GC", lp.lnum); + return 0; } /** @@ -1430,7 +1449,7 @@ static int fix_size_in_place(struct ubifs_info *c, struct size_entry *e) err = ubi_leb_change(c->ubi, lnum, c->sbuf, len, UBI_UNKNOWN); if (err) goto out; - dbg_rcvry("inode %lu at %d:%d size %lld -> %lld ", + dbg_rcvry("inode %lu at %d:%d size %lld -> %lld", (unsigned long)e->inum, lnum, offs, i_size, e->d_size); return 0; @@ -1479,20 +1498,27 @@ int ubifs_recover_size(struct ubifs_info *c) e->i_size = le64_to_cpu(ino->size); } } + if (e->exists && e->i_size < e->d_size) { - if (!e->inode && c->ro_mount) { + if (c->ro_mount) { /* Fix the inode size and pin it in memory */ struct inode *inode; + struct ubifs_inode *ui; + + ubifs_assert(!e->inode); inode = ubifs_iget(c->vfs_sb, e->inum); if (IS_ERR(inode)) return PTR_ERR(inode); + + ui = ubifs_inode(inode); if (inode->i_size < e->d_size) { dbg_rcvry("ino %lu size %lld -> %lld", (unsigned long)e->inum, - e->d_size, inode->i_size); + inode->i_size, e->d_size); inode->i_size = e->d_size; - ubifs_inode(inode)->ui_size = e->d_size; + ui->ui_size = e->d_size; + ui->synced_i_size = e->d_size; e->inode = inode; this = rb_next(this); continue; @@ -1507,9 +1533,11 @@ int ubifs_recover_size(struct ubifs_info *c) iput(e->inode); } } + this = rb_next(this); rb_erase(&e->rb, &c->size_tree); kfree(e); } + return 0; } diff --git a/fs/ubifs/replay.c b/fs/ubifs/replay.c index eed0fcff8d73..6617280d1679 100644 --- a/fs/ubifs/replay.c +++ b/fs/ubifs/replay.c @@ -33,43 +33,32 @@ */ #include "ubifs.h" - -/* - * Replay flags. - * - * REPLAY_DELETION: node was deleted - * REPLAY_REF: node is a reference node - */ -enum { - REPLAY_DELETION = 1, - REPLAY_REF = 2, -}; +#include <linux/list_sort.h> /** - * struct replay_entry - replay tree entry. + * struct replay_entry - replay list entry. * @lnum: logical eraseblock number of the node * @offs: node offset * @len: node length + * @deletion: non-zero if this entry corresponds to a node deletion * @sqnum: node sequence number - * @flags: replay flags - * @rb: links the replay tree + * @list: links the replay list * @key: node key * @nm: directory entry name * @old_size: truncation old size * @new_size: truncation new size - * @free: amount of free space in a bud - * @dirty: amount of dirty space in a bud from padding and deletion nodes * - * UBIFS journal replay must compare node sequence numbers, which means it must - * build a tree of node information to insert into the TNC. + * The replay process first scans all buds and builds the replay list, then + * sorts the replay list in nodes sequence number order, and then inserts all + * the replay entries to the TNC. */ struct replay_entry { int lnum; int offs; int len; + unsigned int deletion:1; unsigned long long sqnum; - int flags; - struct rb_node rb; + struct list_head list; union ubifs_key key; union { struct qstr nm; @@ -77,10 +66,6 @@ struct replay_entry { loff_t old_size; loff_t new_size; }; - struct { - int free; - int dirty; - }; }; }; @@ -88,57 +73,64 @@ struct replay_entry { * struct bud_entry - entry in the list of buds to replay. * @list: next bud in the list * @bud: bud description object - * @free: free bytes in the bud * @sqnum: reference node sequence number + * @free: free bytes in the bud + * @dirty: dirty bytes in the bud */ struct bud_entry { struct list_head list; struct ubifs_bud *bud; - int free; unsigned long long sqnum; + int free; + int dirty; }; /** * set_bud_lprops - set free and dirty space used by a bud. * @c: UBIFS file-system description object - * @r: replay entry of bud + * @b: bud entry which describes the bud + * + * This function makes sure the LEB properties of bud @b are set correctly + * after the replay. Returns zero in case of success and a negative error code + * in case of failure. */ -static int set_bud_lprops(struct ubifs_info *c, struct replay_entry *r) +static int set_bud_lprops(struct ubifs_info *c, struct bud_entry *b) { const struct ubifs_lprops *lp; int err = 0, dirty; ubifs_get_lprops(c); - lp = ubifs_lpt_lookup_dirty(c, r->lnum); + lp = ubifs_lpt_lookup_dirty(c, b->bud->lnum); if (IS_ERR(lp)) { err = PTR_ERR(lp); goto out; } dirty = lp->dirty; - if (r->offs == 0 && (lp->free != c->leb_size || lp->dirty != 0)) { + if (b->bud->start == 0 && (lp->free != c->leb_size || lp->dirty != 0)) { /* * The LEB was added to the journal with a starting offset of * zero which means the LEB must have been empty. The LEB - * property values should be lp->free == c->leb_size and - * lp->dirty == 0, but that is not the case. The reason is that - * the LEB was garbage collected. The garbage collector resets - * the free and dirty space without recording it anywhere except - * lprops, so if there is not a commit then lprops does not have - * that information next time the file system is mounted. + * property values should be @lp->free == @c->leb_size and + * @lp->dirty == 0, but that is not the case. The reason is that + * the LEB had been garbage collected before it became the bud, + * and there was not commit inbetween. The garbage collector + * resets the free and dirty space without recording it + * anywhere except lprops, so if there was no commit then + * lprops does not have that information. * * We do not need to adjust free space because the scan has told * us the exact value which is recorded in the replay entry as - * r->free. + * @b->free. * * However we do need to subtract from the dirty space the * amount of space that the garbage collector reclaimed, which * is the whole LEB minus the amount of space that was free. */ - dbg_mnt("bud LEB %d was GC'd (%d free, %d dirty)", r->lnum, + dbg_mnt("bud LEB %d was GC'd (%d free, %d dirty)", b->bud->lnum, lp->free, lp->dirty); - dbg_gc("bud LEB %d was GC'd (%d free, %d dirty)", r->lnum, + dbg_gc("bud LEB %d was GC'd (%d free, %d dirty)", b->bud->lnum, lp->free, lp->dirty); dirty -= c->leb_size - lp->free; /* @@ -150,21 +142,48 @@ static int set_bud_lprops(struct ubifs_info *c, struct replay_entry *r) */ if (dirty != 0) dbg_msg("LEB %d lp: %d free %d dirty " - "replay: %d free %d dirty", r->lnum, lp->free, - lp->dirty, r->free, r->dirty); + "replay: %d free %d dirty", b->bud->lnum, + lp->free, lp->dirty, b->free, b->dirty); } - lp = ubifs_change_lp(c, lp, r->free, dirty + r->dirty, + lp = ubifs_change_lp(c, lp, b->free, dirty + b->dirty, lp->flags | LPROPS_TAKEN, 0); if (IS_ERR(lp)) { err = PTR_ERR(lp); goto out; } + + /* Make sure the journal head points to the latest bud */ + err = ubifs_wbuf_seek_nolock(&c->jheads[b->bud->jhead].wbuf, + b->bud->lnum, c->leb_size - b->free, + UBI_SHORTTERM); + out: ubifs_release_lprops(c); return err; } /** + * set_buds_lprops - set free and dirty space for all replayed buds. + * @c: UBIFS file-system description object + * + * This function sets LEB properties for all replayed buds. Returns zero in + * case of success and a negative error code in case of failure. + */ +static int set_buds_lprops(struct ubifs_info *c) +{ + struct bud_entry *b; + int err; + + list_for_each_entry(b, &c->replay_buds, list) { + err = set_bud_lprops(c, b); + if (err) + return err; + } + + return 0; +} + +/** * trun_remove_range - apply a replay entry for a truncation to the TNC. * @c: UBIFS file-system description object * @r: replay entry of truncation @@ -200,24 +219,22 @@ static int trun_remove_range(struct ubifs_info *c, struct replay_entry *r) */ static int apply_replay_entry(struct ubifs_info *c, struct replay_entry *r) { - int err, deletion = ((r->flags & REPLAY_DELETION) != 0); + int err; - dbg_mnt("LEB %d:%d len %d flgs %d sqnum %llu %s", r->lnum, - r->offs, r->len, r->flags, r->sqnum, DBGKEY(&r->key)); + dbg_mnt("LEB %d:%d len %d deletion %d sqnum %llu %s", r->lnum, + r->offs, r->len, r->deletion, r->sqnum, DBGKEY(&r->key)); /* Set c->replay_sqnum to help deal with dangling branches. */ c->replay_sqnum = r->sqnum; - if (r->flags & REPLAY_REF) - err = set_bud_lprops(c, r); - else if (is_hash_key(c, &r->key)) { - if (deletion) + if (is_hash_key(c, &r->key)) { + if (r->deletion) err = ubifs_tnc_remove_nm(c, &r->key, &r->nm); else err = ubifs_tnc_add_nm(c, &r->key, r->lnum, r->offs, r->len, &r->nm); } else { - if (deletion) + if (r->deletion) switch (key_type(c, &r->key)) { case UBIFS_INO_KEY: { @@ -240,7 +257,7 @@ static int apply_replay_entry(struct ubifs_info *c, struct replay_entry *r) return err; if (c->need_recovery) - err = ubifs_recover_size_accum(c, &r->key, deletion, + err = ubifs_recover_size_accum(c, &r->key, r->deletion, r->new_size); } @@ -248,68 +265,77 @@ static int apply_replay_entry(struct ubifs_info *c, struct replay_entry *r) } /** - * destroy_replay_tree - destroy the replay. - * @c: UBIFS file-system description object + * replay_entries_cmp - compare 2 replay entries. + * @priv: UBIFS file-system description object + * @a: first replay entry + * @a: second replay entry * - * Destroy the replay tree. + * This is a comparios function for 'list_sort()' which compares 2 replay + * entries @a and @b by comparing their sequence numer. Returns %1 if @a has + * greater sequence number and %-1 otherwise. */ -static void destroy_replay_tree(struct ubifs_info *c) +static int replay_entries_cmp(void *priv, struct list_head *a, + struct list_head *b) { - struct rb_node *this = c->replay_tree.rb_node; - struct replay_entry *r; - - while (this) { - if (this->rb_left) { - this = this->rb_left; - continue; - } else if (this->rb_right) { - this = this->rb_right; - continue; - } - r = rb_entry(this, struct replay_entry, rb); - this = rb_parent(this); - if (this) { - if (this->rb_left == &r->rb) - this->rb_left = NULL; - else - this->rb_right = NULL; - } - if (is_hash_key(c, &r->key)) - kfree(r->nm.name); - kfree(r); - } - c->replay_tree = RB_ROOT; + struct replay_entry *ra, *rb; + + cond_resched(); + if (a == b) + return 0; + + ra = list_entry(a, struct replay_entry, list); + rb = list_entry(b, struct replay_entry, list); + ubifs_assert(ra->sqnum != rb->sqnum); + if (ra->sqnum > rb->sqnum) + return 1; + return -1; } /** - * apply_replay_tree - apply the replay tree to the TNC. + * apply_replay_list - apply the replay list to the TNC. * @c: UBIFS file-system description object * - * Apply the replay tree. - * Returns zero in case of success and a negative error code in case of - * failure. + * Apply all entries in the replay list to the TNC. Returns zero in case of + * success and a negative error code in case of failure. */ -static int apply_replay_tree(struct ubifs_info *c) +static int apply_replay_list(struct ubifs_info *c) { - struct rb_node *this = rb_first(&c->replay_tree); + struct replay_entry *r; + int err; - while (this) { - struct replay_entry *r; - int err; + list_sort(c, &c->replay_list, &replay_entries_cmp); + list_for_each_entry(r, &c->replay_list, list) { cond_resched(); - r = rb_entry(this, struct replay_entry, rb); err = apply_replay_entry(c, r); if (err) return err; - this = rb_next(this); } + return 0; } /** - * insert_node - insert a node to the replay tree. + * destroy_replay_list - destroy the replay. + * @c: UBIFS file-system description object + * + * Destroy the replay list. + */ +static void destroy_replay_list(struct ubifs_info *c) +{ + struct replay_entry *r, *tmp; + + list_for_each_entry_safe(r, tmp, &c->replay_list, list) { + if (is_hash_key(c, &r->key)) + kfree(r->nm.name); + list_del(&r->list); + kfree(r); + } +} + +/** + * insert_node - insert a node to the replay list * @c: UBIFS file-system description object * @lnum: node logical eraseblock number * @offs: node offset @@ -321,39 +347,25 @@ static int apply_replay_tree(struct ubifs_info *c) * @old_size: truncation old size * @new_size: truncation new size * - * This function inserts a scanned non-direntry node to the replay tree. The - * replay tree is an RB-tree containing @struct replay_entry elements which are - * indexed by the sequence number. The replay tree is applied at the very end - * of the replay process. Since the tree is sorted in sequence number order, - * the older modifications are applied first. This function returns zero in - * case of success and a negative error code in case of failure. + * This function inserts a scanned non-direntry node to the replay list. The + * replay list contains @struct replay_entry elements, and we sort this list in + * sequence number order before applying it. The replay list is applied at the + * very end of the replay process. Since the list is sorted in sequence number + * order, the older modifications are applied first. This function returns zero + * in case of success and a negative error code in case of failure. */ static int insert_node(struct ubifs_info *c, int lnum, int offs, int len, union ubifs_key *key, unsigned long long sqnum, int deletion, int *used, loff_t old_size, loff_t new_size) { - struct rb_node **p = &c->replay_tree.rb_node, *parent = NULL; struct replay_entry *r; + dbg_mnt("add LEB %d:%d, key %s", lnum, offs, DBGKEY(key)); + if (key_inum(c, key) >= c->highest_inum) c->highest_inum = key_inum(c, key); - dbg_mnt("add LEB %d:%d, key %s", lnum, offs, DBGKEY(key)); - while (*p) { - parent = *p; - r = rb_entry(parent, struct replay_entry, rb); - if (sqnum < r->sqnum) { - p = &(*p)->rb_left; - continue; - } else if (sqnum > r->sqnum) { - p = &(*p)->rb_right; - continue; - } - ubifs_err("duplicate sqnum in replay"); - return -EINVAL; - } - r = kzalloc(sizeof(struct replay_entry), GFP_KERNEL); if (!r) return -ENOMEM; @@ -363,19 +375,18 @@ static int insert_node(struct ubifs_info *c, int lnum, int offs, int len, r->lnum = lnum; r->offs = offs; r->len = len; + r->deletion = !!deletion; r->sqnum = sqnum; - r->flags = (deletion ? REPLAY_DELETION : 0); + key_copy(c, key, &r->key); r->old_size = old_size; r->new_size = new_size; - key_copy(c, key, &r->key); - rb_link_node(&r->rb, parent, p); - rb_insert_color(&r->rb, &c->replay_tree); + list_add_tail(&r->list, &c->replay_list); return 0; } /** - * insert_dent - insert a directory entry node into the replay tree. + * insert_dent - insert a directory entry node into the replay list. * @c: UBIFS file-system description object * @lnum: node logical eraseblock number * @offs: node offset @@ -387,43 +398,25 @@ static int insert_node(struct ubifs_info *c, int lnum, int offs, int len, * @deletion: non-zero if this is a deletion * @used: number of bytes in use in a LEB * - * This function inserts a scanned directory entry node to the replay tree. - * Returns zero in case of success and a negative error code in case of - * failure. - * - * This function is also used for extended attribute entries because they are - * implemented as directory entry nodes. + * This function inserts a scanned directory entry node or an extended + * attribute entry to the replay list. Returns zero in case of success and a + * negative error code in case of failure. */ static int insert_dent(struct ubifs_info *c, int lnum, int offs, int len, union ubifs_key *key, const char *name, int nlen, unsigned long long sqnum, int deletion, int *used) { - struct rb_node **p = &c->replay_tree.rb_node, *parent = NULL; struct replay_entry *r; char *nbuf; + dbg_mnt("add LEB %d:%d, key %s", lnum, offs, DBGKEY(key)); if (key_inum(c, key) >= c->highest_inum) c->highest_inum = key_inum(c, key); - dbg_mnt("add LEB %d:%d, key %s", lnum, offs, DBGKEY(key)); - while (*p) { - parent = *p; - r = rb_entry(parent, struct replay_entry, rb); - if (sqnum < r->sqnum) { - p = &(*p)->rb_left; - continue; - } - if (sqnum > r->sqnum) { - p = &(*p)->rb_right; - continue; - } - ubifs_err("duplicate sqnum in replay"); - return -EINVAL; - } - r = kzalloc(sizeof(struct replay_entry), GFP_KERNEL); if (!r) return -ENOMEM; + nbuf = kmalloc(nlen + 1, GFP_KERNEL); if (!nbuf) { kfree(r); @@ -435,17 +428,15 @@ static int insert_dent(struct ubifs_info *c, int lnum, int offs, int len, r->lnum = lnum; r->offs = offs; r->len = len; + r->deletion = !!deletion; r->sqnum = sqnum; + key_copy(c, key, &r->key); r->nm.len = nlen; memcpy(nbuf, name, nlen); nbuf[nlen] = '\0'; r->nm.name = nbuf; - r->flags = (deletion ? REPLAY_DELETION : 0); - key_copy(c, key, &r->key); - ubifs_assert(!*p); - rb_link_node(&r->rb, parent, p); - rb_insert_color(&r->rb, &c->replay_tree); + list_add_tail(&r->list, &c->replay_list); return 0; } @@ -482,29 +473,92 @@ int ubifs_validate_entry(struct ubifs_info *c, } /** + * is_last_bud - check if the bud is the last in the journal head. + * @c: UBIFS file-system description object + * @bud: bud description object + * + * This function checks if bud @bud is the last bud in its journal head. This + * information is then used by 'replay_bud()' to decide whether the bud can + * have corruptions or not. Indeed, only last buds can be corrupted by power + * cuts. Returns %1 if this is the last bud, and %0 if not. + */ +static int is_last_bud(struct ubifs_info *c, struct ubifs_bud *bud) +{ + struct ubifs_jhead *jh = &c->jheads[bud->jhead]; + struct ubifs_bud *next; + uint32_t data; + int err; + + if (list_is_last(&bud->list, &jh->buds_list)) + return 1; + + /* + * The following is a quirk to make sure we work correctly with UBIFS + * images used with older UBIFS. + * + * Normally, the last bud will be the last in the journal head's list + * of bud. However, there is one exception if the UBIFS image belongs + * to older UBIFS. This is fairly unlikely: one would need to use old + * UBIFS, then have a power cut exactly at the right point, and then + * try to mount this image with new UBIFS. + * + * The exception is: it is possible to have 2 buds A and B, A goes + * before B, and B is the last, bud B is contains no data, and bud A is + * corrupted at the end. The reason is that in older versions when the + * journal code switched the next bud (from A to B), it first added a + * log reference node for the new bud (B), and only after this it + * synchronized the write-buffer of current bud (A). But later this was + * changed and UBIFS started to always synchronize the write-buffer of + * the bud (A) before writing the log reference for the new bud (B). + * + * But because older UBIFS always synchronized A's write-buffer before + * writing to B, we can recognize this exceptional situation but + * checking the contents of bud B - if it is empty, then A can be + * treated as the last and we can recover it. + * + * TODO: remove this piece of code in a couple of years (today it is + * 16.05.2011). + */ + next = list_entry(bud->list.next, struct ubifs_bud, list); + if (!list_is_last(&next->list, &jh->buds_list)) + return 0; + + err = ubi_read(c->ubi, next->lnum, (char *)&data, + next->start, 4); + if (err) + return 0; + + return data == 0xFFFFFFFF; +} + +/** * replay_bud - replay a bud logical eraseblock. * @c: UBIFS file-system description object - * @lnum: bud logical eraseblock number to replay - * @offs: bud start offset - * @jhead: journal head to which this bud belongs - * @free: amount of free space in the bud is returned here - * @dirty: amount of dirty space from padding and deletion nodes is returned - * here + * @b: bud entry which describes the bud * - * This function returns zero in case of success and a negative error code in - * case of failure. + * This function replays bud @bud, recovers it if needed, and adds all nodes + * from this bud to the replay list. Returns zero in case of success and a + * negative error code in case of failure. */ -static int replay_bud(struct ubifs_info *c, int lnum, int offs, int jhead, - int *free, int *dirty) +static int replay_bud(struct ubifs_info *c, struct bud_entry *b) { - int err = 0, used = 0; + int is_last = is_last_bud(c, b->bud); + int err = 0, used = 0, lnum = b->bud->lnum, offs = b->bud->start; struct ubifs_scan_leb *sleb; struct ubifs_scan_node *snod; - struct ubifs_bud *bud; - dbg_mnt("replay bud LEB %d, head %d", lnum, jhead); - if (c->need_recovery) - sleb = ubifs_recover_leb(c, lnum, offs, c->sbuf, jhead != GCHD); + dbg_mnt("replay bud LEB %d, head %d, offs %d, is_last %d", + lnum, b->bud->jhead, offs, is_last); + + if (c->need_recovery && is_last) + /* + * Recover only last LEBs in the journal heads, because power + * cuts may cause corruptions only in these LEBs, because only + * these LEBs could possibly be written to at the power cut + * time. + */ + sleb = ubifs_recover_leb(c, lnum, offs, c->sbuf, + b->bud->jhead != GCHD); else sleb = ubifs_scan(c, lnum, offs, c->sbuf, 0); if (IS_ERR(sleb)) @@ -620,19 +674,13 @@ static int replay_bud(struct ubifs_info *c, int lnum, int offs, int jhead, goto out; } - bud = ubifs_search_bud(c, lnum); - if (!bud) - BUG(); - + ubifs_assert(ubifs_search_bud(c, lnum)); ubifs_assert(sleb->endpt - offs >= used); ubifs_assert(sleb->endpt % c->min_io_size == 0); - if (sleb->endpt + c->min_io_size <= c->leb_size && !c->ro_mount) - err = ubifs_wbuf_seek_nolock(&c->jheads[jhead].wbuf, lnum, - sleb->endpt, UBI_SHORTTERM); - - *dirty = sleb->endpt - offs - used; - *free = c->leb_size - sleb->endpt; + b->dirty = sleb->endpt - offs - used; + b->free = c->leb_size - sleb->endpt; + dbg_mnt("bud LEB %d replied: dirty %d, free %d", lnum, b->dirty, b->free); out: ubifs_scan_destroy(sleb); @@ -646,55 +694,6 @@ out_dump: } /** - * insert_ref_node - insert a reference node to the replay tree. - * @c: UBIFS file-system description object - * @lnum: node logical eraseblock number - * @offs: node offset - * @sqnum: sequence number - * @free: amount of free space in bud - * @dirty: amount of dirty space from padding and deletion nodes - * - * This function inserts a reference node to the replay tree and returns zero - * in case of success or a negative error code in case of failure. - */ -static int insert_ref_node(struct ubifs_info *c, int lnum, int offs, - unsigned long long sqnum, int free, int dirty) -{ - struct rb_node **p = &c->replay_tree.rb_node, *parent = NULL; - struct replay_entry *r; - - dbg_mnt("add ref LEB %d:%d", lnum, offs); - while (*p) { - parent = *p; - r = rb_entry(parent, struct replay_entry, rb); - if (sqnum < r->sqnum) { - p = &(*p)->rb_left; - continue; - } else if (sqnum > r->sqnum) { - p = &(*p)->rb_right; - continue; - } - ubifs_err("duplicate sqnum in replay tree"); - return -EINVAL; - } - - r = kzalloc(sizeof(struct replay_entry), GFP_KERNEL); - if (!r) - return -ENOMEM; - - r->lnum = lnum; - r->offs = offs; - r->sqnum = sqnum; - r->flags = REPLAY_REF; - r->free = free; - r->dirty = dirty; - - rb_link_node(&r->rb, parent, p); - rb_insert_color(&r->rb, &c->replay_tree); - return 0; -} - -/** * replay_buds - replay all buds. * @c: UBIFS file-system description object * @@ -704,17 +703,16 @@ static int insert_ref_node(struct ubifs_info *c, int lnum, int offs, static int replay_buds(struct ubifs_info *c) { struct bud_entry *b; - int err, uninitialized_var(free), uninitialized_var(dirty); + int err; + unsigned long long prev_sqnum = 0; list_for_each_entry(b, &c->replay_buds, list) { - err = replay_bud(c, b->bud->lnum, b->bud->start, b->bud->jhead, - &free, &dirty); - if (err) - return err; - err = insert_ref_node(c, b->bud->lnum, b->bud->start, b->sqnum, - free, dirty); + err = replay_bud(c, b); if (err) return err; + + ubifs_assert(b->sqnum > prev_sqnum); + prev_sqnum = b->sqnum; } return 0; @@ -1054,25 +1052,29 @@ int ubifs_replay_journal(struct ubifs_info *c) if (err) goto out; - err = apply_replay_tree(c); + err = apply_replay_list(c); + if (err) + goto out; + + err = set_buds_lprops(c); if (err) goto out; /* - * UBIFS budgeting calculations use @c->budg_uncommitted_idx variable - * to roughly estimate index growth. Things like @c->min_idx_lebs + * UBIFS budgeting calculations use @c->bi.uncommitted_idx variable + * to roughly estimate index growth. Things like @c->bi.min_idx_lebs * depend on it. This means we have to initialize it to make sure * budgeting works properly. */ - c->budg_uncommitted_idx = atomic_long_read(&c->dirty_zn_cnt); - c->budg_uncommitted_idx *= c->max_idx_node_sz; + c->bi.uncommitted_idx = atomic_long_read(&c->dirty_zn_cnt); + c->bi.uncommitted_idx *= c->max_idx_node_sz; ubifs_assert(c->bud_bytes <= c->max_bud_bytes || c->need_recovery); dbg_mnt("finished, log head LEB %d:%d, max_sqnum %llu, " "highest_inum %lu", c->lhead_lnum, c->lhead_offs, c->max_sqnum, (unsigned long)c->highest_inum); out: - destroy_replay_tree(c); + destroy_replay_list(c); destroy_bud_list(c); c->replaying = 0; return err; diff --git a/fs/ubifs/sb.c b/fs/ubifs/sb.c index bf31b4729e51..c606f010e8df 100644 --- a/fs/ubifs/sb.c +++ b/fs/ubifs/sb.c @@ -475,7 +475,8 @@ failed: * @c: UBIFS file-system description object * * This function returns a pointer to the superblock node or a negative error - * code. + * code. Note, the user of this function is responsible of kfree()'ing the + * returned superblock buffer. */ struct ubifs_sb_node *ubifs_read_sb_node(struct ubifs_info *c) { @@ -616,6 +617,7 @@ int ubifs_read_superblock(struct ubifs_info *c) c->vfs_sb->s_time_gran = le32_to_cpu(sup->time_gran); memcpy(&c->uuid, &sup->uuid, 16); c->big_lpt = !!(sup_flags & UBIFS_FLG_BIGLPT); + c->space_fixup = !!(sup_flags & UBIFS_FLG_SPACE_FIXUP); /* Automatically increase file system size to the maximum size */ c->old_leb_cnt = c->leb_cnt; @@ -650,3 +652,152 @@ out: kfree(sup); return err; } + +/** + * fixup_leb - fixup/unmap an LEB containing free space. + * @c: UBIFS file-system description object + * @lnum: the LEB number to fix up + * @len: number of used bytes in LEB (starting at offset 0) + * + * This function reads the contents of the given LEB number @lnum, then fixes + * it up, so that empty min. I/O units in the end of LEB are actually erased on + * flash (rather than being just all-0xff real data). If the LEB is completely + * empty, it is simply unmapped. + */ +static int fixup_leb(struct ubifs_info *c, int lnum, int len) +{ + int err; + + ubifs_assert(len >= 0); + ubifs_assert(len % c->min_io_size == 0); + ubifs_assert(len < c->leb_size); + + if (len == 0) { + dbg_mnt("unmap empty LEB %d", lnum); + return ubi_leb_unmap(c->ubi, lnum); + } + + dbg_mnt("fixup LEB %d, data len %d", lnum, len); + err = ubi_read(c->ubi, lnum, c->sbuf, 0, len); + if (err) + return err; + + return ubi_leb_change(c->ubi, lnum, c->sbuf, len, UBI_UNKNOWN); +} + +/** + * fixup_free_space - find & remap all LEBs containing free space. + * @c: UBIFS file-system description object + * + * This function walks through all LEBs in the filesystem and fiexes up those + * containing free/empty space. + */ +static int fixup_free_space(struct ubifs_info *c) +{ + int lnum, err = 0; + struct ubifs_lprops *lprops; + + ubifs_get_lprops(c); + + /* Fixup LEBs in the master area */ + for (lnum = UBIFS_MST_LNUM; lnum < UBIFS_LOG_LNUM; lnum++) { + err = fixup_leb(c, lnum, c->mst_offs + c->mst_node_alsz); + if (err) + goto out; + } + + /* Unmap unused log LEBs */ + lnum = ubifs_next_log_lnum(c, c->lhead_lnum); + while (lnum != c->ltail_lnum) { + err = fixup_leb(c, lnum, 0); + if (err) + goto out; + lnum = ubifs_next_log_lnum(c, lnum); + } + + /* Fixup the current log head */ + err = fixup_leb(c, c->lhead_lnum, c->lhead_offs); + if (err) + goto out; + + /* Fixup LEBs in the LPT area */ + for (lnum = c->lpt_first; lnum <= c->lpt_last; lnum++) { + int free = c->ltab[lnum - c->lpt_first].free; + + if (free > 0) { + err = fixup_leb(c, lnum, c->leb_size - free); + if (err) + goto out; + } + } + + /* Unmap LEBs in the orphans area */ + for (lnum = c->orph_first; lnum <= c->orph_last; lnum++) { + err = fixup_leb(c, lnum, 0); + if (err) + goto out; + } + + /* Fixup LEBs in the main area */ + for (lnum = c->main_first; lnum < c->leb_cnt; lnum++) { + lprops = ubifs_lpt_lookup(c, lnum); + if (IS_ERR(lprops)) { + err = PTR_ERR(lprops); + goto out; + } + + if (lprops->free > 0) { + err = fixup_leb(c, lnum, c->leb_size - lprops->free); + if (err) + goto out; + } + } + +out: + ubifs_release_lprops(c); + return err; +} + +/** + * ubifs_fixup_free_space - find & fix all LEBs with free space. + * @c: UBIFS file-system description object + * + * This function fixes up LEBs containing free space on first mount, if the + * appropriate flag was set when the FS was created. Each LEB with one or more + * empty min. I/O unit (i.e. free-space-count > 0) is re-written, to make sure + * the free space is actually erased. E.g., this is necessary for some NAND + * chips, since the free space may have been programmed like real "0xff" data + * (generating a non-0xff ECC), causing future writes to the not-really-erased + * NAND pages to behave badly. After the space is fixed up, the superblock flag + * is cleared, so that this is skipped for all future mounts. + */ +int ubifs_fixup_free_space(struct ubifs_info *c) +{ + int err; + struct ubifs_sb_node *sup; + + ubifs_assert(c->space_fixup); + ubifs_assert(!c->ro_mount); + + ubifs_msg("start fixing up free space"); + + err = fixup_free_space(c); + if (err) + return err; + + sup = ubifs_read_sb_node(c); + if (IS_ERR(sup)) + return PTR_ERR(sup); + + /* Free-space fixup is no longer required */ + c->space_fixup = 0; + sup->flags &= cpu_to_le32(~UBIFS_FLG_SPACE_FIXUP); + + err = ubifs_write_sb_node(c, sup); + kfree(sup); + if (err) + return err; + + ubifs_msg("free space fixup complete"); + return err; +} diff --git a/fs/ubifs/shrinker.c b/fs/ubifs/shrinker.c index 46961c003236..ca953a945029 100644 --- a/fs/ubifs/shrinker.c +++ b/fs/ubifs/shrinker.c @@ -277,8 +277,9 @@ static int kick_a_thread(void) return 0; } -int ubifs_shrinker(struct shrinker *shrink, int nr, gfp_t gfp_mask) +int ubifs_shrinker(struct shrinker *shrink, struct shrink_control *sc) { + int nr = sc->nr_to_scan; int freed, contention = 0; long clean_zn_cnt = atomic_long_read(&ubifs_clean_zn_cnt); diff --git a/fs/ubifs/super.c b/fs/ubifs/super.c index 6ddd9973e681..1ab0d22e4c94 100644 --- a/fs/ubifs/super.c +++ b/fs/ubifs/super.c @@ -375,14 +375,14 @@ out: ubifs_release_dirty_inode_budget(c, ui); else { /* We've deleted something - clean the "no space" flags */ - c->nospace = c->nospace_rp = 0; + c->bi.nospace = c->bi.nospace_rp = 0; smp_wmb(); } done: end_writeback(inode); } -static void ubifs_dirty_inode(struct inode *inode) +static void ubifs_dirty_inode(struct inode *inode, int flags) { struct ubifs_inode *ui = ubifs_inode(inode); @@ -694,11 +694,11 @@ static int init_constants_sb(struct ubifs_info *c) * be compressed and direntries are of the maximum size. * * Note, data, which may be stored in inodes is budgeted separately, so - * it is not included into 'c->inode_budget'. + * it is not included into 'c->bi.inode_budget'. */ - c->page_budget = UBIFS_MAX_DATA_NODE_SZ * UBIFS_BLOCKS_PER_PAGE; - c->inode_budget = UBIFS_INO_NODE_SZ; - c->dent_budget = UBIFS_MAX_DENT_NODE_SZ; + c->bi.page_budget = UBIFS_MAX_DATA_NODE_SZ * UBIFS_BLOCKS_PER_PAGE; + c->bi.inode_budget = UBIFS_INO_NODE_SZ; + c->bi.dent_budget = UBIFS_MAX_DENT_NODE_SZ; /* * When the amount of flash space used by buds becomes @@ -742,7 +742,7 @@ static void init_constants_master(struct ubifs_info *c) { long long tmp64; - c->min_idx_lebs = ubifs_calc_min_idx_lebs(c); + c->bi.min_idx_lebs = ubifs_calc_min_idx_lebs(c); c->report_rp_size = ubifs_reported_space(c, c->rp_size); /* @@ -1144,8 +1144,8 @@ static int check_free_space(struct ubifs_info *c) { ubifs_assert(c->dark_wm > 0); if (c->lst.total_free + c->lst.total_dirty < c->dark_wm) { - ubifs_err("insufficient free space to mount in read/write mode"); - dbg_dump_budg(c); + ubifs_err("insufficient free space to mount in R/W mode"); + dbg_dump_budg(c, &c->bi); dbg_dump_lprops(c); return -ENOSPC; } @@ -1257,12 +1257,12 @@ static int mount_ubifs(struct ubifs_info *c) goto out_free; } + err = alloc_wbufs(c); + if (err) + goto out_cbuf; + sprintf(c->bgt_name, BGT_NAME_PATTERN, c->vi.ubi_num, c->vi.vol_id); if (!c->ro_mount) { - err = alloc_wbufs(c); - if (err) - goto out_cbuf; - /* Create background thread */ c->bgt = kthread_create(ubifs_bg_thread, c, "%s", c->bgt_name); if (IS_ERR(c->bgt)) { @@ -1304,7 +1304,7 @@ static int mount_ubifs(struct ubifs_info *c) if (err) goto out_lpt; - err = dbg_check_idx_size(c, c->old_idx_sz); + err = dbg_check_idx_size(c, c->bi.old_idx_sz); if (err) goto out_lpt; @@ -1313,7 +1313,7 @@ static int mount_ubifs(struct ubifs_info *c) goto out_journal; /* Calculate 'min_idx_lebs' after journal replay */ - c->min_idx_lebs = ubifs_calc_min_idx_lebs(c); + c->bi.min_idx_lebs = ubifs_calc_min_idx_lebs(c); err = ubifs_mount_orphans(c, c->need_recovery, c->ro_mount); if (err) @@ -1396,6 +1396,12 @@ static int mount_ubifs(struct ubifs_info *c) } else ubifs_assert(c->lst.taken_empty_lebs > 0); + if (!c->ro_mount && c->space_fixup) { + err = ubifs_fixup_free_space(c); + if (err) + goto out_infos; + } + err = dbg_check_filesystem(c); if (err) goto out_infos; @@ -1442,7 +1448,8 @@ static int mount_ubifs(struct ubifs_info *c) c->main_lebs, c->main_first, c->leb_cnt - 1); dbg_msg("index LEBs: %d", c->lst.idx_lebs); dbg_msg("total index bytes: %lld (%lld KiB, %lld MiB)", - c->old_idx_sz, c->old_idx_sz >> 10, c->old_idx_sz >> 20); + c->bi.old_idx_sz, c->bi.old_idx_sz >> 10, + c->bi.old_idx_sz >> 20); dbg_msg("key hash type: %d", c->key_hash_type); dbg_msg("tree fanout: %d", c->fanout); dbg_msg("reserved GC LEB: %d", c->gc_lnum); @@ -1456,7 +1463,7 @@ static int mount_ubifs(struct ubifs_info *c) dbg_msg("node sizes: ref %zu, cmt. start %zu, orph %zu", UBIFS_REF_NODE_SZ, UBIFS_CS_NODE_SZ, UBIFS_ORPH_NODE_SZ); dbg_msg("max. node sizes: data %zu, inode %zu dentry %zu, idx %d", - UBIFS_MAX_DATA_NODE_SZ, UBIFS_MAX_INO_NODE_SZ, + UBIFS_MAX_DATA_NODE_SZ, UBIFS_MAX_INO_NODE_SZ, UBIFS_MAX_DENT_NODE_SZ, ubifs_idx_node_sz(c, c->fanout)); dbg_msg("dead watermark: %d", c->dead_wm); dbg_msg("dark watermark: %d", c->dark_wm); @@ -1568,6 +1575,7 @@ static int ubifs_remount_rw(struct ubifs_info *c) mutex_lock(&c->umount_mutex); dbg_save_space_info(c); c->remounting_rw = 1; + c->ro_mount = 0; err = check_free_space(c); if (err) @@ -1583,6 +1591,7 @@ static int ubifs_remount_rw(struct ubifs_info *c) } sup->leb_cnt = cpu_to_le32(c->leb_cnt); err = ubifs_write_sb_node(c, sup); + kfree(sup); if (err) goto out; } @@ -1630,12 +1639,6 @@ static int ubifs_remount_rw(struct ubifs_info *c) if (err) goto out; - err = alloc_wbufs(c); - if (err) - goto out; - - ubifs_create_buds_lists(c); - /* Create background thread */ c->bgt = kthread_create(ubifs_bg_thread, c, "%s", c->bgt_name); if (IS_ERR(c->bgt)) { @@ -1670,19 +1673,37 @@ static int ubifs_remount_rw(struct ubifs_info *c) if (err) goto out; + dbg_gen("re-mounted read-write"); + c->remounting_rw = 0; + if (c->need_recovery) { c->need_recovery = 0; ubifs_msg("deferred recovery completed"); + } else { + /* + * Do not run the debugging space check if the were doing + * recovery, because when we saved the information we had the + * file-system in a state where the TNC and lprops has been + * modified in memory, but all the I/O operations (including a + * commit) were deferred. So the file-system was in + * "non-committed" state. Now the file-system is in committed + * state, and of course the amount of free space will change + * because, for example, the old index size was imprecise. + */ + err = dbg_check_space_info(c); + } + + if (c->space_fixup) { + err = ubifs_fixup_free_space(c); + if (err) + goto out; } - dbg_gen("re-mounted read-write"); - c->ro_mount = 0; - c->remounting_rw = 0; - err = dbg_check_space_info(c); mutex_unlock(&c->umount_mutex); return err; out: + c->ro_mount = 1; vfree(c->orph_buf); c->orph_buf = NULL; if (c->bgt) { @@ -1732,7 +1753,6 @@ static void ubifs_remount_ro(struct ubifs_info *c) if (err) ubifs_ro_mode(c, err); - free_wbufs(c); vfree(c->orph_buf); c->orph_buf = NULL; kfree(c->write_reserve_buf); @@ -1760,10 +1780,11 @@ static void ubifs_put_super(struct super_block *sb) * of the media. For example, there will be dirty inodes if we failed * to write them back because of I/O errors. */ - ubifs_assert(atomic_long_read(&c->dirty_pg_cnt) == 0); - ubifs_assert(c->budg_idx_growth == 0); - ubifs_assert(c->budg_dd_growth == 0); - ubifs_assert(c->budg_data_growth == 0); + if (!c->ro_error) { + ubifs_assert(c->bi.idx_growth == 0); + ubifs_assert(c->bi.dd_growth == 0); + ubifs_assert(c->bi.data_growth == 0); + } /* * The 'c->umount_lock' prevents races between UBIFS memory shrinker diff --git a/fs/ubifs/tnc.c b/fs/ubifs/tnc.c index de485979ca39..8119b1fd8d94 100644 --- a/fs/ubifs/tnc.c +++ b/fs/ubifs/tnc.c @@ -2557,11 +2557,11 @@ int ubifs_tnc_remove_nm(struct ubifs_info *c, const union ubifs_key *key, if (err) { /* Ensure the znode is dirtied */ if (znode->cnext || !ubifs_zn_dirty(znode)) { - znode = dirty_cow_bottom_up(c, znode); - if (IS_ERR(znode)) { - err = PTR_ERR(znode); - goto out_unlock; - } + znode = dirty_cow_bottom_up(c, znode); + if (IS_ERR(znode)) { + err = PTR_ERR(znode); + goto out_unlock; + } } err = tnc_delete(c, znode, n); } diff --git a/fs/ubifs/tnc_commit.c b/fs/ubifs/tnc_commit.c index 53288e5d604e..41920f357bbf 100644 --- a/fs/ubifs/tnc_commit.c +++ b/fs/ubifs/tnc_commit.c @@ -377,15 +377,13 @@ static int layout_in_gaps(struct ubifs_info *c, int cnt) c->gap_lebs = NULL; return err; } - if (!dbg_force_in_the_gaps_enabled) { + if (dbg_force_in_the_gaps_enabled()) { /* * Do not print scary warnings if the debugging * option which forces in-the-gaps is enabled. */ - ubifs_err("out of space"); - spin_lock(&c->space_lock); - dbg_dump_budg(c); - spin_unlock(&c->space_lock); + ubifs_warn("out of space"); + dbg_dump_budg(c, &c->bi); dbg_dump_lprops(c); } /* Try to commit anyway */ @@ -796,16 +794,16 @@ int ubifs_tnc_start_commit(struct ubifs_info *c, struct ubifs_zbranch *zroot) spin_lock(&c->space_lock); /* * Although we have not finished committing yet, update size of the - * committed index ('c->old_idx_sz') and zero out the index growth + * committed index ('c->bi.old_idx_sz') and zero out the index growth * budget. It is OK to do this now, because we've reserved all the * space which is needed to commit the index, and it is save for the * budgeting subsystem to assume the index is already committed, * even though it is not. */ - ubifs_assert(c->min_idx_lebs == ubifs_calc_min_idx_lebs(c)); - c->old_idx_sz = c->calc_idx_sz; - c->budg_uncommitted_idx = 0; - c->min_idx_lebs = ubifs_calc_min_idx_lebs(c); + ubifs_assert(c->bi.min_idx_lebs == ubifs_calc_min_idx_lebs(c)); + c->bi.old_idx_sz = c->calc_idx_sz; + c->bi.uncommitted_idx = 0; + c->bi.min_idx_lebs = ubifs_calc_min_idx_lebs(c); spin_unlock(&c->space_lock); mutex_unlock(&c->tnc_mutex); diff --git a/fs/ubifs/ubifs-media.h b/fs/ubifs/ubifs-media.h index 191ca7863fe7..e24380cf46ed 100644 --- a/fs/ubifs/ubifs-media.h +++ b/fs/ubifs/ubifs-media.h @@ -408,9 +408,11 @@ enum { * Superblock flags. * * UBIFS_FLG_BIGLPT: if "big" LPT model is used if set + * UBIFS_FLG_SPACE_FIXUP: first-mount "fixup" of free space within LEBs needed */ enum { UBIFS_FLG_BIGLPT = 0x02, + UBIFS_FLG_SPACE_FIXUP = 0x04, }; /** @@ -434,7 +436,7 @@ struct ubifs_ch { __u8 node_type; __u8 group_type; __u8 padding[2]; -} __attribute__ ((packed)); +} __packed; /** * union ubifs_dev_desc - device node descriptor. @@ -448,7 +450,7 @@ struct ubifs_ch { union ubifs_dev_desc { __le32 new; __le64 huge; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_ino_node - inode node. @@ -509,7 +511,7 @@ struct ubifs_ino_node { __le16 compr_type; __u8 padding2[26]; /* Watch 'zero_ino_node_unused()' if changing! */ __u8 data[]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_dent_node - directory entry node. @@ -534,7 +536,7 @@ struct ubifs_dent_node { __le16 nlen; __u8 padding2[4]; /* Watch 'zero_dent_node_unused()' if changing! */ __u8 name[]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_data_node - data node. @@ -555,7 +557,7 @@ struct ubifs_data_node { __le16 compr_type; __u8 padding[2]; /* Watch 'zero_data_node_unused()' if changing! */ __u8 data[]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_trun_node - truncation node. @@ -575,7 +577,7 @@ struct ubifs_trun_node { __u8 padding[12]; /* Watch 'zero_trun_node_unused()' if changing! */ __le64 old_size; __le64 new_size; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_pad_node - padding node. @@ -586,7 +588,7 @@ struct ubifs_trun_node { struct ubifs_pad_node { struct ubifs_ch ch; __le32 pad_len; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_sb_node - superblock node. @@ -644,7 +646,7 @@ struct ubifs_sb_node { __u8 uuid[16]; __le32 ro_compat_version; __u8 padding2[3968]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_mst_node - master node. @@ -711,7 +713,7 @@ struct ubifs_mst_node { __le32 idx_lebs; __le32 leb_cnt; __u8 padding[344]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_ref_node - logical eraseblock reference node. @@ -727,7 +729,7 @@ struct ubifs_ref_node { __le32 offs; __le32 jhead; __u8 padding[28]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_branch - key/reference/length branch @@ -741,7 +743,7 @@ struct ubifs_branch { __le32 offs; __le32 len; __u8 key[]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_idx_node - indexing node. @@ -755,7 +757,7 @@ struct ubifs_idx_node { __le16 child_cnt; __le16 level; __u8 branches[]; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_cs_node - commit start node. @@ -765,7 +767,7 @@ struct ubifs_idx_node { struct ubifs_cs_node { struct ubifs_ch ch; __le64 cmt_no; -} __attribute__ ((packed)); +} __packed; /** * struct ubifs_orph_node - orphan node. @@ -777,6 +779,6 @@ struct ubifs_orph_node { struct ubifs_ch ch; __le64 cmt_no; __le64 inos[]; -} __attribute__ ((packed)); +} __packed; #endif /* __UBIFS_MEDIA_H__ */ diff --git a/fs/ubifs/ubifs.h b/fs/ubifs/ubifs.h index 8c40ad3c6721..a70d7b4ffb25 100644 --- a/fs/ubifs/ubifs.h +++ b/fs/ubifs/ubifs.h @@ -389,9 +389,9 @@ struct ubifs_gced_idx_leb { * The @ui_size is a "shadow" variable for @inode->i_size and UBIFS uses * @ui_size instead of @inode->i_size. The reason for this is that UBIFS cannot * make sure @inode->i_size is always changed under @ui_mutex, because it - * cannot call 'truncate_setsize()' with @ui_mutex locked, because it would deadlock - * with 'ubifs_writepage()' (see file.c). All the other inode fields are - * changed under @ui_mutex, so they do not need "shadow" fields. Note, one + * cannot call 'truncate_setsize()' with @ui_mutex locked, because it would + * deadlock with 'ubifs_writepage()' (see file.c). All the other inode fields + * are changed under @ui_mutex, so they do not need "shadow" fields. Note, one * could consider to rework locking and base it on "shadow" fields. */ struct ubifs_inode { @@ -937,6 +937,40 @@ struct ubifs_mount_opts { unsigned int compr_type:2; }; +/** + * struct ubifs_budg_info - UBIFS budgeting information. + * @idx_growth: amount of bytes budgeted for index growth + * @data_growth: amount of bytes budgeted for cached data + * @dd_growth: amount of bytes budgeted for cached data that will make + * other data dirty + * @uncommitted_idx: amount of bytes were budgeted for growth of the index, but + * which still have to be taken into account because the index + * has not been committed so far + * @old_idx_sz: size of index on flash + * @min_idx_lebs: minimum number of LEBs required for the index + * @nospace: non-zero if the file-system does not have flash space (used as + * optimization) + * @nospace_rp: the same as @nospace, but additionally means that even reserved + * pool is full + * @page_budget: budget for a page (constant, nenver changed after mount) + * @inode_budget: budget for an inode (constant, nenver changed after mount) + * @dent_budget: budget for a directory entry (constant, nenver changed after + * mount) + */ +struct ubifs_budg_info { + long long idx_growth; + long long data_growth; + long long dd_growth; + long long uncommitted_idx; + unsigned long long old_idx_sz; + int min_idx_lebs; + unsigned int nospace:1; + unsigned int nospace_rp:1; + int page_budget; + int inode_budget; + int dent_budget; +}; + struct ubifs_debug_info; /** @@ -980,6 +1014,7 @@ struct ubifs_debug_info; * @cmt_wq: wait queue to sleep on if the log is full and a commit is running * * @big_lpt: flag that LPT is too big to write whole during commit + * @space_fixup: flag indicating that free space in LEBs needs to be cleaned up * @no_chk_data_crc: do not check CRCs when reading data nodes (except during * recovery) * @bulk_read: enable bulk-reads @@ -1057,32 +1092,14 @@ struct ubifs_debug_info; * @dirty_zn_cnt: number of dirty znodes * @clean_zn_cnt: number of clean znodes * - * @budg_idx_growth: amount of bytes budgeted for index growth - * @budg_data_growth: amount of bytes budgeted for cached data - * @budg_dd_growth: amount of bytes budgeted for cached data that will make - * other data dirty - * @budg_uncommitted_idx: amount of bytes were budgeted for growth of the index, - * but which still have to be taken into account because - * the index has not been committed so far - * @space_lock: protects @budg_idx_growth, @budg_data_growth, @budg_dd_growth, - * @budg_uncommited_idx, @min_idx_lebs, @old_idx_sz, @lst, - * @nospace, and @nospace_rp; - * @min_idx_lebs: minimum number of LEBs required for the index - * @old_idx_sz: size of index on flash + * @space_lock: protects @bi and @lst + * @lst: lprops statistics + * @bi: budgeting information * @calc_idx_sz: temporary variable which is used to calculate new index size * (contains accurate new index size at end of TNC commit start) - * @lst: lprops statistics - * @nospace: non-zero if the file-system does not have flash space (used as - * optimization) - * @nospace_rp: the same as @nospace, but additionally means that even reserved - * pool is full - * - * @page_budget: budget for a page - * @inode_budget: budget for an inode - * @dent_budget: budget for a directory entry * * @ref_node_alsz: size of the LEB reference node aligned to the min. flash - * I/O unit + * I/O unit * @mst_node_alsz: master node aligned size * @min_idx_node_sz: minimum indexing node aligned on 8-bytes boundary * @max_idx_node_sz: maximum indexing node aligned on 8-bytes boundary @@ -1189,7 +1206,6 @@ struct ubifs_debug_info; * @replaying: %1 during journal replay * @mounting: %1 while mounting * @remounting_rw: %1 while re-mounting from R/O mode to R/W mode - * @replay_tree: temporary tree used during journal replay * @replay_list: temporary list used during journal replay * @replay_buds: list of buds to replay * @cs_sqnum: sequence number of first node in the log (commit start node) @@ -1238,6 +1254,7 @@ struct ubifs_info { wait_queue_head_t cmt_wq; unsigned int big_lpt:1; + unsigned int space_fixup:1; unsigned int no_chk_data_crc:1; unsigned int bulk_read:1; unsigned int default_compr:2; @@ -1308,21 +1325,10 @@ struct ubifs_info { atomic_long_t dirty_zn_cnt; atomic_long_t clean_zn_cnt; - long long budg_idx_growth; - long long budg_data_growth; - long long budg_dd_growth; - long long budg_uncommitted_idx; spinlock_t space_lock; - int min_idx_lebs; - unsigned long long old_idx_sz; - unsigned long long calc_idx_sz; struct ubifs_lp_stats lst; - unsigned int nospace:1; - unsigned int nospace_rp:1; - - int page_budget; - int inode_budget; - int dent_budget; + struct ubifs_budg_info bi; + unsigned long long calc_idx_sz; int ref_node_alsz; int mst_node_alsz; @@ -1430,7 +1436,6 @@ struct ubifs_info { unsigned int replaying:1; unsigned int mounting:1; unsigned int remounting_rw:1; - struct rb_root replay_tree; struct list_head replay_list; struct list_head replay_buds; unsigned long long cs_sqnum; @@ -1609,7 +1614,7 @@ int ubifs_tnc_start_commit(struct ubifs_info *c, struct ubifs_zbranch *zroot); int ubifs_tnc_end_commit(struct ubifs_info *c); /* shrinker.c */ -int ubifs_shrinker(struct shrinker *shrink, int nr_to_scan, gfp_t gfp_mask); +int ubifs_shrinker(struct shrinker *shrink, struct shrink_control *sc); /* commit.c */ int ubifs_bg_thread(void *info); @@ -1628,6 +1633,7 @@ int ubifs_write_master(struct ubifs_info *c); int ubifs_read_superblock(struct ubifs_info *c); struct ubifs_sb_node *ubifs_read_sb_node(struct ubifs_info *c); int ubifs_write_sb_node(struct ubifs_info *c, struct ubifs_sb_node *sup); +int ubifs_fixup_free_space(struct ubifs_info *c); /* replay.c */ int ubifs_validate_entry(struct ubifs_info *c, diff --git a/fs/ubifs/xattr.c b/fs/ubifs/xattr.c index 3299f469e712..16f19f55e63f 100644 --- a/fs/ubifs/xattr.c +++ b/fs/ubifs/xattr.c @@ -80,8 +80,8 @@ enum { SECURITY_XATTR, }; -static const struct inode_operations none_inode_operations; -static const struct file_operations none_file_operations; +static const struct inode_operations empty_iops; +static const struct file_operations empty_fops; /** * create_xattr - create an extended attribute. @@ -131,8 +131,8 @@ static int create_xattr(struct ubifs_info *c, struct inode *host, /* Re-define all operations to be "nothing" */ inode->i_mapping->a_ops = &empty_aops; - inode->i_op = &none_inode_operations; - inode->i_fop = &none_file_operations; + inode->i_op = &empty_iops; + inode->i_fop = &empty_fops; inode->i_flags |= S_SYNC | S_NOATIME | S_NOCMTIME | S_NOQUOTA; ui = ubifs_inode(inode); diff --git a/fs/ufs/balloc.c b/fs/ufs/balloc.c index 46f7a807bbc1..42694e11c23d 100644 --- a/fs/ufs/balloc.c +++ b/fs/ufs/balloc.c @@ -424,8 +424,7 @@ u64 ufs_new_fragments(struct inode *inode, void *p, u64 fragment, ufs_cpu_to_data_ptr(sb, p, result); *err = 0; UFS_I(inode)->i_lastfrag = - max_t(u32, UFS_I(inode)->i_lastfrag, - fragment + count); + max(UFS_I(inode)->i_lastfrag, fragment + count); ufs_clear_frags(inode, result + oldcount, newcount - oldcount, locked_page != NULL); } @@ -440,7 +439,8 @@ u64 ufs_new_fragments(struct inode *inode, void *p, u64 fragment, result = ufs_add_fragments (inode, tmp, oldcount, newcount, err); if (result) { *err = 0; - UFS_I(inode)->i_lastfrag = max_t(u32, UFS_I(inode)->i_lastfrag, fragment + count); + UFS_I(inode)->i_lastfrag = max(UFS_I(inode)->i_lastfrag, + fragment + count); ufs_clear_frags(inode, result + oldcount, newcount - oldcount, locked_page != NULL); unlock_super(sb); @@ -479,7 +479,8 @@ u64 ufs_new_fragments(struct inode *inode, void *p, u64 fragment, uspi->s_sbbase + result, locked_page); ufs_cpu_to_data_ptr(sb, p, result); *err = 0; - UFS_I(inode)->i_lastfrag = max_t(u32, UFS_I(inode)->i_lastfrag, fragment + count); + UFS_I(inode)->i_lastfrag = max(UFS_I(inode)->i_lastfrag, + fragment + count); unlock_super(sb); if (newcount < request) ufs_free_fragments (inode, result + newcount, request - newcount); diff --git a/fs/ufs/inode.c b/fs/ufs/inode.c index 27a4babe7df0..b4d791a83207 100644 --- a/fs/ufs/inode.c +++ b/fs/ufs/inode.c @@ -78,7 +78,7 @@ static int ufs_block_to_path(struct inode *inode, sector_t i_block, sector_t off /* * Returns the location of the fragment from - * the begining of the filesystem. + * the beginning of the filesystem. */ static u64 ufs_frag_map(struct inode *inode, sector_t frag, bool needs_lock) @@ -409,7 +409,7 @@ out: } /** - * ufs_getfrag_bloc() - `get_block_t' function, interface between UFS and + * ufs_getfrag_block() - `get_block_t' function, interface between UFS and * readpage, writepage and so on */ diff --git a/fs/ufs/super.c b/fs/ufs/super.c index 7693d6293404..3915ade6f9a8 100644 --- a/fs/ufs/super.c +++ b/fs/ufs/super.c @@ -483,9 +483,9 @@ static int ufs_parse_options (char * options, unsigned * mount_options) } /* - * Diffrent types of UFS hold fs_cstotal in different - * places, and use diffrent data structure for it. - * To make things simplier we just copy fs_cstotal to ufs_sb_private_info + * Different types of UFS hold fs_cstotal in different + * places, and use different data structure for it. + * To make things simpler we just copy fs_cstotal to ufs_sb_private_info */ static void ufs_setup_cstotal(struct super_block *sb) { diff --git a/fs/ufs/truncate.c b/fs/ufs/truncate.c index 5f821dbc0579..f04f89fbd4d9 100644 --- a/fs/ufs/truncate.c +++ b/fs/ufs/truncate.c @@ -84,7 +84,7 @@ static int ufs_trunc_direct(struct inode *inode) retry = 0; frag1 = DIRECT_FRAGMENT; - frag4 = min_t(u32, UFS_NDIR_FRAGMENT, ufsi->i_lastfrag); + frag4 = min_t(u64, UFS_NDIR_FRAGMENT, ufsi->i_lastfrag); frag2 = ((frag1 & uspi->s_fpbmask) ? ((frag1 | uspi->s_fpbmask) + 1) : frag1); frag3 = frag4 & ~uspi->s_fpbmask; block1 = block2 = 0; diff --git a/fs/xattr.c b/fs/xattr.c index a19acdb81cd1..f060663ab70c 100644 --- a/fs/xattr.c +++ b/fs/xattr.c @@ -46,18 +46,22 @@ xattr_permission(struct inode *inode, const char *name, int mask) return 0; /* - * The trusted.* namespace can only be accessed by a privileged user. + * The trusted.* namespace can only be accessed by privileged users. */ - if (!strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN)) - return (capable(CAP_SYS_ADMIN) ? 0 : -EPERM); + if (!strncmp(name, XATTR_TRUSTED_PREFIX, XATTR_TRUSTED_PREFIX_LEN)) { + if (!capable(CAP_SYS_ADMIN)) + return (mask & MAY_WRITE) ? -EPERM : -ENODATA; + return 0; + } - /* In user.* namespace, only regular files and directories can have + /* + * In the user.* namespace, only regular files and directories can have * extended attributes. For sticky directories, only the owner and - * privileged user can write attributes. + * privileged users can write attributes. */ if (!strncmp(name, XATTR_USER_PREFIX, XATTR_USER_PREFIX_LEN)) { if (!S_ISREG(inode->i_mode) && !S_ISDIR(inode->i_mode)) - return -EPERM; + return (mask & MAY_WRITE) ? -EPERM : -ENODATA; if (S_ISDIR(inode->i_mode) && (inode->i_mode & S_ISVTX) && (mask & MAY_WRITE) && !inode_owner_or_capable(inode)) return -EPERM; @@ -87,7 +91,11 @@ int __vfs_setxattr_noperm(struct dentry *dentry, const char *name, { struct inode *inode = dentry->d_inode; int error = -EOPNOTSUPP; + int issec = !strncmp(name, XATTR_SECURITY_PREFIX, + XATTR_SECURITY_PREFIX_LEN); + if (issec) + inode->i_flags &= ~S_NOSEC; if (inode->i_op->setxattr) { error = inode->i_op->setxattr(dentry, name, value, size, flags); if (!error) { @@ -95,8 +103,7 @@ int __vfs_setxattr_noperm(struct dentry *dentry, const char *name, security_inode_post_setxattr(dentry, name, value, size, flags); } - } else if (!strncmp(name, XATTR_SECURITY_PREFIX, - XATTR_SECURITY_PREFIX_LEN)) { + } else if (issec) { const char *suffix = name + XATTR_SECURITY_PREFIX_LEN; error = security_inode_setsecurity(inode, suffix, value, size, flags); @@ -666,7 +673,7 @@ generic_setxattr(struct dentry *dentry, const char *name, const void *value, siz handler = xattr_resolve_name(dentry->d_sb->s_xattr, &name); if (!handler) return -EOPNOTSUPP; - return handler->set(dentry, name, value, size, 0, handler->flags); + return handler->set(dentry, name, value, size, flags, handler->flags); } /* diff --git a/fs/xfs/linux-2.6/xfs_aops.c b/fs/xfs/linux-2.6/xfs_aops.c index 52dbd14260ba..79ce38be15a1 100644 --- a/fs/xfs/linux-2.6/xfs_aops.c +++ b/fs/xfs/linux-2.6/xfs_aops.c @@ -1295,7 +1295,7 @@ xfs_get_blocks_direct( * If the private argument is non-NULL __xfs_get_blocks signals us that we * need to issue a transaction to convert the range from unwritten to written * extents. In case this is regular synchronous I/O we just call xfs_end_io - * to do this and we are done. But in case this was a successfull AIO + * to do this and we are done. But in case this was a successful AIO * request this handler is called from interrupt context, from which we * can't start transactions. In that case offload the I/O completion to * the workqueues we also use for buffered I/O completion. diff --git a/fs/xfs/linux-2.6/xfs_buf.c b/fs/xfs/linux-2.6/xfs_buf.c index 596bb2c9de42..5e68099db2a5 100644 --- a/fs/xfs/linux-2.6/xfs_buf.c +++ b/fs/xfs/linux-2.6/xfs_buf.c @@ -33,7 +33,6 @@ #include <linux/migrate.h> #include <linux/backing-dev.h> #include <linux/freezer.h> -#include <linux/list_sort.h> #include "xfs_sb.h" #include "xfs_inum.h" @@ -120,7 +119,7 @@ xfs_buf_lru_add( * The unlocked check is safe here because it only occurs when there are not * b_lru_ref counts left on the inode under the pag->pag_buf_lock. it is there * to optimise the shrinker removing the buffer from the LRU and calling - * xfs_buf_free(). i.e. it removes an unneccessary round trip on the + * xfs_buf_free(). i.e. it removes an unnecessary round trip on the * bt_lru_lock. */ STATIC void @@ -293,7 +292,6 @@ xfs_buf_allocate_memory( size_t nbytes, offset; gfp_t gfp_mask = xb_to_gfp(flags); unsigned short page_count, i; - pgoff_t first; xfs_off_t end; int error; @@ -333,7 +331,6 @@ use_alloc_page: return error; offset = bp->b_offset; - first = bp->b_file_offset >> PAGE_SHIFT; bp->b_flags |= _XBF_PAGES; for (i = 0; i < bp->b_page_count; i++) { @@ -380,7 +377,7 @@ out_free_pages: } /* - * Map buffer into kernel address-space if nessecary. + * Map buffer into kernel address-space if necessary. */ STATIC int _xfs_buf_map_pages( @@ -657,8 +654,6 @@ xfs_buf_readahead( xfs_off_t ioff, size_t isize) { - struct backing_dev_info *bdi; - if (bdi_read_congested(target->bt_bdi)) return; @@ -713,6 +708,27 @@ xfs_buf_get_empty( return bp; } +/* + * Return a buffer allocated as an empty buffer and associated to external + * memory via xfs_buf_associate_memory() back to it's empty state. + */ +void +xfs_buf_set_empty( + struct xfs_buf *bp, + size_t len) +{ + if (bp->b_pages) + _xfs_buf_free_pages(bp); + + bp->b_pages = NULL; + bp->b_page_count = 0; + bp->b_addr = NULL; + bp->b_file_offset = 0; + bp->b_buffer_length = bp->b_count_desired = len; + bp->b_bn = XFS_BUF_DADDR_NULL; + bp->b_flags &= ~XBF_MAPPED; +} + static inline struct page * mem_to_page( void *addr) @@ -919,8 +935,6 @@ xfs_buf_lock( if (atomic_read(&bp->b_pin_count) && (bp->b_flags & XBF_STALE)) xfs_log_force(bp->b_target->bt_mount, 0); - if (atomic_read(&bp->b_io_remaining)) - blk_flush_plug(current); down(&bp->b_sema); XB_SET_OWNER(bp); @@ -1309,8 +1323,6 @@ xfs_buf_iowait( { trace_xfs_buf_iowait(bp, _RET_IP_); - if (atomic_read(&bp->b_io_remaining)) - blk_flush_plug(current); wait_for_completion(&bp->b_iowait); trace_xfs_buf_iowait_done(bp, _RET_IP_); @@ -1410,12 +1422,12 @@ restart: int xfs_buftarg_shrink( struct shrinker *shrink, - int nr_to_scan, - gfp_t mask) + struct shrink_control *sc) { struct xfs_buftarg *btp = container_of(shrink, struct xfs_buftarg, bt_shrinker); struct xfs_buf *bp; + int nr_to_scan = sc->nr_to_scan; LIST_HEAD(dispose); if (!nr_to_scan) @@ -1747,8 +1759,8 @@ xfsbufd( do { long age = xfs_buf_age_centisecs * msecs_to_jiffies(10); long tout = xfs_buf_timer_centisecs * msecs_to_jiffies(10); - int count = 0; struct list_head tmp; + struct blk_plug plug; if (unlikely(freezing(current))) { set_bit(XBT_FORCE_SLEEP, &target->bt_flags); @@ -1764,16 +1776,15 @@ xfsbufd( xfs_buf_delwri_split(target, &tmp, age); list_sort(NULL, &tmp, xfs_buf_cmp); + + blk_start_plug(&plug); while (!list_empty(&tmp)) { struct xfs_buf *bp; bp = list_first_entry(&tmp, struct xfs_buf, b_list); list_del_init(&bp->b_list); xfs_bdstrat_cb(bp); - count++; } - if (count) - blk_flush_plug(current); - + blk_finish_plug(&plug); } while (!kthread_should_stop()); return 0; @@ -1793,6 +1804,7 @@ xfs_flush_buftarg( int pincount = 0; LIST_HEAD(tmp_list); LIST_HEAD(wait_list); + struct blk_plug plug; xfs_buf_runall_queues(xfsconvertd_workqueue); xfs_buf_runall_queues(xfsdatad_workqueue); @@ -1807,6 +1819,8 @@ xfs_flush_buftarg( * we do that after issuing all the IO. */ list_sort(NULL, &tmp_list, xfs_buf_cmp); + + blk_start_plug(&plug); while (!list_empty(&tmp_list)) { bp = list_first_entry(&tmp_list, struct xfs_buf, b_list); ASSERT(target == bp->b_target); @@ -1817,10 +1831,10 @@ xfs_flush_buftarg( } xfs_bdstrat_cb(bp); } + blk_finish_plug(&plug); if (wait) { - /* Expedite and wait for IO to complete. */ - blk_flush_plug(current); + /* Wait for IO to complete. */ while (!list_empty(&wait_list)) { bp = list_first_entry(&wait_list, struct xfs_buf, b_list); diff --git a/fs/xfs/linux-2.6/xfs_buf.h b/fs/xfs/linux-2.6/xfs_buf.h index a9a1c4512645..50a7d5fb3b73 100644 --- a/fs/xfs/linux-2.6/xfs_buf.h +++ b/fs/xfs/linux-2.6/xfs_buf.h @@ -178,6 +178,7 @@ extern xfs_buf_t *xfs_buf_read(xfs_buftarg_t *, xfs_off_t, size_t, xfs_buf_flags_t); extern xfs_buf_t *xfs_buf_get_empty(size_t, xfs_buftarg_t *); +extern void xfs_buf_set_empty(struct xfs_buf *bp, size_t len); extern xfs_buf_t *xfs_buf_get_uncached(struct xfs_buftarg *, size_t, int); extern int xfs_buf_associate_memory(xfs_buf_t *, void *, size_t); extern void xfs_buf_hold(xfs_buf_t *); diff --git a/fs/xfs/linux-2.6/xfs_discard.c b/fs/xfs/linux-2.6/xfs_discard.c index d61611c88012..244e797dae32 100644 --- a/fs/xfs/linux-2.6/xfs_discard.c +++ b/fs/xfs/linux-2.6/xfs_discard.c @@ -191,3 +191,32 @@ xfs_ioc_trim( return -XFS_ERROR(EFAULT); return 0; } + +int +xfs_discard_extents( + struct xfs_mount *mp, + struct list_head *list) +{ + struct xfs_busy_extent *busyp; + int error = 0; + + list_for_each_entry(busyp, list, list) { + trace_xfs_discard_extent(mp, busyp->agno, busyp->bno, + busyp->length); + + error = -blkdev_issue_discard(mp->m_ddev_targp->bt_bdev, + XFS_AGB_TO_DADDR(mp, busyp->agno, busyp->bno), + XFS_FSB_TO_BB(mp, busyp->length), + GFP_NOFS, 0); + if (error && error != EOPNOTSUPP) { + xfs_info(mp, + "discard failed for extent [0x%llu,%u], error %d", + (unsigned long long)busyp->bno, + busyp->length, + error); + return error; + } + } + + return 0; +} diff --git a/fs/xfs/linux-2.6/xfs_discard.h b/fs/xfs/linux-2.6/xfs_discard.h index e82b6dd3e127..344879aea646 100644 --- a/fs/xfs/linux-2.6/xfs_discard.h +++ b/fs/xfs/linux-2.6/xfs_discard.h @@ -2,7 +2,9 @@ #define XFS_DISCARD_H 1 struct fstrim_range; +struct list_head; extern int xfs_ioc_trim(struct xfs_mount *, struct fstrim_range __user *); +extern int xfs_discard_extents(struct xfs_mount *, struct list_head *); #endif /* XFS_DISCARD_H */ diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index 52aadfbed132..f4213ba1ff85 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -381,7 +381,7 @@ xfs_aio_write_isize_update( /* * If this was a direct or synchronous I/O that failed (such as ENOSPC) then - * part of the I/O may have been written to disk before the error occured. In + * part of the I/O may have been written to disk before the error occurred. In * this case the on-disk file size may have been adjusted beyond the in-memory * file size and now needs to be truncated back. */ diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c index b3486dfa5520..54e623bfbb85 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.c +++ b/fs/xfs/linux-2.6/xfs_ioctl32.c @@ -586,7 +586,8 @@ xfs_file_compat_ioctl( case XFS_IOC_RESVSP_32: case XFS_IOC_UNRESVSP_32: case XFS_IOC_RESVSP64_32: - case XFS_IOC_UNRESVSP64_32: { + case XFS_IOC_UNRESVSP64_32: + case XFS_IOC_ZERO_RANGE_32: { struct xfs_flock64 bf; if (xfs_compat_flock64_copyin(&bf, arg)) diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.h b/fs/xfs/linux-2.6/xfs_ioctl32.h index 08b605792a99..80f4060e8970 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.h +++ b/fs/xfs/linux-2.6/xfs_ioctl32.h @@ -184,6 +184,7 @@ typedef struct compat_xfs_flock64 { #define XFS_IOC_UNRESVSP_32 _IOW('X', 41, struct compat_xfs_flock64) #define XFS_IOC_RESVSP64_32 _IOW('X', 42, struct compat_xfs_flock64) #define XFS_IOC_UNRESVSP64_32 _IOW('X', 43, struct compat_xfs_flock64) +#define XFS_IOC_ZERO_RANGE_32 _IOW('X', 57, struct compat_xfs_flock64) typedef struct compat_xfs_fsop_geom_v1 { __u32 blocksize; /* filesystem (data) block size */ diff --git a/fs/xfs/linux-2.6/xfs_iops.c b/fs/xfs/linux-2.6/xfs_iops.c index 9ff7fc603d2f..dd21784525a8 100644 --- a/fs/xfs/linux-2.6/xfs_iops.c +++ b/fs/xfs/linux-2.6/xfs_iops.c @@ -70,7 +70,7 @@ xfs_synchronize_times( /* * If the linux inode is valid, mark it dirty. - * Used when commiting a dirty inode into a transaction so that + * Used when committing a dirty inode into a transaction so that * the inode will get written back by the linux code */ void diff --git a/fs/xfs/linux-2.6/xfs_linux.h b/fs/xfs/linux-2.6/xfs_linux.h index 244be9cbfe78..8633521b3b2e 100644 --- a/fs/xfs/linux-2.6/xfs_linux.h +++ b/fs/xfs/linux-2.6/xfs_linux.h @@ -70,6 +70,7 @@ #include <linux/ctype.h> #include <linux/writeback.h> #include <linux/capability.h> +#include <linux/list_sort.h> #include <asm/page.h> #include <asm/div64.h> diff --git a/fs/xfs/linux-2.6/xfs_message.c b/fs/xfs/linux-2.6/xfs_message.c index 508e06fd7d1e..bd672def95ac 100644 --- a/fs/xfs/linux-2.6/xfs_message.c +++ b/fs/xfs/linux-2.6/xfs_message.c @@ -28,53 +28,32 @@ /* * XFS logging functions */ -static int +static void __xfs_printk( const char *level, const struct xfs_mount *mp, struct va_format *vaf) { - if (mp && mp->m_fsname) - return printk("%sXFS (%s): %pV\n", level, mp->m_fsname, vaf); - return printk("%sXFS: %pV\n", level, vaf); -} - -int xfs_printk( - const char *level, - const struct xfs_mount *mp, - const char *fmt, ...) -{ - struct va_format vaf; - va_list args; - int r; - - va_start(args, fmt); - - vaf.fmt = fmt; - vaf.va = &args; - - r = __xfs_printk(level, mp, &vaf); - va_end(args); - - return r; + if (mp && mp->m_fsname) { + printk("%sXFS (%s): %pV\n", level, mp->m_fsname, vaf); + return; + } + printk("%sXFS: %pV\n", level, vaf); } #define define_xfs_printk_level(func, kern_level) \ -int func(const struct xfs_mount *mp, const char *fmt, ...) \ +void func(const struct xfs_mount *mp, const char *fmt, ...) \ { \ struct va_format vaf; \ va_list args; \ - int r; \ \ va_start(args, fmt); \ \ vaf.fmt = fmt; \ vaf.va = &args; \ \ - r = __xfs_printk(kern_level, mp, &vaf); \ + __xfs_printk(kern_level, mp, &vaf); \ va_end(args); \ - \ - return r; \ } \ define_xfs_printk_level(xfs_emerg, KERN_EMERG); @@ -88,7 +67,7 @@ define_xfs_printk_level(xfs_info, KERN_INFO); define_xfs_printk_level(xfs_debug, KERN_DEBUG); #endif -int +void xfs_alert_tag( const struct xfs_mount *mp, int panic_tag, @@ -97,11 +76,9 @@ xfs_alert_tag( struct va_format vaf; va_list args; int do_panic = 0; - int r; if (xfs_panic_mask && (xfs_panic_mask & panic_tag)) { - xfs_printk(KERN_ALERT, mp, - "XFS: Transforming an alert into a BUG."); + xfs_alert(mp, "Transforming an alert into a BUG."); do_panic = 1; } @@ -110,12 +87,10 @@ xfs_alert_tag( vaf.fmt = fmt; vaf.va = &args; - r = __xfs_printk(KERN_ALERT, mp, &vaf); + __xfs_printk(KERN_ALERT, mp, &vaf); va_end(args); BUG_ON(do_panic); - - return r; } void diff --git a/fs/xfs/linux-2.6/xfs_message.h b/fs/xfs/linux-2.6/xfs_message.h index e77ffa16745b..7fb7ea007672 100644 --- a/fs/xfs/linux-2.6/xfs_message.h +++ b/fs/xfs/linux-2.6/xfs_message.h @@ -3,32 +3,33 @@ struct xfs_mount; -extern int xfs_printk(const char *level, const struct xfs_mount *mp, - const char *fmt, ...) - __attribute__ ((format (printf, 3, 4))); -extern int xfs_emerg(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_emerg(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_alert(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_alert(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_alert_tag(const struct xfs_mount *mp, int tag, +extern void xfs_alert_tag(const struct xfs_mount *mp, int tag, const char *fmt, ...) __attribute__ ((format (printf, 3, 4))); -extern int xfs_crit(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_crit(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_err(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_err(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_warn(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_warn(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_notice(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_notice(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); -extern int xfs_info(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_info(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); #ifdef DEBUG -extern int xfs_debug(const struct xfs_mount *mp, const char *fmt, ...) +extern void xfs_debug(const struct xfs_mount *mp, const char *fmt, ...) __attribute__ ((format (printf, 2, 3))); #else -#define xfs_debug(mp, fmt, ...) (0) +static inline void +__attribute__ ((format (printf, 2, 3))) +xfs_debug(const struct xfs_mount *mp, const char *fmt, ...) +{ +} #endif extern void assfail(char *expr, char *f, int l); diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index 1ba5c451da36..1e3a7ce804dc 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -110,8 +110,10 @@ mempool_t *xfs_ioend_pool; #define MNTOPT_GQUOTANOENF "gqnoenforce"/* group quota limit enforcement */ #define MNTOPT_PQUOTANOENF "pqnoenforce"/* project quota limit enforcement */ #define MNTOPT_QUOTANOENF "qnoenforce" /* same as uqnoenforce */ -#define MNTOPT_DELAYLOG "delaylog" /* Delayed loging enabled */ -#define MNTOPT_NODELAYLOG "nodelaylog" /* Delayed loging disabled */ +#define MNTOPT_DELAYLOG "delaylog" /* Delayed logging enabled */ +#define MNTOPT_NODELAYLOG "nodelaylog" /* Delayed logging disabled */ +#define MNTOPT_DISCARD "discard" /* Discard unused blocks */ +#define MNTOPT_NODISCARD "nodiscard" /* Do not discard unused blocks */ /* * Table driven mount option parser. @@ -355,6 +357,10 @@ xfs_parseargs( mp->m_flags |= XFS_MOUNT_DELAYLOG; } else if (!strcmp(this_char, MNTOPT_NODELAYLOG)) { mp->m_flags &= ~XFS_MOUNT_DELAYLOG; + } else if (!strcmp(this_char, MNTOPT_DISCARD)) { + mp->m_flags |= XFS_MOUNT_DISCARD; + } else if (!strcmp(this_char, MNTOPT_NODISCARD)) { + mp->m_flags &= ~XFS_MOUNT_DISCARD; } else if (!strcmp(this_char, "ihashsize")) { xfs_warn(mp, "ihashsize no longer used, option is deprecated."); @@ -388,6 +394,13 @@ xfs_parseargs( return EINVAL; } + if ((mp->m_flags & XFS_MOUNT_DISCARD) && + !(mp->m_flags & XFS_MOUNT_DELAYLOG)) { + xfs_warn(mp, + "the discard option is incompatible with the nodelaylog option"); + return EINVAL; + } + #ifndef CONFIG_XFS_QUOTA if (XFS_IS_QUOTA_RUNNING(mp)) { xfs_warn(mp, "quota support not available in this kernel."); @@ -488,6 +501,7 @@ xfs_showargs( { XFS_MOUNT_FILESTREAMS, "," MNTOPT_FILESTREAM }, { XFS_MOUNT_GRPID, "," MNTOPT_GRPID }, { XFS_MOUNT_DELAYLOG, "," MNTOPT_DELAYLOG }, + { XFS_MOUNT_DISCARD, "," MNTOPT_DISCARD }, { 0, NULL } }; static struct proc_xfs_info xfs_info_unset[] = { @@ -816,75 +830,6 @@ xfs_setup_devices( return 0; } -/* - * XFS AIL push thread support - */ -void -xfsaild_wakeup( - struct xfs_ail *ailp, - xfs_lsn_t threshold_lsn) -{ - /* only ever move the target forwards */ - if (XFS_LSN_CMP(threshold_lsn, ailp->xa_target) > 0) { - ailp->xa_target = threshold_lsn; - wake_up_process(ailp->xa_task); - } -} - -STATIC int -xfsaild( - void *data) -{ - struct xfs_ail *ailp = data; - xfs_lsn_t last_pushed_lsn = 0; - long tout = 0; /* milliseconds */ - - while (!kthread_should_stop()) { - /* - * for short sleeps indicating congestion, don't allow us to - * get woken early. Otherwise all we do is bang on the AIL lock - * without making progress. - */ - if (tout && tout <= 20) - __set_current_state(TASK_KILLABLE); - else - __set_current_state(TASK_INTERRUPTIBLE); - schedule_timeout(tout ? - msecs_to_jiffies(tout) : MAX_SCHEDULE_TIMEOUT); - - /* swsusp */ - try_to_freeze(); - - ASSERT(ailp->xa_mount->m_log); - if (XFS_FORCED_SHUTDOWN(ailp->xa_mount)) - continue; - - tout = xfsaild_push(ailp, &last_pushed_lsn); - } - - return 0; -} /* xfsaild */ - -int -xfsaild_start( - struct xfs_ail *ailp) -{ - ailp->xa_target = 0; - ailp->xa_task = kthread_run(xfsaild, ailp, "xfsaild/%s", - ailp->xa_mount->m_fsname); - if (IS_ERR(ailp->xa_task)) - return -PTR_ERR(ailp->xa_task); - return 0; -} - -void -xfsaild_stop( - struct xfs_ail *ailp) -{ - kthread_stop(ailp->xa_task); -} - - /* Catch misguided souls that try to use this interface on XFS */ STATIC struct inode * xfs_fs_alloc_inode( @@ -980,7 +925,8 @@ xfs_fs_inode_init_once( */ STATIC void xfs_fs_dirty_inode( - struct inode *inode) + struct inode *inode, + int flags) { barrier(); XFS_I(inode)->i_update_core = 1; @@ -1191,22 +1137,12 @@ xfs_fs_sync_fs( return -error; if (laptop_mode) { - int prev_sync_seq = mp->m_sync_seq; - /* * The disk must be active because we're syncing. * We schedule xfssyncd now (now that the disk is * active) instead of later (when it might not be). */ - wake_up_process(mp->m_sync_task); - /* - * We have to wait for the sync iteration to complete. - * If we don't, the disk activity caused by the sync - * will come after the sync is completed, and that - * triggers another sync from laptop mode. - */ - wait_event(mp->m_wait_single_sync_task, - mp->m_sync_seq != prev_sync_seq); + flush_delayed_work_sync(&mp->m_sync_work); } return 0; @@ -1490,9 +1426,6 @@ xfs_fs_fill_super( spin_lock_init(&mp->m_sb_lock); mutex_init(&mp->m_growlock); atomic_set(&mp->m_active_trans, 0); - INIT_LIST_HEAD(&mp->m_sync_list); - spin_lock_init(&mp->m_sync_lock); - init_waitqueue_head(&mp->m_wait_single_sync_task); mp->m_super = sb; sb->s_fs_info = mp; @@ -1799,6 +1732,38 @@ xfs_destroy_zones(void) } STATIC int __init +xfs_init_workqueues(void) +{ + /* + * max_active is set to 8 to give enough concurency to allow + * multiple work operations on each CPU to run. This allows multiple + * filesystems to be running sync work concurrently, and scales with + * the number of CPUs in the system. + */ + xfs_syncd_wq = alloc_workqueue("xfssyncd", WQ_CPU_INTENSIVE, 8); + if (!xfs_syncd_wq) + goto out; + + xfs_ail_wq = alloc_workqueue("xfsail", WQ_CPU_INTENSIVE, 8); + if (!xfs_ail_wq) + goto out_destroy_syncd; + + return 0; + +out_destroy_syncd: + destroy_workqueue(xfs_syncd_wq); +out: + return -ENOMEM; +} + +STATIC void +xfs_destroy_workqueues(void) +{ + destroy_workqueue(xfs_ail_wq); + destroy_workqueue(xfs_syncd_wq); +} + +STATIC int __init init_xfs_fs(void) { int error; @@ -1813,10 +1778,14 @@ init_xfs_fs(void) if (error) goto out; - error = xfs_mru_cache_init(); + error = xfs_init_workqueues(); if (error) goto out_destroy_zones; + error = xfs_mru_cache_init(); + if (error) + goto out_destroy_wq; + error = xfs_filestream_init(); if (error) goto out_mru_cache_uninit; @@ -1850,6 +1819,8 @@ init_xfs_fs(void) xfs_filestream_uninit(); out_mru_cache_uninit: xfs_mru_cache_uninit(); + out_destroy_wq: + xfs_destroy_workqueues(); out_destroy_zones: xfs_destroy_zones(); out: @@ -1866,6 +1837,7 @@ exit_xfs_fs(void) xfs_buf_terminate(); xfs_filestream_uninit(); xfs_mru_cache_uninit(); + xfs_destroy_workqueues(); xfs_destroy_zones(); } diff --git a/fs/xfs/linux-2.6/xfs_sync.c b/fs/xfs/linux-2.6/xfs_sync.c index 594cd822d84d..8ecad5ff9f9b 100644 --- a/fs/xfs/linux-2.6/xfs_sync.c +++ b/fs/xfs/linux-2.6/xfs_sync.c @@ -22,6 +22,7 @@ #include "xfs_log.h" #include "xfs_inum.h" #include "xfs_trans.h" +#include "xfs_trans_priv.h" #include "xfs_sb.h" #include "xfs_ag.h" #include "xfs_mount.h" @@ -39,6 +40,8 @@ #include <linux/kthread.h> #include <linux/freezer.h> +struct workqueue_struct *xfs_syncd_wq; /* sync workqueue */ + /* * The inode lookup is done in batches to keep the amount of lock traffic and * radix tree lookups to a minimum. The batch size is a trade off between @@ -264,6 +267,16 @@ xfs_sync_inode_attr( error = xfs_iflush(ip, flags); + /* + * We don't want to try again on non-blocking flushes that can't run + * again immediately. If an inode really must be written, then that's + * what the SYNC_WAIT flag is for. + */ + if (error == EAGAIN) { + ASSERT(!(flags & SYNC_WAIT)); + error = 0; + } + out_unlock: xfs_iunlock(ip, XFS_ILOCK_SHARED); return error; @@ -401,7 +414,7 @@ xfs_quiesce_fs( /* * Second stage of a quiesce. The data is already synced, now we have to take * care of the metadata. New transactions are already blocked, so we need to - * wait for any remaining transactions to drain out before proceding. + * wait for any remaining transactions to drain out before proceeding. */ void xfs_quiesce_attr( @@ -431,62 +444,12 @@ xfs_quiesce_attr( xfs_unmountfs_writesb(mp); } -/* - * Enqueue a work item to be picked up by the vfs xfssyncd thread. - * Doing this has two advantages: - * - It saves on stack space, which is tight in certain situations - * - It can be used (with care) as a mechanism to avoid deadlocks. - * Flushing while allocating in a full filesystem requires both. - */ -STATIC void -xfs_syncd_queue_work( - struct xfs_mount *mp, - void *data, - void (*syncer)(struct xfs_mount *, void *), - struct completion *completion) +static void +xfs_syncd_queue_sync( + struct xfs_mount *mp) { - struct xfs_sync_work *work; - - work = kmem_alloc(sizeof(struct xfs_sync_work), KM_SLEEP); - INIT_LIST_HEAD(&work->w_list); - work->w_syncer = syncer; - work->w_data = data; - work->w_mount = mp; - work->w_completion = completion; - spin_lock(&mp->m_sync_lock); - list_add_tail(&work->w_list, &mp->m_sync_list); - spin_unlock(&mp->m_sync_lock); - wake_up_process(mp->m_sync_task); -} - -/* - * Flush delayed allocate data, attempting to free up reserved space - * from existing allocations. At this point a new allocation attempt - * has failed with ENOSPC and we are in the process of scratching our - * heads, looking about for more room... - */ -STATIC void -xfs_flush_inodes_work( - struct xfs_mount *mp, - void *arg) -{ - struct inode *inode = arg; - xfs_sync_data(mp, SYNC_TRYLOCK); - xfs_sync_data(mp, SYNC_TRYLOCK | SYNC_WAIT); - iput(inode); -} - -void -xfs_flush_inodes( - xfs_inode_t *ip) -{ - struct inode *inode = VFS_I(ip); - DECLARE_COMPLETION_ONSTACK(completion); - - igrab(inode); - xfs_syncd_queue_work(ip->i_mount, inode, xfs_flush_inodes_work, &completion); - wait_for_completion(&completion); - xfs_log_force(ip->i_mount, XFS_LOG_SYNC); + queue_delayed_work(xfs_syncd_wq, &mp->m_sync_work, + msecs_to_jiffies(xfs_syncd_centisecs * 10)); } /* @@ -496,9 +459,10 @@ xfs_flush_inodes( */ STATIC void xfs_sync_worker( - struct xfs_mount *mp, - void *unused) + struct work_struct *work) { + struct xfs_mount *mp = container_of(to_delayed_work(work), + struct xfs_mount, m_sync_work); int error; if (!(mp->m_flags & XFS_MOUNT_RDONLY)) { @@ -508,73 +472,106 @@ xfs_sync_worker( error = xfs_fs_log_dummy(mp); else xfs_log_force(mp, 0); - xfs_reclaim_inodes(mp, 0); error = xfs_qm_sync(mp, SYNC_TRYLOCK); + + /* start pushing all the metadata that is currently dirty */ + xfs_ail_push_all(mp->m_ail); } - mp->m_sync_seq++; - wake_up(&mp->m_wait_single_sync_task); + + /* queue us up again */ + xfs_syncd_queue_sync(mp); } -STATIC int -xfssyncd( - void *arg) +/* + * Queue a new inode reclaim pass if there are reclaimable inodes and there + * isn't a reclaim pass already in progress. By default it runs every 5s based + * on the xfs syncd work default of 30s. Perhaps this should have it's own + * tunable, but that can be done if this method proves to be ineffective or too + * aggressive. + */ +static void +xfs_syncd_queue_reclaim( + struct xfs_mount *mp) { - struct xfs_mount *mp = arg; - long timeleft; - xfs_sync_work_t *work, *n; - LIST_HEAD (tmp); - - set_freezable(); - timeleft = xfs_syncd_centisecs * msecs_to_jiffies(10); - for (;;) { - if (list_empty(&mp->m_sync_list)) - timeleft = schedule_timeout_interruptible(timeleft); - /* swsusp */ - try_to_freeze(); - if (kthread_should_stop() && list_empty(&mp->m_sync_list)) - break; - spin_lock(&mp->m_sync_lock); - /* - * We can get woken by laptop mode, to do a sync - - * that's the (only!) case where the list would be - * empty with time remaining. - */ - if (!timeleft || list_empty(&mp->m_sync_list)) { - if (!timeleft) - timeleft = xfs_syncd_centisecs * - msecs_to_jiffies(10); - INIT_LIST_HEAD(&mp->m_sync_work.w_list); - list_add_tail(&mp->m_sync_work.w_list, - &mp->m_sync_list); - } - list_splice_init(&mp->m_sync_list, &tmp); - spin_unlock(&mp->m_sync_lock); + /* + * We can have inodes enter reclaim after we've shut down the syncd + * workqueue during unmount, so don't allow reclaim work to be queued + * during unmount. + */ + if (!(mp->m_super->s_flags & MS_ACTIVE)) + return; - list_for_each_entry_safe(work, n, &tmp, w_list) { - (*work->w_syncer)(mp, work->w_data); - list_del(&work->w_list); - if (work == &mp->m_sync_work) - continue; - if (work->w_completion) - complete(work->w_completion); - kmem_free(work); - } + rcu_read_lock(); + if (radix_tree_tagged(&mp->m_perag_tree, XFS_ICI_RECLAIM_TAG)) { + queue_delayed_work(xfs_syncd_wq, &mp->m_reclaim_work, + msecs_to_jiffies(xfs_syncd_centisecs / 6 * 10)); } + rcu_read_unlock(); +} - return 0; +/* + * This is a fast pass over the inode cache to try to get reclaim moving on as + * many inodes as possible in a short period of time. It kicks itself every few + * seconds, as well as being kicked by the inode cache shrinker when memory + * goes low. It scans as quickly as possible avoiding locked inodes or those + * already being flushed, and once done schedules a future pass. + */ +STATIC void +xfs_reclaim_worker( + struct work_struct *work) +{ + struct xfs_mount *mp = container_of(to_delayed_work(work), + struct xfs_mount, m_reclaim_work); + + xfs_reclaim_inodes(mp, SYNC_TRYLOCK); + xfs_syncd_queue_reclaim(mp); +} + +/* + * Flush delayed allocate data, attempting to free up reserved space + * from existing allocations. At this point a new allocation attempt + * has failed with ENOSPC and we are in the process of scratching our + * heads, looking about for more room. + * + * Queue a new data flush if there isn't one already in progress and + * wait for completion of the flush. This means that we only ever have one + * inode flush in progress no matter how many ENOSPC events are occurring and + * so will prevent the system from bogging down due to every concurrent + * ENOSPC event scanning all the active inodes in the system for writeback. + */ +void +xfs_flush_inodes( + struct xfs_inode *ip) +{ + struct xfs_mount *mp = ip->i_mount; + + queue_work(xfs_syncd_wq, &mp->m_flush_work); + flush_work_sync(&mp->m_flush_work); +} + +STATIC void +xfs_flush_worker( + struct work_struct *work) +{ + struct xfs_mount *mp = container_of(work, + struct xfs_mount, m_flush_work); + + xfs_sync_data(mp, SYNC_TRYLOCK); + xfs_sync_data(mp, SYNC_TRYLOCK | SYNC_WAIT); } int xfs_syncd_init( struct xfs_mount *mp) { - mp->m_sync_work.w_syncer = xfs_sync_worker; - mp->m_sync_work.w_mount = mp; - mp->m_sync_work.w_completion = NULL; - mp->m_sync_task = kthread_run(xfssyncd, mp, "xfssyncd/%s", mp->m_fsname); - if (IS_ERR(mp->m_sync_task)) - return -PTR_ERR(mp->m_sync_task); + INIT_WORK(&mp->m_flush_work, xfs_flush_worker); + INIT_DELAYED_WORK(&mp->m_sync_work, xfs_sync_worker); + INIT_DELAYED_WORK(&mp->m_reclaim_work, xfs_reclaim_worker); + + xfs_syncd_queue_sync(mp); + xfs_syncd_queue_reclaim(mp); + return 0; } @@ -582,7 +579,9 @@ void xfs_syncd_stop( struct xfs_mount *mp) { - kthread_stop(mp->m_sync_task); + cancel_delayed_work_sync(&mp->m_sync_work); + cancel_delayed_work_sync(&mp->m_reclaim_work); + cancel_work_sync(&mp->m_flush_work); } void @@ -601,6 +600,10 @@ __xfs_inode_set_reclaim_tag( XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino), XFS_ICI_RECLAIM_TAG); spin_unlock(&ip->i_mount->m_perag_lock); + + /* schedule periodic background inode reclaim */ + xfs_syncd_queue_reclaim(ip->i_mount); + trace_xfs_perag_set_reclaim(ip->i_mount, pag->pag_agno, -1, _RET_IP_); } @@ -933,6 +936,7 @@ restart: XFS_LOOKUP_BATCH, XFS_ICI_RECLAIM_TAG); if (!nr_found) { + done = 1; rcu_read_unlock(); break; } @@ -1017,25 +1021,37 @@ xfs_reclaim_inodes( } /* - * Shrinker infrastructure. + * Inode cache shrinker. + * + * When called we make sure that there is a background (fast) inode reclaim in + * progress, while we will throttle the speed of reclaim via doiing synchronous + * reclaim of inodes. That means if we come across dirty inodes, we wait for + * them to be cleaned, which we hope will not be very long due to the + * background walker having already kicked the IO off on those dirty inodes. */ static int xfs_reclaim_inode_shrink( struct shrinker *shrink, - int nr_to_scan, - gfp_t gfp_mask) + struct shrink_control *sc) { struct xfs_mount *mp; struct xfs_perag *pag; xfs_agnumber_t ag; int reclaimable; + int nr_to_scan = sc->nr_to_scan; + gfp_t gfp_mask = sc->gfp_mask; mp = container_of(shrink, struct xfs_mount, m_inode_shrink); if (nr_to_scan) { + /* kick background reclaimer and push the AIL */ + xfs_syncd_queue_reclaim(mp); + xfs_ail_push_all(mp->m_ail); + if (!(gfp_mask & __GFP_FS)) return -1; - xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK, &nr_to_scan); + xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT, + &nr_to_scan); /* terminate if we don't exhaust the scan */ if (nr_to_scan > 0) return -1; diff --git a/fs/xfs/linux-2.6/xfs_sync.h b/fs/xfs/linux-2.6/xfs_sync.h index 32ba6628290c..e3a6ad27415f 100644 --- a/fs/xfs/linux-2.6/xfs_sync.h +++ b/fs/xfs/linux-2.6/xfs_sync.h @@ -32,6 +32,8 @@ typedef struct xfs_sync_work { #define SYNC_WAIT 0x0001 /* wait for i/o to complete */ #define SYNC_TRYLOCK 0x0002 /* only try to lock inodes */ +extern struct workqueue_struct *xfs_syncd_wq; /* sync workqueue */ + int xfs_syncd_init(struct xfs_mount *mp); void xfs_syncd_stop(struct xfs_mount *mp); diff --git a/fs/xfs/linux-2.6/xfs_trace.h b/fs/xfs/linux-2.6/xfs_trace.h index 2d0bcb479075..d48b7a579ae1 100644 --- a/fs/xfs/linux-2.6/xfs_trace.h +++ b/fs/xfs/linux-2.6/xfs_trace.h @@ -1151,44 +1151,7 @@ TRACE_EVENT(xfs_bunmap, ); -#define XFS_BUSY_SYNC \ - { 0, "async" }, \ - { 1, "sync" } - -TRACE_EVENT(xfs_alloc_busy, - TP_PROTO(struct xfs_trans *trans, xfs_agnumber_t agno, - xfs_agblock_t agbno, xfs_extlen_t len, int sync), - TP_ARGS(trans, agno, agbno, len, sync), - TP_STRUCT__entry( - __field(dev_t, dev) - __field(struct xfs_trans *, tp) - __field(int, tid) - __field(xfs_agnumber_t, agno) - __field(xfs_agblock_t, agbno) - __field(xfs_extlen_t, len) - __field(int, sync) - ), - TP_fast_assign( - __entry->dev = trans->t_mountp->m_super->s_dev; - __entry->tp = trans; - __entry->tid = trans->t_ticket->t_tid; - __entry->agno = agno; - __entry->agbno = agbno; - __entry->len = len; - __entry->sync = sync; - ), - TP_printk("dev %d:%d trans 0x%p tid 0x%x agno %u agbno %u len %u %s", - MAJOR(__entry->dev), MINOR(__entry->dev), - __entry->tp, - __entry->tid, - __entry->agno, - __entry->agbno, - __entry->len, - __print_symbolic(__entry->sync, XFS_BUSY_SYNC)) - -); - -TRACE_EVENT(xfs_alloc_unbusy, +DECLARE_EVENT_CLASS(xfs_busy_class, TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, xfs_agblock_t agbno, xfs_extlen_t len), TP_ARGS(mp, agno, agbno, len), @@ -1210,35 +1173,45 @@ TRACE_EVENT(xfs_alloc_unbusy, __entry->agbno, __entry->len) ); +#define DEFINE_BUSY_EVENT(name) \ +DEFINE_EVENT(xfs_busy_class, name, \ + TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, \ + xfs_agblock_t agbno, xfs_extlen_t len), \ + TP_ARGS(mp, agno, agbno, len)) +DEFINE_BUSY_EVENT(xfs_alloc_busy); +DEFINE_BUSY_EVENT(xfs_alloc_busy_enomem); +DEFINE_BUSY_EVENT(xfs_alloc_busy_force); +DEFINE_BUSY_EVENT(xfs_alloc_busy_reuse); +DEFINE_BUSY_EVENT(xfs_alloc_busy_clear); -#define XFS_BUSY_STATES \ - { 0, "missing" }, \ - { 1, "found" } - -TRACE_EVENT(xfs_alloc_busysearch, +TRACE_EVENT(xfs_alloc_busy_trim, TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, - xfs_agblock_t agbno, xfs_extlen_t len, int found), - TP_ARGS(mp, agno, agbno, len, found), + xfs_agblock_t agbno, xfs_extlen_t len, + xfs_agblock_t tbno, xfs_extlen_t tlen), + TP_ARGS(mp, agno, agbno, len, tbno, tlen), TP_STRUCT__entry( __field(dev_t, dev) __field(xfs_agnumber_t, agno) __field(xfs_agblock_t, agbno) __field(xfs_extlen_t, len) - __field(int, found) + __field(xfs_agblock_t, tbno) + __field(xfs_extlen_t, tlen) ), TP_fast_assign( __entry->dev = mp->m_super->s_dev; __entry->agno = agno; __entry->agbno = agbno; __entry->len = len; - __entry->found = found; + __entry->tbno = tbno; + __entry->tlen = tlen; ), - TP_printk("dev %d:%d agno %u agbno %u len %u %s", + TP_printk("dev %d:%d agno %u agbno %u len %u tbno %u tlen %u", MAJOR(__entry->dev), MINOR(__entry->dev), __entry->agno, __entry->agbno, __entry->len, - __print_symbolic(__entry->found, XFS_BUSY_STATES)) + __entry->tbno, + __entry->tlen) ); TRACE_EVENT(xfs_trans_commit_lsn, @@ -1418,7 +1391,7 @@ DECLARE_EVENT_CLASS(xfs_alloc_class, __entry->wasfromfl, __entry->isfl, __entry->userdata, - __entry->firstblock) + (unsigned long long)__entry->firstblock) ) #define DEFINE_ALLOC_EVENT(name) \ @@ -1433,11 +1406,14 @@ DEFINE_ALLOC_EVENT(xfs_alloc_near_first); DEFINE_ALLOC_EVENT(xfs_alloc_near_greater); DEFINE_ALLOC_EVENT(xfs_alloc_near_lesser); DEFINE_ALLOC_EVENT(xfs_alloc_near_error); +DEFINE_ALLOC_EVENT(xfs_alloc_near_noentry); +DEFINE_ALLOC_EVENT(xfs_alloc_near_busy); DEFINE_ALLOC_EVENT(xfs_alloc_size_neither); DEFINE_ALLOC_EVENT(xfs_alloc_size_noentry); DEFINE_ALLOC_EVENT(xfs_alloc_size_nominleft); DEFINE_ALLOC_EVENT(xfs_alloc_size_done); DEFINE_ALLOC_EVENT(xfs_alloc_size_error); +DEFINE_ALLOC_EVENT(xfs_alloc_size_busy); DEFINE_ALLOC_EVENT(xfs_alloc_small_freelist); DEFINE_ALLOC_EVENT(xfs_alloc_small_notenough); DEFINE_ALLOC_EVENT(xfs_alloc_small_done); diff --git a/fs/xfs/quota/xfs_dquot.c b/fs/xfs/quota/xfs_dquot.c index 7e2416478503..6fa214603819 100644 --- a/fs/xfs/quota/xfs_dquot.c +++ b/fs/xfs/quota/xfs_dquot.c @@ -600,7 +600,7 @@ xfs_qm_dqread( /* * Reservation counters are defined as reservation plus current usage - * to avoid having to add everytime. + * to avoid having to add every time. */ dqp->q_res_bcount = be64_to_cpu(ddqp->d_bcount); dqp->q_res_icount = be64_to_cpu(ddqp->d_icount); diff --git a/fs/xfs/quota/xfs_qm.c b/fs/xfs/quota/xfs_qm.c index 254ee062bd7d..b94dace4e785 100644 --- a/fs/xfs/quota/xfs_qm.c +++ b/fs/xfs/quota/xfs_qm.c @@ -60,7 +60,7 @@ STATIC void xfs_qm_list_destroy(xfs_dqlist_t *); STATIC int xfs_qm_init_quotainos(xfs_mount_t *); STATIC int xfs_qm_init_quotainfo(xfs_mount_t *); -STATIC int xfs_qm_shake(struct shrinker *, int, gfp_t); +STATIC int xfs_qm_shake(struct shrinker *, struct shrink_control *); static struct shrinker xfs_qm_shaker = { .shrink = xfs_qm_shake, @@ -461,12 +461,10 @@ xfs_qm_dqflush_all( struct xfs_quotainfo *q = mp->m_quotainfo; int recl; struct xfs_dquot *dqp; - int niters; int error; if (!q) return 0; - niters = 0; again: mutex_lock(&q->qi_dqlist_lock); list_for_each_entry(dqp, &q->qi_dqlist, q_mplist) { @@ -1314,14 +1312,9 @@ xfs_qm_dqiter_bufs( { xfs_buf_t *bp; int error; - int notcommitted; - int incr; int type; ASSERT(blkcnt > 0); - notcommitted = 0; - incr = (blkcnt > XFS_QM_MAX_DQCLUSTER_LOGSZ) ? - XFS_QM_MAX_DQCLUSTER_LOGSZ : blkcnt; type = flags & XFS_QMOPT_UQUOTA ? XFS_DQ_USER : (flags & XFS_QMOPT_PQUOTA ? XFS_DQ_PROJ : XFS_DQ_GROUP); error = 0; @@ -2016,10 +2009,10 @@ xfs_qm_shake_freelist( STATIC int xfs_qm_shake( struct shrinker *shrink, - int nr_to_scan, - gfp_t gfp_mask) + struct shrink_control *sc) { int ndqused, nfree, n; + gfp_t gfp_mask = sc->gfp_mask; if (!kmem_shake_allow(gfp_mask)) return 0; diff --git a/fs/xfs/quota/xfs_qm.h b/fs/xfs/quota/xfs_qm.h index c9446f1c726d..567b29b9f1b3 100644 --- a/fs/xfs/quota/xfs_qm.h +++ b/fs/xfs/quota/xfs_qm.h @@ -65,11 +65,6 @@ extern kmem_zone_t *qm_dqtrxzone; * block in the dquot/xqm code. */ #define XFS_DQUOT_CLUSTER_SIZE_FSB (xfs_filblks_t)1 -/* - * When doing a quotacheck, we log dquot clusters of this many FSBs at most - * in a single transaction. We don't want to ask for too huge a log reservation. - */ -#define XFS_QM_MAX_DQCLUSTER_LOGSZ 3 typedef xfs_dqhash_t xfs_dqlist_t; diff --git a/fs/xfs/quota/xfs_qm_bhv.c b/fs/xfs/quota/xfs_qm_bhv.c index 774d7ec6df8e..a0a829addca9 100644 --- a/fs/xfs/quota/xfs_qm_bhv.c +++ b/fs/xfs/quota/xfs_qm_bhv.c @@ -134,7 +134,7 @@ xfs_qm_newmount( */ if (quotaondisk && !XFS_QM_NEED_QUOTACHECK(mp)) { /* - * If an error occured, qm_mount_quotas code + * If an error occurred, qm_mount_quotas code * has already disabled quotas. So, just finish * mounting, and get on with the boring life * without disk quotas. diff --git a/fs/xfs/quota/xfs_qm_syscalls.c b/fs/xfs/quota/xfs_qm_syscalls.c index c82f06778a27..2dadb15d5ca9 100644 --- a/fs/xfs/quota/xfs_qm_syscalls.c +++ b/fs/xfs/quota/xfs_qm_syscalls.c @@ -172,7 +172,7 @@ xfs_qm_scall_quotaoff( /* * Next we make the changes in the quota flag in the mount struct. * This isn't protected by a particular lock directly, because we - * don't want to take a mrlock everytime we depend on quotas being on. + * don't want to take a mrlock every time we depend on quotas being on. */ mp->m_qflags &= ~(flags); @@ -313,14 +313,12 @@ xfs_qm_scall_quotaon( { int error; uint qf; - uint accflags; __int64_t sbflags; flags &= (XFS_ALL_QUOTA_ACCT | XFS_ALL_QUOTA_ENFD); /* * Switching on quota accounting must be done at mount time. */ - accflags = flags & XFS_ALL_QUOTA_ACCT; flags &= ~(XFS_ALL_QUOTA_ACCT); sbflags = 0; @@ -354,7 +352,7 @@ xfs_qm_scall_quotaon( return XFS_ERROR(EINVAL); } /* - * If everything's upto-date incore, then don't waste time. + * If everything's up to-date incore, then don't waste time. */ if ((mp->m_qflags & flags) == flags) return XFS_ERROR(EEXIST); diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h index 58632cc17f2d..6530769a999b 100644 --- a/fs/xfs/xfs_ag.h +++ b/fs/xfs/xfs_ag.h @@ -187,7 +187,9 @@ struct xfs_busy_extent { xfs_agnumber_t agno; xfs_agblock_t bno; xfs_extlen_t length; - xlog_tid_t tid; /* transaction that created this */ + unsigned int flags; +#define XFS_ALLOC_BUSY_DISCARDED 0x01 /* undergoing a discard op. */ +#define XFS_ALLOC_BUSY_SKIP_DISCARD 0x02 /* do not discard */ }; /* diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index 4bc3c649aee4..95862bbff56b 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -41,19 +41,13 @@ #define XFSA_FIXUP_BNO_OK 1 #define XFSA_FIXUP_CNT_OK 2 -/* - * Prototypes for per-ag allocation routines - */ - STATIC int xfs_alloc_ag_vextent_exact(xfs_alloc_arg_t *); STATIC int xfs_alloc_ag_vextent_near(xfs_alloc_arg_t *); STATIC int xfs_alloc_ag_vextent_size(xfs_alloc_arg_t *); STATIC int xfs_alloc_ag_vextent_small(xfs_alloc_arg_t *, - xfs_btree_cur_t *, xfs_agblock_t *, xfs_extlen_t *, int *); - -/* - * Internal functions. - */ + xfs_btree_cur_t *, xfs_agblock_t *, xfs_extlen_t *, int *); +STATIC void xfs_alloc_busy_trim(struct xfs_alloc_arg *, + xfs_agblock_t, xfs_extlen_t, xfs_agblock_t *, xfs_extlen_t *); /* * Lookup the record equal to [bno, len] in the btree given by cur. @@ -154,19 +148,21 @@ xfs_alloc_compute_aligned( xfs_extlen_t *reslen) /* result length */ { xfs_agblock_t bno; - xfs_extlen_t diff; xfs_extlen_t len; - if (args->alignment > 1 && foundlen >= args->minlen) { - bno = roundup(foundbno, args->alignment); - diff = bno - foundbno; - len = diff >= foundlen ? 0 : foundlen - diff; + /* Trim busy sections out of found extent */ + xfs_alloc_busy_trim(args, foundbno, foundlen, &bno, &len); + + if (args->alignment > 1 && len >= args->minlen) { + xfs_agblock_t aligned_bno = roundup(bno, args->alignment); + xfs_extlen_t diff = aligned_bno - bno; + + *resbno = aligned_bno; + *reslen = diff >= len ? 0 : len - diff; } else { - bno = foundbno; - len = foundlen; + *resbno = bno; + *reslen = len; } - *resbno = bno; - *reslen = len; } /* @@ -280,7 +276,6 @@ xfs_alloc_fix_minleft( return 1; agf = XFS_BUF_TO_AGF(args->agbp); diff = be32_to_cpu(agf->agf_freeblks) - + be32_to_cpu(agf->agf_flcount) - args->len - args->minleft; if (diff >= 0) return 1; @@ -541,16 +536,8 @@ xfs_alloc_ag_vextent( if (error) return error; - /* - * Search the busylist for these blocks and mark the - * transaction as synchronous if blocks are found. This - * avoids the need to block due to a synchronous log - * force to ensure correct ordering as the synchronous - * transaction will guarantee that for us. - */ - if (xfs_alloc_busy_search(args->mp, args->agno, - args->agbno, args->len)) - xfs_trans_set_sync(args->tp); + ASSERT(!xfs_alloc_busy_search(args->mp, args->agno, + args->agbno, args->len)); } if (!args->isfl) { @@ -577,14 +564,14 @@ xfs_alloc_ag_vextent_exact( { xfs_btree_cur_t *bno_cur;/* by block-number btree cursor */ xfs_btree_cur_t *cnt_cur;/* by count btree cursor */ - xfs_agblock_t end; /* end of allocated extent */ int error; xfs_agblock_t fbno; /* start block of found extent */ - xfs_agblock_t fend; /* end block of found extent */ xfs_extlen_t flen; /* length of found extent */ + xfs_agblock_t tbno; /* start block of trimmed extent */ + xfs_extlen_t tlen; /* length of trimmed extent */ + xfs_agblock_t tend; /* end block of trimmed extent */ + xfs_agblock_t end; /* end of allocated extent */ int i; /* success/failure of operation */ - xfs_agblock_t maxend; /* end of maximal extent */ - xfs_agblock_t minend; /* end of minimal extent */ xfs_extlen_t rlen; /* length of returned extent */ ASSERT(args->alignment == 1); @@ -614,14 +601,22 @@ xfs_alloc_ag_vextent_exact( goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); ASSERT(fbno <= args->agbno); - minend = args->agbno + args->minlen; - maxend = args->agbno + args->maxlen; - fend = fbno + flen; /* - * Give up if the freespace isn't long enough for the minimum request. + * Check for overlapping busy extents. + */ + xfs_alloc_busy_trim(args, fbno, flen, &tbno, &tlen); + + /* + * Give up if the start of the extent is busy, or the freespace isn't + * long enough for the minimum request. */ - if (fend < minend) + if (tbno > args->agbno) + goto not_found; + if (tlen < args->minlen) + goto not_found; + tend = tbno + tlen; + if (tend < args->agbno + args->minlen) goto not_found; /* @@ -630,14 +625,14 @@ xfs_alloc_ag_vextent_exact( * * Fix the length according to mod and prod if given. */ - end = XFS_AGBLOCK_MIN(fend, maxend); + end = XFS_AGBLOCK_MIN(tend, args->agbno + args->maxlen); args->len = end - args->agbno; xfs_alloc_fix_len(args); if (!xfs_alloc_fix_minleft(args)) goto not_found; rlen = args->len; - ASSERT(args->agbno + rlen <= fend); + ASSERT(args->agbno + rlen <= tend); end = args->agbno + rlen; /* @@ -686,11 +681,11 @@ xfs_alloc_find_best_extent( struct xfs_btree_cur **scur, /* searching cursor */ xfs_agblock_t gdiff, /* difference for search comparison */ xfs_agblock_t *sbno, /* extent found by search */ - xfs_extlen_t *slen, - xfs_extlen_t *slena, /* aligned length */ + xfs_extlen_t *slen, /* extent length */ + xfs_agblock_t *sbnoa, /* aligned extent found by search */ + xfs_extlen_t *slena, /* aligned extent length */ int dir) /* 0 = search right, 1 = search left */ { - xfs_agblock_t bno; xfs_agblock_t new; xfs_agblock_t sdiff; int error; @@ -708,16 +703,16 @@ xfs_alloc_find_best_extent( if (error) goto error0; XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - xfs_alloc_compute_aligned(args, *sbno, *slen, &bno, slena); + xfs_alloc_compute_aligned(args, *sbno, *slen, sbnoa, slena); /* * The good extent is closer than this one. */ if (!dir) { - if (bno >= args->agbno + gdiff) + if (*sbnoa >= args->agbno + gdiff) goto out_use_good; } else { - if (bno <= args->agbno - gdiff) + if (*sbnoa <= args->agbno - gdiff) goto out_use_good; } @@ -729,8 +724,8 @@ xfs_alloc_find_best_extent( xfs_alloc_fix_len(args); sdiff = xfs_alloc_compute_diff(args->agbno, args->len, - args->alignment, *sbno, - *slen, &new); + args->alignment, *sbnoa, + *slena, &new); /* * Choose closer size and invalidate other cursor. @@ -780,7 +775,7 @@ xfs_alloc_ag_vextent_near( xfs_agblock_t gtbnoa; /* aligned ... */ xfs_extlen_t gtdiff; /* difference to right side entry */ xfs_extlen_t gtlen; /* length of right side entry */ - xfs_extlen_t gtlena = 0; /* aligned ... */ + xfs_extlen_t gtlena; /* aligned ... */ xfs_agblock_t gtnew; /* useful start bno of right side */ int error; /* error code */ int i; /* result code, temporary */ @@ -789,9 +784,10 @@ xfs_alloc_ag_vextent_near( xfs_agblock_t ltbnoa; /* aligned ... */ xfs_extlen_t ltdiff; /* difference to left side entry */ xfs_extlen_t ltlen; /* length of left side entry */ - xfs_extlen_t ltlena = 0; /* aligned ... */ + xfs_extlen_t ltlena; /* aligned ... */ xfs_agblock_t ltnew; /* useful start bno of left side */ xfs_extlen_t rlen; /* length of returned extent */ + int forced = 0; #if defined(DEBUG) && defined(__KERNEL__) /* * Randomly don't execute the first algorithm. @@ -800,13 +796,20 @@ xfs_alloc_ag_vextent_near( dofirst = random32() & 1; #endif + +restart: + bno_cur_lt = NULL; + bno_cur_gt = NULL; + ltlen = 0; + gtlena = 0; + ltlena = 0; + /* * Get a cursor for the by-size btree. */ cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, args->agno, XFS_BTNUM_CNT); - ltlen = 0; - bno_cur_lt = bno_cur_gt = NULL; + /* * See if there are any free extents as big as maxlen. */ @@ -822,11 +825,13 @@ xfs_alloc_ag_vextent_near( goto error0; if (i == 0 || ltlen == 0) { xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); + trace_xfs_alloc_near_noentry(args); return 0; } ASSERT(i == 1); } args->wasfromfl = 0; + /* * First algorithm. * If the requested extent is large wrt the freespaces available @@ -890,7 +895,7 @@ xfs_alloc_ag_vextent_near( if (args->len < blen) continue; ltdiff = xfs_alloc_compute_diff(args->agbno, args->len, - args->alignment, ltbno, ltlen, <new); + args->alignment, ltbnoa, ltlena, <new); if (ltnew != NULLAGBLOCK && (args->len > blen || ltdiff < bdiff)) { bdiff = ltdiff; @@ -1042,11 +1047,12 @@ xfs_alloc_ag_vextent_near( args->len = XFS_EXTLEN_MIN(ltlena, args->maxlen); xfs_alloc_fix_len(args); ltdiff = xfs_alloc_compute_diff(args->agbno, args->len, - args->alignment, ltbno, ltlen, <new); + args->alignment, ltbnoa, ltlena, <new); error = xfs_alloc_find_best_extent(args, &bno_cur_lt, &bno_cur_gt, - ltdiff, >bno, >len, >lena, + ltdiff, >bno, >len, + >bnoa, >lena, 0 /* search right */); } else { ASSERT(gtlena >= args->minlen); @@ -1057,11 +1063,12 @@ xfs_alloc_ag_vextent_near( args->len = XFS_EXTLEN_MIN(gtlena, args->maxlen); xfs_alloc_fix_len(args); gtdiff = xfs_alloc_compute_diff(args->agbno, args->len, - args->alignment, gtbno, gtlen, >new); + args->alignment, gtbnoa, gtlena, >new); error = xfs_alloc_find_best_extent(args, &bno_cur_gt, &bno_cur_lt, - gtdiff, <bno, <len, <lena, + gtdiff, <bno, <len, + <bnoa, <lena, 1 /* search left */); } @@ -1073,6 +1080,12 @@ xfs_alloc_ag_vextent_near( * If we couldn't get anything, give up. */ if (bno_cur_lt == NULL && bno_cur_gt == NULL) { + if (!forced++) { + trace_xfs_alloc_near_busy(args); + xfs_log_force(args->mp, XFS_LOG_SYNC); + goto restart; + } + trace_xfs_alloc_size_neither(args); args->agbno = NULLAGBLOCK; return 0; @@ -1107,12 +1120,13 @@ xfs_alloc_ag_vextent_near( return 0; } rlen = args->len; - (void)xfs_alloc_compute_diff(args->agbno, rlen, args->alignment, ltbno, - ltlen, <new); + (void)xfs_alloc_compute_diff(args->agbno, rlen, args->alignment, + ltbnoa, ltlena, <new); ASSERT(ltnew >= ltbno); - ASSERT(ltnew + rlen <= ltbno + ltlen); + ASSERT(ltnew + rlen <= ltbnoa + ltlena); ASSERT(ltnew + rlen <= be32_to_cpu(XFS_BUF_TO_AGF(args->agbp)->agf_length)); args->agbno = ltnew; + if ((error = xfs_alloc_fixup_trees(cnt_cur, bno_cur_lt, ltbno, ltlen, ltnew, rlen, XFSA_FIXUP_BNO_OK))) goto error0; @@ -1155,26 +1169,35 @@ xfs_alloc_ag_vextent_size( int i; /* temp status variable */ xfs_agblock_t rbno; /* returned block number */ xfs_extlen_t rlen; /* length of returned extent */ + int forced = 0; +restart: /* * Allocate and initialize a cursor for the by-size btree. */ cnt_cur = xfs_allocbt_init_cursor(args->mp, args->tp, args->agbp, args->agno, XFS_BTNUM_CNT); bno_cur = NULL; + /* * Look for an entry >= maxlen+alignment-1 blocks. */ if ((error = xfs_alloc_lookup_ge(cnt_cur, 0, args->maxlen + args->alignment - 1, &i))) goto error0; + /* - * If none, then pick up the last entry in the tree unless the - * tree is empty. + * If none or we have busy extents that we cannot allocate from, then + * we have to settle for a smaller extent. In the case that there are + * no large extents, this will return the last entry in the tree unless + * the tree is empty. In the case that there are only busy large + * extents, this will return the largest small extent unless there + * are no smaller extents available. */ - if (!i) { - if ((error = xfs_alloc_ag_vextent_small(args, cnt_cur, &fbno, - &flen, &i))) + if (!i || forced > 1) { + error = xfs_alloc_ag_vextent_small(args, cnt_cur, + &fbno, &flen, &i); + if (error) goto error0; if (i == 0 || flen == 0) { xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); @@ -1182,22 +1205,56 @@ xfs_alloc_ag_vextent_size( return 0; } ASSERT(i == 1); + xfs_alloc_compute_aligned(args, fbno, flen, &rbno, &rlen); + } else { + /* + * Search for a non-busy extent that is large enough. + * If we are at low space, don't check, or if we fall of + * the end of the btree, turn off the busy check and + * restart. + */ + for (;;) { + error = xfs_alloc_get_rec(cnt_cur, &fbno, &flen, &i); + if (error) + goto error0; + XFS_WANT_CORRUPTED_GOTO(i == 1, error0); + + xfs_alloc_compute_aligned(args, fbno, flen, + &rbno, &rlen); + + if (rlen >= args->maxlen) + break; + + error = xfs_btree_increment(cnt_cur, 0, &i); + if (error) + goto error0; + if (i == 0) { + /* + * Our only valid extents must have been busy. + * Make it unbusy by forcing the log out and + * retrying. If we've been here before, forcing + * the log isn't making the extents available, + * which means they have probably been freed in + * this transaction. In that case, we have to + * give up on them and we'll attempt a minlen + * allocation the next time around. + */ + xfs_btree_del_cursor(cnt_cur, + XFS_BTREE_NOERROR); + trace_xfs_alloc_size_busy(args); + if (!forced++) + xfs_log_force(args->mp, XFS_LOG_SYNC); + goto restart; + } + } } - /* - * There's a freespace as big as maxlen+alignment-1, get it. - */ - else { - if ((error = xfs_alloc_get_rec(cnt_cur, &fbno, &flen, &i))) - goto error0; - XFS_WANT_CORRUPTED_GOTO(i == 1, error0); - } + /* * In the first case above, we got the last entry in the * by-size btree. Now we check to see if the space hits maxlen * once aligned; if not, we search left for something better. * This can't happen in the second case above. */ - xfs_alloc_compute_aligned(args, fbno, flen, &rbno, &rlen); rlen = XFS_EXTLEN_MIN(args->maxlen, rlen); XFS_WANT_CORRUPTED_GOTO(rlen == 0 || (rlen <= flen && rbno + rlen <= fbno + flen), error0); @@ -1251,13 +1308,19 @@ xfs_alloc_ag_vextent_size( * Fix up the length. */ args->len = rlen; - xfs_alloc_fix_len(args); - if (rlen < args->minlen || !xfs_alloc_fix_minleft(args)) { - xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); - trace_xfs_alloc_size_nominleft(args); - args->agbno = NULLAGBLOCK; - return 0; + if (rlen < args->minlen) { + if (!forced++) { + xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); + trace_xfs_alloc_size_busy(args); + xfs_log_force(args->mp, XFS_LOG_SYNC); + goto restart; + } + goto out_nominleft; } + xfs_alloc_fix_len(args); + + if (!xfs_alloc_fix_minleft(args)) + goto out_nominleft; rlen = args->len; XFS_WANT_CORRUPTED_GOTO(rlen <= flen, error0); /* @@ -1287,6 +1350,12 @@ error0: if (bno_cur) xfs_btree_del_cursor(bno_cur, XFS_BTREE_ERROR); return error; + +out_nominleft: + xfs_btree_del_cursor(cnt_cur, XFS_BTREE_NOERROR); + trace_xfs_alloc_size_nominleft(args); + args->agbno = NULLAGBLOCK; + return 0; } /* @@ -1326,6 +1395,9 @@ xfs_alloc_ag_vextent_small( if (error) goto error0; if (fbno != NULLAGBLOCK) { + xfs_alloc_busy_reuse(args->mp, args->agno, fbno, 1, + args->userdata); + if (args->userdata) { xfs_buf_t *bp; @@ -1617,18 +1689,6 @@ xfs_free_ag_extent( trace_xfs_free_extent(mp, agno, bno, len, isfl, haveleft, haveright); - /* - * Since blocks move to the free list without the coordination - * used in xfs_bmap_finish, we can't allow block to be available - * for reallocation and non-transaction writing (user data) - * until we know that the transaction that moved it to the free - * list is permanently on disk. We track the blocks by declaring - * these blocks as "busy"; the busy list is maintained on a per-ag - * basis and each transaction records which entries should be removed - * when the iclog commits to disk. If a busy block is allocated, - * the iclog is pushed up to the LSN that freed the block. - */ - xfs_alloc_busy_insert(tp, agno, bno, len); return 0; error0: @@ -1923,21 +1983,6 @@ xfs_alloc_get_freelist( xfs_alloc_log_agf(tp, agbp, logflags); *bnop = bno; - /* - * As blocks are freed, they are added to the per-ag busy list and - * remain there until the freeing transaction is committed to disk. - * Now that we have allocated blocks, this list must be searched to see - * if a block is being reused. If one is, then the freeing transaction - * must be pushed to disk before this transaction. - * - * We do this by setting the current transaction to a sync transaction - * which guarantees that the freeing transaction is on disk before this - * transaction. This is done instead of a synchronous log force here so - * that we don't sit and wait with the AGF locked in the transaction - * during the log force. - */ - if (xfs_alloc_busy_search(mp, be32_to_cpu(agf->agf_seqno), bno, 1)) - xfs_trans_set_sync(tp); return 0; } @@ -2395,131 +2440,54 @@ xfs_free_extent( memset(&args, 0, sizeof(xfs_alloc_arg_t)); args.tp = tp; args.mp = tp->t_mountp; + + /* + * validate that the block number is legal - the enables us to detect + * and handle a silent filesystem corruption rather than crashing. + */ args.agno = XFS_FSB_TO_AGNO(args.mp, bno); - ASSERT(args.agno < args.mp->m_sb.sb_agcount); + if (args.agno >= args.mp->m_sb.sb_agcount) + return EFSCORRUPTED; + args.agbno = XFS_FSB_TO_AGBNO(args.mp, bno); + if (args.agbno >= args.mp->m_sb.sb_agblocks) + return EFSCORRUPTED; + args.pag = xfs_perag_get(args.mp, args.agno); - if ((error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING))) + ASSERT(args.pag); + + error = xfs_alloc_fix_freelist(&args, XFS_ALLOC_FLAG_FREEING); + if (error) goto error0; -#ifdef DEBUG - ASSERT(args.agbp != NULL); - ASSERT((args.agbno + len) <= - be32_to_cpu(XFS_BUF_TO_AGF(args.agbp)->agf_length)); -#endif + + /* validate the extent size is legal now we have the agf locked */ + if (args.agbno + len > + be32_to_cpu(XFS_BUF_TO_AGF(args.agbp)->agf_length)) { + error = EFSCORRUPTED; + goto error0; + } + error = xfs_free_ag_extent(tp, args.agbp, args.agno, args.agbno, len, 0); + if (!error) + xfs_alloc_busy_insert(tp, args.agno, args.agbno, len, 0); error0: xfs_perag_put(args.pag); return error; } - -/* - * AG Busy list management - * The busy list contains block ranges that have been freed but whose - * transactions have not yet hit disk. If any block listed in a busy - * list is reused, the transaction that freed it must be forced to disk - * before continuing to use the block. - * - * xfs_alloc_busy_insert - add to the per-ag busy list - * xfs_alloc_busy_clear - remove an item from the per-ag busy list - * xfs_alloc_busy_search - search for a busy extent - */ - -/* - * Insert a new extent into the busy tree. - * - * The busy extent tree is indexed by the start block of the busy extent. - * there can be multiple overlapping ranges in the busy extent tree but only - * ever one entry at a given start block. The reason for this is that - * multi-block extents can be freed, then smaller chunks of that extent - * allocated and freed again before the first transaction commit is on disk. - * If the exact same start block is freed a second time, we have to wait for - * that busy extent to pass out of the tree before the new extent is inserted. - * There are two main cases we have to handle here. - * - * The first case is a transaction that triggers a "free - allocate - free" - * cycle. This can occur during btree manipulations as a btree block is freed - * to the freelist, then allocated from the free list, then freed again. In - * this case, the second extxpnet free is what triggers the duplicate and as - * such the transaction IDs should match. Because the extent was allocated in - * this transaction, the transaction must be marked as synchronous. This is - * true for all cases where the free/alloc/free occurs in the one transaction, - * hence the addition of the ASSERT(tp->t_flags & XFS_TRANS_SYNC) to this case. - * This serves to catch violations of the second case quite effectively. - * - * The second case is where the free/alloc/free occur in different - * transactions. In this case, the thread freeing the extent the second time - * can't mark the extent busy immediately because it is already tracked in a - * transaction that may be committing. When the log commit for the existing - * busy extent completes, the busy extent will be removed from the tree. If we - * allow the second busy insert to continue using that busy extent structure, - * it can be freed before this transaction is safely in the log. Hence our - * only option in this case is to force the log to remove the existing busy - * extent from the list before we insert the new one with the current - * transaction ID. - * - * The problem we are trying to avoid in the free-alloc-free in separate - * transactions is most easily described with a timeline: - * - * Thread 1 Thread 2 Thread 3 xfslogd - * xact alloc - * free X - * mark busy - * commit xact - * free xact - * xact alloc - * alloc X - * busy search - * mark xact sync - * commit xact - * free xact - * force log - * checkpoint starts - * .... - * xact alloc - * free X - * mark busy - * finds match - * *** KABOOM! *** - * .... - * log IO completes - * unbusy X - * checkpoint completes - * - * By issuing a log force in thread 3 @ "KABOOM", the thread will block until - * the checkpoint completes, and the busy extent it matched will have been - * removed from the tree when it is woken. Hence it can then continue safely. - * - * However, to ensure this matching process is robust, we need to use the - * transaction ID for identifying transaction, as delayed logging results in - * the busy extent and transaction lifecycles being different. i.e. the busy - * extent is active for a lot longer than the transaction. Hence the - * transaction structure can be freed and reallocated, then mark the same - * extent busy again in the new transaction. In this case the new transaction - * will have a different tid but can have the same address, and hence we need - * to check against the tid. - * - * Future: for delayed logging, we could avoid the log force if the extent was - * first freed in the current checkpoint sequence. This, however, requires the - * ability to pin the current checkpoint in memory until this transaction - * commits to ensure that both the original free and the current one combine - * logically into the one checkpoint. If the checkpoint sequences are - * different, however, we still need to wait on a log force. - */ void xfs_alloc_busy_insert( struct xfs_trans *tp, xfs_agnumber_t agno, xfs_agblock_t bno, - xfs_extlen_t len) + xfs_extlen_t len, + unsigned int flags) { struct xfs_busy_extent *new; struct xfs_busy_extent *busyp; struct xfs_perag *pag; struct rb_node **rbp; - struct rb_node *parent; - int match; - + struct rb_node *parent = NULL; new = kmem_zalloc(sizeof(struct xfs_busy_extent), KM_MAYFAIL); if (!new) { @@ -2528,7 +2496,7 @@ xfs_alloc_busy_insert( * block, make this a synchronous transaction to insure that * the block is not reused before this transaction commits. */ - trace_xfs_alloc_busy(tp, agno, bno, len, 1); + trace_xfs_alloc_busy_enomem(tp->t_mountp, agno, bno, len); xfs_trans_set_sync(tp); return; } @@ -2536,66 +2504,29 @@ xfs_alloc_busy_insert( new->agno = agno; new->bno = bno; new->length = len; - new->tid = xfs_log_get_trans_ident(tp); - INIT_LIST_HEAD(&new->list); + new->flags = flags; /* trace before insert to be able to see failed inserts */ - trace_xfs_alloc_busy(tp, agno, bno, len, 0); + trace_xfs_alloc_busy(tp->t_mountp, agno, bno, len); pag = xfs_perag_get(tp->t_mountp, new->agno); -restart: spin_lock(&pag->pagb_lock); rbp = &pag->pagb_tree.rb_node; - parent = NULL; - busyp = NULL; - match = 0; - while (*rbp && match >= 0) { + while (*rbp) { parent = *rbp; busyp = rb_entry(parent, struct xfs_busy_extent, rb_node); if (new->bno < busyp->bno) { - /* may overlap, but exact start block is lower */ rbp = &(*rbp)->rb_left; - if (new->bno + new->length > busyp->bno) - match = busyp->tid == new->tid ? 1 : -1; + ASSERT(new->bno + new->length <= busyp->bno); } else if (new->bno > busyp->bno) { - /* may overlap, but exact start block is higher */ rbp = &(*rbp)->rb_right; - if (bno < busyp->bno + busyp->length) - match = busyp->tid == new->tid ? 1 : -1; + ASSERT(bno >= busyp->bno + busyp->length); } else { - match = busyp->tid == new->tid ? 1 : -1; - break; + ASSERT(0); } } - if (match < 0) { - /* overlap marked busy in different transaction */ - spin_unlock(&pag->pagb_lock); - xfs_log_force(tp->t_mountp, XFS_LOG_SYNC); - goto restart; - } - if (match > 0) { - /* - * overlap marked busy in same transaction. Update if exact - * start block match, otherwise combine the busy extents into - * a single range. - */ - if (busyp->bno == new->bno) { - busyp->length = max(busyp->length, new->length); - spin_unlock(&pag->pagb_lock); - ASSERT(tp->t_flags & XFS_TRANS_SYNC); - xfs_perag_put(pag); - kmem_free(new); - return; - } - rb_erase(&busyp->rb_node, &pag->pagb_tree); - new->length = max(busyp->bno + busyp->length, - new->bno + new->length) - - min(busyp->bno, new->bno); - new->bno = min(busyp->bno, new->bno); - } else - busyp = NULL; rb_link_node(&new->rb_node, parent, rbp); rb_insert_color(&new->rb_node, &pag->pagb_tree); @@ -2603,7 +2534,6 @@ restart: list_add(&new->list, &tp->t_busy); spin_unlock(&pag->pagb_lock); xfs_perag_put(pag); - kmem_free(busyp); } /* @@ -2652,31 +2582,466 @@ xfs_alloc_busy_search( } } spin_unlock(&pag->pagb_lock); - trace_xfs_alloc_busysearch(mp, agno, bno, len, !!match); xfs_perag_put(pag); return match; } +/* + * The found free extent [fbno, fend] overlaps part or all of the given busy + * extent. If the overlap covers the beginning, the end, or all of the busy + * extent, the overlapping portion can be made unbusy and used for the + * allocation. We can't split a busy extent because we can't modify a + * transaction/CIL context busy list, but we can update an entries block + * number or length. + * + * Returns true if the extent can safely be reused, or false if the search + * needs to be restarted. + */ +STATIC bool +xfs_alloc_busy_update_extent( + struct xfs_mount *mp, + struct xfs_perag *pag, + struct xfs_busy_extent *busyp, + xfs_agblock_t fbno, + xfs_extlen_t flen, + bool userdata) +{ + xfs_agblock_t fend = fbno + flen; + xfs_agblock_t bbno = busyp->bno; + xfs_agblock_t bend = bbno + busyp->length; + + /* + * This extent is currently being discarded. Give the thread + * performing the discard a chance to mark the extent unbusy + * and retry. + */ + if (busyp->flags & XFS_ALLOC_BUSY_DISCARDED) { + spin_unlock(&pag->pagb_lock); + delay(1); + spin_lock(&pag->pagb_lock); + return false; + } + + /* + * If there is a busy extent overlapping a user allocation, we have + * no choice but to force the log and retry the search. + * + * Fortunately this does not happen during normal operation, but + * only if the filesystem is very low on space and has to dip into + * the AGFL for normal allocations. + */ + if (userdata) + goto out_force_log; + + if (bbno < fbno && bend > fend) { + /* + * Case 1: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +---------+ + * fbno fend + */ + + /* + * We would have to split the busy extent to be able to track + * it correct, which we cannot do because we would have to + * modify the list of busy extents attached to the transaction + * or CIL context, which is immutable. + * + * Force out the log to clear the busy extent and retry the + * search. + */ + goto out_force_log; + } else if (bbno >= fbno && bend <= fend) { + /* + * Case 2: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +-----------------+ + * fbno fend + * + * Case 3: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +--------------------------+ + * fbno fend + * + * Case 4: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +--------------------------+ + * fbno fend + * + * Case 5: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +-----------------------------------+ + * fbno fend + * + */ + + /* + * The busy extent is fully covered by the extent we are + * allocating, and can simply be removed from the rbtree. + * However we cannot remove it from the immutable list + * tracking busy extents in the transaction or CIL context, + * so set the length to zero to mark it invalid. + * + * We also need to restart the busy extent search from the + * tree root, because erasing the node can rearrange the + * tree topology. + */ + rb_erase(&busyp->rb_node, &pag->pagb_tree); + busyp->length = 0; + return false; + } else if (fend < bend) { + /* + * Case 6: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +---------+ + * fbno fend + * + * Case 7: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +------------------+ + * fbno fend + * + */ + busyp->bno = fend; + } else if (bbno < fbno) { + /* + * Case 8: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +-------------+ + * fbno fend + * + * Case 9: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +----------------------+ + * fbno fend + */ + busyp->length = fbno - busyp->bno; + } else { + ASSERT(0); + } + + trace_xfs_alloc_busy_reuse(mp, pag->pag_agno, fbno, flen); + return true; + +out_force_log: + spin_unlock(&pag->pagb_lock); + xfs_log_force(mp, XFS_LOG_SYNC); + trace_xfs_alloc_busy_force(mp, pag->pag_agno, fbno, flen); + spin_lock(&pag->pagb_lock); + return false; +} + + +/* + * For a given extent [fbno, flen], make sure we can reuse it safely. + */ void -xfs_alloc_busy_clear( +xfs_alloc_busy_reuse( struct xfs_mount *mp, - struct xfs_busy_extent *busyp) + xfs_agnumber_t agno, + xfs_agblock_t fbno, + xfs_extlen_t flen, + bool userdata) { struct xfs_perag *pag; + struct rb_node *rbp; - trace_xfs_alloc_unbusy(mp, busyp->agno, busyp->bno, - busyp->length); + ASSERT(flen > 0); - ASSERT(xfs_alloc_busy_search(mp, busyp->agno, busyp->bno, - busyp->length) == 1); + pag = xfs_perag_get(mp, agno); + spin_lock(&pag->pagb_lock); +restart: + rbp = pag->pagb_tree.rb_node; + while (rbp) { + struct xfs_busy_extent *busyp = + rb_entry(rbp, struct xfs_busy_extent, rb_node); + xfs_agblock_t bbno = busyp->bno; + xfs_agblock_t bend = bbno + busyp->length; - list_del_init(&busyp->list); + if (fbno + flen <= bbno) { + rbp = rbp->rb_left; + continue; + } else if (fbno >= bend) { + rbp = rbp->rb_right; + continue; + } - pag = xfs_perag_get(mp, busyp->agno); - spin_lock(&pag->pagb_lock); - rb_erase(&busyp->rb_node, &pag->pagb_tree); + if (!xfs_alloc_busy_update_extent(mp, pag, busyp, fbno, flen, + userdata)) + goto restart; + } spin_unlock(&pag->pagb_lock); xfs_perag_put(pag); +} + +/* + * For a given extent [fbno, flen], search the busy extent list to find a + * subset of the extent that is not busy. If *rlen is smaller than + * args->minlen no suitable extent could be found, and the higher level + * code needs to force out the log and retry the allocation. + */ +STATIC void +xfs_alloc_busy_trim( + struct xfs_alloc_arg *args, + xfs_agblock_t bno, + xfs_extlen_t len, + xfs_agblock_t *rbno, + xfs_extlen_t *rlen) +{ + xfs_agblock_t fbno; + xfs_extlen_t flen; + struct rb_node *rbp; + ASSERT(len > 0); + + spin_lock(&args->pag->pagb_lock); +restart: + fbno = bno; + flen = len; + rbp = args->pag->pagb_tree.rb_node; + while (rbp && flen >= args->minlen) { + struct xfs_busy_extent *busyp = + rb_entry(rbp, struct xfs_busy_extent, rb_node); + xfs_agblock_t fend = fbno + flen; + xfs_agblock_t bbno = busyp->bno; + xfs_agblock_t bend = bbno + busyp->length; + + if (fend <= bbno) { + rbp = rbp->rb_left; + continue; + } else if (fbno >= bend) { + rbp = rbp->rb_right; + continue; + } + + /* + * If this is a metadata allocation, try to reuse the busy + * extent instead of trimming the allocation. + */ + if (!args->userdata && + !(busyp->flags & XFS_ALLOC_BUSY_DISCARDED)) { + if (!xfs_alloc_busy_update_extent(args->mp, args->pag, + busyp, fbno, flen, + false)) + goto restart; + continue; + } + + if (bbno <= fbno) { + /* start overlap */ + + /* + * Case 1: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +---------+ + * fbno fend + * + * Case 2: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +-------------+ + * fbno fend + * + * Case 3: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +-------------+ + * fbno fend + * + * Case 4: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +-----------------+ + * fbno fend + * + * No unbusy region in extent, return failure. + */ + if (fend <= bend) + goto fail; + + /* + * Case 5: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +----------------------+ + * fbno fend + * + * Case 6: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +--------------------------+ + * fbno fend + * + * Needs to be trimmed to: + * +-------+ + * fbno fend + */ + fbno = bend; + } else if (bend >= fend) { + /* end overlap */ + + /* + * Case 7: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +------------------+ + * fbno fend + * + * Case 8: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +--------------------------+ + * fbno fend + * + * Needs to be trimmed to: + * +-------+ + * fbno fend + */ + fend = bbno; + } else { + /* middle overlap */ + + /* + * Case 9: + * bbno bend + * +BBBBBBBBBBBBBBBBB+ + * +-----------------------------------+ + * fbno fend + * + * Can be trimmed to: + * +-------+ OR +-------+ + * fbno fend fbno fend + * + * Backward allocation leads to significant + * fragmentation of directories, which degrades + * directory performance, therefore we always want to + * choose the option that produces forward allocation + * patterns. + * Preferring the lower bno extent will make the next + * request use "fend" as the start of the next + * allocation; if the segment is no longer busy at + * that point, we'll get a contiguous allocation, but + * even if it is still busy, we will get a forward + * allocation. + * We try to avoid choosing the segment at "bend", + * because that can lead to the next allocation + * taking the segment at "fbno", which would be a + * backward allocation. We only use the segment at + * "fbno" if it is much larger than the current + * requested size, because in that case there's a + * good chance subsequent allocations will be + * contiguous. + */ + if (bbno - fbno >= args->maxlen) { + /* left candidate fits perfect */ + fend = bbno; + } else if (fend - bend >= args->maxlen * 4) { + /* right candidate has enough free space */ + fbno = bend; + } else if (bbno - fbno >= args->minlen) { + /* left candidate fits minimum requirement */ + fend = bbno; + } else { + goto fail; + } + } + + flen = fend - fbno; + } + spin_unlock(&args->pag->pagb_lock); + + if (fbno != bno || flen != len) { + trace_xfs_alloc_busy_trim(args->mp, args->agno, bno, len, + fbno, flen); + } + *rbno = fbno; + *rlen = flen; + return; +fail: + /* + * Return a zero extent length as failure indications. All callers + * re-check if the trimmed extent satisfies the minlen requirement. + */ + spin_unlock(&args->pag->pagb_lock); + trace_xfs_alloc_busy_trim(args->mp, args->agno, bno, len, fbno, 0); + *rbno = fbno; + *rlen = 0; +} + +static void +xfs_alloc_busy_clear_one( + struct xfs_mount *mp, + struct xfs_perag *pag, + struct xfs_busy_extent *busyp) +{ + if (busyp->length) { + trace_xfs_alloc_busy_clear(mp, busyp->agno, busyp->bno, + busyp->length); + rb_erase(&busyp->rb_node, &pag->pagb_tree); + } + + list_del_init(&busyp->list); kmem_free(busyp); } + +/* + * Remove all extents on the passed in list from the busy extents tree. + * If do_discard is set skip extents that need to be discarded, and mark + * these as undergoing a discard operation instead. + */ +void +xfs_alloc_busy_clear( + struct xfs_mount *mp, + struct list_head *list, + bool do_discard) +{ + struct xfs_busy_extent *busyp, *n; + struct xfs_perag *pag = NULL; + xfs_agnumber_t agno = NULLAGNUMBER; + + list_for_each_entry_safe(busyp, n, list, list) { + if (busyp->agno != agno) { + if (pag) { + spin_unlock(&pag->pagb_lock); + xfs_perag_put(pag); + } + pag = xfs_perag_get(mp, busyp->agno); + spin_lock(&pag->pagb_lock); + agno = busyp->agno; + } + + if (do_discard && busyp->length && + !(busyp->flags & XFS_ALLOC_BUSY_SKIP_DISCARD)) + busyp->flags = XFS_ALLOC_BUSY_DISCARDED; + else + xfs_alloc_busy_clear_one(mp, pag, busyp); + } + + if (pag) { + spin_unlock(&pag->pagb_lock); + xfs_perag_put(pag); + } +} + +/* + * Callback for list_sort to sort busy extents by the AG they reside in. + */ +int +xfs_busy_extent_ag_cmp( + void *priv, + struct list_head *a, + struct list_head *b) +{ + return container_of(a, struct xfs_busy_extent, list)->agno - + container_of(b, struct xfs_busy_extent, list)->agno; +} diff --git a/fs/xfs/xfs_alloc.h b/fs/xfs/xfs_alloc.h index d0b3bc72005b..2f52b924be79 100644 --- a/fs/xfs/xfs_alloc.h +++ b/fs/xfs/xfs_alloc.h @@ -137,14 +137,28 @@ xfs_alloc_longest_free_extent(struct xfs_mount *mp, #ifdef __KERNEL__ void xfs_alloc_busy_insert(struct xfs_trans *tp, xfs_agnumber_t agno, - xfs_agblock_t bno, xfs_extlen_t len); + xfs_agblock_t bno, xfs_extlen_t len, unsigned int flags); void -xfs_alloc_busy_clear(struct xfs_mount *mp, struct xfs_busy_extent *busyp); +xfs_alloc_busy_clear(struct xfs_mount *mp, struct list_head *list, + bool do_discard); int xfs_alloc_busy_search(struct xfs_mount *mp, xfs_agnumber_t agno, xfs_agblock_t bno, xfs_extlen_t len); + +void +xfs_alloc_busy_reuse(struct xfs_mount *mp, xfs_agnumber_t agno, + xfs_agblock_t fbno, xfs_extlen_t flen, bool userdata); + +int +xfs_busy_extent_ag_cmp(void *priv, struct list_head *a, struct list_head *b); + +static inline void xfs_alloc_busy_sort(struct list_head *list) +{ + list_sort(NULL, list, xfs_busy_extent_ag_cmp); +} + #endif /* __KERNEL__ */ /* diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c index 3916925e2584..2b3518826a69 100644 --- a/fs/xfs/xfs_alloc_btree.c +++ b/fs/xfs/xfs_alloc_btree.c @@ -95,6 +95,8 @@ xfs_allocbt_alloc_block( return 0; } + xfs_alloc_busy_reuse(cur->bc_mp, cur->bc_private.a.agno, bno, 1, false); + xfs_trans_agbtree_delta(cur->bc_tp, 1); new->s = cpu_to_be32(bno); @@ -118,18 +120,8 @@ xfs_allocbt_free_block( if (error) return error; - /* - * Since blocks move to the free list without the coordination used in - * xfs_bmap_finish, we can't allow block to be available for - * reallocation and non-transaction writing (user data) until we know - * that the transaction that moved it to the free list is permanently - * on disk. We track the blocks by declaring these blocks as "busy"; - * the busy list is maintained on a per-ag basis and each transaction - * records which entries should be removed when the iclog commits to - * disk. If a busy block is allocated, the iclog is pushed up to the - * LSN that freed the block. - */ - xfs_alloc_busy_insert(cur->bc_tp, be32_to_cpu(agf->agf_seqno), bno, 1); + xfs_alloc_busy_insert(cur->bc_tp, be32_to_cpu(agf->agf_seqno), bno, 1, + XFS_ALLOC_BUSY_SKIP_DISCARD); xfs_trans_agbtree_delta(cur->bc_tp, -1); return 0; } diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index fa00788de2f5..e546a33214c9 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -89,36 +89,19 @@ xfs_bmap_add_attrfork_local( int *flags); /* inode logging flags */ /* - * Called by xfs_bmapi to update file extent records and the btree - * after allocating space (or doing a delayed allocation). - */ -STATIC int /* error */ -xfs_bmap_add_extent( - xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ - xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ - xfs_bmbt_irec_t *new, /* new data to add to file extents */ - xfs_fsblock_t *first, /* pointer to firstblock variable */ - xfs_bmap_free_t *flist, /* list of extents to be freed */ - int *logflagsp, /* inode logging flags */ - int whichfork, /* data or attr fork */ - int rsvd); /* OK to allocate reserved blocks */ - -/* * Called by xfs_bmap_add_extent to handle cases converting a delayed * allocation to a real allocation. */ STATIC int /* error */ xfs_bmap_add_extent_delay_real( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ xfs_filblks_t *dnew, /* new delayed-alloc indirect blocks */ xfs_fsblock_t *first, /* pointer to firstblock variable */ xfs_bmap_free_t *flist, /* list of extents to be freed */ - int *logflagsp, /* inode logging flags */ - int rsvd); /* OK to allocate reserved blocks */ + int *logflagsp); /* inode logging flags */ /* * Called by xfs_bmap_add_extent to handle cases converting a hole @@ -127,10 +110,9 @@ xfs_bmap_add_extent_delay_real( STATIC int /* error */ xfs_bmap_add_extent_hole_delay( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ - int *logflagsp,/* inode logging flags */ - int rsvd); /* OK to allocate reserved blocks */ + int *logflagsp); /* inode logging flags */ /* * Called by xfs_bmap_add_extent to handle cases converting a hole @@ -139,7 +121,7 @@ xfs_bmap_add_extent_hole_delay( STATIC int /* error */ xfs_bmap_add_extent_hole_real( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_btree_cur_t *cur, /* if null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp, /* inode logging flags */ @@ -152,7 +134,7 @@ xfs_bmap_add_extent_hole_real( STATIC int /* error */ xfs_bmap_add_extent_unwritten_real( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp); /* inode logging flags */ @@ -180,22 +162,6 @@ xfs_bmap_btree_to_extents( int whichfork); /* data or attr fork */ /* - * Called by xfs_bmapi to update file extent records and the btree - * after removing space (or undoing a delayed allocation). - */ -STATIC int /* error */ -xfs_bmap_del_extent( - xfs_inode_t *ip, /* incore inode pointer */ - xfs_trans_t *tp, /* current trans pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ - xfs_bmap_free_t *flist, /* list of extents to be freed */ - xfs_btree_cur_t *cur, /* if null, not a btree */ - xfs_bmbt_irec_t *new, /* new data to add to file extents */ - int *logflagsp,/* inode logging flags */ - int whichfork, /* data or attr fork */ - int rsvd); /* OK to allocate reserved blocks */ - -/* * Remove the entry "free" from the free item list. Prev points to the * previous entry, unless "free" is the head of the list. */ @@ -474,14 +440,13 @@ xfs_bmap_add_attrfork_local( STATIC int /* error */ xfs_bmap_add_extent( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ xfs_fsblock_t *first, /* pointer to firstblock variable */ xfs_bmap_free_t *flist, /* list of extents to be freed */ int *logflagsp, /* inode logging flags */ - int whichfork, /* data or attr fork */ - int rsvd) /* OK to use reserved data blocks */ + int whichfork) /* data or attr fork */ { xfs_btree_cur_t *cur; /* btree cursor or null */ xfs_filblks_t da_new; /* new count del alloc blocks used */ @@ -492,23 +457,27 @@ xfs_bmap_add_extent( xfs_extnum_t nextents; /* number of extents in file now */ XFS_STATS_INC(xs_add_exlist); + cur = *curp; ifp = XFS_IFORK_PTR(ip, whichfork); nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t); - ASSERT(idx <= nextents); da_old = da_new = 0; error = 0; + + ASSERT(*idx >= 0); + ASSERT(*idx <= nextents); + /* * This is the first extent added to a new/empty file. * Special case this one, so other routines get to assume there are * already extents in the list. */ if (nextents == 0) { - xfs_iext_insert(ip, 0, 1, new, + xfs_iext_insert(ip, *idx, 1, new, whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0); ASSERT(cur == NULL); - ifp->if_lastex = 0; + if (!isnullstartblock(new->br_startblock)) { XFS_IFORK_NEXT_SET(ip, whichfork, 1); logflags = XFS_ILOG_CORE | xfs_ilog_fext(whichfork); @@ -522,27 +491,25 @@ xfs_bmap_add_extent( if (cur) ASSERT((cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL) == 0); - if ((error = xfs_bmap_add_extent_hole_delay(ip, idx, new, - &logflags, rsvd))) - goto done; + error = xfs_bmap_add_extent_hole_delay(ip, idx, new, + &logflags); } /* * Real allocation off the end of the file. */ - else if (idx == nextents) { + else if (*idx == nextents) { if (cur) ASSERT((cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL) == 0); - if ((error = xfs_bmap_add_extent_hole_real(ip, idx, cur, new, - &logflags, whichfork))) - goto done; + error = xfs_bmap_add_extent_hole_real(ip, idx, cur, new, + &logflags, whichfork); } else { xfs_bmbt_irec_t prev; /* old extent at offset idx */ /* * Get the record referred to by idx. */ - xfs_bmbt_get_all(xfs_iext_get_ext(ifp, idx), &prev); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &prev); /* * If it's a real allocation record, and the new allocation ends * after the start of the referred to record, then we're filling @@ -557,22 +524,18 @@ xfs_bmap_add_extent( if (cur) ASSERT(cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL); - if ((error = xfs_bmap_add_extent_delay_real(ip, - idx, &cur, new, &da_new, first, flist, - &logflags, rsvd))) - goto done; - } else if (new->br_state == XFS_EXT_NORM) { - ASSERT(new->br_state == XFS_EXT_NORM); - if ((error = xfs_bmap_add_extent_unwritten_real( - ip, idx, &cur, new, &logflags))) - goto done; + error = xfs_bmap_add_extent_delay_real(ip, + idx, &cur, new, &da_new, + first, flist, &logflags); } else { - ASSERT(new->br_state == XFS_EXT_UNWRITTEN); - if ((error = xfs_bmap_add_extent_unwritten_real( - ip, idx, &cur, new, &logflags))) + ASSERT(new->br_state == XFS_EXT_NORM || + new->br_state == XFS_EXT_UNWRITTEN); + + error = xfs_bmap_add_extent_unwritten_real(ip, + idx, &cur, new, &logflags); + if (error) goto done; } - ASSERT(*curp == cur || *curp == NULL); } /* * Otherwise we're filling in a hole with an allocation. @@ -581,13 +544,15 @@ xfs_bmap_add_extent( if (cur) ASSERT((cur->bc_private.b.flags & XFS_BTCUR_BPRV_WASDEL) == 0); - if ((error = xfs_bmap_add_extent_hole_real(ip, idx, cur, - new, &logflags, whichfork))) - goto done; + error = xfs_bmap_add_extent_hole_real(ip, idx, cur, + new, &logflags, whichfork); } } + if (error) + goto done; ASSERT(*curp == cur || *curp == NULL); + /* * Convert to a btree if necessary. */ @@ -615,7 +580,7 @@ xfs_bmap_add_extent( ASSERT(nblks <= da_old); if (nblks < da_old) xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS, - (int64_t)(da_old - nblks), rsvd); + (int64_t)(da_old - nblks), 0); } /* * Clear out the allocated field, done with it now in any case. @@ -640,14 +605,13 @@ done: STATIC int /* error */ xfs_bmap_add_extent_delay_real( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ xfs_filblks_t *dnew, /* new delayed-alloc indirect blocks */ xfs_fsblock_t *first, /* pointer to firstblock variable */ xfs_bmap_free_t *flist, /* list of extents to be freed */ - int *logflagsp, /* inode logging flags */ - int rsvd) /* OK to use reserved data block allocation */ + int *logflagsp) /* inode logging flags */ { xfs_btree_cur_t *cur; /* btree cursor */ int diff; /* temp value */ @@ -673,7 +637,7 @@ xfs_bmap_add_extent_delay_real( */ cur = *curp; ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK); - ep = xfs_iext_get_ext(ifp, idx); + ep = xfs_iext_get_ext(ifp, *idx); xfs_bmbt_get_all(ep, &PREV); new_endoff = new->br_startoff + new->br_blockcount; ASSERT(PREV.br_startoff <= new->br_startoff); @@ -692,9 +656,9 @@ xfs_bmap_add_extent_delay_real( * Check and set flags if this segment has a left neighbor. * Don't set contiguous if the combined extent would be too large. */ - if (idx > 0) { + if (*idx > 0) { state |= BMAP_LEFT_VALID; - xfs_bmbt_get_all(xfs_iext_get_ext(ifp, idx - 1), &LEFT); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &LEFT); if (isnullstartblock(LEFT.br_startblock)) state |= BMAP_LEFT_DELAY; @@ -712,9 +676,9 @@ xfs_bmap_add_extent_delay_real( * Don't set contiguous if the combined extent would be too large. * Also check for all-three-contiguous being too large. */ - if (idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) { + if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) { state |= BMAP_RIGHT_VALID; - xfs_bmbt_get_all(xfs_iext_get_ext(ifp, idx + 1), &RIGHT); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT); if (isnullstartblock(RIGHT.br_startblock)) state |= BMAP_RIGHT_DELAY; @@ -745,14 +709,14 @@ xfs_bmap_add_extent_delay_real( * Filling in all of a previously delayed allocation extent. * The left and right neighbors are both contiguous with new. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + --*idx; + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), LEFT.br_blockcount + PREV.br_blockcount + RIGHT.br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - xfs_iext_remove(ip, idx, 2, state); - ip->i_df.if_lastex = idx - 1; + xfs_iext_remove(ip, *idx + 1, 2, state); ip->i_d.di_nextents--; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -784,13 +748,14 @@ xfs_bmap_add_extent_delay_real( * Filling in all of a previously delayed allocation extent. * The left neighbor is contiguous, the right is not. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + --*idx; + + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), LEFT.br_blockcount + PREV.br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ip->i_df.if_lastex = idx - 1; - xfs_iext_remove(ip, idx, 1, state); + xfs_iext_remove(ip, *idx + 1, 1, state); if (cur == NULL) rval = XFS_ILOG_DEXT; else { @@ -814,14 +779,13 @@ xfs_bmap_add_extent_delay_real( * Filling in all of a previously delayed allocation extent. * The right neighbor is contiguous, the left is not. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_startblock(ep, new->br_startblock); xfs_bmbt_set_blockcount(ep, PREV.br_blockcount + RIGHT.br_blockcount); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ip->i_df.if_lastex = idx; - xfs_iext_remove(ip, idx + 1, 1, state); + xfs_iext_remove(ip, *idx + 1, 1, state); if (cur == NULL) rval = XFS_ILOG_DEXT; else { @@ -837,6 +801,7 @@ xfs_bmap_add_extent_delay_real( RIGHT.br_blockcount, PREV.br_state))) goto done; } + *dnew = 0; break; @@ -846,11 +811,10 @@ xfs_bmap_add_extent_delay_real( * Neither the left nor right neighbors are contiguous with * the new one. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_startblock(ep, new->br_startblock); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ip->i_df.if_lastex = idx; ip->i_d.di_nextents++; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -866,6 +830,7 @@ xfs_bmap_add_extent_delay_real( goto done; XFS_WANT_CORRUPTED_GOTO(i == 1, done); } + *dnew = 0; break; @@ -874,17 +839,16 @@ xfs_bmap_add_extent_delay_real( * Filling in the first part of a previous delayed allocation. * The left neighbor is contiguous. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + trace_xfs_bmap_pre_update(ip, *idx - 1, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx - 1), LEFT.br_blockcount + new->br_blockcount); xfs_bmbt_set_startoff(ep, PREV.br_startoff + new->br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx - 1, state, _THIS_IP_); temp = PREV.br_blockcount - new->br_blockcount; - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, temp); - ip->i_df.if_lastex = idx - 1; if (cur == NULL) rval = XFS_ILOG_DEXT; else { @@ -904,7 +868,9 @@ xfs_bmap_add_extent_delay_real( temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp), startblockval(PREV.br_startblock)); xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + + --*idx; *dnew = temp; break; @@ -913,12 +879,11 @@ xfs_bmap_add_extent_delay_real( * Filling in the first part of a previous delayed allocation. * The left neighbor is not contiguous. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_startoff(ep, new_endoff); temp = PREV.br_blockcount - new->br_blockcount; xfs_bmbt_set_blockcount(ep, temp); - xfs_iext_insert(ip, idx, 1, new, state); - ip->i_df.if_lastex = idx; + xfs_iext_insert(ip, *idx, 1, new, state); ip->i_d.di_nextents++; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -946,9 +911,10 @@ xfs_bmap_add_extent_delay_real( temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp), startblockval(PREV.br_startblock) - (cur ? cur->bc_private.b.allocated : 0)); - ep = xfs_iext_get_ext(ifp, idx + 1); + ep = xfs_iext_get_ext(ifp, *idx + 1); xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); - trace_xfs_bmap_post_update(ip, idx + 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx + 1, state, _THIS_IP_); + *dnew = temp; break; @@ -958,15 +924,13 @@ xfs_bmap_add_extent_delay_real( * The right neighbor is contiguous with the new allocation. */ temp = PREV.br_blockcount - new->br_blockcount; - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); - trace_xfs_bmap_pre_update(ip, idx + 1, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx + 1, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, temp); - xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, idx + 1), + xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx + 1), new->br_startoff, new->br_startblock, new->br_blockcount + RIGHT.br_blockcount, RIGHT.br_state); - trace_xfs_bmap_post_update(ip, idx + 1, state, _THIS_IP_); - ip->i_df.if_lastex = idx + 1; + trace_xfs_bmap_post_update(ip, *idx + 1, state, _THIS_IP_); if (cur == NULL) rval = XFS_ILOG_DEXT; else { @@ -983,10 +947,14 @@ xfs_bmap_add_extent_delay_real( RIGHT.br_state))) goto done; } + temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp), startblockval(PREV.br_startblock)); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + + ++*idx; *dnew = temp; break; @@ -996,10 +964,9 @@ xfs_bmap_add_extent_delay_real( * The right neighbor is not contiguous. */ temp = PREV.br_blockcount - new->br_blockcount; - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, temp); - xfs_iext_insert(ip, idx + 1, 1, new, state); - ip->i_df.if_lastex = idx + 1; + xfs_iext_insert(ip, *idx + 1, 1, new, state); ip->i_d.di_nextents++; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1027,9 +994,11 @@ xfs_bmap_add_extent_delay_real( temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp), startblockval(PREV.br_startblock) - (cur ? cur->bc_private.b.allocated : 0)); - ep = xfs_iext_get_ext(ifp, idx); + ep = xfs_iext_get_ext(ifp, *idx); xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + + ++*idx; *dnew = temp; break; @@ -1056,7 +1025,7 @@ xfs_bmap_add_extent_delay_real( */ temp = new->br_startoff - PREV.br_startoff; temp2 = PREV.br_startoff + PREV.br_blockcount - new_endoff; - trace_xfs_bmap_pre_update(ip, idx, 0, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, 0, _THIS_IP_); xfs_bmbt_set_blockcount(ep, temp); /* truncate PREV */ LEFT = *new; RIGHT.br_state = PREV.br_state; @@ -1065,8 +1034,7 @@ xfs_bmap_add_extent_delay_real( RIGHT.br_startoff = new_endoff; RIGHT.br_blockcount = temp2; /* insert LEFT (r[0]) and RIGHT (r[1]) at the same time */ - xfs_iext_insert(ip, idx + 1, 2, &LEFT, state); - ip->i_df.if_lastex = idx + 1; + xfs_iext_insert(ip, *idx + 1, 2, &LEFT, state); ip->i_d.di_nextents++; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1097,7 +1065,7 @@ xfs_bmap_add_extent_delay_real( (cur ? cur->bc_private.b.allocated : 0)); if (diff > 0 && xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS, - -((int64_t)diff), rsvd)) { + -((int64_t)diff), 0)) { /* * Ick gross gag me with a spoon. */ @@ -1109,7 +1077,7 @@ xfs_bmap_add_extent_delay_real( if (!diff || !xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS, - -((int64_t)diff), rsvd)) + -((int64_t)diff), 0)) break; } if (temp2) { @@ -1118,18 +1086,20 @@ xfs_bmap_add_extent_delay_real( if (!diff || !xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS, - -((int64_t)diff), rsvd)) + -((int64_t)diff), 0)) break; } } } - ep = xfs_iext_get_ext(ifp, idx); + ep = xfs_iext_get_ext(ifp, *idx); xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); - trace_xfs_bmap_pre_update(ip, idx + 2, state, _THIS_IP_); - xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, idx + 2), + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx + 2, state, _THIS_IP_); + xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx + 2), nullstartblock((int)temp2)); - trace_xfs_bmap_post_update(ip, idx + 2, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx + 2, state, _THIS_IP_); + + ++*idx; *dnew = temp + temp2; break; @@ -1161,7 +1131,7 @@ done: STATIC int /* error */ xfs_bmap_add_extent_unwritten_real( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_btree_cur_t **curp, /* if *curp is null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp) /* inode logging flags */ @@ -1188,7 +1158,7 @@ xfs_bmap_add_extent_unwritten_real( error = 0; cur = *curp; ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK); - ep = xfs_iext_get_ext(ifp, idx); + ep = xfs_iext_get_ext(ifp, *idx); xfs_bmbt_get_all(ep, &PREV); newext = new->br_state; oldext = (newext == XFS_EXT_UNWRITTEN) ? @@ -1211,9 +1181,9 @@ xfs_bmap_add_extent_unwritten_real( * Check and set flags if this segment has a left neighbor. * Don't set contiguous if the combined extent would be too large. */ - if (idx > 0) { + if (*idx > 0) { state |= BMAP_LEFT_VALID; - xfs_bmbt_get_all(xfs_iext_get_ext(ifp, idx - 1), &LEFT); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &LEFT); if (isnullstartblock(LEFT.br_startblock)) state |= BMAP_LEFT_DELAY; @@ -1231,9 +1201,9 @@ xfs_bmap_add_extent_unwritten_real( * Don't set contiguous if the combined extent would be too large. * Also check for all-three-contiguous being too large. */ - if (idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) { + if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t) - 1) { state |= BMAP_RIGHT_VALID; - xfs_bmbt_get_all(xfs_iext_get_ext(ifp, idx + 1), &RIGHT); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx + 1), &RIGHT); if (isnullstartblock(RIGHT.br_startblock)) state |= BMAP_RIGHT_DELAY; } @@ -1262,14 +1232,15 @@ xfs_bmap_add_extent_unwritten_real( * Setting all of a previous oldext extent to newext. * The left and right neighbors are both contiguous with new. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + --*idx; + + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), LEFT.br_blockcount + PREV.br_blockcount + RIGHT.br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - xfs_iext_remove(ip, idx, 2, state); - ip->i_df.if_lastex = idx - 1; + xfs_iext_remove(ip, *idx + 1, 2, state); ip->i_d.di_nextents -= 2; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1305,13 +1276,14 @@ xfs_bmap_add_extent_unwritten_real( * Setting all of a previous oldext extent to newext. * The left neighbor is contiguous, the right is not. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + --*idx; + + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), LEFT.br_blockcount + PREV.br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ip->i_df.if_lastex = idx - 1; - xfs_iext_remove(ip, idx, 1, state); + xfs_iext_remove(ip, *idx + 1, 1, state); ip->i_d.di_nextents--; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1341,13 +1313,12 @@ xfs_bmap_add_extent_unwritten_real( * Setting all of a previous oldext extent to newext. * The right neighbor is contiguous, the left is not. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, PREV.br_blockcount + RIGHT.br_blockcount); xfs_bmbt_set_state(ep, newext); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); - ip->i_df.if_lastex = idx; - xfs_iext_remove(ip, idx + 1, 1, state); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + xfs_iext_remove(ip, *idx + 1, 1, state); ip->i_d.di_nextents--; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1378,11 +1349,10 @@ xfs_bmap_add_extent_unwritten_real( * Neither the left nor right neighbors are contiguous with * the new one. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_state(ep, newext); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ip->i_df.if_lastex = idx; if (cur == NULL) rval = XFS_ILOG_DEXT; else { @@ -1404,21 +1374,22 @@ xfs_bmap_add_extent_unwritten_real( * Setting the first part of a previous oldext extent to newext. * The left neighbor is contiguous. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + trace_xfs_bmap_pre_update(ip, *idx - 1, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx - 1), LEFT.br_blockcount + new->br_blockcount); xfs_bmbt_set_startoff(ep, PREV.br_startoff + new->br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx - 1, state, _THIS_IP_); - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_startblock(ep, new->br_startblock + new->br_blockcount); xfs_bmbt_set_blockcount(ep, PREV.br_blockcount - new->br_blockcount); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + + --*idx; - ip->i_df.if_lastex = idx - 1; if (cur == NULL) rval = XFS_ILOG_DEXT; else { @@ -1449,17 +1420,16 @@ xfs_bmap_add_extent_unwritten_real( * Setting the first part of a previous oldext extent to newext. * The left neighbor is not contiguous. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); ASSERT(ep && xfs_bmbt_get_state(ep) == oldext); xfs_bmbt_set_startoff(ep, new_endoff); xfs_bmbt_set_blockcount(ep, PREV.br_blockcount - new->br_blockcount); xfs_bmbt_set_startblock(ep, new->br_startblock + new->br_blockcount); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - xfs_iext_insert(ip, idx, 1, new, state); - ip->i_df.if_lastex = idx; + xfs_iext_insert(ip, *idx, 1, new, state); ip->i_d.di_nextents++; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1488,17 +1458,19 @@ xfs_bmap_add_extent_unwritten_real( * Setting the last part of a previous oldext extent to newext. * The right neighbor is contiguous with the new allocation. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); - trace_xfs_bmap_pre_update(ip, idx + 1, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, PREV.br_blockcount - new->br_blockcount); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); - xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, idx + 1), + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + + ++*idx; + + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx), new->br_startoff, new->br_startblock, new->br_blockcount + RIGHT.br_blockcount, newext); - trace_xfs_bmap_post_update(ip, idx + 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ip->i_df.if_lastex = idx + 1; if (cur == NULL) rval = XFS_ILOG_DEXT; else { @@ -1528,13 +1500,14 @@ xfs_bmap_add_extent_unwritten_real( * Setting the last part of a previous oldext extent to newext. * The right neighbor is not contiguous. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, PREV.br_blockcount - new->br_blockcount); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + + ++*idx; + xfs_iext_insert(ip, *idx, 1, new, state); - xfs_iext_insert(ip, idx + 1, 1, new, state); - ip->i_df.if_lastex = idx + 1; ip->i_d.di_nextents++; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1568,10 +1541,10 @@ xfs_bmap_add_extent_unwritten_real( * newext. Contiguity is impossible here. * One extent becomes three extents. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, new->br_startoff - PREV.br_startoff); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); r[0] = *new; r[1].br_startoff = new_endoff; @@ -1579,8 +1552,10 @@ xfs_bmap_add_extent_unwritten_real( PREV.br_startoff + PREV.br_blockcount - new_endoff; r[1].br_startblock = new->br_startblock + new->br_blockcount; r[1].br_state = oldext; - xfs_iext_insert(ip, idx + 1, 2, &r[0], state); - ip->i_df.if_lastex = idx + 1; + + ++*idx; + xfs_iext_insert(ip, *idx, 2, &r[0], state); + ip->i_d.di_nextents += 2; if (cur == NULL) rval = XFS_ILOG_CORE | XFS_ILOG_DEXT; @@ -1650,12 +1625,10 @@ done: STATIC int /* error */ xfs_bmap_add_extent_hole_delay( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ - int *logflagsp, /* inode logging flags */ - int rsvd) /* OK to allocate reserved blocks */ + int *logflagsp) /* inode logging flags */ { - xfs_bmbt_rec_host_t *ep; /* extent record for idx */ xfs_ifork_t *ifp; /* inode fork pointer */ xfs_bmbt_irec_t left; /* left neighbor extent entry */ xfs_filblks_t newlen=0; /* new indirect size */ @@ -1665,16 +1638,15 @@ xfs_bmap_add_extent_hole_delay( xfs_filblks_t temp=0; /* temp for indirect calculations */ ifp = XFS_IFORK_PTR(ip, XFS_DATA_FORK); - ep = xfs_iext_get_ext(ifp, idx); state = 0; ASSERT(isnullstartblock(new->br_startblock)); /* * Check and set flags if this segment has a left neighbor */ - if (idx > 0) { + if (*idx > 0) { state |= BMAP_LEFT_VALID; - xfs_bmbt_get_all(xfs_iext_get_ext(ifp, idx - 1), &left); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left); if (isnullstartblock(left.br_startblock)) state |= BMAP_LEFT_DELAY; @@ -1684,9 +1656,9 @@ xfs_bmap_add_extent_hole_delay( * Check and set flags if the current (right) segment exists. * If it doesn't exist, we're converting the hole at end-of-file. */ - if (idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) { + if (*idx < ip->i_df.if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) { state |= BMAP_RIGHT_VALID; - xfs_bmbt_get_all(ep, &right); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right); if (isnullstartblock(right.br_startblock)) state |= BMAP_RIGHT_DELAY; @@ -1719,21 +1691,21 @@ xfs_bmap_add_extent_hole_delay( * on the left and on the right. * Merge all three into a single extent record. */ + --*idx; temp = left.br_blockcount + new->br_blockcount + right.br_blockcount; - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), temp); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp); oldlen = startblockval(left.br_startblock) + startblockval(new->br_startblock) + startblockval(right.br_startblock); newlen = xfs_bmap_worst_indlen(ip, temp); - xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, idx - 1), + xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx), nullstartblock((int)newlen)); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - xfs_iext_remove(ip, idx, 1, state); - ip->i_df.if_lastex = idx - 1; + xfs_iext_remove(ip, *idx + 1, 1, state); break; case BMAP_LEFT_CONTIG: @@ -1742,17 +1714,17 @@ xfs_bmap_add_extent_hole_delay( * on the left. * Merge the new allocation with the left neighbor. */ + --*idx; temp = left.br_blockcount + new->br_blockcount; - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), temp); + + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), temp); oldlen = startblockval(left.br_startblock) + startblockval(new->br_startblock); newlen = xfs_bmap_worst_indlen(ip, temp); - xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, idx - 1), + xfs_bmbt_set_startblock(xfs_iext_get_ext(ifp, *idx), nullstartblock((int)newlen)); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); - - ip->i_df.if_lastex = idx - 1; + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); break; case BMAP_RIGHT_CONTIG: @@ -1761,16 +1733,15 @@ xfs_bmap_add_extent_hole_delay( * on the right. * Merge the new allocation with the right neighbor. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); temp = new->br_blockcount + right.br_blockcount; oldlen = startblockval(new->br_startblock) + startblockval(right.br_startblock); newlen = xfs_bmap_worst_indlen(ip, temp); - xfs_bmbt_set_allf(ep, new->br_startoff, + xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx), + new->br_startoff, nullstartblock((int)newlen), temp, right.br_state); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); - - ip->i_df.if_lastex = idx; + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); break; case 0: @@ -1780,14 +1751,13 @@ xfs_bmap_add_extent_hole_delay( * Insert a new entry. */ oldlen = newlen = 0; - xfs_iext_insert(ip, idx, 1, new, state); - ip->i_df.if_lastex = idx; + xfs_iext_insert(ip, *idx, 1, new, state); break; } if (oldlen != newlen) { ASSERT(oldlen > newlen); xfs_icsb_modify_counters(ip->i_mount, XFS_SBS_FDBLOCKS, - (int64_t)(oldlen - newlen), rsvd); + (int64_t)(oldlen - newlen), 0); /* * Nothing to do for disk quota accounting here. */ @@ -1803,13 +1773,12 @@ xfs_bmap_add_extent_hole_delay( STATIC int /* error */ xfs_bmap_add_extent_hole_real( xfs_inode_t *ip, /* incore inode pointer */ - xfs_extnum_t idx, /* extent number to update/insert */ + xfs_extnum_t *idx, /* extent number to update/insert */ xfs_btree_cur_t *cur, /* if null, not a btree */ xfs_bmbt_irec_t *new, /* new data to add to file extents */ int *logflagsp, /* inode logging flags */ int whichfork) /* data or attr fork */ { - xfs_bmbt_rec_host_t *ep; /* pointer to extent entry ins. point */ int error; /* error return value */ int i; /* temp state */ xfs_ifork_t *ifp; /* inode fork pointer */ @@ -1819,8 +1788,7 @@ xfs_bmap_add_extent_hole_real( int state; /* state bits, accessed thru macros */ ifp = XFS_IFORK_PTR(ip, whichfork); - ASSERT(idx <= ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)); - ep = xfs_iext_get_ext(ifp, idx); + ASSERT(*idx <= ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)); state = 0; if (whichfork == XFS_ATTR_FORK) @@ -1829,9 +1797,9 @@ xfs_bmap_add_extent_hole_real( /* * Check and set flags if this segment has a left neighbor. */ - if (idx > 0) { + if (*idx > 0) { state |= BMAP_LEFT_VALID; - xfs_bmbt_get_all(xfs_iext_get_ext(ifp, idx - 1), &left); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx - 1), &left); if (isnullstartblock(left.br_startblock)) state |= BMAP_LEFT_DELAY; } @@ -1840,9 +1808,9 @@ xfs_bmap_add_extent_hole_real( * Check and set flags if this segment has a current value. * Not true if we're inserting into the "hole" at eof. */ - if (idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) { + if (*idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)) { state |= BMAP_RIGHT_VALID; - xfs_bmbt_get_all(ep, &right); + xfs_bmbt_get_all(xfs_iext_get_ext(ifp, *idx), &right); if (isnullstartblock(right.br_startblock)) state |= BMAP_RIGHT_DELAY; } @@ -1879,14 +1847,15 @@ xfs_bmap_add_extent_hole_real( * left and on the right. * Merge all three into a single extent record. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + --*idx; + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), left.br_blockcount + new->br_blockcount + right.br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + + xfs_iext_remove(ip, *idx + 1, 1, state); - xfs_iext_remove(ip, idx, 1, state); - ifp->if_lastex = idx - 1; XFS_IFORK_NEXT_SET(ip, whichfork, XFS_IFORK_NEXTENTS(ip, whichfork) - 1); if (cur == NULL) { @@ -1921,12 +1890,12 @@ xfs_bmap_add_extent_hole_real( * on the left. * Merge the new allocation with the left neighbor. */ - trace_xfs_bmap_pre_update(ip, idx - 1, state, _THIS_IP_); - xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, idx - 1), + --*idx; + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_blockcount(xfs_iext_get_ext(ifp, *idx), left.br_blockcount + new->br_blockcount); - trace_xfs_bmap_post_update(ip, idx - 1, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ifp->if_lastex = idx - 1; if (cur == NULL) { rval = xfs_ilog_fext(whichfork); } else { @@ -1952,13 +1921,13 @@ xfs_bmap_add_extent_hole_real( * on the right. * Merge the new allocation with the right neighbor. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); - xfs_bmbt_set_allf(ep, new->br_startoff, new->br_startblock, + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); + xfs_bmbt_set_allf(xfs_iext_get_ext(ifp, *idx), + new->br_startoff, new->br_startblock, new->br_blockcount + right.br_blockcount, right.br_state); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); - ifp->if_lastex = idx; if (cur == NULL) { rval = xfs_ilog_fext(whichfork); } else { @@ -1984,8 +1953,7 @@ xfs_bmap_add_extent_hole_real( * real allocation. * Insert a new entry. */ - xfs_iext_insert(ip, idx, 1, new, state); - ifp->if_lastex = idx; + xfs_iext_insert(ip, *idx, 1, new, state); XFS_IFORK_NEXT_SET(ip, whichfork, XFS_IFORK_NEXTENTS(ip, whichfork) + 1); if (cur == NULL) { @@ -2833,13 +2801,12 @@ STATIC int /* error */ xfs_bmap_del_extent( xfs_inode_t *ip, /* incore inode pointer */ xfs_trans_t *tp, /* current transaction pointer */ - xfs_extnum_t idx, /* extent number to update/delete */ + xfs_extnum_t *idx, /* extent number to update/delete */ xfs_bmap_free_t *flist, /* list of extents to be freed */ xfs_btree_cur_t *cur, /* if null, not a btree */ xfs_bmbt_irec_t *del, /* data to remove from extents */ int *logflagsp, /* inode logging flags */ - int whichfork, /* data or attr fork */ - int rsvd) /* OK to allocate reserved blocks */ + int whichfork) /* data or attr fork */ { xfs_filblks_t da_new; /* new delay-alloc indirect blocks */ xfs_filblks_t da_old; /* old delay-alloc indirect blocks */ @@ -2870,10 +2837,10 @@ xfs_bmap_del_extent( mp = ip->i_mount; ifp = XFS_IFORK_PTR(ip, whichfork); - ASSERT((idx >= 0) && (idx < ifp->if_bytes / + ASSERT((*idx >= 0) && (*idx < ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t))); ASSERT(del->br_blockcount > 0); - ep = xfs_iext_get_ext(ifp, idx); + ep = xfs_iext_get_ext(ifp, *idx); xfs_bmbt_get_all(ep, &got); ASSERT(got.br_startoff <= del->br_startoff); del_endoff = del->br_startoff + del->br_blockcount; @@ -2947,11 +2914,12 @@ xfs_bmap_del_extent( /* * Matches the whole extent. Delete the entry. */ - xfs_iext_remove(ip, idx, 1, + xfs_iext_remove(ip, *idx, 1, whichfork == XFS_ATTR_FORK ? BMAP_ATTRFORK : 0); - ifp->if_lastex = idx; + --*idx; if (delay) break; + XFS_IFORK_NEXT_SET(ip, whichfork, XFS_IFORK_NEXTENTS(ip, whichfork) - 1); flags |= XFS_ILOG_CORE; @@ -2968,21 +2936,20 @@ xfs_bmap_del_extent( /* * Deleting the first part of the extent. */ - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_startoff(ep, del_endoff); temp = got.br_blockcount - del->br_blockcount; xfs_bmbt_set_blockcount(ep, temp); - ifp->if_lastex = idx; if (delay) { temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp), da_old); xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); da_new = temp; break; } xfs_bmbt_set_startblock(ep, del_endblock); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); if (!cur) { flags |= xfs_ilog_fext(whichfork); break; @@ -2998,18 +2965,17 @@ xfs_bmap_del_extent( * Deleting the last part of the extent. */ temp = got.br_blockcount - del->br_blockcount; - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, temp); - ifp->if_lastex = idx; if (delay) { temp = XFS_FILBLKS_MIN(xfs_bmap_worst_indlen(ip, temp), da_old); xfs_bmbt_set_startblock(ep, nullstartblock((int)temp)); - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); da_new = temp; break; } - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); if (!cur) { flags |= xfs_ilog_fext(whichfork); break; @@ -3026,7 +2992,7 @@ xfs_bmap_del_extent( * Deleting the middle of the extent. */ temp = del->br_startoff - got.br_startoff; - trace_xfs_bmap_pre_update(ip, idx, state, _THIS_IP_); + trace_xfs_bmap_pre_update(ip, *idx, state, _THIS_IP_); xfs_bmbt_set_blockcount(ep, temp); new.br_startoff = del_endoff; temp2 = got_endoff - del_endoff; @@ -3113,9 +3079,9 @@ xfs_bmap_del_extent( } } } - trace_xfs_bmap_post_update(ip, idx, state, _THIS_IP_); - xfs_iext_insert(ip, idx + 1, 1, &new, state); - ifp->if_lastex = idx + 1; + trace_xfs_bmap_post_update(ip, *idx, state, _THIS_IP_); + xfs_iext_insert(ip, *idx + 1, 1, &new, state); + ++*idx; break; } /* @@ -3142,7 +3108,7 @@ xfs_bmap_del_extent( ASSERT(da_old >= da_new); if (da_old > da_new) { xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, - (int64_t)(da_old - da_new), rsvd); + (int64_t)(da_old - da_new), 0); } done: *logflagsp = flags; @@ -4562,29 +4528,24 @@ xfs_bmapi( if (rt) { error = xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS, - -((int64_t)extsz), (flags & - XFS_BMAPI_RSVBLOCKS)); + -((int64_t)extsz), 0); } else { error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, - -((int64_t)alen), (flags & - XFS_BMAPI_RSVBLOCKS)); + -((int64_t)alen), 0); } if (!error) { error = xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, - -((int64_t)indlen), (flags & - XFS_BMAPI_RSVBLOCKS)); + -((int64_t)indlen), 0); if (error && rt) xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS, - (int64_t)extsz, (flags & - XFS_BMAPI_RSVBLOCKS)); + (int64_t)extsz, 0); else if (error) xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, - (int64_t)alen, (flags & - XFS_BMAPI_RSVBLOCKS)); + (int64_t)alen, 0); } if (error) { @@ -4701,13 +4662,12 @@ xfs_bmapi( if (!wasdelay && (flags & XFS_BMAPI_PREALLOC)) got.br_state = XFS_EXT_UNWRITTEN; } - error = xfs_bmap_add_extent(ip, lastx, &cur, &got, + error = xfs_bmap_add_extent(ip, &lastx, &cur, &got, firstblock, flist, &tmp_logflags, - whichfork, (flags & XFS_BMAPI_RSVBLOCKS)); + whichfork); logflags |= tmp_logflags; if (error) goto error0; - lastx = ifp->if_lastex; ep = xfs_iext_get_ext(ifp, lastx); nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t); xfs_bmbt_get_all(ep, &got); @@ -4803,13 +4763,12 @@ xfs_bmapi( mval->br_state = (mval->br_state == XFS_EXT_UNWRITTEN) ? XFS_EXT_NORM : XFS_EXT_UNWRITTEN; - error = xfs_bmap_add_extent(ip, lastx, &cur, mval, + error = xfs_bmap_add_extent(ip, &lastx, &cur, mval, firstblock, flist, &tmp_logflags, - whichfork, (flags & XFS_BMAPI_RSVBLOCKS)); + whichfork); logflags |= tmp_logflags; if (error) goto error0; - lastx = ifp->if_lastex; ep = xfs_iext_get_ext(ifp, lastx); nextents = ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t); xfs_bmbt_get_all(ep, &got); @@ -4868,14 +4827,14 @@ xfs_bmapi( /* * Else go on to the next record. */ - ep = xfs_iext_get_ext(ifp, ++lastx); prev = got; - if (lastx >= nextents) - eof = 1; - else + if (++lastx < nextents) { + ep = xfs_iext_get_ext(ifp, lastx); xfs_bmbt_get_all(ep, &got); + } else { + eof = 1; + } } - ifp->if_lastex = lastx; *nmap = n; /* * Transform from btree to extents, give it cur. @@ -4984,7 +4943,6 @@ xfs_bmapi_single( ASSERT(!isnullstartblock(got.br_startblock)); ASSERT(bno < got.br_startoff + got.br_blockcount); *fsb = got.br_startblock + (bno - got.br_startoff); - ifp->if_lastex = lastx; return 0; } @@ -5026,7 +4984,6 @@ xfs_bunmapi( int tmp_logflags; /* partial logging flags */ int wasdel; /* was a delayed alloc extent */ int whichfork; /* data or attribute fork */ - int rsvd; /* OK to allocate reserved blocks */ xfs_fsblock_t sum; trace_xfs_bunmap(ip, bno, len, flags, _RET_IP_); @@ -5044,7 +5001,7 @@ xfs_bunmapi( mp = ip->i_mount; if (XFS_FORCED_SHUTDOWN(mp)) return XFS_ERROR(EIO); - rsvd = (flags & XFS_BMAPI_RSVBLOCKS) != 0; + ASSERT(len > 0); ASSERT(nexts >= 0); ASSERT(ifp->if_ext_max == @@ -5160,9 +5117,9 @@ xfs_bunmapi( del.br_blockcount = mod; } del.br_state = XFS_EXT_UNWRITTEN; - error = xfs_bmap_add_extent(ip, lastx, &cur, &del, + error = xfs_bmap_add_extent(ip, &lastx, &cur, &del, firstblock, flist, &logflags, - XFS_DATA_FORK, 0); + XFS_DATA_FORK); if (error) goto error0; goto nodelete; @@ -5188,9 +5145,12 @@ xfs_bunmapi( */ ASSERT(bno >= del.br_blockcount); bno -= del.br_blockcount; - if (bno < got.br_startoff) { - if (--lastx >= 0) - xfs_bmbt_get_all(--ep, &got); + if (got.br_startoff > bno) { + if (--lastx >= 0) { + ep = xfs_iext_get_ext(ifp, + lastx); + xfs_bmbt_get_all(ep, &got); + } } continue; } else if (del.br_state == XFS_EXT_UNWRITTEN) { @@ -5214,18 +5174,19 @@ xfs_bunmapi( prev.br_startoff = start; } prev.br_state = XFS_EXT_UNWRITTEN; - error = xfs_bmap_add_extent(ip, lastx - 1, &cur, + lastx--; + error = xfs_bmap_add_extent(ip, &lastx, &cur, &prev, firstblock, flist, &logflags, - XFS_DATA_FORK, 0); + XFS_DATA_FORK); if (error) goto error0; goto nodelete; } else { ASSERT(del.br_state == XFS_EXT_NORM); del.br_state = XFS_EXT_UNWRITTEN; - error = xfs_bmap_add_extent(ip, lastx, &cur, + error = xfs_bmap_add_extent(ip, &lastx, &cur, &del, firstblock, flist, &logflags, - XFS_DATA_FORK, 0); + XFS_DATA_FORK); if (error) goto error0; goto nodelete; @@ -5240,13 +5201,13 @@ xfs_bunmapi( rtexts = XFS_FSB_TO_B(mp, del.br_blockcount); do_div(rtexts, mp->m_sb.sb_rextsize); xfs_mod_incore_sb(mp, XFS_SBS_FREXTENTS, - (int64_t)rtexts, rsvd); + (int64_t)rtexts, 0); (void)xfs_trans_reserve_quota_nblks(NULL, ip, -((long)del.br_blockcount), 0, XFS_QMOPT_RES_RTBLKS); } else { xfs_icsb_modify_counters(mp, XFS_SBS_FDBLOCKS, - (int64_t)del.br_blockcount, rsvd); + (int64_t)del.br_blockcount, 0); (void)xfs_trans_reserve_quota_nblks(NULL, ip, -((long)del.br_blockcount), 0, XFS_QMOPT_RES_REGBLKS); @@ -5277,31 +5238,29 @@ xfs_bunmapi( error = XFS_ERROR(ENOSPC); goto error0; } - error = xfs_bmap_del_extent(ip, tp, lastx, flist, cur, &del, - &tmp_logflags, whichfork, rsvd); + error = xfs_bmap_del_extent(ip, tp, &lastx, flist, cur, &del, + &tmp_logflags, whichfork); logflags |= tmp_logflags; if (error) goto error0; bno = del.br_startoff - 1; nodelete: - lastx = ifp->if_lastex; /* * If not done go on to the next (previous) record. - * Reset ep in case the extents array was re-alloced. */ - ep = xfs_iext_get_ext(ifp, lastx); if (bno != (xfs_fileoff_t)-1 && bno >= start) { - if (lastx >= XFS_IFORK_NEXTENTS(ip, whichfork) || - xfs_bmbt_get_startoff(ep) > bno) { - if (--lastx >= 0) - ep = xfs_iext_get_ext(ifp, lastx); - } - if (lastx >= 0) + if (lastx >= 0) { + ep = xfs_iext_get_ext(ifp, lastx); + if (xfs_bmbt_get_startoff(ep) > bno) { + if (--lastx >= 0) + ep = xfs_iext_get_ext(ifp, + lastx); + } xfs_bmbt_get_all(ep, &got); + } extno++; } } - ifp->if_lastex = lastx; *done = bno == (xfs_fileoff_t)-1 || bno < start || lastx < 0; ASSERT(ifp->if_ext_max == XFS_IFORK_SIZE(ip, whichfork) / (uint)sizeof(xfs_bmbt_rec_t)); diff --git a/fs/xfs/xfs_bmap.h b/fs/xfs/xfs_bmap.h index 3651191daea1..c62234bde053 100644 --- a/fs/xfs/xfs_bmap.h +++ b/fs/xfs/xfs_bmap.h @@ -69,7 +69,6 @@ typedef struct xfs_bmap_free #define XFS_BMAPI_ENTIRE 0x004 /* return entire extent, not trimmed */ #define XFS_BMAPI_METADATA 0x008 /* mapping metadata not user data */ #define XFS_BMAPI_ATTRFORK 0x010 /* use attribute fork not data */ -#define XFS_BMAPI_RSVBLOCKS 0x020 /* OK to alloc. reserved data blocks */ #define XFS_BMAPI_PREALLOC 0x040 /* preallocation op: unwritten space */ #define XFS_BMAPI_IGSTATE 0x080 /* Ignore state - */ /* combine contig. space */ @@ -87,7 +86,6 @@ typedef struct xfs_bmap_free { XFS_BMAPI_ENTIRE, "ENTIRE" }, \ { XFS_BMAPI_METADATA, "METADATA" }, \ { XFS_BMAPI_ATTRFORK, "ATTRFORK" }, \ - { XFS_BMAPI_RSVBLOCKS, "RSVBLOCKS" }, \ { XFS_BMAPI_PREALLOC, "PREALLOC" }, \ { XFS_BMAPI_IGSTATE, "IGSTATE" }, \ { XFS_BMAPI_CONTIG, "CONTIG" }, \ diff --git a/fs/xfs/xfs_buf_item.c b/fs/xfs/xfs_buf_item.c index e5413d96f1af..7b7e005e3dcc 100644 --- a/fs/xfs/xfs_buf_item.c +++ b/fs/xfs/xfs_buf_item.c @@ -992,7 +992,7 @@ xfs_buf_iodone_callbacks( lasttarg = XFS_BUF_TARGET(bp); /* - * If the write was asynchronous then noone will be looking for the + * If the write was asynchronous then no one will be looking for the * error. Clear the error state and write the buffer out again. * * During sync or umount we'll write all pending buffers again diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index be628677c288..9a84a85c03b1 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -202,7 +202,7 @@ xfs_swap_extents( xfs_inode_t *tip, /* tmp inode */ xfs_swapext_t *sxp) { - xfs_mount_t *mp; + xfs_mount_t *mp = ip->i_mount; xfs_trans_t *tp; xfs_bstat_t *sbp = &sxp->sx_stat; xfs_ifork_t *tempifp, *ifp, *tifp; @@ -212,16 +212,12 @@ xfs_swap_extents( int taforkblks = 0; __uint64_t tmp; - mp = ip->i_mount; - tempifp = kmem_alloc(sizeof(xfs_ifork_t), KM_MAYFAIL); if (!tempifp) { error = XFS_ERROR(ENOMEM); goto out; } - sbp = &sxp->sx_stat; - /* * we have to do two separate lock calls here to keep lockdep * happy. If we try to get all the locks in one call, lock will diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 742c8330994a..a098a20ca63e 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -920,7 +920,6 @@ xfs_iread_extents( /* * We know that the size is valid (it's checked in iformat_btree) */ - ifp->if_lastex = NULLEXTNUM; ifp->if_bytes = ifp->if_real_bytes = 0; ifp->if_flags |= XFS_IFEXTENTS; xfs_iext_add(ifp, 0, nextents); @@ -1354,7 +1353,7 @@ xfs_itruncate_start( return 0; } last_byte = xfs_file_last_byte(ip); - trace_xfs_itruncate_start(ip, flags, new_size, toss_start, last_byte); + trace_xfs_itruncate_start(ip, new_size, flags, toss_start, last_byte); if (last_byte > toss_start) { if (flags & XFS_ITRUNC_DEFINITE) { xfs_tosspages(ip, toss_start, @@ -1470,7 +1469,7 @@ xfs_itruncate_finish( * file but the log buffers containing the free and reallocation * don't, then we'd end up with garbage in the blocks being freed. * As long as we make the new_size permanent before actually - * freeing any blocks it doesn't matter if they get writtten to. + * freeing any blocks it doesn't matter if they get written to. * * The callers must signal into us whether or not the size * setting here must be synchronous. There are a few cases @@ -2558,12 +2557,9 @@ xfs_iflush_fork( case XFS_DINODE_FMT_EXTENTS: ASSERT((ifp->if_flags & XFS_IFEXTENTS) || !(iip->ili_format.ilf_fields & extflag[whichfork])); - ASSERT((xfs_iext_get_ext(ifp, 0) != NULL) || - (ifp->if_bytes == 0)); - ASSERT((xfs_iext_get_ext(ifp, 0) == NULL) || - (ifp->if_bytes > 0)); if ((iip->ili_format.ilf_fields & extflag[whichfork]) && (ifp->if_bytes > 0)) { + ASSERT(xfs_iext_get_ext(ifp, 0)); ASSERT(XFS_IFORK_NEXTENTS(ip, whichfork) > 0); (void)xfs_iextents_copy(ip, (xfs_bmbt_rec_t *)cp, whichfork); @@ -2789,7 +2785,7 @@ xfs_iflush( /* * We can't flush the inode until it is unpinned, so wait for it if we - * are allowed to block. We know noone new can pin it, because we are + * are allowed to block. We know no one new can pin it, because we are * holding the inode lock shared and you need to hold it exclusively to * pin the inode. * @@ -3112,6 +3108,8 @@ xfs_iext_get_ext( xfs_extnum_t idx) /* index of target extent */ { ASSERT(idx >= 0); + ASSERT(idx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t)); + if ((ifp->if_flags & XFS_IFEXTIREC) && (idx == 0)) { return ifp->if_u1.if_ext_irec->er_extbuf; } else if (ifp->if_flags & XFS_IFEXTIREC) { @@ -3191,7 +3189,6 @@ xfs_iext_add( } ifp->if_u1.if_extents = ifp->if_u2.if_inline_ext; ifp->if_real_bytes = 0; - ifp->if_lastex = nextents + ext_diff; } /* * Otherwise use a linear (direct) extent list. @@ -3886,8 +3883,10 @@ xfs_iext_idx_to_irec( xfs_extnum_t page_idx = *idxp; /* extent index in target list */ ASSERT(ifp->if_flags & XFS_IFEXTIREC); - ASSERT(page_idx >= 0 && page_idx <= - ifp->if_bytes / (uint)sizeof(xfs_bmbt_rec_t)); + ASSERT(page_idx >= 0); + ASSERT(page_idx <= ifp->if_bytes / sizeof(xfs_bmbt_rec_t)); + ASSERT(page_idx < ifp->if_bytes / sizeof(xfs_bmbt_rec_t) || realloc); + nlists = ifp->if_real_bytes / XFS_IEXT_BUFSZ; erp_idx = 0; low = 0; diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index f753200cef8d..3ae6d58e5473 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -67,7 +67,6 @@ typedef struct xfs_ifork { short if_broot_bytes; /* bytes allocated for root */ unsigned char if_flags; /* per-fork flags */ unsigned char if_ext_max; /* max # of extent records */ - xfs_extnum_t if_lastex; /* last if_extents used */ union { xfs_bmbt_rec_host_t *if_extents;/* linear map file exts */ xfs_ext_irec_t *if_ext_irec; /* irec map file exts */ @@ -111,7 +110,7 @@ struct xfs_imap { * Generally, we do not want to hold the i_rlock while holding the * i_ilock. Hierarchy is i_iolock followed by i_rlock. * - * xfs_iptr_t contains all the inode fields upto and including the + * xfs_iptr_t contains all the inode fields up to and including the * i_mnext and i_mprev fields, it is used as a marker in the inode * chain off the mount structure by xfs_sync calls. */ @@ -336,7 +335,7 @@ xfs_iflags_test_and_clear(xfs_inode_t *ip, unsigned short flags) /* * Project quota id helpers (previously projid was 16bit only - * and using two 16bit values to hold new 32bit projid was choosen + * and using two 16bit values to hold new 32bit projid was chosen * to retain compatibility with "old" filesystems). */ static inline prid_t diff --git a/fs/xfs/xfs_inode_item.c b/fs/xfs/xfs_inode_item.c index 46cc40131d4a..09983a3344a5 100644 --- a/fs/xfs/xfs_inode_item.c +++ b/fs/xfs/xfs_inode_item.c @@ -198,6 +198,41 @@ xfs_inode_item_size( } /* + * xfs_inode_item_format_extents - convert in-core extents to on-disk form + * + * For either the data or attr fork in extent format, we need to endian convert + * the in-core extent as we place them into the on-disk inode. In this case, we + * need to do this conversion before we write the extents into the log. Because + * we don't have the disk inode to write into here, we allocate a buffer and + * format the extents into it via xfs_iextents_copy(). We free the buffer in + * the unlock routine after the copy for the log has been made. + * + * In the case of the data fork, the in-core and on-disk fork sizes can be + * different due to delayed allocation extents. We only log on-disk extents + * here, so always use the physical fork size to determine the size of the + * buffer we need to allocate. + */ +STATIC void +xfs_inode_item_format_extents( + struct xfs_inode *ip, + struct xfs_log_iovec *vecp, + int whichfork, + int type) +{ + xfs_bmbt_rec_t *ext_buffer; + + ext_buffer = kmem_alloc(XFS_IFORK_SIZE(ip, whichfork), KM_SLEEP); + if (whichfork == XFS_DATA_FORK) + ip->i_itemp->ili_extents_buf = ext_buffer; + else + ip->i_itemp->ili_aextents_buf = ext_buffer; + + vecp->i_addr = ext_buffer; + vecp->i_len = xfs_iextents_copy(ip, ext_buffer, whichfork); + vecp->i_type = type; +} + +/* * This is called to fill in the vector of log iovecs for the * given inode log item. It fills the first item with an inode * log format structure, the second with the on-disk inode structure, @@ -213,7 +248,6 @@ xfs_inode_item_format( struct xfs_inode *ip = iip->ili_inode; uint nvecs; size_t data_bytes; - xfs_bmbt_rec_t *ext_buffer; xfs_mount_t *mp; vecp->i_addr = &iip->ili_format; @@ -320,22 +354,8 @@ xfs_inode_item_format( } else #endif { - /* - * There are delayed allocation extents - * in the inode, or we need to convert - * the extents to on disk format. - * Use xfs_iextents_copy() - * to copy only the real extents into - * a separate buffer. We'll free the - * buffer in the unlock routine. - */ - ext_buffer = kmem_alloc(ip->i_df.if_bytes, - KM_SLEEP); - iip->ili_extents_buf = ext_buffer; - vecp->i_addr = ext_buffer; - vecp->i_len = xfs_iextents_copy(ip, ext_buffer, - XFS_DATA_FORK); - vecp->i_type = XLOG_REG_TYPE_IEXT; + xfs_inode_item_format_extents(ip, vecp, + XFS_DATA_FORK, XLOG_REG_TYPE_IEXT); } ASSERT(vecp->i_len <= ip->i_df.if_bytes); iip->ili_format.ilf_dsize = vecp->i_len; @@ -445,19 +465,12 @@ xfs_inode_item_format( */ vecp->i_addr = ip->i_afp->if_u1.if_extents; vecp->i_len = ip->i_afp->if_bytes; + vecp->i_type = XLOG_REG_TYPE_IATTR_EXT; #else ASSERT(iip->ili_aextents_buf == NULL); - /* - * Need to endian flip before logging - */ - ext_buffer = kmem_alloc(ip->i_afp->if_bytes, - KM_SLEEP); - iip->ili_aextents_buf = ext_buffer; - vecp->i_addr = ext_buffer; - vecp->i_len = xfs_iextents_copy(ip, ext_buffer, - XFS_ATTR_FORK); + xfs_inode_item_format_extents(ip, vecp, + XFS_ATTR_FORK, XLOG_REG_TYPE_IATTR_EXT); #endif - vecp->i_type = XLOG_REG_TYPE_IATTR_EXT; iip->ili_format.ilf_asize = vecp->i_len; vecp++; nvecs++; @@ -957,7 +970,6 @@ xfs_iflush_abort( { xfs_inode_log_item_t *iip = ip->i_itemp; - iip = ip->i_itemp; if (iip) { struct xfs_ail *ailp = iip->ili_item.li_ailp; if (iip->ili_item.li_flags & XFS_LI_IN_AIL) { diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index dc1882adaf54..751e94fe1f77 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -204,7 +204,6 @@ xfs_bulkstat( xfs_agi_t *agi; /* agi header data */ xfs_agino_t agino; /* inode # in allocation group */ xfs_agnumber_t agno; /* allocation group number */ - xfs_daddr_t bno; /* inode cluster start daddr */ int chunkidx; /* current index into inode chunk */ int clustidx; /* current index into inode cluster */ xfs_btree_cur_t *cur; /* btree cursor for ialloc btree */ @@ -463,7 +462,6 @@ xfs_bulkstat( mp->m_sb.sb_inopblog); } ino = XFS_AGINO_TO_INO(mp, agno, agino); - bno = XFS_AGB_TO_DADDR(mp, agno, agbno); /* * Skip if this inode is free. */ diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 25efa9b8a602..211930246f20 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -761,7 +761,7 @@ xfs_log_need_covered(xfs_mount_t *mp) break; case XLOG_STATE_COVER_NEED: case XLOG_STATE_COVER_NEED2: - if (!xfs_trans_ail_tail(log->l_ailp) && + if (!xfs_ail_min_lsn(log->l_ailp) && xlog_iclogs_empty(log)) { if (log->l_covered_state == XLOG_STATE_COVER_NEED) log->l_covered_state = XLOG_STATE_COVER_DONE; @@ -801,7 +801,7 @@ xlog_assign_tail_lsn( xfs_lsn_t tail_lsn; struct log *log = mp->m_log; - tail_lsn = xfs_trans_ail_tail(mp->m_ail); + tail_lsn = xfs_ail_min_lsn(mp->m_ail); if (!tail_lsn) tail_lsn = atomic64_read(&log->l_last_sync_lsn); @@ -1239,7 +1239,7 @@ xlog_grant_push_ail( * the filesystem is shutting down. */ if (!XLOG_FORCED_SHUTDOWN(log)) - xfs_trans_ail_push(log->l_ailp, threshold_lsn); + xfs_ail_push(log->l_ailp, threshold_lsn); } /* @@ -1449,6 +1449,13 @@ xlog_dealloc_log(xlog_t *log) xlog_cil_destroy(log); + /* + * always need to ensure that the extra buffer does not point to memory + * owned by another log buffer before we free it. + */ + xfs_buf_set_empty(log->l_xbuf, log->l_iclog_size); + xfs_buf_free(log->l_xbuf); + iclog = log->l_iclog; for (i=0; i<log->l_iclog_bufs; i++) { xfs_buf_free(iclog->ic_bp); @@ -1458,7 +1465,6 @@ xlog_dealloc_log(xlog_t *log) } spinlock_destroy(&log->l_icloglock); - xfs_buf_free(log->l_xbuf); log->l_mp->m_log = NULL; kmem_free(log); } /* xlog_dealloc_log */ @@ -3248,13 +3254,6 @@ xfs_log_ticket_get( return ticket; } -xlog_tid_t -xfs_log_get_trans_ident( - struct xfs_trans *tp) -{ - return tp->t_ticket->t_tid; -} - /* * Allocate and initialise a new log ticket. */ @@ -3407,6 +3406,17 @@ xlog_verify_dest_ptr( xfs_emerg(log->l_mp, "%s: invalid ptr", __func__); } +/* + * Check to make sure the grant write head didn't just over lap the tail. If + * the cycles are the same, we can't be overlapping. Otherwise, make sure that + * the cycles differ by exactly one and check the byte count. + * + * This check is run unlocked, so can give false positives. Rather than assert + * on failures, use a warn-once flag and a panic tag to allow the admin to + * determine if they want to panic the machine when such an error occurs. For + * debug kernels this will have the same effect as using an assert but, unlinke + * an assert, it can be turned off at runtime. + */ STATIC void xlog_verify_grant_tail( struct log *log) @@ -3414,17 +3424,22 @@ xlog_verify_grant_tail( int tail_cycle, tail_blocks; int cycle, space; - /* - * Check to make sure the grant write head didn't just over lap the - * tail. If the cycles are the same, we can't be overlapping. - * Otherwise, make sure that the cycles differ by exactly one and - * check the byte count. - */ xlog_crack_grant_head(&log->l_grant_write_head, &cycle, &space); xlog_crack_atomic_lsn(&log->l_tail_lsn, &tail_cycle, &tail_blocks); if (tail_cycle != cycle) { - ASSERT(cycle - 1 == tail_cycle); - ASSERT(space <= BBTOB(tail_blocks)); + if (cycle - 1 != tail_cycle && + !(log->l_flags & XLOG_TAIL_WARN)) { + xfs_alert_tag(log->l_mp, XFS_PTAG_LOGRES, + "%s: cycle - 1 != tail_cycle", __func__); + log->l_flags |= XLOG_TAIL_WARN; + } + + if (space > BBTOB(tail_blocks) && + !(log->l_flags & XLOG_TAIL_WARN)) { + xfs_alert_tag(log->l_mp, XFS_PTAG_LOGRES, + "%s: space > BBTOB(tail_blocks)", __func__); + log->l_flags |= XLOG_TAIL_WARN; + } } } diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h index 3bd3291ef8d2..78c9039994af 100644 --- a/fs/xfs/xfs_log.h +++ b/fs/xfs/xfs_log.h @@ -189,8 +189,6 @@ void xlog_iodone(struct xfs_buf *); struct xlog_ticket *xfs_log_ticket_get(struct xlog_ticket *ticket); void xfs_log_ticket_put(struct xlog_ticket *ticket); -xlog_tid_t xfs_log_get_trans_ident(struct xfs_trans *tp); - void xfs_log_commit_cil(struct xfs_mount *mp, struct xfs_trans *tp, struct xfs_log_vec *log_vector, xfs_lsn_t *commit_lsn, int flags); diff --git a/fs/xfs/xfs_log_cil.c b/fs/xfs/xfs_log_cil.c index 9ca59be08977..c7755d5a5fbe 100644 --- a/fs/xfs/xfs_log_cil.c +++ b/fs/xfs/xfs_log_cil.c @@ -29,6 +29,7 @@ #include "xfs_mount.h" #include "xfs_error.h" #include "xfs_alloc.h" +#include "xfs_discard.h" /* * Perform initial CIL structure initialisation. If the CIL is not @@ -361,19 +362,28 @@ xlog_cil_committed( int abort) { struct xfs_cil_ctx *ctx = args; - struct xfs_busy_extent *busyp, *n; + struct xfs_mount *mp = ctx->cil->xc_log->l_mp; xfs_trans_committed_bulk(ctx->cil->xc_log->l_ailp, ctx->lv_chain, ctx->start_lsn, abort); - list_for_each_entry_safe(busyp, n, &ctx->busy_extents, list) - xfs_alloc_busy_clear(ctx->cil->xc_log->l_mp, busyp); + xfs_alloc_busy_sort(&ctx->busy_extents); + xfs_alloc_busy_clear(mp, &ctx->busy_extents, + (mp->m_flags & XFS_MOUNT_DISCARD) && !abort); spin_lock(&ctx->cil->xc_cil_lock); list_del(&ctx->committing); spin_unlock(&ctx->cil->xc_cil_lock); xlog_cil_free_logvec(ctx->lv_chain); + + if (!list_empty(&ctx->busy_extents)) { + ASSERT(mp->m_flags & XFS_MOUNT_DISCARD); + + xfs_discard_extents(mp, &ctx->busy_extents); + xfs_alloc_busy_clear(mp, &ctx->busy_extents, false); + } + kmem_free(ctx); } diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h index 15dbf1f9c2be..2d3b6a498d63 100644 --- a/fs/xfs/xfs_log_priv.h +++ b/fs/xfs/xfs_log_priv.h @@ -144,6 +144,9 @@ static inline uint xlog_get_client_id(__be32 i) #define XLOG_RECOVERY_NEEDED 0x4 /* log was recovered */ #define XLOG_IO_ERROR 0x8 /* log hit an I/O error, and being shutdown */ +#define XLOG_TAIL_WARN 0x10 /* log tail verify warning issued */ + +typedef __uint32_t xlog_tid_t; #ifdef __KERNEL__ /* @@ -570,7 +573,7 @@ int xlog_write(struct log *log, struct xfs_log_vec *log_vector, * When we crack an atomic LSN, we sample it first so that the value will not * change while we are cracking it into the component values. This means we * will always get consistent component values to work from. This should always - * be used to smaple and crack LSNs taht are stored and updated in atomic + * be used to sample and crack LSNs that are stored and updated in atomic * variables. */ static inline void diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index 0c4a5618e7af..04142caedb2b 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -101,7 +101,7 @@ xlog_get_bp( /* * We do log I/O in units of log sectors (a power-of-2 * multiple of the basic block size), so we round up the - * requested size to acommodate the basic blocks required + * requested size to accommodate the basic blocks required * for complete log sectors. * * In addition, the buffer may be used for a non-sector- @@ -112,7 +112,7 @@ xlog_get_bp( * an issue. Nor will this be a problem if the log I/O is * done in basic blocks (sector size 1). But otherwise we * extend the buffer by one extra log sector to ensure - * there's space to accomodate this possiblility. + * there's space to accommodate this possibility. */ if (nbblks > 1 && log->l_sectBBsize > 1) nbblks += log->l_sectBBsize; @@ -205,6 +205,35 @@ xlog_bread( } /* + * Read at an offset into the buffer. Returns with the buffer in it's original + * state regardless of the result of the read. + */ +STATIC int +xlog_bread_offset( + xlog_t *log, + xfs_daddr_t blk_no, /* block to read from */ + int nbblks, /* blocks to read */ + xfs_buf_t *bp, + xfs_caddr_t offset) +{ + xfs_caddr_t orig_offset = XFS_BUF_PTR(bp); + int orig_len = bp->b_buffer_length; + int error, error2; + + error = XFS_BUF_SET_PTR(bp, offset, BBTOB(nbblks)); + if (error) + return error; + + error = xlog_bread_noalign(log, blk_no, nbblks, bp); + + /* must reset buffer pointer even on error */ + error2 = XFS_BUF_SET_PTR(bp, orig_offset, orig_len); + if (error) + return error; + return error2; +} + +/* * Write out the buffer at the given block for the given number of blocks. * The buffer is kept locked across the write and is returned locked. * This can only be used for synchronous log writes. @@ -1229,20 +1258,12 @@ xlog_write_log_records( */ ealign = round_down(end_block, sectbb); if (j == 0 && (start_block + endcount > ealign)) { - offset = XFS_BUF_PTR(bp); - balign = BBTOB(ealign - start_block); - error = XFS_BUF_SET_PTR(bp, offset + balign, - BBTOB(sectbb)); + offset = XFS_BUF_PTR(bp) + BBTOB(ealign - start_block); + error = xlog_bread_offset(log, ealign, sectbb, + bp, offset); if (error) break; - error = xlog_bread_noalign(log, ealign, sectbb, bp); - if (error) - break; - - error = XFS_BUF_SET_PTR(bp, offset, bufblks); - if (error) - break; } offset = xlog_align(log, start_block, endcount, bp); @@ -3448,19 +3469,9 @@ xlog_do_recovery_pass( * - order is important. */ wrapped_hblks = hblks - split_hblks; - error = XFS_BUF_SET_PTR(hbp, - offset + BBTOB(split_hblks), - BBTOB(hblks - split_hblks)); - if (error) - goto bread_err2; - - error = xlog_bread_noalign(log, 0, - wrapped_hblks, hbp); - if (error) - goto bread_err2; - - error = XFS_BUF_SET_PTR(hbp, offset, - BBTOB(hblks)); + error = xlog_bread_offset(log, 0, + wrapped_hblks, hbp, + offset + BBTOB(split_hblks)); if (error) goto bread_err2; } @@ -3511,19 +3522,9 @@ xlog_do_recovery_pass( * _first_, then the log start (LR header end) * - order is important. */ - error = XFS_BUF_SET_PTR(dbp, - offset + BBTOB(split_bblks), - BBTOB(bblks - split_bblks)); - if (error) - goto bread_err2; - - error = xlog_bread_noalign(log, wrapped_hblks, - bblks - split_bblks, - dbp); - if (error) - goto bread_err2; - - error = XFS_BUF_SET_PTR(dbp, offset, h_size); + error = xlog_bread_offset(log, 0, + bblks - split_bblks, hbp, + offset + BBTOB(split_bblks)); if (error) goto bread_err2; } diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index bb3f9a7b24ed..b49b82363d20 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -1900,7 +1900,7 @@ xfs_mod_incore_sb_batch( uint nmsb, int rsvd) { - xfs_mod_sb_t *msbp = &msb[0]; + xfs_mod_sb_t *msbp; int error = 0; /* @@ -1910,7 +1910,7 @@ xfs_mod_incore_sb_batch( * changes will be atomic. */ spin_lock(&mp->m_sb_lock); - for (msbp = &msbp[0]; msbp < (msb + nmsb); msbp++) { + for (msbp = msb; msbp < (msb + nmsb); msbp++) { ASSERT(msbp->msb_field < XFS_SBS_ICOUNT || msbp->msb_field > XFS_SBS_FDBLOCKS); diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index a62e8971539d..3d68bb267c5f 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -203,12 +203,9 @@ typedef struct xfs_mount { struct mutex m_icsb_mutex; /* balancer sync lock */ #endif struct xfs_mru_cache *m_filestream; /* per-mount filestream data */ - struct task_struct *m_sync_task; /* generalised sync thread */ - xfs_sync_work_t m_sync_work; /* work item for VFS_SYNC */ - struct list_head m_sync_list; /* sync thread work item list */ - spinlock_t m_sync_lock; /* work item list lock */ - int m_sync_seq; /* sync thread generation no. */ - wait_queue_head_t m_wait_single_sync_task; + struct delayed_work m_sync_work; /* background sync work */ + struct delayed_work m_reclaim_work; /* background inode reclaim */ + struct work_struct m_flush_work; /* background inode flush */ __int64_t m_update_flags; /* sb flags we need to update on the next remount,rw */ struct shrinker m_inode_shrink; /* inode reclaim shrinker */ @@ -227,6 +224,7 @@ typedef struct xfs_mount { #define XFS_MOUNT_FS_SHUTDOWN (1ULL << 4) /* atomic stop of all filesystem operations, typically for disk errors in metadata */ +#define XFS_MOUNT_DISCARD (1ULL << 5) /* discard unused blocks */ #define XFS_MOUNT_RETERR (1ULL << 6) /* return alignment errors to user */ #define XFS_MOUNT_NOALIGN (1ULL << 7) /* turn off stripe alignment diff --git a/fs/xfs/xfs_trans.c b/fs/xfs/xfs_trans.c index 76922793f64f..7c7bc2b786bd 100644 --- a/fs/xfs/xfs_trans.c +++ b/fs/xfs/xfs_trans.c @@ -608,10 +608,8 @@ STATIC void xfs_trans_free( struct xfs_trans *tp) { - struct xfs_busy_extent *busyp, *n; - - list_for_each_entry_safe(busyp, n, &tp->t_busy, list) - xfs_alloc_busy_clear(tp->t_mountp, busyp); + xfs_alloc_busy_sort(&tp->t_busy); + xfs_alloc_busy_clear(tp->t_mountp, &tp->t_busy, false); atomic_dec(&tp->t_mountp->m_active_trans); xfs_trans_free_dqinfo(tp); diff --git a/fs/xfs/xfs_trans_ail.c b/fs/xfs/xfs_trans_ail.c index 12aff9584e29..5fc2380092c8 100644 --- a/fs/xfs/xfs_trans_ail.c +++ b/fs/xfs/xfs_trans_ail.c @@ -28,74 +28,138 @@ #include "xfs_trans_priv.h" #include "xfs_error.h" -STATIC void xfs_ail_splice(struct xfs_ail *, struct list_head *, xfs_lsn_t); -STATIC void xfs_ail_delete(struct xfs_ail *, xfs_log_item_t *); -STATIC xfs_log_item_t * xfs_ail_min(struct xfs_ail *); -STATIC xfs_log_item_t * xfs_ail_next(struct xfs_ail *, xfs_log_item_t *); +struct workqueue_struct *xfs_ail_wq; /* AIL workqueue */ #ifdef DEBUG -STATIC void xfs_ail_check(struct xfs_ail *, xfs_log_item_t *); -#else +/* + * Check that the list is sorted as it should be. + */ +STATIC void +xfs_ail_check( + struct xfs_ail *ailp, + xfs_log_item_t *lip) +{ + xfs_log_item_t *prev_lip; + + if (list_empty(&ailp->xa_ail)) + return; + + /* + * Check the next and previous entries are valid. + */ + ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); + prev_lip = list_entry(lip->li_ail.prev, xfs_log_item_t, li_ail); + if (&prev_lip->li_ail != &ailp->xa_ail) + ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); + + prev_lip = list_entry(lip->li_ail.next, xfs_log_item_t, li_ail); + if (&prev_lip->li_ail != &ailp->xa_ail) + ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) >= 0); + + +#ifdef XFS_TRANS_DEBUG + /* + * Walk the list checking lsn ordering, and that every entry has the + * XFS_LI_IN_AIL flag set. This is really expensive, so only do it + * when specifically debugging the transaction subsystem. + */ + prev_lip = list_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); + list_for_each_entry(lip, &ailp->xa_ail, li_ail) { + if (&prev_lip->li_ail != &ailp->xa_ail) + ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); + ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); + prev_lip = lip; + } +#endif /* XFS_TRANS_DEBUG */ +} +#else /* !DEBUG */ #define xfs_ail_check(a,l) #endif /* DEBUG */ +/* + * Return a pointer to the first item in the AIL. If the AIL is empty, then + * return NULL. + */ +static xfs_log_item_t * +xfs_ail_min( + struct xfs_ail *ailp) +{ + if (list_empty(&ailp->xa_ail)) + return NULL; + + return list_first_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); +} + + /* + * Return a pointer to the last item in the AIL. If the AIL is empty, then + * return NULL. + */ +static xfs_log_item_t * +xfs_ail_max( + struct xfs_ail *ailp) +{ + if (list_empty(&ailp->xa_ail)) + return NULL; + + return list_entry(ailp->xa_ail.prev, xfs_log_item_t, li_ail); +} + +/* + * Return a pointer to the item which follows the given item in the AIL. If + * the given item is the last item in the list, then return NULL. + */ +static xfs_log_item_t * +xfs_ail_next( + struct xfs_ail *ailp, + xfs_log_item_t *lip) +{ + if (lip->li_ail.next == &ailp->xa_ail) + return NULL; + + return list_first_entry(&lip->li_ail, xfs_log_item_t, li_ail); +} /* - * This is called by the log manager code to determine the LSN - * of the tail of the log. This is exactly the LSN of the first - * item in the AIL. If the AIL is empty, then this function - * returns 0. + * This is called by the log manager code to determine the LSN of the tail of + * the log. This is exactly the LSN of the first item in the AIL. If the AIL + * is empty, then this function returns 0. * - * We need the AIL lock in order to get a coherent read of the - * lsn of the last item in the AIL. + * We need the AIL lock in order to get a coherent read of the lsn of the last + * item in the AIL. */ xfs_lsn_t -xfs_trans_ail_tail( +xfs_ail_min_lsn( struct xfs_ail *ailp) { - xfs_lsn_t lsn; + xfs_lsn_t lsn = 0; xfs_log_item_t *lip; spin_lock(&ailp->xa_lock); lip = xfs_ail_min(ailp); - if (lip == NULL) { - lsn = (xfs_lsn_t)0; - } else { + if (lip) lsn = lip->li_lsn; - } spin_unlock(&ailp->xa_lock); return lsn; } /* - * xfs_trans_push_ail - * - * This routine is called to move the tail of the AIL forward. It does this by - * trying to flush items in the AIL whose lsns are below the given - * threshold_lsn. - * - * the push is run asynchronously in a separate thread, so we return the tail - * of the log right now instead of the tail after the push. This means we will - * either continue right away, or we will sleep waiting on the async thread to - * do its work. - * - * We do this unlocked - we only need to know whether there is anything in the - * AIL at the time we are called. We don't need to access the contents of - * any of the objects, so the lock is not needed. + * Return the maximum lsn held in the AIL, or zero if the AIL is empty. */ -void -xfs_trans_ail_push( - struct xfs_ail *ailp, - xfs_lsn_t threshold_lsn) +static xfs_lsn_t +xfs_ail_max_lsn( + struct xfs_ail *ailp) { - xfs_log_item_t *lip; + xfs_lsn_t lsn = 0; + xfs_log_item_t *lip; - lip = xfs_ail_min(ailp); - if (lip && !XFS_FORCED_SHUTDOWN(ailp->xa_mount)) { - if (XFS_LSN_CMP(threshold_lsn, ailp->xa_target) > 0) - xfsaild_wakeup(ailp, threshold_lsn); - } + spin_lock(&ailp->xa_lock); + lip = xfs_ail_max(ailp); + if (lip) + lsn = lip->li_lsn; + spin_unlock(&ailp->xa_lock); + + return lsn; } /* @@ -236,35 +300,78 @@ out: } /* - * xfsaild_push does the work of pushing on the AIL. Returning a timeout of - * zero indicates that the caller should sleep until woken. + * splice the log item list into the AIL at the given LSN. */ -long -xfsaild_push( - struct xfs_ail *ailp, - xfs_lsn_t *last_lsn) +static void +xfs_ail_splice( + struct xfs_ail *ailp, + struct list_head *list, + xfs_lsn_t lsn) { - long tout = 0; - xfs_lsn_t last_pushed_lsn = *last_lsn; - xfs_lsn_t target = ailp->xa_target; - xfs_lsn_t lsn; - xfs_log_item_t *lip; - int flush_log, count, stuck; - xfs_mount_t *mp = ailp->xa_mount; + xfs_log_item_t *next_lip; + + /* If the list is empty, just insert the item. */ + if (list_empty(&ailp->xa_ail)) { + list_splice(list, &ailp->xa_ail); + return; + } + + list_for_each_entry_reverse(next_lip, &ailp->xa_ail, li_ail) { + if (XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0) + break; + } + + ASSERT(&next_lip->li_ail == &ailp->xa_ail || + XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0); + + list_splice_init(list, &next_lip->li_ail); +} + +/* + * Delete the given item from the AIL. Return a pointer to the item. + */ +static void +xfs_ail_delete( + struct xfs_ail *ailp, + xfs_log_item_t *lip) +{ + xfs_ail_check(ailp, lip); + list_del(&lip->li_ail); + xfs_trans_ail_cursor_clear(ailp, lip); +} + +/* + * xfs_ail_worker does the work of pushing on the AIL. It will requeue itself + * to run at a later time if there is more work to do to complete the push. + */ +STATIC void +xfs_ail_worker( + struct work_struct *work) +{ + struct xfs_ail *ailp = container_of(to_delayed_work(work), + struct xfs_ail, xa_work); + xfs_mount_t *mp = ailp->xa_mount; struct xfs_ail_cursor *cur = &ailp->xa_cursors; - int push_xfsbufd = 0; + xfs_log_item_t *lip; + xfs_lsn_t lsn; + xfs_lsn_t target; + long tout = 10; + int flush_log = 0; + int stuck = 0; + int count = 0; + int push_xfsbufd = 0; spin_lock(&ailp->xa_lock); + target = ailp->xa_target; xfs_trans_ail_cursor_init(ailp, cur); - lip = xfs_trans_ail_cursor_first(ailp, cur, *last_lsn); + lip = xfs_trans_ail_cursor_first(ailp, cur, ailp->xa_last_pushed_lsn); if (!lip || XFS_FORCED_SHUTDOWN(mp)) { /* * AIL is empty or our push has reached the end. */ xfs_trans_ail_cursor_done(ailp, cur); spin_unlock(&ailp->xa_lock); - *last_lsn = 0; - return tout; + goto out_done; } XFS_STATS_INC(xs_push_ail); @@ -281,8 +388,7 @@ xfsaild_push( * lots of contention on the AIL lists. */ lsn = lip->li_lsn; - flush_log = stuck = count = 0; - while ((XFS_LSN_CMP(lip->li_lsn, target) < 0)) { + while ((XFS_LSN_CMP(lip->li_lsn, target) <= 0)) { int lock_result; /* * If we can lock the item without sleeping, unlock the AIL @@ -301,13 +407,13 @@ xfsaild_push( case XFS_ITEM_SUCCESS: XFS_STATS_INC(xs_push_ail_success); IOP_PUSH(lip); - last_pushed_lsn = lsn; + ailp->xa_last_pushed_lsn = lsn; break; case XFS_ITEM_PUSHBUF: XFS_STATS_INC(xs_push_ail_pushbuf); IOP_PUSHBUF(lip); - last_pushed_lsn = lsn; + ailp->xa_last_pushed_lsn = lsn; push_xfsbufd = 1; break; @@ -319,7 +425,7 @@ xfsaild_push( case XFS_ITEM_LOCKED: XFS_STATS_INC(xs_push_ail_locked); - last_pushed_lsn = lsn; + ailp->xa_last_pushed_lsn = lsn; stuck++; break; @@ -374,9 +480,27 @@ xfsaild_push( wake_up_process(mp->m_ddev_targp->bt_task); } + /* assume we have more work to do in a short while */ +out_done: if (!count) { /* We're past our target or empty, so idle */ - last_pushed_lsn = 0; + ailp->xa_last_pushed_lsn = 0; + + /* + * We clear the XFS_AIL_PUSHING_BIT first before checking + * whether the target has changed. If the target has changed, + * this pushes the requeue race directly onto the result of the + * atomic test/set bit, so we are guaranteed that either the + * the pusher that changed the target or ourselves will requeue + * the work (but not both). + */ + clear_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags); + smp_rmb(); + if (XFS_LSN_CMP(ailp->xa_target, target) == 0 || + test_and_set_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags)) + return; + + tout = 50; } else if (XFS_LSN_CMP(lsn, target) >= 0) { /* * We reached the target so wait a bit longer for I/O to @@ -384,7 +508,7 @@ xfsaild_push( * start the next scan from the start of the AIL. */ tout = 50; - last_pushed_lsn = 0; + ailp->xa_last_pushed_lsn = 0; } else if ((stuck * 100) / count > 90) { /* * Either there is a lot of contention on the AIL or we @@ -396,14 +520,61 @@ xfsaild_push( * continuing from where we were. */ tout = 20; - } else { - /* more to do, but wait a short while before continuing */ - tout = 10; } - *last_lsn = last_pushed_lsn; - return tout; + + /* There is more to do, requeue us. */ + queue_delayed_work(xfs_syncd_wq, &ailp->xa_work, + msecs_to_jiffies(tout)); +} + +/* + * This routine is called to move the tail of the AIL forward. It does this by + * trying to flush items in the AIL whose lsns are below the given + * threshold_lsn. + * + * The push is run asynchronously in a workqueue, which means the caller needs + * to handle waiting on the async flush for space to become available. + * We don't want to interrupt any push that is in progress, hence we only queue + * work if we set the pushing bit approriately. + * + * We do this unlocked - we only need to know whether there is anything in the + * AIL at the time we are called. We don't need to access the contents of + * any of the objects, so the lock is not needed. + */ +void +xfs_ail_push( + struct xfs_ail *ailp, + xfs_lsn_t threshold_lsn) +{ + xfs_log_item_t *lip; + + lip = xfs_ail_min(ailp); + if (!lip || XFS_FORCED_SHUTDOWN(ailp->xa_mount) || + XFS_LSN_CMP(threshold_lsn, ailp->xa_target) <= 0) + return; + + /* + * Ensure that the new target is noticed in push code before it clears + * the XFS_AIL_PUSHING_BIT. + */ + smp_wmb(); + xfs_trans_ail_copy_lsn(ailp, &ailp->xa_target, &threshold_lsn); + if (!test_and_set_bit(XFS_AIL_PUSHING_BIT, &ailp->xa_flags)) + queue_delayed_work(xfs_syncd_wq, &ailp->xa_work, 0); } +/* + * Push out all items in the AIL immediately + */ +void +xfs_ail_push_all( + struct xfs_ail *ailp) +{ + xfs_lsn_t threshold_lsn = xfs_ail_max_lsn(ailp); + + if (threshold_lsn) + xfs_ail_push(ailp, threshold_lsn); +} /* * This is to be called when an item is unlocked that may have @@ -615,7 +786,6 @@ xfs_trans_ail_init( xfs_mount_t *mp) { struct xfs_ail *ailp; - int error; ailp = kmem_zalloc(sizeof(struct xfs_ail), KM_MAYFAIL); if (!ailp) @@ -624,15 +794,9 @@ xfs_trans_ail_init( ailp->xa_mount = mp; INIT_LIST_HEAD(&ailp->xa_ail); spin_lock_init(&ailp->xa_lock); - error = xfsaild_start(ailp); - if (error) - goto out_free_ailp; + INIT_DELAYED_WORK(&ailp->xa_work, xfs_ail_worker); mp->m_ail = ailp; return 0; - -out_free_ailp: - kmem_free(ailp); - return error; } void @@ -641,124 +805,6 @@ xfs_trans_ail_destroy( { struct xfs_ail *ailp = mp->m_ail; - xfsaild_stop(ailp); + cancel_delayed_work_sync(&ailp->xa_work); kmem_free(ailp); } - -/* - * splice the log item list into the AIL at the given LSN. - */ -STATIC void -xfs_ail_splice( - struct xfs_ail *ailp, - struct list_head *list, - xfs_lsn_t lsn) -{ - xfs_log_item_t *next_lip; - - /* - * If the list is empty, just insert the item. - */ - if (list_empty(&ailp->xa_ail)) { - list_splice(list, &ailp->xa_ail); - return; - } - - list_for_each_entry_reverse(next_lip, &ailp->xa_ail, li_ail) { - if (XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0) - break; - } - - ASSERT((&next_lip->li_ail == &ailp->xa_ail) || - (XFS_LSN_CMP(next_lip->li_lsn, lsn) <= 0)); - - list_splice_init(list, &next_lip->li_ail); - return; -} - -/* - * Delete the given item from the AIL. Return a pointer to the item. - */ -STATIC void -xfs_ail_delete( - struct xfs_ail *ailp, - xfs_log_item_t *lip) -{ - xfs_ail_check(ailp, lip); - list_del(&lip->li_ail); - xfs_trans_ail_cursor_clear(ailp, lip); -} - -/* - * Return a pointer to the first item in the AIL. - * If the AIL is empty, then return NULL. - */ -STATIC xfs_log_item_t * -xfs_ail_min( - struct xfs_ail *ailp) -{ - if (list_empty(&ailp->xa_ail)) - return NULL; - - return list_first_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); -} - -/* - * Return a pointer to the item which follows - * the given item in the AIL. If the given item - * is the last item in the list, then return NULL. - */ -STATIC xfs_log_item_t * -xfs_ail_next( - struct xfs_ail *ailp, - xfs_log_item_t *lip) -{ - if (lip->li_ail.next == &ailp->xa_ail) - return NULL; - - return list_first_entry(&lip->li_ail, xfs_log_item_t, li_ail); -} - -#ifdef DEBUG -/* - * Check that the list is sorted as it should be. - */ -STATIC void -xfs_ail_check( - struct xfs_ail *ailp, - xfs_log_item_t *lip) -{ - xfs_log_item_t *prev_lip; - - if (list_empty(&ailp->xa_ail)) - return; - - /* - * Check the next and previous entries are valid. - */ - ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); - prev_lip = list_entry(lip->li_ail.prev, xfs_log_item_t, li_ail); - if (&prev_lip->li_ail != &ailp->xa_ail) - ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); - - prev_lip = list_entry(lip->li_ail.next, xfs_log_item_t, li_ail); - if (&prev_lip->li_ail != &ailp->xa_ail) - ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) >= 0); - - -#ifdef XFS_TRANS_DEBUG - /* - * Walk the list checking lsn ordering, and that every entry has the - * XFS_LI_IN_AIL flag set. This is really expensive, so only do it - * when specifically debugging the transaction subsystem. - */ - prev_lip = list_entry(&ailp->xa_ail, xfs_log_item_t, li_ail); - list_for_each_entry(lip, &ailp->xa_ail, li_ail) { - if (&prev_lip->li_ail != &ailp->xa_ail) - ASSERT(XFS_LSN_CMP(prev_lip->li_lsn, lip->li_lsn) <= 0); - ASSERT((lip->li_flags & XFS_LI_IN_AIL) != 0); - prev_lip = lip; - } -#endif /* XFS_TRANS_DEBUG */ -} -#endif /* DEBUG */ diff --git a/fs/xfs/xfs_trans_inode.c b/fs/xfs/xfs_trans_inode.c index 16084d8ea231..048b0c689d3e 100644 --- a/fs/xfs/xfs_trans_inode.c +++ b/fs/xfs/xfs_trans_inode.c @@ -81,7 +81,7 @@ xfs_trans_ijoin( * * * Grabs a reference to the inode which will be dropped when the transaction - * is commited. The inode will also be unlocked at that point. The inode + * is committed. The inode will also be unlocked at that point. The inode * must be locked, and it cannot be associated with any transaction. */ void diff --git a/fs/xfs/xfs_trans_priv.h b/fs/xfs/xfs_trans_priv.h index 35162c238fa3..6b164e9e9a1f 100644 --- a/fs/xfs/xfs_trans_priv.h +++ b/fs/xfs/xfs_trans_priv.h @@ -65,16 +65,22 @@ struct xfs_ail_cursor { struct xfs_ail { struct xfs_mount *xa_mount; struct list_head xa_ail; - uint xa_gen; - struct task_struct *xa_task; xfs_lsn_t xa_target; struct xfs_ail_cursor xa_cursors; spinlock_t xa_lock; + struct delayed_work xa_work; + xfs_lsn_t xa_last_pushed_lsn; + unsigned long xa_flags; }; +#define XFS_AIL_PUSHING_BIT 0 + /* * From xfs_trans_ail.c */ + +extern struct workqueue_struct *xfs_ail_wq; /* AIL workqueue */ + void xfs_trans_ail_update_bulk(struct xfs_ail *ailp, struct xfs_log_item **log_items, int nr_items, xfs_lsn_t lsn) __releases(ailp->xa_lock); @@ -98,12 +104,13 @@ xfs_trans_ail_delete( xfs_trans_ail_delete_bulk(ailp, &lip, 1); } -void xfs_trans_ail_push(struct xfs_ail *, xfs_lsn_t); +void xfs_ail_push(struct xfs_ail *, xfs_lsn_t); +void xfs_ail_push_all(struct xfs_ail *); +xfs_lsn_t xfs_ail_min_lsn(struct xfs_ail *ailp); + void xfs_trans_unlocked_item(struct xfs_ail *, xfs_log_item_t *); -xfs_lsn_t xfs_trans_ail_tail(struct xfs_ail *ailp); - struct xfs_log_item *xfs_trans_ail_cursor_first(struct xfs_ail *ailp, struct xfs_ail_cursor *cur, xfs_lsn_t lsn); @@ -112,11 +119,6 @@ struct xfs_log_item *xfs_trans_ail_cursor_next(struct xfs_ail *ailp, void xfs_trans_ail_cursor_done(struct xfs_ail *ailp, struct xfs_ail_cursor *cur); -long xfsaild_push(struct xfs_ail *, xfs_lsn_t *); -void xfsaild_wakeup(struct xfs_ail *, xfs_lsn_t); -int xfsaild_start(struct xfs_ail *); -void xfsaild_stop(struct xfs_ail *); - #if BITS_PER_LONG != 64 static inline void xfs_trans_ail_copy_lsn( diff --git a/fs/xfs/xfs_types.h b/fs/xfs/xfs_types.h index 26d1867d8156..65584b55607d 100644 --- a/fs/xfs/xfs_types.h +++ b/fs/xfs/xfs_types.h @@ -73,8 +73,6 @@ typedef __int32_t xfs_tid_t; /* transaction identifier */ typedef __uint32_t xfs_dablk_t; /* dir/attr block number (in file) */ typedef __uint32_t xfs_dahash_t; /* dir/attr hash value */ -typedef __uint32_t xlog_tid_t; /* transaction ID type */ - /* * These types are 64 bits on disk but are either 32 or 64 bits in memory. * Disk based types: diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index c48b4217ec47..b7a5fe7c52c8 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -953,7 +953,7 @@ xfs_release( * If we previously truncated this file and removed old data * in the process, we want to initiate "early" writeout on * the last close. This is an attempt to combat the notorious - * NULL files problem which is particularly noticable from a + * NULL files problem which is particularly noticeable from a * truncate down, buffered (re-)write (delalloc), followed by * a crash. What we are effectively doing here is * significantly reducing the time window where we'd otherwise @@ -982,7 +982,7 @@ xfs_release( * * Further, check if the inode is being opened, written and * closed frequently and we have delayed allocation blocks - * oustanding (e.g. streaming writes from the NFS server), + * outstanding (e.g. streaming writes from the NFS server), * truncating the blocks past EOF will cause fragmentation to * occur. * |