diff options
Diffstat (limited to 'fs')
194 files changed, 9468 insertions, 5056 deletions
diff --git a/fs/Kconfig b/fs/Kconfig index f95ae3a027f3..eaff24a19502 100644 --- a/fs/Kconfig +++ b/fs/Kconfig @@ -28,8 +28,8 @@ config FS_MBCACHE tristate default y if EXT2_FS=y && EXT2_FS_XATTR default y if EXT3_FS=y && EXT3_FS_XATTR - default y if EXT4_FS=y && EXT4_FS_XATTR - default m if EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS_XATTR + default y if EXT4_FS=y + default m if EXT2_FS_XATTR || EXT3_FS_XATTR || EXT4_FS source "fs/reiserfs/Kconfig" source "fs/jfs/Kconfig" diff --git a/fs/autofs4/expire.c b/fs/autofs4/expire.c index 842d00048a65..01443ce43ee7 100644 --- a/fs/autofs4/expire.c +++ b/fs/autofs4/expire.c @@ -548,15 +548,6 @@ int autofs4_do_expire_multi(struct super_block *sb, struct vfsmount *mnt, spin_lock(&sbi->fs_lock); ino->flags &= ~AUTOFS_INF_EXPIRING; - spin_lock(&dentry->d_lock); - if (!ret) { - if ((IS_ROOT(dentry) || - (autofs_type_indirect(sbi->type) && - IS_ROOT(dentry->d_parent))) && - !(dentry->d_flags & DCACHE_NEED_AUTOMOUNT)) - __managed_dentry_set_automount(dentry); - } - spin_unlock(&dentry->d_lock); complete_all(&ino->expire_complete); spin_unlock(&sbi->fs_lock); dput(dentry); diff --git a/fs/autofs4/root.c b/fs/autofs4/root.c index 91b11650722e..c93447604da8 100644 --- a/fs/autofs4/root.c +++ b/fs/autofs4/root.c @@ -124,13 +124,10 @@ static int autofs4_dir_open(struct inode *inode, struct file *file) * it. */ spin_lock(&sbi->lookup_lock); - spin_lock(&dentry->d_lock); - if (!d_mountpoint(dentry) && list_empty(&dentry->d_subdirs)) { - spin_unlock(&dentry->d_lock); + if (!d_mountpoint(dentry) && simple_empty(dentry)) { spin_unlock(&sbi->lookup_lock); return -ENOENT; } - spin_unlock(&dentry->d_lock); spin_unlock(&sbi->lookup_lock); out: @@ -355,7 +352,6 @@ static struct vfsmount *autofs4_d_automount(struct path *path) status = autofs4_mount_wait(dentry); if (status) return ERR_PTR(status); - spin_lock(&sbi->fs_lock); goto done; } @@ -364,8 +360,11 @@ static struct vfsmount *autofs4_d_automount(struct path *path) * having d_mountpoint() true, so there's no need to call back * to the daemon. */ - if (dentry->d_inode && S_ISLNK(dentry->d_inode->i_mode)) + if (dentry->d_inode && S_ISLNK(dentry->d_inode->i_mode)) { + spin_unlock(&sbi->fs_lock); goto done; + } + if (!d_mountpoint(dentry)) { /* * It's possible that user space hasn't removed directories @@ -379,15 +378,13 @@ static struct vfsmount *autofs4_d_automount(struct path *path) * require user space behave. */ if (sbi->version > 4) { - if (have_submounts(dentry)) + if (have_submounts(dentry)) { + spin_unlock(&sbi->fs_lock); goto done; + } } else { - spin_lock(&dentry->d_lock); - if (!list_empty(&dentry->d_subdirs)) { - spin_unlock(&dentry->d_lock); + if (!simple_empty(dentry)) goto done; - } - spin_unlock(&dentry->d_lock); } ino->flags |= AUTOFS_INF_PENDING; spin_unlock(&sbi->fs_lock); @@ -399,28 +396,8 @@ static struct vfsmount *autofs4_d_automount(struct path *path) return ERR_PTR(status); } } -done: - if (!(ino->flags & AUTOFS_INF_EXPIRING)) { - /* - * Any needed mounting has been completed and the path - * updated so clear DCACHE_NEED_AUTOMOUNT so we don't - * call ->d_automount() on rootless multi-mounts since - * it can lead to an incorrect ELOOP error return. - * - * Only clear DMANAGED_AUTOMOUNT for rootless multi-mounts and - * symlinks as in all other cases the dentry will be covered by - * an actual mount so ->d_automount() won't be called during - * the follow. - */ - spin_lock(&dentry->d_lock); - if ((!d_mountpoint(dentry) && - !list_empty(&dentry->d_subdirs)) || - (dentry->d_inode && S_ISLNK(dentry->d_inode->i_mode))) - __managed_dentry_clear_automount(dentry); - spin_unlock(&dentry->d_lock); - } spin_unlock(&sbi->fs_lock); - +done: /* Mount succeeded, check if we ended up with a new dentry */ dentry = autofs4_mountpoint_changed(path); if (!dentry) @@ -432,6 +409,8 @@ done: int autofs4_d_manage(struct dentry *dentry, bool rcu_walk) { struct autofs_sb_info *sbi = autofs4_sbi(dentry->d_sb); + struct autofs_info *ino = autofs4_dentry_ino(dentry); + int status; DPRINTK("dentry=%p %.*s", dentry, dentry->d_name.len, dentry->d_name.name); @@ -456,7 +435,32 @@ int autofs4_d_manage(struct dentry *dentry, bool rcu_walk) * This dentry may be under construction so wait on mount * completion. */ - return autofs4_mount_wait(dentry); + status = autofs4_mount_wait(dentry); + if (status) + return status; + + spin_lock(&sbi->fs_lock); + /* + * If the dentry has been selected for expire while we slept + * on the lock then it might go away. We'll deal with that in + * ->d_automount() and wait on a new mount if the expire + * succeeds or return here if it doesn't (since there's no + * mount to follow with a rootless multi-mount). + */ + if (!(ino->flags & AUTOFS_INF_EXPIRING)) { + /* + * Any needed mounting has been completed and the path + * updated so check if this is a rootless multi-mount so + * we can avoid needless calls ->d_automount() and avoid + * an incorrect ELOOP error return. + */ + if ((!d_mountpoint(dentry) && !simple_empty(dentry)) || + (dentry->d_inode && S_ISLNK(dentry->d_inode->i_mode))) + status = -EISDIR; + } + spin_unlock(&sbi->fs_lock); + + return status; } /* Lookups in the root directory */ @@ -599,9 +603,7 @@ static int autofs4_dir_unlink(struct inode *dir, struct dentry *dentry) spin_lock(&sbi->lookup_lock); __autofs4_add_expiring(dentry); - spin_lock(&dentry->d_lock); - __d_drop(dentry); - spin_unlock(&dentry->d_lock); + d_drop(dentry); spin_unlock(&sbi->lookup_lock); return 0; @@ -672,15 +674,12 @@ static int autofs4_dir_rmdir(struct inode *dir, struct dentry *dentry) return -EACCES; spin_lock(&sbi->lookup_lock); - spin_lock(&dentry->d_lock); - if (!list_empty(&dentry->d_subdirs)) { - spin_unlock(&dentry->d_lock); + if (!simple_empty(dentry)) { spin_unlock(&sbi->lookup_lock); return -ENOTEMPTY; } __autofs4_add_expiring(dentry); - __d_drop(dentry); - spin_unlock(&dentry->d_lock); + d_drop(dentry); spin_unlock(&sbi->lookup_lock); if (sbi->version < 5) diff --git a/fs/binfmt_aout.c b/fs/binfmt_aout.c index 0e7a6f81ae36..6043567b95c2 100644 --- a/fs/binfmt_aout.c +++ b/fs/binfmt_aout.c @@ -30,7 +30,7 @@ #include <asm/cacheflush.h> #include <asm/a.out-core.h> -static int load_aout_binary(struct linux_binprm *, struct pt_regs * regs); +static int load_aout_binary(struct linux_binprm *); static int load_aout_library(struct file*); #ifdef CONFIG_COREDUMP @@ -201,8 +201,9 @@ static unsigned long __user *create_aout_tables(char __user *p, struct linux_bin * libraries. There is no binary dependent code anywhere else. */ -static int load_aout_binary(struct linux_binprm * bprm, struct pt_regs * regs) +static int load_aout_binary(struct linux_binprm * bprm) { + struct pt_regs *regs = current_pt_regs(); struct exec ex; unsigned long error; unsigned long fd_offset; diff --git a/fs/binfmt_elf.c b/fs/binfmt_elf.c index fbd9f60bd763..6d7d1647a68c 100644 --- a/fs/binfmt_elf.c +++ b/fs/binfmt_elf.c @@ -44,7 +44,7 @@ #define user_siginfo_t siginfo_t #endif -static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs); +static int load_elf_binary(struct linux_binprm *bprm); static int load_elf_library(struct file *); static unsigned long elf_map(struct file *, unsigned long, struct elf_phdr *, int, int, unsigned long); @@ -558,7 +558,7 @@ static unsigned long randomize_stack_top(unsigned long stack_top) #endif } -static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) +static int load_elf_binary(struct linux_binprm *bprm) { struct file *interpreter = NULL; /* to shut gcc up */ unsigned long load_addr = 0, load_bias = 0; @@ -575,6 +575,7 @@ static int load_elf_binary(struct linux_binprm *bprm, struct pt_regs *regs) unsigned long reloc_func_desc __maybe_unused = 0; int executable_stack = EXSTACK_DEFAULT; unsigned long def_flags = 0; + struct pt_regs *regs = current_pt_regs(); struct { struct elfhdr elf_ex; struct elfhdr interp_elf_ex; diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index a46049154107..dc84732e554f 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -56,7 +56,7 @@ typedef char *elf_caddr_t; MODULE_LICENSE("GPL"); -static int load_elf_fdpic_binary(struct linux_binprm *, struct pt_regs *); +static int load_elf_fdpic_binary(struct linux_binprm *); static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *, struct file *); static int elf_fdpic_map_file(struct elf_fdpic_params *, struct file *, struct mm_struct *, const char *); @@ -164,10 +164,10 @@ static int elf_fdpic_fetch_phdrs(struct elf_fdpic_params *params, /* * load an fdpic binary into various bits of memory */ -static int load_elf_fdpic_binary(struct linux_binprm *bprm, - struct pt_regs *regs) +static int load_elf_fdpic_binary(struct linux_binprm *bprm) { struct elf_fdpic_params exec_params, interp_params; + struct pt_regs *regs = current_pt_regs(); struct elf_phdr *phdr; unsigned long stack_size, entryaddr; #ifdef ELF_FDPIC_PLAT_INIT diff --git a/fs/binfmt_em86.c b/fs/binfmt_em86.c index 2790c7e1912e..4e6cce57d113 100644 --- a/fs/binfmt_em86.c +++ b/fs/binfmt_em86.c @@ -22,7 +22,7 @@ #define EM86_INTERP "/usr/bin/em86" #define EM86_I_NAME "em86" -static int load_em86(struct linux_binprm *bprm,struct pt_regs *regs) +static int load_em86(struct linux_binprm *bprm) { char *interp, *i_name, *i_arg; struct file * file; @@ -90,7 +90,7 @@ static int load_em86(struct linux_binprm *bprm,struct pt_regs *regs) if (retval < 0) return retval; - return search_binary_handler(bprm, regs); + return search_binary_handler(bprm); } static struct linux_binfmt em86_format = { diff --git a/fs/binfmt_flat.c b/fs/binfmt_flat.c index e280352b28f9..b56371981d16 100644 --- a/fs/binfmt_flat.c +++ b/fs/binfmt_flat.c @@ -88,7 +88,7 @@ struct lib_info { static int load_flat_shared_library(int id, struct lib_info *p); #endif -static int load_flat_binary(struct linux_binprm *, struct pt_regs * regs); +static int load_flat_binary(struct linux_binprm *); static int flat_core_dump(struct coredump_params *cprm); static struct linux_binfmt flat_format = { @@ -858,9 +858,10 @@ out: * libraries. There is no binary dependent code anywhere else. */ -static int load_flat_binary(struct linux_binprm * bprm, struct pt_regs * regs) +static int load_flat_binary(struct linux_binprm * bprm) { struct lib_info libinfo; + struct pt_regs *regs = current_pt_regs(); unsigned long p = bprm->p; unsigned long stack_len; unsigned long start_addr; diff --git a/fs/binfmt_misc.c b/fs/binfmt_misc.c index 790b3cddca67..b0b70fbea06c 100644 --- a/fs/binfmt_misc.c +++ b/fs/binfmt_misc.c @@ -104,7 +104,7 @@ static Node *check_file(struct linux_binprm *bprm) /* * the loader itself */ -static int load_misc_binary(struct linux_binprm *bprm, struct pt_regs *regs) +static int load_misc_binary(struct linux_binprm *bprm) { Node *fmt; struct file * interp_file = NULL; @@ -199,7 +199,7 @@ static int load_misc_binary(struct linux_binprm *bprm, struct pt_regs *regs) bprm->recursion_depth++; - retval = search_binary_handler (bprm, regs); + retval = search_binary_handler(bprm); if (retval < 0) goto _error; diff --git a/fs/binfmt_script.c b/fs/binfmt_script.c index d3b8c1f63155..8c954997e7f7 100644 --- a/fs/binfmt_script.c +++ b/fs/binfmt_script.c @@ -14,7 +14,7 @@ #include <linux/err.h> #include <linux/fs.h> -static int load_script(struct linux_binprm *bprm,struct pt_regs *regs) +static int load_script(struct linux_binprm *bprm) { const char *i_arg, *i_name; char *cp; @@ -95,7 +95,7 @@ static int load_script(struct linux_binprm *bprm,struct pt_regs *regs) retval = prepare_binprm(bprm); if (retval < 0) return retval; - return search_binary_handler(bprm,regs); + return search_binary_handler(bprm); } static struct linux_binfmt script_format = { diff --git a/fs/binfmt_som.c b/fs/binfmt_som.c index 4517aaff61b4..4e00ed68d4a6 100644 --- a/fs/binfmt_som.c +++ b/fs/binfmt_som.c @@ -35,7 +35,7 @@ #include <linux/elf.h> -static int load_som_binary(struct linux_binprm * bprm, struct pt_regs * regs); +static int load_som_binary(struct linux_binprm * bprm); static int load_som_library(struct file *); /* @@ -180,13 +180,14 @@ out: */ static int -load_som_binary(struct linux_binprm * bprm, struct pt_regs * regs) +load_som_binary(struct linux_binprm * bprm) { int retval; unsigned int size; unsigned long som_entry; struct som_hdr *som_ex; struct som_exec_auxhdr *hpuxhdr; + struct pt_regs *regs = current_pt_regs(); /* Get the exec-header */ som_ex = (struct som_hdr *) bprm->buf; diff --git a/fs/block_dev.c b/fs/block_dev.c index 1a1e5e3b1eaf..ab3a456f6650 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -70,19 +70,6 @@ static void bdev_inode_switch_bdi(struct inode *inode, spin_unlock(&dst->wb.list_lock); } -sector_t blkdev_max_block(struct block_device *bdev) -{ - sector_t retval = ~((sector_t)0); - loff_t sz = i_size_read(bdev->bd_inode); - - if (sz) { - unsigned int size = block_size(bdev); - unsigned int sizebits = blksize_bits(size); - retval = (sz >> sizebits); - } - return retval; -} - /* Kill _all_ buffers and pagecache , dirty or not.. */ void kill_bdev(struct block_device *bdev) { @@ -116,8 +103,6 @@ EXPORT_SYMBOL(invalidate_bdev); int set_blocksize(struct block_device *bdev, int size) { - struct address_space *mapping; - /* Size must be a power of two, and between 512 and PAGE_SIZE */ if (size > PAGE_SIZE || size < 512 || !is_power_of_2(size)) return -EINVAL; @@ -126,19 +111,6 @@ int set_blocksize(struct block_device *bdev, int size) if (size < bdev_logical_block_size(bdev)) return -EINVAL; - /* Prevent starting I/O or mapping the device */ - percpu_down_write(&bdev->bd_block_size_semaphore); - - /* Check that the block device is not memory mapped */ - mapping = bdev->bd_inode->i_mapping; - mutex_lock(&mapping->i_mmap_mutex); - if (mapping_mapped(mapping)) { - mutex_unlock(&mapping->i_mmap_mutex); - percpu_up_write(&bdev->bd_block_size_semaphore); - return -EBUSY; - } - mutex_unlock(&mapping->i_mmap_mutex); - /* Don't change the size if it is same as current */ if (bdev->bd_block_size != size) { sync_blockdev(bdev); @@ -146,9 +118,6 @@ int set_blocksize(struct block_device *bdev, int size) bdev->bd_inode->i_blkbits = blksize_bits(size); kill_bdev(bdev); } - - percpu_up_write(&bdev->bd_block_size_semaphore); - return 0; } @@ -181,52 +150,12 @@ static int blkdev_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh, int create) { - if (iblock >= blkdev_max_block(I_BDEV(inode))) { - if (create) - return -EIO; - - /* - * for reads, we're just trying to fill a partial page. - * return a hole, they will have to call get_block again - * before they can fill it, and they will get -EIO at that - * time - */ - return 0; - } bh->b_bdev = I_BDEV(inode); bh->b_blocknr = iblock; set_buffer_mapped(bh); return 0; } -static int -blkdev_get_blocks(struct inode *inode, sector_t iblock, - struct buffer_head *bh, int create) -{ - sector_t end_block = blkdev_max_block(I_BDEV(inode)); - unsigned long max_blocks = bh->b_size >> inode->i_blkbits; - - if ((iblock + max_blocks) > end_block) { - max_blocks = end_block - iblock; - if ((long)max_blocks <= 0) { - if (create) - return -EIO; /* write fully beyond EOF */ - /* - * It is a read which is fully beyond EOF. We return - * a !buffer_mapped buffer - */ - max_blocks = 0; - } - } - - bh->b_bdev = I_BDEV(inode); - bh->b_blocknr = iblock; - bh->b_size = max_blocks << inode->i_blkbits; - if (max_blocks) - set_buffer_mapped(bh); - return 0; -} - static ssize_t blkdev_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, loff_t offset, unsigned long nr_segs) @@ -235,7 +164,7 @@ blkdev_direct_IO(int rw, struct kiocb *iocb, const struct iovec *iov, struct inode *inode = file->f_mapping->host; return __blockdev_direct_IO(rw, iocb, inode, I_BDEV(inode), iov, offset, - nr_segs, blkdev_get_blocks, NULL, NULL, 0); + nr_segs, blkdev_get_block, NULL, NULL, 0); } int __sync_blockdev(struct block_device *bdev, int wait) @@ -459,12 +388,6 @@ static struct inode *bdev_alloc_inode(struct super_block *sb) struct bdev_inode *ei = kmem_cache_alloc(bdev_cachep, GFP_KERNEL); if (!ei) return NULL; - - if (unlikely(percpu_init_rwsem(&ei->bdev.bd_block_size_semaphore))) { - kmem_cache_free(bdev_cachep, ei); - return NULL; - } - return &ei->vfs_inode; } @@ -473,8 +396,6 @@ static void bdev_i_callback(struct rcu_head *head) struct inode *inode = container_of(head, struct inode, i_rcu); struct bdev_inode *bdi = BDEV_I(inode); - percpu_free_rwsem(&bdi->bdev.bd_block_size_semaphore); - kmem_cache_free(bdev_cachep, bdi); } @@ -1593,22 +1514,6 @@ static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg) return blkdev_ioctl(bdev, mode, cmd, arg); } -ssize_t blkdev_aio_read(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) -{ - ssize_t ret; - struct block_device *bdev = I_BDEV(iocb->ki_filp->f_mapping->host); - - percpu_down_read(&bdev->bd_block_size_semaphore); - - ret = generic_file_aio_read(iocb, iov, nr_segs, pos); - - percpu_up_read(&bdev->bd_block_size_semaphore); - - return ret; -} -EXPORT_SYMBOL_GPL(blkdev_aio_read); - /* * Write data to the block device. Only intended for the block device itself * and the raw driver which basically is a fake block device. @@ -1620,16 +1525,12 @@ ssize_t blkdev_aio_write(struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, loff_t pos) { struct file *file = iocb->ki_filp; - struct block_device *bdev = I_BDEV(file->f_mapping->host); struct blk_plug plug; ssize_t ret; BUG_ON(iocb->ki_pos != pos); blk_start_plug(&plug); - - percpu_down_read(&bdev->bd_block_size_semaphore); - ret = __generic_file_aio_write(iocb, iov, nr_segs, &iocb->ki_pos); if (ret > 0 || ret == -EIOCBQUEUED) { ssize_t err; @@ -1638,62 +1539,27 @@ ssize_t blkdev_aio_write(struct kiocb *iocb, const struct iovec *iov, if (err < 0 && ret > 0) ret = err; } - - percpu_up_read(&bdev->bd_block_size_semaphore); - blk_finish_plug(&plug); - return ret; } EXPORT_SYMBOL_GPL(blkdev_aio_write); -static int blkdev_mmap(struct file *file, struct vm_area_struct *vma) -{ - int ret; - struct block_device *bdev = I_BDEV(file->f_mapping->host); - - percpu_down_read(&bdev->bd_block_size_semaphore); - - ret = generic_file_mmap(file, vma); - - percpu_up_read(&bdev->bd_block_size_semaphore); - - return ret; -} - -static ssize_t blkdev_splice_read(struct file *file, loff_t *ppos, - struct pipe_inode_info *pipe, size_t len, - unsigned int flags) -{ - ssize_t ret; - struct block_device *bdev = I_BDEV(file->f_mapping->host); - - percpu_down_read(&bdev->bd_block_size_semaphore); - - ret = generic_file_splice_read(file, ppos, pipe, len, flags); - - percpu_up_read(&bdev->bd_block_size_semaphore); - - return ret; -} - -static ssize_t blkdev_splice_write(struct pipe_inode_info *pipe, - struct file *file, loff_t *ppos, size_t len, - unsigned int flags) +static ssize_t blkdev_aio_read(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos) { - ssize_t ret; - struct block_device *bdev = I_BDEV(file->f_mapping->host); - - percpu_down_read(&bdev->bd_block_size_semaphore); - - ret = generic_file_splice_write(pipe, file, ppos, len, flags); + struct file *file = iocb->ki_filp; + struct inode *bd_inode = file->f_mapping->host; + loff_t size = i_size_read(bd_inode); - percpu_up_read(&bdev->bd_block_size_semaphore); + if (pos >= size) + return 0; - return ret; + size -= pos; + if (size < INT_MAX) + nr_segs = iov_shorten((struct iovec *)iov, nr_segs, size); + return generic_file_aio_read(iocb, iov, nr_segs, pos); } - /* * Try to release a page associated with block device when the system * is under memory pressure. @@ -1724,16 +1590,16 @@ const struct file_operations def_blk_fops = { .llseek = block_llseek, .read = do_sync_read, .write = do_sync_write, - .aio_read = blkdev_aio_read, + .aio_read = blkdev_aio_read, .aio_write = blkdev_aio_write, - .mmap = blkdev_mmap, + .mmap = generic_file_mmap, .fsync = blkdev_fsync, .unlocked_ioctl = block_ioctl, #ifdef CONFIG_COMPAT .compat_ioctl = compat_blkdev_ioctl, #endif - .splice_read = blkdev_splice_read, - .splice_write = blkdev_splice_write, + .splice_read = generic_file_splice_read, + .splice_write = generic_file_splice_write, }; int ioctl_by_bdev(struct block_device *bdev, unsigned cmd, unsigned long arg) diff --git a/fs/btrfs/ctree.h b/fs/btrfs/ctree.h index c72ead869507..596617ecd329 100644 --- a/fs/btrfs/ctree.h +++ b/fs/btrfs/ctree.h @@ -413,7 +413,7 @@ struct btrfs_root_backup { __le64 bytes_used; __le64 num_devices; /* future */ - __le64 unsed_64[4]; + __le64 unused_64[4]; u8 tree_root_level; u8 chunk_root_level; diff --git a/fs/btrfs/disk-io.c b/fs/btrfs/disk-io.c index 7cda51995c1e..22a0439e5a86 100644 --- a/fs/btrfs/disk-io.c +++ b/fs/btrfs/disk-io.c @@ -3416,8 +3416,8 @@ void btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr) 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); + balance_dirty_pages_ratelimited( + root->fs_info->btree_inode->i_mapping); } return; } @@ -3437,8 +3437,8 @@ void __btrfs_btree_balance_dirty(struct btrfs_root *root, unsigned long nr) 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); + balance_dirty_pages_ratelimited( + root->fs_info->btree_inode->i_mapping); } return; } diff --git a/fs/btrfs/extent-tree.c b/fs/btrfs/extent-tree.c index 3d3e2c17d8d1..06b2635073f3 100644 --- a/fs/btrfs/extent-tree.c +++ b/fs/btrfs/extent-tree.c @@ -3888,7 +3888,7 @@ static int flush_space(struct btrfs_root *root, * @root - the root we're allocating for * @block_rsv - the block_rsv we're allocating for * @orig_bytes - the number of bytes we want - * @flush - wether or not we can flush to make our reservation + * @flush - whether or not we can flush to make our reservation * * This will reserve orgi_bytes number of bytes from the space info associated * with the block_rsv. If there is not enough space it will make an attempt to diff --git a/fs/btrfs/extent_map.c b/fs/btrfs/extent_map.c index b8cbc8d5c7f7..ce9f79216723 100644 --- a/fs/btrfs/extent_map.c +++ b/fs/btrfs/extent_map.c @@ -234,12 +234,11 @@ static void try_merge_map(struct extent_map_tree *tree, struct extent_map *em) } /** - * unpint_extent_cache - unpin an extent from the cache + * unpin_extent_cache - unpin an extent from the cache * @tree: tree to unpin the extent in * @start: logical offset in the file * @len: length of the extent * @gen: generation that this extent has been modified in - * @prealloc: if this is set we need to clear the prealloc flag * * Called after an extent has been written to disk properly. Set the generation * to the generation that actually added the file item to the inode so we know diff --git a/fs/btrfs/file.c b/fs/btrfs/file.c index 9ab1bed88116..a8ee75cb96ee 100644 --- a/fs/btrfs/file.c +++ b/fs/btrfs/file.c @@ -1346,8 +1346,7 @@ static noinline ssize_t __btrfs_buffered_write(struct file *file, cond_resched(); - balance_dirty_pages_ratelimited_nr(inode->i_mapping, - dirty_pages); + balance_dirty_pages_ratelimited(inode->i_mapping); if (dirty_pages < (root->leafsize >> PAGE_CACHE_SHIFT) + 1) btrfs_btree_balance_dirty(root, 1); diff --git a/fs/btrfs/ioctl.c b/fs/btrfs/ioctl.c index 8fcf9a59c28d..5b3429ab8ec1 100644 --- a/fs/btrfs/ioctl.c +++ b/fs/btrfs/ioctl.c @@ -1225,7 +1225,7 @@ int btrfs_defrag_file(struct inode *inode, struct file *file, } defrag_count += ret; - balance_dirty_pages_ratelimited_nr(inode->i_mapping, ret); + balance_dirty_pages_ratelimited(inode->i_mapping); mutex_unlock(&inode->i_mutex); if (newer_than) { diff --git a/fs/btrfs/ordered-data.h b/fs/btrfs/ordered-data.h index dd27a0b46a37..853fc7beedfa 100644 --- a/fs/btrfs/ordered-data.h +++ b/fs/btrfs/ordered-data.h @@ -76,7 +76,7 @@ struct btrfs_ordered_sum { #define BTRFS_ORDERED_IOERR 6 /* We had an io error when writing this out */ -#define BTRFS_ORDERED_UPDATED_ISIZE 7 /* indicates wether this ordered extent +#define BTRFS_ORDERED_UPDATED_ISIZE 7 /* indicates whether this ordered extent * has done its due diligence in updating * the isize. */ diff --git a/fs/btrfs/volumes.c b/fs/btrfs/volumes.c index 0f5ebb72a5ea..e3c6ee3cc2ba 100644 --- a/fs/btrfs/volumes.c +++ b/fs/btrfs/volumes.c @@ -4294,7 +4294,7 @@ int btrfs_map_bio(struct btrfs_root *root, int rw, struct bio *bio, rcu_read_lock(); name = rcu_dereference(dev->name); - pr_debug("btrfs_map_bio: rw %d, secor=%llu, dev=%lu " + pr_debug("btrfs_map_bio: rw %d, sector=%llu, dev=%lu " "(%s id %llu), size=%u\n", rw, (u64)bio->bi_sector, (u_long)dev->bdev->bd_dev, name->str, dev->devid, bio->bi_size); diff --git a/fs/buffer.c b/fs/buffer.c index b5f044283edb..c017a2dfb909 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -46,8 +46,7 @@ static int fsync_buffers_list(spinlock_t *lock, struct list_head *list); #define BH_ENTRY(list) list_entry((list), struct buffer_head, b_assoc_buffers) -inline void -init_buffer(struct buffer_head *bh, bh_end_io_t *handler, void *private) +void init_buffer(struct buffer_head *bh, bh_end_io_t *handler, void *private) { bh->b_end_io = handler; bh->b_private = private; @@ -555,7 +554,7 @@ void emergency_thaw_all(void) */ int sync_mapping_buffers(struct address_space *mapping) { - struct address_space *buffer_mapping = mapping->assoc_mapping; + struct address_space *buffer_mapping = mapping->private_data; if (buffer_mapping == NULL || list_empty(&mapping->private_list)) return 0; @@ -588,10 +587,10 @@ void mark_buffer_dirty_inode(struct buffer_head *bh, struct inode *inode) struct address_space *buffer_mapping = bh->b_page->mapping; mark_buffer_dirty(bh); - if (!mapping->assoc_mapping) { - mapping->assoc_mapping = buffer_mapping; + if (!mapping->private_data) { + mapping->private_data = buffer_mapping; } else { - BUG_ON(mapping->assoc_mapping != buffer_mapping); + BUG_ON(mapping->private_data != buffer_mapping); } if (!bh->b_assoc_map) { spin_lock(&buffer_mapping->private_lock); @@ -788,7 +787,7 @@ void invalidate_inode_buffers(struct inode *inode) if (inode_has_buffers(inode)) { struct address_space *mapping = &inode->i_data; struct list_head *list = &mapping->private_list; - struct address_space *buffer_mapping = mapping->assoc_mapping; + struct address_space *buffer_mapping = mapping->private_data; spin_lock(&buffer_mapping->private_lock); while (!list_empty(list)) @@ -811,7 +810,7 @@ int remove_inode_buffers(struct inode *inode) if (inode_has_buffers(inode)) { struct address_space *mapping = &inode->i_data; struct list_head *list = &mapping->private_list; - struct address_space *buffer_mapping = mapping->assoc_mapping; + struct address_space *buffer_mapping = mapping->private_data; spin_lock(&buffer_mapping->private_lock); while (!list_empty(list)) { @@ -850,13 +849,10 @@ try_again: if (!bh) goto no_grow; - bh->b_bdev = NULL; bh->b_this_page = head; bh->b_blocknr = -1; head = bh; - bh->b_state = 0; - atomic_set(&bh->b_count, 0); bh->b_size = size; /* Link the buffer to its page */ @@ -911,6 +907,18 @@ link_dev_buffers(struct page *page, struct buffer_head *head) attach_page_buffers(page, head); } +static sector_t blkdev_max_block(struct block_device *bdev, unsigned int size) +{ + sector_t retval = ~((sector_t)0); + loff_t sz = i_size_read(bdev->bd_inode); + + if (sz) { + unsigned int sizebits = blksize_bits(size); + retval = (sz >> sizebits); + } + return retval; +} + /* * Initialise the state of a blockdev page's buffers. */ @@ -921,7 +929,7 @@ init_page_buffers(struct page *page, struct block_device *bdev, struct buffer_head *head = page_buffers(page); struct buffer_head *bh = head; int uptodate = PageUptodate(page); - sector_t end_block = blkdev_max_block(I_BDEV(bdev->bd_inode)); + sector_t end_block = blkdev_max_block(I_BDEV(bdev->bd_inode), size); do { if (!buffer_mapped(bh)) { @@ -1553,6 +1561,28 @@ void unmap_underlying_metadata(struct block_device *bdev, sector_t block) EXPORT_SYMBOL(unmap_underlying_metadata); /* + * Size is a power-of-two in the range 512..PAGE_SIZE, + * and the case we care about most is PAGE_SIZE. + * + * So this *could* possibly be written with those + * constraints in mind (relevant mostly if some + * architecture has a slow bit-scan instruction) + */ +static inline int block_size_bits(unsigned int blocksize) +{ + return ilog2(blocksize); +} + +static struct buffer_head *create_page_buffers(struct page *page, struct inode *inode, unsigned int b_state) +{ + BUG_ON(!PageLocked(page)); + + if (!page_has_buffers(page)) + create_empty_buffers(page, 1 << ACCESS_ONCE(inode->i_blkbits), b_state); + return page_buffers(page); +} + +/* * NOTE! All mapped/uptodate combinations are valid: * * Mapped Uptodate Meaning @@ -1589,19 +1619,13 @@ static int __block_write_full_page(struct inode *inode, struct page *page, sector_t block; sector_t last_block; struct buffer_head *bh, *head; - const unsigned blocksize = 1 << inode->i_blkbits; + unsigned int blocksize, bbits; int nr_underway = 0; int write_op = (wbc->sync_mode == WB_SYNC_ALL ? WRITE_SYNC : WRITE); - BUG_ON(!PageLocked(page)); - - last_block = (i_size_read(inode) - 1) >> inode->i_blkbits; - - if (!page_has_buffers(page)) { - create_empty_buffers(page, blocksize, + head = create_page_buffers(page, inode, (1 << BH_Dirty)|(1 << BH_Uptodate)); - } /* * Be very careful. We have no exclusion from __set_page_dirty_buffers @@ -1613,9 +1637,12 @@ static int __block_write_full_page(struct inode *inode, struct page *page, * handle that here by just cleaning them. */ - block = (sector_t)page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits); - head = page_buffers(page); bh = head; + blocksize = bh->b_size; + bbits = block_size_bits(blocksize); + + block = (sector_t)page->index << (PAGE_CACHE_SHIFT - bbits); + last_block = (i_size_read(inode) - 1) >> bbits; /* * Get all the dirty buffers mapped to disk addresses and @@ -1806,12 +1833,10 @@ int __block_write_begin(struct page *page, loff_t pos, unsigned len, BUG_ON(to > PAGE_CACHE_SIZE); BUG_ON(from > to); - blocksize = 1 << inode->i_blkbits; - if (!page_has_buffers(page)) - create_empty_buffers(page, blocksize, 0); - head = page_buffers(page); + head = create_page_buffers(page, inode, 0); + blocksize = head->b_size; + bbits = block_size_bits(blocksize); - bbits = inode->i_blkbits; block = (sector_t)page->index << (PAGE_CACHE_SHIFT - bbits); for(bh = head, block_start = 0; bh != head || !block_start; @@ -1881,11 +1906,11 @@ static int __block_commit_write(struct inode *inode, struct page *page, unsigned blocksize; struct buffer_head *bh, *head; - blocksize = 1 << inode->i_blkbits; + bh = head = page_buffers(page); + blocksize = bh->b_size; - for(bh = head = page_buffers(page), block_start = 0; - bh != head || !block_start; - block_start=block_end, bh = bh->b_this_page) { + block_start = 0; + do { block_end = block_start + blocksize; if (block_end <= from || block_start >= to) { if (!buffer_uptodate(bh)) @@ -1895,7 +1920,10 @@ static int __block_commit_write(struct inode *inode, struct page *page, mark_buffer_dirty(bh); } clear_buffer_new(bh); - } + + block_start = block_end; + bh = bh->b_this_page; + } while (bh != head); /* * If this is a partial write which happened to make all buffers @@ -2020,7 +2048,6 @@ EXPORT_SYMBOL(generic_write_end); int block_is_partially_uptodate(struct page *page, read_descriptor_t *desc, unsigned long from) { - struct inode *inode = page->mapping->host; unsigned block_start, block_end, blocksize; unsigned to; struct buffer_head *bh, *head; @@ -2029,13 +2056,13 @@ int block_is_partially_uptodate(struct page *page, read_descriptor_t *desc, if (!page_has_buffers(page)) return 0; - blocksize = 1 << inode->i_blkbits; + head = page_buffers(page); + blocksize = head->b_size; to = min_t(unsigned, PAGE_CACHE_SIZE - from, desc->count); to = from + to; if (from < blocksize && to > PAGE_CACHE_SIZE - blocksize) return 0; - head = page_buffers(page); bh = head; block_start = 0; do { @@ -2068,18 +2095,16 @@ int block_read_full_page(struct page *page, get_block_t *get_block) struct inode *inode = page->mapping->host; sector_t iblock, lblock; struct buffer_head *bh, *head, *arr[MAX_BUF_PER_PAGE]; - unsigned int blocksize; + unsigned int blocksize, bbits; int nr, i; int fully_mapped = 1; - BUG_ON(!PageLocked(page)); - blocksize = 1 << inode->i_blkbits; - if (!page_has_buffers(page)) - create_empty_buffers(page, blocksize, 0); - head = page_buffers(page); + head = create_page_buffers(page, inode, 0); + blocksize = head->b_size; + bbits = block_size_bits(blocksize); - iblock = (sector_t)page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits); - lblock = (i_size_read(inode)+blocksize-1) >> inode->i_blkbits; + iblock = (sector_t)page->index << (PAGE_CACHE_SHIFT - bbits); + lblock = (i_size_read(inode)+blocksize-1) >> bbits; bh = head; nr = 0; i = 0; @@ -2864,6 +2889,55 @@ static void end_bio_bh_io_sync(struct bio *bio, int err) bio_put(bio); } +/* + * This allows us to do IO even on the odd last sectors + * of a device, even if the bh block size is some multiple + * of the physical sector size. + * + * We'll just truncate the bio to the size of the device, + * and clear the end of the buffer head manually. + * + * Truly out-of-range accesses will turn into actual IO + * errors, this only handles the "we need to be able to + * do IO at the final sector" case. + */ +static void guard_bh_eod(int rw, struct bio *bio, struct buffer_head *bh) +{ + sector_t maxsector; + unsigned bytes; + + maxsector = i_size_read(bio->bi_bdev->bd_inode) >> 9; + if (!maxsector) + return; + + /* + * If the *whole* IO is past the end of the device, + * let it through, and the IO layer will turn it into + * an EIO. + */ + if (unlikely(bio->bi_sector >= maxsector)) + return; + + maxsector -= bio->bi_sector; + bytes = bio->bi_size; + if (likely((bytes >> 9) <= maxsector)) + return; + + /* Uhhuh. We've got a bh that straddles the device size! */ + bytes = maxsector << 9; + + /* Truncate the bio.. */ + bio->bi_size = bytes; + bio->bi_io_vec[0].bv_len = bytes; + + /* ..and clear the end of the buffer for reads */ + if ((rw & RW_MASK) == READ) { + void *kaddr = kmap_atomic(bh->b_page); + memset(kaddr + bh_offset(bh) + bytes, 0, bh->b_size - bytes); + kunmap_atomic(kaddr); + } +} + int submit_bh(int rw, struct buffer_head * bh) { struct bio *bio; @@ -2900,6 +2974,9 @@ int submit_bh(int rw, struct buffer_head * bh) bio->bi_end_io = end_bio_bh_io_sync; bio->bi_private = bh; + /* Take care of bh's that straddle the end of the device */ + guard_bh_eod(rw, bio, bh); + bio_get(bio); submit_bio(rw, bio); diff --git a/fs/cifs/Kconfig b/fs/cifs/Kconfig index 2075ddfffa73..21ff76c22a17 100644 --- a/fs/cifs/Kconfig +++ b/fs/cifs/Kconfig @@ -122,9 +122,17 @@ config CIFS_ACL Allows fetching CIFS/NTFS ACL from the server. The DACL blob is handed over to the application/caller. +config CIFS_DEBUG + bool "Enable CIFS debugging routines" + default y + depends on CIFS + help + Enabling this option adds helpful debugging messages to + the cifs code which increases the size of the cifs module. + If unsure, say Y. config CIFS_DEBUG2 bool "Enable additional CIFS debugging routines" - depends on CIFS + depends on CIFS_DEBUG help Enabling this option adds a few more debugging routines to the cifs code which slightly increases the size of diff --git a/fs/cifs/README b/fs/cifs/README index 22ab7b5b8da7..2d5622f60e11 100644 --- a/fs/cifs/README +++ b/fs/cifs/README @@ -480,7 +480,7 @@ A partial list of the supported mount options follows: Unicode on the wire. nomapchars Do not translate any of these seven characters (default). nocase Request case insensitive path name matching (case - sensitive is the default if the server suports it). + sensitive is the default if the server supports it). (mount option "ignorecase" is identical to "nocase") posixpaths If CIFS Unix extensions are supported, attempt to negotiate posix path name support which allows certain diff --git a/fs/cifs/cifs_debug.h b/fs/cifs/cifs_debug.h index c0c68bb492d7..86e92ef2abc1 100644 --- a/fs/cifs/cifs_debug.h +++ b/fs/cifs/cifs_debug.h @@ -18,7 +18,6 @@ * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA * */ -#define CIFS_DEBUG /* BB temporary */ #ifndef _H_CIFS_DEBUG #define _H_CIFS_DEBUG @@ -37,49 +36,43 @@ void dump_smb(void *, int); #define CIFS_RC 0x02 #define CIFS_TIMER 0x04 +extern int cifsFYI; +extern int cifsERROR; + /* * debug ON * -------- */ -#ifdef CIFS_DEBUG +#ifdef CONFIG_CIFS_DEBUG /* information message: e.g., configuration, major event */ -extern int cifsFYI; -#define cifsfyi(fmt, arg...) \ +#define cifsfyi(fmt, ...) \ do { \ if (cifsFYI & CIFS_INFO) \ - printk(KERN_DEBUG "%s: " fmt "\n", __FILE__, ##arg); \ + printk(KERN_DEBUG "%s: " fmt "\n", \ + __FILE__, ##__VA_ARGS__); \ } while (0) -#define cFYI(set, fmt, arg...) \ -do { \ - if (set) \ - cifsfyi(fmt, ##arg); \ +#define cFYI(set, fmt, ...) \ +do { \ + if (set) \ + cifsfyi(fmt, ##__VA_ARGS__); \ } while (0) -#define cifswarn(fmt, arg...) \ - printk(KERN_WARNING fmt "\n", ##arg) +#define cifswarn(fmt, ...) \ + printk(KERN_WARNING fmt "\n", ##__VA_ARGS__) -/* debug event message: */ -extern int cifsERROR; - -#define cEVENT(fmt, arg...) \ +/* error event message: e.g., i/o error */ +#define cifserror(fmt, ...) \ do { \ if (cifsERROR) \ - printk(KERN_EVENT "%s: " fmt "\n", __FILE__, ##arg); \ -} while (0) - -/* error event message: e.g., i/o error */ -#define cifserror(fmt, arg...) \ -do { \ - if (cifsERROR) \ - printk(KERN_ERR "CIFS VFS: " fmt "\n", ##arg); \ + printk(KERN_ERR "CIFS VFS: " fmt "\n", ##__VA_ARGS__); \ } while (0) -#define cERROR(set, fmt, arg...) \ -do { \ - if (set) \ - cifserror(fmt, ##arg); \ +#define cERROR(set, fmt, ...) \ +do { \ + if (set) \ + cifserror(fmt, ##__VA_ARGS__); \ } while (0) /* @@ -87,10 +80,27 @@ do { \ * --------- */ #else /* _CIFS_DEBUG */ -#define cERROR(set, fmt, arg...) -#define cEVENT(fmt, arg...) -#define cFYI(set, fmt, arg...) -#define cifserror(fmt, arg...) +#define cifsfyi(fmt, ...) \ +do { \ + if (0) \ + printk(KERN_DEBUG "%s: " fmt "\n", \ + __FILE__, ##__VA_ARGS__); \ +} while (0) +#define cFYI(set, fmt, ...) \ +do { \ + if (0 && set) \ + cifsfyi(fmt, ##__VA_ARGS__); \ +} while (0) +#define cifserror(fmt, ...) \ +do { \ + if (0) \ + printk(KERN_ERR "CIFS VFS: " fmt "\n", ##__VA_ARGS__); \ +} while (0) +#define cERROR(set, fmt, ...) \ +do { \ + if (0 && set) \ + cifserror(fmt, ##__VA_ARGS__); \ +} while (0) #endif /* _CIFS_DEBUG */ #endif /* _H_CIFS_DEBUG */ diff --git a/fs/cifs/cifsacl.c b/fs/cifs/cifsacl.c index 0fb15bbbe43c..5cbd00e74067 100644 --- a/fs/cifs/cifsacl.c +++ b/fs/cifs/cifsacl.c @@ -42,135 +42,27 @@ static const struct cifs_sid sid_authusers = { /* group users */ static const struct cifs_sid sid_user = {1, 2 , {0, 0, 0, 0, 0, 5}, {} }; -const struct cred *root_cred; - -static void -shrink_idmap_tree(struct rb_root *root, int nr_to_scan, int *nr_rem, - int *nr_del) -{ - 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); - } - } -} - -/* - * 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); - - root = &siduidtree; - spin_lock(&uidsidlock); - shrink_idmap_tree(root, nr_to_scan, &nr_rem, &nr_del); - spin_unlock(&uidsidlock); - - root = &sidgidtree; - spin_lock(&gidsidlock); - shrink_idmap_tree(root, nr_to_scan, &nr_rem, &nr_del); - spin_unlock(&gidsidlock); - - return nr_rem; -} - -static void -sid_rb_insert(struct rb_root *root, unsigned long cid, - struct cifs_sid_id **psidid, char *typestr) -{ - 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; - if (cid > lsidid->id) { - linkto = &(node->rb_left); - node = node->rb_left; - } - if (cid < lsidid->id) { - linkto = &(node->rb_right); - node = node->rb_right; - } - } - - (*psidid)->id = cid; - (*psidid)->time = jiffies - (SID_MAP_RETRY + 1); - (*psidid)->refcount = 0; - - sprintf((*psidid)->sidstr, "%s", typestr); - strptr = (*psidid)->sidstr + strlen((*psidid)->sidstr); - sprintf(strptr, "%ld", cid); - - 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); -} - -static struct cifs_sid_id * -sid_rb_search(struct rb_root *root, unsigned long cid) -{ - struct rb_node *node = root->rb_node; - struct cifs_sid_id *lsidid; - - while (node) { - lsidid = rb_entry(node, struct cifs_sid_id, rbnode); - if (cid > lsidid->id) - node = node->rb_left; - else if (cid < lsidid->id) - node = node->rb_right; - else /* node found */ - return lsidid; - } - - return NULL; -} - -static struct shrinker cifs_shrinker = { - .shrink = cifs_idmap_shrinker, - .seeks = DEFAULT_SEEKS, -}; +static const struct cred *root_cred; static int cifs_idmap_key_instantiate(struct key *key, struct key_preparsed_payload *prep) { char *payload; + /* + * If the payload is less than or equal to the size of a pointer, then + * an allocation here is wasteful. Just copy the data directly to the + * payload.value union member instead. + * + * With this however, you must check the datalen before trying to + * dereference payload.data! + */ + if (prep->datalen <= sizeof(key->payload)) { + key->payload.value = 0; + memcpy(&key->payload.value, prep->data, prep->datalen); + key->datalen = prep->datalen; + return 0; + } payload = kmalloc(prep->datalen, GFP_KERNEL); if (!payload) return -ENOMEM; @@ -184,10 +76,11 @@ cifs_idmap_key_instantiate(struct key *key, struct key_preparsed_payload *prep) static inline void cifs_idmap_key_destroy(struct key *key) { - kfree(key->payload.data); + if (key->datalen > sizeof(key->payload)) + kfree(key->payload.data); } -struct key_type cifs_idmap_key_type = { +static struct key_type cifs_idmap_key_type = { .name = "cifs.idmap", .instantiate = cifs_idmap_key_instantiate, .destroy = cifs_idmap_key_destroy, @@ -195,221 +88,174 @@ struct key_type cifs_idmap_key_type = { .match = user_match, }; -static void -sid_to_str(struct cifs_sid *sidptr, char *sidstr) +static char * +sid_to_key_str(struct cifs_sid *sidptr, unsigned int type) { - int i; - unsigned long saval; - char *strptr; + int i, len; + unsigned int saval; + char *sidstr, *strptr; + unsigned long long id_auth_val; + + /* 3 bytes for prefix */ + sidstr = kmalloc(3 + SID_STRING_BASE_SIZE + + (SID_STRING_SUBAUTH_SIZE * sidptr->num_subauth), + GFP_KERNEL); + if (!sidstr) + return sidstr; strptr = sidstr; + len = sprintf(strptr, "%cs:S-%hhu", type == SIDOWNER ? 'o' : 'g', + sidptr->revision); + strptr += len; + + /* The authority field is a single 48-bit number */ + id_auth_val = (unsigned long long)sidptr->authority[5]; + id_auth_val |= (unsigned long long)sidptr->authority[4] << 8; + id_auth_val |= (unsigned long long)sidptr->authority[3] << 16; + id_auth_val |= (unsigned long long)sidptr->authority[2] << 24; + id_auth_val |= (unsigned long long)sidptr->authority[1] << 32; + id_auth_val |= (unsigned long long)sidptr->authority[0] << 48; - sprintf(strptr, "%s", "S"); - strptr = sidstr + strlen(sidstr); - - sprintf(strptr, "-%d", sidptr->revision); - strptr = sidstr + strlen(sidstr); + /* + * MS-DTYP states that if the authority is >= 2^32, then it should be + * expressed as a hex value. + */ + if (id_auth_val <= UINT_MAX) + len = sprintf(strptr, "-%llu", id_auth_val); + else + len = sprintf(strptr, "-0x%llx", id_auth_val); - for (i = 0; i < 6; ++i) { - if (sidptr->authority[i]) { - sprintf(strptr, "-%d", sidptr->authority[i]); - strptr = sidstr + strlen(sidstr); - } - } + strptr += len; for (i = 0; i < sidptr->num_subauth; ++i) { saval = le32_to_cpu(sidptr->sub_auth[i]); - sprintf(strptr, "-%ld", saval); - strptr = sidstr + strlen(sidstr); + len = sprintf(strptr, "-%u", saval); + strptr += len; } -} -static void -cifs_copy_sid(struct cifs_sid *dst, const struct cifs_sid *src) -{ - memcpy(dst, src, sizeof(*dst)); - dst->num_subauth = min_t(u8, src->num_subauth, NUM_SUBAUTHS); + return sidstr; } -static void -id_rb_insert(struct rb_root *root, struct cifs_sid *sidptr, - struct cifs_sid_id **psidid, char *typestr) +/* + * if the two SIDs (roughly equivalent to a UUID for a user or group) are + * the same returns zero, if they do not match returns non-zero. + */ +static int +compare_sids(const struct cifs_sid *ctsid, const struct cifs_sid *cwsid) { - 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; - } - } - - cifs_copy_sid(&(*psidid)->sid, sidptr); - (*psidid)->time = jiffies - (SID_MAP_RETRY + 1); - (*psidid)->refcount = 0; + int i; + int num_subauth, num_sat, num_saw; - sprintf((*psidid)->sidstr, "%s", typestr); - strptr = (*psidid)->sidstr + strlen((*psidid)->sidstr); - sid_to_str(&(*psidid)->sid, strptr); + if ((!ctsid) || (!cwsid)) + return 1; - clear_bit(SID_ID_PENDING, &(*psidid)->state); - clear_bit(SID_ID_MAPPED, &(*psidid)->state); + /* compare the revision */ + if (ctsid->revision != cwsid->revision) { + if (ctsid->revision > cwsid->revision) + return 1; + else + return -1; + } - rb_link_node(&(*psidid)->rbnode, parent, linkto); - rb_insert_color(&(*psidid)->rbnode, root); -} + /* compare all of the six auth values */ + for (i = 0; i < NUM_AUTHS; ++i) { + if (ctsid->authority[i] != cwsid->authority[i]) { + if (ctsid->authority[i] > cwsid->authority[i]) + return 1; + else + return -1; + } + } -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; + /* 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 (i = 0; i < num_subauth; ++i) { + 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 NULL; + return 0; /* sids compare/match */ } -static int -sidid_pending_wait(void *unused) +static void +cifs_copy_sid(struct cifs_sid *dst, const struct cifs_sid *src) { - schedule(); - return signal_pending(current) ? -ERESTARTSYS : 0; + int i; + + dst->revision = src->revision; + dst->num_subauth = min_t(u8, src->num_subauth, SID_MAX_SUB_AUTHORITIES); + for (i = 0; i < NUM_AUTHS; ++i) + dst->authority[i] = src->authority[i]; + for (i = 0; i < dst->num_subauth; ++i) + dst->sub_auth[i] = src->sub_auth[i]; } static int -id_to_sid(unsigned long cid, uint sidtype, struct cifs_sid *ssid) +id_to_sid(unsigned int cid, uint sidtype, struct cifs_sid *ssid) { - int rc = 0; + int rc; struct key *sidkey; + struct cifs_sid *ksid; + unsigned int ksid_size; + char desc[3 + 10 + 1]; /* 3 byte prefix + 10 bytes for value + NULL */ const struct cred *saved_cred; - struct cifs_sid *lsid; - struct cifs_sid_id *psidid, *npsidid; - struct rb_root *cidtree; - spinlock_t *cidlock; - - if (sidtype == SIDOWNER) { - cidlock = &siduidlock; - cidtree = &uidtree; - } else if (sidtype == SIDGROUP) { - cidlock = &sidgidlock; - cidtree = &gidtree; - } else - return -EINVAL; - spin_lock(cidlock); - psidid = sid_rb_search(cidtree, cid); - - 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; - } + rc = snprintf(desc, sizeof(desc), "%ci:%u", + sidtype == SIDOWNER ? 'o' : 'g', cid); + if (rc >= sizeof(desc)) + return -EINVAL; - spin_lock(cidlock); - psidid = sid_rb_search(cidtree, cid); - if (psidid) { /* node happened to get inserted meanwhile */ - ++psidid->refcount; - spin_unlock(cidlock); - kfree(npsidid->sidstr); - kfree(npsidid); - } else { - psidid = npsidid; - sid_rb_insert(cidtree, cid, &psidid, - sidtype == SIDOWNER ? "oi:" : "gi:"); - ++psidid->refcount; - spin_unlock(cidlock); - } - } else { - ++psidid->refcount; - spin_unlock(cidlock); + rc = 0; + saved_cred = override_creds(root_cred); + sidkey = request_key(&cifs_idmap_key_type, desc, ""); + if (IS_ERR(sidkey)) { + rc = -EINVAL; + cFYI(1, "%s: Can't map %cid %u to a SID", __func__, + sidtype == SIDOWNER ? 'u' : 'g', cid); + goto out_revert_creds; + } else if (sidkey->datalen < CIFS_SID_BASE_SIZE) { + rc = -EIO; + cFYI(1, "%s: Downcall contained malformed key " + "(datalen=%hu)", __func__, sidkey->datalen); + goto invalidate_key; } /* - * 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 . + * A sid is usually too large to be embedded in payload.value, but if + * there are no subauthorities and the host has 8-byte pointers, then + * it could be. */ - if (test_bit(SID_ID_MAPPED, &psidid->state)) { - cifs_copy_sid(ssid, &psidid->sid); - psidid->time = jiffies; /* update ts for accessing */ - goto id_sid_out; - } - - if (time_after(psidid->time + SID_MAP_RETRY, jiffies)) { - rc = -EINVAL; - goto id_sid_out; + ksid = sidkey->datalen <= sizeof(sidkey->payload) ? + (struct cifs_sid *)&sidkey->payload.value : + (struct cifs_sid *)sidkey->payload.data; + + ksid_size = CIFS_SID_BASE_SIZE + (ksid->num_subauth * sizeof(__le32)); + if (ksid_size > sidkey->datalen) { + rc = -EIO; + cFYI(1, "%s: Downcall contained malformed key (datalen=%hu, " + "ksid_size=%u)", __func__, sidkey->datalen, ksid_size); + goto invalidate_key; } - if (!test_and_set_bit(SID_ID_PENDING, &psidid->state)) { - saved_cred = override_creds(root_cred); - sidkey = request_key(&cifs_idmap_key_type, psidid->sidstr, ""); - if (IS_ERR(sidkey)) { - rc = -EINVAL; - cFYI(1, "%s: Can't map and id to a SID", __func__); - } else if (sidkey->datalen < sizeof(struct cifs_sid)) { - rc = -EIO; - cFYI(1, "%s: Downcall contained malformed key " - "(datalen=%hu)", __func__, sidkey->datalen); - } else { - lsid = (struct cifs_sid *)sidkey->payload.data; - cifs_copy_sid(&psidid->sid, lsid); - cifs_copy_sid(ssid, &psidid->sid); - set_bit(SID_ID_MAPPED, &psidid->state); - key_put(sidkey); - kfree(psidid->sidstr); - } - psidid->time = jiffies; /* update ts for accessing */ - revert_creds(saved_cred); - 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; - return rc; - } - if (test_bit(SID_ID_MAPPED, &psidid->state)) - cifs_copy_sid(ssid, &psidid->sid); - else - rc = -EINVAL; - } -id_sid_out: - --psidid->refcount; + cifs_copy_sid(ssid, ksid); +out_key_put: + key_put(sidkey); +out_revert_creds: + revert_creds(saved_cred); return rc; + +invalidate_key: + key_invalidate(sidkey); + goto out_key_put; } static int @@ -417,111 +263,67 @@ 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; + struct key *sidkey; + char *sidstr; 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); - } + uid_t fuid = cifs_sb->mnt_uid; + gid_t fgid = cifs_sb->mnt_gid; /* - * 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 we have too many subauthorities, then something is really wrong. + * Just return an error. */ - if (test_bit(SID_ID_MAPPED, &psidid->state)) { - cid = psidid->id; - psidid->time = jiffies; /* update ts for accessing */ - goto sid_to_id_out; + if (unlikely(psid->num_subauth > SID_MAX_SUB_AUTHORITIES)) { + cFYI(1, "%s: %u subauthorities is too many!", __func__, + psid->num_subauth); + return -EIO; } - 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; + sidstr = sid_to_key_str(psid, sidtype); + if (!sidstr) + return -ENOMEM; + + saved_cred = override_creds(root_cred); + sidkey = request_key(&cifs_idmap_key_type, sidstr, ""); + if (IS_ERR(sidkey)) { + rc = -EINVAL; + cFYI(1, "%s: Can't map SID %s to a %cid", __func__, sidstr, + sidtype == SIDOWNER ? 'u' : 'g'); + goto out_revert_creds; + } + + /* + * FIXME: Here we assume that uid_t and gid_t are same size. It's + * probably a safe assumption but might be better to check based on + * sidtype. + */ + if (sidkey->datalen != sizeof(uid_t)) { + rc = -EIO; + cFYI(1, "%s: Downcall contained malformed key " + "(datalen=%hu)", __func__, sidkey->datalen); + key_invalidate(sidkey); + goto out_key_put; } -sid_to_id_out: - --psidid->refcount; /* decremented without spinlock */ if (sidtype == SIDOWNER) - fattr->cf_uid = cid; + memcpy(&fuid, &sidkey->payload.value, sizeof(uid_t)); else - fattr->cf_gid = cid; + memcpy(&fgid, &sidkey->payload.value, sizeof(gid_t)); + +out_key_put: + key_put(sidkey); +out_revert_creds: + revert_creds(saved_cred); + kfree(sidstr); + /* + * Note that we return 0 here unconditionally. If the mapping + * fails then we just fall back to using the mnt_uid/mnt_gid. + */ + if (sidtype == SIDOWNER) + fattr->cf_uid = fuid; + else + fattr->cf_gid = fgid; return 0; } @@ -544,19 +346,15 @@ init_cifs_idmap(void) 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); + keyring = keyring_alloc(".cifs_idmap", 0, 0, cred, + (KEY_POS_ALL & ~KEY_POS_SETATTR) | + KEY_USR_VIEW | KEY_USR_READ, + KEY_ALLOC_NOT_IN_QUOTA, NULL); 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; @@ -568,17 +366,6 @@ init_cifs_idmap(void) 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; - - spin_lock_init(&uidsidlock); - siduidtree = RB_ROOT; - spin_lock_init(&gidsidlock); - sidgidtree = RB_ROOT; - register_shrinker(&cifs_shrinker); - cFYI(1, "cifs idmap keyring: %d", key_serial(keyring)); return 0; @@ -595,89 +382,9 @@ 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", 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); - - root = &siduidtree; - spin_lock(&uidsidlock); - while ((node = rb_first(root))) - rb_erase(node, root); - spin_unlock(&uidsidlock); - - root = &sidgidtree; - spin_lock(&gidsidlock); - while ((node = rb_first(root))) - rb_erase(node, root); - spin_unlock(&gidsidlock); -} - -/* if the two SIDs (roughly equivalent to a UUID for a user or group) are - the same returns 1, if they do not match returns 0 */ -int compare_sids(const struct cifs_sid *ctsid, const struct cifs_sid *cwsid) -{ - int i; - int num_subauth, num_sat, num_saw; - - if ((!ctsid) || (!cwsid)) - return 1; - - /* compare the revision */ - 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]) { - if (ctsid->authority[i] > cwsid->authority[i]) - return 1; - else - return -1; - } - } - - /* 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 (i = 0; i < num_subauth; ++i) { - 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 0; /* sids compare/match */ -} - - /* copy ntsd, owner sid, and group sid from a security descriptor to another */ static void copy_sec_desc(const struct cifs_ntsd *pntsd, struct cifs_ntsd *pnntsd, __u32 sidsoffset) @@ -811,7 +518,7 @@ static __u16 fill_ace_for_sid(struct cifs_ace *pntace, pntace->sid.revision = psid->revision; pntace->sid.num_subauth = psid->num_subauth; - for (i = 0; i < 6; i++) + for (i = 0; i < NUM_AUTHS; i++) pntace->sid.authority[i] = psid->authority[i]; for (i = 0; i < psid->num_subauth; i++) pntace->sid.sub_auth[i] = psid->sub_auth[i]; @@ -987,8 +694,8 @@ static int parse_sid(struct cifs_sid *psid, char *end_of_acl) return -EINVAL; } - if (psid->num_subauth) { #ifdef CONFIG_CIFS_DEBUG2 + if (psid->num_subauth) { int i; cFYI(1, "SID revision %d num_auth %d", psid->revision, psid->num_subauth); @@ -1002,8 +709,8 @@ static int parse_sid(struct cifs_sid *psid, char *end_of_acl) num auths and therefore go off the end */ cFYI(1, "RID 0x%x", le32_to_cpu(psid->sub_auth[psid->num_subauth-1])); -#endif } +#endif return 0; } @@ -1307,42 +1014,39 @@ id_mode_to_cifs_acl(struct inode *inode, const char *path, __u64 nmode, /* Get the security descriptor */ pntsd = get_cifs_acl(CIFS_SB(inode->i_sb), inode, path, &secdesclen); - - /* Add three ACEs for owner, group, everyone getting rid of - other ACEs as chmod disables ACEs and set the security descriptor */ - if (IS_ERR(pntsd)) { rc = PTR_ERR(pntsd); cERROR(1, "%s: error %d getting sec desc", __func__, rc); - } else { - /* allocate memory for the smb header, - set security descriptor request security descriptor - parameters, and secuirty descriptor itself */ - - secdesclen = secdesclen < DEFSECDESCLEN ? - DEFSECDESCLEN : secdesclen; - pnntsd = kmalloc(secdesclen, GFP_KERNEL); - if (!pnntsd) { - cERROR(1, "Unable to allocate security descriptor"); - kfree(pntsd); - return -ENOMEM; - } + goto out; + } - rc = build_sec_desc(pntsd, pnntsd, secdesclen, nmode, uid, gid, - &aclflag); + /* + * Add three ACEs for owner, group, everyone getting rid of other ACEs + * as chmod disables ACEs and set the security descriptor. Allocate + * memory for the smb header, set security descriptor request security + * descriptor parameters, and secuirty descriptor itself + */ + secdesclen = max_t(u32, secdesclen, DEFAULT_SEC_DESC_LEN); + pnntsd = kmalloc(secdesclen, GFP_KERNEL); + if (!pnntsd) { + cERROR(1, "Unable to allocate security descriptor"); + kfree(pntsd); + return -ENOMEM; + } - cFYI(DBG2, "build_sec_desc rc: %d", rc); + rc = build_sec_desc(pntsd, pnntsd, secdesclen, nmode, uid, gid, + &aclflag); - if (!rc) { - /* Set the security descriptor */ - rc = set_cifs_acl(pnntsd, secdesclen, inode, - path, aclflag); - cFYI(DBG2, "set_cifs_acl rc: %d", rc); - } + cFYI(DBG2, "build_sec_desc rc: %d", rc); - kfree(pnntsd); - kfree(pntsd); + if (!rc) { + /* Set the security descriptor */ + rc = set_cifs_acl(pnntsd, secdesclen, inode, path, aclflag); + cFYI(DBG2, "set_cifs_acl rc: %d", rc); } + kfree(pnntsd); + kfree(pntsd); +out: return rc; } diff --git a/fs/cifs/cifsacl.h b/fs/cifs/cifsacl.h index 5c902c7ce524..4f3884835267 100644 --- a/fs/cifs/cifsacl.h +++ b/fs/cifs/cifsacl.h @@ -23,11 +23,8 @@ #define _CIFSACL_H -#define NUM_AUTHS 6 /* number of authority fields */ -#define NUM_SUBAUTHS 5 /* number of sub authority fields */ -#define NUM_WK_SIDS 7 /* number of well known sids */ -#define SIDNAMELENGTH 20 /* long enough for the ones we care about */ -#define DEFSECDESCLEN 192 /* sec desc len contaiting a dacl with three aces */ +#define NUM_AUTHS (6) /* number of authority fields */ +#define SID_MAX_SUB_AUTHORITIES (15) /* max number of sub authority fields */ #define READ_BIT 0x4 #define WRITE_BIT 0x2 @@ -41,12 +38,32 @@ #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 */ +/* + * Security Descriptor length containing DACL with 3 ACEs (one each for + * owner, group and world). + */ +#define DEFAULT_SEC_DESC_LEN (sizeof(struct cifs_ntsd) + \ + sizeof(struct cifs_acl) + \ + (sizeof(struct cifs_ace) * 3)) + +/* + * Maximum size of a string representation of a SID: + * + * The fields are unsigned values in decimal. So: + * + * u8: max 3 bytes in decimal + * u32: max 10 bytes in decimal + * + * "S-" + 3 bytes for version field + 15 for authority field + NULL terminator + * + * For authority field, max is when all 6 values are non-zero and it must be + * represented in hex. So "-0x" + 12 hex digits. + * + * Add 11 bytes for each subauthority field (10 bytes each + 1 for '-') + */ +#define SID_STRING_BASE_SIZE (2 + 3 + 15 + 1) +#define SID_STRING_SUBAUTH_SIZE (11) /* size of a single subauth string */ struct cifs_ntsd { __le16 revision; /* revision level */ @@ -60,10 +77,13 @@ struct cifs_ntsd { struct cifs_sid { __u8 revision; /* revision level */ __u8 num_subauth; - __u8 authority[6]; - __le32 sub_auth[5]; /* sub_auth[num_subauth] */ + __u8 authority[NUM_AUTHS]; + __le32 sub_auth[SID_MAX_SUB_AUTHORITIES]; /* sub_auth[num_subauth] */ } __attribute__((packed)); +/* size of a struct cifs_sid, sans sub_auth array */ +#define CIFS_SID_BASE_SIZE (1 + 1 + NUM_AUTHS) + struct cifs_acl { __le16 revision; /* revision level */ __le16 size; @@ -78,26 +98,4 @@ struct cifs_ace { struct cifs_sid sid; /* ie UUID of user or group who gets these perms */ } __attribute__((packed)); -struct cifs_wksid { - struct cifs_sid cifssid; - char sidname[SIDNAMELENGTH]; -} __attribute__((packed)); - -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/cifsfs.c b/fs/cifs/cifsfs.c index e7931cc55d0c..210f0af83fc4 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -64,24 +64,23 @@ unsigned int global_secflags = CIFSSEC_DEF; unsigned int sign_CIFS_PDUs = 1; static const struct super_operations cifs_super_ops; unsigned int CIFSMaxBufSize = CIFS_MAX_MSGSIZE; -module_param(CIFSMaxBufSize, int, 0); +module_param(CIFSMaxBufSize, uint, 0); MODULE_PARM_DESC(CIFSMaxBufSize, "Network buffer size (not including header). " "Default: 16384 Range: 8192 to 130048"); unsigned int cifs_min_rcv = CIFS_MIN_RCV_POOL; -module_param(cifs_min_rcv, int, 0); +module_param(cifs_min_rcv, uint, 0); MODULE_PARM_DESC(cifs_min_rcv, "Network buffers in pool. Default: 4 Range: " "1 to 64"); unsigned int cifs_min_small = 30; -module_param(cifs_min_small, int, 0); +module_param(cifs_min_small, uint, 0); MODULE_PARM_DESC(cifs_min_small, "Small network buffers in pool. Default: 30 " "Range: 2 to 256"); unsigned int cifs_max_pending = CIFS_MAX_REQ; -module_param(cifs_max_pending, int, 0444); +module_param(cifs_max_pending, uint, 0444); MODULE_PARM_DESC(cifs_max_pending, "Simultaneous requests to server. " "Default: 32767 Range: 2 to 32767."); module_param(enable_oplocks, bool, 0644); -MODULE_PARM_DESC(enable_oplocks, "Enable or disable oplocks (bool). Default:" - "y/Y/1"); +MODULE_PARM_DESC(enable_oplocks, "Enable or disable oplocks. Default: y/Y/1"); extern mempool_t *cifs_sm_req_poolp; extern mempool_t *cifs_req_poolp; @@ -230,6 +229,7 @@ cifs_alloc_inode(struct super_block *sb) cifs_set_oplock_level(cifs_inode, 0); cifs_inode->delete_pending = false; cifs_inode->invalid_mapping = false; + cifs_inode->leave_pages_clean = false; cifs_inode->vfs_inode.i_blkbits = 14; /* 2**14 = CIFS_MAX_MSGSIZE */ cifs_inode->server_eof = 0; cifs_inode->uniqueid = 0; @@ -540,8 +540,8 @@ cifs_get_root(struct smb_vol *vol, struct super_block *sb) char *s, *p; char sep; - full_path = build_path_to_root(vol, cifs_sb, - cifs_sb_master_tcon(cifs_sb)); + full_path = cifs_build_path_to_root(vol, cifs_sb, + cifs_sb_master_tcon(cifs_sb)); if (full_path == NULL) return ERR_PTR(-ENOMEM); @@ -1205,7 +1205,6 @@ exit_cifs(void) unregister_filesystem(&cifs_fs_type); cifs_dfs_release_automount_timer(); #ifdef CONFIG_CIFS_ACL - cifs_destroy_idmaptrees(); exit_cifs_idmap(); #endif #ifdef CONFIG_CIFS_UPCALL diff --git a/fs/cifs/cifsglob.h b/fs/cifs/cifsglob.h index f5af2527fc69..aea1eec64911 100644 --- a/fs/cifs/cifsglob.h +++ b/fs/cifs/cifsglob.h @@ -178,6 +178,7 @@ struct smb_rqst { enum smb_version { Smb_1 = 1, + Smb_20, Smb_21, Smb_30, }; @@ -280,9 +281,6 @@ struct smb_version_operations { /* set attributes */ int (*set_file_info)(struct inode *, const char *, FILE_BASIC_INFO *, const unsigned int); - /* build a full path to the root of the mount */ - char * (*build_path_to_root)(struct smb_vol *, struct cifs_sb_info *, - struct cifs_tcon *); /* check if we can send an echo or nor */ bool (*can_echo)(struct TCP_Server_Info *); /* send echo request */ @@ -369,6 +367,8 @@ struct smb_version_operations { void (*set_lease_key)(struct inode *, struct cifs_fid *fid); /* generate new lease key */ void (*new_lease_key)(struct cifs_fid *fid); + int (*calc_signature)(struct smb_rqst *rqst, + struct TCP_Server_Info *server); }; struct smb_version_values { @@ -396,7 +396,6 @@ struct smb_vol { 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 */ @@ -444,11 +443,11 @@ struct smb_vol { unsigned int rsize; unsigned int wsize; bool sockopt_tcp_nodelay:1; - unsigned short int port; unsigned long actimeo; /* attribute cache timeout (jiffies) */ struct smb_version_operations *ops; struct smb_version_values *vals; char *prepath; + struct sockaddr_storage dstaddr; /* destination address */ struct sockaddr_storage srcaddr; /* allow binding to a local IP */ struct nls_table *local_nls; }; @@ -1031,6 +1030,7 @@ struct cifsInodeInfo { bool clientCanCacheAll; /* read and writebehind oplock */ bool delete_pending; /* DELETE_ON_CLOSE is set */ bool invalid_mapping; /* pagecache is invalid */ + bool leave_pages_clean; /* protected by i_mutex, not set pages dirty */ unsigned long time; /* jiffies of last update of inode */ u64 server_eof; /* current file size on server -- protected by i_lock */ u64 uniqueid; /* server inode number */ @@ -1067,30 +1067,16 @@ static inline char CIFS_DIR_SEP(const struct cifs_sb_info *cifs_sb) static inline void convert_delimiter(char *path, char delim) { - int i; - char old_delim; - - if (path == NULL) - return; + char old_delim, *pos; if (delim == '/') old_delim = '\\'; else old_delim = '/'; - for (i = 0; path[i] != '\0'; i++) { - if (path[i] == old_delim) - path[i] = delim; - } -} - -static inline char * -build_path_to_root(struct smb_vol *vol, struct cifs_sb_info *cifs_sb, - struct cifs_tcon *tcon) -{ - if (!vol->ops->build_path_to_root) - return NULL; - return vol->ops->build_path_to_root(vol, cifs_sb, tcon); + pos = path; + while ((pos = strchr(pos, old_delim))) + *pos = delim; } #ifdef CONFIG_CIFS_STATS @@ -1362,7 +1348,7 @@ require use of the stronger protocol */ #define CIFSSEC_MUST_SEAL 0x40040 /* not supported yet */ #define CIFSSEC_MUST_NTLMSSP 0x80080 /* raw ntlmssp with ntlmv2 */ -#define CIFSSEC_DEF (CIFSSEC_MAY_SIGN | CIFSSEC_MAY_NTLM | CIFSSEC_MAY_NTLMV2 | CIFSSEC_MAY_NTLMSSP) +#define CIFSSEC_DEF (CIFSSEC_MAY_SIGN | CIFSSEC_MAY_NTLMSSP) #define CIFSSEC_MAX (CIFSSEC_MUST_SIGN | CIFSSEC_MUST_NTLMV2) #define CIFSSEC_AUTH_MASK (CIFSSEC_MAY_NTLM | CIFSSEC_MAY_NTLMV2 | CIFSSEC_MAY_LANMAN | CIFSSEC_MAY_PLNTXT | CIFSSEC_MAY_KRB5 | CIFSSEC_MAY_NTLMSSP) /* @@ -1506,6 +1492,6 @@ extern struct smb_version_values smb20_values; extern struct smb_version_operations smb21_operations; extern struct smb_version_values smb21_values; #define SMB30_VERSION_STRING "3.0" -/*extern struct smb_version_operations smb30_operations; */ /* not needed yet */ +extern struct smb_version_operations smb30_operations; extern struct smb_version_values smb30_values; #endif /* _CIFS_GLOB_H */ diff --git a/fs/cifs/cifsproto.h b/fs/cifs/cifsproto.h index 5144e9fbeb8c..1988c1baa224 100644 --- a/fs/cifs/cifsproto.h +++ b/fs/cifs/cifsproto.h @@ -58,8 +58,10 @@ do { \ } 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 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, @@ -107,9 +109,7 @@ extern unsigned int smbCalcSize(void *buf); 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 void cifs_set_port(struct sockaddr *addr, const unsigned short int port); extern int map_smb_to_linux_error(char *buf, bool logErr); extern void header_assemble(struct smb_hdr *, char /* command */ , const struct cifs_tcon *, int /* length of @@ -185,7 +185,7 @@ extern void cifs_mark_open_files_invalid(struct cifs_tcon *tcon); extern bool cifs_find_lock_conflict(struct cifsFileInfo *cfile, __u64 offset, __u64 length, __u8 type, struct cifsLockInfo **conf_lock, - bool rw_check); + int rw_check); extern void cifs_add_pending_open(struct cifs_fid *fid, struct tcon_link *tlink, struct cifs_pending_open *open); diff --git a/fs/cifs/connect.c b/fs/cifs/connect.c index 5c670b998ffb..7635b5db26a7 100644 --- a/fs/cifs/connect.c +++ b/fs/cifs/connect.c @@ -186,6 +186,7 @@ static const match_table_t cifs_mount_option_tokens = { { Opt_user, "user=%s" }, { Opt_user, "username=%s" }, { Opt_blank_pass, "pass=" }, + { Opt_blank_pass, "password=" }, { Opt_pass, "pass=%s" }, { Opt_pass, "password=%s" }, { Opt_blank_ip, "ip=" }, @@ -274,6 +275,7 @@ static const match_table_t cifs_cacheflavor_tokens = { static const match_table_t cifs_smb_version_tokens = { { Smb_1, SMB1_VERSION_STRING }, + { Smb_20, SMB20_VERSION_STRING}, { Smb_21, SMB21_VERSION_STRING }, { Smb_30, SMB30_VERSION_STRING }, }; @@ -1074,12 +1076,16 @@ cifs_parse_smb_version(char *value, struct smb_vol *vol) vol->vals = &smb1_values; break; #ifdef CONFIG_CIFS_SMB2 + case Smb_20: + vol->ops = &smb21_operations; /* currently identical with 2.1 */ + vol->vals = &smb20_values; + break; case Smb_21: vol->ops = &smb21_operations; vol->vals = &smb21_values; break; case Smb_30: - vol->ops = &smb21_operations; /* currently identical with 2.1 */ + vol->ops = &smb30_operations; vol->vals = &smb30_values; break; #endif @@ -1090,6 +1096,52 @@ cifs_parse_smb_version(char *value, struct smb_vol *vol) return 0; } +/* + * Parse a devname into substrings and populate the vol->UNC and vol->prepath + * fields with the result. Returns 0 on success and an error otherwise. + */ +static int +cifs_parse_devname(const char *devname, struct smb_vol *vol) +{ + char *pos; + const char *delims = "/\\"; + size_t len; + + /* make sure we have a valid UNC double delimiter prefix */ + len = strspn(devname, delims); + if (len != 2) + return -EINVAL; + + /* find delimiter between host and sharename */ + pos = strpbrk(devname + 2, delims); + if (!pos) + return -EINVAL; + + /* skip past delimiter */ + ++pos; + + /* now go until next delimiter or end of string */ + len = strcspn(pos, delims); + + /* move "pos" up to delimiter or NULL */ + pos += len; + vol->UNC = kstrndup(devname, pos - devname, GFP_KERNEL); + if (!vol->UNC) + return -ENOMEM; + + convert_delimiter(vol->UNC, '\\'); + + /* If pos is NULL, or is a bogus trailing delimiter then no prepath */ + if (!*pos++ || !*pos) + return 0; + + vol->prepath = kstrdup(pos, GFP_KERNEL); + if (!vol->prepath) + return -ENOMEM; + + return 0; +} + static int cifs_parse_mount_options(const char *mountdata, const char *devname, struct smb_vol *vol) @@ -1108,11 +1160,17 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, char *string = NULL; char *tmp_end, *value; char delim; + bool got_ip = false; + unsigned short port = 0; + struct sockaddr *dstaddr = (struct sockaddr *)&vol->dstaddr; separator[0] = ','; separator[1] = 0; delim = separator[0]; + /* ensure we always start with zeroed-out smb_vol */ + memset(vol, 0, sizeof(*vol)); + /* * does not have to be perfect mapping since field is * informational, only used for servers that do not support @@ -1169,6 +1227,16 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, vol->backupuid_specified = false; /* no backup intent for a user */ vol->backupgid_specified = false; /* no backup intent for a group */ + /* + * For now, we ignore -EINVAL errors under the assumption that the + * unc= and prefixpath= options will be usable. + */ + if (cifs_parse_devname(devname, vol) == -ENOMEM) { + printk(KERN_ERR "CIFS: Unable to allocate memory to parse " + "device string.\n"); + goto out_nomem; + } + while ((data = strsep(&options, separator)) != NULL) { substring_t args[MAX_OPT_ARGS]; unsigned long option; @@ -1416,12 +1484,12 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, vol->dir_mode = option; break; case Opt_port: - if (get_option_ul(args, &option)) { - cERROR(1, "%s: Invalid port value", - __func__); + if (get_option_ul(args, &option) || + option > USHRT_MAX) { + cERROR(1, "%s: Invalid port value", __func__); goto cifs_parse_mount_err; } - vol->port = option; + port = (unsigned short)option; break; case Opt_rsize: if (get_option_ul(args, &option)) { @@ -1537,53 +1605,48 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, vol->password[j] = '\0'; break; case Opt_blank_ip: - vol->UNCip = NULL; + /* FIXME: should this be an error instead? */ + got_ip = false; break; case Opt_ip: string = match_strdup(args); if (string == NULL) goto out_nomem; - if (strnlen(string, INET6_ADDRSTRLEN) > - INET6_ADDRSTRLEN) { - printk(KERN_WARNING "CIFS: ip address " - "too long\n"); - goto cifs_parse_mount_err; - } - vol->UNCip = kstrdup(string, GFP_KERNEL); - if (!vol->UNCip) { - printk(KERN_WARNING "CIFS: no memory " - "for UNC IP\n"); + if (!cifs_convert_address(dstaddr, string, + strlen(string))) { + printk(KERN_ERR "CIFS: bad ip= option (%s).\n", + string); goto cifs_parse_mount_err; } + got_ip = true; break; case Opt_unc: - string = match_strdup(args); - if (string == NULL) + string = vol->UNC; + vol->UNC = match_strdup(args); + if (vol->UNC == NULL) { + kfree(string); goto out_nomem; - - temp_len = strnlen(string, 300); - if (temp_len == 300) { - printk(KERN_WARNING "CIFS: UNC name too long\n"); - goto cifs_parse_mount_err; } - vol->UNC = kmalloc(temp_len+1, GFP_KERNEL); - if (vol->UNC == NULL) { - printk(KERN_WARNING "CIFS: no memory for UNC\n"); - goto cifs_parse_mount_err; - } - strcpy(vol->UNC, string); - - if (strncmp(string, "//", 2) == 0) { - vol->UNC[0] = '\\'; - vol->UNC[1] = '\\'; - } else if (strncmp(string, "\\\\", 2) != 0) { - printk(KERN_WARNING "CIFS: UNC Path does not " - "begin with // or \\\\\n"); + convert_delimiter(vol->UNC, '\\'); + if (vol->UNC[0] != '\\' || vol->UNC[1] != '\\') { + kfree(string); + printk(KERN_ERR "CIFS: UNC Path does not " + "begin with // or \\\\\n"); goto cifs_parse_mount_err; } + /* Compare old unc= option to new one */ + if (!string || strcmp(string, vol->UNC)) + printk(KERN_WARNING "CIFS: the value of the " + "unc= mount option does not match the " + "device string. Using the unc= option " + "for now. In 3.10, that option will " + "be ignored and the contents of the " + "device string will be used " + "instead. (%s != %s)\n", string, + vol->UNC); break; case Opt_domain: string = match_strdup(args); @@ -1618,31 +1681,26 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, } break; case Opt_prefixpath: - string = match_strdup(args); - if (string == NULL) - goto out_nomem; - - temp_len = strnlen(string, 1024); - if (string[0] != '/') - temp_len++; /* missing leading slash */ - if (temp_len > 1024) { - printk(KERN_WARNING "CIFS: prefix too long\n"); - goto cifs_parse_mount_err; - } + /* skip over any leading delimiter */ + if (*args[0].from == '/' || *args[0].from == '\\') + args[0].from++; - vol->prepath = kmalloc(temp_len+1, GFP_KERNEL); + string = vol->prepath; + vol->prepath = match_strdup(args); if (vol->prepath == NULL) { - printk(KERN_WARNING "CIFS: no memory " - "for path prefix\n"); - goto cifs_parse_mount_err; + kfree(string); + goto out_nomem; } - - if (string[0] != '/') { - vol->prepath[0] = '/'; - strcpy(vol->prepath+1, string); - } else - strcpy(vol->prepath, string); - + /* Compare old prefixpath= option to new one */ + if (!string || strcmp(string, vol->prepath)) + printk(KERN_WARNING "CIFS: the value of the " + "prefixpath= mount option does not " + "match the device string. Using the " + "prefixpath= option for now. In 3.10, " + "that option will be ignored and the " + "contents of the device string will be " + "used instead.(%s != %s)\n", string, + vol->prepath); break; case Opt_iocharset: string = match_strdup(args); @@ -1799,9 +1857,30 @@ cifs_parse_mount_options(const char *mountdata, const char *devname, goto cifs_parse_mount_err; } #endif + if (!vol->UNC) { + cERROR(1, "CIFS mount error: No usable UNC path provided in " + "device string or in unc= option!"); + goto cifs_parse_mount_err; + } - if (vol->UNCip == NULL) - vol->UNCip = &vol->UNC[2]; + /* make sure UNC has a share name */ + if (!strchr(vol->UNC + 3, '\\')) { + cERROR(1, "Malformed UNC. Unable to find share name."); + goto cifs_parse_mount_err; + } + + if (!got_ip) { + /* No ip= option specified? Try to get it from UNC */ + if (!cifs_convert_address(dstaddr, &vol->UNC[2], + strlen(&vol->UNC[2]))) { + printk(KERN_ERR "Unable to determine destination " + "address.\n"); + goto cifs_parse_mount_err; + } + } + + /* set the port that we got earlier */ + cifs_set_port(dstaddr, port); if (uid_specified) vol->override_uid = override_uid; @@ -1972,9 +2051,10 @@ match_security(struct TCP_Server_Info *server, struct smb_vol *vol) return true; } -static int match_server(struct TCP_Server_Info *server, struct sockaddr *addr, - struct smb_vol *vol) +static int match_server(struct TCP_Server_Info *server, struct smb_vol *vol) { + struct sockaddr *addr = (struct sockaddr *)&vol->dstaddr; + if ((server->vals != vol->vals) || (server->ops != vol->ops)) return 0; @@ -1995,13 +2075,13 @@ static int match_server(struct TCP_Server_Info *server, struct sockaddr *addr, } static struct TCP_Server_Info * -cifs_find_tcp_session(struct sockaddr *addr, struct smb_vol *vol) +cifs_find_tcp_session(struct smb_vol *vol) { struct TCP_Server_Info *server; spin_lock(&cifs_tcp_ses_lock); list_for_each_entry(server, &cifs_tcp_ses_list, tcp_ses_list) { - if (!match_server(server, addr, vol)) + if (!match_server(server, vol)) continue; ++server->srv_count; @@ -2051,40 +2131,12 @@ static struct TCP_Server_Info * cifs_get_tcp_session(struct smb_vol *volume_info) { struct TCP_Server_Info *tcp_ses = NULL; - struct sockaddr_storage addr; - struct sockaddr_in *sin_server = (struct sockaddr_in *) &addr; - struct sockaddr_in6 *sin_server6 = (struct sockaddr_in6 *) &addr; int rc; - memset(&addr, 0, sizeof(struct sockaddr_storage)); - - cFYI(1, "UNC: %s ip: %s", volume_info->UNC, volume_info->UNCip); - - if (volume_info->UNCip && volume_info->UNC) { - rc = cifs_fill_sockaddr((struct sockaddr *)&addr, - volume_info->UNCip, - strlen(volume_info->UNCip), - volume_info->port); - if (!rc) { - /* we failed translating address */ - rc = -EINVAL; - goto out_err; - } - } else if (volume_info->UNCip) { - /* BB using ip addr as tcp_ses name to connect to the - DFS root below */ - cERROR(1, "Connecting to DFS root not implemented yet"); - rc = -EINVAL; - goto out_err; - } else /* which tcp_sess DFS root would we conect to */ { - cERROR(1, "CIFS mount error: No UNC path (e.g. -o " - "unc=//192.168.1.100/public) specified"); - rc = -EINVAL; - goto out_err; - } + cFYI(1, "UNC: %s", volume_info->UNC); /* see if we already have a matching tcp_ses */ - tcp_ses = cifs_find_tcp_session((struct sockaddr *)&addr, volume_info); + tcp_ses = cifs_find_tcp_session(volume_info); if (tcp_ses) return tcp_ses; @@ -2129,27 +2181,18 @@ cifs_get_tcp_session(struct smb_vol *volume_info) INIT_LIST_HEAD(&tcp_ses->tcp_ses_list); INIT_LIST_HEAD(&tcp_ses->smb_ses_list); INIT_DELAYED_WORK(&tcp_ses->echo, cifs_echo_request); - + memcpy(&tcp_ses->srcaddr, &volume_info->srcaddr, + sizeof(tcp_ses->srcaddr)); + memcpy(&tcp_ses->dstaddr, &volume_info->dstaddr, + sizeof(tcp_ses->dstaddr)); /* * at this point we are the only ones with the pointer * to the struct since the kernel thread not created yet * no need to spinlock this init of tcpStatus or srv_count */ tcp_ses->tcpStatus = CifsNew; - memcpy(&tcp_ses->srcaddr, &volume_info->srcaddr, - sizeof(tcp_ses->srcaddr)); ++tcp_ses->srv_count; - if (addr.ss_family == AF_INET6) { - cFYI(1, "attempting ipv6 connect"); - /* BB should we allow ipv6 on port 139? */ - /* other OS never observed in Wild doing 139 with v6 */ - memcpy(&tcp_ses->dstaddr, sin_server6, - sizeof(struct sockaddr_in6)); - } else - memcpy(&tcp_ses->dstaddr, sin_server, - sizeof(struct sockaddr_in)); - rc = ip_connect(tcp_ses); if (rc < 0) { cERROR(1, "Error connecting to socket. Aborting operation"); @@ -2397,8 +2440,6 @@ cifs_set_cifscreds(struct smb_vol *vol __attribute__((unused)), } #endif /* CONFIG_KEYS */ -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) { @@ -2475,14 +2516,6 @@ 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 3.3"); - } ses->overrideSecFlg = volume_info->secFlg; mutex_lock(&ses->session_mutex); @@ -2598,13 +2631,6 @@ cifs_get_tcon(struct cifs_ses *ses, struct smb_vol *volume_info) } } - if (strchr(volume_info->UNC + 3, '\\') == NULL - && strchr(volume_info->UNC + 3, '/') == NULL) { - cERROR(1, "Missing share name"); - rc = -ENODEV; - goto out_fail; - } - /* * BB Do we need to wrap session_mutex around this TCon call and Unix * SetFS as we do on SessSetup and reconnect? @@ -2718,11 +2744,8 @@ cifs_match_super(struct super_block *sb, void *data) 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)); @@ -2736,17 +2759,7 @@ cifs_match_super(struct super_block *sb, void *data) 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) || + if (!match_server(tcp_srv, volume_info) || !match_session(ses, volume_info) || !match_tcon(tcon, volume_info->UNC)) { rc = 0; @@ -3261,8 +3274,6 @@ cleanup_volume_info_contents(struct smb_vol *volume_info) { kfree(volume_info->username); kzfree(volume_info->password); - if (volume_info->UNCip != volume_info->UNC + 2) - kfree(volume_info->UNCip); kfree(volume_info->UNC); kfree(volume_info->domainname); kfree(volume_info->iocharset); @@ -3280,14 +3291,16 @@ cifs_cleanup_volume_info(struct smb_vol *volume_info) #ifdef CONFIG_CIFS_DFS_UPCALL -/* build_path_to_root returns full path to root when - * we do not have an exiting connection (tcon) */ +/* + * cifs_build_path_to_root returns full path to root when we do not have an + * exiting connection (tcon) + */ static char * build_unc_path_to_root(const struct smb_vol *vol, const struct cifs_sb_info *cifs_sb) { char *full_path, *pos; - unsigned int pplen = vol->prepath ? strlen(vol->prepath) : 0; + unsigned int pplen = vol->prepath ? strlen(vol->prepath) + 1 : 0; unsigned int unc_len = strnlen(vol->UNC, MAX_TREE_SIZE + 1); full_path = kmalloc(unc_len + pplen + 1, GFP_KERNEL); @@ -3298,6 +3311,7 @@ build_unc_path_to_root(const struct smb_vol *vol, pos = full_path + unc_len; if (pplen) { + *pos++ = CIFS_DIR_SEP(cifs_sb); strncpy(pos, vol->prepath, pplen); pos += pplen; } @@ -3353,7 +3367,6 @@ expand_dfs_referral(const unsigned int xid, struct cifs_ses *ses, mdata = NULL; } else { cleanup_volume_info_contents(volume_info); - memset(volume_info, '\0', sizeof(*volume_info)); rc = cifs_setup_volume_info(volume_info, mdata, fake_devname); } @@ -3375,7 +3388,6 @@ cifs_setup_volume_info(struct smb_vol *volume_info, char *mount_data, if (cifs_parse_mount_options(mount_data, devname, volume_info)) return -EINVAL; - if (volume_info->nullauth) { cFYI(1, "Anonymous login"); kfree(volume_info->username); @@ -3412,7 +3424,7 @@ cifs_get_volume_info(char *mount_data, const char *devname) int rc; struct smb_vol *volume_info; - volume_info = kzalloc(sizeof(struct smb_vol), GFP_KERNEL); + volume_info = kmalloc(sizeof(struct smb_vol), GFP_KERNEL); if (!volume_info) return ERR_PTR(-ENOMEM); @@ -3537,8 +3549,10 @@ remote_path_check: rc = -ENOSYS; goto mount_fail_check; } - /* build_path_to_root works only when we have a valid tcon */ - full_path = build_path_to_root(volume_info, cifs_sb, tcon); + /* + * cifs_build_path_to_root works only when we have a valid tcon + */ + full_path = cifs_build_path_to_root(volume_info, cifs_sb, tcon); if (full_path == NULL) { rc = -ENOMEM; goto mount_fail_check; diff --git a/fs/cifs/dir.c b/fs/cifs/dir.c index d3671f2acb29..8719bbe0dcc3 100644 --- a/fs/cifs/dir.c +++ b/fs/cifs/dir.c @@ -44,6 +44,38 @@ renew_parental_timestamps(struct dentry *direntry) } while (!IS_ROOT(direntry)); } +char * +cifs_build_path_to_root(struct smb_vol *vol, struct cifs_sb_info *cifs_sb, + struct cifs_tcon *tcon) +{ + int pplen = vol->prepath ? strlen(vol->prepath) + 1 : 0; + int dfsplen; + char *full_path = NULL; + + /* if no prefix path, simply set path to the root of share to "" */ + if (pplen == 0) { + full_path = kzalloc(1, GFP_KERNEL); + return full_path; + } + + if (tcon->Flags & SMB_SHARE_IS_IN_DFS) + dfsplen = strnlen(tcon->treeName, MAX_TREE_SIZE + 1); + else + dfsplen = 0; + + full_path = kmalloc(dfsplen + pplen + 1, GFP_KERNEL); + if (full_path == NULL) + return full_path; + + if (dfsplen) + strncpy(full_path, tcon->treeName, dfsplen); + full_path[dfsplen] = CIFS_DIR_SEP(cifs_sb); + strncpy(full_path + dfsplen + 1, vol->prepath, pplen); + convert_delimiter(full_path, CIFS_DIR_SEP(cifs_sb)); + full_path[dfsplen + pplen] = 0; /* add trailing null */ + return full_path; +} + /* Note: caller must free return buffer */ char * build_path_from_dentry(struct dentry *direntry) diff --git a/fs/cifs/file.c b/fs/cifs/file.c index edb25b4bbb95..0a6677ba212b 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -505,16 +505,36 @@ out: return rc; } +static int cifs_push_posix_locks(struct cifsFileInfo *cfile); + /* * Try to reacquire byte range locks that were released when session - * to server was lost + * to server was lost. */ -static int cifs_relock_file(struct cifsFileInfo *cifsFile) +static int +cifs_relock_file(struct cifsFileInfo *cfile) { + struct cifs_sb_info *cifs_sb = CIFS_SB(cfile->dentry->d_sb); + struct cifsInodeInfo *cinode = CIFS_I(cfile->dentry->d_inode); + struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); int rc = 0; - /* BB list all locks open on this file and relock */ + /* we are going to update can_cache_brlcks here - need a write access */ + down_write(&cinode->lock_sem); + if (cinode->can_cache_brlcks) { + /* can cache locks - no need to push them */ + up_write(&cinode->lock_sem); + return rc; + } + if (cap_unix(tcon->ses) && + (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) && + ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0)) + rc = cifs_push_posix_locks(cfile); + else + rc = tcon->ses->server->ops->push_mand_locks(cfile); + + up_write(&cinode->lock_sem); return rc; } @@ -739,10 +759,15 @@ cifs_del_lock_waiters(struct cifsLockInfo *lock) } } +#define CIFS_LOCK_OP 0 +#define CIFS_READ_OP 1 +#define CIFS_WRITE_OP 2 + +/* @rw_check : 0 - no op, 1 - read, 2 - write */ static bool cifs_find_fid_lock_conflict(struct cifs_fid_locks *fdlocks, __u64 offset, __u64 length, __u8 type, struct cifsFileInfo *cfile, - struct cifsLockInfo **conf_lock, bool rw_check) + struct cifsLockInfo **conf_lock, int rw_check) { struct cifsLockInfo *li; struct cifsFileInfo *cur_cfile = fdlocks->cfile; @@ -752,9 +777,13 @@ cifs_find_fid_lock_conflict(struct cifs_fid_locks *fdlocks, __u64 offset, if (offset + length <= li->offset || offset >= li->offset + li->length) continue; - if (rw_check && server->ops->compare_fids(cfile, cur_cfile) && - current->tgid == li->pid) - continue; + if (rw_check != CIFS_LOCK_OP && current->tgid == li->pid && + server->ops->compare_fids(cfile, cur_cfile)) { + /* shared lock prevents write op through the same fid */ + if (!(li->type & server->vals->shared_lock_type) || + rw_check != CIFS_WRITE_OP) + continue; + } if ((type & server->vals->shared_lock_type) && ((server->ops->compare_fids(cfile, cur_cfile) && current->tgid == li->pid) || type == li->type)) @@ -769,7 +798,7 @@ cifs_find_fid_lock_conflict(struct cifs_fid_locks *fdlocks, __u64 offset, bool cifs_find_lock_conflict(struct cifsFileInfo *cfile, __u64 offset, __u64 length, __u8 type, struct cifsLockInfo **conf_lock, - bool rw_check) + int rw_check) { bool rc = false; struct cifs_fid_locks *cur; @@ -805,7 +834,7 @@ cifs_lock_test(struct cifsFileInfo *cfile, __u64 offset, __u64 length, down_read(&cinode->lock_sem); exist = cifs_find_lock_conflict(cfile, offset, length, type, - &conf_lock, false); + &conf_lock, CIFS_LOCK_OP); if (exist) { flock->fl_start = conf_lock->offset; flock->fl_end = conf_lock->offset + conf_lock->length - 1; @@ -852,7 +881,7 @@ try_again: down_write(&cinode->lock_sem); exist = cifs_find_lock_conflict(cfile, lock->offset, lock->length, - lock->type, &conf_lock, false); + lock->type, &conf_lock, CIFS_LOCK_OP); if (!exist && cinode->can_cache_brlcks) { list_add_tail(&lock->llist, &cfile->llist->locks); up_write(&cinode->lock_sem); @@ -948,7 +977,6 @@ cifs_push_mandatory_locks(struct cifsFileInfo *cfile) int rc = 0, stored_rc; struct cifsLockInfo *li, *tmp; struct cifs_tcon *tcon; - struct cifsInodeInfo *cinode = CIFS_I(cfile->dentry->d_inode); unsigned int num, max_num, max_buf; LOCKING_ANDX_RANGE *buf, *cur; int types[] = {LOCKING_ANDX_LARGE_FILES, @@ -958,21 +986,12 @@ cifs_push_mandatory_locks(struct cifsFileInfo *cfile) xid = get_xid(); tcon = tlink_tcon(cfile->tlink); - /* we are going to update can_cache_brlcks here - need a write access */ - down_write(&cinode->lock_sem); - if (!cinode->can_cache_brlcks) { - up_write(&cinode->lock_sem); - free_xid(xid); - return rc; - } - /* * Accessing maxBuf is racy with cifs_reconnect - need to store value * and check it for zero before using. */ max_buf = tcon->ses->server->maxBuf; if (!max_buf) { - up_write(&cinode->lock_sem); free_xid(xid); return -EINVAL; } @@ -981,7 +1000,6 @@ cifs_push_mandatory_locks(struct cifsFileInfo *cfile) sizeof(LOCKING_ANDX_RANGE); buf = kzalloc(max_num * sizeof(LOCKING_ANDX_RANGE), GFP_KERNEL); if (!buf) { - up_write(&cinode->lock_sem); free_xid(xid); return -ENOMEM; } @@ -1018,9 +1036,6 @@ cifs_push_mandatory_locks(struct cifsFileInfo *cfile) } } - cinode->can_cache_brlcks = false; - up_write(&cinode->lock_sem); - kfree(buf); free_xid(xid); return rc; @@ -1043,7 +1058,6 @@ struct lock_to_push { static int cifs_push_posix_locks(struct cifsFileInfo *cfile) { - struct cifsInodeInfo *cinode = CIFS_I(cfile->dentry->d_inode); struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); struct file_lock *flock, **before; unsigned int count = 0, i = 0; @@ -1054,14 +1068,6 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile) xid = get_xid(); - /* we are going to update can_cache_brlcks here - need a write access */ - down_write(&cinode->lock_sem); - if (!cinode->can_cache_brlcks) { - up_write(&cinode->lock_sem); - free_xid(xid); - return rc; - } - lock_flocks(); cifs_for_each_lock(cfile->dentry->d_inode, before) { if ((*before)->fl_flags & FL_POSIX) @@ -1127,9 +1133,6 @@ cifs_push_posix_locks(struct cifsFileInfo *cfile) } out: - cinode->can_cache_brlcks = false; - up_write(&cinode->lock_sem); - free_xid(xid); return rc; err_out: @@ -1144,14 +1147,27 @@ static int cifs_push_locks(struct cifsFileInfo *cfile) { struct cifs_sb_info *cifs_sb = CIFS_SB(cfile->dentry->d_sb); + struct cifsInodeInfo *cinode = CIFS_I(cfile->dentry->d_inode); struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); + int rc = 0; + + /* we are going to update can_cache_brlcks here - need a write access */ + down_write(&cinode->lock_sem); + if (!cinode->can_cache_brlcks) { + up_write(&cinode->lock_sem); + return rc; + } if (cap_unix(tcon->ses) && (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) && ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0)) - return cifs_push_posix_locks(cfile); + rc = cifs_push_posix_locks(cfile); + else + rc = tcon->ses->server->ops->push_mand_locks(cfile); - return tcon->ses->server->ops->push_mand_locks(cfile); + cinode->can_cache_brlcks = false; + up_write(&cinode->lock_sem); + return rc; } static void @@ -1436,16 +1452,18 @@ cifs_setlk(struct file *file, struct file_lock *flock, __u32 type, return -ENOMEM; rc = cifs_lock_add_if(cfile, lock, wait_flag); - if (rc < 0) + if (rc < 0) { kfree(lock); - if (rc <= 0) + return rc; + } + if (!rc) goto out; rc = server->ops->mand_lock(xid, cfile, flock->fl_start, length, type, 1, 0, wait_flag); if (rc) { kfree(lock); - goto out; + return rc; } cifs_lock_add(cfile, lock); @@ -1794,7 +1812,6 @@ static int cifs_writepages(struct address_space *mapping, struct TCP_Server_Info *server; struct page *page; int rc = 0; - loff_t isize = i_size_read(mapping->host); /* * If wsize is smaller than the page cache size, default to writing @@ -1899,7 +1916,7 @@ retry: */ set_page_writeback(page); - if (page_offset(page) >= isize) { + if (page_offset(page) >= i_size_read(mapping->host)) { done = true; unlock_page(page); end_page_writeback(page); @@ -1932,7 +1949,8 @@ retry: wdata->offset = page_offset(wdata->pages[0]); wdata->pagesz = PAGE_CACHE_SIZE; wdata->tailsz = - min(isize - page_offset(wdata->pages[nr_pages - 1]), + min(i_size_read(mapping->host) - + page_offset(wdata->pages[nr_pages - 1]), (loff_t)PAGE_CACHE_SIZE); wdata->bytes = ((nr_pages - 1) * PAGE_CACHE_SIZE) + wdata->tailsz; @@ -2085,7 +2103,15 @@ static int cifs_write_end(struct file *file, struct address_space *mapping, } else { rc = copied; pos += copied; - set_page_dirty(page); + /* + * When we use strict cache mode and cifs_strict_writev was run + * with level II oplock (indicated by leave_pages_clean field of + * CIFS_I(inode)), we can leave pages clean - cifs_strict_writev + * sent the data to the server itself. + */ + if (!CIFS_I(inode)->leave_pages_clean || + !(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_STRICT_IO)) + set_page_dirty(page); } if (rc > 0) { @@ -2436,8 +2462,8 @@ ssize_t cifs_user_writev(struct kiocb *iocb, const struct iovec *iov, } static ssize_t -cifs_writev(struct kiocb *iocb, const struct iovec *iov, - unsigned long nr_segs, loff_t pos) +cifs_pagecache_writev(struct kiocb *iocb, const struct iovec *iov, + unsigned long nr_segs, loff_t pos, bool cache_ex) { struct file *file = iocb->ki_filp; struct cifsFileInfo *cfile = (struct cifsFileInfo *)file->private_data; @@ -2457,10 +2483,14 @@ cifs_writev(struct kiocb *iocb, const struct iovec *iov, down_read(&cinode->lock_sem); if (!cifs_find_lock_conflict(cfile, pos, iov_length(iov, nr_segs), server->vals->exclusive_lock_type, NULL, - true)) { + CIFS_WRITE_OP)) { mutex_lock(&inode->i_mutex); + if (!cache_ex) + cinode->leave_pages_clean = true; rc = __generic_file_aio_write(iocb, iov, nr_segs, - &iocb->ki_pos); + &iocb->ki_pos); + if (!cache_ex) + cinode->leave_pages_clean = false; mutex_unlock(&inode->i_mutex); } @@ -2487,42 +2517,62 @@ cifs_strict_writev(struct kiocb *iocb, const struct iovec *iov, struct cifsFileInfo *cfile = (struct cifsFileInfo *) iocb->ki_filp->private_data; struct cifs_tcon *tcon = tlink_tcon(cfile->tlink); - -#ifdef CONFIG_CIFS_SMB2 + ssize_t written, written2; /* - * If we have an oplock for read and want to write a data to the file - * we need to store it in the page cache and then push it to the server - * to be sure the next read will get a valid data. + * We need to store clientCanCacheAll here to prevent race + * conditions - this value can be changed during an execution + * of generic_file_aio_write. For CIFS it can be changed from + * true to false only, but for SMB2 it can be changed both from + * true to false and vice versa. So, we can end up with a data + * stored in the cache, not marked dirty and not sent to the + * server if this value changes its state from false to true + * after cifs_write_end. */ - if (!cinode->clientCanCacheAll && cinode->clientCanCacheRead) { - ssize_t written; - int rc; - - written = generic_file_aio_write(iocb, iov, nr_segs, pos); - rc = filemap_fdatawrite(inode->i_mapping); - if (rc) - return (ssize_t)rc; + bool cache_ex = cinode->clientCanCacheAll; + bool cache_read = cinode->clientCanCacheRead; + int rc; + loff_t saved_pos; - return written; + if (cache_ex) { + if (cap_unix(tcon->ses) && + ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0) && + (CIFS_UNIX_FCNTL_CAP & le64_to_cpu( + tcon->fsUnixInfo.Capability))) + return generic_file_aio_write(iocb, iov, nr_segs, pos); + return cifs_pagecache_writev(iocb, iov, nr_segs, pos, cache_ex); } -#endif /* - * For non-oplocked files in strict cache mode we need to write the data - * to the server exactly from the pos to pos+len-1 rather than flush all - * affected pages because it may cause a error with mandatory locks on - * these pages but not on the region from pos to ppos+len-1. + * For files without exclusive oplock in strict cache mode we need to + * write the data to the server exactly from the pos to pos+len-1 rather + * than flush all affected pages because it may cause a error with + * mandatory locks on these pages but not on the region from pos to + * ppos+len-1. */ + written = cifs_user_writev(iocb, iov, nr_segs, pos); + if (!cache_read || written <= 0) + return written; - if (!cinode->clientCanCacheAll) - return cifs_user_writev(iocb, iov, nr_segs, pos); - + saved_pos = iocb->ki_pos; + iocb->ki_pos = pos; + /* we have a read oplock - need to store a data in the page cache */ if (cap_unix(tcon->ses) && - (CIFS_UNIX_FCNTL_CAP & le64_to_cpu(tcon->fsUnixInfo.Capability)) && - ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0)) - return generic_file_aio_write(iocb, iov, nr_segs, pos); - - return cifs_writev(iocb, iov, nr_segs, pos); + ((cifs_sb->mnt_cifs_flags & CIFS_MOUNT_NOPOSIXBRL) == 0) && + (CIFS_UNIX_FCNTL_CAP & le64_to_cpu( + tcon->fsUnixInfo.Capability))) + written2 = generic_file_aio_write(iocb, iov, nr_segs, pos); + else + written2 = cifs_pagecache_writev(iocb, iov, nr_segs, pos, + cache_ex); + /* errors occured during writing - invalidate the page cache */ + if (written2 < 0) { + rc = cifs_invalidate_mapping(inode); + if (rc) + written = (ssize_t)rc; + else + iocb->ki_pos = saved_pos; + } + return written; } static struct cifs_readdata * @@ -2892,7 +2942,7 @@ cifs_strict_readv(struct kiocb *iocb, const struct iovec *iov, down_read(&cinode->lock_sem); if (!cifs_find_lock_conflict(cfile, pos, iov_length(iov, nr_segs), tcon->ses->server->vals->shared_lock_type, - NULL, true)) + NULL, CIFS_READ_OP)) rc = generic_file_aio_read(iocb, iov, nr_segs, pos); up_read(&cinode->lock_sem); return rc; @@ -3536,7 +3586,7 @@ void cifs_oplock_break(struct work_struct *work) if (cinode->clientCanCacheRead == 0) { rc = filemap_fdatawait(inode->i_mapping); mapping_set_error(inode->i_mapping, rc); - invalidate_remote_inode(inode); + cifs_invalidate_mapping(inode); } cFYI(1, "Oplock flush inode %p rc %d", inode, rc); } diff --git a/fs/cifs/inode.c b/fs/cifs/inode.c index afdff79651f1..ed6208ff85a7 100644 --- a/fs/cifs/inode.c +++ b/fs/cifs/inode.c @@ -1791,11 +1791,12 @@ int cifs_getattr(struct vfsmount *mnt, struct dentry *dentry, 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 on a multiuser mount without unix extensions or cifsacl being + * enabled, 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) && + !(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) && !tcon->unix_ext) { if (!(cifs_sb->mnt_cifs_flags & CIFS_MOUNT_OVERR_UID)) stat->uid = current_fsuid(); diff --git a/fs/cifs/netmisc.c b/fs/cifs/netmisc.c index d5ce9e26696c..a82bc51fdc82 100644 --- a/fs/cifs/netmisc.c +++ b/fs/cifs/netmisc.c @@ -204,7 +204,7 @@ cifs_convert_address(struct sockaddr *dst, const char *src, int len) return rc; } -int +void cifs_set_port(struct sockaddr *addr, const unsigned short int port) { switch (addr->sa_family) { @@ -214,19 +214,7 @@ cifs_set_port(struct sockaddr *addr, const unsigned short int port) case AF_INET6: ((struct sockaddr_in6 *)addr)->sin6_port = htons(port); break; - default: - return 0; } - return 1; -} - -int -cifs_fill_sockaddr(struct sockaddr *dst, const char *src, int len, - const unsigned short int port) -{ - if (!cifs_convert_address(dst, src, len)) - return 0; - return cifs_set_port(dst, port); } /***************************************************************************** diff --git a/fs/cifs/readdir.c b/fs/cifs/readdir.c index f9b5d3d6cf33..6002fdc920ae 100644 --- a/fs/cifs/readdir.c +++ b/fs/cifs/readdir.c @@ -66,18 +66,20 @@ static inline void dump_cifs_file_struct(struct file *file, char *label) #endif /* DEBUG2 */ /* + * Attempt to preload the dcache with the results from the FIND_FIRST/NEXT + * * Find the dentry that matches "name". If there isn't one, create one. If it's * a negative dentry or the uniqueid changed, then drop it and recreate it. */ -static struct dentry * -cifs_readdir_lookup(struct dentry *parent, struct qstr *name, +static void +cifs_prime_dcache(struct dentry *parent, struct qstr *name, struct cifs_fattr *fattr) { struct dentry *dentry, *alias; struct inode *inode; struct super_block *sb = parent->d_inode->i_sb; - cFYI(1, "For %s", name->name); + cFYI(1, "%s: for %s", __func__, name->name); if (parent->d_op && parent->d_op->d_hash) parent->d_op->d_hash(parent, parent->d_inode, name); @@ -86,35 +88,33 @@ cifs_readdir_lookup(struct dentry *parent, struct qstr *name, dentry = d_lookup(parent, name); if (dentry) { + int err; + inode = dentry->d_inode; /* update inode in place if i_ino didn't change */ if (inode && CIFS_I(inode)->uniqueid == fattr->cf_uniqueid) { cifs_fattr_to_inode(inode, fattr); - return dentry; + goto out; } - d_drop(dentry); + err = d_invalidate(dentry); dput(dentry); + if (err) + return; } dentry = d_alloc(parent, name); - if (dentry == NULL) - return NULL; + if (!dentry) + return; inode = cifs_iget(sb, fattr); - if (!inode) { - dput(dentry); - return NULL; - } + if (!inode) + goto out; alias = d_materialise_unique(dentry, inode); - if (alias != NULL) { - dput(dentry); - if (IS_ERR(alias)) - return NULL; - dentry = alias; - } - - return dentry; + if (alias && !IS_ERR(alias)) + dput(alias); +out: + dput(dentry); } static void @@ -134,6 +134,16 @@ cifs_fill_common_info(struct cifs_fattr *fattr, struct cifs_sb_info *cifs_sb) if (fattr->cf_cifsattrs & ATTR_READONLY) fattr->cf_mode &= ~S_IWUGO; + /* + * We of course don't get ACL info in FIND_FIRST/NEXT results, so + * mark it for revalidation so that "ls -l" will look right. It might + * be super-slow, but if we don't do this then the ownership of files + * may look wrong since the inodes may not have timed out by the time + * "ls" does a stat() call on them. + */ + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_CIFS_ACL) + fattr->cf_flags |= CIFS_FATTR_NEED_REVAL; + if (cifs_sb->mnt_cifs_flags & CIFS_MOUNT_UNX_EMUL && fattr->cf_cifsattrs & ATTR_SYSTEM) { if (fattr->cf_eof == 0) { @@ -649,7 +659,6 @@ static int cifs_filldir(char *find_entry, struct file *file, filldir_t filldir, struct cifs_sb_info *cifs_sb = CIFS_SB(sb); struct cifs_dirent de = { NULL, }; struct cifs_fattr fattr; - struct dentry *dentry; struct qstr name; int rc = 0; ino_t ino; @@ -720,13 +729,11 @@ static int cifs_filldir(char *find_entry, struct file *file, filldir_t filldir, */ fattr.cf_flags |= CIFS_FATTR_NEED_REVAL; - ino = cifs_uniqueid_to_ino_t(fattr.cf_uniqueid); - dentry = cifs_readdir_lookup(file->f_dentry, &name, &fattr); + cifs_prime_dcache(file->f_dentry, &name, &fattr); + ino = cifs_uniqueid_to_ino_t(fattr.cf_uniqueid); rc = filldir(dirent, name.name, name.len, file->f_pos, ino, fattr.cf_dtype); - - dput(dentry); return rc; } diff --git a/fs/cifs/smb1ops.c b/fs/cifs/smb1ops.c index 56cc4be87807..a5d234c8d5d9 100644 --- a/fs/cifs/smb1ops.c +++ b/fs/cifs/smb1ops.c @@ -575,37 +575,6 @@ cifs_query_file_info(const unsigned int xid, struct cifs_tcon *tcon, return CIFSSMBQFileInfo(xid, tcon, fid->netfid, data); } -static char * -cifs_build_path_to_root(struct smb_vol *vol, struct cifs_sb_info *cifs_sb, - struct cifs_tcon *tcon) -{ - int pplen = vol->prepath ? strlen(vol->prepath) : 0; - int dfsplen; - char *full_path = NULL; - - /* if no prefix path, simply set path to the root of share to "" */ - if (pplen == 0) { - full_path = kzalloc(1, GFP_KERNEL); - return full_path; - } - - if (tcon->Flags & SMB_SHARE_IS_IN_DFS) - dfsplen = strnlen(tcon->treeName, MAX_TREE_SIZE + 1); - else - dfsplen = 0; - - full_path = kmalloc(dfsplen + pplen + 1, GFP_KERNEL); - if (full_path == NULL) - return full_path; - - if (dfsplen) - strncpy(full_path, tcon->treeName, dfsplen); - strncpy(full_path + dfsplen, vol->prepath, pplen); - convert_delimiter(full_path, CIFS_DIR_SEP(cifs_sb)); - full_path[dfsplen + pplen] = 0; /* add trailing null */ - return full_path; -} - static void cifs_clear_stats(struct cifs_tcon *tcon) { @@ -766,7 +735,6 @@ smb_set_file_info(struct inode *inode, const char *full_path, struct cifs_sb_info *cifs_sb = CIFS_SB(inode->i_sb); struct tcon_link *tlink = NULL; struct cifs_tcon *tcon; - FILE_BASIC_INFO info_buf; /* if the file is already open for write, just use that fileid */ open_file = find_writable_file(cinode, true); @@ -817,7 +785,7 @@ smb_set_file_info(struct inode *inode, const char *full_path, netpid = current->tgid; set_via_filehandle: - rc = CIFSSMBSetFileInfo(xid, tcon, &info_buf, netfid, netpid); + rc = CIFSSMBSetFileInfo(xid, tcon, buf, netfid, netpid); if (!rc) cinode->cifsAttrs = le32_to_cpu(buf->Attributes); @@ -944,7 +912,6 @@ struct smb_version_operations smb1_operations = { .set_path_size = CIFSSMBSetEOF, .set_file_size = CIFSSMBSetFileSize, .set_file_info = smb_set_file_info, - .build_path_to_root = cifs_build_path_to_root, .echo = CIFSSMBEcho, .mkdir = CIFSSMBMkDir, .mkdir_setinfo = cifs_mkdir_setinfo, diff --git a/fs/cifs/smb2file.c b/fs/cifs/smb2file.c index a93eec30a50d..71e6aed4b382 100644 --- a/fs/cifs/smb2file.c +++ b/fs/cifs/smb2file.c @@ -260,13 +260,6 @@ smb2_push_mandatory_locks(struct cifsFileInfo *cfile) struct cifs_fid_locks *fdlocks; xid = get_xid(); - /* we are going to update can_cache_brlcks here - need a write access */ - down_write(&cinode->lock_sem); - if (!cinode->can_cache_brlcks) { - up_write(&cinode->lock_sem); - free_xid(xid); - return rc; - } /* * Accessing maxBuf is racy with cifs_reconnect - need to store value @@ -274,7 +267,6 @@ smb2_push_mandatory_locks(struct cifsFileInfo *cfile) */ max_buf = tlink_tcon(cfile->tlink)->ses->server->maxBuf; if (!max_buf) { - up_write(&cinode->lock_sem); free_xid(xid); return -EINVAL; } @@ -282,7 +274,6 @@ smb2_push_mandatory_locks(struct cifsFileInfo *cfile) max_num = max_buf / sizeof(struct smb2_lock_element); buf = kzalloc(max_num * sizeof(struct smb2_lock_element), GFP_KERNEL); if (!buf) { - up_write(&cinode->lock_sem); free_xid(xid); return -ENOMEM; } @@ -293,10 +284,7 @@ smb2_push_mandatory_locks(struct cifsFileInfo *cfile) rc = stored_rc; } - cinode->can_cache_brlcks = false; kfree(buf); - - up_write(&cinode->lock_sem); free_xid(xid); return rc; } diff --git a/fs/cifs/smb2ops.c b/fs/cifs/smb2ops.c index 4d9dbe0b7385..d79de7bc4435 100644 --- a/fs/cifs/smb2ops.c +++ b/fs/cifs/smb2ops.c @@ -262,23 +262,6 @@ smb2_query_file_info(const unsigned int xid, struct cifs_tcon *tcon, return rc; } -static char * -smb2_build_path_to_root(struct smb_vol *vol, struct cifs_sb_info *cifs_sb, - struct cifs_tcon *tcon) -{ - int pplen = vol->prepath ? strlen(vol->prepath) : 0; - char *full_path = NULL; - - /* if no prefix path, simply set path to the root of share to "" */ - if (pplen == 0) { - full_path = kzalloc(2, GFP_KERNEL); - return full_path; - } - - cERROR(1, "prefixpath is not supported for SMB2 now"); - return NULL; -} - static bool smb2_can_echo(struct TCP_Server_Info *server) { @@ -613,7 +596,6 @@ struct smb_version_operations smb21_operations = { .set_path_size = smb2_set_path_size, .set_file_size = smb2_set_file_size, .set_file_info = smb2_set_file_info, - .build_path_to_root = smb2_build_path_to_root, .mkdir = smb2_mkdir, .mkdir_setinfo = smb2_mkdir_setinfo, .rmdir = smb2_rmdir, @@ -641,6 +623,91 @@ struct smb_version_operations smb21_operations = { .get_lease_key = smb2_get_lease_key, .set_lease_key = smb2_set_lease_key, .new_lease_key = smb2_new_lease_key, + .calc_signature = smb2_calc_signature, +}; + + +struct smb_version_operations smb30_operations = { + .compare_fids = smb2_compare_fids, + .setup_request = smb2_setup_request, + .setup_async_request = smb2_setup_async_request, + .check_receive = smb2_check_receive, + .add_credits = smb2_add_credits, + .set_credits = smb2_set_credits, + .get_credits_field = smb2_get_credits_field, + .get_credits = smb2_get_credits, + .get_next_mid = smb2_get_next_mid, + .read_data_offset = smb2_read_data_offset, + .read_data_length = smb2_read_data_length, + .map_error = map_smb2_to_linux_error, + .find_mid = smb2_find_mid, + .check_message = smb2_check_message, + .dump_detail = smb2_dump_detail, + .clear_stats = smb2_clear_stats, + .print_stats = smb2_print_stats, + .is_oplock_break = smb2_is_valid_oplock_break, + .need_neg = smb2_need_neg, + .negotiate = smb2_negotiate, + .negotiate_wsize = smb2_negotiate_wsize, + .negotiate_rsize = smb2_negotiate_rsize, + .sess_setup = SMB2_sess_setup, + .logoff = SMB2_logoff, + .tree_connect = SMB2_tcon, + .tree_disconnect = SMB2_tdis, + .is_path_accessible = smb2_is_path_accessible, + .can_echo = smb2_can_echo, + .echo = SMB2_echo, + .query_path_info = smb2_query_path_info, + .get_srv_inum = smb2_get_srv_inum, + .query_file_info = smb2_query_file_info, + .set_path_size = smb2_set_path_size, + .set_file_size = smb2_set_file_size, + .set_file_info = smb2_set_file_info, + .mkdir = smb2_mkdir, + .mkdir_setinfo = smb2_mkdir_setinfo, + .rmdir = smb2_rmdir, + .unlink = smb2_unlink, + .rename = smb2_rename_path, + .create_hardlink = smb2_create_hardlink, + .open = smb2_open_file, + .set_fid = smb2_set_fid, + .close = smb2_close_file, + .flush = smb2_flush_file, + .async_readv = smb2_async_readv, + .async_writev = smb2_async_writev, + .sync_read = smb2_sync_read, + .sync_write = smb2_sync_write, + .query_dir_first = smb2_query_dir_first, + .query_dir_next = smb2_query_dir_next, + .close_dir = smb2_close_dir, + .calc_smb_size = smb2_calc_size, + .is_status_pending = smb2_is_status_pending, + .oplock_response = smb2_oplock_response, + .queryfs = smb2_queryfs, + .mand_lock = smb2_mand_lock, + .mand_unlock_range = smb2_unlock_range, + .push_mand_locks = smb2_push_mandatory_locks, + .get_lease_key = smb2_get_lease_key, + .set_lease_key = smb2_set_lease_key, + .new_lease_key = smb2_new_lease_key, + .calc_signature = smb3_calc_signature, +}; + +struct smb_version_values smb20_values = { + .version_string = SMB20_VERSION_STRING, + .protocol_id = SMB20_PROT_ID, + .req_capabilities = 0, /* MBZ */ + .large_lock_type = 0, + .exclusive_lock_type = SMB2_LOCKFLAG_EXCLUSIVE_LOCK, + .shared_lock_type = SMB2_LOCKFLAG_SHARED_LOCK, + .unlock_lock_type = SMB2_LOCKFLAG_UNLOCK, + .header_size = sizeof(struct smb2_hdr), + .max_header_size = MAX_SMB2_HDR_SIZE, + .read_rsp_size = sizeof(struct smb2_read_rsp) - 1, + .lock_cmd = SMB2_LOCK, + .cap_unix = 0, + .cap_nt_find = SMB2_NT_FIND, + .cap_large_files = SMB2_LARGE_FILES, }; struct smb_version_values smb21_values = { diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c index cf33622cdac8..41d9d0725f0f 100644 --- a/fs/cifs/smb2pdu.c +++ b/fs/cifs/smb2pdu.c @@ -425,7 +425,7 @@ SMB2_negotiate(const unsigned int xid, struct cifs_ses *ses) } cFYI(1, "sec_flags 0x%x", sec_flags); - if (sec_flags & CIFSSEC_MUST_SIGN) { + if ((sec_flags & CIFSSEC_MUST_SIGN) == CIFSSEC_MUST_SIGN) { cFYI(1, "Signing required"); if (!(server->sec_mode & (SMB2_NEGOTIATE_SIGNING_REQUIRED | SMB2_NEGOTIATE_SIGNING_ENABLED))) { @@ -612,7 +612,8 @@ ssetup_ntlmssp_authenticate: /* BB add code to build os and lm fields */ - rc = SendReceive2(xid, ses, iov, 2, &resp_buftype, CIFS_LOG_ERROR); + rc = SendReceive2(xid, ses, iov, 2, &resp_buftype, + CIFS_LOG_ERROR | CIFS_NEG_OP); kfree(security_blob); rsp = (struct smb2_sess_setup_rsp *)iov[0].iov_base; diff --git a/fs/cifs/smb2proto.h b/fs/cifs/smb2proto.h index 7d25f8b14f93..2aa3535e38ce 100644 --- a/fs/cifs/smb2proto.h +++ b/fs/cifs/smb2proto.h @@ -47,6 +47,10 @@ extern struct mid_q_entry *smb2_setup_request(struct cifs_ses *ses, struct smb_rqst *rqst); extern struct mid_q_entry *smb2_setup_async_request( struct TCP_Server_Info *server, struct smb_rqst *rqst); +extern int smb2_calc_signature(struct smb_rqst *rqst, + struct TCP_Server_Info *server); +extern int smb3_calc_signature(struct smb_rqst *rqst, + struct TCP_Server_Info *server); extern void smb2_echo_request(struct work_struct *work); extern __le32 smb2_get_lease_state(struct cifsInodeInfo *cinode); extern __u8 smb2_map_lease_to_oplock(__le32 lease_state); diff --git a/fs/cifs/smb2transport.c b/fs/cifs/smb2transport.c index 2a5fdf26f79f..8dd73e61d762 100644 --- a/fs/cifs/smb2transport.c +++ b/fs/cifs/smb2transport.c @@ -39,7 +39,7 @@ #include "smb2status.h" #include "smb2glob.h" -static int +int smb2_calc_signature(struct smb_rqst *rqst, struct TCP_Server_Info *server) { int i, rc; @@ -116,6 +116,13 @@ smb2_calc_signature(struct smb_rqst *rqst, struct TCP_Server_Info *server) return rc; } +int +smb3_calc_signature(struct smb_rqst *rqst, struct TCP_Server_Info *server) +{ + cFYI(1, "smb3 signatures not supported yet"); + return -EOPNOTSUPP; +} + /* must be called with server->srv_mutex held */ static int smb2_sign_rqst(struct smb_rqst *rqst, struct TCP_Server_Info *server) @@ -132,7 +139,7 @@ smb2_sign_rqst(struct smb_rqst *rqst, struct TCP_Server_Info *server) return rc; } - rc = smb2_calc_signature(rqst, server); + rc = server->ops->calc_signature(rqst, server); return rc; } @@ -168,7 +175,7 @@ smb2_verify_signature(struct smb_rqst *rqst, struct TCP_Server_Info *server) memset(smb2_pdu->Signature, 0, SMB2_SIGNATURE_SIZE); mutex_lock(&server->srv_mutex); - rc = smb2_calc_signature(rqst, server); + rc = server->ops->calc_signature(rqst, server); mutex_unlock(&server->srv_mutex); if (rc) diff --git a/fs/compat_ioctl.c b/fs/compat_ioctl.c index 4c6285fff598..e2f57a007029 100644 --- a/fs/compat_ioctl.c +++ b/fs/compat_ioctl.c @@ -844,6 +844,9 @@ COMPATIBLE_IOCTL(TIOCGDEV) COMPATIBLE_IOCTL(TIOCCBRK) COMPATIBLE_IOCTL(TIOCGSID) COMPATIBLE_IOCTL(TIOCGICOUNT) +COMPATIBLE_IOCTL(TIOCGPKT) +COMPATIBLE_IOCTL(TIOCGPTLCK) +COMPATIBLE_IOCTL(TIOCGEXCL) /* Little t */ COMPATIBLE_IOCTL(TIOCGETD) COMPATIBLE_IOCTL(TIOCSETD) diff --git a/fs/coredump.c b/fs/coredump.c index ce47379bfa61..177493272a61 100644 --- a/fs/coredump.c +++ b/fs/coredump.c @@ -458,7 +458,7 @@ static int umh_pipe_setup(struct subprocess_info *info, struct cred *new) return err; } -void do_coredump(siginfo_t *siginfo, struct pt_regs *regs) +void do_coredump(siginfo_t *siginfo) { struct core_state core_state; struct core_name cn; @@ -474,7 +474,7 @@ void do_coredump(siginfo_t *siginfo, struct pt_regs *regs) static atomic_t core_dump_count = ATOMIC_INIT(0); struct coredump_params cprm = { .siginfo = siginfo, - .regs = regs, + .regs = signal_pt_regs(), .limit = rlimit(RLIMIT_CORE), /* * We must use the same mm->flags while dumping core to avoid diff --git a/fs/debugfs/inode.c b/fs/debugfs/inode.c index b607d92cdf24..153bb1e42e63 100644 --- a/fs/debugfs/inode.c +++ b/fs/debugfs/inode.c @@ -59,7 +59,6 @@ static struct inode *debugfs_get_inode(struct super_block *sb, umode_t mode, dev case S_IFDIR: inode->i_op = &simple_dir_inode_operations; inode->i_fop = &simple_dir_operations; - inode->i_private = NULL; /* directory inodes start off with i_nlink == 2 * (for "." entry) */ diff --git a/fs/devpts/inode.c b/fs/devpts/inode.c index 14afbabe6546..472e6befc54d 100644 --- a/fs/devpts/inode.c +++ b/fs/devpts/inode.c @@ -545,37 +545,38 @@ void devpts_kill_index(struct inode *ptmx_inode, int idx) mutex_unlock(&allocated_ptys_lock); } -int devpts_pty_new(struct inode *ptmx_inode, struct tty_struct *tty) +/** + * devpts_pty_new -- create a new inode in /dev/pts/ + * @ptmx_inode: inode of the master + * @device: major+minor of the node to be created + * @index: used as a name of the node + * @priv: what's given back by devpts_get_priv + * + * The created inode is returned. Remove it from /dev/pts/ by devpts_pty_kill. + */ +struct inode *devpts_pty_new(struct inode *ptmx_inode, dev_t device, int index, + void *priv) { - /* tty layer puts index from devpts_new_index() in here */ - int number = tty->index; - struct tty_driver *driver = tty->driver; - dev_t device = MKDEV(driver->major, driver->minor_start+number); struct dentry *dentry; struct super_block *sb = pts_sb_from_inode(ptmx_inode); - struct inode *inode = new_inode(sb); + struct inode *inode; struct dentry *root = sb->s_root; struct pts_fs_info *fsi = DEVPTS_SB(sb); struct pts_mount_opts *opts = &fsi->mount_opts; - int ret = 0; char s[12]; - /* We're supposed to be given the slave end of a pty */ - BUG_ON(driver->type != TTY_DRIVER_TYPE_PTY); - BUG_ON(driver->subtype != PTY_TYPE_SLAVE); - + inode = new_inode(sb); if (!inode) - return -ENOMEM; + return ERR_PTR(-ENOMEM); - inode->i_ino = number + 3; + inode->i_ino = index + 3; inode->i_uid = opts->setuid ? opts->uid : current_fsuid(); inode->i_gid = opts->setgid ? opts->gid : current_fsgid(); inode->i_mtime = inode->i_atime = inode->i_ctime = CURRENT_TIME; init_special_inode(inode, S_IFCHR|opts->mode, device); - inode->i_private = tty; - tty->driver_data = inode; + inode->i_private = priv; - sprintf(s, "%d", number); + sprintf(s, "%d", index); mutex_lock(&root->d_inode->i_mutex); @@ -585,18 +586,24 @@ int devpts_pty_new(struct inode *ptmx_inode, struct tty_struct *tty) fsnotify_create(root->d_inode, dentry); } else { iput(inode); - ret = -ENOMEM; + inode = ERR_PTR(-ENOMEM); } mutex_unlock(&root->d_inode->i_mutex); - return ret; + return inode; } -struct tty_struct *devpts_get_tty(struct inode *pts_inode, int number) +/** + * devpts_get_priv -- get private data for a slave + * @pts_inode: inode of the slave + * + * Returns whatever was passed as priv in devpts_pty_new for a given inode. + */ +void *devpts_get_priv(struct inode *pts_inode) { struct dentry *dentry; - struct tty_struct *tty; + void *priv = NULL; BUG_ON(pts_inode->i_rdev == MKDEV(TTYAUX_MAJOR, PTMX_MINOR)); @@ -605,18 +612,22 @@ struct tty_struct *devpts_get_tty(struct inode *pts_inode, int number) if (!dentry) return NULL; - tty = NULL; if (pts_inode->i_sb->s_magic == DEVPTS_SUPER_MAGIC) - tty = (struct tty_struct *)pts_inode->i_private; + priv = pts_inode->i_private; dput(dentry); - return tty; + return priv; } -void devpts_pty_kill(struct tty_struct *tty) +/** + * devpts_pty_kill -- remove inode form /dev/pts/ + * @inode: inode of the slave to be removed + * + * This is an inverse operation of devpts_pty_new. + */ +void devpts_pty_kill(struct inode *inode) { - struct inode *inode = tty->driver_data; struct super_block *sb = pts_sb_from_inode(inode); struct dentry *root = sb->s_root; struct dentry *dentry; diff --git a/fs/direct-io.c b/fs/direct-io.c index f86c720dba0e..cf5b44b10c67 100644 --- a/fs/direct-io.c +++ b/fs/direct-io.c @@ -540,6 +540,7 @@ static int get_more_blocks(struct dio *dio, struct dio_submit *sdio, sector_t fs_endblk; /* Into file, in filesystem-sized blocks */ unsigned long fs_count; /* Number of filesystem-sized blocks */ int create; + unsigned int i_blkbits = sdio->blkbits + sdio->blkfactor; /* * If there was a memory error and we've overwritten all the @@ -554,7 +555,7 @@ static int get_more_blocks(struct dio *dio, struct dio_submit *sdio, fs_count = fs_endblk - fs_startblk + 1; map_bh->b_state = 0; - map_bh->b_size = fs_count << dio->inode->i_blkbits; + map_bh->b_size = fs_count << i_blkbits; /* * For writes inside i_size on a DIO_SKIP_HOLES filesystem we @@ -1053,7 +1054,8 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, int seg; size_t size; unsigned long addr; - unsigned blkbits = inode->i_blkbits; + unsigned i_blkbits = ACCESS_ONCE(inode->i_blkbits); + unsigned blkbits = i_blkbits; unsigned blocksize_mask = (1 << blkbits) - 1; ssize_t retval = -EINVAL; loff_t end = offset; @@ -1149,7 +1151,7 @@ do_blockdev_direct_IO(int rw, struct kiocb *iocb, struct inode *inode, dio->inode = inode; dio->rw = rw; sdio.blkbits = blkbits; - sdio.blkfactor = inode->i_blkbits - blkbits; + sdio.blkfactor = i_blkbits - blkbits; sdio.block_in_file = offset >> blkbits; sdio.get_block = get_block; diff --git a/fs/dlm/Kconfig b/fs/dlm/Kconfig index 1897eb1b4b6a..e4242c3f8486 100644 --- a/fs/dlm/Kconfig +++ b/fs/dlm/Kconfig @@ -1,6 +1,6 @@ menuconfig DLM tristate "Distributed Lock Manager (DLM)" - depends on EXPERIMENTAL && INET + depends on INET depends on SYSFS && CONFIGFS_FS && (IPV6 || IPV6=n) select IP_SCTP help diff --git a/fs/dlm/dlm_internal.h b/fs/dlm/dlm_internal.h index 871c1abf6029..77c0f70f8fe8 100644 --- a/fs/dlm/dlm_internal.h +++ b/fs/dlm/dlm_internal.h @@ -337,6 +337,7 @@ enum rsb_flags { RSB_NEW_MASTER2, RSB_RECOVER_CONVERT, RSB_RECOVER_GRANT, + RSB_RECOVER_LVB_INVAL, }; static inline void rsb_set_flag(struct dlm_rsb *r, enum rsb_flags flag) diff --git a/fs/dlm/lock.c b/fs/dlm/lock.c index b56950758188..a579f30f237d 100644 --- a/fs/dlm/lock.c +++ b/fs/dlm/lock.c @@ -5393,6 +5393,13 @@ static void purge_dead_list(struct dlm_ls *ls, struct dlm_rsb *r, if ((lkb->lkb_nodeid == nodeid_gone) || dlm_is_removed(ls, lkb->lkb_nodeid)) { + /* tell recover_lvb to invalidate the lvb + because a node holding EX/PW failed */ + if ((lkb->lkb_exflags & DLM_LKF_VALBLK) && + (lkb->lkb_grmode >= DLM_LOCK_PW)) { + rsb_set_flag(r, RSB_RECOVER_LVB_INVAL); + } + del_lkb(r, lkb); /* this put should free the lkb */ @@ -6025,15 +6032,18 @@ static int orphan_proc_lock(struct dlm_ls *ls, struct dlm_lkb *lkb) return error; } -/* The force flag allows the unlock to go ahead even if the lkb isn't granted. - Regardless of what rsb queue the lock is on, it's removed and freed. */ +/* The FORCEUNLOCK flag allows the unlock to go ahead even if the lkb isn't + granted. Regardless of what rsb queue the lock is on, it's removed and + freed. The IVVALBLK flag causes the lvb on the resource to be invalidated + if our lock is PW/EX (it's ignored if our granted mode is smaller.) */ static int unlock_proc_lock(struct dlm_ls *ls, struct dlm_lkb *lkb) { struct dlm_args args; int error; - set_unlock_args(DLM_LKF_FORCEUNLOCK, lkb->lkb_ua, &args); + set_unlock_args(DLM_LKF_FORCEUNLOCK | DLM_LKF_IVVALBLK, + lkb->lkb_ua, &args); error = unlock_lock(ls, lkb, &args); if (error == -DLM_EUNLOCK) diff --git a/fs/dlm/lowcomms.c b/fs/dlm/lowcomms.c index 331ea4f94efd..dd87a31bcc21 100644 --- a/fs/dlm/lowcomms.c +++ b/fs/dlm/lowcomms.c @@ -1385,7 +1385,6 @@ void *dlm_lowcomms_get_buffer(int nodeid, int len, gfp_t allocation, char **ppc) struct connection *con; struct writequeue_entry *e; int offset = 0; - int users = 0; con = nodeid2con(nodeid, allocation); if (!con) @@ -1399,7 +1398,7 @@ void *dlm_lowcomms_get_buffer(int nodeid, int len, gfp_t allocation, char **ppc) } else { offset = e->end; e->end += len; - users = e->users++; + e->users++; } spin_unlock(&con->writequeue_lock); @@ -1414,7 +1413,7 @@ void *dlm_lowcomms_get_buffer(int nodeid, int len, gfp_t allocation, char **ppc) spin_lock(&con->writequeue_lock); offset = e->end; e->end += len; - users = e->users++; + e->users++; list_add_tail(&e->list, &con->writequeue); spin_unlock(&con->writequeue_lock); goto got_one; diff --git a/fs/dlm/recover.c b/fs/dlm/recover.c index 4a7a76e42fc3..aedea28a86a1 100644 --- a/fs/dlm/recover.c +++ b/fs/dlm/recover.c @@ -717,8 +717,14 @@ void dlm_recovered_lock(struct dlm_rsb *r) * the VALNOTVALID flag if necessary, and determining the correct lvb contents * based on the lvb's of the locks held on the rsb. * - * RSB_VALNOTVALID is set if there are only NL/CR locks on the rsb. If it - * was already set prior to recovery, it's not cleared, regardless of locks. + * RSB_VALNOTVALID is set in two cases: + * + * 1. we are master, but not new, and we purged an EX/PW lock held by a + * failed node (in dlm_recover_purge which set RSB_RECOVER_LVB_INVAL) + * + * 2. we are a new master, and there are only NL/CR locks left. + * (We could probably improve this by only invaliding in this way when + * the previous master left uncleanly. VMS docs mention that.) * * The LVB contents are only considered for changing when this is a new master * of the rsb (NEW_MASTER2). Then, the rsb's lvb is taken from any lkb with @@ -734,6 +740,19 @@ static void recover_lvb(struct dlm_rsb *r) int big_lock_exists = 0; int lvblen = r->res_ls->ls_lvblen; + if (!rsb_flag(r, RSB_NEW_MASTER2) && + rsb_flag(r, RSB_RECOVER_LVB_INVAL)) { + /* case 1 above */ + rsb_set_flag(r, RSB_VALNOTVALID); + return; + } + + if (!rsb_flag(r, RSB_NEW_MASTER2)) + return; + + /* we are the new master, so figure out if VALNOTVALID should + be set, and set the rsb lvb from the best lkb available. */ + list_for_each_entry(lkb, &r->res_grantqueue, lkb_statequeue) { if (!(lkb->lkb_exflags & DLM_LKF_VALBLK)) continue; @@ -772,13 +791,10 @@ static void recover_lvb(struct dlm_rsb *r) if (!lock_lvb_exists) goto out; + /* lvb is invalidated if only NL/CR locks remain */ if (!big_lock_exists) rsb_set_flag(r, RSB_VALNOTVALID); - /* don't mess with the lvb unless we're the new master */ - if (!rsb_flag(r, RSB_NEW_MASTER2)) - goto out; - if (!r->res_lvbptr) { r->res_lvbptr = dlm_allocate_lvb(r->res_ls); if (!r->res_lvbptr) @@ -852,12 +868,19 @@ void dlm_recover_rsbs(struct dlm_ls *ls) if (is_master(r)) { if (rsb_flag(r, RSB_RECOVER_CONVERT)) recover_conversion(r); + + /* recover lvb before granting locks so the updated + lvb/VALNOTVALID is presented in the completion */ + recover_lvb(r); + if (rsb_flag(r, RSB_NEW_MASTER2)) recover_grant(r); - recover_lvb(r); count++; + } else { + rsb_clear_flag(r, RSB_VALNOTVALID); } rsb_clear_flag(r, RSB_RECOVER_CONVERT); + rsb_clear_flag(r, RSB_RECOVER_LVB_INVAL); rsb_clear_flag(r, RSB_NEW_MASTER2); unlock_rsb(r); } diff --git a/fs/exec.c b/fs/exec.c index 0039055b1fc6..721a29929511 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -1349,7 +1349,7 @@ EXPORT_SYMBOL(remove_arg_zero); /* * cycle the list of binary formats handler, until one recognizes the image */ -int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) +int search_binary_handler(struct linux_binprm *bprm) { unsigned int depth = bprm->recursion_depth; int try,retval; @@ -1374,13 +1374,13 @@ int search_binary_handler(struct linux_binprm *bprm,struct pt_regs *regs) for (try=0; try<2; try++) { read_lock(&binfmt_lock); list_for_each_entry(fmt, &formats, lh) { - int (*fn)(struct linux_binprm *, struct pt_regs *) = fmt->load_binary; + int (*fn)(struct linux_binprm *) = fmt->load_binary; if (!fn) continue; if (!try_module_get(fmt->module)) continue; read_unlock(&binfmt_lock); - retval = fn(bprm, regs); + retval = fn(bprm); /* * Restore the depth counter to its starting value * in this call, so we don't have to rely on every @@ -1439,8 +1439,7 @@ EXPORT_SYMBOL(search_binary_handler); */ static int do_execve_common(const char *filename, struct user_arg_ptr argv, - struct user_arg_ptr envp, - struct pt_regs *regs) + struct user_arg_ptr envp) { struct linux_binprm *bprm; struct file *file; @@ -1524,7 +1523,7 @@ static int do_execve_common(const char *filename, if (retval < 0) goto out; - retval = search_binary_handler(bprm,regs); + retval = search_binary_handler(bprm); if (retval < 0) goto out; @@ -1566,19 +1565,17 @@ out_ret: int do_execve(const char *filename, const char __user *const __user *__argv, - const char __user *const __user *__envp, - struct pt_regs *regs) + const char __user *const __user *__envp) { struct user_arg_ptr argv = { .ptr.native = __argv }; struct user_arg_ptr envp = { .ptr.native = __envp }; - return do_execve_common(filename, argv, envp, regs); + return do_execve_common(filename, argv, envp); } #ifdef CONFIG_COMPAT -int compat_do_execve(const char *filename, +static int compat_do_execve(const char *filename, const compat_uptr_t __user *__argv, - const compat_uptr_t __user *__envp, - struct pt_regs *regs) + const compat_uptr_t __user *__envp) { struct user_arg_ptr argv = { .is_compat = true, @@ -1588,7 +1585,7 @@ int compat_do_execve(const char *filename, .is_compat = true, .ptr.compat = __envp, }; - return do_execve_common(filename, argv, envp, regs); + return do_execve_common(filename, argv, envp); } #endif @@ -1669,7 +1666,7 @@ SYSCALL_DEFINE3(execve, struct filename *path = getname(filename); int error = PTR_ERR(path); if (!IS_ERR(path)) { - error = do_execve(path->name, argv, envp, current_pt_regs()); + error = do_execve(path->name, argv, envp); putname(path); } return error; @@ -1682,8 +1679,7 @@ asmlinkage long compat_sys_execve(const char __user * filename, struct filename *path = getname(filename); int error = PTR_ERR(path); if (!IS_ERR(path)) { - error = compat_do_execve(path->name, argv, envp, - current_pt_regs()); + error = compat_do_execve(path->name, argv, envp); putname(path); } return error; @@ -1696,12 +1692,9 @@ int kernel_execve(const char *filename, const char *const argv[], const char *const envp[]) { - struct pt_regs *p = current_pt_regs(); - int ret; - - ret = do_execve(filename, + int ret = do_execve(filename, (const char __user *const __user *)argv, - (const char __user *const __user *)envp, p); + (const char __user *const __user *)envp); if (ret < 0) return ret; @@ -1709,6 +1702,6 @@ int kernel_execve(const char *filename, * We were successful. We won't be returning to our caller, but * instead to user space by manipulating the kernel stack. */ - ret_from_kernel_execve(p); + ret_from_kernel_execve(current_pt_regs()); } #endif diff --git a/fs/ext4/Kconfig b/fs/ext4/Kconfig index c22f17021b6e..0a475c881852 100644 --- a/fs/ext4/Kconfig +++ b/fs/ext4/Kconfig @@ -39,22 +39,8 @@ config EXT4_USE_FOR_EXT23 compiled kernel size by using one file system driver for ext2, ext3, and ext4 file systems. -config EXT4_FS_XATTR - bool "Ext4 extended attributes" - depends on EXT4_FS - default y - 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). - - If unsure, say N. - - You need this for POSIX ACL support on ext4. - config EXT4_FS_POSIX_ACL bool "Ext4 POSIX Access Control Lists" - depends on EXT4_FS_XATTR select FS_POSIX_ACL help POSIX Access Control Lists (ACLs) support permissions for users and @@ -67,7 +53,6 @@ config EXT4_FS_POSIX_ACL config EXT4_FS_SECURITY bool "Ext4 Security Labels" - depends on EXT4_FS_XATTR help Security labels support alternative access control models implemented by security modules like SELinux. This option diff --git a/fs/ext4/Makefile b/fs/ext4/Makefile index 56fd8f865930..0310fec2ee3d 100644 --- a/fs/ext4/Makefile +++ b/fs/ext4/Makefile @@ -7,8 +7,8 @@ obj-$(CONFIG_EXT4_FS) += ext4.o ext4-y := balloc.o bitmap.o dir.o file.o fsync.o ialloc.o inode.o page-io.o \ ioctl.o namei.o super.o symlink.o hash.o resize.o extents.o \ ext4_jbd2.o migrate.o mballoc.o block_validity.o move_extent.o \ - mmp.o indirect.o + mmp.o indirect.o extents_status.o xattr.o xattr_user.o \ + xattr_trusted.o inline.o -ext4-$(CONFIG_EXT4_FS_XATTR) += xattr.o xattr_user.o xattr_trusted.o ext4-$(CONFIG_EXT4_FS_POSIX_ACL) += acl.o ext4-$(CONFIG_EXT4_FS_SECURITY) += xattr_security.o diff --git a/fs/ext4/acl.c b/fs/ext4/acl.c index d3c5b88fd89f..e6e0d988439b 100644 --- a/fs/ext4/acl.c +++ b/fs/ext4/acl.c @@ -423,8 +423,10 @@ ext4_xattr_set_acl(struct dentry *dentry, const char *name, const void *value, retry: handle = ext4_journal_start(inode, EXT4_DATA_TRANS_BLOCKS(inode->i_sb)); - if (IS_ERR(handle)) - return PTR_ERR(handle); + if (IS_ERR(handle)) { + error = PTR_ERR(handle); + goto release_and_out; + } error = ext4_set_acl(handle, inode, type, acl); ext4_journal_stop(handle); if (error == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) diff --git a/fs/ext4/dir.c b/fs/ext4/dir.c index 8e07d2a5a139..b8d877f6c1fa 100644 --- a/fs/ext4/dir.c +++ b/fs/ext4/dir.c @@ -27,23 +27,11 @@ #include <linux/slab.h> #include <linux/rbtree.h> #include "ext4.h" - -static unsigned char ext4_filetype_table[] = { - DT_UNKNOWN, DT_REG, DT_DIR, DT_CHR, DT_BLK, DT_FIFO, DT_SOCK, DT_LNK -}; +#include "xattr.h" static int ext4_dx_readdir(struct file *filp, void *dirent, filldir_t filldir); -static unsigned char get_dtype(struct super_block *sb, int filetype) -{ - if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FILETYPE) || - (filetype >= EXT4_FT_MAX)) - return DT_UNKNOWN; - - return (ext4_filetype_table[filetype]); -} - /** * Check if the given dir-inode refers to an htree-indexed directory * (or a directory which chould potentially get coverted to use htree @@ -68,11 +56,14 @@ static int is_dx_dir(struct inode *inode) * Return 0 if the directory entry is OK, and 1 if there is a problem * * Note: this is the opposite of what ext2 and ext3 historically returned... + * + * bh passed here can be an inode block or a dir data block, depending + * on the inode inline data flag. */ int __ext4_check_dir_entry(const char *function, unsigned int line, struct inode *dir, struct file *filp, struct ext4_dir_entry_2 *de, - struct buffer_head *bh, + struct buffer_head *bh, char *buf, int size, unsigned int offset) { const char *error_msg = NULL; @@ -85,9 +76,8 @@ int __ext4_check_dir_entry(const char *function, unsigned int line, error_msg = "rec_len % 4 != 0"; else if (unlikely(rlen < EXT4_DIR_REC_LEN(de->name_len))) error_msg = "rec_len is too small for name_len"; - else if (unlikely(((char *) de - bh->b_data) + rlen > - dir->i_sb->s_blocksize)) - error_msg = "directory entry across blocks"; + else if (unlikely(((char *) de - buf) + rlen > size)) + error_msg = "directory entry across range"; else if (unlikely(le32_to_cpu(de->inode) > le32_to_cpu(EXT4_SB(dir->i_sb)->s_es->s_inodes_count))) error_msg = "inode out of bounds"; @@ -98,14 +88,14 @@ int __ext4_check_dir_entry(const char *function, unsigned int line, ext4_error_file(filp, function, line, bh->b_blocknr, "bad entry in directory: %s - offset=%u(%u), " "inode=%u, rec_len=%d, name_len=%d", - error_msg, (unsigned) (offset % bh->b_size), + error_msg, (unsigned) (offset % size), offset, le32_to_cpu(de->inode), rlen, de->name_len); else ext4_error_inode(dir, function, line, bh->b_blocknr, "bad entry in directory: %s - offset=%u(%u), " "inode=%u, rec_len=%d, name_len=%d", - error_msg, (unsigned) (offset % bh->b_size), + error_msg, (unsigned) (offset % size), offset, le32_to_cpu(de->inode), rlen, de->name_len); @@ -125,6 +115,14 @@ static int ext4_readdir(struct file *filp, int ret = 0; int dir_has_error = 0; + if (ext4_has_inline_data(inode)) { + int has_inline_data = 1; + ret = ext4_read_inline_dir(filp, dirent, filldir, + &has_inline_data); + if (has_inline_data) + return ret; + } + if (is_dx_dir(inode)) { err = ext4_dx_readdir(filp, dirent, filldir); if (err != ERR_BAD_DX_DIR) { @@ -221,8 +219,9 @@ revalidate: while (!error && filp->f_pos < inode->i_size && offset < sb->s_blocksize) { de = (struct ext4_dir_entry_2 *) (bh->b_data + offset); - if (ext4_check_dir_entry(inode, filp, de, - bh, offset)) { + if (ext4_check_dir_entry(inode, filp, de, bh, + bh->b_data, bh->b_size, + offset)) { /* * On error, skip the f_pos to the next block */ diff --git a/fs/ext4/ext4.h b/fs/ext4/ext4.h index 3c20de1d59d0..8462eb3c33aa 100644 --- a/fs/ext4/ext4.h +++ b/fs/ext4/ext4.h @@ -57,6 +57,16 @@ #define ext4_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) #endif +/* + * Turn on EXT_DEBUG to get lots of info about extents operations. + */ +#define EXT_DEBUG__ +#ifdef EXT_DEBUG +#define ext_debug(fmt, ...) printk(fmt, ##__VA_ARGS__) +#else +#define ext_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) +#endif + #define EXT4_ERROR_INODE(inode, fmt, a...) \ ext4_error_inode((inode), __func__, __LINE__, 0, (fmt), ## a) @@ -392,6 +402,7 @@ struct flex_groups { #define EXT4_EXTENTS_FL 0x00080000 /* Inode uses extents */ #define EXT4_EA_INODE_FL 0x00200000 /* Inode used for large EA */ #define EXT4_EOFBLOCKS_FL 0x00400000 /* Blocks allocated beyond EOF */ +#define EXT4_INLINE_DATA_FL 0x10000000 /* Inode has inline data. */ #define EXT4_RESERVED_FL 0x80000000 /* reserved for ext4 lib */ #define EXT4_FL_USER_VISIBLE 0x004BDFFF /* User visible flags */ @@ -448,28 +459,26 @@ enum { EXT4_INODE_EXTENTS = 19, /* Inode uses extents */ EXT4_INODE_EA_INODE = 21, /* Inode used for large EA */ EXT4_INODE_EOFBLOCKS = 22, /* Blocks allocated beyond EOF */ + EXT4_INODE_INLINE_DATA = 28, /* Data in inode. */ EXT4_INODE_RESERVED = 31, /* reserved for ext4 lib */ }; -#define TEST_FLAG_VALUE(FLAG) (EXT4_##FLAG##_FL == (1 << EXT4_INODE_##FLAG)) -#define CHECK_FLAG_VALUE(FLAG) if (!TEST_FLAG_VALUE(FLAG)) { \ - printk(KERN_EMERG "EXT4 flag fail: " #FLAG ": %d %d\n", \ - EXT4_##FLAG##_FL, EXT4_INODE_##FLAG); BUG_ON(1); } - -/* - * Since it's pretty easy to mix up bit numbers and hex values, and we - * can't do a compile-time test for ENUM values, we use a run-time - * test to make sure that EXT4_XXX_FL is consistent with respect to - * EXT4_INODE_XXX. If all is well the printk and BUG_ON will all drop - * out so it won't cost any extra space in the compiled kernel image. - * But it's important that these values are the same, since we are - * using EXT4_INODE_XXX to test for the flag values, but EXT4_XX_FL - * must be consistent with the values of FS_XXX_FL defined in - * include/linux/fs.h and the on-disk values found in ext2, ext3, and - * ext4 filesystems, and of course the values defined in e2fsprogs. +/* + * Since it's pretty easy to mix up bit numbers and hex values, we use a + * build-time check to make sure that EXT4_XXX_FL is consistent with respect to + * EXT4_INODE_XXX. If all is well, the macros will be dropped, so, it won't cost + * any extra space in the compiled kernel image, otherwise, the build will fail. + * It's important that these values are the same, since we are using + * EXT4_INODE_XXX to test for flag values, but EXT4_XXX_FL must be consistent + * with the values of FS_XXX_FL defined in include/linux/fs.h and the on-disk + * values found in ext2, ext3 and ext4 filesystems, and of course the values + * defined in e2fsprogs. * * It's not paranoia if the Murphy's Law really *is* out to get you. :-) */ +#define TEST_FLAG_VALUE(FLAG) (EXT4_##FLAG##_FL == (1 << EXT4_INODE_##FLAG)) +#define CHECK_FLAG_VALUE(FLAG) BUILD_BUG_ON(!TEST_FLAG_VALUE(FLAG)) + static inline void ext4_check_flag_values(void) { CHECK_FLAG_VALUE(SECRM); @@ -494,6 +503,7 @@ static inline void ext4_check_flag_values(void) CHECK_FLAG_VALUE(EXTENTS); CHECK_FLAG_VALUE(EA_INODE); CHECK_FLAG_VALUE(EOFBLOCKS); + CHECK_FLAG_VALUE(INLINE_DATA); CHECK_FLAG_VALUE(RESERVED); } @@ -811,6 +821,8 @@ struct ext4_ext_cache { __u32 ec_len; /* must be 32bit to return holes */ }; +#include "extents_status.h" + /* * fourth extended file system inode data in memory */ @@ -833,7 +845,6 @@ struct ext4_inode_info { #endif unsigned long i_flags; -#ifdef CONFIG_EXT4_FS_XATTR /* * Extended attributes can be read independently of the main file * data. Taking i_mutex even when reading would cause contention @@ -842,7 +853,6 @@ struct ext4_inode_info { * EAs. */ struct rw_semaphore xattr_sem; -#endif struct list_head i_orphan; /* unlinked but open inodes */ @@ -888,6 +898,10 @@ struct ext4_inode_info { struct list_head i_prealloc_list; spinlock_t i_prealloc_lock; + /* extents status tree */ + struct ext4_es_tree i_es_tree; + rwlock_t i_es_lock; + /* ialloc */ ext4_group_t i_last_alloc_group; @@ -902,6 +916,10 @@ struct ext4_inode_info { /* on-disk additional length */ __u16 i_extra_isize; + /* Indicate the inline data space. */ + u16 i_inline_off; + u16 i_inline_size; + #ifdef CONFIG_QUOTA /* quota space reservation, managed internally by quota code */ qsize_t i_reserved_quota; @@ -1360,6 +1378,7 @@ enum { EXT4_STATE_DELALLOC_RESERVED, /* blks already reserved for delalloc */ EXT4_STATE_DIOREAD_LOCK, /* Disable support for dio read nolocking */ + EXT4_STATE_MAY_INLINE_DATA, /* may have in-inode data */ }; #define EXT4_INODE_BIT_FNS(name, field, offset) \ @@ -1481,7 +1500,7 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei) #define EXT4_FEATURE_INCOMPAT_DIRDATA 0x1000 /* data in dirent */ #define EXT4_FEATURE_INCOMPAT_BG_USE_META_CSUM 0x2000 /* use crc32c for bg */ #define EXT4_FEATURE_INCOMPAT_LARGEDIR 0x4000 /* >2GB or 3-lvl htree */ -#define EXT4_FEATURE_INCOMPAT_INLINEDATA 0x8000 /* data in inode */ +#define EXT4_FEATURE_INCOMPAT_INLINE_DATA 0x8000 /* data in inode */ #define EXT2_FEATURE_COMPAT_SUPP EXT4_FEATURE_COMPAT_EXT_ATTR #define EXT2_FEATURE_INCOMPAT_SUPP (EXT4_FEATURE_INCOMPAT_FILETYPE| \ @@ -1505,7 +1524,8 @@ static inline void ext4_clear_state_flags(struct ext4_inode_info *ei) EXT4_FEATURE_INCOMPAT_EXTENTS| \ EXT4_FEATURE_INCOMPAT_64BIT| \ EXT4_FEATURE_INCOMPAT_FLEX_BG| \ - EXT4_FEATURE_INCOMPAT_MMP) + EXT4_FEATURE_INCOMPAT_MMP | \ + EXT4_FEATURE_INCOMPAT_INLINE_DATA) #define EXT4_FEATURE_RO_COMPAT_SUPP (EXT4_FEATURE_RO_COMPAT_SPARSE_SUPER| \ EXT4_FEATURE_RO_COMPAT_LARGE_FILE| \ EXT4_FEATURE_RO_COMPAT_GDT_CSUM| \ @@ -1592,6 +1612,11 @@ struct ext4_dir_entry_tail { __le32 det_checksum; /* crc32c(uuid+inum+dirblock) */ }; +#define EXT4_DIRENT_TAIL(block, blocksize) \ + ((struct ext4_dir_entry_tail *)(((void *)(block)) + \ + ((blocksize) - \ + sizeof(struct ext4_dir_entry_tail)))) + /* * Ext4 directory file types. Only the low 3 bits are used. The * other bits are reserved for now. @@ -1936,14 +1961,42 @@ ext4_fsblk_t ext4_inode_to_goal_block(struct inode *); extern int __ext4_check_dir_entry(const char *, unsigned int, struct inode *, struct file *, struct ext4_dir_entry_2 *, - struct buffer_head *, unsigned int); -#define ext4_check_dir_entry(dir, filp, de, bh, offset) \ + struct buffer_head *, char *, int, + unsigned int); +#define ext4_check_dir_entry(dir, filp, de, bh, buf, size, offset) \ unlikely(__ext4_check_dir_entry(__func__, __LINE__, (dir), (filp), \ - (de), (bh), (offset))) + (de), (bh), (buf), (size), (offset))) extern int ext4_htree_store_dirent(struct file *dir_file, __u32 hash, __u32 minor_hash, struct ext4_dir_entry_2 *dirent); extern void ext4_htree_free_dir_info(struct dir_private_info *p); +extern int ext4_find_dest_de(struct inode *dir, struct inode *inode, + struct buffer_head *bh, + void *buf, int buf_size, + const char *name, int namelen, + struct ext4_dir_entry_2 **dest_de); +void ext4_insert_dentry(struct inode *inode, + struct ext4_dir_entry_2 *de, + int buf_size, + const char *name, int namelen); +static inline void ext4_update_dx_flag(struct inode *inode) +{ + if (!EXT4_HAS_COMPAT_FEATURE(inode->i_sb, + EXT4_FEATURE_COMPAT_DIR_INDEX)) + ext4_clear_inode_flag(inode, EXT4_INODE_INDEX); +} +static unsigned char ext4_filetype_table[] = { + DT_UNKNOWN, DT_REG, DT_DIR, DT_CHR, DT_BLK, DT_FIFO, DT_SOCK, DT_LNK +}; + +static inline unsigned char get_dtype(struct super_block *sb, int filetype) +{ + if (!EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_FILETYPE) || + (filetype >= EXT4_FT_MAX)) + return DT_UNKNOWN; + + return ext4_filetype_table[filetype]; +} /* fsync.c */ extern int ext4_sync_file(struct file *, loff_t, loff_t, int); @@ -1994,8 +2047,23 @@ struct buffer_head *ext4_getblk(handle_t *, struct inode *, ext4_lblk_t, int, int *); struct buffer_head *ext4_bread(handle_t *, struct inode *, ext4_lblk_t, int, int *); +int ext4_get_block_write(struct inode *inode, sector_t iblock, + struct buffer_head *bh_result, int create); int ext4_get_block(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create); +int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, + struct buffer_head *bh, int create); +int ext4_walk_page_buffers(handle_t *handle, + struct buffer_head *head, + unsigned from, + unsigned to, + int *partial, + int (*fn)(handle_t *handle, + struct buffer_head *bh)); +int do_journal_get_write_access(handle_t *handle, + struct buffer_head *bh); +#define FALL_BACK_TO_NONDELALLOC 1 +#define CONVERT_INLINE_DATA 2 extern struct inode *ext4_iget(struct super_block *, unsigned long); extern int ext4_write_inode(struct inode *, struct writeback_control *); @@ -2050,6 +2118,20 @@ extern int ext4_orphan_add(handle_t *, struct inode *); extern int ext4_orphan_del(handle_t *, struct inode *); extern int ext4_htree_fill_tree(struct file *dir_file, __u32 start_hash, __u32 start_minor_hash, __u32 *next_hash); +extern int search_dir(struct buffer_head *bh, + char *search_buf, + int buf_size, + struct inode *dir, + const struct qstr *d_name, + unsigned int offset, + struct ext4_dir_entry_2 **res_dir); +extern int ext4_generic_delete_entry(handle_t *handle, + struct inode *dir, + struct ext4_dir_entry_2 *de_del, + struct buffer_head *bh, + void *entry_buf, + int buf_size, + int csum_size); /* resize.c */ extern int ext4_group_add(struct super_block *sb, @@ -2376,6 +2458,15 @@ extern void ext4_unwritten_wait(struct inode *inode); extern const struct inode_operations ext4_dir_inode_operations; extern const struct inode_operations ext4_special_inode_operations; extern struct dentry *ext4_get_parent(struct dentry *child); +extern struct ext4_dir_entry_2 *ext4_init_dot_dotdot(struct inode *inode, + struct ext4_dir_entry_2 *de, + int blocksize, int csum_size, + unsigned int parent_ino, int dotdot_real_len); +extern void initialize_dirent_tail(struct ext4_dir_entry_tail *t, + unsigned int blocksize); +extern int ext4_handle_dirty_dirent_node(handle_t *handle, + struct inode *inode, + struct buffer_head *bh); /* symlink.c */ extern const struct inode_operations ext4_symlink_inode_operations; @@ -2393,6 +2484,9 @@ extern int ext4_check_blockref(const char *, unsigned int, struct inode *, __le32 *, unsigned int); /* extents.c */ +struct ext4_ext_path; +struct ext4_extent; + extern int ext4_ext_tree_init(handle_t *handle, struct inode *); extern int ext4_ext_writepage_trans_blocks(struct inode *, int); extern int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, @@ -2410,8 +2504,27 @@ extern int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset, ssize_t len); extern int ext4_map_blocks(handle_t *handle, struct inode *inode, struct ext4_map_blocks *map, int flags); +extern int ext4_ext_calc_metadata_amount(struct inode *inode, + ext4_lblk_t lblocks); +extern int ext4_extent_tree_init(handle_t *, struct inode *); +extern int ext4_ext_calc_credits_for_single_extent(struct inode *inode, + int num, + struct ext4_ext_path *path); +extern int ext4_can_extents_be_merged(struct inode *inode, + struct ext4_extent *ex1, + struct ext4_extent *ex2); +extern int ext4_ext_insert_extent(handle_t *, struct inode *, + struct ext4_ext_path *, + struct ext4_extent *, int); +extern struct ext4_ext_path *ext4_ext_find_extent(struct inode *, ext4_lblk_t, + struct ext4_ext_path *); +extern void ext4_ext_drop_refs(struct ext4_ext_path *); +extern int ext4_ext_check_inode(struct inode *inode); +extern int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk); extern int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len); + + /* move_extent.c */ extern int ext4_move_extents(struct file *o_filp, struct file *d_filp, __u64 start_orig, __u64 start_donor, @@ -2445,17 +2558,13 @@ enum ext4_state_bits { * never, ever appear in a buffer_head's state * flag. See EXT4_MAP_FROM_CLUSTER to see where * this is used. */ - BH_Da_Mapped, /* Delayed allocated block that now has a mapping. This - * flag is set when ext4_map_blocks is called on a - * delayed allocated block to get its real mapping. */ }; BUFFER_FNS(Uninit, uninit) TAS_BUFFER_FNS(Uninit, uninit) -BUFFER_FNS(Da_Mapped, da_mapped) /* - * Add new method to test wether block and inode bitmaps are properly + * Add new method to test whether block and inode bitmaps are properly * initialized. With uninit_bg reading the block from disk is not enough * to mark the bitmap uptodate. We need to also zero-out the bitmap */ @@ -2503,6 +2612,4 @@ extern void ext4_resize_end(struct super_block *sb); #endif /* __KERNEL__ */ -#include "ext4_extents.h" - #endif /* _EXT4_H */ diff --git a/fs/ext4/ext4_extents.h b/fs/ext4/ext4_extents.h index cb1b2c919963..487fda12bc00 100644 --- a/fs/ext4/ext4_extents.h +++ b/fs/ext4/ext4_extents.h @@ -43,16 +43,6 @@ #define CHECK_BINSEARCH__ /* - * Turn on EXT_DEBUG to get lots of info about extents operations. - */ -#define EXT_DEBUG__ -#ifdef EXT_DEBUG -#define ext_debug(fmt, ...) printk(fmt, ##__VA_ARGS__) -#else -#define ext_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) -#endif - -/* * If EXT_STATS is defined then stats numbers are collected. * These number will be displayed at umount time. */ @@ -144,20 +134,6 @@ struct ext4_ext_path { */ /* - * to be called by ext4_ext_walk_space() - * negative retcode - error - * positive retcode - signal for ext4_ext_walk_space(), see below - * callback must return valid extent (passed or newly created) - */ -typedef int (*ext_prepare_callback)(struct inode *, ext4_lblk_t, - struct ext4_ext_cache *, - struct ext4_extent *, void *); - -#define EXT_CONTINUE 0 -#define EXT_BREAK 1 -#define EXT_REPEAT 2 - -/* * Maximum number of logical blocks in a file; ext4_extent's ee_block is * __le32. */ @@ -300,21 +276,5 @@ static inline void ext4_idx_store_pblock(struct ext4_extent_idx *ix, 0xffff); } -extern int ext4_ext_calc_metadata_amount(struct inode *inode, - ext4_lblk_t lblocks); -extern int ext4_extent_tree_init(handle_t *, struct inode *); -extern int ext4_ext_calc_credits_for_single_extent(struct inode *inode, - int num, - struct ext4_ext_path *path); -extern int ext4_can_extents_be_merged(struct inode *inode, - struct ext4_extent *ex1, - struct ext4_extent *ex2); -extern int ext4_ext_insert_extent(handle_t *, struct inode *, struct ext4_ext_path *, struct ext4_extent *, int); -extern struct ext4_ext_path *ext4_ext_find_extent(struct inode *, ext4_lblk_t, - struct ext4_ext_path *); -extern void ext4_ext_drop_refs(struct ext4_ext_path *); -extern int ext4_ext_check_inode(struct inode *inode); -extern int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk, - int search_hint_reverse); #endif /* _EXT4_EXTENTS */ diff --git a/fs/ext4/ext4_jbd2.h b/fs/ext4/ext4_jbd2.h index 56d258c18303..7177f9b21cb2 100644 --- a/fs/ext4/ext4_jbd2.h +++ b/fs/ext4/ext4_jbd2.h @@ -254,13 +254,6 @@ static inline void ext4_handle_sync(handle_t *handle) handle->h_sync = 1; } -static inline void ext4_handle_release_buffer(handle_t *handle, - struct buffer_head *bh) -{ - if (ext4_handle_valid(handle)) - jbd2_journal_release_buffer(handle, bh); -} - static inline int ext4_handle_is_aborted(handle_t *handle) { if (ext4_handle_valid(handle)) diff --git a/fs/ext4/extents.c b/fs/ext4/extents.c index 7011ac967208..26af22832a84 100644 --- a/fs/ext4/extents.c +++ b/fs/ext4/extents.c @@ -41,6 +41,8 @@ #include <asm/uaccess.h> #include <linux/fiemap.h> #include "ext4_jbd2.h" +#include "ext4_extents.h" +#include "xattr.h" #include <trace/events/ext4.h> @@ -109,6 +111,9 @@ static int ext4_split_extent_at(handle_t *handle, int split_flag, int flags); +static int ext4_find_delayed_extent(struct inode *inode, + struct ext4_ext_cache *newex); + static int ext4_ext_truncate_extend_restart(handle_t *handle, struct inode *inode, int needed) @@ -1959,27 +1964,33 @@ cleanup: return err; } -static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block, - ext4_lblk_t num, ext_prepare_callback func, - void *cbdata) +static int ext4_fill_fiemap_extents(struct inode *inode, + ext4_lblk_t block, ext4_lblk_t num, + struct fiemap_extent_info *fieinfo) { struct ext4_ext_path *path = NULL; - struct ext4_ext_cache cbex; + struct ext4_ext_cache newex; struct ext4_extent *ex; - ext4_lblk_t next, start = 0, end = 0; + ext4_lblk_t next, next_del, start = 0, end = 0; ext4_lblk_t last = block + num; - int depth, exists, err = 0; - - BUG_ON(func == NULL); - BUG_ON(inode == NULL); + int exists, depth = 0, err = 0; + unsigned int flags = 0; + unsigned char blksize_bits = inode->i_sb->s_blocksize_bits; while (block < last && block != EXT_MAX_BLOCKS) { num = last - block; /* find extent for this block */ down_read(&EXT4_I(inode)->i_data_sem); + + if (path && ext_depth(inode) != depth) { + /* depth was changed. we have to realloc path */ + kfree(path); + path = NULL; + } + path = ext4_ext_find_extent(inode, block, path); - up_read(&EXT4_I(inode)->i_data_sem); if (IS_ERR(path)) { + up_read(&EXT4_I(inode)->i_data_sem); err = PTR_ERR(path); path = NULL; break; @@ -1987,13 +1998,16 @@ static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block, depth = ext_depth(inode); if (unlikely(path[depth].p_hdr == NULL)) { + up_read(&EXT4_I(inode)->i_data_sem); EXT4_ERROR_INODE(inode, "path[%d].p_hdr == NULL", depth); err = -EIO; break; } ex = path[depth].p_ext; next = ext4_ext_next_allocated_block(path); + ext4_ext_drop_refs(path); + flags = 0; exists = 0; if (!ex) { /* there is no extent yet, so try to allocate @@ -2030,40 +2044,64 @@ static int ext4_ext_walk_space(struct inode *inode, ext4_lblk_t block, BUG_ON(end <= start); if (!exists) { - cbex.ec_block = start; - cbex.ec_len = end - start; - cbex.ec_start = 0; + newex.ec_block = start; + newex.ec_len = end - start; + newex.ec_start = 0; } else { - cbex.ec_block = le32_to_cpu(ex->ee_block); - cbex.ec_len = ext4_ext_get_actual_len(ex); - cbex.ec_start = ext4_ext_pblock(ex); + newex.ec_block = le32_to_cpu(ex->ee_block); + newex.ec_len = ext4_ext_get_actual_len(ex); + newex.ec_start = ext4_ext_pblock(ex); + if (ext4_ext_is_uninitialized(ex)) + flags |= FIEMAP_EXTENT_UNWRITTEN; } - if (unlikely(cbex.ec_len == 0)) { - EXT4_ERROR_INODE(inode, "cbex.ec_len == 0"); - err = -EIO; - break; + /* + * Find delayed extent and update newex accordingly. We call + * it even in !exists case to find out whether newex is the + * last existing extent or not. + */ + next_del = ext4_find_delayed_extent(inode, &newex); + if (!exists && next_del) { + exists = 1; + flags |= FIEMAP_EXTENT_DELALLOC; } - err = func(inode, next, &cbex, ex, cbdata); - ext4_ext_drop_refs(path); + up_read(&EXT4_I(inode)->i_data_sem); - if (err < 0) + if (unlikely(newex.ec_len == 0)) { + EXT4_ERROR_INODE(inode, "newex.ec_len == 0"); + err = -EIO; break; + } - if (err == EXT_REPEAT) - continue; - else if (err == EXT_BREAK) { - err = 0; - break; + /* This is possible iff next == next_del == EXT_MAX_BLOCKS */ + if (next == next_del) { + flags |= FIEMAP_EXTENT_LAST; + if (unlikely(next_del != EXT_MAX_BLOCKS || + next != EXT_MAX_BLOCKS)) { + EXT4_ERROR_INODE(inode, + "next extent == %u, next " + "delalloc extent = %u", + next, next_del); + err = -EIO; + break; + } } - if (ext_depth(inode) != depth) { - /* depth was changed. we have to realloc path */ - kfree(path); - path = NULL; + if (exists) { + err = fiemap_fill_next_extent(fieinfo, + (__u64)newex.ec_block << blksize_bits, + (__u64)newex.ec_start << blksize_bits, + (__u64)newex.ec_len << blksize_bits, + flags); + if (err < 0) + break; + if (err == 1) { + err = 0; + break; + } } - block = cbex.ec_block + cbex.ec_len; + block = newex.ec_block + newex.ec_len; } if (path) { @@ -2156,7 +2194,6 @@ ext4_ext_in_cache(struct inode *inode, ext4_lblk_t block, struct ext4_extent *ex) { struct ext4_ext_cache *cex; - struct ext4_sb_info *sbi; int ret = 0; /* @@ -2164,7 +2201,6 @@ 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) @@ -2273,7 +2309,13 @@ int ext4_ext_calc_credits_for_single_extent(struct inode *inode, int nrblocks, int ext4_ext_index_trans_blocks(struct inode *inode, int nrblocks, int chunk) { int index; - int depth = ext_depth(inode); + int depth; + + /* If we are converting the inline data, only one is needed here. */ + if (ext4_has_inline_data(inode)) + return 1; + + depth = ext_depth(inode); if (chunk) index = depth * 2; @@ -3461,115 +3503,34 @@ out: /** * ext4_find_delalloc_range: find delayed allocated block in the given range. * - * Goes through the buffer heads in the range [lblk_start, lblk_end] and returns - * whether there are any buffers marked for delayed allocation. It returns '1' - * on the first delalloc'ed buffer head found. If no buffer head in the given - * range is marked for delalloc, it returns 0. - * lblk_start should always be <= lblk_end. - * search_hint_reverse is to indicate that searching in reverse from lblk_end to - * lblk_start might be more efficient (i.e., we will likely hit the delalloc'ed - * block sooner). This is useful when blocks are truncated sequentially from - * lblk_start towards lblk_end. + * Return 1 if there is a delalloc block in the range, otherwise 0. */ static int ext4_find_delalloc_range(struct inode *inode, ext4_lblk_t lblk_start, - ext4_lblk_t lblk_end, - int search_hint_reverse) + ext4_lblk_t lblk_end) { - struct address_space *mapping = inode->i_mapping; - struct buffer_head *head, *bh = NULL; - struct page *page; - ext4_lblk_t i, pg_lblk; - pgoff_t index; - - if (!test_opt(inode->i_sb, DELALLOC)) - return 0; - - /* reverse search wont work if fs block size is less than page size */ - if (inode->i_blkbits < PAGE_CACHE_SHIFT) - search_hint_reverse = 0; + struct extent_status es; - if (search_hint_reverse) - i = lblk_end; + es.start = lblk_start; + ext4_es_find_extent(inode, &es); + if (es.len == 0) + return 0; /* there is no delay extent in this tree */ + else if (es.start <= lblk_start && lblk_start < es.start + es.len) + return 1; + else if (lblk_start <= es.start && es.start <= lblk_end) + return 1; else - i = lblk_start; - - index = i >> (PAGE_CACHE_SHIFT - inode->i_blkbits); - - while ((i >= lblk_start) && (i <= lblk_end)) { - page = find_get_page(mapping, index); - if (!page) - goto nextpage; - - if (!page_has_buffers(page)) - goto nextpage; - - head = page_buffers(page); - if (!head) - goto nextpage; - - bh = head; - pg_lblk = index << (PAGE_CACHE_SHIFT - - inode->i_blkbits); - do { - if (unlikely(pg_lblk < lblk_start)) { - /* - * This is possible when fs block size is less - * than page size and our cluster starts/ends in - * middle of the page. So we need to skip the - * initial few blocks till we reach the 'lblk' - */ - pg_lblk++; - continue; - } - - /* Check if the buffer is delayed allocated and that it - * is not yet mapped. (when da-buffers are mapped during - * their writeout, their da_mapped bit is set.) - */ - if (buffer_delay(bh) && !buffer_da_mapped(bh)) { - page_cache_release(page); - trace_ext4_find_delalloc_range(inode, - lblk_start, lblk_end, - search_hint_reverse, - 1, i); - return 1; - } - if (search_hint_reverse) - i--; - else - i++; - } while ((i >= lblk_start) && (i <= lblk_end) && - ((bh = bh->b_this_page) != head)); -nextpage: - if (page) - page_cache_release(page); - /* - * Move to next page. 'i' will be the first lblk in the next - * page. - */ - if (search_hint_reverse) - index--; - else - index++; - i = index << (PAGE_CACHE_SHIFT - inode->i_blkbits); - } - - trace_ext4_find_delalloc_range(inode, lblk_start, lblk_end, - search_hint_reverse, 0, 0); - return 0; + return 0; } -int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk, - int search_hint_reverse) +int ext4_find_delalloc_cluster(struct inode *inode, ext4_lblk_t lblk) { struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); ext4_lblk_t lblk_start, lblk_end; lblk_start = lblk & (~(sbi->s_cluster_ratio - 1)); lblk_end = lblk_start + sbi->s_cluster_ratio - 1; - return ext4_find_delalloc_range(inode, lblk_start, lblk_end, - search_hint_reverse); + return ext4_find_delalloc_range(inode, lblk_start, lblk_end); } /** @@ -3630,7 +3591,7 @@ get_reserved_cluster_alloc(struct inode *inode, ext4_lblk_t lblk_start, lblk_from = lblk_start & (~(sbi->s_cluster_ratio - 1)); lblk_to = lblk_from + c_offset - 1; - if (ext4_find_delalloc_range(inode, lblk_from, lblk_to, 0)) + if (ext4_find_delalloc_range(inode, lblk_from, lblk_to)) allocated_clusters--; } @@ -3640,7 +3601,7 @@ get_reserved_cluster_alloc(struct inode *inode, ext4_lblk_t lblk_start, lblk_from = lblk_start + num_blks; lblk_to = lblk_from + (sbi->s_cluster_ratio - c_offset) - 1; - if (ext4_find_delalloc_range(inode, lblk_from, lblk_to, 0)) + if (ext4_find_delalloc_range(inode, lblk_from, lblk_to)) allocated_clusters--; } @@ -3663,8 +3624,8 @@ ext4_ext_handle_uninitialized_extents(handle_t *handle, struct inode *inode, flags, allocated); ext4_ext_show_leaf(inode, path); - trace_ext4_ext_handle_uninitialized_extents(inode, map, allocated, - newblock); + trace_ext4_ext_handle_uninitialized_extents(inode, map, flags, + allocated, newblock); /* get_block() before submit the IO, split the extent */ if ((flags & EXT4_GET_BLOCKS_PRE_IO)) { @@ -3911,7 +3872,7 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, struct ext4_extent newex, *ex, *ex2; struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); ext4_fsblk_t newblock = 0; - int free_on_err = 0, err = 0, depth, ret; + int free_on_err = 0, err = 0, depth; unsigned int allocated = 0, offset = 0; unsigned int allocated_clusters = 0; struct ext4_allocation_request ar; @@ -3927,7 +3888,7 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, if (ext4_ext_in_cache(inode, map->m_lblk, &newex)) { if (!newex.ee_start_lo && !newex.ee_start_hi) { if ((sbi->s_cluster_ratio > 1) && - ext4_find_delalloc_cluster(inode, map->m_lblk, 0)) + ext4_find_delalloc_cluster(inode, map->m_lblk)) map->m_flags |= EXT4_MAP_FROM_CLUSTER; if ((flags & EXT4_GET_BLOCKS_CREATE) == 0) { @@ -4007,15 +3968,15 @@ int ext4_ext_map_blocks(handle_t *handle, struct inode *inode, ee_len, ee_start); goto out; } - ret = ext4_ext_handle_uninitialized_extents( + allocated = ext4_ext_handle_uninitialized_extents( handle, inode, map, path, flags, allocated, newblock); - return ret; + goto out3; } } if ((sbi->s_cluster_ratio > 1) && - ext4_find_delalloc_cluster(inode, map->m_lblk, 0)) + ext4_find_delalloc_cluster(inode, map->m_lblk)) map->m_flags |= EXT4_MAP_FROM_CLUSTER; /* @@ -4284,8 +4245,8 @@ out2: kfree(path); } - trace_ext4_ext_map_blocks_exit(inode, map->m_lblk, - newblock, map->m_len, err ? err : allocated); +out3: + trace_ext4_ext_map_blocks_exit(inode, map, err ? err : allocated); return err ? err : allocated; } @@ -4344,6 +4305,8 @@ void ext4_ext_truncate(struct inode *inode) last_block = (inode->i_size + sb->s_blocksize - 1) >> EXT4_BLOCK_SIZE_BITS(sb); + err = ext4_es_remove_extent(inode, last_block, + EXT_MAX_BLOCKS - last_block); err = ext4_ext_remove_space(inode, last_block, EXT_MAX_BLOCKS - 1); /* In a multi-transaction truncate, we only make the final @@ -4434,6 +4397,10 @@ long ext4_fallocate(struct file *file, int mode, loff_t offset, loff_t len) if (mode & FALLOC_FL_PUNCH_HOLE) return ext4_punch_hole(file, offset, len); + ret = ext4_convert_inline_data(inode); + if (ret) + return ret; + trace_ext4_fallocate_enter(inode, offset, len, mode); map.m_lblk = offset >> blkbits; /* @@ -4572,206 +4539,43 @@ int ext4_convert_unwritten_extents(struct inode *inode, loff_t offset, } /* - * Callback function called for each extent to gather FIEMAP information. + * If newex is not existing extent (newex->ec_start equals zero) find + * delayed extent at start of newex and update newex accordingly and + * return start of the next delayed extent. + * + * If newex is existing extent (newex->ec_start is not equal zero) + * return start of next delayed extent or EXT_MAX_BLOCKS if no delayed + * extent found. Leave newex unmodified. */ -static int ext4_ext_fiemap_cb(struct inode *inode, ext4_lblk_t next, - struct ext4_ext_cache *newex, struct ext4_extent *ex, - void *data) +static int ext4_find_delayed_extent(struct inode *inode, + struct ext4_ext_cache *newex) { - __u64 logical; - __u64 physical; - __u64 length; - __u32 flags = 0; - int ret = 0; - struct fiemap_extent_info *fieinfo = data; - unsigned char blksize_bits; + struct extent_status es; + ext4_lblk_t next_del; - blksize_bits = inode->i_sb->s_blocksize_bits; - logical = (__u64)newex->ec_block << blksize_bits; + es.start = newex->ec_block; + next_del = ext4_es_find_extent(inode, &es); if (newex->ec_start == 0) { /* * No extent in extent-tree contains block @newex->ec_start, * then the block may stay in 1)a hole or 2)delayed-extent. - * - * Holes or delayed-extents are processed as follows. - * 1. lookup dirty pages with specified range in pagecache. - * If no page is got, then there is no delayed-extent and - * return with EXT_CONTINUE. - * 2. find the 1st mapped buffer, - * 3. check if the mapped buffer is both in the request range - * and a delayed buffer. If not, there is no delayed-extent, - * then return. - * 4. a delayed-extent is found, the extent will be collected. */ - ext4_lblk_t end = 0; - 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; - unsigned int nr_pages = PAGE_SIZE / sizeof(struct page *); - - pages = kmalloc(PAGE_SIZE, GFP_KERNEL); - if (pages == NULL) - return -ENOMEM; - - offset = logical >> PAGE_SHIFT; -repeat: - last_offset = offset; - head = NULL; - ret = find_get_pages_tag(inode->i_mapping, &offset, - PAGECACHE_TAG_DIRTY, nr_pages, pages); - - if (!(flags & FIEMAP_EXTENT_DELALLOC)) { - /* First time, try to find a mapped buffer. */ - if (ret == 0) { -out: - for (index = 0; index < ret; index++) - page_cache_release(pages[index]); - /* just a hole. */ - kfree(pages); - return EXT_CONTINUE; - } - index = 0; - -next_page: - /* Try to find the 1st mapped buffer. */ - end = ((__u64)pages[index]->index << PAGE_SHIFT) >> - blksize_bits; - if (!page_has_buffers(pages[index])) - goto out; - head = page_buffers(pages[index]); - if (!head) - goto out; - - index++; - bh = head; - do { - 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. */ - goto found_mapped_buffer; - } - - bh = bh->b_this_page; - end++; - } while (bh != head); - - /* 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: - if (bh != NULL && buffer_delay(bh)) { - /* 1st or contiguous delayed buffer found. */ - if (!(flags & FIEMAP_EXTENT_DELALLOC)) { - /* - * 1st delayed buffer found, record - * the start of extent. - */ - flags |= FIEMAP_EXTENT_DELALLOC; - newex->ec_block = end; - logical = (__u64)end << blksize_bits; - } - /* Find contiguous delayed buffers. */ - do { - if (!buffer_delay(bh)) - goto found_delayed_extent; - bh = bh->b_this_page; - end++; - } while (bh != head); - - for (; index < ret; index++) { - if (!page_has_buffers(pages[index])) { - bh = NULL; - break; - } - head = page_buffers(pages[index]); - if (!head) { - bh = NULL; - break; - } - - if (pages[index]->index != - pages[start_index]->index + index - - start_index) { - /* Blocks are not contiguous. */ - bh = NULL; - break; - } - bh = head; - do { - if (!buffer_delay(bh)) - /* Delayed-extent ends. */ - goto found_delayed_extent; - bh = bh->b_this_page; - end++; - } while (bh != head); - } - } else if (!(flags & FIEMAP_EXTENT_DELALLOC)) - /* a hole found. */ - goto out; + if (es.len == 0) + /* A hole found. */ + return 0; -found_delayed_extent: - newex->ec_len = min(end - newex->ec_block, - (ext4_lblk_t)EXT_INIT_MAX_LEN); - if (ret == nr_pages && bh != NULL && - newex->ec_len < EXT_INIT_MAX_LEN && - buffer_delay(bh)) { - /* Have not collected an extent and continue. */ - for (index = 0; index < ret; index++) - page_cache_release(pages[index]); - goto repeat; + if (es.start > newex->ec_block) { + /* A hole found. */ + newex->ec_len = min(es.start - newex->ec_block, + newex->ec_len); + return 0; } - for (index = 0; index < ret; index++) - page_cache_release(pages[index]); - kfree(pages); + newex->ec_len = es.start + es.len - newex->ec_block; } - physical = (__u64)newex->ec_start << blksize_bits; - length = (__u64)newex->ec_len << blksize_bits; - - if (ex && ext4_ext_is_uninitialized(ex)) - flags |= FIEMAP_EXTENT_UNWRITTEN; - - if (next == EXT_MAX_BLOCKS) - flags |= FIEMAP_EXTENT_LAST; - - ret = fiemap_fill_next_extent(fieinfo, logical, physical, - length, flags); - if (ret < 0) - return ret; - if (ret == 1) - return EXT_BREAK; - return EXT_CONTINUE; + return next_del; } /* fiemap flags we can handle specified here */ #define EXT4_FIEMAP_FLAGS (FIEMAP_FLAG_SYNC|FIEMAP_FLAG_XATTR) @@ -4971,6 +4775,8 @@ int ext4_ext_punch_hole(struct file *file, loff_t offset, loff_t length) ext4_ext_invalidate_cache(inode); ext4_discard_preallocations(inode); + err = ext4_es_remove_extent(inode, first_block, + stop_block - first_block); err = ext4_ext_remove_space(inode, first_block, stop_block - 1); ext4_ext_invalidate_cache(inode); @@ -4991,12 +4797,22 @@ out_mutex: mutex_unlock(&inode->i_mutex); return err; } + int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, __u64 start, __u64 len) { ext4_lblk_t start_blk; int error = 0; + if (ext4_has_inline_data(inode)) { + int has_inline = 1; + + error = ext4_inline_data_fiemap(inode, fieinfo, &has_inline); + + if (has_inline) + return error; + } + /* fallback to generic here if not in extents fmt */ if (!(ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS))) return generic_block_fiemap(inode, fieinfo, start, len, @@ -5018,11 +4834,11 @@ int ext4_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, len_blks = ((ext4_lblk_t) last_blk) - start_blk + 1; /* - * Walk the extent tree gathering extent information. - * ext4_ext_fiemap_cb will push extents back to user. + * Walk the extent tree gathering extent information + * and pushing extents back to the user. */ - error = ext4_ext_walk_space(inode, start_blk, len_blks, - ext4_ext_fiemap_cb, fieinfo); + error = ext4_fill_fiemap_extents(inode, start_blk, + len_blks, fieinfo); } return error; diff --git a/fs/ext4/extents_status.c b/fs/ext4/extents_status.c new file mode 100644 index 000000000000..564d981a2fcc --- /dev/null +++ b/fs/ext4/extents_status.c @@ -0,0 +1,500 @@ +/* + * fs/ext4/extents_status.c + * + * Written by Yongqiang Yang <xiaoqiangnk@gmail.com> + * Modified by + * Allison Henderson <achender@linux.vnet.ibm.com> + * Hugh Dickins <hughd@google.com> + * Zheng Liu <wenqing.lz@taobao.com> + * + * Ext4 extents status tree core functions. + */ +#include <linux/rbtree.h> +#include "ext4.h" +#include "extents_status.h" +#include "ext4_extents.h" + +#include <trace/events/ext4.h> + +/* + * According to previous discussion in Ext4 Developer Workshop, we + * will introduce a new structure called io tree to track all extent + * status in order to solve some problems that we have met + * (e.g. Reservation space warning), and provide extent-level locking. + * Delay extent tree is the first step to achieve this goal. It is + * original built by Yongqiang Yang. At that time it is called delay + * extent tree, whose goal is only track delay extent in memory to + * simplify the implementation of fiemap and bigalloc, and introduce + * lseek SEEK_DATA/SEEK_HOLE support. That is why it is still called + * delay extent tree at the following comment. But for better + * understand what it does, it has been rename to extent status tree. + * + * Currently the first step has been done. All delay extents are + * tracked in the tree. It maintains the delay extent when a delay + * allocation is issued, and the delay extent is written out or + * invalidated. Therefore the implementation of fiemap and bigalloc + * are simplified, and SEEK_DATA/SEEK_HOLE are introduced. + * + * The following comment describes the implemenmtation of extent + * status tree and future works. + */ + +/* + * extents status tree implementation for ext4. + * + * + * ========================================================================== + * Extents status encompass delayed extents and extent locks + * + * 1. Why delayed extent implementation ? + * + * Without delayed extent, ext4 identifies a delayed extent by looking + * up page cache, this has several deficiencies - complicated, buggy, + * and inefficient code. + * + * FIEMAP, SEEK_HOLE/DATA, bigalloc, punch hole and writeout all need + * to know if a block or a range of blocks are belonged to a delayed + * extent. + * + * Let us have a look at how they do without delayed extents implementation. + * -- FIEMAP + * FIEMAP looks up page cache to identify delayed allocations from holes. + * + * -- SEEK_HOLE/DATA + * SEEK_HOLE/DATA has the same problem as FIEMAP. + * + * -- bigalloc + * bigalloc looks up page cache to figure out if a block is + * already under delayed allocation or not to determine whether + * quota reserving is needed for the cluster. + * + * -- punch hole + * punch hole looks up page cache to identify a delayed extent. + * + * -- writeout + * Writeout looks up whole page cache to see if a buffer is + * mapped, If there are not very many delayed buffers, then it is + * time comsuming. + * + * With delayed extents implementation, FIEMAP, SEEK_HOLE/DATA, + * bigalloc and writeout can figure out if a block or a range of + * blocks is under delayed allocation(belonged to a delayed extent) or + * not by searching the delayed extent tree. + * + * + * ========================================================================== + * 2. ext4 delayed extents impelmentation + * + * -- delayed extent + * A delayed extent is a range of blocks which are contiguous + * logically and under delayed allocation. Unlike extent in + * ext4, delayed extent in ext4 is a in-memory struct, there is + * no corresponding on-disk data. There is no limit on length of + * delayed extent, so a delayed extent can contain as many blocks + * as they are contiguous logically. + * + * -- delayed extent tree + * Every inode has a delayed extent tree and all under delayed + * allocation blocks are added to the tree as delayed extents. + * Delayed extents in the tree are ordered by logical block no. + * + * -- operations on a delayed extent tree + * There are three operations on a delayed extent tree: find next + * delayed extent, adding a space(a range of blocks) and removing + * a space. + * + * -- race on a delayed extent tree + * Delayed extent tree is protected inode->i_es_lock. + * + * + * ========================================================================== + * 3. performance analysis + * -- overhead + * 1. There is a cache extent for write access, so if writes are + * not very random, adding space operaions are in O(1) time. + * + * -- gain + * 2. Code is much simpler, more readable, more maintainable and + * more efficient. + * + * + * ========================================================================== + * 4. TODO list + * -- Track all extent status + * + * -- Improve get block process + * + * -- Extent-level locking + */ + +static struct kmem_cache *ext4_es_cachep; + +int __init ext4_init_es(void) +{ + ext4_es_cachep = KMEM_CACHE(extent_status, SLAB_RECLAIM_ACCOUNT); + if (ext4_es_cachep == NULL) + return -ENOMEM; + return 0; +} + +void ext4_exit_es(void) +{ + if (ext4_es_cachep) + kmem_cache_destroy(ext4_es_cachep); +} + +void ext4_es_init_tree(struct ext4_es_tree *tree) +{ + tree->root = RB_ROOT; + tree->cache_es = NULL; +} + +#ifdef ES_DEBUG__ +static void ext4_es_print_tree(struct inode *inode) +{ + struct ext4_es_tree *tree; + struct rb_node *node; + + printk(KERN_DEBUG "status extents for inode %lu:", inode->i_ino); + tree = &EXT4_I(inode)->i_es_tree; + node = rb_first(&tree->root); + while (node) { + struct extent_status *es; + es = rb_entry(node, struct extent_status, rb_node); + printk(KERN_DEBUG " [%u/%u)", es->start, es->len); + node = rb_next(node); + } + printk(KERN_DEBUG "\n"); +} +#else +#define ext4_es_print_tree(inode) +#endif + +static inline ext4_lblk_t extent_status_end(struct extent_status *es) +{ + BUG_ON(es->start + es->len < es->start); + return es->start + es->len - 1; +} + +/* + * search through the tree for an delayed extent with a given offset. If + * it can't be found, try to find next extent. + */ +static struct extent_status *__es_tree_search(struct rb_root *root, + ext4_lblk_t offset) +{ + struct rb_node *node = root->rb_node; + struct extent_status *es = NULL; + + while (node) { + es = rb_entry(node, struct extent_status, rb_node); + if (offset < es->start) + node = node->rb_left; + else if (offset > extent_status_end(es)) + node = node->rb_right; + else + return es; + } + + if (es && offset < es->start) + return es; + + if (es && offset > extent_status_end(es)) { + node = rb_next(&es->rb_node); + return node ? rb_entry(node, struct extent_status, rb_node) : + NULL; + } + + return NULL; +} + +/* + * ext4_es_find_extent: find the 1st delayed extent covering @es->start + * if it exists, otherwise, the next extent after @es->start. + * + * @inode: the inode which owns delayed extents + * @es: delayed extent that we found + * + * Returns the first block of the next extent after es, otherwise + * EXT_MAX_BLOCKS if no delay extent is found. + * Delayed extent is returned via @es. + */ +ext4_lblk_t ext4_es_find_extent(struct inode *inode, struct extent_status *es) +{ + struct ext4_es_tree *tree = NULL; + struct extent_status *es1 = NULL; + struct rb_node *node; + ext4_lblk_t ret = EXT_MAX_BLOCKS; + + trace_ext4_es_find_extent_enter(inode, es->start); + + read_lock(&EXT4_I(inode)->i_es_lock); + tree = &EXT4_I(inode)->i_es_tree; + + /* find delay extent in cache firstly */ + if (tree->cache_es) { + es1 = tree->cache_es; + if (in_range(es->start, es1->start, es1->len)) { + es_debug("%u cached by [%u/%u)\n", + es->start, es1->start, es1->len); + goto out; + } + } + + es->len = 0; + es1 = __es_tree_search(&tree->root, es->start); + +out: + if (es1) { + tree->cache_es = es1; + es->start = es1->start; + es->len = es1->len; + node = rb_next(&es1->rb_node); + if (node) { + es1 = rb_entry(node, struct extent_status, rb_node); + ret = es1->start; + } + } + + read_unlock(&EXT4_I(inode)->i_es_lock); + + trace_ext4_es_find_extent_exit(inode, es, ret); + return ret; +} + +static struct extent_status * +ext4_es_alloc_extent(ext4_lblk_t start, ext4_lblk_t len) +{ + struct extent_status *es; + es = kmem_cache_alloc(ext4_es_cachep, GFP_ATOMIC); + if (es == NULL) + return NULL; + es->start = start; + es->len = len; + return es; +} + +static void ext4_es_free_extent(struct extent_status *es) +{ + kmem_cache_free(ext4_es_cachep, es); +} + +static struct extent_status * +ext4_es_try_to_merge_left(struct ext4_es_tree *tree, struct extent_status *es) +{ + struct extent_status *es1; + struct rb_node *node; + + node = rb_prev(&es->rb_node); + if (!node) + return es; + + es1 = rb_entry(node, struct extent_status, rb_node); + if (es->start == extent_status_end(es1) + 1) { + es1->len += es->len; + rb_erase(&es->rb_node, &tree->root); + ext4_es_free_extent(es); + es = es1; + } + + return es; +} + +static struct extent_status * +ext4_es_try_to_merge_right(struct ext4_es_tree *tree, struct extent_status *es) +{ + struct extent_status *es1; + struct rb_node *node; + + node = rb_next(&es->rb_node); + if (!node) + return es; + + es1 = rb_entry(node, struct extent_status, rb_node); + if (es1->start == extent_status_end(es) + 1) { + es->len += es1->len; + rb_erase(node, &tree->root); + ext4_es_free_extent(es1); + } + + return es; +} + +static int __es_insert_extent(struct ext4_es_tree *tree, ext4_lblk_t offset, + ext4_lblk_t len) +{ + struct rb_node **p = &tree->root.rb_node; + struct rb_node *parent = NULL; + struct extent_status *es; + ext4_lblk_t end = offset + len - 1; + + BUG_ON(end < offset); + es = tree->cache_es; + if (es && offset == (extent_status_end(es) + 1)) { + es_debug("cached by [%u/%u)\n", es->start, es->len); + es->len += len; + es = ext4_es_try_to_merge_right(tree, es); + goto out; + } else if (es && es->start == end + 1) { + es_debug("cached by [%u/%u)\n", es->start, es->len); + es->start = offset; + es->len += len; + es = ext4_es_try_to_merge_left(tree, es); + goto out; + } else if (es && es->start <= offset && + end <= extent_status_end(es)) { + es_debug("cached by [%u/%u)\n", es->start, es->len); + goto out; + } + + while (*p) { + parent = *p; + es = rb_entry(parent, struct extent_status, rb_node); + + if (offset < es->start) { + if (es->start == end + 1) { + es->start = offset; + es->len += len; + es = ext4_es_try_to_merge_left(tree, es); + goto out; + } + p = &(*p)->rb_left; + } else if (offset > extent_status_end(es)) { + if (offset == extent_status_end(es) + 1) { + es->len += len; + es = ext4_es_try_to_merge_right(tree, es); + goto out; + } + p = &(*p)->rb_right; + } else { + if (extent_status_end(es) <= end) + es->len = offset - es->start + len; + goto out; + } + } + + es = ext4_es_alloc_extent(offset, len); + if (!es) + return -ENOMEM; + rb_link_node(&es->rb_node, parent, p); + rb_insert_color(&es->rb_node, &tree->root); + +out: + tree->cache_es = es; + return 0; +} + +/* + * ext4_es_insert_extent() adds a space to a delayed extent tree. + * Caller holds inode->i_es_lock. + * + * ext4_es_insert_extent is called by ext4_da_write_begin and + * ext4_es_remove_extent. + * + * Return 0 on success, error code on failure. + */ +int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t offset, + ext4_lblk_t len) +{ + struct ext4_es_tree *tree; + int err = 0; + + trace_ext4_es_insert_extent(inode, offset, len); + es_debug("add [%u/%u) to extent status tree of inode %lu\n", + offset, len, inode->i_ino); + + write_lock(&EXT4_I(inode)->i_es_lock); + tree = &EXT4_I(inode)->i_es_tree; + err = __es_insert_extent(tree, offset, len); + write_unlock(&EXT4_I(inode)->i_es_lock); + + ext4_es_print_tree(inode); + + return err; +} + +/* + * ext4_es_remove_extent() removes a space from a delayed extent tree. + * Caller holds inode->i_es_lock. + * + * Return 0 on success, error code on failure. + */ +int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t offset, + ext4_lblk_t len) +{ + struct rb_node *node; + struct ext4_es_tree *tree; + struct extent_status *es; + struct extent_status orig_es; + ext4_lblk_t len1, len2, end; + int err = 0; + + trace_ext4_es_remove_extent(inode, offset, len); + es_debug("remove [%u/%u) from extent status tree of inode %lu\n", + offset, len, inode->i_ino); + + end = offset + len - 1; + BUG_ON(end < offset); + write_lock(&EXT4_I(inode)->i_es_lock); + tree = &EXT4_I(inode)->i_es_tree; + es = __es_tree_search(&tree->root, offset); + if (!es) + goto out; + if (es->start > end) + goto out; + + /* Simply invalidate cache_es. */ + tree->cache_es = NULL; + + orig_es.start = es->start; + orig_es.len = es->len; + len1 = offset > es->start ? offset - es->start : 0; + len2 = extent_status_end(es) > end ? + extent_status_end(es) - end : 0; + if (len1 > 0) + es->len = len1; + if (len2 > 0) { + if (len1 > 0) { + err = __es_insert_extent(tree, end + 1, len2); + if (err) { + es->start = orig_es.start; + es->len = orig_es.len; + goto out; + } + } else { + es->start = end + 1; + es->len = len2; + } + goto out; + } + + if (len1 > 0) { + node = rb_next(&es->rb_node); + if (node) + es = rb_entry(node, struct extent_status, rb_node); + else + es = NULL; + } + + while (es && extent_status_end(es) <= end) { + node = rb_next(&es->rb_node); + rb_erase(&es->rb_node, &tree->root); + ext4_es_free_extent(es); + if (!node) { + es = NULL; + break; + } + es = rb_entry(node, struct extent_status, rb_node); + } + + if (es && es->start < end + 1) { + len1 = extent_status_end(es) - end; + es->start = end + 1; + es->len = len1; + } + +out: + write_unlock(&EXT4_I(inode)->i_es_lock); + ext4_es_print_tree(inode); + return err; +} diff --git a/fs/ext4/extents_status.h b/fs/ext4/extents_status.h new file mode 100644 index 000000000000..077f82db092a --- /dev/null +++ b/fs/ext4/extents_status.h @@ -0,0 +1,45 @@ +/* + * fs/ext4/extents_status.h + * + * Written by Yongqiang Yang <xiaoqiangnk@gmail.com> + * Modified by + * Allison Henderson <achender@linux.vnet.ibm.com> + * Zheng Liu <wenqing.lz@taobao.com> + * + */ + +#ifndef _EXT4_EXTENTS_STATUS_H +#define _EXT4_EXTENTS_STATUS_H + +/* + * Turn on ES_DEBUG__ to get lots of info about extent status operations. + */ +#ifdef ES_DEBUG__ +#define es_debug(fmt, ...) printk(fmt, ##__VA_ARGS__) +#else +#define es_debug(fmt, ...) no_printk(fmt, ##__VA_ARGS__) +#endif + +struct extent_status { + struct rb_node rb_node; + ext4_lblk_t start; /* first block extent covers */ + ext4_lblk_t len; /* length of extent in block */ +}; + +struct ext4_es_tree { + struct rb_root root; + struct extent_status *cache_es; /* recently accessed extent */ +}; + +extern int __init ext4_init_es(void); +extern void ext4_exit_es(void); +extern void ext4_es_init_tree(struct ext4_es_tree *tree); + +extern int ext4_es_insert_extent(struct inode *inode, ext4_lblk_t start, + ext4_lblk_t len); +extern int ext4_es_remove_extent(struct inode *inode, ext4_lblk_t start, + ext4_lblk_t len); +extern ext4_lblk_t ext4_es_find_extent(struct inode *inode, + struct extent_status *es); + +#endif /* _EXT4_EXTENTS_STATUS_H */ diff --git a/fs/ext4/file.c b/fs/ext4/file.c index bf3966bccd34..b64a60bf105a 100644 --- a/fs/ext4/file.c +++ b/fs/ext4/file.c @@ -24,6 +24,7 @@ #include <linux/mount.h> #include <linux/path.h> #include <linux/quotaops.h> +#include <linux/pagevec.h> #include "ext4.h" #include "ext4_jbd2.h" #include "xattr.h" @@ -286,6 +287,324 @@ static int ext4_file_open(struct inode * inode, struct file * filp) } /* + * Here we use ext4_map_blocks() to get a block mapping for a extent-based + * file rather than ext4_ext_walk_space() because we can introduce + * SEEK_DATA/SEEK_HOLE for block-mapped and extent-mapped file at the same + * function. When extent status tree has been fully implemented, it will + * track all extent status for a file and we can directly use it to + * retrieve the offset for SEEK_DATA/SEEK_HOLE. + */ + +/* + * When we retrieve the offset for SEEK_DATA/SEEK_HOLE, we would need to + * lookup page cache to check whether or not there has some data between + * [startoff, endoff] because, if this range contains an unwritten extent, + * we determine this extent as a data or a hole according to whether the + * page cache has data or not. + */ +static int ext4_find_unwritten_pgoff(struct inode *inode, + int origin, + struct ext4_map_blocks *map, + loff_t *offset) +{ + struct pagevec pvec; + unsigned int blkbits; + pgoff_t index; + pgoff_t end; + loff_t endoff; + loff_t startoff; + loff_t lastoff; + int found = 0; + + blkbits = inode->i_sb->s_blocksize_bits; + startoff = *offset; + lastoff = startoff; + endoff = (map->m_lblk + map->m_len) << blkbits; + + index = startoff >> PAGE_CACHE_SHIFT; + end = endoff >> PAGE_CACHE_SHIFT; + + pagevec_init(&pvec, 0); + do { + int i, num; + unsigned long nr_pages; + + num = min_t(pgoff_t, end - index, PAGEVEC_SIZE); + nr_pages = pagevec_lookup(&pvec, inode->i_mapping, index, + (pgoff_t)num); + if (nr_pages == 0) { + if (origin == SEEK_DATA) + break; + + BUG_ON(origin != SEEK_HOLE); + /* + * If this is the first time to go into the loop and + * offset is not beyond the end offset, it will be a + * hole at this offset + */ + if (lastoff == startoff || lastoff < endoff) + found = 1; + break; + } + + /* + * If this is the first time to go into the loop and + * offset is smaller than the first page offset, it will be a + * hole at this offset. + */ + if (lastoff == startoff && origin == SEEK_HOLE && + lastoff < page_offset(pvec.pages[0])) { + found = 1; + break; + } + + for (i = 0; i < nr_pages; i++) { + struct page *page = pvec.pages[i]; + struct buffer_head *bh, *head; + + /* + * If the current offset is not beyond the end of given + * range, it will be a hole. + */ + if (lastoff < endoff && origin == SEEK_HOLE && + page->index > end) { + found = 1; + *offset = lastoff; + goto out; + } + + lock_page(page); + + if (unlikely(page->mapping != inode->i_mapping)) { + unlock_page(page); + continue; + } + + if (!page_has_buffers(page)) { + unlock_page(page); + continue; + } + + if (page_has_buffers(page)) { + lastoff = page_offset(page); + bh = head = page_buffers(page); + do { + if (buffer_uptodate(bh) || + buffer_unwritten(bh)) { + if (origin == SEEK_DATA) + found = 1; + } else { + if (origin == SEEK_HOLE) + found = 1; + } + if (found) { + *offset = max_t(loff_t, + startoff, lastoff); + unlock_page(page); + goto out; + } + lastoff += bh->b_size; + bh = bh->b_this_page; + } while (bh != head); + } + + lastoff = page_offset(page) + PAGE_SIZE; + unlock_page(page); + } + + /* + * The no. of pages is less than our desired, that would be a + * hole in there. + */ + if (nr_pages < num && origin == SEEK_HOLE) { + found = 1; + *offset = lastoff; + break; + } + + index = pvec.pages[i - 1]->index + 1; + pagevec_release(&pvec); + } while (index <= end); + +out: + pagevec_release(&pvec); + return found; +} + +/* + * ext4_seek_data() retrieves the offset for SEEK_DATA. + */ +static loff_t ext4_seek_data(struct file *file, loff_t offset, loff_t maxsize) +{ + struct inode *inode = file->f_mapping->host; + struct ext4_map_blocks map; + struct extent_status es; + ext4_lblk_t start, last, end; + loff_t dataoff, isize; + int blkbits; + int ret = 0; + + mutex_lock(&inode->i_mutex); + + isize = i_size_read(inode); + if (offset >= isize) { + mutex_unlock(&inode->i_mutex); + return -ENXIO; + } + + blkbits = inode->i_sb->s_blocksize_bits; + start = offset >> blkbits; + last = start; + end = isize >> blkbits; + dataoff = offset; + + do { + map.m_lblk = last; + map.m_len = end - last + 1; + ret = ext4_map_blocks(NULL, inode, &map, 0); + if (ret > 0 && !(map.m_flags & EXT4_MAP_UNWRITTEN)) { + if (last != start) + dataoff = last << blkbits; + break; + } + + /* + * If there is a delay extent at this offset, + * it will be as a data. + */ + es.start = last; + (void)ext4_es_find_extent(inode, &es); + if (last >= es.start && + last < es.start + es.len) { + if (last != start) + dataoff = last << blkbits; + break; + } + + /* + * If there is a unwritten extent at this offset, + * it will be as a data or a hole according to page + * cache that has data or not. + */ + if (map.m_flags & EXT4_MAP_UNWRITTEN) { + int unwritten; + unwritten = ext4_find_unwritten_pgoff(inode, SEEK_DATA, + &map, &dataoff); + if (unwritten) + break; + } + + last++; + dataoff = last << blkbits; + } while (last <= end); + + mutex_unlock(&inode->i_mutex); + + if (dataoff > isize) + return -ENXIO; + + if (dataoff < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) + return -EINVAL; + if (dataoff > maxsize) + return -EINVAL; + + if (dataoff != file->f_pos) { + file->f_pos = dataoff; + file->f_version = 0; + } + + return dataoff; +} + +/* + * ext4_seek_hole() retrieves the offset for SEEK_HOLE. + */ +static loff_t ext4_seek_hole(struct file *file, loff_t offset, loff_t maxsize) +{ + struct inode *inode = file->f_mapping->host; + struct ext4_map_blocks map; + struct extent_status es; + ext4_lblk_t start, last, end; + loff_t holeoff, isize; + int blkbits; + int ret = 0; + + mutex_lock(&inode->i_mutex); + + isize = i_size_read(inode); + if (offset >= isize) { + mutex_unlock(&inode->i_mutex); + return -ENXIO; + } + + blkbits = inode->i_sb->s_blocksize_bits; + start = offset >> blkbits; + last = start; + end = isize >> blkbits; + holeoff = offset; + + do { + map.m_lblk = last; + map.m_len = end - last + 1; + ret = ext4_map_blocks(NULL, inode, &map, 0); + if (ret > 0 && !(map.m_flags & EXT4_MAP_UNWRITTEN)) { + last += ret; + holeoff = last << blkbits; + continue; + } + + /* + * If there is a delay extent at this offset, + * we will skip this extent. + */ + es.start = last; + (void)ext4_es_find_extent(inode, &es); + if (last >= es.start && + last < es.start + es.len) { + last = es.start + es.len; + holeoff = last << blkbits; + continue; + } + + /* + * If there is a unwritten extent at this offset, + * it will be as a data or a hole according to page + * cache that has data or not. + */ + if (map.m_flags & EXT4_MAP_UNWRITTEN) { + int unwritten; + unwritten = ext4_find_unwritten_pgoff(inode, SEEK_HOLE, + &map, &holeoff); + if (!unwritten) { + last += ret; + holeoff = last << blkbits; + continue; + } + } + + /* find a hole */ + break; + } while (last <= end); + + mutex_unlock(&inode->i_mutex); + + if (holeoff > isize) + holeoff = isize; + + if (holeoff < 0 && !(file->f_mode & FMODE_UNSIGNED_OFFSET)) + return -EINVAL; + if (holeoff > maxsize) + return -EINVAL; + + if (holeoff != file->f_pos) { + file->f_pos = holeoff; + file->f_version = 0; + } + + return holeoff; +} + +/* * ext4_llseek() handles both block-mapped and extent-mapped maxbytes values * by calling generic_file_llseek_size() with the appropriate maxbytes * value for each. @@ -300,8 +619,19 @@ loff_t ext4_llseek(struct file *file, loff_t offset, int origin) else maxbytes = inode->i_sb->s_maxbytes; - return generic_file_llseek_size(file, offset, origin, - maxbytes, i_size_read(inode)); + switch (origin) { + case SEEK_SET: + case SEEK_CUR: + case SEEK_END: + return generic_file_llseek_size(file, offset, origin, + maxbytes, i_size_read(inode)); + case SEEK_DATA: + return ext4_seek_data(file, offset, maxbytes); + case SEEK_HOLE: + return ext4_seek_hole(file, offset, maxbytes); + } + + return -EINVAL; } const struct file_operations ext4_file_operations = { @@ -326,12 +656,10 @@ const struct file_operations ext4_file_operations = { const struct inode_operations ext4_file_inode_operations = { .setattr = ext4_setattr, .getattr = ext4_getattr, -#ifdef CONFIG_EXT4_FS_XATTR .setxattr = generic_setxattr, .getxattr = generic_getxattr, .listxattr = ext4_listxattr, .removexattr = generic_removexattr, -#endif .get_acl = ext4_get_acl, .fiemap = ext4_fiemap, }; diff --git a/fs/ext4/fsync.c b/fs/ext4/fsync.c index be1d89f385b4..dfbc1fe96674 100644 --- a/fs/ext4/fsync.c +++ b/fs/ext4/fsync.c @@ -44,7 +44,6 @@ */ static int ext4_sync_parent(struct inode *inode) { - struct writeback_control wbc; struct dentry *dentry = NULL; struct inode *next; int ret = 0; @@ -66,10 +65,7 @@ static int ext4_sync_parent(struct inode *inode) 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); + ret = sync_inode_metadata(inode, 1); if (ret) break; } diff --git a/fs/ext4/ialloc.c b/fs/ext4/ialloc.c index 3a100e7a62a8..3f32c8012447 100644 --- a/fs/ext4/ialloc.c +++ b/fs/ext4/ialloc.c @@ -762,7 +762,6 @@ got: BUFFER_TRACE(block_bitmap_bh, "dirty block bitmap"); err = ext4_handle_dirty_metadata(handle, NULL, block_bitmap_bh); - brelse(block_bitmap_bh); /* recheck and clear flag under lock if we still need to */ ext4_lock_group(sb, group); @@ -775,6 +774,7 @@ got: ext4_group_desc_csum_set(sb, group, gdp); } ext4_unlock_group(sb, group); + brelse(block_bitmap_bh); if (err) goto fail; @@ -902,6 +902,10 @@ got: ei->i_extra_isize = EXT4_SB(sb)->s_want_extra_isize; + ei->i_inline_off = 0; + if (EXT4_HAS_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_INLINE_DATA)) + ext4_set_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + ret = inode; dquot_initialize(inode); err = dquot_alloc_inode(inode); diff --git a/fs/ext4/indirect.c b/fs/ext4/indirect.c index 792e388e7b44..20862f96e8ae 100644 --- a/fs/ext4/indirect.c +++ b/fs/ext4/indirect.c @@ -22,6 +22,7 @@ #include "ext4_jbd2.h" #include "truncate.h" +#include "ext4_extents.h" /* Needed for EXT_MAX_BLOCKS */ #include <trace/events/ext4.h> @@ -755,8 +756,7 @@ cleanup: partial--; } out: - trace_ext4_ind_map_blocks_exit(inode, map->m_lblk, - map->m_pblk, map->m_len, err); + trace_ext4_ind_map_blocks_exit(inode, map, err); return err; } @@ -1412,6 +1412,7 @@ void ext4_ind_truncate(struct inode *inode) down_write(&ei->i_data_sem); ext4_discard_preallocations(inode); + ext4_es_remove_extent(inode, last_block, EXT_MAX_BLOCKS - last_block); /* * The orphan list entry will now protect us from any crash which diff --git a/fs/ext4/inline.c b/fs/ext4/inline.c new file mode 100644 index 000000000000..387c47c6cda9 --- /dev/null +++ b/fs/ext4/inline.c @@ -0,0 +1,1884 @@ +/* + * Copyright (c) 2012 Taobao. + * Written by Tao Ma <boyu.mt@taobao.com> + * + * This program is free software; you can redistribute it and/or modify it + * under the terms of version 2.1 of the GNU Lesser General Public License + * 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 "ext4_jbd2.h" +#include "ext4.h" +#include "xattr.h" +#include "truncate.h" +#include <linux/fiemap.h> + +#define EXT4_XATTR_SYSTEM_DATA "data" +#define EXT4_MIN_INLINE_DATA_SIZE ((sizeof(__le32) * EXT4_N_BLOCKS)) +#define EXT4_INLINE_DOTDOT_SIZE 4 + +int ext4_get_inline_size(struct inode *inode) +{ + if (EXT4_I(inode)->i_inline_off) + return EXT4_I(inode)->i_inline_size; + + return 0; +} + +static int get_max_inline_xattr_value_size(struct inode *inode, + struct ext4_iloc *iloc) +{ + struct ext4_xattr_ibody_header *header; + struct ext4_xattr_entry *entry; + struct ext4_inode *raw_inode; + int free, min_offs; + + min_offs = EXT4_SB(inode->i_sb)->s_inode_size - + EXT4_GOOD_OLD_INODE_SIZE - + EXT4_I(inode)->i_extra_isize - + sizeof(struct ext4_xattr_ibody_header); + + /* + * We need to subtract another sizeof(__u32) since an in-inode xattr + * needs an empty 4 bytes to indicate the gap between the xattr entry + * and the name/value pair. + */ + if (!ext4_test_inode_state(inode, EXT4_STATE_XATTR)) + return EXT4_XATTR_SIZE(min_offs - + EXT4_XATTR_LEN(strlen(EXT4_XATTR_SYSTEM_DATA)) - + EXT4_XATTR_ROUND - sizeof(__u32)); + + raw_inode = ext4_raw_inode(iloc); + header = IHDR(inode, raw_inode); + entry = IFIRST(header); + + /* Compute min_offs. */ + for (; !IS_LAST_ENTRY(entry); entry = EXT4_XATTR_NEXT(entry)) { + if (!entry->e_value_block && entry->e_value_size) { + size_t offs = le16_to_cpu(entry->e_value_offs); + if (offs < min_offs) + min_offs = offs; + } + } + free = min_offs - + ((void *)entry - (void *)IFIRST(header)) - sizeof(__u32); + + if (EXT4_I(inode)->i_inline_off) { + entry = (struct ext4_xattr_entry *) + ((void *)raw_inode + EXT4_I(inode)->i_inline_off); + + free += le32_to_cpu(entry->e_value_size); + goto out; + } + + free -= EXT4_XATTR_LEN(strlen(EXT4_XATTR_SYSTEM_DATA)); + + if (free > EXT4_XATTR_ROUND) + free = EXT4_XATTR_SIZE(free - EXT4_XATTR_ROUND); + else + free = 0; + +out: + return free; +} + +/* + * Get the maximum size we now can store in an inode. + * If we can't find the space for a xattr entry, don't use the space + * of the extents since we have no space to indicate the inline data. + */ +int ext4_get_max_inline_size(struct inode *inode) +{ + int error, max_inline_size; + struct ext4_iloc iloc; + + if (EXT4_I(inode)->i_extra_isize == 0) + return 0; + + error = ext4_get_inode_loc(inode, &iloc); + if (error) { + ext4_error_inode(inode, __func__, __LINE__, 0, + "can't get inode location %lu", + inode->i_ino); + return 0; + } + + down_read(&EXT4_I(inode)->xattr_sem); + max_inline_size = get_max_inline_xattr_value_size(inode, &iloc); + up_read(&EXT4_I(inode)->xattr_sem); + + brelse(iloc.bh); + + if (!max_inline_size) + return 0; + + return max_inline_size + EXT4_MIN_INLINE_DATA_SIZE; +} + +int ext4_has_inline_data(struct inode *inode) +{ + return ext4_test_inode_flag(inode, EXT4_INODE_INLINE_DATA) && + EXT4_I(inode)->i_inline_off; +} + +/* + * this function does not take xattr_sem, which is OK because it is + * currently only used in a code path coming form ext4_iget, before + * the new inode has been unlocked + */ +int ext4_find_inline_data_nolock(struct inode *inode) +{ + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + int error; + + if (EXT4_I(inode)->i_extra_isize == 0) + return 0; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + if (!is.s.not_found) { + EXT4_I(inode)->i_inline_off = (u16)((void *)is.s.here - + (void *)ext4_raw_inode(&is.iloc)); + EXT4_I(inode)->i_inline_size = EXT4_MIN_INLINE_DATA_SIZE + + le32_to_cpu(is.s.here->e_value_size); + ext4_set_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + } +out: + brelse(is.iloc.bh); + return error; +} + +static int ext4_read_inline_data(struct inode *inode, void *buffer, + unsigned int len, + struct ext4_iloc *iloc) +{ + struct ext4_xattr_entry *entry; + struct ext4_xattr_ibody_header *header; + int cp_len = 0; + struct ext4_inode *raw_inode; + + if (!len) + return 0; + + BUG_ON(len > EXT4_I(inode)->i_inline_size); + + cp_len = len < EXT4_MIN_INLINE_DATA_SIZE ? + len : EXT4_MIN_INLINE_DATA_SIZE; + + raw_inode = ext4_raw_inode(iloc); + memcpy(buffer, (void *)(raw_inode->i_block), cp_len); + + len -= cp_len; + buffer += cp_len; + + if (!len) + goto out; + + header = IHDR(inode, raw_inode); + entry = (struct ext4_xattr_entry *)((void *)raw_inode + + EXT4_I(inode)->i_inline_off); + len = min_t(unsigned int, len, + (unsigned int)le32_to_cpu(entry->e_value_size)); + + memcpy(buffer, + (void *)IFIRST(header) + le16_to_cpu(entry->e_value_offs), len); + cp_len += len; + +out: + return cp_len; +} + +/* + * write the buffer to the inline inode. + * If 'create' is set, we don't need to do the extra copy in the xattr + * value since it is already handled by ext4_xattr_ibody_inline_set. + * That saves us one memcpy. + */ +void ext4_write_inline_data(struct inode *inode, struct ext4_iloc *iloc, + void *buffer, loff_t pos, unsigned int len) +{ + struct ext4_xattr_entry *entry; + struct ext4_xattr_ibody_header *header; + struct ext4_inode *raw_inode; + int cp_len = 0; + + BUG_ON(!EXT4_I(inode)->i_inline_off); + BUG_ON(pos + len > EXT4_I(inode)->i_inline_size); + + raw_inode = ext4_raw_inode(iloc); + buffer += pos; + + if (pos < EXT4_MIN_INLINE_DATA_SIZE) { + cp_len = pos + len > EXT4_MIN_INLINE_DATA_SIZE ? + EXT4_MIN_INLINE_DATA_SIZE - pos : len; + memcpy((void *)raw_inode->i_block + pos, buffer, cp_len); + + len -= cp_len; + buffer += cp_len; + pos += cp_len; + } + + if (!len) + return; + + pos -= EXT4_MIN_INLINE_DATA_SIZE; + header = IHDR(inode, raw_inode); + entry = (struct ext4_xattr_entry *)((void *)raw_inode + + EXT4_I(inode)->i_inline_off); + + memcpy((void *)IFIRST(header) + le16_to_cpu(entry->e_value_offs) + pos, + buffer, len); +} + +static int ext4_create_inline_data(handle_t *handle, + struct inode *inode, unsigned len) +{ + int error; + void *value = NULL; + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + error = ext4_journal_get_write_access(handle, is.iloc.bh); + if (error) + goto out; + + if (len > EXT4_MIN_INLINE_DATA_SIZE) { + value = EXT4_ZERO_XATTR_VALUE; + len -= EXT4_MIN_INLINE_DATA_SIZE; + } else { + value = ""; + len = 0; + } + + /* Insert the the xttr entry. */ + i.value = value; + i.value_len = len; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + BUG_ON(!is.s.not_found); + + error = ext4_xattr_ibody_inline_set(handle, inode, &i, &is); + if (error) { + if (error == -ENOSPC) + ext4_clear_inode_state(inode, + EXT4_STATE_MAY_INLINE_DATA); + goto out; + } + + memset((void *)ext4_raw_inode(&is.iloc)->i_block, + 0, EXT4_MIN_INLINE_DATA_SIZE); + + EXT4_I(inode)->i_inline_off = (u16)((void *)is.s.here - + (void *)ext4_raw_inode(&is.iloc)); + EXT4_I(inode)->i_inline_size = len + EXT4_MIN_INLINE_DATA_SIZE; + ext4_clear_inode_flag(inode, EXT4_INODE_EXTENTS); + ext4_set_inode_flag(inode, EXT4_INODE_INLINE_DATA); + get_bh(is.iloc.bh); + error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); + +out: + brelse(is.iloc.bh); + return error; +} + +static int ext4_update_inline_data(handle_t *handle, struct inode *inode, + unsigned int len) +{ + int error; + void *value = NULL; + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + + /* If the old space is ok, write the data directly. */ + if (len <= EXT4_I(inode)->i_inline_size) + return 0; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + BUG_ON(is.s.not_found); + + len -= EXT4_MIN_INLINE_DATA_SIZE; + value = kzalloc(len, GFP_NOFS); + if (!value) + goto out; + + error = ext4_xattr_ibody_get(inode, i.name_index, i.name, + value, len); + if (error == -ENODATA) + goto out; + + error = ext4_journal_get_write_access(handle, is.iloc.bh); + if (error) + goto out; + + /* Update the xttr entry. */ + i.value = value; + i.value_len = len; + + error = ext4_xattr_ibody_inline_set(handle, inode, &i, &is); + if (error) + goto out; + + EXT4_I(inode)->i_inline_off = (u16)((void *)is.s.here - + (void *)ext4_raw_inode(&is.iloc)); + EXT4_I(inode)->i_inline_size = EXT4_MIN_INLINE_DATA_SIZE + + le32_to_cpu(is.s.here->e_value_size); + ext4_set_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + get_bh(is.iloc.bh); + error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); + +out: + kfree(value); + brelse(is.iloc.bh); + return error; +} + +int ext4_prepare_inline_data(handle_t *handle, struct inode *inode, + unsigned int len) +{ + int ret, size; + struct ext4_inode_info *ei = EXT4_I(inode); + + if (!ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) + return -ENOSPC; + + size = ext4_get_max_inline_size(inode); + if (size < len) + return -ENOSPC; + + down_write(&EXT4_I(inode)->xattr_sem); + + if (ei->i_inline_off) + ret = ext4_update_inline_data(handle, inode, len); + else + ret = ext4_create_inline_data(handle, inode, len); + + up_write(&EXT4_I(inode)->xattr_sem); + + return ret; +} + +static int ext4_destroy_inline_data_nolock(handle_t *handle, + struct inode *inode) +{ + struct ext4_inode_info *ei = EXT4_I(inode); + struct ext4_xattr_ibody_find is = { + .s = { .not_found = 0, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + .value = NULL, + .value_len = 0, + }; + int error; + + if (!ei->i_inline_off) + return 0; + + error = ext4_get_inode_loc(inode, &is.iloc); + if (error) + return error; + + error = ext4_xattr_ibody_find(inode, &i, &is); + if (error) + goto out; + + error = ext4_journal_get_write_access(handle, is.iloc.bh); + if (error) + goto out; + + error = ext4_xattr_ibody_inline_set(handle, inode, &i, &is); + if (error) + goto out; + + memset((void *)ext4_raw_inode(&is.iloc)->i_block, + 0, EXT4_MIN_INLINE_DATA_SIZE); + + if (EXT4_HAS_INCOMPAT_FEATURE(inode->i_sb, + EXT4_FEATURE_INCOMPAT_EXTENTS)) { + if (S_ISDIR(inode->i_mode) || + S_ISREG(inode->i_mode) || S_ISLNK(inode->i_mode)) { + ext4_set_inode_flag(inode, EXT4_INODE_EXTENTS); + ext4_ext_tree_init(handle, inode); + } + } + ext4_clear_inode_flag(inode, EXT4_INODE_INLINE_DATA); + + get_bh(is.iloc.bh); + error = ext4_mark_iloc_dirty(handle, inode, &is.iloc); + + EXT4_I(inode)->i_inline_off = 0; + EXT4_I(inode)->i_inline_size = 0; + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); +out: + brelse(is.iloc.bh); + if (error == -ENODATA) + error = 0; + return error; +} + +static int ext4_read_inline_page(struct inode *inode, struct page *page) +{ + void *kaddr; + int ret = 0; + size_t len; + struct ext4_iloc iloc; + + BUG_ON(!PageLocked(page)); + BUG_ON(!ext4_has_inline_data(inode)); + BUG_ON(page->index); + + if (!EXT4_I(inode)->i_inline_off) { + ext4_warning(inode->i_sb, "inode %lu doesn't have inline data.", + inode->i_ino); + goto out; + } + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + goto out; + + len = min_t(size_t, ext4_get_inline_size(inode), i_size_read(inode)); + kaddr = kmap_atomic(page); + ret = ext4_read_inline_data(inode, kaddr, len, &iloc); + flush_dcache_page(page); + kunmap_atomic(kaddr); + zero_user_segment(page, len, PAGE_CACHE_SIZE); + SetPageUptodate(page); + brelse(iloc.bh); + +out: + return ret; +} + +int ext4_readpage_inline(struct inode *inode, struct page *page) +{ + int ret = 0; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + up_read(&EXT4_I(inode)->xattr_sem); + return -EAGAIN; + } + + /* + * Current inline data can only exist in the 1st page, + * So for all the other pages, just set them uptodate. + */ + if (!page->index) + ret = ext4_read_inline_page(inode, page); + else if (!PageUptodate(page)) { + zero_user_segment(page, 0, PAGE_CACHE_SIZE); + SetPageUptodate(page); + } + + up_read(&EXT4_I(inode)->xattr_sem); + + unlock_page(page); + return ret >= 0 ? 0 : ret; +} + +static int ext4_convert_inline_data_to_extent(struct address_space *mapping, + struct inode *inode, + unsigned flags) +{ + int ret, needed_blocks; + handle_t *handle = NULL; + int retries = 0, sem_held = 0; + struct page *page = NULL; + unsigned from, to; + struct ext4_iloc iloc; + + if (!ext4_has_inline_data(inode)) { + /* + * clear the flag so that no new write + * will trap here again. + */ + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + return 0; + } + + needed_blocks = ext4_writepage_trans_blocks(inode); + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + +retry: + handle = ext4_journal_start(inode, needed_blocks); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + handle = NULL; + goto out; + } + + /* We cannot recurse into the filesystem as the transaction is already + * started */ + flags |= AOP_FLAG_NOFS; + + page = grab_cache_page_write_begin(mapping, 0, flags); + if (!page) { + ret = -ENOMEM; + goto out; + } + + down_write(&EXT4_I(inode)->xattr_sem); + sem_held = 1; + /* If some one has already done this for us, just exit. */ + if (!ext4_has_inline_data(inode)) { + ret = 0; + goto out; + } + + from = 0; + to = ext4_get_inline_size(inode); + if (!PageUptodate(page)) { + ret = ext4_read_inline_page(inode, page); + if (ret < 0) + goto out; + } + + ret = ext4_destroy_inline_data_nolock(handle, inode); + if (ret) + goto out; + + if (ext4_should_dioread_nolock(inode)) + ret = __block_write_begin(page, from, to, ext4_get_block_write); + else + ret = __block_write_begin(page, from, to, ext4_get_block); + + if (!ret && ext4_should_journal_data(inode)) { + ret = ext4_walk_page_buffers(handle, page_buffers(page), + from, to, NULL, + do_journal_get_write_access); + } + + if (ret) { + unlock_page(page); + page_cache_release(page); + ext4_orphan_add(handle, inode); + up_write(&EXT4_I(inode)->xattr_sem); + sem_held = 0; + ext4_journal_stop(handle); + handle = NULL; + ext4_truncate_failed_write(inode); + /* + * If truncate failed early the inode might + * still be on the orphan list; we need to + * make sure the inode is removed from the + * orphan list in that case. + */ + if (inode->i_nlink) + ext4_orphan_del(NULL, inode); + } + + if (ret == -ENOSPC && ext4_should_retry_alloc(inode->i_sb, &retries)) + goto retry; + + block_commit_write(page, from, to); +out: + if (page) { + unlock_page(page); + page_cache_release(page); + } + if (sem_held) + up_write(&EXT4_I(inode)->xattr_sem); + if (handle) + ext4_journal_stop(handle); + brelse(iloc.bh); + return ret; +} + +/* + * Try to write data in the inode. + * If the inode has inline data, check whether the new write can be + * in the inode also. If not, create the page the handle, move the data + * to the page make it update and let the later codes create extent for it. + */ +int ext4_try_to_write_inline_data(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + unsigned flags, + struct page **pagep) +{ + int ret; + handle_t *handle; + struct page *page; + struct ext4_iloc iloc; + + if (pos + len > ext4_get_max_inline_size(inode)) + goto convert; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + /* + * The possible write could happen in the inode, + * so try to reserve the space in inode first. + */ + handle = ext4_journal_start(inode, 1); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + handle = NULL; + goto out; + } + + ret = ext4_prepare_inline_data(handle, inode, pos + len); + if (ret && ret != -ENOSPC) + goto out; + + /* We don't have space in inline inode, so convert it to extent. */ + if (ret == -ENOSPC) { + ext4_journal_stop(handle); + brelse(iloc.bh); + goto convert; + } + + flags |= AOP_FLAG_NOFS; + + page = grab_cache_page_write_begin(mapping, 0, flags); + if (!page) { + ret = -ENOMEM; + goto out; + } + + *pagep = page; + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + ret = 0; + unlock_page(page); + page_cache_release(page); + goto out_up_read; + } + + if (!PageUptodate(page)) { + ret = ext4_read_inline_page(inode, page); + if (ret < 0) + goto out_up_read; + } + + ret = 1; + handle = NULL; +out_up_read: + up_read(&EXT4_I(inode)->xattr_sem); +out: + if (handle) + ext4_journal_stop(handle); + brelse(iloc.bh); + return ret; +convert: + return ext4_convert_inline_data_to_extent(mapping, + inode, flags); +} + +int ext4_write_inline_data_end(struct inode *inode, loff_t pos, unsigned len, + unsigned copied, struct page *page) +{ + int ret; + void *kaddr; + struct ext4_iloc iloc; + + if (unlikely(copied < len)) { + if (!PageUptodate(page)) { + copied = 0; + goto out; + } + } + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) { + ext4_std_error(inode->i_sb, ret); + copied = 0; + goto out; + } + + down_write(&EXT4_I(inode)->xattr_sem); + BUG_ON(!ext4_has_inline_data(inode)); + + kaddr = kmap_atomic(page); + ext4_write_inline_data(inode, &iloc, kaddr, pos, len); + kunmap_atomic(kaddr); + SetPageUptodate(page); + /* clear page dirty so that writepages wouldn't work for us. */ + ClearPageDirty(page); + + up_write(&EXT4_I(inode)->xattr_sem); + brelse(iloc.bh); +out: + return copied; +} + +struct buffer_head * +ext4_journalled_write_inline_data(struct inode *inode, + unsigned len, + struct page *page) +{ + int ret; + void *kaddr; + struct ext4_iloc iloc; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) { + ext4_std_error(inode->i_sb, ret); + return NULL; + } + + down_write(&EXT4_I(inode)->xattr_sem); + kaddr = kmap_atomic(page); + ext4_write_inline_data(inode, &iloc, kaddr, 0, len); + kunmap_atomic(kaddr); + up_write(&EXT4_I(inode)->xattr_sem); + + return iloc.bh; +} + +/* + * Try to make the page cache and handle ready for the inline data case. + * We can call this function in 2 cases: + * 1. The inode is created and the first write exceeds inline size. We can + * clear the inode state safely. + * 2. The inode has inline data, then we need to read the data, make it + * update and dirty so that ext4_da_writepages can handle it. We don't + * need to start the journal since the file's metatdata isn't changed now. + */ +static int ext4_da_convert_inline_data_to_extent(struct address_space *mapping, + struct inode *inode, + unsigned flags, + void **fsdata) +{ + int ret = 0, inline_size; + struct page *page; + + page = grab_cache_page_write_begin(mapping, 0, flags); + if (!page) + return -ENOMEM; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + goto out; + } + + inline_size = ext4_get_inline_size(inode); + + if (!PageUptodate(page)) { + ret = ext4_read_inline_page(inode, page); + if (ret < 0) + goto out; + } + + ret = __block_write_begin(page, 0, inline_size, + ext4_da_get_block_prep); + if (ret) { + ext4_truncate_failed_write(inode); + goto out; + } + + SetPageDirty(page); + SetPageUptodate(page); + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + *fsdata = (void *)CONVERT_INLINE_DATA; + +out: + up_read(&EXT4_I(inode)->xattr_sem); + if (page) { + unlock_page(page); + page_cache_release(page); + } + return ret; +} + +/* + * Prepare the write for the inline data. + * If the the data can be written into the inode, we just read + * the page and make it uptodate, and start the journal. + * Otherwise read the page, makes it dirty so that it can be + * handle in writepages(the i_disksize update is left to the + * normal ext4_da_write_end). + */ +int ext4_da_write_inline_data_begin(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + unsigned flags, + struct page **pagep, + void **fsdata) +{ + int ret, inline_size; + handle_t *handle; + struct page *page; + struct ext4_iloc iloc; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + handle = ext4_journal_start(inode, 1); + if (IS_ERR(handle)) { + ret = PTR_ERR(handle); + handle = NULL; + goto out; + } + + inline_size = ext4_get_max_inline_size(inode); + + ret = -ENOSPC; + if (inline_size >= pos + len) { + ret = ext4_prepare_inline_data(handle, inode, pos + len); + if (ret && ret != -ENOSPC) + goto out; + } + + if (ret == -ENOSPC) { + ret = ext4_da_convert_inline_data_to_extent(mapping, + inode, + flags, + fsdata); + goto out; + } + + /* + * We cannot recurse into the filesystem as the transaction + * is already started. + */ + flags |= AOP_FLAG_NOFS; + + page = grab_cache_page_write_begin(mapping, 0, flags); + if (!page) { + ret = -ENOMEM; + goto out; + } + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + ret = 0; + goto out_release_page; + } + + if (!PageUptodate(page)) { + ret = ext4_read_inline_page(inode, page); + if (ret < 0) + goto out_release_page; + } + + up_read(&EXT4_I(inode)->xattr_sem); + *pagep = page; + handle = NULL; + brelse(iloc.bh); + return 1; +out_release_page: + up_read(&EXT4_I(inode)->xattr_sem); + unlock_page(page); + page_cache_release(page); +out: + if (handle) + ext4_journal_stop(handle); + brelse(iloc.bh); + return ret; +} + +int ext4_da_write_inline_data_end(struct inode *inode, loff_t pos, + unsigned len, unsigned copied, + struct page *page) +{ + int i_size_changed = 0; + + copied = ext4_write_inline_data_end(inode, pos, len, copied, page); + + /* + * No need to use i_size_read() here, the i_size + * cannot change under us because we hold i_mutex. + * + * But it's important to update i_size while still holding page lock: + * page writeout could otherwise come in and zero beyond i_size. + */ + if (pos+copied > inode->i_size) { + i_size_write(inode, pos+copied); + i_size_changed = 1; + } + unlock_page(page); + page_cache_release(page); + + /* + * Don't mark the inode dirty under page lock. First, it unnecessarily + * makes the holding time of page lock longer. Second, it forces lock + * ordering of page lock and transaction start for journaling + * filesystems. + */ + if (i_size_changed) + mark_inode_dirty(inode); + + return copied; +} + +#ifdef INLINE_DIR_DEBUG +void ext4_show_inline_dir(struct inode *dir, struct buffer_head *bh, + void *inline_start, int inline_size) +{ + int offset; + unsigned short de_len; + struct ext4_dir_entry_2 *de = inline_start; + void *dlimit = inline_start + inline_size; + + trace_printk("inode %lu\n", dir->i_ino); + offset = 0; + while ((void *)de < dlimit) { + de_len = ext4_rec_len_from_disk(de->rec_len, inline_size); + trace_printk("de: off %u rlen %u name %*.s nlen %u ino %u\n", + offset, de_len, de->name_len, de->name, + de->name_len, le32_to_cpu(de->inode)); + if (ext4_check_dir_entry(dir, NULL, de, bh, + inline_start, inline_size, offset)) + BUG(); + + offset += de_len; + de = (struct ext4_dir_entry_2 *) ((char *) de + de_len); + } +} +#else +#define ext4_show_inline_dir(dir, bh, inline_start, inline_size) +#endif + +/* + * Add a new entry into a inline dir. + * It will return -ENOSPC if no space is available, and -EIO + * and -EEXIST if directory entry already exists. + */ +static int ext4_add_dirent_to_inline(handle_t *handle, + struct dentry *dentry, + struct inode *inode, + struct ext4_iloc *iloc, + void *inline_start, int inline_size) +{ + struct inode *dir = dentry->d_parent->d_inode; + const char *name = dentry->d_name.name; + int namelen = dentry->d_name.len; + unsigned short reclen; + int err; + struct ext4_dir_entry_2 *de; + + reclen = EXT4_DIR_REC_LEN(namelen); + err = ext4_find_dest_de(dir, inode, iloc->bh, + inline_start, inline_size, + name, namelen, &de); + if (err) + return err; + + err = ext4_journal_get_write_access(handle, iloc->bh); + if (err) + return err; + ext4_insert_dentry(inode, de, inline_size, name, namelen); + + ext4_show_inline_dir(dir, iloc->bh, inline_start, inline_size); + + /* + * XXX shouldn't update any times until successful + * completion of syscall, but too many callers depend + * on this. + * + * XXX similarly, too many callers depend on + * ext4_new_inode() setting the times, but error + * recovery deletes the inode, so the worst that can + * happen is that the times are slightly out of date + * and/or different from the directory change time. + */ + dir->i_mtime = dir->i_ctime = ext4_current_time(dir); + ext4_update_dx_flag(dir); + dir->i_version++; + ext4_mark_inode_dirty(handle, dir); + return 1; +} + +static void *ext4_get_inline_xattr_pos(struct inode *inode, + struct ext4_iloc *iloc) +{ + struct ext4_xattr_entry *entry; + struct ext4_xattr_ibody_header *header; + + BUG_ON(!EXT4_I(inode)->i_inline_off); + + header = IHDR(inode, ext4_raw_inode(iloc)); + entry = (struct ext4_xattr_entry *)((void *)ext4_raw_inode(iloc) + + EXT4_I(inode)->i_inline_off); + + return (void *)IFIRST(header) + le16_to_cpu(entry->e_value_offs); +} + +/* Set the final de to cover the whole block. */ +static void ext4_update_final_de(void *de_buf, int old_size, int new_size) +{ + struct ext4_dir_entry_2 *de, *prev_de; + void *limit; + int de_len; + + de = (struct ext4_dir_entry_2 *)de_buf; + if (old_size) { + limit = de_buf + old_size; + do { + prev_de = de; + de_len = ext4_rec_len_from_disk(de->rec_len, old_size); + de_buf += de_len; + de = (struct ext4_dir_entry_2 *)de_buf; + } while (de_buf < limit); + + prev_de->rec_len = ext4_rec_len_to_disk(de_len + new_size - + old_size, new_size); + } else { + /* this is just created, so create an empty entry. */ + de->inode = 0; + de->rec_len = ext4_rec_len_to_disk(new_size, new_size); + } +} + +static int ext4_update_inline_dir(handle_t *handle, struct inode *dir, + struct ext4_iloc *iloc) +{ + int ret; + int old_size = EXT4_I(dir)->i_inline_size - EXT4_MIN_INLINE_DATA_SIZE; + int new_size = get_max_inline_xattr_value_size(dir, iloc); + + if (new_size - old_size <= EXT4_DIR_REC_LEN(1)) + return -ENOSPC; + + ret = ext4_update_inline_data(handle, dir, + new_size + EXT4_MIN_INLINE_DATA_SIZE); + if (ret) + return ret; + + ext4_update_final_de(ext4_get_inline_xattr_pos(dir, iloc), old_size, + EXT4_I(dir)->i_inline_size - + EXT4_MIN_INLINE_DATA_SIZE); + dir->i_size = EXT4_I(dir)->i_disksize = EXT4_I(dir)->i_inline_size; + return 0; +} + +static void ext4_restore_inline_data(handle_t *handle, struct inode *inode, + struct ext4_iloc *iloc, + void *buf, int inline_size) +{ + ext4_create_inline_data(handle, inode, inline_size); + ext4_write_inline_data(inode, iloc, buf, 0, inline_size); + ext4_set_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); +} + +static int ext4_finish_convert_inline_dir(handle_t *handle, + struct inode *inode, + struct buffer_head *dir_block, + void *buf, + int inline_size) +{ + int err, csum_size = 0, header_size = 0; + struct ext4_dir_entry_2 *de; + struct ext4_dir_entry_tail *t; + void *target = dir_block->b_data; + + /* + * First create "." and ".." and then copy the dir information + * back to the block. + */ + de = (struct ext4_dir_entry_2 *)target; + de = ext4_init_dot_dotdot(inode, de, + inode->i_sb->s_blocksize, csum_size, + le32_to_cpu(((struct ext4_dir_entry_2 *)buf)->inode), 1); + header_size = (void *)de - target; + + memcpy((void *)de, buf + EXT4_INLINE_DOTDOT_SIZE, + inline_size - EXT4_INLINE_DOTDOT_SIZE); + + if (EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb, + EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) + csum_size = sizeof(struct ext4_dir_entry_tail); + + inode->i_size = inode->i_sb->s_blocksize; + i_size_write(inode, inode->i_sb->s_blocksize); + EXT4_I(inode)->i_disksize = inode->i_sb->s_blocksize; + ext4_update_final_de(dir_block->b_data, + inline_size - EXT4_INLINE_DOTDOT_SIZE + header_size, + inode->i_sb->s_blocksize - csum_size); + + if (csum_size) { + t = EXT4_DIRENT_TAIL(dir_block->b_data, + inode->i_sb->s_blocksize); + initialize_dirent_tail(t, inode->i_sb->s_blocksize); + } + set_buffer_uptodate(dir_block); + err = ext4_handle_dirty_dirent_node(handle, inode, dir_block); + if (err) + goto out; + set_buffer_verified(dir_block); +out: + return err; +} + +static int ext4_convert_inline_data_nolock(handle_t *handle, + struct inode *inode, + struct ext4_iloc *iloc) +{ + int error; + void *buf = NULL; + struct buffer_head *data_bh = NULL; + struct ext4_map_blocks map; + int inline_size; + + inline_size = ext4_get_inline_size(inode); + buf = kmalloc(inline_size, GFP_NOFS); + if (!buf) { + error = -ENOMEM; + goto out; + } + + error = ext4_read_inline_data(inode, buf, inline_size, iloc); + if (error < 0) + goto out; + + error = ext4_destroy_inline_data_nolock(handle, inode); + if (error) + goto out; + + map.m_lblk = 0; + map.m_len = 1; + map.m_flags = 0; + error = ext4_map_blocks(handle, inode, &map, EXT4_GET_BLOCKS_CREATE); + if (error < 0) + goto out_restore; + if (!(map.m_flags & EXT4_MAP_MAPPED)) { + error = -EIO; + goto out_restore; + } + + data_bh = sb_getblk(inode->i_sb, map.m_pblk); + if (!data_bh) { + error = -EIO; + goto out_restore; + } + + lock_buffer(data_bh); + error = ext4_journal_get_create_access(handle, data_bh); + if (error) { + unlock_buffer(data_bh); + error = -EIO; + goto out_restore; + } + memset(data_bh->b_data, 0, inode->i_sb->s_blocksize); + + if (!S_ISDIR(inode->i_mode)) { + memcpy(data_bh->b_data, buf, inline_size); + set_buffer_uptodate(data_bh); + error = ext4_handle_dirty_metadata(handle, + inode, data_bh); + } else { + error = ext4_finish_convert_inline_dir(handle, inode, data_bh, + buf, inline_size); + } + + unlock_buffer(data_bh); +out_restore: + if (error) + ext4_restore_inline_data(handle, inode, iloc, buf, inline_size); + +out: + brelse(data_bh); + kfree(buf); + return error; +} + +/* + * Try to add the new entry to the inline data. + * If succeeds, return 0. If not, extended the inline dir and copied data to + * the new created block. + */ +int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, + struct inode *inode) +{ + int ret, inline_size; + void *inline_start; + struct ext4_iloc iloc; + struct inode *dir = dentry->d_parent->d_inode; + + ret = ext4_get_inode_loc(dir, &iloc); + if (ret) + return ret; + + down_write(&EXT4_I(dir)->xattr_sem); + if (!ext4_has_inline_data(dir)) + goto out; + + inline_start = (void *)ext4_raw_inode(&iloc)->i_block + + EXT4_INLINE_DOTDOT_SIZE; + inline_size = EXT4_MIN_INLINE_DATA_SIZE - EXT4_INLINE_DOTDOT_SIZE; + + ret = ext4_add_dirent_to_inline(handle, dentry, inode, &iloc, + inline_start, inline_size); + if (ret != -ENOSPC) + goto out; + + /* check whether it can be inserted to inline xattr space. */ + inline_size = EXT4_I(dir)->i_inline_size - + EXT4_MIN_INLINE_DATA_SIZE; + if (!inline_size) { + /* Try to use the xattr space.*/ + ret = ext4_update_inline_dir(handle, dir, &iloc); + if (ret && ret != -ENOSPC) + goto out; + + inline_size = EXT4_I(dir)->i_inline_size - + EXT4_MIN_INLINE_DATA_SIZE; + } + + if (inline_size) { + inline_start = ext4_get_inline_xattr_pos(dir, &iloc); + + ret = ext4_add_dirent_to_inline(handle, dentry, inode, &iloc, + inline_start, inline_size); + + if (ret != -ENOSPC) + goto out; + } + + /* + * The inline space is filled up, so create a new block for it. + * As the extent tree will be created, we have to save the inline + * dir first. + */ + ret = ext4_convert_inline_data_nolock(handle, dir, &iloc); + +out: + ext4_mark_inode_dirty(handle, dir); + up_write(&EXT4_I(dir)->xattr_sem); + brelse(iloc.bh); + return ret; +} + +int ext4_read_inline_dir(struct file *filp, + void *dirent, filldir_t filldir, + int *has_inline_data) +{ + int error = 0; + unsigned int offset, parent_ino; + int i, stored; + struct ext4_dir_entry_2 *de; + struct super_block *sb; + struct inode *inode = filp->f_path.dentry->d_inode; + int ret, inline_size = 0; + struct ext4_iloc iloc; + void *dir_buf = NULL; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + up_read(&EXT4_I(inode)->xattr_sem); + *has_inline_data = 0; + goto out; + } + + inline_size = ext4_get_inline_size(inode); + dir_buf = kmalloc(inline_size, GFP_NOFS); + if (!dir_buf) { + ret = -ENOMEM; + up_read(&EXT4_I(inode)->xattr_sem); + goto out; + } + + ret = ext4_read_inline_data(inode, dir_buf, inline_size, &iloc); + up_read(&EXT4_I(inode)->xattr_sem); + if (ret < 0) + goto out; + + sb = inode->i_sb; + stored = 0; + parent_ino = le32_to_cpu(((struct ext4_dir_entry_2 *)dir_buf)->inode); + + while (!error && !stored && filp->f_pos < inode->i_size) { +revalidate: + /* + * If the version has changed since the last call to + * readdir(2), then we might be pointing to an invalid + * dirent right now. Scan from the start of the inline + * dir to make sure. + */ + if (filp->f_version != inode->i_version) { + for (i = 0; + i < inode->i_size && i < offset;) { + if (!i) { + /* skip "." and ".." if needed. */ + i += EXT4_INLINE_DOTDOT_SIZE; + continue; + } + de = (struct ext4_dir_entry_2 *) + (dir_buf + i); + /* It's too expensive to do a full + * dirent test each time round this + * loop, but we do have to test at + * least that it is non-zero. A + * failure will be detected in the + * dirent test below. */ + if (ext4_rec_len_from_disk(de->rec_len, + inline_size) < EXT4_DIR_REC_LEN(1)) + break; + i += ext4_rec_len_from_disk(de->rec_len, + inline_size); + } + offset = i; + filp->f_pos = offset; + filp->f_version = inode->i_version; + } + + while (!error && filp->f_pos < inode->i_size) { + if (filp->f_pos == 0) { + error = filldir(dirent, ".", 1, 0, inode->i_ino, + DT_DIR); + if (error) + break; + stored++; + + error = filldir(dirent, "..", 2, 0, parent_ino, + DT_DIR); + if (error) + break; + stored++; + + filp->f_pos = offset = EXT4_INLINE_DOTDOT_SIZE; + continue; + } + + de = (struct ext4_dir_entry_2 *)(dir_buf + offset); + if (ext4_check_dir_entry(inode, filp, de, + iloc.bh, dir_buf, + inline_size, offset)) { + ret = stored; + goto out; + } + offset += ext4_rec_len_from_disk(de->rec_len, + inline_size); + if (le32_to_cpu(de->inode)) { + /* We might block in the next section + * if the data destination is + * currently swapped out. So, use a + * version stamp to detect whether or + * not the directory has been modified + * during the copy operation. + */ + u64 version = filp->f_version; + + error = filldir(dirent, de->name, + de->name_len, + filp->f_pos, + le32_to_cpu(de->inode), + get_dtype(sb, de->file_type)); + if (error) + break; + if (version != filp->f_version) + goto revalidate; + stored++; + } + filp->f_pos += ext4_rec_len_from_disk(de->rec_len, + inline_size); + } + offset = 0; + } +out: + kfree(dir_buf); + brelse(iloc.bh); + return ret; +} + +struct buffer_head *ext4_get_first_inline_block(struct inode *inode, + struct ext4_dir_entry_2 **parent_de, + int *retval) +{ + struct ext4_iloc iloc; + + *retval = ext4_get_inode_loc(inode, &iloc); + if (*retval) + return NULL; + + *parent_de = (struct ext4_dir_entry_2 *)ext4_raw_inode(&iloc)->i_block; + + return iloc.bh; +} + +/* + * Try to create the inline data for the new dir. + * If it succeeds, return 0, otherwise return the error. + * In case of ENOSPC, the caller should create the normal disk layout dir. + */ +int ext4_try_create_inline_dir(handle_t *handle, struct inode *parent, + struct inode *inode) +{ + int ret, inline_size = EXT4_MIN_INLINE_DATA_SIZE; + struct ext4_iloc iloc; + struct ext4_dir_entry_2 *de; + + ret = ext4_get_inode_loc(inode, &iloc); + if (ret) + return ret; + + ret = ext4_prepare_inline_data(handle, inode, inline_size); + if (ret) + goto out; + + /* + * For inline dir, we only save the inode information for the ".." + * and create a fake dentry to cover the left space. + */ + de = (struct ext4_dir_entry_2 *)ext4_raw_inode(&iloc)->i_block; + de->inode = cpu_to_le32(parent->i_ino); + de = (struct ext4_dir_entry_2 *)((void *)de + EXT4_INLINE_DOTDOT_SIZE); + de->inode = 0; + de->rec_len = ext4_rec_len_to_disk( + inline_size - EXT4_INLINE_DOTDOT_SIZE, + inline_size); + set_nlink(inode, 2); + inode->i_size = EXT4_I(inode)->i_disksize = inline_size; +out: + brelse(iloc.bh); + return ret; +} + +struct buffer_head *ext4_find_inline_entry(struct inode *dir, + const struct qstr *d_name, + struct ext4_dir_entry_2 **res_dir, + int *has_inline_data) +{ + int ret; + struct ext4_iloc iloc; + void *inline_start; + int inline_size; + + if (ext4_get_inode_loc(dir, &iloc)) + return NULL; + + down_read(&EXT4_I(dir)->xattr_sem); + if (!ext4_has_inline_data(dir)) { + *has_inline_data = 0; + goto out; + } + + inline_start = (void *)ext4_raw_inode(&iloc)->i_block + + EXT4_INLINE_DOTDOT_SIZE; + inline_size = EXT4_MIN_INLINE_DATA_SIZE - EXT4_INLINE_DOTDOT_SIZE; + ret = search_dir(iloc.bh, inline_start, inline_size, + dir, d_name, 0, res_dir); + if (ret == 1) + goto out_find; + if (ret < 0) + goto out; + + if (ext4_get_inline_size(dir) == EXT4_MIN_INLINE_DATA_SIZE) + goto out; + + inline_start = ext4_get_inline_xattr_pos(dir, &iloc); + inline_size = ext4_get_inline_size(dir) - EXT4_MIN_INLINE_DATA_SIZE; + + ret = search_dir(iloc.bh, inline_start, inline_size, + dir, d_name, 0, res_dir); + if (ret == 1) + goto out_find; + +out: + brelse(iloc.bh); + iloc.bh = NULL; +out_find: + up_read(&EXT4_I(dir)->xattr_sem); + return iloc.bh; +} + +int ext4_delete_inline_entry(handle_t *handle, + struct inode *dir, + struct ext4_dir_entry_2 *de_del, + struct buffer_head *bh, + int *has_inline_data) +{ + int err, inline_size; + struct ext4_iloc iloc; + void *inline_start; + + err = ext4_get_inode_loc(dir, &iloc); + if (err) + return err; + + down_write(&EXT4_I(dir)->xattr_sem); + if (!ext4_has_inline_data(dir)) { + *has_inline_data = 0; + goto out; + } + + if ((void *)de_del - ((void *)ext4_raw_inode(&iloc)->i_block) < + EXT4_MIN_INLINE_DATA_SIZE) { + inline_start = (void *)ext4_raw_inode(&iloc)->i_block + + EXT4_INLINE_DOTDOT_SIZE; + inline_size = EXT4_MIN_INLINE_DATA_SIZE - + EXT4_INLINE_DOTDOT_SIZE; + } else { + inline_start = ext4_get_inline_xattr_pos(dir, &iloc); + inline_size = ext4_get_inline_size(dir) - + EXT4_MIN_INLINE_DATA_SIZE; + } + + err = ext4_journal_get_write_access(handle, bh); + if (err) + goto out; + + err = ext4_generic_delete_entry(handle, dir, de_del, bh, + inline_start, inline_size, 0); + if (err) + goto out; + + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + err = ext4_mark_inode_dirty(handle, dir); + if (unlikely(err)) + goto out; + + ext4_show_inline_dir(dir, iloc.bh, inline_start, inline_size); +out: + up_write(&EXT4_I(dir)->xattr_sem); + brelse(iloc.bh); + if (err != -ENOENT) + ext4_std_error(dir->i_sb, err); + return err; +} + +/* + * Get the inline dentry at offset. + */ +static inline struct ext4_dir_entry_2 * +ext4_get_inline_entry(struct inode *inode, + struct ext4_iloc *iloc, + unsigned int offset, + void **inline_start, + int *inline_size) +{ + void *inline_pos; + + BUG_ON(offset > ext4_get_inline_size(inode)); + + if (offset < EXT4_MIN_INLINE_DATA_SIZE) { + inline_pos = (void *)ext4_raw_inode(iloc)->i_block; + *inline_size = EXT4_MIN_INLINE_DATA_SIZE; + } else { + inline_pos = ext4_get_inline_xattr_pos(inode, iloc); + offset -= EXT4_MIN_INLINE_DATA_SIZE; + *inline_size = ext4_get_inline_size(inode) - + EXT4_MIN_INLINE_DATA_SIZE; + } + + if (inline_start) + *inline_start = inline_pos; + return (struct ext4_dir_entry_2 *)(inline_pos + offset); +} + +int empty_inline_dir(struct inode *dir, int *has_inline_data) +{ + int err, inline_size; + struct ext4_iloc iloc; + void *inline_pos; + unsigned int offset; + struct ext4_dir_entry_2 *de; + int ret = 1; + + err = ext4_get_inode_loc(dir, &iloc); + if (err) { + EXT4_ERROR_INODE(dir, "error %d getting inode %lu block", + err, dir->i_ino); + return 1; + } + + down_read(&EXT4_I(dir)->xattr_sem); + if (!ext4_has_inline_data(dir)) { + *has_inline_data = 0; + goto out; + } + + de = (struct ext4_dir_entry_2 *)ext4_raw_inode(&iloc)->i_block; + if (!le32_to_cpu(de->inode)) { + ext4_warning(dir->i_sb, + "bad inline directory (dir #%lu) - no `..'", + dir->i_ino); + ret = 1; + goto out; + } + + offset = EXT4_INLINE_DOTDOT_SIZE; + while (offset < dir->i_size) { + de = ext4_get_inline_entry(dir, &iloc, offset, + &inline_pos, &inline_size); + if (ext4_check_dir_entry(dir, NULL, de, + iloc.bh, inline_pos, + inline_size, offset)) { + ext4_warning(dir->i_sb, + "bad inline directory (dir #%lu) - " + "inode %u, rec_len %u, name_len %d" + "inline size %d\n", + dir->i_ino, le32_to_cpu(de->inode), + le16_to_cpu(de->rec_len), de->name_len, + inline_size); + ret = 1; + goto out; + } + if (le32_to_cpu(de->inode)) { + ret = 0; + goto out; + } + offset += ext4_rec_len_from_disk(de->rec_len, inline_size); + } + +out: + up_read(&EXT4_I(dir)->xattr_sem); + brelse(iloc.bh); + return ret; +} + +int ext4_destroy_inline_data(handle_t *handle, struct inode *inode) +{ + int ret; + + down_write(&EXT4_I(inode)->xattr_sem); + ret = ext4_destroy_inline_data_nolock(handle, inode); + up_write(&EXT4_I(inode)->xattr_sem); + + return ret; +} + +int ext4_inline_data_fiemap(struct inode *inode, + struct fiemap_extent_info *fieinfo, + int *has_inline) +{ + __u64 physical = 0; + __u64 length; + __u32 flags = FIEMAP_EXTENT_DATA_INLINE | FIEMAP_EXTENT_LAST; + int error = 0; + struct ext4_iloc iloc; + + down_read(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + *has_inline = 0; + goto out; + } + + error = ext4_get_inode_loc(inode, &iloc); + if (error) + goto out; + + physical = iloc.bh->b_blocknr << inode->i_sb->s_blocksize_bits; + physical += (char *)ext4_raw_inode(&iloc) - iloc.bh->b_data; + physical += offsetof(struct ext4_inode, i_block); + length = i_size_read(inode); + + if (physical) + error = fiemap_fill_next_extent(fieinfo, 0, physical, + length, flags); + brelse(iloc.bh); +out: + up_read(&EXT4_I(inode)->xattr_sem); + return (error < 0 ? error : 0); +} + +/* + * Called during xattr set, and if we can sparse space 'needed', + * just create the extent tree evict the data to the outer block. + * + * We use jbd2 instead of page cache to move data to the 1st block + * so that the whole transaction can be committed as a whole and + * the data isn't lost because of the delayed page cache write. + */ +int ext4_try_to_evict_inline_data(handle_t *handle, + struct inode *inode, + int needed) +{ + int error; + struct ext4_xattr_entry *entry; + struct ext4_xattr_ibody_header *header; + struct ext4_inode *raw_inode; + struct ext4_iloc iloc; + + error = ext4_get_inode_loc(inode, &iloc); + if (error) + return error; + + raw_inode = ext4_raw_inode(&iloc); + header = IHDR(inode, raw_inode); + entry = (struct ext4_xattr_entry *)((void *)raw_inode + + EXT4_I(inode)->i_inline_off); + if (EXT4_XATTR_LEN(entry->e_name_len) + + EXT4_XATTR_SIZE(le32_to_cpu(entry->e_value_size)) < needed) { + error = -ENOSPC; + goto out; + } + + error = ext4_convert_inline_data_nolock(handle, inode, &iloc); +out: + brelse(iloc.bh); + return error; +} + +void ext4_inline_data_truncate(struct inode *inode, int *has_inline) +{ + handle_t *handle; + int inline_size, value_len, needed_blocks; + size_t i_size; + void *value = NULL; + struct ext4_xattr_ibody_find is = { + .s = { .not_found = -ENODATA, }, + }; + struct ext4_xattr_info i = { + .name_index = EXT4_XATTR_INDEX_SYSTEM, + .name = EXT4_XATTR_SYSTEM_DATA, + }; + + + needed_blocks = ext4_writepage_trans_blocks(inode); + handle = ext4_journal_start(inode, needed_blocks); + if (IS_ERR(handle)) + return; + + down_write(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + *has_inline = 0; + ext4_journal_stop(handle); + return; + } + + if (ext4_orphan_add(handle, inode)) + goto out; + + if (ext4_get_inode_loc(inode, &is.iloc)) + goto out; + + down_write(&EXT4_I(inode)->i_data_sem); + i_size = inode->i_size; + inline_size = ext4_get_inline_size(inode); + EXT4_I(inode)->i_disksize = i_size; + + if (i_size < inline_size) { + /* Clear the content in the xattr space. */ + if (inline_size > EXT4_MIN_INLINE_DATA_SIZE) { + if (ext4_xattr_ibody_find(inode, &i, &is)) + goto out_error; + + BUG_ON(is.s.not_found); + + value_len = le32_to_cpu(is.s.here->e_value_size); + value = kmalloc(value_len, GFP_NOFS); + if (!value) + goto out_error; + + if (ext4_xattr_ibody_get(inode, i.name_index, i.name, + value, value_len)) + goto out_error; + + i.value = value; + i.value_len = i_size > EXT4_MIN_INLINE_DATA_SIZE ? + i_size - EXT4_MIN_INLINE_DATA_SIZE : 0; + if (ext4_xattr_ibody_inline_set(handle, inode, &i, &is)) + goto out_error; + } + + /* Clear the content within i_blocks. */ + if (i_size < EXT4_MIN_INLINE_DATA_SIZE) + memset(ext4_raw_inode(&is.iloc)->i_block + i_size, 0, + EXT4_MIN_INLINE_DATA_SIZE - i_size); + + EXT4_I(inode)->i_inline_size = i_size < + EXT4_MIN_INLINE_DATA_SIZE ? + EXT4_MIN_INLINE_DATA_SIZE : i_size; + } + +out_error: + up_write(&EXT4_I(inode)->i_data_sem); +out: + brelse(is.iloc.bh); + up_write(&EXT4_I(inode)->xattr_sem); + kfree(value); + if (inode->i_nlink) + ext4_orphan_del(handle, inode); + + inode->i_mtime = inode->i_ctime = ext4_current_time(inode); + ext4_mark_inode_dirty(handle, inode); + if (IS_SYNC(inode)) + ext4_handle_sync(handle); + + ext4_journal_stop(handle); + return; +} + +int ext4_convert_inline_data(struct inode *inode) +{ + int error, needed_blocks; + handle_t *handle; + struct ext4_iloc iloc; + + if (!ext4_has_inline_data(inode)) { + ext4_clear_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA); + return 0; + } + + needed_blocks = ext4_writepage_trans_blocks(inode); + + iloc.bh = NULL; + error = ext4_get_inode_loc(inode, &iloc); + if (error) + return error; + + handle = ext4_journal_start(inode, needed_blocks); + if (IS_ERR(handle)) { + error = PTR_ERR(handle); + goto out_free; + } + + down_write(&EXT4_I(inode)->xattr_sem); + if (!ext4_has_inline_data(inode)) { + up_write(&EXT4_I(inode)->xattr_sem); + goto out; + } + + error = ext4_convert_inline_data_nolock(handle, inode, &iloc); + up_write(&EXT4_I(inode)->xattr_sem); +out: + ext4_journal_stop(handle); +out_free: + brelse(iloc.bh); + return error; +} diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index b3c243b9afa5..cb1c1ab2720b 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -484,49 +484,6 @@ static pgoff_t ext4_num_dirty_pages(struct inode *inode, pgoff_t idx, } /* - * Sets the BH_Da_Mapped bit on the buffer heads corresponding to the given map. - */ -static void set_buffers_da_mapped(struct inode *inode, - struct ext4_map_blocks *map) -{ - struct address_space *mapping = inode->i_mapping; - struct pagevec pvec; - int i, nr_pages; - pgoff_t index, end; - - index = map->m_lblk >> (PAGE_CACHE_SHIFT - inode->i_blkbits); - end = (map->m_lblk + map->m_len - 1) >> - (PAGE_CACHE_SHIFT - inode->i_blkbits); - - pagevec_init(&pvec, 0); - while (index <= end) { - nr_pages = pagevec_lookup(&pvec, mapping, index, - min(end - index + 1, - (pgoff_t)PAGEVEC_SIZE)); - if (nr_pages == 0) - break; - for (i = 0; i < nr_pages; i++) { - struct page *page = pvec.pages[i]; - struct buffer_head *bh, *head; - - if (unlikely(page->mapping != mapping) || - !PageDirty(page)) - break; - - if (page_has_buffers(page)) { - bh = head = page_buffers(page); - do { - set_buffer_da_mapped(bh); - bh = bh->b_this_page; - } while (bh != head); - } - index++; - } - pagevec_release(&pvec); - } -} - -/* * The ext4_map_blocks() function tries to look up the requested blocks, * and returns if the blocks are already mapped. * @@ -574,7 +531,16 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, up_read((&EXT4_I(inode)->i_data_sem)); if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { - int ret = check_block_validity(inode, map); + int ret; + if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) { + /* delayed alloc may be allocated by fallocate and + * coverted to initialized by directIO. + * we need to handle delayed extent here. + */ + down_write((&EXT4_I(inode)->i_data_sem)); + goto delayed_mapped; + } + ret = check_block_validity(inode, map); if (ret != 0) return ret; } @@ -652,12 +618,15 @@ int ext4_map_blocks(handle_t *handle, struct inode *inode, if (flags & EXT4_GET_BLOCKS_DELALLOC_RESERVE) { ext4_clear_inode_state(inode, EXT4_STATE_DELALLOC_RESERVED); - /* If we have successfully mapped the delayed allocated blocks, - * set the BH_Da_Mapped bit on them. Its important to do this - * under the protection of i_data_sem. - */ - if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) - set_buffers_da_mapped(inode, map); + if (retval > 0 && map->m_flags & EXT4_MAP_MAPPED) { + int ret; +delayed_mapped: + /* delayed allocation blocks has been allocated */ + ret = ext4_es_remove_extent(inode, map->m_lblk, + map->m_len); + if (ret < 0) + retval = ret; + } } up_write((&EXT4_I(inode)->i_data_sem)); @@ -680,10 +649,13 @@ static int _ext4_get_block(struct inode *inode, sector_t iblock, int ret = 0, started = 0; int dio_credits; + if (ext4_has_inline_data(inode)) + return -ERANGE; + map.m_lblk = iblock; map.m_len = bh->b_size >> inode->i_blkbits; - if (flags && !handle) { + if (flags && !(flags & EXT4_GET_BLOCKS_NO_LOCK) && !handle) { /* Direct IO write... */ if (map.m_len > DIO_MAX_BLOCKS) map.m_len = DIO_MAX_BLOCKS; @@ -798,13 +770,13 @@ struct buffer_head *ext4_bread(handle_t *handle, struct inode *inode, return NULL; } -static int walk_page_buffers(handle_t *handle, - struct buffer_head *head, - unsigned from, - unsigned to, - int *partial, - int (*fn)(handle_t *handle, - struct buffer_head *bh)) +int ext4_walk_page_buffers(handle_t *handle, + struct buffer_head *head, + unsigned from, + unsigned to, + int *partial, + int (*fn)(handle_t *handle, + struct buffer_head *bh)) { struct buffer_head *bh; unsigned block_start, block_end; @@ -854,8 +826,8 @@ static int walk_page_buffers(handle_t *handle, * is elevated. We'll still have enough credits for the tiny quotafile * write. */ -static int do_journal_get_write_access(handle_t *handle, - struct buffer_head *bh) +int do_journal_get_write_access(handle_t *handle, + struct buffer_head *bh) { int dirty = buffer_dirty(bh); int ret; @@ -878,7 +850,7 @@ static int do_journal_get_write_access(handle_t *handle, return ret; } -static int ext4_get_block_write(struct inode *inode, sector_t iblock, +static int ext4_get_block_write_nolock(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create); static int ext4_write_begin(struct file *file, struct address_space *mapping, loff_t pos, unsigned len, unsigned flags, @@ -902,6 +874,17 @@ static int ext4_write_begin(struct file *file, struct address_space *mapping, from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; + if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) { + ret = ext4_try_to_write_inline_data(mapping, inode, pos, len, + flags, pagep); + if (ret < 0) + goto out; + if (ret == 1) { + ret = 0; + goto out; + } + } + retry: handle = ext4_journal_start(inode, needed_blocks); if (IS_ERR(handle)) { @@ -919,6 +902,7 @@ retry: ret = -ENOMEM; goto out; } + *pagep = page; if (ext4_should_dioread_nolock(inode)) @@ -927,8 +911,9 @@ retry: ret = __block_write_begin(page, pos, len, ext4_get_block); if (!ret && ext4_should_journal_data(inode)) { - ret = walk_page_buffers(handle, page_buffers(page), - from, to, NULL, do_journal_get_write_access); + ret = ext4_walk_page_buffers(handle, page_buffers(page), + from, to, NULL, + do_journal_get_write_access); } if (ret) { @@ -983,7 +968,12 @@ static int ext4_generic_write_end(struct file *file, struct inode *inode = mapping->host; handle_t *handle = ext4_journal_current_handle(); - copied = block_write_end(file, mapping, pos, len, copied, page, fsdata); + if (ext4_has_inline_data(inode)) + copied = ext4_write_inline_data_end(inode, pos, len, + copied, page); + else + copied = block_write_end(file, mapping, pos, + len, copied, page, fsdata); /* * No need to use i_size_read() here, the i_size @@ -1134,16 +1124,21 @@ static int ext4_journalled_write_end(struct file *file, BUG_ON(!ext4_handle_valid(handle)); - if (copied < len) { - if (!PageUptodate(page)) - copied = 0; - page_zero_new_buffers(page, from+copied, to); - } + if (ext4_has_inline_data(inode)) + copied = ext4_write_inline_data_end(inode, pos, len, + copied, page); + else { + if (copied < len) { + if (!PageUptodate(page)) + copied = 0; + page_zero_new_buffers(page, from+copied, to); + } - ret = walk_page_buffers(handle, page_buffers(page), from, - to, &partial, write_end_fn); - if (!partial) - SetPageUptodate(page); + ret = ext4_walk_page_buffers(handle, page_buffers(page), from, + to, &partial, write_end_fn); + if (!partial) + SetPageUptodate(page); + } new_i_size = pos + copied; if (new_i_size > inode->i_size) i_size_write(inode, pos+copied); @@ -1301,6 +1296,7 @@ static void ext4_da_page_release_reservation(struct page *page, struct inode *inode = page->mapping->host; struct ext4_sb_info *sbi = EXT4_SB(inode->i_sb); int num_clusters; + ext4_fsblk_t lblk; head = page_buffers(page); bh = head; @@ -1310,20 +1306,23 @@ static void ext4_da_page_release_reservation(struct page *page, if ((offset <= curr_off) && (buffer_delay(bh))) { to_release++; clear_buffer_delay(bh); - clear_buffer_da_mapped(bh); } curr_off = next_off; } while ((bh = bh->b_this_page) != head); + if (to_release) { + lblk = page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits); + ext4_es_remove_extent(inode, lblk, to_release); + } + /* If we have released all the blocks belonging to a cluster, then we * need to release the reserved space for that cluster. */ num_clusters = EXT4_NUM_B2C(sbi, to_release); while (num_clusters > 0) { - ext4_fsblk_t lblk; lblk = (page->index << (PAGE_CACHE_SHIFT - inode->i_blkbits)) + ((num_clusters - 1) << sbi->s_cluster_bits); if (sbi->s_cluster_ratio == 1 || - !ext4_find_delalloc_cluster(inode, lblk, 1)) + !ext4_find_delalloc_cluster(inode, lblk)) ext4_da_release_space(inode, 1); num_clusters--; @@ -1429,8 +1428,6 @@ static int mpage_da_submit_io(struct mpage_da_data *mpd, clear_buffer_delay(bh); bh->b_blocknr = pblock; } - if (buffer_da_mapped(bh)) - clear_buffer_da_mapped(bh); if (buffer_unwritten(bh) || buffer_mapped(bh)) BUG_ON(bh->b_blocknr != pblock); @@ -1500,9 +1497,16 @@ static void ext4_da_block_invalidatepages(struct mpage_da_data *mpd) struct pagevec pvec; struct inode *inode = mpd->inode; struct address_space *mapping = inode->i_mapping; + ext4_lblk_t start, last; index = mpd->first_page; end = mpd->next_page - 1; + + start = index << (PAGE_CACHE_SHIFT - inode->i_blkbits); + last = end << (PAGE_CACHE_SHIFT - inode->i_blkbits); + ext4_es_remove_extent(inode, start, last - start + 1); + + pagevec_init(&pvec, 0); while (index <= end) { nr_pages = pagevec_lookup(&pvec, mapping, index, PAGEVEC_SIZE); if (nr_pages == 0) @@ -1656,15 +1660,6 @@ static void mpage_da_map_and_submit(struct mpage_da_data *mpd) for (i = 0; i < map.m_len; i++) unmap_underlying_metadata(bdev, map.m_pblk + i); - - if (ext4_should_order_data(mpd->inode)) { - err = ext4_jbd2_file_inode(handle, mpd->inode); - if (err) { - /* Only if the journal is aborted */ - mpd->retval = err; - goto submit_io; - } - } } /* @@ -1795,7 +1790,19 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, * file system block. */ down_read((&EXT4_I(inode)->i_data_sem)); - if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) + if (ext4_has_inline_data(inode)) { + /* + * We will soon create blocks for this page, and let + * us pretend as if the blocks aren't allocated yet. + * In case of clusters, we have to handle the work + * of mapping from cluster so that the reserved space + * is calculated properly. + */ + if ((EXT4_SB(inode->i_sb)->s_cluster_ratio > 1) && + ext4_find_delalloc_cluster(inode, map->m_lblk)) + map->m_flags |= EXT4_MAP_FROM_CLUSTER; + retval = 0; + } else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) retval = ext4_ext_map_blocks(NULL, inode, map, 0); else retval = ext4_ind_map_blocks(NULL, inode, map, 0); @@ -1814,6 +1821,10 @@ static int ext4_da_map_blocks(struct inode *inode, sector_t iblock, goto out_unlock; } + retval = ext4_es_insert_extent(inode, map->m_lblk, map->m_len); + if (retval) + goto out_unlock; + /* Clear EXT4_MAP_FROM_CLUSTER flag since its purpose is served * and it should not appear on the bh->b_state. */ @@ -1842,8 +1853,8 @@ out_unlock: * We also have b_blocknr = physicalblock mapping unwritten extent and b_bdev * initialized properly. */ -static int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, - struct buffer_head *bh, int create) +int ext4_da_get_block_prep(struct inode *inode, sector_t iblock, + struct buffer_head *bh, int create) { struct ext4_map_blocks map; int ret = 0; @@ -1917,15 +1928,29 @@ static int __ext4_journalled_writepage(struct page *page, { struct address_space *mapping = page->mapping; struct inode *inode = mapping->host; - struct buffer_head *page_bufs; + struct buffer_head *page_bufs = NULL; handle_t *handle = NULL; - int ret = 0; - int err; + int ret = 0, err = 0; + int inline_data = ext4_has_inline_data(inode); + struct buffer_head *inode_bh = NULL; ClearPageChecked(page); - page_bufs = page_buffers(page); - BUG_ON(!page_bufs); - walk_page_buffers(handle, page_bufs, 0, len, NULL, bget_one); + + if (inline_data) { + BUG_ON(page->index != 0); + BUG_ON(len > ext4_get_max_inline_size(inode)); + inode_bh = ext4_journalled_write_inline_data(inode, len, page); + if (inode_bh == NULL) + goto out; + } else { + page_bufs = page_buffers(page); + if (!page_bufs) { + BUG(); + goto out; + } + ext4_walk_page_buffers(handle, page_bufs, 0, len, + NULL, bget_one); + } /* As soon as we unlock the page, it can go away, but we have * references to buffers so we are safe */ unlock_page(page); @@ -1938,11 +1963,18 @@ static int __ext4_journalled_writepage(struct page *page, BUG_ON(!ext4_handle_valid(handle)); - ret = walk_page_buffers(handle, page_bufs, 0, len, NULL, - do_journal_get_write_access); + if (inline_data) { + ret = ext4_journal_get_write_access(handle, inode_bh); + + err = ext4_handle_dirty_metadata(handle, inode, inode_bh); - err = walk_page_buffers(handle, page_bufs, 0, len, NULL, - write_end_fn); + } else { + ret = ext4_walk_page_buffers(handle, page_bufs, 0, len, NULL, + do_journal_get_write_access); + + err = ext4_walk_page_buffers(handle, page_bufs, 0, len, NULL, + write_end_fn); + } if (ret == 0) ret = err; EXT4_I(inode)->i_datasync_tid = handle->h_transaction->t_tid; @@ -1950,9 +1982,12 @@ static int __ext4_journalled_writepage(struct page *page, if (!ret) ret = err; - walk_page_buffers(handle, page_bufs, 0, len, NULL, bput_one); + if (!ext4_has_inline_data(inode)) + ext4_walk_page_buffers(handle, page_bufs, 0, len, + NULL, bput_one); ext4_set_inode_state(inode, EXT4_STATE_JDATA); out: + brelse(inode_bh); return ret; } @@ -2029,8 +2064,8 @@ static int ext4_writepage(struct page *page, commit_write = 1; } page_bufs = page_buffers(page); - if (walk_page_buffers(NULL, page_bufs, 0, len, NULL, - ext4_bh_delay_or_unwritten)) { + if (ext4_walk_page_buffers(NULL, page_bufs, 0, len, NULL, + ext4_bh_delay_or_unwritten)) { /* * We don't want to do block allocation, so redirty * the page and return. We may reach here when we do @@ -2096,7 +2131,8 @@ static int ext4_da_writepages_trans_blocks(struct inode *inode) * mpage_da_map_and_submit to map a single contiguous memory region * and then write them. */ -static int write_cache_pages_da(struct address_space *mapping, +static int write_cache_pages_da(handle_t *handle, + struct address_space *mapping, struct writeback_control *wbc, struct mpage_da_data *mpd, pgoff_t *done_index) @@ -2175,6 +2211,17 @@ static int write_cache_pages_da(struct address_space *mapping, wait_on_page_writeback(page); BUG_ON(PageWriteback(page)); + /* + * If we have inline data and arrive here, it means that + * we will soon create the block for the 1st page, so + * we'd better clear the inline data here. + */ + if (ext4_has_inline_data(inode)) { + BUG_ON(ext4_test_inode_state(inode, + EXT4_STATE_MAY_INLINE_DATA)); + ext4_destroy_inline_data(handle, inode); + } + if (mpd->next_page != page->index) mpd->first_page = page->index; mpd->next_page = page->index + 1; @@ -2381,7 +2428,8 @@ retry: * contiguous region of logical blocks that need * blocks to be allocated by ext4 and submit them. */ - ret = write_cache_pages_da(mapping, wbc, &mpd, &done_index); + ret = write_cache_pages_da(handle, mapping, + wbc, &mpd, &done_index); /* * If we have a contiguous extent of pages and we * haven't done the I/O yet, map the blocks and submit @@ -2445,7 +2493,6 @@ out_writepages: return ret; } -#define FALL_BACK_TO_NONDELALLOC 1 static int ext4_nonda_switch(struct super_block *sb) { s64 free_blocks, dirty_blocks; @@ -2502,6 +2549,19 @@ static int ext4_da_write_begin(struct file *file, struct address_space *mapping, } *fsdata = (void *)0; trace_ext4_da_write_begin(inode, pos, len, flags); + + if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) { + ret = ext4_da_write_inline_data_begin(mapping, inode, + pos, len, flags, + pagep, fsdata); + if (ret < 0) + goto out; + if (ret == 1) { + ret = 0; + goto out; + } + } + retry: /* * With delayed allocation, we don't log the i_disksize update @@ -2603,22 +2663,13 @@ static int ext4_da_write_end(struct file *file, * changes. So let's piggyback the i_disksize mark_inode_dirty * into that. */ - new_i_size = pos + copied; if (copied && new_i_size > EXT4_I(inode)->i_disksize) { - if (ext4_da_should_update_i_disksize(page, end)) { + if (ext4_has_inline_data(inode) || + ext4_da_should_update_i_disksize(page, end)) { down_write(&EXT4_I(inode)->i_data_sem); - if (new_i_size > EXT4_I(inode)->i_disksize) { - /* - * Updating i_disksize when extending file - * without needing block allocation - */ - if (ext4_should_order_data(inode)) - ret = ext4_jbd2_file_inode(handle, - inode); - + if (new_i_size > EXT4_I(inode)->i_disksize) EXT4_I(inode)->i_disksize = new_i_size; - } up_write(&EXT4_I(inode)->i_data_sem); /* We need to mark inode dirty even if * new_i_size is less that inode->i_size @@ -2627,8 +2678,16 @@ static int ext4_da_write_end(struct file *file, ext4_mark_inode_dirty(handle, inode); } } - ret2 = generic_write_end(file, mapping, pos, len, copied, + + if (write_mode != CONVERT_INLINE_DATA && + ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA) && + ext4_has_inline_data(inode)) + ret2 = ext4_da_write_inline_data_end(inode, pos, len, copied, + page); + else + ret2 = generic_write_end(file, mapping, pos, len, copied, page, fsdata); + copied = ret2; if (ret2 < 0) ret = ret2; @@ -2721,6 +2780,12 @@ static sector_t ext4_bmap(struct address_space *mapping, sector_t block) journal_t *journal; int err; + /* + * We can get here for an inline file via the FIBMAP ioctl + */ + if (ext4_has_inline_data(inode)) + return 0; + if (mapping_tagged(mapping, PAGECACHE_TAG_DIRTY) && test_opt(inode->i_sb, DELALLOC)) { /* @@ -2766,14 +2831,30 @@ static sector_t ext4_bmap(struct address_space *mapping, sector_t block) static int ext4_readpage(struct file *file, struct page *page) { + int ret = -EAGAIN; + struct inode *inode = page->mapping->host; + trace_ext4_readpage(page); - return mpage_readpage(page, ext4_get_block); + + if (ext4_has_inline_data(inode)) + ret = ext4_readpage_inline(inode, page); + + if (ret == -EAGAIN) + return mpage_readpage(page, ext4_get_block); + + return ret; } static int ext4_readpages(struct file *file, struct address_space *mapping, struct list_head *pages, unsigned nr_pages) { + struct inode *inode = mapping->host; + + /* If the file has inline data, no need to do readpages. */ + if (ext4_has_inline_data(inode)) + return 0; + return mpage_readpages(mapping, pages, nr_pages, ext4_get_block); } @@ -2840,7 +2921,7 @@ static int ext4_releasepage(struct page *page, gfp_t wait) * We allocate an uinitialized extent if blocks haven't been allocated. * The extent will be converted to initialized after the IO is complete. */ -static int ext4_get_block_write(struct inode *inode, sector_t iblock, +int ext4_get_block_write(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create) { ext4_debug("ext4_get_block_write: inode %lu, create flag %d\n", @@ -2850,29 +2931,12 @@ static int ext4_get_block_write(struct inode *inode, sector_t iblock, } static int ext4_get_block_write_nolock(struct inode *inode, sector_t iblock, - struct buffer_head *bh_result, int flags) + struct buffer_head *bh_result, int create) { - handle_t *handle = ext4_journal_current_handle(); - struct ext4_map_blocks map; - int ret = 0; - - ext4_debug("ext4_get_block_write_nolock: inode %lu, flag %d\n", - inode->i_ino, flags); - - flags = EXT4_GET_BLOCKS_NO_LOCK; - - map.m_lblk = iblock; - map.m_len = bh_result->b_size >> inode->i_blkbits; - - ret = ext4_map_blocks(handle, inode, &map, flags); - if (ret > 0) { - map_bh(bh_result, inode->i_sb, map.m_pblk); - bh_result->b_state = (bh_result->b_state & ~EXT4_MAP_FLAGS) | - map.m_flags; - bh_result->b_size = inode->i_sb->s_blocksize * map.m_len; - ret = 0; - } - return ret; + ext4_debug("ext4_get_block_write_nolock: inode %lu, create flag %d\n", + inode->i_ino, create); + return _ext4_get_block(inode, iblock, bh_result, + EXT4_GET_BLOCKS_NO_LOCK); } static void ext4_end_io_dio(struct kiocb *iocb, loff_t offset, @@ -2978,10 +3042,10 @@ retry: * fall back to buffered IO. * * For holes, we fallocate those blocks, mark them as uninitialized - * If those blocks were preallocated, we mark sure they are splited, but + * If those blocks were preallocated, we mark sure they are split, but * still keep the range to write as uninitialized. * - * The unwrritten extents will be converted to written when DIO is completed. + * The unwritten 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 conversion * when async direct IO completed. @@ -2999,125 +3063,120 @@ static ssize_t ext4_ext_direct_IO(int rw, struct kiocb *iocb, struct inode *inode = file->f_mapping->host; ssize_t ret; size_t count = iov_length(iov, nr_segs); - + int overwrite = 0; + get_block_t *get_block_func = NULL; + int dio_flags = 0; loff_t final_size = offset + count; - if (rw == WRITE && final_size <= inode->i_size) { - int overwrite = 0; - BUG_ON(iocb->private == NULL); + /* Use the old path for reads and writes beyond i_size. */ + if (rw != WRITE || final_size > inode->i_size) + return ext4_ind_direct_IO(rw, iocb, iov, offset, nr_segs); - /* If we do a overwrite dio, i_mutex locking can be released */ - overwrite = *((int *)iocb->private); + BUG_ON(iocb->private == NULL); - if (overwrite) { - atomic_inc(&inode->i_dio_count); - down_read(&EXT4_I(inode)->i_data_sem); - mutex_unlock(&inode->i_mutex); - } + /* If we do a overwrite dio, i_mutex locking can be released */ + overwrite = *((int *)iocb->private); - /* - * We could direct write to holes and fallocate. - * - * Allocated blocks to fill the hole are marked as uninitialized - * to prevent parallel buffered read to expose the stale data - * before DIO complete the data IO. - * - * As to previously fallocated extents, ext4 get_block - * will just simply mark the buffer mapped but still - * keep the extents uninitialized. - * - * for non AIO case, we will convert those unwritten extents - * to written after return back from blockdev_direct_IO. - * - * for async DIO, the conversion needs to be defered when - * the IO is completed. The ext4 end_io callback function - * will be called to take care of the conversion work. - * Here for async case, we allocate an io_end structure to - * hook to the iocb. - */ - iocb->private = NULL; - ext4_inode_aio_set(inode, NULL); - if (!is_sync_kiocb(iocb)) { - ext4_io_end_t *io_end = - ext4_init_io_end(inode, GFP_NOFS); - if (!io_end) { - ret = -ENOMEM; - goto retake_lock; - } - io_end->flag |= EXT4_IO_END_DIRECT; - iocb->private = io_end; - /* - * we save the io structure for current async - * direct IO, so that later ext4_map_blocks() - * could flag the io structure whether there - * is a unwritten extents needs to be converted - * when IO is completed. - */ - ext4_inode_aio_set(inode, io_end); - } + if (overwrite) { + atomic_inc(&inode->i_dio_count); + down_read(&EXT4_I(inode)->i_data_sem); + mutex_unlock(&inode->i_mutex); + } - if (overwrite) - ret = __blockdev_direct_IO(rw, iocb, inode, - inode->i_sb->s_bdev, iov, - offset, nr_segs, - ext4_get_block_write_nolock, - ext4_end_io_dio, - NULL, - 0); - else - ret = __blockdev_direct_IO(rw, iocb, inode, - inode->i_sb->s_bdev, iov, - offset, nr_segs, - ext4_get_block_write, - ext4_end_io_dio, - NULL, - DIO_LOCKING); - if (iocb->private) - ext4_inode_aio_set(inode, NULL); + /* + * We could direct write to holes and fallocate. + * + * Allocated blocks to fill the hole are marked as + * uninitialized to prevent parallel buffered read to expose + * the stale data before DIO complete the data IO. + * + * As to previously fallocated extents, ext4 get_block will + * just simply mark the buffer mapped but still keep the + * extents uninitialized. + * + * For non AIO case, we will convert those unwritten extents + * to written after return back from blockdev_direct_IO. + * + * For async DIO, the conversion needs to be deferred when the + * IO is completed. The ext4 end_io callback function will be + * called to take care of the conversion work. Here for async + * case, we allocate an io_end structure to hook to the iocb. + */ + iocb->private = NULL; + ext4_inode_aio_set(inode, NULL); + if (!is_sync_kiocb(iocb)) { + ext4_io_end_t *io_end = ext4_init_io_end(inode, GFP_NOFS); + if (!io_end) { + ret = -ENOMEM; + goto retake_lock; + } + io_end->flag |= EXT4_IO_END_DIRECT; + iocb->private = io_end; /* - * The io_end structure takes a reference to the inode, - * that structure needs to be destroyed and the - * reference to the inode need to be dropped, when IO is - * complete, even with 0 byte write, or failed. - * - * In the successful AIO DIO case, the io_end structure will be - * desctroyed and the reference to the inode will be dropped - * after the end_io call back function is called. - * - * In the case there is 0 byte write, or error case, since - * VFS direct IO won't invoke the end_io call back function, - * we need to free the end_io structure here. + * we save the io structure for current async direct + * IO, so that later ext4_map_blocks() could flag the + * io structure whether there is a unwritten extents + * needs to be converted when IO is completed. */ - if (ret != -EIOCBQUEUED && ret <= 0 && iocb->private) { - ext4_free_io_end(iocb->private); - iocb->private = NULL; - } else if (ret > 0 && !overwrite && ext4_test_inode_state(inode, - EXT4_STATE_DIO_UNWRITTEN)) { - int err; - /* - * for non AIO case, since the IO is already - * completed, we could do the conversion right here - */ - err = ext4_convert_unwritten_extents(inode, - offset, ret); - if (err < 0) - ret = err; - ext4_clear_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN); - } + ext4_inode_aio_set(inode, io_end); + } - retake_lock: - /* take i_mutex locking again if we do a ovewrite dio */ - if (overwrite) { - inode_dio_done(inode); - up_read(&EXT4_I(inode)->i_data_sem); - mutex_lock(&inode->i_mutex); - } + if (overwrite) { + get_block_func = ext4_get_block_write_nolock; + } else { + get_block_func = ext4_get_block_write; + dio_flags = DIO_LOCKING; + } + ret = __blockdev_direct_IO(rw, iocb, inode, + inode->i_sb->s_bdev, iov, + offset, nr_segs, + get_block_func, + ext4_end_io_dio, + NULL, + dio_flags); + + if (iocb->private) + ext4_inode_aio_set(inode, NULL); + /* + * The io_end structure takes a reference to the inode, that + * structure needs to be destroyed and the reference to the + * inode need to be dropped, when IO is complete, even with 0 + * byte write, or failed. + * + * In the successful AIO DIO case, the io_end structure will + * be destroyed and the reference to the inode will be dropped + * after the end_io call back function is called. + * + * In the case there is 0 byte write, or error case, since VFS + * direct IO won't invoke the end_io call back function, we + * need to free the end_io structure here. + */ + if (ret != -EIOCBQUEUED && ret <= 0 && iocb->private) { + ext4_free_io_end(iocb->private); + iocb->private = NULL; + } else if (ret > 0 && !overwrite && ext4_test_inode_state(inode, + EXT4_STATE_DIO_UNWRITTEN)) { + int err; + /* + * for non AIO case, since the IO is already + * completed, we could do the conversion right here + */ + err = ext4_convert_unwritten_extents(inode, + offset, ret); + if (err < 0) + ret = err; + ext4_clear_inode_state(inode, EXT4_STATE_DIO_UNWRITTEN); + } - return ret; +retake_lock: + /* take i_mutex locking again if we do a ovewrite dio */ + if (overwrite) { + inode_dio_done(inode); + up_read(&EXT4_I(inode)->i_data_sem); + mutex_lock(&inode->i_mutex); } - /* for write the the end of file case, we fall back to old way */ - return ext4_ind_direct_IO(rw, iocb, iov, offset, nr_segs); + return ret; } static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, @@ -3134,6 +3193,10 @@ static ssize_t ext4_direct_IO(int rw, struct kiocb *iocb, if (ext4_should_journal_data(inode)) return 0; + /* Let buffer I/O handle the inline data case. */ + if (ext4_has_inline_data(inode)) + return 0; + trace_ext4_direct_IO_enter(inode, offset, iov_length(iov, nr_segs), rw); if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) ret = ext4_ext_direct_IO(rw, iocb, iov, offset, nr_segs); @@ -3531,6 +3594,14 @@ void ext4_truncate(struct inode *inode) if (inode->i_size == 0 && !test_opt(inode->i_sb, NO_AUTO_DA_ALLOC)) ext4_set_inode_state(inode, EXT4_STATE_DA_ALLOC_CLOSE); + if (ext4_has_inline_data(inode)) { + int has_inline = 1; + + ext4_inline_data_truncate(inode, &has_inline); + if (has_inline) + return; + } + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) ext4_ext_truncate(inode); else @@ -3756,6 +3827,19 @@ static blkcnt_t ext4_inode_blocks(struct ext4_inode *raw_inode, } } +static inline void ext4_iget_extra_inode(struct inode *inode, + struct ext4_inode *raw_inode, + struct ext4_inode_info *ei) +{ + __le32 *magic = (void *)raw_inode + + EXT4_GOOD_OLD_INODE_SIZE + ei->i_extra_isize; + if (*magic == cpu_to_le32(EXT4_XATTR_MAGIC)) { + ext4_set_inode_state(inode, EXT4_STATE_XATTR); + ext4_find_inline_data_nolock(inode); + } else + EXT4_I(inode)->i_inline_off = 0; +} + struct inode *ext4_iget(struct super_block *sb, unsigned long ino) { struct ext4_iloc iloc; @@ -3826,6 +3910,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) set_nlink(inode, le16_to_cpu(raw_inode->i_links_count)); ext4_clear_state_flags(ei); /* Only relevant on 32-bit archs */ + ei->i_inline_off = 0; ei->i_dir_start_lookup = 0; ei->i_dtime = le32_to_cpu(raw_inode->i_dtime); /* We now have enough fields to check if the inode was active or not. @@ -3898,11 +3983,7 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) ei->i_extra_isize = sizeof(struct ext4_inode) - EXT4_GOOD_OLD_INODE_SIZE; } else { - __le32 *magic = (void *)raw_inode + - EXT4_GOOD_OLD_INODE_SIZE + - ei->i_extra_isize; - if (*magic == cpu_to_le32(EXT4_XATTR_MAGIC)) - ext4_set_inode_state(inode, EXT4_STATE_XATTR); + ext4_iget_extra_inode(inode, raw_inode, ei); } } @@ -3925,17 +4006,19 @@ struct inode *ext4_iget(struct super_block *sb, unsigned long ino) ei->i_file_acl); ret = -EIO; goto bad_inode; - } else if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { - if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || - (S_ISLNK(inode->i_mode) && - !ext4_inode_is_fast_symlink(inode))) - /* Validate extent which is part of inode */ - ret = ext4_ext_check_inode(inode); - } else if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || - (S_ISLNK(inode->i_mode) && - !ext4_inode_is_fast_symlink(inode))) { - /* Validate block references which are part of inode */ - ret = ext4_ind_check_inode(inode); + } else if (!ext4_has_inline_data(inode)) { + if (ext4_test_inode_flag(inode, EXT4_INODE_EXTENTS)) { + if ((S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || + (S_ISLNK(inode->i_mode) && + !ext4_inode_is_fast_symlink(inode)))) + /* Validate extent which is part of inode */ + ret = ext4_ext_check_inode(inode); + } else if (S_ISREG(inode->i_mode) || S_ISDIR(inode->i_mode) || + (S_ISLNK(inode->i_mode) && + !ext4_inode_is_fast_symlink(inode))) { + /* Validate block references which are part of inode */ + ret = ext4_ind_check_inode(inode); + } } if (ret) goto bad_inode; @@ -4122,9 +4205,10 @@ static int ext4_do_update_inode(handle_t *handle, cpu_to_le32(new_encode_dev(inode->i_rdev)); raw_inode->i_block[2] = 0; } - } else + } else if (!ext4_has_inline_data(inode)) { for (block = 0; block < EXT4_N_BLOCKS; block++) raw_inode->i_block[block] = ei->i_data[block]; + } raw_inode->i_disk_version = cpu_to_le32(inode->i_version); if (ei->i_extra_isize) { @@ -4811,8 +4895,9 @@ int ext4_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) * journal_start/journal_stop which can block and take a long time */ if (page_has_buffers(page)) { - if (!walk_page_buffers(NULL, page_buffers(page), 0, len, NULL, - ext4_bh_unmapped)) { + if (!ext4_walk_page_buffers(NULL, page_buffers(page), + 0, len, NULL, + ext4_bh_unmapped)) { /* Wait so that we don't change page under IO */ wait_on_page_writeback(page); ret = VM_FAULT_LOCKED; @@ -4833,7 +4918,7 @@ retry_alloc: } ret = __block_page_mkwrite(vma, vmf, get_block); if (!ret && ext4_should_journal_data(inode)) { - if (walk_page_buffers(handle, page_buffers(page), 0, + if (ext4_walk_page_buffers(handle, page_buffers(page), 0, PAGE_CACHE_SIZE, NULL, do_journal_get_write_access)) { unlock_page(page); ret = VM_FAULT_SIGBUS; diff --git a/fs/ext4/mballoc.c b/fs/ext4/mballoc.c index 526e55358606..1bf6fe785c4f 100644 --- a/fs/ext4/mballoc.c +++ b/fs/ext4/mballoc.c @@ -1373,7 +1373,7 @@ static int mb_find_extent(struct ext4_buddy *e4b, int block, ex->fe_start += next; while (needed > ex->fe_len && - (buddy = mb_find_buddy(e4b, order, &max))) { + mb_find_buddy(e4b, order, &max)) { if (block + 1 >= max) break; @@ -2607,9 +2607,17 @@ static void ext4_free_data_callback(struct super_block *sb, mb_debug(1, "gonna free %u blocks in group %u (0x%p):", entry->efd_count, entry->efd_group, entry); - if (test_opt(sb, DISCARD)) - ext4_issue_discard(sb, entry->efd_group, - entry->efd_start_cluster, entry->efd_count); + if (test_opt(sb, DISCARD)) { + err = ext4_issue_discard(sb, entry->efd_group, + entry->efd_start_cluster, + entry->efd_count); + if (err && err != -EOPNOTSUPP) + ext4_msg(sb, KERN_WARNING, "discard request in" + " group:%d block:%d count:%d failed" + " with %d", entry->efd_group, + entry->efd_start_cluster, + entry->efd_count, err); + } err = ext4_mb_load_buddy(sb, entry->efd_group, &e4b); /* we expect to find existing buddy because it's pinned */ @@ -4310,8 +4318,10 @@ ext4_fsblk_t ext4_mb_new_blocks(handle_t *handle, repeat: /* allocate space in core */ *errp = ext4_mb_regular_allocator(ac); - if (*errp) + if (*errp) { + ext4_discard_allocated_blocks(ac); goto errout; + } /* as we've just preallocated more space than * user requested orinally, we store allocated @@ -4333,10 +4343,10 @@ repeat: ac->ac_b_ex.fe_len = 0; ac->ac_status = AC_STATUS_CONTINUE; goto repeat; - } else if (*errp) - errout: + } else if (*errp) { ext4_discard_allocated_blocks(ac); - else { + goto errout; + } else { block = ext4_grp_offs_to_block(sb, &ac->ac_b_ex); ar->len = ac->ac_b_ex.fe_len; } @@ -4347,6 +4357,7 @@ repeat: *errp = -ENOSPC; } +errout: if (*errp) { ac->ac_b_ex.fe_len = 0; ar->len = 0; @@ -4656,8 +4667,16 @@ do_more: * with group lock held. generate_buddy look at * them with group lock_held */ - if (test_opt(sb, DISCARD)) - ext4_issue_discard(sb, block_group, bit, count); + if (test_opt(sb, DISCARD)) { + err = ext4_issue_discard(sb, block_group, bit, count); + if (err && err != -EOPNOTSUPP) + ext4_msg(sb, KERN_WARNING, "discard request in" + " group:%d block:%d count:%lu failed" + " with %d", block_group, bit, count, + err); + } + + ext4_lock_group(sb, block_group); mb_clear_bits(bitmap_bh->b_data, bit, count_clusters); mb_free_blocks(inode, &e4b, bit, count_clusters); @@ -4851,10 +4870,11 @@ error_return: * one will allocate those blocks, mark it as used in buddy bitmap. This must * be called with under the group lock. */ -static void ext4_trim_extent(struct super_block *sb, int start, int count, +static int 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; trace_ext4_trim_extent(sb, group, start, count); @@ -4870,9 +4890,10 @@ static void ext4_trim_extent(struct super_block *sb, int start, int count, */ mb_mark_used(e4b, &ex); ext4_unlock_group(sb, group); - ext4_issue_discard(sb, group, start, count); + ret = ext4_issue_discard(sb, group, start, count); ext4_lock_group(sb, group); mb_free_blocks(NULL, e4b, start, ex.fe_len); + return ret; } /** @@ -4901,7 +4922,7 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group, void *bitmap; ext4_grpblk_t next, count = 0, free_count = 0; struct ext4_buddy e4b; - int ret; + int ret = 0; trace_ext4_trim_all_free(sb, group, start, max); @@ -4928,8 +4949,11 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group, next = mb_find_next_bit(bitmap, max + 1, start); if ((next - start) >= minblocks) { - ext4_trim_extent(sb, start, - next - start, group, &e4b); + ret = ext4_trim_extent(sb, start, + next - start, group, &e4b); + if (ret && ret != -EOPNOTSUPP) + break; + ret = 0; count += next - start; } free_count += next - start; @@ -4950,8 +4974,10 @@ ext4_trim_all_free(struct super_block *sb, ext4_group_t group, break; } - if (!ret) + if (!ret) { + ret = count; EXT4_MB_GRP_SET_TRIMMED(e4b.bd_info); + } out: ext4_unlock_group(sb, group); ext4_mb_unload_buddy(&e4b); @@ -4959,7 +4985,7 @@ out: ext4_debug("trimmed %d blocks in the group %d\n", count, group); - return count; + return ret; } /** diff --git a/fs/ext4/migrate.c b/fs/ext4/migrate.c index f1bb32ec0169..db8226d595fa 100644 --- a/fs/ext4/migrate.c +++ b/fs/ext4/migrate.c @@ -14,6 +14,7 @@ #include <linux/slab.h> #include "ext4_jbd2.h" +#include "ext4_extents.h" /* * The contiguous blocks details which can be diff --git a/fs/ext4/move_extent.c b/fs/ext4/move_extent.c index 292daeeed455..d9cc5ee42f53 100644 --- a/fs/ext4/move_extent.c +++ b/fs/ext4/move_extent.c @@ -18,6 +18,7 @@ #include <linux/slab.h> #include "ext4_jbd2.h" #include "ext4.h" +#include "ext4_extents.h" /** * get_ext_path - Find an extent path for designated logical block number. diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index 6d600a69fc9d..cac448282331 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -202,13 +202,8 @@ static int ext4_dx_add_entry(handle_t *handle, struct dentry *dentry, struct inode *inode); /* checksumming functions */ -#define EXT4_DIRENT_TAIL(block, blocksize) \ - ((struct ext4_dir_entry_tail *)(((void *)(block)) + \ - ((blocksize) - \ - sizeof(struct ext4_dir_entry_tail)))) - -static void initialize_dirent_tail(struct ext4_dir_entry_tail *t, - unsigned int blocksize) +void initialize_dirent_tail(struct ext4_dir_entry_tail *t, + unsigned int blocksize) { memset(t, 0, sizeof(struct ext4_dir_entry_tail)); t->det_rec_len = ext4_rec_len_to_disk( @@ -261,6 +256,12 @@ static __le32 ext4_dirent_csum(struct inode *inode, return cpu_to_le32(csum); } +static void warn_no_space_for_csum(struct inode *inode) +{ + ext4_warning(inode->i_sb, "no space in directory inode %lu leaf for " + "checksum. Please run e2fsck -D.", inode->i_ino); +} + int ext4_dirent_csum_verify(struct inode *inode, struct ext4_dir_entry *dirent) { struct ext4_dir_entry_tail *t; @@ -271,8 +272,7 @@ int ext4_dirent_csum_verify(struct inode *inode, struct ext4_dir_entry *dirent) t = get_dirent_tail(inode, dirent); if (!t) { - EXT4_ERROR_INODE(inode, "metadata_csum set but no space in dir " - "leaf for checksum. Please run e2fsck -D."); + warn_no_space_for_csum(inode); return 0; } @@ -294,8 +294,7 @@ static void ext4_dirent_csum_set(struct inode *inode, t = get_dirent_tail(inode, dirent); if (!t) { - EXT4_ERROR_INODE(inode, "metadata_csum set but no space in dir " - "leaf for checksum. Please run e2fsck -D."); + warn_no_space_for_csum(inode); return; } @@ -303,9 +302,9 @@ static void ext4_dirent_csum_set(struct inode *inode, (void *)t - (void *)dirent); } -static inline int ext4_handle_dirty_dirent_node(handle_t *handle, - struct inode *inode, - struct buffer_head *bh) +int ext4_handle_dirty_dirent_node(handle_t *handle, + struct inode *inode, + struct buffer_head *bh) { ext4_dirent_csum_set(inode, (struct ext4_dir_entry *)bh->b_data); return ext4_handle_dirty_metadata(handle, inode, bh); @@ -377,8 +376,7 @@ static int ext4_dx_csum_verify(struct inode *inode, count = le16_to_cpu(c->count); if (count_offset + (limit * sizeof(struct dx_entry)) > EXT4_BLOCK_SIZE(inode->i_sb) - sizeof(struct dx_tail)) { - EXT4_ERROR_INODE(inode, "metadata_csum set but no space for " - "tree checksum found. Run e2fsck -D."); + warn_no_space_for_csum(inode); return 1; } t = (struct dx_tail *)(((struct dx_entry *)c) + limit); @@ -408,8 +406,7 @@ static void ext4_dx_csum_set(struct inode *inode, struct ext4_dir_entry *dirent) count = le16_to_cpu(c->count); if (count_offset + (limit * sizeof(struct dx_entry)) > EXT4_BLOCK_SIZE(inode->i_sb) - sizeof(struct dx_tail)) { - EXT4_ERROR_INODE(inode, "metadata_csum set but no space for " - "tree checksum. Run e2fsck -D."); + warn_no_space_for_csum(inode); return; } t = (struct dx_tail *)(((struct dx_entry *)c) + limit); @@ -890,6 +887,7 @@ static int htree_dirblock_to_tree(struct file *dir_file, EXT4_DIR_REC_LEN(0)); for (; de < top; de = ext4_next_entry(de, dir->i_sb->s_blocksize)) { if (ext4_check_dir_entry(dir, NULL, de, bh, + bh->b_data, bh->b_size, (block<<EXT4_BLOCK_SIZE_BITS(dir->i_sb)) + ((char *)de - bh->b_data))) { /* On error, skip the f_pos to the next block. */ @@ -1007,6 +1005,15 @@ errout: return (err); } +static inline int search_dirblock(struct buffer_head *bh, + struct inode *dir, + const struct qstr *d_name, + unsigned int offset, + struct ext4_dir_entry_2 **res_dir) +{ + return search_dir(bh, bh->b_data, dir->i_sb->s_blocksize, dir, + d_name, offset, res_dir); +} /* * Directory block splitting, compacting @@ -1081,13 +1088,6 @@ static void dx_insert_block(struct dx_frame *frame, u32 hash, ext4_lblk_t block) dx_set_count(entries, count + 1); } -static void ext4_update_dx_flag(struct inode *inode) -{ - if (!EXT4_HAS_COMPAT_FEATURE(inode->i_sb, - EXT4_FEATURE_COMPAT_DIR_INDEX)) - ext4_clear_inode_flag(inode, EXT4_INODE_INDEX); -} - /* * NOTE! unlike strncmp, ext4_match returns 1 for success, 0 for failure. * @@ -1107,11 +1107,13 @@ static inline int ext4_match (int len, const char * const name, /* * Returns 0 if not found, -1 on failure, and 1 on success */ -static inline int search_dirblock(struct buffer_head *bh, - struct inode *dir, - const struct qstr *d_name, - unsigned int offset, - struct ext4_dir_entry_2 ** res_dir) +int search_dir(struct buffer_head *bh, + char *search_buf, + int buf_size, + struct inode *dir, + const struct qstr *d_name, + unsigned int offset, + struct ext4_dir_entry_2 **res_dir) { struct ext4_dir_entry_2 * de; char * dlimit; @@ -1119,8 +1121,8 @@ static inline int search_dirblock(struct buffer_head *bh, const char *name = d_name->name; int namelen = d_name->len; - de = (struct ext4_dir_entry_2 *) bh->b_data; - dlimit = bh->b_data + dir->i_sb->s_blocksize; + de = (struct ext4_dir_entry_2 *)search_buf; + dlimit = search_buf + buf_size; while ((char *) de < dlimit) { /* this code is executed quadratically often */ /* do minimal checking `by hand' */ @@ -1128,7 +1130,8 @@ static inline int search_dirblock(struct buffer_head *bh, if ((char *) de + namelen <= dlimit && ext4_match (namelen, name, de)) { /* found a match - just to be sure, do a full check */ - if (ext4_check_dir_entry(dir, NULL, de, bh, offset)) + if (ext4_check_dir_entry(dir, NULL, de, bh, bh->b_data, + bh->b_size, offset)) return -1; *res_dir = de; return 1; @@ -1144,6 +1147,21 @@ static inline int search_dirblock(struct buffer_head *bh, return 0; } +static int is_dx_internal_node(struct inode *dir, ext4_lblk_t block, + struct ext4_dir_entry *de) +{ + struct super_block *sb = dir->i_sb; + + if (!is_dx(dir)) + return 0; + if (block == 0) + return 1; + if (de->inode == 0 && + ext4_rec_len_from_disk(de->rec_len, sb->s_blocksize) == + sb->s_blocksize) + return 1; + return 0; +} /* * ext4_find_entry() @@ -1158,7 +1176,8 @@ static inline int search_dirblock(struct buffer_head *bh, */ static struct buffer_head * ext4_find_entry (struct inode *dir, const struct qstr *d_name, - struct ext4_dir_entry_2 ** res_dir) + struct ext4_dir_entry_2 **res_dir, + int *inlined) { struct super_block *sb; struct buffer_head *bh_use[NAMEI_RA_SIZE]; @@ -1179,6 +1198,18 @@ static struct buffer_head * ext4_find_entry (struct inode *dir, namelen = d_name->len; if (namelen > EXT4_NAME_LEN) return NULL; + + if (ext4_has_inline_data(dir)) { + int has_inline_data = 1; + ret = ext4_find_inline_entry(dir, d_name, res_dir, + &has_inline_data); + if (has_inline_data) { + if (inlined) + *inlined = 1; + return ret; + } + } + if ((namelen <= 2) && (name[0] == '.') && (name[1] == '.' || name[1] == '\0')) { /* @@ -1244,6 +1275,8 @@ restart: goto next; } if (!buffer_verified(bh) && + !is_dx_internal_node(dir, block, + (struct ext4_dir_entry *)bh->b_data) && !ext4_dirent_csum_verify(dir, (struct ext4_dir_entry *)bh->b_data)) { EXT4_ERROR_INODE(dir, "checksumming directory " @@ -1361,7 +1394,7 @@ static struct dentry *ext4_lookup(struct inode *dir, struct dentry *dentry, unsi if (dentry->d_name.len > EXT4_NAME_LEN) return ERR_PTR(-ENAMETOOLONG); - bh = ext4_find_entry(dir, &dentry->d_name, &de); + bh = ext4_find_entry(dir, &dentry->d_name, &de, NULL); inode = NULL; if (bh) { __u32 ino = le32_to_cpu(de->inode); @@ -1395,7 +1428,7 @@ struct dentry *ext4_get_parent(struct dentry *child) struct ext4_dir_entry_2 * de; struct buffer_head *bh; - bh = ext4_find_entry(child->d_inode, &dotdot, &de); + bh = ext4_find_entry(child->d_inode, &dotdot, &de, NULL); if (!bh) return ERR_PTR(-ENOENT); ino = le32_to_cpu(de->inode); @@ -1593,6 +1626,63 @@ errout: return NULL; } +int ext4_find_dest_de(struct inode *dir, struct inode *inode, + struct buffer_head *bh, + void *buf, int buf_size, + const char *name, int namelen, + struct ext4_dir_entry_2 **dest_de) +{ + struct ext4_dir_entry_2 *de; + unsigned short reclen = EXT4_DIR_REC_LEN(namelen); + int nlen, rlen; + unsigned int offset = 0; + char *top; + + de = (struct ext4_dir_entry_2 *)buf; + top = buf + buf_size - reclen; + while ((char *) de <= top) { + if (ext4_check_dir_entry(dir, NULL, de, bh, + buf, buf_size, offset)) + return -EIO; + if (ext4_match(namelen, name, de)) + return -EEXIST; + nlen = EXT4_DIR_REC_LEN(de->name_len); + rlen = ext4_rec_len_from_disk(de->rec_len, buf_size); + if ((de->inode ? rlen - nlen : rlen) >= reclen) + break; + de = (struct ext4_dir_entry_2 *)((char *)de + rlen); + offset += rlen; + } + if ((char *) de > top) + return -ENOSPC; + + *dest_de = de; + return 0; +} + +void ext4_insert_dentry(struct inode *inode, + struct ext4_dir_entry_2 *de, + int buf_size, + const char *name, int namelen) +{ + + int nlen, rlen; + + nlen = EXT4_DIR_REC_LEN(de->name_len); + rlen = ext4_rec_len_from_disk(de->rec_len, buf_size); + if (de->inode) { + struct ext4_dir_entry_2 *de1 = + (struct ext4_dir_entry_2 *)((char *)de + nlen); + de1->rec_len = ext4_rec_len_to_disk(rlen - nlen, buf_size); + de->rec_len = ext4_rec_len_to_disk(nlen, buf_size); + de = de1; + } + de->file_type = EXT4_FT_UNKNOWN; + de->inode = cpu_to_le32(inode->i_ino); + ext4_set_de_type(inode->i_sb, de, inode->i_mode); + de->name_len = namelen; + memcpy(de->name, name, namelen); +} /* * Add a new entry into a directory (leaf) block. If de is non-NULL, * it points to a directory entry which is guaranteed to be large @@ -1608,12 +1698,10 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, struct inode *dir = dentry->d_parent->d_inode; const char *name = dentry->d_name.name; int namelen = dentry->d_name.len; - unsigned int offset = 0; unsigned int blocksize = dir->i_sb->s_blocksize; unsigned short reclen; - int nlen, rlen, err; - char *top; int csum_size = 0; + int err; if (EXT4_HAS_RO_COMPAT_FEATURE(inode->i_sb, EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) @@ -1621,22 +1709,11 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, reclen = EXT4_DIR_REC_LEN(namelen); if (!de) { - de = (struct ext4_dir_entry_2 *)bh->b_data; - top = bh->b_data + (blocksize - csum_size) - reclen; - while ((char *) de <= top) { - if (ext4_check_dir_entry(dir, NULL, de, bh, offset)) - return -EIO; - if (ext4_match(namelen, name, de)) - return -EEXIST; - nlen = EXT4_DIR_REC_LEN(de->name_len); - rlen = ext4_rec_len_from_disk(de->rec_len, blocksize); - if ((de->inode? rlen - nlen: rlen) >= reclen) - break; - de = (struct ext4_dir_entry_2 *)((char *)de + rlen); - offset += rlen; - } - if ((char *) de > top) - return -ENOSPC; + err = ext4_find_dest_de(dir, inode, + bh, bh->b_data, blocksize - csum_size, + name, namelen, &de); + if (err) + return err; } BUFFER_TRACE(bh, "get_write_access"); err = ext4_journal_get_write_access(handle, bh); @@ -1646,19 +1723,8 @@ static int add_dirent_to_buf(handle_t *handle, struct dentry *dentry, } /* By now the buffer is marked for journaling */ - nlen = EXT4_DIR_REC_LEN(de->name_len); - rlen = ext4_rec_len_from_disk(de->rec_len, blocksize); - if (de->inode) { - struct ext4_dir_entry_2 *de1 = (struct ext4_dir_entry_2 *)((char *)de + nlen); - de1->rec_len = ext4_rec_len_to_disk(rlen - nlen, blocksize); - de->rec_len = ext4_rec_len_to_disk(nlen, blocksize); - de = de1; - } - de->file_type = EXT4_FT_UNKNOWN; - de->inode = cpu_to_le32(inode->i_ino); - ext4_set_de_type(dir->i_sb, de, inode->i_mode); - de->name_len = namelen; - memcpy(de->name, name, namelen); + ext4_insert_dentry(inode, de, blocksize, name, namelen); + /* * XXX shouldn't update any times until successful * completion of syscall, but too many callers depend @@ -1831,6 +1897,17 @@ static int ext4_add_entry(handle_t *handle, struct dentry *dentry, blocksize = sb->s_blocksize; if (!dentry->d_name.len) return -EINVAL; + + if (ext4_has_inline_data(dir)) { + retval = ext4_try_add_inline_entry(handle, dentry, inode); + if (retval < 0) + return retval; + if (retval == 1) { + retval = 0; + return retval; + } + } + if (is_dx(dir)) { retval = ext4_dx_add_entry(handle, dentry, inode); if (!retval || (retval != ERR_BAD_DX_DIR)) @@ -2036,36 +2113,29 @@ cleanup: } /* - * ext4_delete_entry deletes a directory entry by merging it with the - * previous entry + * ext4_generic_delete_entry deletes a directory entry by merging it + * with the previous entry */ -static int ext4_delete_entry(handle_t *handle, - struct inode *dir, - struct ext4_dir_entry_2 *de_del, - struct buffer_head *bh) +int ext4_generic_delete_entry(handle_t *handle, + struct inode *dir, + struct ext4_dir_entry_2 *de_del, + struct buffer_head *bh, + void *entry_buf, + int buf_size, + int csum_size) { struct ext4_dir_entry_2 *de, *pde; unsigned int blocksize = dir->i_sb->s_blocksize; - int csum_size = 0; - int i, err; - - if (EXT4_HAS_RO_COMPAT_FEATURE(dir->i_sb, - EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) - csum_size = sizeof(struct ext4_dir_entry_tail); + int i; i = 0; pde = NULL; - de = (struct ext4_dir_entry_2 *) bh->b_data; - while (i < bh->b_size - csum_size) { - if (ext4_check_dir_entry(dir, NULL, de, bh, i)) + de = (struct ext4_dir_entry_2 *)entry_buf; + while (i < buf_size - csum_size) { + if (ext4_check_dir_entry(dir, NULL, de, bh, + bh->b_data, bh->b_size, i)) return -EIO; if (de == de_del) { - BUFFER_TRACE(bh, "get_write_access"); - err = ext4_journal_get_write_access(handle, bh); - if (unlikely(err)) { - ext4_std_error(dir->i_sb, err); - return err; - } if (pde) pde->rec_len = ext4_rec_len_to_disk( ext4_rec_len_from_disk(pde->rec_len, @@ -2076,12 +2146,6 @@ static int ext4_delete_entry(handle_t *handle, else de->inode = 0; dir->i_version++; - BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); - err = ext4_handle_dirty_dirent_node(handle, dir, bh); - if (unlikely(err)) { - ext4_std_error(dir->i_sb, err); - return err; - } return 0; } i += ext4_rec_len_from_disk(de->rec_len, blocksize); @@ -2091,6 +2155,48 @@ static int ext4_delete_entry(handle_t *handle, return -ENOENT; } +static int ext4_delete_entry(handle_t *handle, + struct inode *dir, + struct ext4_dir_entry_2 *de_del, + struct buffer_head *bh) +{ + int err, csum_size = 0; + + if (ext4_has_inline_data(dir)) { + int has_inline_data = 1; + err = ext4_delete_inline_entry(handle, dir, de_del, bh, + &has_inline_data); + if (has_inline_data) + return err; + } + + if (EXT4_HAS_RO_COMPAT_FEATURE(dir->i_sb, + EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) + csum_size = sizeof(struct ext4_dir_entry_tail); + + BUFFER_TRACE(bh, "get_write_access"); + err = ext4_journal_get_write_access(handle, bh); + if (unlikely(err)) + goto out; + + err = ext4_generic_delete_entry(handle, dir, de_del, + bh, bh->b_data, + dir->i_sb->s_blocksize, csum_size); + if (err) + goto out; + + BUFFER_TRACE(bh, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_dirent_node(handle, dir, bh); + if (unlikely(err)) + goto out; + + return 0; +out: + if (err != -ENOENT) + ext4_std_error(dir->i_sb, err); + return err; +} + /* * DIR_NLINK feature is set if 1) nlinks > EXT4_LINK_MAX or 2) nlinks == 2, * since this indicates that nlinks count was previously 1. @@ -2211,21 +2317,95 @@ retry: return err; } -static int ext4_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) +struct ext4_dir_entry_2 *ext4_init_dot_dotdot(struct inode *inode, + struct ext4_dir_entry_2 *de, + int blocksize, int csum_size, + unsigned int parent_ino, int dotdot_real_len) +{ + de->inode = cpu_to_le32(inode->i_ino); + de->name_len = 1; + de->rec_len = ext4_rec_len_to_disk(EXT4_DIR_REC_LEN(de->name_len), + blocksize); + strcpy(de->name, "."); + ext4_set_de_type(inode->i_sb, de, S_IFDIR); + + de = ext4_next_entry(de, blocksize); + de->inode = cpu_to_le32(parent_ino); + de->name_len = 2; + if (!dotdot_real_len) + de->rec_len = ext4_rec_len_to_disk(blocksize - + (csum_size + EXT4_DIR_REC_LEN(1)), + blocksize); + else + de->rec_len = ext4_rec_len_to_disk( + EXT4_DIR_REC_LEN(de->name_len), blocksize); + strcpy(de->name, ".."); + ext4_set_de_type(inode->i_sb, de, S_IFDIR); + + return ext4_next_entry(de, blocksize); +} + +static int ext4_init_new_dir(handle_t *handle, struct inode *dir, + struct inode *inode) { - handle_t *handle; - struct inode *inode; struct buffer_head *dir_block = NULL; struct ext4_dir_entry_2 *de; struct ext4_dir_entry_tail *t; unsigned int blocksize = dir->i_sb->s_blocksize; int csum_size = 0; - int err, retries = 0; + int err; if (EXT4_HAS_RO_COMPAT_FEATURE(dir->i_sb, EXT4_FEATURE_RO_COMPAT_METADATA_CSUM)) csum_size = sizeof(struct ext4_dir_entry_tail); + if (ext4_test_inode_state(inode, EXT4_STATE_MAY_INLINE_DATA)) { + err = ext4_try_create_inline_dir(handle, dir, inode); + if (err < 0 && err != -ENOSPC) + goto out; + if (!err) + goto out; + } + + inode->i_size = EXT4_I(inode)->i_disksize = blocksize; + dir_block = ext4_bread(handle, inode, 0, 1, &err); + if (!(dir_block = ext4_bread(handle, inode, 0, 1, &err))) { + if (!err) { + err = -EIO; + ext4_error(inode->i_sb, + "Directory hole detected on inode %lu\n", + inode->i_ino); + } + goto out; + } + BUFFER_TRACE(dir_block, "get_write_access"); + err = ext4_journal_get_write_access(handle, dir_block); + if (err) + goto out; + de = (struct ext4_dir_entry_2 *)dir_block->b_data; + ext4_init_dot_dotdot(inode, de, blocksize, csum_size, dir->i_ino, 0); + set_nlink(inode, 2); + if (csum_size) { + t = EXT4_DIRENT_TAIL(dir_block->b_data, blocksize); + initialize_dirent_tail(t, blocksize); + } + + BUFFER_TRACE(dir_block, "call ext4_handle_dirty_metadata"); + err = ext4_handle_dirty_dirent_node(handle, inode, dir_block); + if (err) + goto out; + set_buffer_verified(dir_block); +out: + brelse(dir_block); + return err; +} + +static int ext4_mkdir(struct inode *dir, struct dentry *dentry, umode_t mode) +{ + handle_t *handle; + struct inode *inode; + int err, retries = 0; + if (EXT4_DIR_LINK_MAX(dir)) return -EMLINK; @@ -2249,47 +2429,9 @@ retry: inode->i_op = &ext4_dir_inode_operations; inode->i_fop = &ext4_dir_operations; - inode->i_size = EXT4_I(inode)->i_disksize = inode->i_sb->s_blocksize; - if (!(dir_block = ext4_bread(handle, inode, 0, 1, &err))) { - if (!err) { - err = -EIO; - ext4_error(inode->i_sb, - "Directory hole detected on inode %lu\n", - inode->i_ino); - } - goto out_clear_inode; - } - BUFFER_TRACE(dir_block, "get_write_access"); - err = ext4_journal_get_write_access(handle, dir_block); - if (err) - goto out_clear_inode; - de = (struct ext4_dir_entry_2 *) dir_block->b_data; - de->inode = cpu_to_le32(inode->i_ino); - de->name_len = 1; - de->rec_len = ext4_rec_len_to_disk(EXT4_DIR_REC_LEN(de->name_len), - blocksize); - strcpy(de->name, "."); - ext4_set_de_type(dir->i_sb, de, S_IFDIR); - de = ext4_next_entry(de, blocksize); - de->inode = cpu_to_le32(dir->i_ino); - de->rec_len = ext4_rec_len_to_disk(blocksize - - (csum_size + EXT4_DIR_REC_LEN(1)), - blocksize); - de->name_len = 2; - strcpy(de->name, ".."); - ext4_set_de_type(dir->i_sb, de, S_IFDIR); - set_nlink(inode, 2); - - if (csum_size) { - t = EXT4_DIRENT_TAIL(dir_block->b_data, blocksize); - initialize_dirent_tail(t, blocksize); - } - - BUFFER_TRACE(dir_block, "call ext4_handle_dirty_metadata"); - err = ext4_handle_dirty_dirent_node(handle, inode, dir_block); + err = ext4_init_new_dir(handle, dir, inode); if (err) goto out_clear_inode; - set_buffer_verified(dir_block); err = ext4_mark_inode_dirty(handle, inode); if (!err) err = ext4_add_entry(handle, dentry, inode); @@ -2309,7 +2451,6 @@ out_clear_inode: unlock_new_inode(inode); d_instantiate(dentry, inode); out_stop: - brelse(dir_block); ext4_journal_stop(handle); if (err == -ENOSPC && ext4_should_retry_alloc(dir->i_sb, &retries)) goto retry; @@ -2327,6 +2468,14 @@ static int empty_dir(struct inode *inode) struct super_block *sb; int err = 0; + if (ext4_has_inline_data(inode)) { + int has_inline_data = 1; + + err = empty_inline_dir(inode, &has_inline_data); + if (has_inline_data) + return err; + } + sb = inode->i_sb; if (inode->i_size < EXT4_DIR_REC_LEN(1) + EXT4_DIR_REC_LEN(2) || !(bh = ext4_bread(NULL, inode, 0, 0, &err))) { @@ -2393,7 +2542,8 @@ static int empty_dir(struct inode *inode) set_buffer_verified(bh); de = (struct ext4_dir_entry_2 *) bh->b_data; } - if (ext4_check_dir_entry(inode, NULL, de, bh, offset)) { + if (ext4_check_dir_entry(inode, NULL, de, bh, + bh->b_data, bh->b_size, offset)) { de = (struct ext4_dir_entry_2 *)(bh->b_data + sb->s_blocksize); offset = (offset | (sb->s_blocksize - 1)) + 1; @@ -2579,7 +2729,7 @@ static int ext4_rmdir(struct inode *dir, struct dentry *dentry) return PTR_ERR(handle); retval = -ENOENT; - bh = ext4_find_entry(dir, &dentry->d_name, &de); + bh = ext4_find_entry(dir, &dentry->d_name, &de, NULL); if (!bh) goto end_rmdir; @@ -2644,7 +2794,7 @@ static int ext4_unlink(struct inode *dir, struct dentry *dentry) ext4_handle_sync(handle); retval = -ENOENT; - bh = ext4_find_entry(dir, &dentry->d_name, &de); + bh = ext4_find_entry(dir, &dentry->d_name, &de, NULL); if (!bh) goto end_unlink; @@ -2826,8 +2976,39 @@ retry: return err; } -#define PARENT_INO(buffer, size) \ - (ext4_next_entry((struct ext4_dir_entry_2 *)(buffer), size)->inode) + +/* + * Try to find buffer head where contains the parent block. + * It should be the inode block if it is inlined or the 1st block + * if it is a normal dir. + */ +static struct buffer_head *ext4_get_first_dir_block(handle_t *handle, + struct inode *inode, + int *retval, + struct ext4_dir_entry_2 **parent_de, + int *inlined) +{ + struct buffer_head *bh; + + if (!ext4_has_inline_data(inode)) { + if (!(bh = ext4_bread(handle, inode, 0, 0, retval))) { + if (!*retval) { + *retval = -EIO; + ext4_error(inode->i_sb, + "Directory hole detected on inode %lu\n", + inode->i_ino); + } + return NULL; + } + *parent_de = ext4_next_entry( + (struct ext4_dir_entry_2 *)bh->b_data, + inode->i_sb->s_blocksize); + return bh; + } + + *inlined = 1; + return ext4_get_first_inline_block(inode, parent_de, retval); +} /* * Anybody can rename anything with this: the permission checks are left to the @@ -2841,6 +3022,8 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, struct buffer_head *old_bh, *new_bh, *dir_bh; struct ext4_dir_entry_2 *old_de, *new_de; int retval, force_da_alloc = 0; + int inlined = 0, new_inlined = 0; + struct ext4_dir_entry_2 *parent_de; dquot_initialize(old_dir); dquot_initialize(new_dir); @@ -2860,7 +3043,7 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, if (IS_DIRSYNC(old_dir) || IS_DIRSYNC(new_dir)) ext4_handle_sync(handle); - old_bh = ext4_find_entry(old_dir, &old_dentry->d_name, &old_de); + old_bh = ext4_find_entry(old_dir, &old_dentry->d_name, &old_de, NULL); /* * Check for inode number is _not_ due to possible IO errors. * We might rmdir the source, keep it as pwd of some process @@ -2873,7 +3056,8 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, goto end_rename; new_inode = new_dentry->d_inode; - new_bh = ext4_find_entry(new_dir, &new_dentry->d_name, &new_de); + new_bh = ext4_find_entry(new_dir, &new_dentry->d_name, + &new_de, &new_inlined); if (new_bh) { if (!new_inode) { brelse(new_bh); @@ -2887,22 +3071,17 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, goto end_rename; } retval = -EIO; - if (!(dir_bh = ext4_bread(handle, old_inode, 0, 0, &retval))) { - if (!retval) { - retval = -EIO; - ext4_error(old_inode->i_sb, - "Directory hole detected on inode %lu\n", - old_inode->i_ino); - } + dir_bh = ext4_get_first_dir_block(handle, old_inode, + &retval, &parent_de, + &inlined); + if (!dir_bh) goto end_rename; - } - if (!buffer_verified(dir_bh) && + if (!inlined && !buffer_verified(dir_bh) && !ext4_dirent_csum_verify(old_inode, (struct ext4_dir_entry *)dir_bh->b_data)) goto end_rename; set_buffer_verified(dir_bh); - if (le32_to_cpu(PARENT_INO(dir_bh->b_data, - old_dir->i_sb->s_blocksize)) != old_dir->i_ino) + if (le32_to_cpu(parent_de->inode) != old_dir->i_ino) goto end_rename; retval = -EMLINK; if (!new_inode && new_dir != old_dir && @@ -2931,10 +3110,13 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, ext4_current_time(new_dir); ext4_mark_inode_dirty(handle, new_dir); BUFFER_TRACE(new_bh, "call ext4_handle_dirty_metadata"); - retval = ext4_handle_dirty_dirent_node(handle, new_dir, new_bh); - if (unlikely(retval)) { - ext4_std_error(new_dir->i_sb, retval); - goto end_rename; + if (!new_inlined) { + retval = ext4_handle_dirty_dirent_node(handle, + new_dir, new_bh); + if (unlikely(retval)) { + ext4_std_error(new_dir->i_sb, retval); + goto end_rename; + } } brelse(new_bh); new_bh = NULL; @@ -2962,7 +3144,8 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, struct buffer_head *old_bh2; struct ext4_dir_entry_2 *old_de2; - old_bh2 = ext4_find_entry(old_dir, &old_dentry->d_name, &old_de2); + old_bh2 = ext4_find_entry(old_dir, &old_dentry->d_name, + &old_de2, NULL); if (old_bh2) { retval = ext4_delete_entry(handle, old_dir, old_de2, old_bh2); @@ -2982,17 +3165,19 @@ static int ext4_rename(struct inode *old_dir, struct dentry *old_dentry, old_dir->i_ctime = old_dir->i_mtime = ext4_current_time(old_dir); ext4_update_dx_flag(old_dir); if (dir_bh) { - PARENT_INO(dir_bh->b_data, new_dir->i_sb->s_blocksize) = - cpu_to_le32(new_dir->i_ino); + parent_de->inode = cpu_to_le32(new_dir->i_ino); BUFFER_TRACE(dir_bh, "call ext4_handle_dirty_metadata"); - if (is_dx(old_inode)) { - retval = ext4_handle_dirty_dx_node(handle, - old_inode, - dir_bh); + if (!inlined) { + if (is_dx(old_inode)) { + retval = ext4_handle_dirty_dx_node(handle, + old_inode, + dir_bh); + } else { + retval = ext4_handle_dirty_dirent_node(handle, + old_inode, dir_bh); + } } else { - retval = ext4_handle_dirty_dirent_node(handle, - old_inode, - dir_bh); + retval = ext4_mark_inode_dirty(handle, old_inode); } if (retval) { ext4_std_error(old_dir->i_sb, retval); @@ -3043,23 +3228,19 @@ const struct inode_operations ext4_dir_inode_operations = { .mknod = ext4_mknod, .rename = ext4_rename, .setattr = ext4_setattr, -#ifdef CONFIG_EXT4_FS_XATTR .setxattr = generic_setxattr, .getxattr = generic_getxattr, .listxattr = ext4_listxattr, .removexattr = generic_removexattr, -#endif .get_acl = ext4_get_acl, .fiemap = ext4_fiemap, }; const struct inode_operations ext4_special_inode_operations = { .setattr = ext4_setattr, -#ifdef CONFIG_EXT4_FS_XATTR .setxattr = generic_setxattr, .getxattr = generic_getxattr, .listxattr = ext4_listxattr, .removexattr = generic_removexattr, -#endif .get_acl = ext4_get_acl, }; diff --git a/fs/ext4/page-io.c b/fs/ext4/page-io.c index 68e896e12a67..0016fbca2a40 100644 --- a/fs/ext4/page-io.c +++ b/fs/ext4/page-io.c @@ -27,7 +27,6 @@ #include "ext4_jbd2.h" #include "xattr.h" #include "acl.h" -#include "ext4_extents.h" static struct kmem_cache *io_page_cachep, *io_end_cachep; @@ -111,7 +110,7 @@ static int ext4_end_io(ext4_io_end_t *io) inode_dio_done(inode); /* Wake up anyone waiting on unwritten extent conversion */ if (atomic_dec_and_test(&EXT4_I(inode)->i_unwritten)) - wake_up_all(ext4_ioend_wq(io->inode)); + wake_up_all(ext4_ioend_wq(inode)); return ret; } diff --git a/fs/ext4/resize.c b/fs/ext4/resize.c index 47bf06a2765d..d99387b89edd 100644 --- a/fs/ext4/resize.c +++ b/fs/ext4/resize.c @@ -783,7 +783,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, err = ext4_journal_get_write_access(handle, gdb_bh); if (unlikely(err)) - goto exit_sbh; + goto exit_dind; err = ext4_journal_get_write_access(handle, dind); if (unlikely(err)) @@ -792,7 +792,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, /* ext4_reserve_inode_write() gets a reference on the iloc */ err = ext4_reserve_inode_write(handle, inode, &iloc); if (unlikely(err)) - goto exit_dindj; + goto exit_dind; n_group_desc = ext4_kvmalloc((gdb_num + 1) * sizeof(struct buffer_head *), @@ -846,12 +846,7 @@ static int add_new_gdb(handle_t *handle, struct inode *inode, exit_inode: ext4_kvfree(n_group_desc); - /* ext4_handle_release_buffer(handle, iloc.bh); */ brelse(iloc.bh); -exit_dindj: - /* ext4_handle_release_buffer(handle, dind); */ -exit_sbh: - /* ext4_handle_release_buffer(handle, EXT4_SB(sb)->s_sbh); */ exit_dind: brelse(dind); exit_bh: @@ -969,14 +964,8 @@ static int reserve_backup_gdb(handle_t *handle, struct inode *inode, } for (i = 0; i < reserved_gdb; i++) { - if ((err = ext4_journal_get_write_access(handle, primary[i]))) { - /* - int j; - for (j = 0; j < i; j++) - ext4_handle_release_buffer(handle, primary[j]); - */ + if ((err = ext4_journal_get_write_access(handle, primary[i]))) goto exit_bh; - } } if ((err = ext4_reserve_inode_write(handle, inode, &iloc))) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 80928f716850..3cdb0a2fc648 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -45,7 +45,7 @@ #include <linux/freezer.h> #include "ext4.h" -#include "ext4_extents.h" +#include "ext4_extents.h" /* Needed for trace points definition */ #include "ext4_jbd2.h" #include "xattr.h" #include "acl.h" @@ -939,10 +939,11 @@ static struct inode *ext4_alloc_inode(struct super_block *sb) return NULL; ei->vfs_inode.i_version = 1; - ei->vfs_inode.i_data.writeback_index = 0; memset(&ei->i_cached_extent, 0, sizeof(struct ext4_ext_cache)); INIT_LIST_HEAD(&ei->i_prealloc_list); spin_lock_init(&ei->i_prealloc_lock); + ext4_es_init_tree(&ei->i_es_tree); + rwlock_init(&ei->i_es_lock); ei->i_reserved_data_blocks = 0; ei->i_reserved_meta_blocks = 0; ei->i_allocated_meta_blocks = 0; @@ -996,9 +997,7 @@ static void init_once(void *foo) struct ext4_inode_info *ei = (struct ext4_inode_info *) foo; INIT_LIST_HEAD(&ei->i_orphan); -#ifdef CONFIG_EXT4_FS_XATTR init_rwsem(&ei->xattr_sem); -#endif init_rwsem(&ei->i_data_sem); inode_init_once(&ei->vfs_inode); } @@ -1031,6 +1030,7 @@ void ext4_clear_inode(struct inode *inode) clear_inode(inode); dquot_drop(inode); ext4_discard_preallocations(inode); + ext4_es_remove_extent(inode, 0, EXT_MAX_BLOCKS); if (EXT4_I(inode)->jinode) { jbd2_journal_release_jbd_inode(EXT4_JOURNAL(inode), EXT4_I(inode)->jinode); @@ -1447,13 +1447,8 @@ static const struct mount_opts { {Opt_data_journal, EXT4_MOUNT_JOURNAL_DATA, MOPT_DATAJ}, {Opt_data_ordered, EXT4_MOUNT_ORDERED_DATA, MOPT_DATAJ}, {Opt_data_writeback, EXT4_MOUNT_WRITEBACK_DATA, MOPT_DATAJ}, -#ifdef CONFIG_EXT4_FS_XATTR {Opt_user_xattr, EXT4_MOUNT_XATTR_USER, MOPT_SET}, {Opt_nouser_xattr, EXT4_MOUNT_XATTR_USER, MOPT_CLEAR}, -#else - {Opt_user_xattr, 0, MOPT_NOSUPPORT}, - {Opt_nouser_xattr, 0, MOPT_NOSUPPORT}, -#endif #ifdef CONFIG_EXT4_FS_POSIX_ACL {Opt_acl, EXT4_MOUNT_POSIX_ACL, MOPT_SET}, {Opt_noacl, EXT4_MOUNT_POSIX_ACL, MOPT_CLEAR}, @@ -3202,7 +3197,6 @@ int ext4_calculate_overhead(struct super_block *sb) ext4_fsblk_t overhead = 0; char *buf = (char *) get_zeroed_page(GFP_KERNEL); - memset(buf, 0, PAGE_SIZE); if (!buf) return -ENOMEM; @@ -3256,7 +3250,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) unsigned int i; int needs_recovery, has_huge_files, has_bigalloc; __u64 blocks_count; - int err; + int err = 0; unsigned int journal_ioprio = DEFAULT_JOURNAL_IOPRIO; ext4_group_t first_not_zeroed; @@ -3272,9 +3266,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) } sb->s_fs_info = sbi; sbi->s_sb = sb; - sbi->s_mount_opt = 0; - sbi->s_resuid = make_kuid(&init_user_ns, EXT4_DEF_RESUID); - sbi->s_resgid = make_kgid(&init_user_ns, EXT4_DEF_RESGID); sbi->s_inode_readahead_blks = EXT4_DEF_INODE_READAHEAD_BLKS; sbi->s_sb_block = sb_block; if (sb->s_bdev->bd_part) @@ -3285,6 +3276,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) for (cp = sb->s_id; (cp = strchr(cp, '/'));) *cp = '!'; + /* -EINVAL is default */ ret = -EINVAL; blocksize = sb_min_blocksize(sb, EXT4_MIN_BLOCK_SIZE); if (!blocksize) { @@ -3369,9 +3361,7 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) if (def_mount_opts & EXT4_DEFM_UID16) set_opt(sb, NO_UID32); /* xattr user namespace & acls are now defaulted on */ -#ifdef CONFIG_EXT4_FS_XATTR set_opt(sb, XATTR_USER); -#endif #ifdef CONFIG_EXT4_FS_POSIX_ACL set_opt(sb, POSIX_ACL); #endif @@ -3662,7 +3652,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) " too large to mount safely on this system"); if (sizeof(sector_t) < 8) ext4_msg(sb, KERN_WARNING, "CONFIG_LBDAF not enabled"); - ret = err; goto failed_mount; } @@ -3770,7 +3759,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) } if (err) { ext4_msg(sb, KERN_ERR, "insufficient memory"); - ret = err; goto failed_mount3; } @@ -3801,7 +3789,6 @@ static int ext4_fill_super(struct super_block *sb, void *data, int silent) INIT_LIST_HEAD(&sbi->s_orphan); /* unlinked but open files */ mutex_init(&sbi->s_orphan_lock); - sbi->s_resize_flags = 0; sb->s_root = NULL; @@ -3897,8 +3884,8 @@ no_journal: if (es->s_overhead_clusters) sbi->s_overhead = le32_to_cpu(es->s_overhead_clusters); else { - ret = ext4_calculate_overhead(sb); - if (ret) + err = ext4_calculate_overhead(sb); + if (err) goto failed_mount_wq; } @@ -3910,6 +3897,7 @@ no_journal: alloc_workqueue("ext4-dio-unwritten", WQ_MEM_RECLAIM | WQ_UNBOUND, 1); if (!EXT4_SB(sb)->dio_unwritten_wq) { printk(KERN_ERR "EXT4-fs: failed to create DIO workqueue\n"); + ret = -ENOMEM; goto failed_mount_wq; } @@ -4012,12 +4000,20 @@ no_journal: /* Enable quota usage during mount. */ if (EXT4_HAS_RO_COMPAT_FEATURE(sb, EXT4_FEATURE_RO_COMPAT_QUOTA) && !(sb->s_flags & MS_RDONLY)) { - ret = ext4_enable_quotas(sb); - if (ret) + err = ext4_enable_quotas(sb); + if (err) goto failed_mount7; } #endif /* CONFIG_QUOTA */ + if (test_opt(sb, DISCARD)) { + struct request_queue *q = bdev_get_queue(sb->s_bdev); + if (!blk_queue_discard(q)) + ext4_msg(sb, KERN_WARNING, + "mounting with \"discard\" option, but " + "the device does not support discard"); + } + ext4_msg(sb, KERN_INFO, "mounted filesystem with%s. " "Opts: %s%s%s", descr, sbi->s_es->s_mount_opts, *sbi->s_es->s_mount_opts ? "; " : "", orig_data); @@ -4084,7 +4080,7 @@ out_fail: kfree(sbi); out_free_orig: kfree(orig_data); - return ret; + return err ? err : ret; } /* @@ -4790,7 +4786,7 @@ 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) - EXT4_C2B(sbi, sbi->s_overhead); + buf->f_blocks = ext4_blocks_count(es) - EXT4_C2B(sbi, overhead); bfree = percpu_counter_sum_positive(&sbi->s_freeclusters_counter) - percpu_counter_sum_positive(&sbi->s_dirtyclusters_counter); /* prevent underflow in case that few free space is available */ @@ -5282,6 +5278,7 @@ static int __init ext4_init_fs(void) ext4_li_info = NULL; mutex_init(&ext4_li_mtx); + /* Build-time check for flags consistency */ ext4_check_flag_values(); for (i = 0; i < EXT4_WQ_HASH_SZ; i++) { @@ -5289,9 +5286,14 @@ static int __init ext4_init_fs(void) init_waitqueue_head(&ext4__ioend_wq[i]); } - err = ext4_init_pageio(); + err = ext4_init_es(); if (err) return err; + + err = ext4_init_pageio(); + if (err) + goto out7; + err = ext4_init_system_zone(); if (err) goto out6; @@ -5341,6 +5343,9 @@ out5: ext4_exit_system_zone(); out6: ext4_exit_pageio(); +out7: + ext4_exit_es(); + return err; } diff --git a/fs/ext4/symlink.c b/fs/ext4/symlink.c index ed9354aff279..ff3711932018 100644 --- a/fs/ext4/symlink.c +++ b/fs/ext4/symlink.c @@ -35,22 +35,18 @@ const struct inode_operations ext4_symlink_inode_operations = { .follow_link = page_follow_link_light, .put_link = page_put_link, .setattr = ext4_setattr, -#ifdef CONFIG_EXT4_FS_XATTR .setxattr = generic_setxattr, .getxattr = generic_getxattr, .listxattr = ext4_listxattr, .removexattr = generic_removexattr, -#endif }; const struct inode_operations ext4_fast_symlink_inode_operations = { .readlink = generic_readlink, .follow_link = ext4_follow_link, .setattr = ext4_setattr, -#ifdef CONFIG_EXT4_FS_XATTR .setxattr = generic_setxattr, .getxattr = generic_getxattr, .listxattr = ext4_listxattr, .removexattr = generic_removexattr, -#endif }; diff --git a/fs/ext4/xattr.c b/fs/ext4/xattr.c index 2cdb98d62980..3a91ebc2b66f 100644 --- a/fs/ext4/xattr.c +++ b/fs/ext4/xattr.c @@ -61,11 +61,6 @@ #include "xattr.h" #include "acl.h" -#define BHDR(bh) ((struct ext4_xattr_header *)((bh)->b_data)) -#define ENTRY(ptr) ((struct ext4_xattr_entry *)(ptr)) -#define BFIRST(bh) ENTRY(BHDR(bh)+1) -#define IS_LAST_ENTRY(entry) (*(__u32 *)(entry) == 0) - #ifdef EXT4_XATTR_DEBUG # define ea_idebug(inode, f...) do { \ printk(KERN_DEBUG "inode %s:%lu: ", \ @@ -312,7 +307,7 @@ cleanup: return error; } -static int +int ext4_xattr_ibody_get(struct inode *inode, int name_index, const char *name, void *buffer, size_t buffer_size) { @@ -581,21 +576,6 @@ static size_t ext4_xattr_free_space(struct ext4_xattr_entry *last, return (*min_offs - ((void *)last - base) - sizeof(__u32)); } -struct ext4_xattr_info { - int name_index; - const char *name; - const void *value; - size_t value_len; -}; - -struct ext4_xattr_search { - struct ext4_xattr_entry *first; - void *base; - void *end; - struct ext4_xattr_entry *here; - int not_found; -}; - static int ext4_xattr_set_entry(struct ext4_xattr_info *i, struct ext4_xattr_search *s) { @@ -648,9 +628,14 @@ ext4_xattr_set_entry(struct ext4_xattr_info *i, struct ext4_xattr_search *s) size. Just replace. */ s->here->e_value_size = cpu_to_le32(i->value_len); - memset(val + size - EXT4_XATTR_PAD, 0, - EXT4_XATTR_PAD); /* Clear pad bytes. */ - memcpy(val, i->value, i->value_len); + if (i->value == EXT4_ZERO_XATTR_VALUE) { + memset(val, 0, size); + } else { + /* Clear pad bytes first. */ + memset(val + size - EXT4_XATTR_PAD, 0, + EXT4_XATTR_PAD); + memcpy(val, i->value, i->value_len); + } return 0; } @@ -689,9 +674,14 @@ ext4_xattr_set_entry(struct ext4_xattr_info *i, struct ext4_xattr_search *s) size_t size = EXT4_XATTR_SIZE(i->value_len); void *val = s->base + min_offs - size; s->here->e_value_offs = cpu_to_le16(min_offs - size); - memset(val + size - EXT4_XATTR_PAD, 0, - EXT4_XATTR_PAD); /* Clear the pad bytes. */ - memcpy(val, i->value, i->value_len); + if (i->value == EXT4_ZERO_XATTR_VALUE) { + memset(val, 0, size); + } else { + /* Clear the pad bytes first. */ + memset(val + size - EXT4_XATTR_PAD, 0, + EXT4_XATTR_PAD); + memcpy(val, i->value, i->value_len); + } } } return 0; @@ -794,7 +784,6 @@ ext4_xattr_block_set(handle_t *handle, struct inode *inode, int offset = (char *)s->here - bs->bh->b_data; unlock_buffer(bs->bh); - ext4_handle_release_buffer(handle, bs->bh); if (ce) { mb_cache_entry_release(ce); ce = NULL; @@ -950,14 +939,8 @@ bad_block: #undef header } -struct ext4_xattr_ibody_find { - struct ext4_xattr_search s; - struct ext4_iloc iloc; -}; - -static int -ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i, - struct ext4_xattr_ibody_find *is) +int ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is) { struct ext4_xattr_ibody_header *header; struct ext4_inode *raw_inode; @@ -985,10 +968,47 @@ ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i, return 0; } -static int -ext4_xattr_ibody_set(handle_t *handle, struct inode *inode, - struct ext4_xattr_info *i, - struct ext4_xattr_ibody_find *is) +int ext4_xattr_ibody_inline_set(handle_t *handle, struct inode *inode, + struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is) +{ + struct ext4_xattr_ibody_header *header; + struct ext4_xattr_search *s = &is->s; + int error; + + if (EXT4_I(inode)->i_extra_isize == 0) + return -ENOSPC; + error = ext4_xattr_set_entry(i, s); + if (error) { + if (error == -ENOSPC && + ext4_has_inline_data(inode)) { + error = ext4_try_to_evict_inline_data(handle, inode, + EXT4_XATTR_LEN(strlen(i->name) + + EXT4_XATTR_SIZE(i->value_len))); + if (error) + return error; + error = ext4_xattr_ibody_find(inode, i, is); + if (error) + return error; + error = ext4_xattr_set_entry(i, s); + } + if (error) + return error; + } + header = IHDR(inode, ext4_raw_inode(&is->iloc)); + if (!IS_LAST_ENTRY(s->first)) { + header->h_magic = cpu_to_le32(EXT4_XATTR_MAGIC); + ext4_set_inode_state(inode, EXT4_STATE_XATTR); + } else { + header->h_magic = cpu_to_le32(0); + ext4_clear_inode_state(inode, EXT4_STATE_XATTR); + } + return 0; +} + +static int ext4_xattr_ibody_set(handle_t *handle, struct inode *inode, + struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is) { struct ext4_xattr_ibody_header *header; struct ext4_xattr_search *s = &is->s; @@ -1144,9 +1164,17 @@ ext4_xattr_set(struct inode *inode, int name_index, const char *name, { handle_t *handle; int error, retries = 0; + int credits = EXT4_DATA_TRANS_BLOCKS(inode->i_sb); retry: - handle = ext4_journal_start(inode, EXT4_DATA_TRANS_BLOCKS(inode->i_sb)); + /* + * In case of inline data, we may push out the data to a block, + * So reserve the journal space first. + */ + if (ext4_has_inline_data(inode)) + credits += ext4_writepage_trans_blocks(inode) + 1; + + handle = ext4_journal_start(inode, credits); if (IS_ERR(handle)) { error = PTR_ERR(handle); } else { diff --git a/fs/ext4/xattr.h b/fs/ext4/xattr.h index 91f31ca7d9af..69eda787a96a 100644 --- a/fs/ext4/xattr.h +++ b/fs/ext4/xattr.h @@ -21,6 +21,7 @@ #define EXT4_XATTR_INDEX_TRUSTED 4 #define EXT4_XATTR_INDEX_LUSTRE 5 #define EXT4_XATTR_INDEX_SECURITY 6 +#define EXT4_XATTR_INDEX_SYSTEM 7 struct ext4_xattr_header { __le32 h_magic; /* magic number for identification */ @@ -65,7 +66,32 @@ struct ext4_xattr_entry { EXT4_I(inode)->i_extra_isize)) #define IFIRST(hdr) ((struct ext4_xattr_entry *)((hdr)+1)) -# ifdef CONFIG_EXT4_FS_XATTR +#define BHDR(bh) ((struct ext4_xattr_header *)((bh)->b_data)) +#define ENTRY(ptr) ((struct ext4_xattr_entry *)(ptr)) +#define BFIRST(bh) ENTRY(BHDR(bh)+1) +#define IS_LAST_ENTRY(entry) (*(__u32 *)(entry) == 0) + +#define EXT4_ZERO_XATTR_VALUE ((void *)-1) + +struct ext4_xattr_info { + int name_index; + const char *name; + const void *value; + size_t value_len; +}; + +struct ext4_xattr_search { + struct ext4_xattr_entry *first; + void *base; + void *end; + struct ext4_xattr_entry *here; + int not_found; +}; + +struct ext4_xattr_ibody_find { + struct ext4_xattr_search s; + struct ext4_iloc iloc; +}; extern const struct xattr_handler ext4_xattr_user_handler; extern const struct xattr_handler ext4_xattr_trusted_handler; @@ -90,60 +116,82 @@ extern void ext4_exit_xattr(void); extern const struct xattr_handler *ext4_xattr_handlers[]; -# else /* CONFIG_EXT4_FS_XATTR */ - -static inline int -ext4_xattr_get(struct inode *inode, int name_index, const char *name, - void *buffer, size_t size, int flags) -{ - return -EOPNOTSUPP; -} - -static inline int -ext4_xattr_set(struct inode *inode, int name_index, const char *name, - const void *value, size_t size, int flags) -{ - return -EOPNOTSUPP; -} - -static inline int -ext4_xattr_set_handle(handle_t *handle, struct inode *inode, int name_index, - const char *name, const void *value, size_t size, int flags) -{ - return -EOPNOTSUPP; -} - -static inline void -ext4_xattr_delete_inode(handle_t *handle, struct inode *inode) -{ -} - -static inline void -ext4_xattr_put_super(struct super_block *sb) -{ -} - -static __init inline int -ext4_init_xattr(void) -{ - return 0; -} - -static inline void -ext4_exit_xattr(void) -{ -} - -static inline int -ext4_expand_extra_isize_ea(struct inode *inode, int new_extra_isize, - struct ext4_inode *raw_inode, handle_t *handle) -{ - return -EOPNOTSUPP; -} - -#define ext4_xattr_handlers NULL - -# endif /* CONFIG_EXT4_FS_XATTR */ +extern int ext4_xattr_ibody_find(struct inode *inode, struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is); +extern int ext4_xattr_ibody_get(struct inode *inode, int name_index, + const char *name, + void *buffer, size_t buffer_size); +extern int ext4_xattr_ibody_inline_set(handle_t *handle, struct inode *inode, + struct ext4_xattr_info *i, + struct ext4_xattr_ibody_find *is); + +extern int ext4_has_inline_data(struct inode *inode); +extern int ext4_get_inline_size(struct inode *inode); +extern int ext4_get_max_inline_size(struct inode *inode); +extern int ext4_find_inline_data_nolock(struct inode *inode); +extern void ext4_write_inline_data(struct inode *inode, + struct ext4_iloc *iloc, + void *buffer, loff_t pos, + unsigned int len); +extern int ext4_prepare_inline_data(handle_t *handle, struct inode *inode, + unsigned int len); +extern int ext4_init_inline_data(handle_t *handle, struct inode *inode, + unsigned int len); +extern int ext4_destroy_inline_data(handle_t *handle, struct inode *inode); + +extern int ext4_readpage_inline(struct inode *inode, struct page *page); +extern int ext4_try_to_write_inline_data(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + unsigned flags, + struct page **pagep); +extern int ext4_write_inline_data_end(struct inode *inode, + loff_t pos, unsigned len, + unsigned copied, + struct page *page); +extern struct buffer_head * +ext4_journalled_write_inline_data(struct inode *inode, + unsigned len, + struct page *page); +extern int ext4_da_write_inline_data_begin(struct address_space *mapping, + struct inode *inode, + loff_t pos, unsigned len, + unsigned flags, + struct page **pagep, + void **fsdata); +extern int ext4_da_write_inline_data_end(struct inode *inode, loff_t pos, + unsigned len, unsigned copied, + struct page *page); +extern int ext4_try_add_inline_entry(handle_t *handle, struct dentry *dentry, + struct inode *inode); +extern int ext4_try_create_inline_dir(handle_t *handle, + struct inode *parent, + struct inode *inode); +extern int ext4_read_inline_dir(struct file *filp, + void *dirent, filldir_t filldir, + int *has_inline_data); +extern struct buffer_head *ext4_find_inline_entry(struct inode *dir, + const struct qstr *d_name, + struct ext4_dir_entry_2 **res_dir, + int *has_inline_data); +extern int ext4_delete_inline_entry(handle_t *handle, + struct inode *dir, + struct ext4_dir_entry_2 *de_del, + struct buffer_head *bh, + int *has_inline_data); +extern int empty_inline_dir(struct inode *dir, int *has_inline_data); +extern struct buffer_head *ext4_get_first_inline_block(struct inode *inode, + struct ext4_dir_entry_2 **parent_de, + int *retval); +extern int ext4_inline_data_fiemap(struct inode *inode, + struct fiemap_extent_info *fieinfo, + int *has_inline); +extern int ext4_try_to_evict_inline_data(handle_t *handle, + struct inode *inode, + int needed); +extern void ext4_inline_data_truncate(struct inode *inode, int *has_inline); + +extern int ext4_convert_inline_data(struct inode *inode); #ifdef CONFIG_EXT4_FS_SECURITY extern int ext4_init_security(handle_t *handle, struct inode *inode, diff --git a/fs/fhandle.c b/fs/fhandle.c index f775bfdd6e4a..cccdc874bb55 100644 --- a/fs/fhandle.c +++ b/fs/fhandle.c @@ -22,7 +22,7 @@ static long do_sys_name_to_handle(struct path *path, struct file_handle *handle = NULL; /* - * We need t make sure wether the file system + * We need to make sure whether the file system * support decoding of the file handle */ if (!path->dentry->d_sb->s_export_op || @@ -40,7 +40,7 @@ static long do_sys_name_to_handle(struct path *path, if (!handle) return -ENOMEM; - /* convert handle size to multiple of sizeof(u32) */ + /* convert handle size to multiple of sizeof(u32) */ handle_dwords = f_handle.handle_bytes >> 2; /* we ask for a non connected handle */ diff --git a/fs/file.c b/fs/file.c index 7cb71b992603..15cb8618e95d 100644 --- a/fs/file.c +++ b/fs/file.c @@ -519,12 +519,6 @@ struct files_struct init_files = { .file_lock = __SPIN_LOCK_UNLOCKED(init_task.file_lock), }; -void daemonize_descriptors(void) -{ - atomic_inc(&init_files.count); - reset_files_struct(&init_files); -} - /* * allocate a file descriptor, mark it busy. */ @@ -994,16 +988,18 @@ int iterate_fd(struct files_struct *files, unsigned n, const void *p) { struct fdtable *fdt; - struct file *file; int res = 0; if (!files) return 0; spin_lock(&files->file_lock); - fdt = files_fdtable(files); - while (!res && n < fdt->max_fds) { - file = rcu_dereference_check_fdtable(files, fdt->fd[n++]); - if (file) - res = f(p, file, n); + for (fdt = files_fdtable(files); n < fdt->max_fds; n++) { + struct file *file; + file = rcu_dereference_check_fdtable(files, fdt->fd[n]); + if (!file) + continue; + res = f(p, file, n); + if (res) + break; } spin_unlock(&files->file_lock); return res; diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index 51ea267d444c..310972b72a66 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -228,6 +228,8 @@ static void requeue_io(struct inode *inode, struct bdi_writeback *wb) static void inode_sync_complete(struct inode *inode) { inode->i_state &= ~I_SYNC; + /* If inode is clean an unused, put it into LRU now... */ + inode_add_lru(inode); /* Waiters must see I_SYNC cleared before being woken up */ smp_mb(); wake_up_bit(&inode->i_state, __I_SYNC); @@ -1032,7 +1034,7 @@ int bdi_writeback_thread(void *data) while (!kthread_freezable_should_stop(NULL)) { /* * Remove own delayed wake-up timer, since we are already awake - * and we'll take care of the preriodic write-back. + * and we'll take care of the periodic write-back. */ del_timer(&wb->wakeup_timer); diff --git a/fs/fs_struct.c b/fs/fs_struct.c index 5df4775fea03..fe6ca583bbc0 100644 --- a/fs/fs_struct.c +++ b/fs/fs_struct.c @@ -164,27 +164,3 @@ struct fs_struct init_fs = { .seq = SEQCNT_ZERO, .umask = 0022, }; - -void daemonize_fs_struct(void) -{ - struct fs_struct *fs = current->fs; - - if (fs) { - int kill; - - task_lock(current); - - spin_lock(&init_fs.lock); - init_fs.users++; - spin_unlock(&init_fs.lock); - - spin_lock(&fs->lock); - current->fs = &init_fs; - kill = !--fs->users; - spin_unlock(&fs->lock); - - task_unlock(current); - if (kill) - free_fs_struct(fs); - } -} diff --git a/fs/gfs2/aops.c b/fs/gfs2/aops.c index 01c4975da4bc..30de4f2a2ea9 100644 --- a/fs/gfs2/aops.c +++ b/fs/gfs2/aops.c @@ -643,7 +643,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, goto out_unlock; requested = data_blocks + ind_blocks; - error = gfs2_inplace_reserve(ip, requested); + error = gfs2_inplace_reserve(ip, requested, 0); if (error) goto out_qunlock; } diff --git a/fs/gfs2/bmap.c b/fs/gfs2/bmap.c index 1fd3ae237bdd..a68e91bcef3d 100644 --- a/fs/gfs2/bmap.c +++ b/fs/gfs2/bmap.c @@ -991,6 +991,41 @@ unlock: return err; } +/** + * gfs2_journaled_truncate - Wrapper for truncate_pagecache for jdata files + * @inode: The inode being truncated + * @oldsize: The original (larger) size + * @newsize: The new smaller size + * + * With jdata files, we have to journal a revoke for each block which is + * truncated. As a result, we need to split this into separate transactions + * if the number of pages being truncated gets too large. + */ + +#define GFS2_JTRUNC_REVOKES 8192 + +static int gfs2_journaled_truncate(struct inode *inode, u64 oldsize, u64 newsize) +{ + struct gfs2_sbd *sdp = GFS2_SB(inode); + u64 max_chunk = GFS2_JTRUNC_REVOKES * sdp->sd_vfs->s_blocksize; + u64 chunk; + int error; + + while (oldsize != newsize) { + chunk = oldsize - newsize; + if (chunk > max_chunk) + chunk = max_chunk; + truncate_pagecache(inode, oldsize, oldsize - chunk); + oldsize -= chunk; + gfs2_trans_end(sdp); + error = gfs2_trans_begin(sdp, RES_DINODE, GFS2_JTRUNC_REVOKES); + if (error) + return error; + } + + return 0; +} + static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) { struct gfs2_inode *ip = GFS2_I(inode); @@ -1000,8 +1035,10 @@ static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) int journaled = gfs2_is_jdata(ip); int error; - error = gfs2_trans_begin(sdp, - RES_DINODE + (journaled ? RES_JDATA : 0), 0); + if (journaled) + error = gfs2_trans_begin(sdp, RES_DINODE + RES_JDATA, GFS2_JTRUNC_REVOKES); + else + error = gfs2_trans_begin(sdp, RES_DINODE, 0); if (error) return error; @@ -1026,7 +1063,16 @@ static int trunc_start(struct inode *inode, u64 oldsize, u64 newsize) ip->i_inode.i_mtime = ip->i_inode.i_ctime = CURRENT_TIME; gfs2_dinode_out(ip, dibh->b_data); - truncate_pagecache(inode, oldsize, newsize); + if (journaled) + error = gfs2_journaled_truncate(inode, oldsize, newsize); + else + truncate_pagecache(inode, oldsize, newsize); + + if (error) { + brelse(dibh); + return error; + } + out_brelse: brelse(dibh); out: @@ -1178,7 +1224,7 @@ static int do_grow(struct inode *inode, u64 size) if (error) return error; - error = gfs2_inplace_reserve(ip, 1); + error = gfs2_inplace_reserve(ip, 1, 0); if (error) goto do_grow_qunlock; unstuff = 1; diff --git a/fs/gfs2/dir.c b/fs/gfs2/dir.c index 259b088cfc4c..9a35670fdc38 100644 --- a/fs/gfs2/dir.c +++ b/fs/gfs2/dir.c @@ -1676,16 +1676,11 @@ int gfs2_dir_add(struct inode *inode, const struct qstr *name, be16_add_cpu(&leaf->lf_entries, 1); } brelse(bh); - error = gfs2_meta_inode_buffer(ip, &bh); - if (error) - break; - 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); + mark_inode_dirty(inode); error = 0; break; } diff --git a/fs/gfs2/file.c b/fs/gfs2/file.c index e056b4ce4877..dfe2d8cb9b2c 100644 --- a/fs/gfs2/file.c +++ b/fs/gfs2/file.c @@ -432,7 +432,7 @@ static int gfs2_page_mkwrite(struct vm_area_struct *vma, struct vm_fault *vmf) if (ret) goto out_unlock; gfs2_write_calc_reserv(ip, PAGE_CACHE_SIZE, &data_blocks, &ind_blocks); - ret = gfs2_inplace_reserve(ip, data_blocks + ind_blocks); + ret = gfs2_inplace_reserve(ip, data_blocks + ind_blocks, 0); if (ret) goto out_quota_unlock; @@ -825,7 +825,7 @@ static long gfs2_fallocate(struct file *file, int mode, loff_t offset, retry: gfs2_write_calc_reserv(ip, bytes, &data_blocks, &ind_blocks); - error = gfs2_inplace_reserve(ip, data_blocks + ind_blocks); + error = gfs2_inplace_reserve(ip, data_blocks + ind_blocks, 0); if (error) { if (error == -ENOSPC && bytes > sdp->sd_sb.sb_bsize) { bytes >>= 1; diff --git a/fs/gfs2/glock.c b/fs/gfs2/glock.c index e6c2fd53cab2..992c5c0cb504 100644 --- a/fs/gfs2/glock.c +++ b/fs/gfs2/glock.c @@ -55,8 +55,6 @@ struct gfs2_glock_iter { typedef void (*glock_examiner) (struct gfs2_glock * gl); -static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl); -#define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { __dump_glock(NULL, gl); BUG(); } } while(0) static void do_xmote(struct gfs2_glock *gl, struct gfs2_holder *gh, unsigned int target); static struct dentry *gfs2_root; @@ -107,10 +105,12 @@ static void gfs2_glock_dealloc(struct rcu_head *rcu) { struct gfs2_glock *gl = container_of(rcu, struct gfs2_glock, gl_rcu); - if (gl->gl_ops->go_flags & GLOF_ASPACE) + if (gl->gl_ops->go_flags & GLOF_ASPACE) { kmem_cache_free(gfs2_glock_aspace_cachep, gl); - else + } else { + kfree(gl->gl_lksb.sb_lvbptr); kmem_cache_free(gfs2_glock_cachep, gl); + } } void gfs2_glock_free(struct gfs2_glock *gl) @@ -537,8 +537,8 @@ __acquires(&gl->gl_spin) (lck_flags & (LM_FLAG_TRY|LM_FLAG_TRY_1CB))) clear_bit(GLF_BLOCKING, &gl->gl_flags); spin_unlock(&gl->gl_spin); - if (glops->go_xmote_th) - glops->go_xmote_th(gl); + if (glops->go_sync) + glops->go_sync(gl); if (test_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags)) glops->go_inval(gl, target == LM_ST_DEFERRED ? 0 : DIO_METADATA); clear_bit(GLF_INVALIDATE_IN_PROGRESS, &gl->gl_flags); @@ -547,7 +547,10 @@ __acquires(&gl->gl_spin) if (sdp->sd_lockstruct.ls_ops->lm_lock) { /* lock_dlm */ ret = sdp->sd_lockstruct.ls_ops->lm_lock(gl, target, lck_flags); - GLOCK_BUG_ON(gl, ret); + if (ret) { + printk(KERN_ERR "GFS2: lm_lock ret %d\n", ret); + GLOCK_BUG_ON(gl, 1); + } } else { /* lock_nolock */ finish_xmote(gl, target); if (queue_delayed_work(glock_workqueue, &gl->gl_work, 0) == 0) @@ -736,6 +739,16 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, if (!gl) return -ENOMEM; + memset(&gl->gl_lksb, 0, sizeof(struct dlm_lksb)); + + if (glops->go_flags & GLOF_LVB) { + gl->gl_lksb.sb_lvbptr = kzalloc(GFS2_MIN_LVB_SIZE, GFP_KERNEL); + if (!gl->gl_lksb.sb_lvbptr) { + kmem_cache_free(cachep, gl); + return -ENOMEM; + } + } + atomic_inc(&sdp->sd_glock_disposal); gl->gl_sbd = sdp; gl->gl_flags = 0; @@ -753,9 +766,6 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, preempt_enable(); gl->gl_stats.stats[GFS2_LKS_DCOUNT] = 0; gl->gl_stats.stats[GFS2_LKS_QCOUNT] = 0; - memset(&gl->gl_lksb, 0, sizeof(struct dlm_lksb)); - memset(gl->gl_lvb, 0, 32 * sizeof(char)); - gl->gl_lksb.sb_lvbptr = gl->gl_lvb; gl->gl_tchange = jiffies; gl->gl_object = NULL; gl->gl_hold_time = GL_GLOCK_DFT_HOLD; @@ -768,7 +778,7 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, mapping->host = s->s_bdev->bd_inode; mapping->flags = 0; mapping_set_gfp_mask(mapping, GFP_NOFS); - mapping->assoc_mapping = NULL; + mapping->private_data = NULL; mapping->backing_dev_info = s->s_bdi; mapping->writeback_index = 0; } @@ -777,6 +787,7 @@ int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, tmp = search_bucket(hash, sdp, &name); if (tmp) { spin_unlock_bucket(hash); + kfree(gl->gl_lksb.sb_lvbptr); kmem_cache_free(cachep, gl); atomic_dec(&sdp->sd_glock_disposal); gl = tmp; @@ -1013,7 +1024,7 @@ trap_recursive: printk(KERN_ERR "pid: %d\n", pid_nr(gh->gh_owner_pid)); printk(KERN_ERR "lock type: %d req lock state : %d\n", gh->gh_gl->gl_name.ln_type, gh->gh_state); - __dump_glock(NULL, gl); + gfs2_dump_glock(NULL, gl); BUG(); } @@ -1508,7 +1519,7 @@ static int dump_glock(struct seq_file *seq, struct gfs2_glock *gl) { int ret; spin_lock(&gl->gl_spin); - ret = __dump_glock(seq, gl); + ret = gfs2_dump_glock(seq, gl); spin_unlock(&gl->gl_spin); return ret; } @@ -1528,6 +1539,7 @@ static void dump_glock_func(struct gfs2_glock *gl) void gfs2_gl_hash_clear(struct gfs2_sbd *sdp) { + set_bit(SDF_SKIP_DLM_UNLOCK, &sdp->sd_flags); glock_hash_walk(clear_glock, sdp); flush_workqueue(glock_workqueue); wait_event(sdp->sd_glock_wait, atomic_read(&sdp->sd_glock_disposal) == 0); @@ -1655,7 +1667,7 @@ static const char *gflags2str(char *buf, const struct gfs2_glock *gl) } /** - * __dump_glock - print information about a glock + * gfs2_dump_glock - print information about a glock * @seq: The seq_file struct * @gl: the glock * @@ -1672,7 +1684,7 @@ static const char *gflags2str(char *buf, const struct gfs2_glock *gl) * Returns: 0 on success, -ENOBUFS when we run out of space */ -static int __dump_glock(struct seq_file *seq, const struct gfs2_glock *gl) +int gfs2_dump_glock(struct seq_file *seq, const struct gfs2_glock *gl) { const struct gfs2_glock_operations *glops = gl->gl_ops; unsigned long long dtime; diff --git a/fs/gfs2/glock.h b/fs/gfs2/glock.h index 307ac31df781..fd580b7861d5 100644 --- a/fs/gfs2/glock.h +++ b/fs/gfs2/glock.h @@ -178,33 +178,33 @@ static inline struct address_space *gfs2_glock2aspace(struct gfs2_glock *gl) return NULL; } -int gfs2_glock_get(struct gfs2_sbd *sdp, - u64 number, const struct gfs2_glock_operations *glops, - int create, struct gfs2_glock **glp); -void gfs2_glock_hold(struct gfs2_glock *gl); -void gfs2_glock_put_nolock(struct gfs2_glock *gl); -void gfs2_glock_put(struct gfs2_glock *gl); -void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, unsigned flags, - struct gfs2_holder *gh); -void gfs2_holder_reinit(unsigned int state, unsigned flags, - struct gfs2_holder *gh); -void gfs2_holder_uninit(struct gfs2_holder *gh); -int gfs2_glock_nq(struct gfs2_holder *gh); -int gfs2_glock_poll(struct gfs2_holder *gh); -int gfs2_glock_wait(struct gfs2_holder *gh); -void gfs2_glock_dq(struct gfs2_holder *gh); -void gfs2_glock_dq_wait(struct gfs2_holder *gh); - -void gfs2_glock_dq_uninit(struct gfs2_holder *gh); -int gfs2_glock_nq_num(struct gfs2_sbd *sdp, - u64 number, const struct gfs2_glock_operations *glops, - unsigned int state, int flags, struct gfs2_holder *gh); - -int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs); -void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs); -void gfs2_glock_dq_uninit_m(unsigned int num_gh, struct gfs2_holder *ghs); - -__printf(2, 3) +extern int gfs2_glock_get(struct gfs2_sbd *sdp, u64 number, + const struct gfs2_glock_operations *glops, + int create, struct gfs2_glock **glp); +extern void gfs2_glock_hold(struct gfs2_glock *gl); +extern void gfs2_glock_put_nolock(struct gfs2_glock *gl); +extern void gfs2_glock_put(struct gfs2_glock *gl); +extern void gfs2_holder_init(struct gfs2_glock *gl, unsigned int state, + unsigned flags, struct gfs2_holder *gh); +extern void gfs2_holder_reinit(unsigned int state, unsigned flags, + struct gfs2_holder *gh); +extern void gfs2_holder_uninit(struct gfs2_holder *gh); +extern int gfs2_glock_nq(struct gfs2_holder *gh); +extern int gfs2_glock_poll(struct gfs2_holder *gh); +extern int gfs2_glock_wait(struct gfs2_holder *gh); +extern void gfs2_glock_dq(struct gfs2_holder *gh); +extern void gfs2_glock_dq_wait(struct gfs2_holder *gh); +extern void gfs2_glock_dq_uninit(struct gfs2_holder *gh); +extern int gfs2_glock_nq_num(struct gfs2_sbd *sdp, u64 number, + const struct gfs2_glock_operations *glops, + unsigned int state, int flags, + struct gfs2_holder *gh); +extern int gfs2_glock_nq_m(unsigned int num_gh, struct gfs2_holder *ghs); +extern void gfs2_glock_dq_m(unsigned int num_gh, struct gfs2_holder *ghs); +extern void gfs2_glock_dq_uninit_m(unsigned int num_gh, struct gfs2_holder *ghs); +extern int gfs2_dump_glock(struct seq_file *seq, const struct gfs2_glock *gl); +#define GLOCK_BUG_ON(gl,x) do { if (unlikely(x)) { gfs2_dump_glock(NULL, gl); BUG(); } } while(0) +extern __printf(2, 3) void gfs2_print_dbg(struct seq_file *seq, const char *fmt, ...); /** diff --git a/fs/gfs2/glops.c b/fs/gfs2/glops.c index 32cc4fde975c..78d4184ffc7d 100644 --- a/fs/gfs2/glops.c +++ b/fs/gfs2/glops.c @@ -74,7 +74,7 @@ static void __gfs2_ail_flush(struct gfs2_glock *gl, bool fsync) gfs2_trans_add_revoke(sdp, bd); } - BUG_ON(!fsync && atomic_read(&gl->gl_ail_count)); + GLOCK_BUG_ON(gl, !fsync && atomic_read(&gl->gl_ail_count)); spin_unlock(&sdp->sd_ail_lock); gfs2_log_unlock(sdp); } @@ -96,7 +96,7 @@ static void gfs2_ail_empty_gl(struct gfs2_glock *gl) tr.tr_ip = (unsigned long)__builtin_return_address(0); sb_start_intwrite(sdp->sd_vfs); gfs2_log_reserve(sdp, tr.tr_reserved); - BUG_ON(current->journal_info); + WARN_ON_ONCE(current->journal_info); current->journal_info = &tr; __gfs2_ail_flush(gl, 0); @@ -139,7 +139,7 @@ static void rgrp_go_sync(struct gfs2_glock *gl) if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags)) return; - BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE); + GLOCK_BUG_ON(gl, gl->gl_state != LM_ST_EXCLUSIVE); gfs2_log_flush(gl->gl_sbd, gl); filemap_fdatawrite(metamapping); @@ -168,7 +168,7 @@ static void rgrp_go_inval(struct gfs2_glock *gl, int flags) { struct address_space *mapping = gfs2_glock2aspace(gl); - BUG_ON(!(flags & DIO_METADATA)); + WARN_ON_ONCE(!(flags & DIO_METADATA)); gfs2_assert_withdraw(gl->gl_sbd, !atomic_read(&gl->gl_ail_count)); truncate_inode_pages(mapping, 0); @@ -197,7 +197,7 @@ static void inode_go_sync(struct gfs2_glock *gl) if (!test_and_clear_bit(GLF_DIRTY, &gl->gl_flags)) return; - BUG_ON(gl->gl_state != LM_ST_EXCLUSIVE); + GLOCK_BUG_ON(gl, gl->gl_state != LM_ST_EXCLUSIVE); gfs2_log_flush(gl->gl_sbd, gl); filemap_fdatawrite(metamapping); @@ -536,7 +536,7 @@ const struct gfs2_glock_operations gfs2_meta_glops = { }; const struct gfs2_glock_operations gfs2_inode_glops = { - .go_xmote_th = inode_go_sync, + .go_sync = inode_go_sync, .go_inval = inode_go_inval, .go_demote_ok = inode_go_demote_ok, .go_lock = inode_go_lock, @@ -546,17 +546,17 @@ const struct gfs2_glock_operations gfs2_inode_glops = { }; const struct gfs2_glock_operations gfs2_rgrp_glops = { - .go_xmote_th = rgrp_go_sync, + .go_sync = rgrp_go_sync, .go_inval = rgrp_go_inval, .go_lock = gfs2_rgrp_go_lock, .go_unlock = gfs2_rgrp_go_unlock, .go_dump = gfs2_rgrp_dump, .go_type = LM_TYPE_RGRP, - .go_flags = GLOF_ASPACE, + .go_flags = GLOF_ASPACE | GLOF_LVB, }; const struct gfs2_glock_operations gfs2_trans_glops = { - .go_xmote_th = trans_go_sync, + .go_sync = trans_go_sync, .go_xmote_bh = trans_go_xmote_bh, .go_demote_ok = trans_go_demote_ok, .go_type = LM_TYPE_NONDISK, @@ -577,6 +577,7 @@ const struct gfs2_glock_operations gfs2_nondisk_glops = { const struct gfs2_glock_operations gfs2_quota_glops = { .go_type = LM_TYPE_QUOTA, + .go_flags = GLOF_LVB, }; const struct gfs2_glock_operations gfs2_journal_glops = { diff --git a/fs/gfs2/incore.h b/fs/gfs2/incore.h index 3d469d37345e..c373a24fedd9 100644 --- a/fs/gfs2/incore.h +++ b/fs/gfs2/incore.h @@ -205,7 +205,7 @@ struct lm_lockname { struct gfs2_glock_operations { - void (*go_xmote_th) (struct gfs2_glock *gl); + void (*go_sync) (struct gfs2_glock *gl); int (*go_xmote_bh) (struct gfs2_glock *gl, struct gfs2_holder *gh); void (*go_inval) (struct gfs2_glock *gl, int flags); int (*go_demote_ok) (const struct gfs2_glock *gl); @@ -216,6 +216,7 @@ struct gfs2_glock_operations { const int go_type; const unsigned long go_flags; #define GLOF_ASPACE 1 +#define GLOF_LVB 2 }; enum { @@ -321,7 +322,6 @@ struct gfs2_glock { ktime_t gl_dstamp; struct gfs2_lkstats gl_stats; struct dlm_lksb gl_lksb; - char gl_lvb[32]; unsigned long gl_tchange; void *gl_object; @@ -539,6 +539,7 @@ enum { SDF_DEMOTE = 5, SDF_NOJOURNALID = 6, SDF_RORECOVERY = 7, /* read only recovery */ + SDF_SKIP_DLM_UNLOCK = 8, }; #define GFS2_FSNAME_LEN 256 @@ -621,6 +622,7 @@ struct gfs2_sbd { u32 sd_hash_bsize_shift; u32 sd_hash_ptrs; /* Number of pointers in a hash block */ u32 sd_qc_per_block; + u32 sd_blocks_per_bitmap; u32 sd_max_dirres; /* Max blocks needed to add a directory entry */ u32 sd_max_height; /* Max height of a file's metadata tree */ u64 sd_heightsize[GFS2_MAX_META_HEIGHT + 1]; diff --git a/fs/gfs2/inode.c b/fs/gfs2/inode.c index 381893ceefa4..2b6f5698ef18 100644 --- a/fs/gfs2/inode.c +++ b/fs/gfs2/inode.c @@ -364,34 +364,34 @@ static int create_ok(struct gfs2_inode *dip, const struct qstr *name, return 0; } -static void munge_mode_uid_gid(struct gfs2_inode *dip, umode_t *mode, - unsigned int *uid, unsigned int *gid) +static void munge_mode_uid_gid(const struct gfs2_inode *dip, + struct inode *inode) { if (GFS2_SB(&dip->i_inode)->sd_args.ar_suiddir && (dip->i_inode.i_mode & S_ISUID) && dip->i_inode.i_uid) { - if (S_ISDIR(*mode)) - *mode |= S_ISUID; + if (S_ISDIR(inode->i_mode)) + inode->i_mode |= S_ISUID; else if (dip->i_inode.i_uid != current_fsuid()) - *mode &= ~07111; - *uid = dip->i_inode.i_uid; + inode->i_mode &= ~07111; + inode->i_uid = dip->i_inode.i_uid; } else - *uid = current_fsuid(); + inode->i_uid = current_fsuid(); if (dip->i_inode.i_mode & S_ISGID) { - if (S_ISDIR(*mode)) - *mode |= S_ISGID; - *gid = dip->i_inode.i_gid; + if (S_ISDIR(inode->i_mode)) + inode->i_mode |= S_ISGID; + inode->i_gid = dip->i_inode.i_gid; } else - *gid = current_fsgid(); + inode->i_gid = current_fsgid(); } -static int alloc_dinode(struct gfs2_inode *dip, u64 *no_addr, u64 *generation) +static int alloc_dinode(struct gfs2_inode *ip, u32 flags) { - struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); + struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); int error; int dblocks = 1; - error = gfs2_inplace_reserve(dip, RES_DINODE); + error = gfs2_inplace_reserve(ip, RES_DINODE, flags); if (error) goto out; @@ -399,12 +399,15 @@ static int alloc_dinode(struct gfs2_inode *dip, u64 *no_addr, u64 *generation) if (error) goto out_ipreserv; - error = gfs2_alloc_blocks(dip, no_addr, &dblocks, 1, generation); + error = gfs2_alloc_blocks(ip, &ip->i_no_addr, &dblocks, 1, &ip->i_generation); + ip->i_no_formal_ino = ip->i_generation; + ip->i_inode.i_ino = ip->i_no_addr; + ip->i_goal = ip->i_no_addr; gfs2_trans_end(sdp); out_ipreserv: - gfs2_inplace_release(dip); + gfs2_inplace_release(ip); out: return error; } @@ -429,52 +432,42 @@ static void gfs2_init_dir(struct buffer_head *dibh, /** * init_dinode - Fill in a new dinode structure * @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: 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) + * @ip: The inode * @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, umode_t mode, - unsigned int uid, unsigned int gid, - const u64 *generation, dev_t dev, const char *symname, - unsigned size, struct buffer_head **bhp) +static void init_dinode(struct gfs2_inode *dip, struct gfs2_inode *ip, + const char *symname, struct buffer_head **bhp) { struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); struct gfs2_dinode *di; struct buffer_head *dibh; struct timespec tv = CURRENT_TIME; - dibh = gfs2_meta_new(gl, inum->no_addr); - gfs2_trans_add_bh(gl, dibh, 1); + dibh = gfs2_meta_new(ip->i_gl, ip->i_no_addr); + gfs2_trans_add_bh(ip->i_gl, dibh, 1); gfs2_metatype_set(dibh, GFS2_METATYPE_DI, GFS2_FORMAT_DI); gfs2_buffer_clear_tail(dibh, sizeof(struct gfs2_dinode)); di = (struct gfs2_dinode *)dibh->b_data; - di->di_num.no_formal_ino = cpu_to_be64(inum->no_formal_ino); - di->di_num.no_addr = cpu_to_be64(inum->no_addr); - di->di_mode = cpu_to_be32(mode); - di->di_uid = cpu_to_be32(uid); - di->di_gid = cpu_to_be32(gid); + di->di_num.no_formal_ino = cpu_to_be64(ip->i_no_formal_ino); + di->di_num.no_addr = cpu_to_be64(ip->i_no_addr); + di->di_mode = cpu_to_be32(ip->i_inode.i_mode); + di->di_uid = cpu_to_be32(ip->i_inode.i_uid); + di->di_gid = cpu_to_be32(ip->i_inode.i_gid); di->di_nlink = 0; - di->di_size = cpu_to_be64(size); + di->di_size = cpu_to_be64(ip->i_inode.i_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)); - di->di_minor = cpu_to_be32(MINOR(dev)); - di->di_goal_meta = di->di_goal_data = cpu_to_be64(inum->no_addr); - di->di_generation = cpu_to_be64(*generation); + di->di_major = cpu_to_be32(MAJOR(ip->i_inode.i_rdev)); + di->di_minor = cpu_to_be32(MINOR(ip->i_inode.i_rdev)); + di->di_goal_meta = di->di_goal_data = cpu_to_be64(ip->i_no_addr); + di->di_generation = cpu_to_be64(ip->i_generation); di->di_flags = 0; di->__pad1 = 0; - di->di_payload_format = cpu_to_be32(S_ISDIR(mode) ? GFS2_FORMAT_DE : 0); + di->di_payload_format = cpu_to_be32(S_ISDIR(ip->i_inode.i_mode) ? GFS2_FORMAT_DE : 0); di->di_height = 0; di->__pad2 = 0; di->__pad3 = 0; @@ -487,7 +480,7 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, di->di_ctime_nsec = cpu_to_be32(tv.tv_nsec); memset(&di->di_reserved, 0, sizeof(di->di_reserved)); - switch(mode & S_IFMT) { + switch(ip->i_inode.i_mode & S_IFMT) { case S_IFREG: if ((dip->i_diskflags & GFS2_DIF_INHERIT_JDATA) || gfs2_tune_get(sdp, gt_new_files_jdata)) @@ -502,7 +495,7 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, gfs2_init_dir(dibh, dip); break; case S_IFLNK: - memcpy(dibh->b_data + sizeof(struct gfs2_dinode), symname, size); + memcpy(dibh->b_data + sizeof(struct gfs2_dinode), symname, ip->i_inode.i_size); break; } @@ -511,25 +504,22 @@ static void init_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, *bhp = dibh; } -static int make_dinode(struct gfs2_inode *dip, struct gfs2_glock *gl, - umode_t mode, const struct gfs2_inum_host *inum, - const u64 *generation, dev_t dev, const char *symname, - unsigned int size, struct buffer_head **bhp) +static int make_dinode(struct gfs2_inode *dip, struct gfs2_inode *ip, + const char *symname, struct buffer_head **bhp) { + struct inode *inode = &ip->i_inode; struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - unsigned int uid, gid; int error; - munge_mode_uid_gid(dip, &mode, &uid, &gid); error = gfs2_rindex_update(sdp); if (error) return error; - error = gfs2_quota_lock(dip, uid, gid); + error = gfs2_quota_lock(dip, inode->i_uid, inode->i_gid); if (error) return error; - error = gfs2_quota_check(dip, uid, gid); + error = gfs2_quota_check(dip, inode->i_uid, inode->i_gid); if (error) goto out_quota; @@ -537,8 +527,8 @@ 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, symname, size, bhp); - gfs2_quota_change(dip, +1, uid, gid); + init_dinode(dip, ip, symname, bhp); + gfs2_quota_change(dip, +1, inode->i_uid, inode->i_gid); gfs2_trans_end(sdp); out_quota: @@ -570,7 +560,7 @@ static int link_dinode(struct gfs2_inode *dip, const struct qstr *name, if (error) goto fail_quota_locks; - error = gfs2_inplace_reserve(dip, sdp->sd_max_dirres); + error = gfs2_inplace_reserve(dip, sdp->sd_max_dirres, 0); if (error) goto fail_quota_locks; @@ -657,19 +647,14 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, struct inode *inode = NULL; struct gfs2_inode *dip = GFS2_I(dir), *ip; struct gfs2_sbd *sdp = GFS2_SB(&dip->i_inode); - struct gfs2_inum_host inum = { .no_addr = 0, .no_formal_ino = 0 }; + struct gfs2_glock *io_gl; int error; - u64 generation; struct buffer_head *bh = NULL; + u32 aflags = 0; if (!name->len || name->len > GFS2_FNAMESIZE) return -ENAMETOOLONG; - /* We need a reservation to allocate the new dinode block. The - directory ip temporarily points to the reservation, but this is - being done to get a set of contiguous blocks for the new dinode. - Since this is a create, we don't have a sizehint yet, so it will - have to use the minimum reservation size. */ error = gfs2_rs_alloc(dip); if (error) return error; @@ -688,45 +673,72 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, if (error) goto fail_gunlock; - error = alloc_dinode(dip, &inum.no_addr, &generation); + inode = new_inode(sdp->sd_vfs); + if (!inode) { + gfs2_glock_dq_uninit(ghs); + return -ENOMEM; + } + ip = GFS2_I(inode); + error = gfs2_rs_alloc(ip); if (error) - goto fail_gunlock; - inum.no_formal_ino = generation; + goto fail_free_inode; + + set_bit(GIF_INVALID, &ip->i_flags); + inode->i_mode = mode; + inode->i_rdev = dev; + inode->i_size = size; + munge_mode_uid_gid(dip, inode); + ip->i_goal = dip->i_goal; - error = gfs2_glock_nq_num(sdp, inum.no_addr, &gfs2_inode_glops, - LM_ST_EXCLUSIVE, GL_SKIP, ghs + 1); + if ((GFS2_I(sdp->sd_root_dir->d_inode) == dip) || + (dip->i_diskflags & GFS2_DIF_TOPDIR)) + aflags |= GFS2_AF_ORLOV; + + error = alloc_dinode(ip, aflags); if (error) - goto fail_gunlock; + goto fail_free_inode; - error = make_dinode(dip, ghs[1].gh_gl, mode, &inum, &generation, dev, symname, size, &bh); + error = gfs2_glock_get(sdp, ip->i_no_addr, &gfs2_inode_glops, CREATE, &ip->i_gl); if (error) - goto fail_gunlock2; + goto fail_free_inode; - inode = gfs2_inode_lookup(dir->i_sb, IF2DT(mode), inum.no_addr, - inum.no_formal_ino, 0); - if (IS_ERR(inode)) + ip->i_gl->gl_object = ip; + error = gfs2_glock_nq_init(ip->i_gl, LM_ST_EXCLUSIVE, GL_SKIP, ghs + 1); + if (error) + goto fail_free_inode; + + error = make_dinode(dip, ip, symname, &bh); + if (error) goto fail_gunlock2; - ip = GFS2_I(inode); - error = gfs2_inode_refresh(ip); + error = gfs2_glock_get(sdp, ip->i_no_addr, &gfs2_iopen_glops, CREATE, &io_gl); if (error) goto fail_gunlock2; - error = gfs2_rs_alloc(ip); + error = gfs2_glock_nq_init(io_gl, LM_ST_SHARED, GL_EXACT, &ip->i_iopen_gh); if (error) goto fail_gunlock2; + ip->i_iopen_gh.gh_gl->gl_object = ip; + gfs2_glock_put(io_gl); + gfs2_set_iop(inode); + insert_inode_hash(inode); + + error = gfs2_inode_refresh(ip); + if (error) + goto fail_gunlock3; + error = gfs2_acl_create(dip, inode); if (error) - goto fail_gunlock2; + goto fail_gunlock3; error = gfs2_security_init(dip, ip, name); if (error) - goto fail_gunlock2; + goto fail_gunlock3; error = link_dinode(dip, name, ip); if (error) - goto fail_gunlock2; + goto fail_gunlock3; if (bh) brelse(bh); @@ -739,8 +751,20 @@ static int gfs2_create_inode(struct inode *dir, struct dentry *dentry, d_instantiate(dentry, inode); return 0; +fail_gunlock3: + gfs2_glock_dq_uninit(ghs + 1); + if (ip->i_gl) + gfs2_glock_put(ip->i_gl); + goto fail_gunlock; + fail_gunlock2: gfs2_glock_dq_uninit(ghs + 1); +fail_free_inode: + if (ip->i_gl) + gfs2_glock_put(ip->i_gl); + gfs2_rs_delete(ip); + free_inode_nonrcu(inode); + inode = NULL; fail_gunlock: gfs2_glock_dq_uninit(ghs); if (inode && !IS_ERR(inode)) { @@ -748,7 +772,6 @@ fail_gunlock: iput(inode); } fail: - gfs2_rs_delete(dip); if (bh) brelse(bh); return error; @@ -884,7 +907,7 @@ static int gfs2_link(struct dentry *old_dentry, struct inode *dir, if (error) goto out_gunlock; - error = gfs2_inplace_reserve(dip, sdp->sd_max_dirres); + error = gfs2_inplace_reserve(dip, sdp->sd_max_dirres, 0); if (error) goto out_gunlock_q; @@ -977,7 +1000,6 @@ static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, * 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 @@ -987,8 +1009,7 @@ static int gfs2_unlink_ok(struct gfs2_inode *dip, const struct qstr *name, */ static int gfs2_unlink_inode(struct gfs2_inode *dip, - const struct dentry *dentry, - struct buffer_head *bh) + const struct dentry *dentry) { struct inode *inode = dentry->d_inode; struct gfs2_inode *ip = GFS2_I(inode); @@ -1028,7 +1049,6 @@ static int gfs2_unlink(struct inode *dir, struct dentry *dentry) 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; int error; @@ -1077,14 +1097,9 @@ static int gfs2_unlink(struct inode *dir, struct dentry *dentry) 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); + error = gfs2_unlink_inode(dip, dentry); out_end_trans: gfs2_trans_end(sdp); @@ -1365,7 +1380,7 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry, if (error) goto out_gunlock; - error = gfs2_inplace_reserve(ndip, sdp->sd_max_dirres); + error = gfs2_inplace_reserve(ndip, sdp->sd_max_dirres, 0); if (error) goto out_gunlock_q; @@ -1384,14 +1399,8 @@ static int gfs2_rename(struct inode *odir, struct dentry *odentry, /* 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 (nip) + error = gfs2_unlink_inode(ndip, ndentry); if (dir_rename) { error = gfs2_dir_mvino(ip, &gfs2_qdotdot, ndip, DT_DIR); diff --git a/fs/gfs2/lock_dlm.c b/fs/gfs2/lock_dlm.c index 0fb6539b0c8c..8dad6b093716 100644 --- a/fs/gfs2/lock_dlm.c +++ b/fs/gfs2/lock_dlm.c @@ -120,8 +120,8 @@ static void gdlm_ast(void *arg) gfs2_update_reply_times(gl); BUG_ON(gl->gl_lksb.sb_flags & DLM_SBF_DEMOTED); - if (gl->gl_lksb.sb_flags & DLM_SBF_VALNOTVALID) - memset(gl->gl_lvb, 0, GDLM_LVB_SIZE); + if ((gl->gl_lksb.sb_flags & DLM_SBF_VALNOTVALID) && gl->gl_lksb.sb_lvbptr) + memset(gl->gl_lksb.sb_lvbptr, 0, GDLM_LVB_SIZE); switch (gl->gl_lksb.sb_status) { case -DLM_EUNLOCK: /* Unlocked, so glock can be freed */ @@ -203,8 +203,10 @@ static int make_mode(const unsigned int lmstate) static u32 make_flags(struct gfs2_glock *gl, const unsigned int gfs_flags, const int req) { - u32 lkf = DLM_LKF_VALBLK; - u32 lkid = gl->gl_lksb.sb_lkid; + u32 lkf = 0; + + if (gl->gl_lksb.sb_lvbptr) + lkf |= DLM_LKF_VALBLK; if (gfs_flags & LM_FLAG_TRY) lkf |= DLM_LKF_NOQUEUE; @@ -228,7 +230,7 @@ static u32 make_flags(struct gfs2_glock *gl, const unsigned int gfs_flags, BUG(); } - if (lkid != 0) { + if (gl->gl_lksb.sb_lkid != 0) { lkf |= DLM_LKF_CONVERT; if (test_bit(GLF_BLOCKING, &gl->gl_flags)) lkf |= DLM_LKF_QUECVT; @@ -289,6 +291,14 @@ static void gdlm_put_lock(struct gfs2_glock *gl) gfs2_glstats_inc(gl, GFS2_LKS_DCOUNT); gfs2_sbstats_inc(gl, GFS2_LKS_DCOUNT); gfs2_update_request_times(gl); + + /* don't want to skip dlm_unlock writing the lvb when lock is ex */ + if (test_bit(SDF_SKIP_DLM_UNLOCK, &sdp->sd_flags) && + gl->gl_lksb.sb_lvbptr && (gl->gl_state != LM_ST_EXCLUSIVE)) { + gfs2_glock_free(gl); + return; + } + error = dlm_unlock(ls->ls_dlm, gl->gl_lksb.sb_lkid, DLM_LKF_VALBLK, NULL, gl); if (error) { diff --git a/fs/gfs2/ops_fstype.c b/fs/gfs2/ops_fstype.c index e443966c8106..0e3554edb8f2 100644 --- a/fs/gfs2/ops_fstype.c +++ b/fs/gfs2/ops_fstype.c @@ -278,6 +278,9 @@ static int gfs2_read_sb(struct gfs2_sbd *sdp, int silent) sdp->sd_qc_per_block = (sdp->sd_sb.sb_bsize - sizeof(struct gfs2_meta_header)) / sizeof(struct gfs2_quota_change); + sdp->sd_blocks_per_bitmap = (sdp->sd_sb.sb_bsize - + sizeof(struct gfs2_meta_header)) + * GFS2_NBBY; /* not the rgrp bitmap, subsequent bitmaps only */ /* Compute maximum reservation required to add a entry to a directory */ diff --git a/fs/gfs2/quota.c b/fs/gfs2/quota.c index c5af8e18f27a..ae55e248c3b7 100644 --- a/fs/gfs2/quota.c +++ b/fs/gfs2/quota.c @@ -816,7 +816,7 @@ static int do_sync(unsigned int num_qd, struct gfs2_quota_data **qda) blocks = num_qd * data_blocks + RES_DINODE + num_qd + 3; reserved = 1 + (nalloc * (data_blocks + ind_blocks)); - error = gfs2_inplace_reserve(ip, reserved); + error = gfs2_inplace_reserve(ip, reserved, 0); if (error) goto out_alloc; @@ -869,7 +869,7 @@ static int update_qd(struct gfs2_sbd *sdp, struct gfs2_quota_data *qd) if (error < 0) return error; - qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb; + qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; qlvb->qb_magic = cpu_to_be32(GFS2_MAGIC); qlvb->__pad = 0; qlvb->qb_limit = q.qu_limit; @@ -893,7 +893,7 @@ restart: if (error) return error; - qd->qd_qb = *(struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb; + qd->qd_qb = *(struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; if (force_refresh || qd->qd_qb.qb_magic != cpu_to_be32(GFS2_MAGIC)) { gfs2_glock_dq_uninit(q_gh); @@ -1506,7 +1506,7 @@ static int gfs2_get_dqblk(struct super_block *sb, struct kqid qid, if (error) goto out; - qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lvb; + qlvb = (struct gfs2_quota_lvb *)qd->qd_gl->gl_lksb.sb_lvbptr; fdq->d_version = FS_DQUOT_VERSION; fdq->d_flags = (type == QUOTA_USER) ? FS_USER_QUOTA : FS_GROUP_QUOTA; fdq->d_id = from_kqid(&init_user_ns, qid); @@ -1605,7 +1605,7 @@ static int gfs2_set_dqblk(struct super_block *sb, struct kqid qid, gfs2_write_calc_reserv(ip, sizeof(struct gfs2_quota), &data_blocks, &ind_blocks); blocks = 1 + data_blocks + ind_blocks; - error = gfs2_inplace_reserve(ip, blocks); + error = gfs2_inplace_reserve(ip, blocks, 0); if (error) goto out_i; blocks += gfs2_rg_blocks(ip, blocks); diff --git a/fs/gfs2/rgrp.c b/fs/gfs2/rgrp.c index 38fe18f2f055..37ee061d899e 100644 --- a/fs/gfs2/rgrp.c +++ b/fs/gfs2/rgrp.c @@ -16,6 +16,7 @@ #include <linux/prefetch.h> #include <linux/blkdev.h> #include <linux/rbtree.h> +#include <linux/random.h> #include "gfs2.h" #include "incore.h" @@ -251,22 +252,25 @@ static u32 gfs2_bitfit(const u8 *buf, const unsigned int len, static int gfs2_rbm_from_block(struct gfs2_rbm *rbm, u64 block) { u64 rblock = block - rbm->rgd->rd_data0; - u32 goal = (u32)rblock; - int x; + u32 x; if (WARN_ON_ONCE(rblock > UINT_MAX)) return -EINVAL; if (block >= rbm->rgd->rd_data0 + rbm->rgd->rd_data) return -E2BIG; - for (x = 0; x < rbm->rgd->rd_length; x++) { - rbm->bi = rbm->rgd->rd_bits + x; - if (goal < (rbm->bi->bi_start + rbm->bi->bi_len) * GFS2_NBBY) { - rbm->offset = goal - (rbm->bi->bi_start * GFS2_NBBY); - break; - } - } + rbm->bi = rbm->rgd->rd_bits; + rbm->offset = (u32)(rblock); + /* Check if the block is within the first block */ + if (rbm->offset < (rbm->bi->bi_start + rbm->bi->bi_len) * GFS2_NBBY) + return 0; + /* Adjust for the size diff between gfs2_meta_header and gfs2_rgrp */ + rbm->offset += (sizeof(struct gfs2_rgrp) - + sizeof(struct gfs2_meta_header)) * GFS2_NBBY; + x = rbm->offset / rbm->rgd->rd_sbd->sd_blocks_per_bitmap; + rbm->offset -= x * rbm->rgd->rd_sbd->sd_blocks_per_bitmap; + rbm->bi += x; return 0; } @@ -875,7 +879,7 @@ static int read_rindex_entry(struct gfs2_inode *ip) goto fail; rgd->rd_gl->gl_object = rgd; - rgd->rd_rgl = (struct gfs2_rgrp_lvb *)rgd->rd_gl->gl_lvb; + rgd->rd_rgl = (struct gfs2_rgrp_lvb *)rgd->rd_gl->gl_lksb.sb_lvbptr; rgd->rd_flags &= ~GFS2_RDF_UPTODATE; if (rgd->rd_data > sdp->sd_max_rg_data) sdp->sd_max_rg_data = rgd->rd_data; @@ -1678,13 +1682,105 @@ static void try_rgrp_unlink(struct gfs2_rgrpd *rgd, u64 *last_unlinked, u64 skip return; } +/** + * gfs2_rgrp_congested - Use stats to figure out whether an rgrp is congested + * @rgd: The rgrp in question + * @loops: An indication of how picky we can be (0=very, 1=less so) + * + * This function uses the recently added glock statistics in order to + * figure out whether a parciular resource group is suffering from + * contention from multiple nodes. This is done purely on the basis + * of timings, since this is the only data we have to work with and + * our aim here is to reject a resource group which is highly contended + * but (very important) not to do this too often in order to ensure that + * we do not land up introducing fragmentation by changing resource + * groups when not actually required. + * + * The calculation is fairly simple, we want to know whether the SRTTB + * (i.e. smoothed round trip time for blocking operations) to acquire + * the lock for this rgrp's glock is significantly greater than the + * time taken for resource groups on average. We introduce a margin in + * the form of the variable @var which is computed as the sum of the two + * respective variences, and multiplied by a factor depending on @loops + * and whether we have a lot of data to base the decision on. This is + * then tested against the square difference of the means in order to + * decide whether the result is statistically significant or not. + * + * Returns: A boolean verdict on the congestion status + */ + +static bool gfs2_rgrp_congested(const struct gfs2_rgrpd *rgd, int loops) +{ + const struct gfs2_glock *gl = rgd->rd_gl; + const struct gfs2_sbd *sdp = gl->gl_sbd; + struct gfs2_lkstats *st; + s64 r_dcount, l_dcount; + s64 r_srttb, l_srttb; + s64 srttb_diff; + s64 sqr_diff; + s64 var; + + preempt_disable(); + st = &this_cpu_ptr(sdp->sd_lkstats)->lkstats[LM_TYPE_RGRP]; + r_srttb = st->stats[GFS2_LKS_SRTTB]; + r_dcount = st->stats[GFS2_LKS_DCOUNT]; + var = st->stats[GFS2_LKS_SRTTVARB] + + gl->gl_stats.stats[GFS2_LKS_SRTTVARB]; + preempt_enable(); + + l_srttb = gl->gl_stats.stats[GFS2_LKS_SRTTB]; + l_dcount = gl->gl_stats.stats[GFS2_LKS_DCOUNT]; + + if ((l_dcount < 1) || (r_dcount < 1) || (r_srttb == 0)) + return false; + + srttb_diff = r_srttb - l_srttb; + sqr_diff = srttb_diff * srttb_diff; + + var *= 2; + if (l_dcount < 8 || r_dcount < 8) + var *= 2; + if (loops == 1) + var *= 2; + + return ((srttb_diff < 0) && (sqr_diff > var)); +} + +/** + * gfs2_rgrp_used_recently + * @rs: The block reservation with the rgrp to test + * @msecs: The time limit in milliseconds + * + * Returns: True if the rgrp glock has been used within the time limit + */ +static bool gfs2_rgrp_used_recently(const struct gfs2_blkreserv *rs, + u64 msecs) +{ + u64 tdiff; + + tdiff = ktime_to_ns(ktime_sub(ktime_get_real(), + rs->rs_rbm.rgd->rd_gl->gl_dstamp)); + + return tdiff > (msecs * 1000 * 1000); +} + +static u32 gfs2_orlov_skip(const struct gfs2_inode *ip) +{ + const struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); + u32 skip; + + get_random_bytes(&skip, sizeof(skip)); + return skip % sdp->sd_rgrps; +} + static bool gfs2_select_rgrp(struct gfs2_rgrpd **pos, const struct gfs2_rgrpd *begin) { struct gfs2_rgrpd *rgd = *pos; + struct gfs2_sbd *sdp = rgd->rd_sbd; rgd = gfs2_rgrpd_get_next(rgd); if (rgd == NULL) - rgd = gfs2_rgrpd_get_next(NULL); + rgd = gfs2_rgrpd_get_first(sdp); *pos = rgd; if (rgd != begin) /* If we didn't wrap */ return true; @@ -1699,14 +1795,15 @@ static bool gfs2_select_rgrp(struct gfs2_rgrpd **pos, const struct gfs2_rgrpd *b * Returns: errno */ -int gfs2_inplace_reserve(struct gfs2_inode *ip, u32 requested) +int gfs2_inplace_reserve(struct gfs2_inode *ip, u32 requested, u32 aflags) { struct gfs2_sbd *sdp = GFS2_SB(&ip->i_inode); struct gfs2_rgrpd *begin = NULL; struct gfs2_blkreserv *rs = ip->i_res; - int error = 0, rg_locked, flags = LM_FLAG_TRY; + int error = 0, rg_locked, flags = 0; u64 last_unlinked = NO_BLOCK; int loops = 0; + u32 skip = 0; if (sdp->sd_args.ar_rgrplvb) flags |= GL_SKIP; @@ -1720,6 +1817,8 @@ int gfs2_inplace_reserve(struct gfs2_inode *ip, u32 requested) } else { rs->rs_rbm.rgd = begin = gfs2_blk2rgrpd(sdp, ip->i_goal, 1); } + if (S_ISDIR(ip->i_inode.i_mode) && (aflags & GFS2_AF_ORLOV)) + skip = gfs2_orlov_skip(ip); if (rs->rs_rbm.rgd == NULL) return -EBADSLT; @@ -1728,13 +1827,20 @@ int gfs2_inplace_reserve(struct gfs2_inode *ip, u32 requested) if (!gfs2_glock_is_locked_by_me(rs->rs_rbm.rgd->rd_gl)) { rg_locked = 0; + if (skip && skip--) + goto next_rgrp; + if (!gfs2_rs_active(rs) && (loops < 2) && + gfs2_rgrp_used_recently(rs, 1000) && + gfs2_rgrp_congested(rs->rs_rbm.rgd, loops)) + goto next_rgrp; error = gfs2_glock_nq_init(rs->rs_rbm.rgd->rd_gl, LM_ST_EXCLUSIVE, flags, &rs->rs_rgd_gh); - if (error == GLR_TRYFAILED) - goto next_rgrp; if (unlikely(error)) return error; + if (!gfs2_rs_active(rs) && (loops < 2) && + gfs2_rgrp_congested(rs->rs_rbm.rgd, loops)) + goto skip_rgrp; if (sdp->sd_args.ar_rgrplvb) { error = update_rgrp_lvb(rs->rs_rbm.rgd); if (unlikely(error)) { @@ -1781,12 +1887,13 @@ next_rgrp: /* Find the next rgrp, and continue looking */ if (gfs2_select_rgrp(&rs->rs_rbm.rgd, begin)) continue; + if (skip) + continue; /* If we've scanned all the rgrps, but found no free blocks * then this checks for some less likely conditions before * trying again. */ - flags &= ~LM_FLAG_TRY; loops++; /* Check that fs hasn't grown if writing to rindex */ if (ip == GFS2_I(sdp->sd_rindex) && !sdp->sd_rindex_uptodate) { diff --git a/fs/gfs2/rgrp.h b/fs/gfs2/rgrp.h index 24077958dcf6..842185853f6b 100644 --- a/fs/gfs2/rgrp.h +++ b/fs/gfs2/rgrp.h @@ -39,7 +39,8 @@ extern void gfs2_rgrp_go_unlock(struct gfs2_holder *gh); extern struct gfs2_alloc *gfs2_alloc_get(struct gfs2_inode *ip); -extern int gfs2_inplace_reserve(struct gfs2_inode *ip, u32 requested); +#define GFS2_AF_ORLOV 1 +extern int gfs2_inplace_reserve(struct gfs2_inode *ip, u32 requested, u32 flags); extern void gfs2_inplace_release(struct gfs2_inode *ip); extern int gfs2_alloc_blocks(struct gfs2_inode *ip, u64 *bn, unsigned int *n, diff --git a/fs/gfs2/trace_gfs2.h b/fs/gfs2/trace_gfs2.h index bbdc78af60ca..2ee13e841e9f 100644 --- a/fs/gfs2/trace_gfs2.h +++ b/fs/gfs2/trace_gfs2.h @@ -486,7 +486,7 @@ TRACE_EVENT(gfs2_block_alloc, ), TP_fast_assign( - __entry->dev = ip->i_gl->gl_sbd->sd_vfs->s_dev; + __entry->dev = rgd->rd_gl->gl_sbd->sd_vfs->s_dev; __entry->start = block; __entry->inum = ip->i_no_addr; __entry->len = len; diff --git a/fs/gfs2/xattr.c b/fs/gfs2/xattr.c index db330e5518cd..76c144b3c9bb 100644 --- a/fs/gfs2/xattr.c +++ b/fs/gfs2/xattr.c @@ -734,7 +734,7 @@ static int ea_alloc_skeleton(struct gfs2_inode *ip, struct gfs2_ea_request *er, if (error) return error; - error = gfs2_inplace_reserve(ip, blks); + error = gfs2_inplace_reserve(ip, blks, 0); if (error) goto out_gunlock_q; diff --git a/fs/hugetlbfs/inode.c b/fs/hugetlbfs/inode.c index c5bc355d8243..78bde32ea951 100644 --- a/fs/hugetlbfs/inode.c +++ b/fs/hugetlbfs/inode.c @@ -1,7 +1,7 @@ /* * hugetlbpage-backed filesystem. Based on ramfs. * - * William Irwin, 2002 + * Nadia Yvette Chambers, 2002 * * Copyright (C) 2002 Linus Torvalds. */ @@ -151,8 +151,8 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, { struct mm_struct *mm = current->mm; struct vm_area_struct *vma; - unsigned long start_addr; struct hstate *h = hstate_file(file); + struct vm_unmapped_area_info info; if (len & ~huge_page_mask(h)) return -EINVAL; @@ -173,39 +173,13 @@ hugetlb_get_unmapped_area(struct file *file, unsigned long addr, return addr; } - if (len > mm->cached_hole_size) - start_addr = mm->free_area_cache; - else { - start_addr = TASK_UNMAPPED_BASE; - mm->cached_hole_size = 0; - } - -full_search: - addr = ALIGN(start_addr, huge_page_size(h)); - - for (vma = find_vma(mm, addr); ; vma = vma->vm_next) { - /* At this point: (!vma || addr < vma->vm_end). */ - if (TASK_SIZE - len < addr) { - /* - * Start a new search - just in case we missed - * some holes. - */ - if (start_addr != TASK_UNMAPPED_BASE) { - start_addr = TASK_UNMAPPED_BASE; - mm->cached_hole_size = 0; - goto full_search; - } - return -ENOMEM; - } - - if (!vma || addr + len <= vma->vm_start) { - mm->free_area_cache = addr + len; - return addr; - } - if (addr + mm->cached_hole_size < vma->vm_start) - mm->cached_hole_size = vma->vm_start - addr; - addr = ALIGN(vma->vm_end, huge_page_size(h)); - } + info.flags = 0; + info.length = len; + info.low_limit = TASK_UNMAPPED_BASE; + info.high_limit = TASK_SIZE; + info.align_mask = PAGE_MASK & ~huge_page_mask(h); + info.align_offset = 0; + return vm_unmapped_area(&info); } #endif @@ -608,11 +582,11 @@ static int hugetlbfs_migrate_page(struct address_space *mapping, int rc; rc = migrate_huge_page_move_mapping(mapping, newpage, page); - if (rc) + if (rc != MIGRATEPAGE_SUCCESS) return rc; migrate_page_copy(newpage, page); - return 0; + return MIGRATEPAGE_SUCCESS; } static int hugetlbfs_statfs(struct dentry *dentry, struct kstatfs *buf) @@ -923,7 +897,7 @@ static struct file_system_type hugetlbfs_fs_type = { .kill_sb = kill_litter_super, }; -static struct vfsmount *hugetlbfs_vfsmount; +static struct vfsmount *hugetlbfs_vfsmount[HUGE_MAX_HSTATE]; static int can_do_hugetlb_shm(void) { @@ -932,9 +906,22 @@ static int can_do_hugetlb_shm(void) return capable(CAP_IPC_LOCK) || in_group_p(shm_group); } +static int get_hstate_idx(int page_size_log) +{ + struct hstate *h; + + if (!page_size_log) + return default_hstate_idx; + h = size_to_hstate(1 << page_size_log); + if (!h) + return -1; + return h - hstates; +} + struct file *hugetlb_file_setup(const char *name, unsigned long addr, size_t size, vm_flags_t acctflag, - struct user_struct **user, int creat_flags) + struct user_struct **user, + int creat_flags, int page_size_log) { int error = -ENOMEM; struct file *file; @@ -944,9 +931,14 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, struct qstr quick_string; struct hstate *hstate; unsigned long num_pages; + int hstate_idx; + + hstate_idx = get_hstate_idx(page_size_log); + if (hstate_idx < 0) + return ERR_PTR(-ENODEV); *user = NULL; - if (!hugetlbfs_vfsmount) + if (!hugetlbfs_vfsmount[hstate_idx]) return ERR_PTR(-ENOENT); if (creat_flags == HUGETLB_SHMFS_INODE && !can_do_hugetlb_shm()) { @@ -963,7 +955,7 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, } } - root = hugetlbfs_vfsmount->mnt_root; + root = hugetlbfs_vfsmount[hstate_idx]->mnt_root; quick_string.name = name; quick_string.len = strlen(quick_string.name); quick_string.hash = 0; @@ -971,7 +963,7 @@ struct file *hugetlb_file_setup(const char *name, unsigned long addr, if (!path.dentry) goto out_shm_unlock; - path.mnt = mntget(hugetlbfs_vfsmount); + path.mnt = mntget(hugetlbfs_vfsmount[hstate_idx]); error = -ENOSPC; inode = hugetlbfs_get_inode(root->d_sb, NULL, S_IFREG | S_IRWXUGO, 0); if (!inode) @@ -1011,8 +1003,9 @@ out_shm_unlock: static int __init init_hugetlbfs_fs(void) { + struct hstate *h; int error; - struct vfsmount *vfsmount; + int i; error = bdi_init(&hugetlbfs_backing_dev_info); if (error) @@ -1029,14 +1022,26 @@ static int __init init_hugetlbfs_fs(void) if (error) goto out; - vfsmount = kern_mount(&hugetlbfs_fs_type); + i = 0; + for_each_hstate(h) { + char buf[50]; + unsigned ps_kb = 1U << (h->order + PAGE_SHIFT - 10); - if (!IS_ERR(vfsmount)) { - hugetlbfs_vfsmount = vfsmount; - return 0; - } + snprintf(buf, sizeof(buf), "pagesize=%uK", ps_kb); + hugetlbfs_vfsmount[i] = kern_mount_data(&hugetlbfs_fs_type, + buf); - error = PTR_ERR(vfsmount); + if (IS_ERR(hugetlbfs_vfsmount[i])) { + pr_err("hugetlb: Cannot mount internal hugetlbfs for " + "page size %uK", ps_kb); + error = PTR_ERR(hugetlbfs_vfsmount[i]); + hugetlbfs_vfsmount[i] = NULL; + } + i++; + } + /* Non default hstates are optional */ + if (!IS_ERR_OR_NULL(hugetlbfs_vfsmount[default_hstate_idx])) + return 0; out: kmem_cache_destroy(hugetlbfs_inode_cachep); @@ -1047,13 +1052,19 @@ static int __init init_hugetlbfs_fs(void) static void __exit exit_hugetlbfs_fs(void) { + struct hstate *h; + int i; + + /* * Make sure all delayed rcu free inodes are flushed before we * destroy cache. */ rcu_barrier(); kmem_cache_destroy(hugetlbfs_inode_cachep); - kern_unmount(hugetlbfs_vfsmount); + i = 0; + for_each_hstate(h) + kern_unmount(hugetlbfs_vfsmount[i++]); unregister_filesystem(&hugetlbfs_fs_type); bdi_destroy(&hugetlbfs_backing_dev_info); } diff --git a/fs/inode.c b/fs/inode.c index b03c71957246..14084b72b259 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -165,7 +165,7 @@ int inode_init_always(struct super_block *sb, struct inode *inode) mapping->host = inode; mapping->flags = 0; mapping_set_gfp_mask(mapping, GFP_HIGHUSER_MOVABLE); - mapping->assoc_mapping = NULL; + mapping->private_data = NULL; mapping->backing_dev_info = &default_backing_dev_info; mapping->writeback_index = 0; @@ -408,6 +408,19 @@ static void inode_lru_list_add(struct inode *inode) spin_unlock(&inode->i_sb->s_inode_lru_lock); } +/* + * Add inode to LRU if needed (inode is unused and clean). + * + * Needs inode->i_lock held. + */ +void inode_add_lru(struct inode *inode) +{ + if (!(inode->i_state & (I_DIRTY | I_SYNC | I_FREEING | I_WILL_FREE)) && + !atomic_read(&inode->i_count) && inode->i_sb->s_flags & MS_ACTIVE) + inode_lru_list_add(inode); +} + + static void inode_lru_list_del(struct inode *inode) { spin_lock(&inode->i_sb->s_inode_lru_lock); @@ -1390,8 +1403,7 @@ static void iput_final(struct inode *inode) if (!drop && (sb->s_flags & MS_ACTIVE)) { inode->i_state |= I_REFERENCED; - if (!(inode->i_state & (I_DIRTY|I_SYNC))) - inode_lru_list_add(inode); + inode_add_lru(inode); spin_unlock(&inode->i_lock); return; } diff --git a/fs/internal.h b/fs/internal.h index 916b7cbf3e3e..2f6af7f645eb 100644 --- a/fs/internal.h +++ b/fs/internal.h @@ -110,6 +110,7 @@ extern int open_check_o_direct(struct file *f); * inode.c */ extern spinlock_t inode_sb_list_lock; +extern void inode_add_lru(struct inode *inode); /* * fs-writeback.c diff --git a/fs/jbd/transaction.c b/fs/jbd/transaction.c index 78b7f84241d4..071d6905f0dd 100644 --- a/fs/jbd/transaction.c +++ b/fs/jbd/transaction.c @@ -1259,7 +1259,7 @@ int journal_forget (handle_t *handle, struct buffer_head *bh) goto not_jbd; } - /* keep track of wether or not this transaction modified us */ + /* keep track of whether or not this transaction modified us */ was_modified = jh->b_modified; /* @@ -1961,7 +1961,9 @@ retry: spin_unlock(&journal->j_list_lock); jbd_unlock_bh_state(bh); spin_unlock(&journal->j_state_lock); + unlock_buffer(bh); log_wait_commit(journal, tid); + lock_buffer(bh); goto retry; } /* diff --git a/fs/jbd2/journal.c b/fs/jbd2/journal.c index 484b8d1c6cb6..dbf41f9452db 100644 --- a/fs/jbd2/journal.c +++ b/fs/jbd2/journal.c @@ -60,7 +60,6 @@ EXPORT_SYMBOL(jbd2_journal_get_create_access); EXPORT_SYMBOL(jbd2_journal_get_undo_access); EXPORT_SYMBOL(jbd2_journal_set_triggers); EXPORT_SYMBOL(jbd2_journal_dirty_metadata); -EXPORT_SYMBOL(jbd2_journal_release_buffer); EXPORT_SYMBOL(jbd2_journal_forget); #if 0 EXPORT_SYMBOL(journal_sync_buffer); diff --git a/fs/jbd2/transaction.c b/fs/jbd2/transaction.c index a74ba4659549..42f6615af0ac 100644 --- a/fs/jbd2/transaction.c +++ b/fs/jbd2/transaction.c @@ -1207,17 +1207,6 @@ out: return ret; } -/* - * jbd2_journal_release_buffer: undo a get_write_access without any buffer - * updates, if the update decided in the end that it didn't need access. - * - */ -void -jbd2_journal_release_buffer(handle_t *handle, struct buffer_head *bh) -{ - BUFFER_TRACE(bh, "entry"); -} - /** * void jbd2_journal_forget() - bforget() for potentially-journaled buffers. * @handle: transaction handle @@ -1261,7 +1250,7 @@ int jbd2_journal_forget (handle_t *handle, struct buffer_head *bh) goto not_jbd; } - /* keep track of wether or not this transaction modified us */ + /* keep track of whether or not this transaction modified us */ was_modified = jh->b_modified; /* diff --git a/fs/logfs/inode.c b/fs/logfs/inode.c index adb90116d36b..af49e2d6941a 100644 --- a/fs/logfs/inode.c +++ b/fs/logfs/inode.c @@ -33,7 +33,7 @@ * are being written out - and waiting for GC to make progress, naturally. * * So we cannot just call iget() or some variant of it, but first have to check - * wether the inode in question might be in I_FREEING state. Therefore we + * whether the inode in question might be in I_FREEING state. Therefore we * maintain our own per-sb list of "almost deleted" inodes and check against * that list first. Normally this should be at most 1-2 entries long. * diff --git a/fs/namei.c b/fs/namei.c index 937f9d50c84b..5f4cdf3ad913 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2131,6 +2131,11 @@ struct dentry *lookup_one_len(const char *name, struct dentry *base, int len) if (!len) return ERR_PTR(-EACCES); + if (unlikely(name[0] == '.')) { + if (len < 2 || (len == 2 && name[1] == '.')) + return ERR_PTR(-EACCES); + } + while (len--) { c = *(const unsigned char *)name++; if (c == '/' || c == '\0') diff --git a/fs/ncpfs/mmap.c b/fs/ncpfs/mmap.c index be20a7e171a0..63d14a99483d 100644 --- a/fs/ncpfs/mmap.c +++ b/fs/ncpfs/mmap.c @@ -89,7 +89,7 @@ static int ncp_file_mmap_fault(struct vm_area_struct *area, /* * If I understand ncp_read_kernel() properly, the above always * fetches from the network, here the analogue of disk. - * -- wli + * -- nyc */ count_vm_event(PGMAJFAULT); mem_cgroup_count_vm_event(area->vm_mm, PGMAJFAULT); diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c index ce8cb926526b..b9e66b7e0c14 100644 --- a/fs/nfs/dir.c +++ b/fs/nfs/dir.c @@ -450,7 +450,8 @@ void nfs_prime_dcache(struct dentry *parent, struct nfs_entry *entry) nfs_refresh_inode(dentry->d_inode, entry->fattr); goto out; } else { - d_drop(dentry); + if (d_invalidate(dentry) != 0) + goto out; dput(dentry); } } @@ -1100,6 +1101,8 @@ out_set_verifier: out_zap_parent: nfs_zap_caches(dir); out_bad: + nfs_free_fattr(fattr); + nfs_free_fhandle(fhandle); nfs_mark_for_revalidate(dir); if (inode && S_ISDIR(inode->i_mode)) { /* Purge readdir caches. */ @@ -1112,8 +1115,6 @@ out_zap_parent: shrink_dcache_parent(dentry); } d_drop(dentry); - nfs_free_fattr(fattr); - nfs_free_fhandle(fhandle); dput(parent); dfprintk(LOOKUPCACHE, "NFS: %s(%s/%s) is invalid\n", __func__, dentry->d_parent->d_name.name, diff --git a/fs/nfs/idmap.c b/fs/nfs/idmap.c index 9cc4a3fbf4b0..bc3968fa81e5 100644 --- a/fs/nfs/idmap.c +++ b/fs/nfs/idmap.c @@ -193,19 +193,15 @@ static int nfs_idmap_init_keyring(void) if (!cred) return -ENOMEM; - keyring = key_alloc(&key_type_keyring, ".id_resolver", 0, 0, cred, - (KEY_POS_ALL & ~KEY_POS_SETATTR) | - KEY_USR_VIEW | KEY_USR_READ, - KEY_ALLOC_NOT_IN_QUOTA); + keyring = keyring_alloc(".id_resolver", 0, 0, cred, + (KEY_POS_ALL & ~KEY_POS_SETATTR) | + KEY_USR_VIEW | KEY_USR_READ, + KEY_ALLOC_NOT_IN_QUOTA, NULL); 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(&key_type_id_resolver); if (ret < 0) goto failed_put_key; diff --git a/fs/nilfs2/page.c b/fs/nilfs2/page.c index 3e7b2a0dc0c8..07f76db04ec7 100644 --- a/fs/nilfs2/page.c +++ b/fs/nilfs2/page.c @@ -431,7 +431,7 @@ void nilfs_mapping_init(struct address_space *mapping, struct inode *inode, mapping->host = inode; mapping->flags = 0; mapping_set_gfp_mask(mapping, GFP_NOFS); - mapping->assoc_mapping = NULL; + mapping->private_data = NULL; mapping->backing_dev_info = bdi; mapping->a_ops = &empty_aops; } diff --git a/fs/notify/fanotify/Kconfig b/fs/notify/fanotify/Kconfig index 7dceff005a67..e5f911bd80d2 100644 --- a/fs/notify/fanotify/Kconfig +++ b/fs/notify/fanotify/Kconfig @@ -4,7 +4,7 @@ config FANOTIFY select ANON_INODES default n ---help--- - Say Y here to enable fanotify suport. fanotify is a file access + Say Y here to enable fanotify support. fanotify is a file access notification system which differs from inotify in that it sends an open file descriptor to the userspace listener along with the event. diff --git a/fs/notify/notification.c b/fs/notify/notification.c index c887b1378f7e..48cb994e4922 100644 --- a/fs/notify/notification.c +++ b/fs/notify/notification.c @@ -18,7 +18,7 @@ /* * Basic idea behind the notification queue: An fsnotify group (like inotify) - * sends the userspace notification about events asyncronously some time after + * sends the userspace notification about events asynchronously some time after * the event happened. When inotify gets an event it will need to add that * event to the group notify queue. Since a single event might need to be on * multiple group's notification queues we can't add the event directly to each diff --git a/fs/ocfs2/file.c b/fs/ocfs2/file.c index 5a4ee77cec51..dda089804942 100644 --- a/fs/ocfs2/file.c +++ b/fs/ocfs2/file.c @@ -2513,18 +2513,15 @@ static ssize_t ocfs2_file_splice_write(struct pipe_inode_info *pipe, ret = sd.num_spliced; if (ret > 0) { - unsigned long nr_pages; int err; - nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - err = generic_write_sync(out, *ppos, ret); if (err) ret = err; else *ppos += ret; - balance_dirty_pages_ratelimited_nr(mapping, nr_pages); + balance_dirty_pages_ratelimited(mapping); } return ret; diff --git a/fs/proc/array.c b/fs/proc/array.c index c1c207c36cae..d3696708fc1a 100644 --- a/fs/proc/array.c +++ b/fs/proc/array.c @@ -438,7 +438,7 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns, min_flt += sig->min_flt; maj_flt += sig->maj_flt; - thread_group_times(task, &utime, &stime); + thread_group_cputime_adjusted(task, &utime, &stime); gtime += sig->gtime; } @@ -454,7 +454,7 @@ static int do_task_stat(struct seq_file *m, struct pid_namespace *ns, if (!whole) { min_flt = task->min_flt; maj_flt = task->maj_flt; - task_times(task, &utime, &stime); + task_cputime_adjusted(task, &utime, &stime); gtime = task->gtime; } diff --git a/fs/proc/base.c b/fs/proc/base.c index 3c231adf8450..aa63d25157b8 100644 --- a/fs/proc/base.c +++ b/fs/proc/base.c @@ -985,7 +985,7 @@ static ssize_t oom_score_adj_read(struct file *file, char __user *buf, { struct task_struct *task = get_proc_task(file->f_path.dentry->d_inode); char buffer[PROC_NUMBUF]; - int oom_score_adj = OOM_SCORE_ADJ_MIN; + short oom_score_adj = OOM_SCORE_ADJ_MIN; unsigned long flags; size_t len; @@ -996,7 +996,7 @@ static ssize_t oom_score_adj_read(struct file *file, char __user *buf, unlock_task_sighand(task, &flags); } put_task_struct(task); - len = snprintf(buffer, sizeof(buffer), "%d\n", oom_score_adj); + len = snprintf(buffer, sizeof(buffer), "%hd\n", oom_score_adj); return simple_read_from_buffer(buf, count, ppos, buffer, len); } @@ -1043,15 +1043,15 @@ static ssize_t oom_score_adj_write(struct file *file, const char __user *buf, goto err_task_lock; } - if (oom_score_adj < task->signal->oom_score_adj_min && + if ((short)oom_score_adj < task->signal->oom_score_adj_min && !capable(CAP_SYS_RESOURCE)) { err = -EACCES; goto err_sighand; } - task->signal->oom_score_adj = oom_score_adj; + task->signal->oom_score_adj = (short)oom_score_adj; if (has_capability_noaudit(current, CAP_SYS_RESOURCE)) - task->signal->oom_score_adj_min = oom_score_adj; + task->signal->oom_score_adj_min = (short)oom_score_adj; trace_oom_score_adj_update(task); err_sighand: @@ -1877,8 +1877,9 @@ static struct dentry *proc_map_files_lookup(struct inode *dir, if (!vma) goto out_no_vma; - result = proc_map_files_instantiate(dir, dentry, task, - (void *)(unsigned long)vma->vm_file->f_mode); + if (vma->vm_file) + result = proc_map_files_instantiate(dir, dentry, task, + (void *)(unsigned long)vma->vm_file->f_mode); out_no_vma: up_read(&mm->mmap_sem); diff --git a/fs/proc/kcore.c b/fs/proc/kcore.c index 86c67eee439f..e96d4f18ca3a 100644 --- a/fs/proc/kcore.c +++ b/fs/proc/kcore.c @@ -249,7 +249,7 @@ static int kcore_update_ram(void) /* Not inialized....update now */ /* find out "max pfn" */ end_pfn = 0; - for_each_node_state(nid, N_HIGH_MEMORY) { + for_each_node_state(nid, N_MEMORY) { unsigned long node_end; node_end = NODE_DATA(nid)->node_start_pfn + NODE_DATA(nid)->node_spanned_pages; diff --git a/fs/proc/proc_sysctl.c b/fs/proc/proc_sysctl.c index a781bdf06694..701580ddfcc3 100644 --- a/fs/proc/proc_sysctl.c +++ b/fs/proc/proc_sysctl.c @@ -378,12 +378,13 @@ static int test_perm(int mode, int op) return -EACCES; } -static int sysctl_perm(struct ctl_table_root *root, struct ctl_table *table, int op) +static int sysctl_perm(struct ctl_table_header *head, struct ctl_table *table, int op) { + struct ctl_table_root *root = head->root; int mode; if (root->permissions) - mode = root->permissions(root, current->nsproxy, table); + mode = root->permissions(head, table); else mode = table->mode; @@ -491,7 +492,7 @@ static ssize_t proc_sys_call_handler(struct file *filp, void __user *buf, * and won't be until we finish. */ error = -EPERM; - if (sysctl_perm(head->root, table, write ? MAY_WRITE : MAY_READ)) + if (sysctl_perm(head, table, write ? MAY_WRITE : MAY_READ)) goto out; /* if that can happen at all, it should be -EINVAL, not -EISDIR */ @@ -717,7 +718,7 @@ static int proc_sys_permission(struct inode *inode, int mask) if (!table) /* global root - r-xr-xr-x */ error = mask & MAY_WRITE ? -EACCES : 0; else /* Use the permissions on the sysctl table entry */ - error = sysctl_perm(head->root, table, mask & ~MAY_NOT_BLOCK); + error = sysctl_perm(head, table, mask & ~MAY_NOT_BLOCK); sysctl_head_finish(head); return error; diff --git a/fs/proc/task_mmu.c b/fs/proc/task_mmu.c index 90c63f9392a5..48775628abbf 100644 --- a/fs/proc/task_mmu.c +++ b/fs/proc/task_mmu.c @@ -643,7 +643,7 @@ static int clear_refs_pte_range(pmd_t *pmd, unsigned long addr, spinlock_t *ptl; struct page *page; - split_huge_page_pmd(walk->mm, pmd); + split_huge_page_pmd(vma, addr, pmd); if (pmd_trans_unstable(pmd)) return 0; @@ -1126,7 +1126,7 @@ static struct page *can_gather_numa_stats(pte_t pte, struct vm_area_struct *vma, return NULL; nid = page_to_nid(page); - if (!node_isset(nid, node_states[N_HIGH_MEMORY])) + if (!node_isset(nid, node_states[N_MEMORY])) return NULL; return page; @@ -1279,7 +1279,7 @@ static int show_numa_map(struct seq_file *m, void *v, int is_pid) if (md->writeback) seq_printf(m, " writeback=%lu", md->writeback); - for_each_node_state(n, N_HIGH_MEMORY) + for_each_node_state(n, N_MEMORY) if (md->node[n]) seq_printf(m, " N%d=%lu", n, md->node[n]); out: diff --git a/fs/pstore/ftrace.c b/fs/pstore/ftrace.c index 2d57e1ac0115..43b12807a51d 100644 --- a/fs/pstore/ftrace.c +++ b/fs/pstore/ftrace.c @@ -28,7 +28,9 @@ #include "internal.h" static void notrace pstore_ftrace_call(unsigned long ip, - unsigned long parent_ip) + unsigned long parent_ip, + struct ftrace_ops *op, + struct pt_regs *regs) { unsigned long flags; struct pstore_ftrace_record rec = {}; diff --git a/fs/pstore/inode.c b/fs/pstore/inode.c index 4ab572e6d277..ed1d8c7212da 100644 --- a/fs/pstore/inode.c +++ b/fs/pstore/inode.c @@ -49,6 +49,7 @@ struct pstore_private { struct pstore_info *psi; enum pstore_type_id type; u64 id; + int count; ssize_t size; char data[]; }; @@ -175,7 +176,8 @@ static int pstore_unlink(struct inode *dir, struct dentry *dentry) struct pstore_private *p = dentry->d_inode->i_private; if (p->psi->erase) - p->psi->erase(p->type, p->id, p->psi); + p->psi->erase(p->type, p->id, p->count, + dentry->d_inode->i_ctime, p->psi); return simple_unlink(dir, dentry); } @@ -270,7 +272,7 @@ int pstore_is_mounted(void) * Load it up with "size" bytes of data from "buf". * Set the mtime & ctime to the date that this record was originally stored. */ -int pstore_mkfile(enum pstore_type_id type, char *psname, u64 id, +int pstore_mkfile(enum pstore_type_id type, char *psname, u64 id, int count, char *data, size_t size, struct timespec time, struct pstore_info *psi) { @@ -306,6 +308,7 @@ int pstore_mkfile(enum pstore_type_id type, char *psname, u64 id, goto fail_alloc; private->type = type; private->id = id; + private->count = count; private->psi = psi; switch (type) { diff --git a/fs/pstore/internal.h b/fs/pstore/internal.h index 4847f588b7d5..937d820f273c 100644 --- a/fs/pstore/internal.h +++ b/fs/pstore/internal.h @@ -50,7 +50,7 @@ extern struct pstore_info *psinfo; extern void pstore_set_kmsg_bytes(int); extern void pstore_get_records(int); extern int pstore_mkfile(enum pstore_type_id, char *psname, u64 id, - char *data, size_t size, + int count, char *data, size_t size, struct timespec time, struct pstore_info *psi); extern int pstore_is_mounted(void); diff --git a/fs/pstore/platform.c b/fs/pstore/platform.c index 947fbe06c3b1..5ea2e77ff023 100644 --- a/fs/pstore/platform.c +++ b/fs/pstore/platform.c @@ -136,7 +136,7 @@ static void pstore_dump(struct kmsg_dumper *dumper, break; ret = psinfo->write(PSTORE_TYPE_DMESG, reason, &id, part, - hsize + len, psinfo); + oopscount, hsize + len, psinfo); if (ret == 0 && reason == KMSG_DUMP_OOPS && pstore_is_mounted()) pstore_new_entry = 1; @@ -173,7 +173,7 @@ static void pstore_console_write(struct console *con, const char *s, unsigned c) spin_lock_irqsave(&psinfo->buf_lock, flags); } memcpy(psinfo->buf, s, c); - psinfo->write(PSTORE_TYPE_CONSOLE, 0, &id, 0, c, psinfo); + psinfo->write(PSTORE_TYPE_CONSOLE, 0, &id, 0, 0, c, psinfo); spin_unlock_irqrestore(&psinfo->buf_lock, flags); s += c; c = e - s; @@ -197,7 +197,7 @@ static void pstore_register_console(void) {} static int pstore_write_compat(enum pstore_type_id type, enum kmsg_dump_reason reason, - u64 *id, unsigned int part, + u64 *id, unsigned int part, int count, size_t size, struct pstore_info *psi) { return psi->write_buf(type, reason, id, part, psinfo->buf, size, psi); @@ -267,6 +267,7 @@ void pstore_get_records(int quiet) char *buf = NULL; ssize_t size; u64 id; + int count; enum pstore_type_id type; struct timespec time; int failed = 0, rc; @@ -278,9 +279,9 @@ void pstore_get_records(int quiet) if (psi->open && psi->open(psi)) goto out; - while ((size = psi->read(&id, &type, &time, &buf, psi)) > 0) { - rc = pstore_mkfile(type, psi->name, id, buf, (size_t)size, - time, psi); + while ((size = psi->read(&id, &type, &count, &time, &buf, psi)) > 0) { + rc = pstore_mkfile(type, psi->name, id, count, buf, + (size_t)size, time, psi); kfree(buf); buf = NULL; if (rc && (rc != -EEXIST || !quiet)) diff --git a/fs/pstore/ram.c b/fs/pstore/ram.c index 1a4f6da58eab..f883e7e74305 100644 --- a/fs/pstore/ram.c +++ b/fs/pstore/ram.c @@ -132,9 +132,8 @@ ramoops_get_next_prz(struct persistent_ram_zone *przs[], uint *c, uint max, } static ssize_t ramoops_pstore_read(u64 *id, enum pstore_type_id *type, - struct timespec *time, - char **buf, - struct pstore_info *psi) + int *count, struct timespec *time, + char **buf, struct pstore_info *psi) { ssize_t size; struct ramoops_context *cxt = psi->data; @@ -189,7 +188,7 @@ static int notrace ramoops_pstore_write_buf(enum pstore_type_id type, struct pstore_info *psi) { struct ramoops_context *cxt = psi->data; - struct persistent_ram_zone *prz = cxt->przs[cxt->dump_write_cnt]; + struct persistent_ram_zone *prz; size_t hlen; if (type == PSTORE_TYPE_CONSOLE) { @@ -226,6 +225,11 @@ static int notrace ramoops_pstore_write_buf(enum pstore_type_id type, if (part != 1) return -ENOSPC; + if (!cxt->przs) + return -ENOSPC; + + prz = cxt->przs[cxt->dump_write_cnt]; + hlen = ramoops_write_kmsg_hdr(prz); if (size + hlen > prz->buffer_size) size = prz->buffer_size - hlen; @@ -236,8 +240,8 @@ static int notrace ramoops_pstore_write_buf(enum pstore_type_id type, return 0; } -static int ramoops_pstore_erase(enum pstore_type_id type, u64 id, - struct pstore_info *psi) +static int ramoops_pstore_erase(enum pstore_type_id type, u64 id, int count, + struct timespec time, struct pstore_info *psi) { struct ramoops_context *cxt = psi->data; struct persistent_ram_zone *prz; @@ -287,8 +291,9 @@ static void ramoops_free_przs(struct ramoops_context *cxt) kfree(cxt->przs); } -static int ramoops_init_przs(struct device *dev, struct ramoops_context *cxt, - phys_addr_t *paddr, size_t dump_mem_sz) +static int __devinit ramoops_init_przs(struct device *dev, + struct ramoops_context *cxt, + phys_addr_t *paddr, size_t dump_mem_sz) { int err = -ENOMEM; int i; @@ -296,6 +301,11 @@ static int ramoops_init_przs(struct device *dev, struct ramoops_context *cxt, if (!cxt->record_size) return 0; + if (*paddr + dump_mem_sz - cxt->phys_addr > cxt->size) { + dev_err(dev, "no room for dumps\n"); + return -ENOMEM; + } + cxt->max_dump_cnt = dump_mem_sz / cxt->record_size; if (!cxt->max_dump_cnt) return -ENOMEM; @@ -326,15 +336,20 @@ fail_prz: return err; } -static int ramoops_init_prz(struct device *dev, struct ramoops_context *cxt, - struct persistent_ram_zone **prz, - phys_addr_t *paddr, size_t sz, u32 sig) +static int __devinit ramoops_init_prz(struct device *dev, + struct ramoops_context *cxt, + struct persistent_ram_zone **prz, + phys_addr_t *paddr, size_t sz, u32 sig) { if (!sz) return 0; - if (*paddr + sz > *paddr + cxt->size) + if (*paddr + sz - cxt->phys_addr > cxt->size) { + dev_err(dev, "no room for mem region (0x%zx@0x%llx) in (0x%lx@0x%llx)\n", + sz, (unsigned long long)*paddr, + cxt->size, (unsigned long long)cxt->phys_addr); return -ENOMEM; + } *prz = persistent_ram_new(*paddr, sz, sig, cxt->ecc_size); if (IS_ERR(*prz)) { @@ -374,10 +389,14 @@ static int __devinit ramoops_probe(struct platform_device *pdev) goto fail_out; } - pdata->mem_size = rounddown_pow_of_two(pdata->mem_size); - pdata->record_size = rounddown_pow_of_two(pdata->record_size); - pdata->console_size = rounddown_pow_of_two(pdata->console_size); - pdata->ftrace_size = rounddown_pow_of_two(pdata->ftrace_size); + if (!is_power_of_2(pdata->mem_size)) + pdata->mem_size = rounddown_pow_of_two(pdata->mem_size); + if (!is_power_of_2(pdata->record_size)) + pdata->record_size = rounddown_pow_of_two(pdata->record_size); + if (!is_power_of_2(pdata->console_size)) + pdata->console_size = rounddown_pow_of_two(pdata->console_size); + if (!is_power_of_2(pdata->ftrace_size)) + pdata->ftrace_size = rounddown_pow_of_two(pdata->ftrace_size); cxt->dump_read_cnt = 0; cxt->size = pdata->mem_size; diff --git a/fs/splice.c b/fs/splice.c index 13e5b4776e7a..8890604e3fcd 100644 --- a/fs/splice.c +++ b/fs/splice.c @@ -1024,17 +1024,14 @@ generic_file_splice_write(struct pipe_inode_info *pipe, struct file *out, ret = sd.num_spliced; if (ret > 0) { - unsigned long nr_pages; int err; - nr_pages = (ret + PAGE_CACHE_SIZE - 1) >> PAGE_CACHE_SHIFT; - err = generic_write_sync(out, *ppos, ret); if (err) ret = err; else *ppos += ret; - balance_dirty_pages_ratelimited_nr(mapping, nr_pages); + balance_dirty_pages_ratelimited(mapping); } sb_end_write(inode->i_sb); diff --git a/fs/sysfs/file.c b/fs/sysfs/file.c index 00012e31829d..602f56db0442 100644 --- a/fs/sysfs/file.c +++ b/fs/sysfs/file.c @@ -485,8 +485,8 @@ const struct file_operations sysfs_file_operations = { .poll = sysfs_poll, }; -int sysfs_attr_ns(struct kobject *kobj, const struct attribute *attr, - const void **pns) +static int sysfs_attr_ns(struct kobject *kobj, const struct attribute *attr, + const void **pns) { struct sysfs_dirent *dir_sd = kobj->sd; const struct sysfs_ops *ops; diff --git a/fs/xfs/Kconfig b/fs/xfs/Kconfig index 6100ec0fa1d4..5a7ffe54f5d5 100644 --- a/fs/xfs/Kconfig +++ b/fs/xfs/Kconfig @@ -2,6 +2,7 @@ config XFS_FS tristate "XFS filesystem support" depends on BLOCK select EXPORTFS + select LIBCRC32C help XFS is a high performance journaling filesystem which originated on the SGI IRIX platform. It is completely multi-threaded, can diff --git a/fs/xfs/Makefile b/fs/xfs/Makefile index d2bf974b1a2f..d02201df855b 100644 --- a/fs/xfs/Makefile +++ b/fs/xfs/Makefile @@ -37,9 +37,8 @@ xfs-y += xfs_aops.o \ xfs_file.o \ xfs_filestream.o \ xfs_fsops.o \ - xfs_fs_subr.o \ xfs_globals.o \ - xfs_iget.o \ + xfs_icache.o \ xfs_ioctl.o \ xfs_iomap.o \ xfs_iops.o \ @@ -47,7 +46,6 @@ xfs-y += xfs_aops.o \ xfs_message.o \ xfs_mru_cache.o \ xfs_super.o \ - xfs_sync.o \ xfs_xattr.o \ xfs_rename.o \ xfs_utils.o \ diff --git a/fs/xfs/uuid.h b/fs/xfs/uuid.h index 4732d71262cc..104db0f3bed6 100644 --- a/fs/xfs/uuid.h +++ b/fs/xfs/uuid.h @@ -26,4 +26,10 @@ extern int uuid_is_nil(uuid_t *uuid); extern int uuid_equal(uuid_t *uuid1, uuid_t *uuid2); extern void uuid_getnodeuniq(uuid_t *uuid, int fsid [2]); +static inline void +uuid_copy(uuid_t *dst, uuid_t *src) +{ + memcpy(dst, src, sizeof(uuid_t)); +} + #endif /* __XFS_SUPPORT_UUID_H__ */ diff --git a/fs/xfs/xfs_ag.h b/fs/xfs/xfs_ag.h index 44d65c1533c0..f2aeedb6a579 100644 --- a/fs/xfs/xfs_ag.h +++ b/fs/xfs/xfs_ag.h @@ -108,6 +108,8 @@ typedef struct xfs_agf { extern int xfs_read_agf(struct xfs_mount *mp, struct xfs_trans *tp, xfs_agnumber_t agno, int flags, struct xfs_buf **bpp); +extern const struct xfs_buf_ops xfs_agf_buf_ops; + /* * Size of the unlinked inode hash table in the agi. */ @@ -161,6 +163,8 @@ typedef struct xfs_agi { extern int xfs_read_agi(struct xfs_mount *mp, struct xfs_trans *tp, xfs_agnumber_t agno, struct xfs_buf **bpp); +extern const struct xfs_buf_ops xfs_agi_buf_ops; + /* * The third a.g. block contains the a.g. freelist, an array * of block pointers to blocks owned by the allocation btree code. @@ -233,6 +237,7 @@ typedef struct xfs_perag { #define XFS_ICI_NO_TAG (-1) /* special flag for an untagged lookup in xfs_inode_ag_iterator */ #define XFS_ICI_RECLAIM_TAG 0 /* inode is to be reclaimed */ +#define XFS_ICI_EOFBLOCKS_TAG 1 /* inode has blocks beyond EOF */ #define XFS_AG_MAXLEVELS(mp) ((mp)->m_ag_maxlevels) #define XFS_MIN_FREELIST_RAW(bl,cl,mp) \ diff --git a/fs/xfs/xfs_alloc.c b/fs/xfs/xfs_alloc.c index 335206a9c698..393055fe3aef 100644 --- a/fs/xfs/xfs_alloc.c +++ b/fs/xfs/xfs_alloc.c @@ -430,6 +430,60 @@ xfs_alloc_fixup_trees( return 0; } +static void +xfs_agfl_verify( + struct xfs_buf *bp) +{ +#ifdef WHEN_CRCS_COME_ALONG + /* + * we cannot actually do any verification of the AGFL because mkfs does + * not initialise the AGFL to zero or NULL. Hence the only valid part of + * the AGFL is what the AGF says is active. We can't get to the AGF, so + * we can't verify just those entries are valid. + * + * This problem goes away when the CRC format change comes along as that + * requires the AGFL to be initialised by mkfs. At that point, we can + * verify the blocks in the agfl -active or not- lie within the bounds + * of the AG. Until then, just leave this check ifdef'd out. + */ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_agfl *agfl = XFS_BUF_TO_AGFL(bp); + int agfl_ok = 1; + + int i; + + for (i = 0; i < XFS_AGFL_SIZE(mp); i++) { + if (be32_to_cpu(agfl->agfl_bno[i]) == NULLAGBLOCK || + be32_to_cpu(agfl->agfl_bno[i]) >= mp->m_sb.sb_agblocks) + agfl_ok = 0; + } + + if (!agfl_ok) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, agfl); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +#endif +} + +static void +xfs_agfl_write_verify( + struct xfs_buf *bp) +{ + xfs_agfl_verify(bp); +} + +static void +xfs_agfl_read_verify( + struct xfs_buf *bp) +{ + xfs_agfl_verify(bp); +} + +const struct xfs_buf_ops xfs_agfl_buf_ops = { + .verify_read = xfs_agfl_read_verify, + .verify_write = xfs_agfl_write_verify, +}; + /* * Read in the allocation group free block array. */ @@ -447,7 +501,7 @@ xfs_alloc_read_agfl( error = xfs_trans_read_buf( mp, tp, mp->m_ddev_targp, XFS_AG_DADDR(mp, agno, XFS_AGFL_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0, &bp); + XFS_FSS_TO_BB(mp, 1), 0, &bp, &xfs_agfl_buf_ops); if (error) return error; ASSERT(!xfs_buf_geterror(bp)); @@ -2091,6 +2145,63 @@ xfs_alloc_put_freelist( return 0; } +static void +xfs_agf_verify( + struct xfs_buf *bp) + { + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_agf *agf; + int agf_ok; + + agf = XFS_BUF_TO_AGF(bp); + + agf_ok = agf->agf_magicnum == cpu_to_be32(XFS_AGF_MAGIC) && + XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) && + be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) && + be32_to_cpu(agf->agf_flfirst) < XFS_AGFL_SIZE(mp) && + be32_to_cpu(agf->agf_fllast) < XFS_AGFL_SIZE(mp) && + be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp); + + /* + * during growfs operations, the perag is not fully initialised, + * so we can't use it for any useful checking. growfs ensures we can't + * use it by using uncached buffers that don't have the perag attached + * so we can detect and avoid this problem. + */ + if (bp->b_pag) + agf_ok = agf_ok && be32_to_cpu(agf->agf_seqno) == + bp->b_pag->pag_agno; + + if (xfs_sb_version_haslazysbcount(&mp->m_sb)) + agf_ok = agf_ok && be32_to_cpu(agf->agf_btreeblks) <= + be32_to_cpu(agf->agf_length); + + if (unlikely(XFS_TEST_ERROR(!agf_ok, mp, XFS_ERRTAG_ALLOC_READ_AGF, + XFS_RANDOM_ALLOC_READ_AGF))) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, agf); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_agf_read_verify( + struct xfs_buf *bp) +{ + xfs_agf_verify(bp); +} + +static void +xfs_agf_write_verify( + struct xfs_buf *bp) +{ + xfs_agf_verify(bp); +} + +const struct xfs_buf_ops xfs_agf_buf_ops = { + .verify_read = xfs_agf_read_verify, + .verify_write = xfs_agf_write_verify, +}; + /* * Read in the allocation group header (free/alloc section). */ @@ -2102,44 +2213,19 @@ xfs_read_agf( int flags, /* XFS_BUF_ */ struct xfs_buf **bpp) /* buffer for the ag freelist header */ { - struct xfs_agf *agf; /* ag freelist header */ - int agf_ok; /* set if agf is consistent */ int error; ASSERT(agno != NULLAGNUMBER); error = xfs_trans_read_buf( mp, tp, mp->m_ddev_targp, XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), flags, bpp); + XFS_FSS_TO_BB(mp, 1), flags, bpp, &xfs_agf_buf_ops); if (error) return error; if (!*bpp) return 0; ASSERT(!(*bpp)->b_error); - agf = XFS_BUF_TO_AGF(*bpp); - - /* - * Validate the magic number of the agf block. - */ - agf_ok = - agf->agf_magicnum == cpu_to_be32(XFS_AGF_MAGIC) && - XFS_AGF_GOOD_VERSION(be32_to_cpu(agf->agf_versionnum)) && - be32_to_cpu(agf->agf_freeblks) <= be32_to_cpu(agf->agf_length) && - be32_to_cpu(agf->agf_flfirst) < XFS_AGFL_SIZE(mp) && - be32_to_cpu(agf->agf_fllast) < XFS_AGFL_SIZE(mp) && - be32_to_cpu(agf->agf_flcount) <= XFS_AGFL_SIZE(mp) && - be32_to_cpu(agf->agf_seqno) == agno; - if (xfs_sb_version_haslazysbcount(&mp->m_sb)) - agf_ok = agf_ok && be32_to_cpu(agf->agf_btreeblks) <= - be32_to_cpu(agf->agf_length); - if (unlikely(XFS_TEST_ERROR(!agf_ok, mp, XFS_ERRTAG_ALLOC_READ_AGF, - XFS_RANDOM_ALLOC_READ_AGF))) { - XFS_CORRUPTION_ERROR("xfs_alloc_read_agf", - XFS_ERRLEVEL_LOW, mp, agf); - xfs_trans_brelse(tp, *bpp); - return XFS_ERROR(EFSCORRUPTED); - } xfs_buf_set_ref(*bpp, XFS_AGF_REF); return 0; } diff --git a/fs/xfs/xfs_alloc.h b/fs/xfs/xfs_alloc.h index feacb061bab7..99d0a6101558 100644 --- a/fs/xfs/xfs_alloc.h +++ b/fs/xfs/xfs_alloc.h @@ -231,4 +231,7 @@ xfs_alloc_get_rec( xfs_extlen_t *len, /* output: length of extent */ int *stat); /* output: success/failure */ +extern const struct xfs_buf_ops xfs_agf_buf_ops; +extern const struct xfs_buf_ops xfs_agfl_buf_ops; + #endif /* __XFS_ALLOC_H__ */ diff --git a/fs/xfs/xfs_alloc_btree.c b/fs/xfs/xfs_alloc_btree.c index f7876c6d6165..b1ddef6b2689 100644 --- a/fs/xfs/xfs_alloc_btree.c +++ b/fs/xfs/xfs_alloc_btree.c @@ -272,6 +272,82 @@ xfs_allocbt_key_diff( return (__int64_t)be32_to_cpu(kp->ar_startblock) - rec->ar_startblock; } +static void +xfs_allocbt_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp); + struct xfs_perag *pag = bp->b_pag; + unsigned int level; + int sblock_ok; /* block passes checks */ + + /* + * magic number and level verification + * + * During growfs operations, we can't verify the exact level as the + * perag is not fully initialised and hence not attached to the buffer. + * In this case, check against the maximum tree depth. + */ + level = be16_to_cpu(block->bb_level); + switch (block->bb_magic) { + case cpu_to_be32(XFS_ABTB_MAGIC): + if (pag) + sblock_ok = level < pag->pagf_levels[XFS_BTNUM_BNOi]; + else + sblock_ok = level < mp->m_ag_maxlevels; + break; + case cpu_to_be32(XFS_ABTC_MAGIC): + if (pag) + sblock_ok = level < pag->pagf_levels[XFS_BTNUM_CNTi]; + else + sblock_ok = level < mp->m_ag_maxlevels; + break; + default: + sblock_ok = 0; + break; + } + + /* numrecs verification */ + sblock_ok = sblock_ok && + be16_to_cpu(block->bb_numrecs) <= mp->m_alloc_mxr[level != 0]; + + /* sibling pointer verification */ + sblock_ok = sblock_ok && + (block->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK) || + be32_to_cpu(block->bb_u.s.bb_leftsib) < mp->m_sb.sb_agblocks) && + block->bb_u.s.bb_leftsib && + (block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) || + be32_to_cpu(block->bb_u.s.bb_rightsib) < mp->m_sb.sb_agblocks) && + block->bb_u.s.bb_rightsib; + + if (!sblock_ok) { + trace_xfs_btree_corrupt(bp, _RET_IP_); + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_allocbt_read_verify( + struct xfs_buf *bp) +{ + xfs_allocbt_verify(bp); +} + +static void +xfs_allocbt_write_verify( + struct xfs_buf *bp) +{ + xfs_allocbt_verify(bp); +} + +const struct xfs_buf_ops xfs_allocbt_buf_ops = { + .verify_read = xfs_allocbt_read_verify, + .verify_write = xfs_allocbt_write_verify, +}; + + #ifdef DEBUG STATIC int xfs_allocbt_keys_inorder( @@ -327,6 +403,7 @@ static const struct xfs_btree_ops xfs_allocbt_ops = { .init_rec_from_cur = xfs_allocbt_init_rec_from_cur, .init_ptr_from_cur = xfs_allocbt_init_ptr_from_cur, .key_diff = xfs_allocbt_key_diff, + .buf_ops = &xfs_allocbt_buf_ops, #ifdef DEBUG .keys_inorder = xfs_allocbt_keys_inorder, .recs_inorder = xfs_allocbt_recs_inorder, diff --git a/fs/xfs/xfs_alloc_btree.h b/fs/xfs/xfs_alloc_btree.h index 359fb86ed876..7e89a2b429dd 100644 --- a/fs/xfs/xfs_alloc_btree.h +++ b/fs/xfs/xfs_alloc_btree.h @@ -93,4 +93,6 @@ extern struct xfs_btree_cur *xfs_allocbt_init_cursor(struct xfs_mount *, xfs_agnumber_t, xfs_btnum_t); extern int xfs_allocbt_maxrecs(struct xfs_mount *, int, int); +extern const struct xfs_buf_ops xfs_allocbt_buf_ops; + #endif /* __XFS_ALLOC_BTREE_H__ */ diff --git a/fs/xfs/xfs_aops.c b/fs/xfs/xfs_aops.c index e57e2daa357c..4111a40ebe1a 100644 --- a/fs/xfs/xfs_aops.c +++ b/fs/xfs/xfs_aops.c @@ -124,7 +124,7 @@ xfs_setfilesize_trans_alloc( ioend->io_append_trans = tp; /* - * We will pass freeze protection with a transaction. So tell lockdep + * We may pass freeze protection with a transaction. So tell lockdep * we released it. */ rwsem_release(&ioend->io_inode->i_sb->s_writers.lock_map[SB_FREEZE_FS-1], @@ -149,11 +149,13 @@ xfs_setfilesize( xfs_fsize_t isize; /* - * The transaction was allocated in the I/O submission thread, - * thus we need to mark ourselves as beeing in a transaction - * manually. + * The transaction may have been allocated in the I/O submission thread, + * thus we need to mark ourselves as beeing in a transaction manually. + * Similarly for freeze protection. */ current_set_flags_nested(&tp->t_pflags, PF_FSTRANS); + rwsem_acquire_read(&VFS_I(ip)->i_sb->s_writers.lock_map[SB_FREEZE_FS-1], + 0, 1, _THIS_IP_); xfs_ilock(ip, XFS_ILOCK_EXCL); isize = xfs_new_eof(ip, ioend->io_offset + ioend->io_size); @@ -187,7 +189,8 @@ xfs_finish_ioend( if (ioend->io_type == XFS_IO_UNWRITTEN) queue_work(mp->m_unwritten_workqueue, &ioend->io_work); - else if (ioend->io_append_trans) + else if (ioend->io_append_trans || + (ioend->io_isdirect && xfs_ioend_is_append(ioend))) queue_work(mp->m_data_workqueue, &ioend->io_work); else xfs_destroy_ioend(ioend); @@ -205,15 +208,6 @@ xfs_end_io( struct xfs_inode *ip = XFS_I(ioend->io_inode); int error = 0; - if (ioend->io_append_trans) { - /* - * We've got freeze protection passed with the transaction. - * Tell lockdep about it. - */ - rwsem_acquire_read( - &ioend->io_inode->i_sb->s_writers.lock_map[SB_FREEZE_FS-1], - 0, 1, _THIS_IP_); - } if (XFS_FORCED_SHUTDOWN(ip->i_mount)) { ioend->io_error = -EIO; goto done; @@ -226,35 +220,31 @@ xfs_end_io( * range to normal written extens after the data I/O has finished. */ if (ioend->io_type == XFS_IO_UNWRITTEN) { + error = xfs_iomap_write_unwritten(ip, ioend->io_offset, + ioend->io_size); + } else if (ioend->io_isdirect && xfs_ioend_is_append(ioend)) { /* - * For buffered I/O we never preallocate a transaction when - * doing the unwritten extent conversion, but for direct I/O - * we do not know if we are converting an unwritten extent - * or not at the point where we preallocate the transaction. + * For direct I/O we do not know if we need to allocate blocks + * or not so we can't preallocate an append transaction as that + * results in nested reservations and log space deadlocks. Hence + * allocate the transaction here. While this is sub-optimal and + * can block IO completion for some time, we're stuck with doing + * it this way until we can pass the ioend to the direct IO + * allocation callbacks and avoid nesting that way. */ - if (ioend->io_append_trans) { - ASSERT(ioend->io_isdirect); - - current_set_flags_nested( - &ioend->io_append_trans->t_pflags, PF_FSTRANS); - xfs_trans_cancel(ioend->io_append_trans, 0); - } - - error = xfs_iomap_write_unwritten(ip, ioend->io_offset, - ioend->io_size); - if (error) { - ioend->io_error = -error; + error = xfs_setfilesize_trans_alloc(ioend); + if (error) goto done; - } + error = xfs_setfilesize(ioend); } else if (ioend->io_append_trans) { error = xfs_setfilesize(ioend); - if (error) - ioend->io_error = -error; } else { ASSERT(!xfs_ioend_is_append(ioend)); } done: + if (error) + ioend->io_error = -error; xfs_destroy_ioend(ioend); } @@ -1432,25 +1422,21 @@ xfs_vm_direct_IO( size_t size = iov_length(iov, nr_segs); /* - * We need to preallocate a transaction for a size update - * here. In the case that this write both updates the size - * and converts at least on unwritten extent we will cancel - * the still clean transaction after the I/O has finished. + * We cannot preallocate a size update transaction here as we + * don't know whether allocation is necessary or not. Hence we + * can only tell IO completion that one is necessary if we are + * not doing unwritten extent conversion. */ iocb->private = ioend = xfs_alloc_ioend(inode, XFS_IO_DIRECT); - if (offset + size > XFS_I(inode)->i_d.di_size) { - ret = xfs_setfilesize_trans_alloc(ioend); - if (ret) - goto out_destroy_ioend; + if (offset + size > XFS_I(inode)->i_d.di_size) ioend->io_isdirect = 1; - } ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov, offset, nr_segs, xfs_get_blocks_direct, xfs_end_io_direct_write, NULL, 0); if (ret != -EIOCBQUEUED && iocb->private) - goto out_trans_cancel; + goto out_destroy_ioend; } else { ret = __blockdev_direct_IO(rw, iocb, inode, bdev, iov, offset, nr_segs, @@ -1460,15 +1446,6 @@ xfs_vm_direct_IO( return ret; -out_trans_cancel: - if (ioend->io_append_trans) { - current_set_flags_nested(&ioend->io_append_trans->t_pflags, - PF_FSTRANS); - rwsem_acquire_read( - &inode->i_sb->s_writers.lock_map[SB_FREEZE_FS-1], - 0, 1, _THIS_IP_); - xfs_trans_cancel(ioend->io_append_trans, 0); - } out_destroy_ioend: xfs_destroy_ioend(ioend); return ret; @@ -1641,7 +1618,7 @@ xfs_vm_bmap( trace_xfs_vm_bmap(XFS_I(inode)); xfs_ilock(ip, XFS_IOLOCK_SHARED); - xfs_flush_pages(ip, (xfs_off_t)0, -1, 0, FI_REMAPF); + filemap_write_and_wait(mapping); xfs_iunlock(ip, XFS_IOLOCK_SHARED); return generic_block_bmap(mapping, block, xfs_get_blocks); } diff --git a/fs/xfs/xfs_attr.c b/fs/xfs/xfs_attr.c index 0ca1f0be62d2..aaf472532b3c 100644 --- a/fs/xfs/xfs_attr.c +++ b/fs/xfs/xfs_attr.c @@ -903,11 +903,9 @@ xfs_attr_leaf_addname(xfs_da_args_t *args) */ dp = args->dp; args->blkno = 0; - error = xfs_da_read_buf(args->trans, args->dp, args->blkno, -1, &bp, - XFS_ATTR_FORK); + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp); if (error) - return(error); - ASSERT(bp != NULL); + return error; /* * Look up the given attribute in the leaf block. Figure out if @@ -1031,12 +1029,12 @@ xfs_attr_leaf_addname(xfs_da_args_t *args) * Read in the block containing the "old" attr, then * remove the "old" attr from that block (neat, huh!) */ - error = xfs_da_read_buf(args->trans, args->dp, args->blkno, -1, - &bp, XFS_ATTR_FORK); + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, + -1, &bp); if (error) - return(error); - ASSERT(bp != NULL); - (void)xfs_attr_leaf_remove(bp, args); + return error; + + xfs_attr_leaf_remove(bp, args); /* * If the result is small enough, shrink it all into the inode. @@ -1100,20 +1098,17 @@ xfs_attr_leaf_removename(xfs_da_args_t *args) */ dp = args->dp; args->blkno = 0; - error = xfs_da_read_buf(args->trans, args->dp, args->blkno, -1, &bp, - XFS_ATTR_FORK); - if (error) { - return(error); - } + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp); + if (error) + return error; - ASSERT(bp != NULL); error = xfs_attr_leaf_lookup_int(bp, args); if (error == ENOATTR) { xfs_trans_brelse(args->trans, bp); return(error); } - (void)xfs_attr_leaf_remove(bp, args); + xfs_attr_leaf_remove(bp, args); /* * If the result is small enough, shrink it all into the inode. @@ -1155,12 +1150,12 @@ xfs_attr_leaf_get(xfs_da_args_t *args) struct xfs_buf *bp; int error; + trace_xfs_attr_leaf_get(args); + args->blkno = 0; - error = xfs_da_read_buf(args->trans, args->dp, args->blkno, -1, &bp, - XFS_ATTR_FORK); + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp); if (error) - return(error); - ASSERT(bp != NULL); + return error; error = xfs_attr_leaf_lookup_int(bp, args); if (error != EEXIST) { @@ -1181,22 +1176,15 @@ xfs_attr_leaf_get(xfs_da_args_t *args) STATIC int xfs_attr_leaf_list(xfs_attr_list_context_t *context) { - xfs_attr_leafblock_t *leaf; int error; struct xfs_buf *bp; + trace_xfs_attr_leaf_list(context); + context->cursor->blkno = 0; - error = xfs_da_read_buf(NULL, context->dp, 0, -1, &bp, XFS_ATTR_FORK); + error = xfs_attr_leaf_read(NULL, context->dp, 0, -1, &bp); if (error) return XFS_ERROR(error); - ASSERT(bp != NULL); - leaf = bp->b_addr; - if (unlikely(leaf->hdr.info.magic != cpu_to_be16(XFS_ATTR_LEAF_MAGIC))) { - XFS_CORRUPTION_ERROR("xfs_attr_leaf_list", XFS_ERRLEVEL_LOW, - context->dp->i_mount, leaf); - xfs_trans_brelse(NULL, bp); - return XFS_ERROR(EFSCORRUPTED); - } error = xfs_attr_leaf_list_int(bp, context); xfs_trans_brelse(NULL, bp); @@ -1600,12 +1588,9 @@ xfs_attr_node_removename(xfs_da_args_t *args) ASSERT(state->path.blk[0].bp); state->path.blk[0].bp = NULL; - error = xfs_da_read_buf(args->trans, args->dp, 0, -1, &bp, - XFS_ATTR_FORK); + error = xfs_attr_leaf_read(args->trans, args->dp, 0, -1, &bp); if (error) goto out; - ASSERT((((xfs_attr_leafblock_t *)bp->b_addr)->hdr.info.magic) == - cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); if ((forkoff = xfs_attr_shortform_allfit(bp, dp))) { xfs_bmap_init(args->flist, args->firstblock); @@ -1653,6 +1638,8 @@ xfs_attr_fillstate(xfs_da_state_t *state) xfs_da_state_blk_t *blk; int level; + trace_xfs_attr_fillstate(state->args); + /* * Roll down the "path" in the state structure, storing the on-disk * block number for those buffers in the "path". @@ -1699,6 +1686,8 @@ xfs_attr_refillstate(xfs_da_state_t *state) xfs_da_state_blk_t *blk; int level, error; + trace_xfs_attr_refillstate(state->args); + /* * Roll down the "path" in the state structure, storing the on-disk * block number for those buffers in the "path". @@ -1707,7 +1696,7 @@ xfs_attr_refillstate(xfs_da_state_t *state) ASSERT((path->active >= 0) && (path->active < XFS_DA_NODE_MAXDEPTH)); for (blk = path->blk, level = 0; level < path->active; blk++, level++) { if (blk->disk_blkno) { - error = xfs_da_read_buf(state->args->trans, + error = xfs_da_node_read(state->args->trans, state->args->dp, blk->blkno, blk->disk_blkno, &blk->bp, XFS_ATTR_FORK); @@ -1726,7 +1715,7 @@ xfs_attr_refillstate(xfs_da_state_t *state) ASSERT((path->active >= 0) && (path->active < XFS_DA_NODE_MAXDEPTH)); for (blk = path->blk, level = 0; level < path->active; blk++, level++) { if (blk->disk_blkno) { - error = xfs_da_read_buf(state->args->trans, + error = xfs_da_node_read(state->args->trans, state->args->dp, blk->blkno, blk->disk_blkno, &blk->bp, XFS_ATTR_FORK); @@ -1755,6 +1744,8 @@ xfs_attr_node_get(xfs_da_args_t *args) int error, retval; int i; + trace_xfs_attr_node_get(args); + state = xfs_da_state_alloc(); state->args = args; state->mp = args->dp->i_mount; @@ -1804,6 +1795,8 @@ xfs_attr_node_list(xfs_attr_list_context_t *context) int error, i; struct xfs_buf *bp; + trace_xfs_attr_node_list(context); + cursor = context->cursor; cursor->initted = 1; @@ -1814,7 +1807,7 @@ xfs_attr_node_list(xfs_attr_list_context_t *context) */ bp = NULL; if (cursor->blkno > 0) { - error = xfs_da_read_buf(NULL, context->dp, cursor->blkno, -1, + error = xfs_da_node_read(NULL, context->dp, cursor->blkno, -1, &bp, XFS_ATTR_FORK); if ((error != 0) && (error != EFSCORRUPTED)) return(error); @@ -1856,17 +1849,11 @@ xfs_attr_node_list(xfs_attr_list_context_t *context) if (bp == NULL) { cursor->blkno = 0; for (;;) { - error = xfs_da_read_buf(NULL, context->dp, + error = xfs_da_node_read(NULL, context->dp, cursor->blkno, -1, &bp, XFS_ATTR_FORK); if (error) return(error); - if (unlikely(bp == NULL)) { - XFS_ERROR_REPORT("xfs_attr_node_list(2)", - XFS_ERRLEVEL_LOW, - context->dp->i_mount); - return(XFS_ERROR(EFSCORRUPTED)); - } node = bp->b_addr; if (node->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)) @@ -1907,14 +1894,6 @@ xfs_attr_node_list(xfs_attr_list_context_t *context) */ for (;;) { leaf = bp->b_addr; - if (unlikely(leaf->hdr.info.magic != - cpu_to_be16(XFS_ATTR_LEAF_MAGIC))) { - XFS_CORRUPTION_ERROR("xfs_attr_node_list(4)", - XFS_ERRLEVEL_LOW, - context->dp->i_mount, leaf); - xfs_trans_brelse(NULL, bp); - return(XFS_ERROR(EFSCORRUPTED)); - } error = xfs_attr_leaf_list_int(bp, context); if (error) { xfs_trans_brelse(NULL, bp); @@ -1924,16 +1903,10 @@ xfs_attr_node_list(xfs_attr_list_context_t *context) break; cursor->blkno = be32_to_cpu(leaf->hdr.info.forw); xfs_trans_brelse(NULL, bp); - error = xfs_da_read_buf(NULL, context->dp, cursor->blkno, -1, - &bp, XFS_ATTR_FORK); + error = xfs_attr_leaf_read(NULL, context->dp, cursor->blkno, -1, + &bp); if (error) - return(error); - if (unlikely((bp == NULL))) { - XFS_ERROR_REPORT("xfs_attr_node_list(5)", - XFS_ERRLEVEL_LOW, - context->dp->i_mount); - return(XFS_ERROR(EFSCORRUPTED)); - } + return error; } xfs_trans_brelse(NULL, bp); return(0); @@ -1959,6 +1932,8 @@ xfs_attr_rmtval_get(xfs_da_args_t *args) int nmap, error, tmp, valuelen, blkcnt, i; xfs_dablk_t lblkno; + trace_xfs_attr_rmtval_get(args); + ASSERT(!(args->flags & ATTR_KERNOVAL)); mp = args->dp->i_mount; @@ -1980,7 +1955,7 @@ xfs_attr_rmtval_get(xfs_da_args_t *args) dblkno = XFS_FSB_TO_DADDR(mp, map[i].br_startblock); blkcnt = XFS_FSB_TO_BB(mp, map[i].br_blockcount); error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, - dblkno, blkcnt, 0, &bp); + dblkno, blkcnt, 0, &bp, NULL); if (error) return(error); @@ -2014,6 +1989,8 @@ xfs_attr_rmtval_set(xfs_da_args_t *args) xfs_dablk_t lblkno; int blkcnt, valuelen, nmap, error, tmp, committed; + trace_xfs_attr_rmtval_set(args); + dp = args->dp; mp = dp->i_mount; src = args->value; @@ -2143,6 +2120,8 @@ xfs_attr_rmtval_remove(xfs_da_args_t *args) xfs_dablk_t lblkno; int valuelen, blkcnt, nmap, error, done, committed; + trace_xfs_attr_rmtval_remove(args); + mp = args->dp->i_mount; /* diff --git a/fs/xfs/xfs_attr_leaf.c b/fs/xfs/xfs_attr_leaf.c index 70eec1829776..ee24993c7d12 100644 --- a/fs/xfs/xfs_attr_leaf.c +++ b/fs/xfs/xfs_attr_leaf.c @@ -57,7 +57,8 @@ STATIC int xfs_attr_leaf_create(xfs_da_args_t *args, xfs_dablk_t which_block, struct xfs_buf **bpp); STATIC int xfs_attr_leaf_add_work(struct xfs_buf *leaf_buffer, xfs_da_args_t *args, int freemap_index); -STATIC void xfs_attr_leaf_compact(xfs_trans_t *tp, struct xfs_buf *leaf_buffer); +STATIC void xfs_attr_leaf_compact(struct xfs_da_args *args, + struct xfs_buf *leaf_buffer); STATIC void xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1, xfs_da_state_blk_t *blk2); @@ -87,6 +88,52 @@ STATIC void xfs_attr_leaf_moveents(xfs_attr_leafblock_t *src_leaf, xfs_mount_t *mp); STATIC int xfs_attr_leaf_entsize(xfs_attr_leafblock_t *leaf, int index); +static void +xfs_attr_leaf_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_attr_leaf_hdr *hdr = bp->b_addr; + int block_ok = 0; + + block_ok = hdr->info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC); + if (!block_ok) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_attr_leaf_read_verify( + struct xfs_buf *bp) +{ + xfs_attr_leaf_verify(bp); +} + +static void +xfs_attr_leaf_write_verify( + struct xfs_buf *bp) +{ + xfs_attr_leaf_verify(bp); +} + +const struct xfs_buf_ops xfs_attr_leaf_buf_ops = { + .verify_read = xfs_attr_leaf_read_verify, + .verify_write = xfs_attr_leaf_write_verify, +}; + +int +xfs_attr_leaf_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t bno, + xfs_daddr_t mappedbno, + struct xfs_buf **bpp) +{ + return xfs_da_read_buf(tp, dp, bno, mappedbno, bpp, + XFS_ATTR_FORK, &xfs_attr_leaf_buf_ops); +} + /*======================================================================== * Namespace helper routines *========================================================================*/ @@ -869,17 +916,16 @@ xfs_attr_leaf_to_node(xfs_da_args_t *args) error = xfs_da_grow_inode(args, &blkno); if (error) goto out; - error = xfs_da_read_buf(args->trans, args->dp, 0, -1, &bp1, - XFS_ATTR_FORK); + error = xfs_attr_leaf_read(args->trans, args->dp, 0, -1, &bp1); if (error) goto out; - ASSERT(bp1 != NULL); + bp2 = NULL; error = xfs_da_get_buf(args->trans, args->dp, blkno, -1, &bp2, XFS_ATTR_FORK); if (error) goto out; - ASSERT(bp2 != NULL); + bp2->b_ops = bp1->b_ops; memcpy(bp2->b_addr, bp1->b_addr, XFS_LBSIZE(dp->i_mount)); bp1 = NULL; xfs_trans_log_buf(args->trans, bp2, 0, XFS_LBSIZE(dp->i_mount) - 1); @@ -933,7 +979,7 @@ xfs_attr_leaf_create( XFS_ATTR_FORK); if (error) return(error); - ASSERT(bp != NULL); + bp->b_ops = &xfs_attr_leaf_buf_ops; leaf = bp->b_addr; memset((char *)leaf, 0, XFS_LBSIZE(dp->i_mount)); hdr = &leaf->hdr; @@ -1071,7 +1117,7 @@ xfs_attr_leaf_add( * Compact the entries to coalesce free space. * This may change the hdr->count via dropping INCOMPLETE entries. */ - xfs_attr_leaf_compact(args->trans, bp); + xfs_attr_leaf_compact(args, bp); /* * After compaction, the block is guaranteed to have only one @@ -1102,6 +1148,8 @@ xfs_attr_leaf_add_work( xfs_mount_t *mp; int tmp, i; + trace_xfs_attr_leaf_add_work(args); + leaf = bp->b_addr; ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); hdr = &leaf->hdr; @@ -1214,15 +1262,17 @@ xfs_attr_leaf_add_work( */ STATIC void xfs_attr_leaf_compact( - struct xfs_trans *trans, - struct xfs_buf *bp) + struct xfs_da_args *args, + struct xfs_buf *bp) { - xfs_attr_leafblock_t *leaf_s, *leaf_d; - xfs_attr_leaf_hdr_t *hdr_s, *hdr_d; - xfs_mount_t *mp; - char *tmpbuffer; + xfs_attr_leafblock_t *leaf_s, *leaf_d; + xfs_attr_leaf_hdr_t *hdr_s, *hdr_d; + struct xfs_trans *trans = args->trans; + struct xfs_mount *mp = trans->t_mountp; + char *tmpbuffer; + + trace_xfs_attr_leaf_compact(args); - mp = trans->t_mountp; tmpbuffer = kmem_alloc(XFS_LBSIZE(mp), KM_SLEEP); ASSERT(tmpbuffer != NULL); memcpy(tmpbuffer, bp->b_addr, XFS_LBSIZE(mp)); @@ -1345,9 +1395,8 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1, max = be16_to_cpu(hdr2->firstused) - sizeof(xfs_attr_leaf_hdr_t); max -= be16_to_cpu(hdr2->count) * sizeof(xfs_attr_leaf_entry_t); - if (space > max) { - xfs_attr_leaf_compact(args->trans, blk2->bp); - } + if (space > max) + xfs_attr_leaf_compact(args, blk2->bp); /* * Move high entries from leaf1 to low end of leaf2. @@ -1378,9 +1427,8 @@ xfs_attr_leaf_rebalance(xfs_da_state_t *state, xfs_da_state_blk_t *blk1, max = be16_to_cpu(hdr1->firstused) - sizeof(xfs_attr_leaf_hdr_t); max -= be16_to_cpu(hdr1->count) * sizeof(xfs_attr_leaf_entry_t); - if (space > max) { - xfs_attr_leaf_compact(args->trans, blk1->bp); - } + if (space > max) + xfs_attr_leaf_compact(args, blk1->bp); /* * Move low entries from leaf2 to high end of leaf1. @@ -1577,6 +1625,8 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action) xfs_dablk_t blkno; struct xfs_buf *bp; + trace_xfs_attr_leaf_toosmall(state->args); + /* * Check for the degenerate case of the block being over 50% full. * If so, it's not worth even looking to see if we might be able @@ -1636,18 +1686,16 @@ xfs_attr_leaf_toosmall(xfs_da_state_t *state, int *action) blkno = be32_to_cpu(info->back); if (blkno == 0) continue; - error = xfs_da_read_buf(state->args->trans, state->args->dp, - blkno, -1, &bp, XFS_ATTR_FORK); + error = xfs_attr_leaf_read(state->args->trans, state->args->dp, + blkno, -1, &bp); if (error) return(error); - ASSERT(bp != NULL); leaf = (xfs_attr_leafblock_t *)info; count = be16_to_cpu(leaf->hdr.count); bytes = state->blocksize - (state->blocksize>>2); bytes -= be16_to_cpu(leaf->hdr.usedbytes); leaf = bp->b_addr; - ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); count += be16_to_cpu(leaf->hdr.count); bytes -= be16_to_cpu(leaf->hdr.usedbytes); bytes -= count * sizeof(xfs_attr_leaf_entry_t); @@ -1702,6 +1750,8 @@ xfs_attr_leaf_remove( int tablesize, tmp, i; xfs_mount_t *mp; + trace_xfs_attr_leaf_remove(args); + leaf = bp->b_addr; ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); hdr = &leaf->hdr; @@ -2511,15 +2561,11 @@ xfs_attr_leaf_clearflag(xfs_da_args_t *args) /* * Set up the operation. */ - error = xfs_da_read_buf(args->trans, args->dp, args->blkno, -1, &bp, - XFS_ATTR_FORK); - if (error) { + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp); + if (error) return(error); - } - ASSERT(bp != NULL); leaf = bp->b_addr; - ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); ASSERT(args->index < be16_to_cpu(leaf->hdr.count)); ASSERT(args->index >= 0); entry = &leaf->entries[ args->index ]; @@ -2576,15 +2622,11 @@ xfs_attr_leaf_setflag(xfs_da_args_t *args) /* * Set up the operation. */ - error = xfs_da_read_buf(args->trans, args->dp, args->blkno, -1, &bp, - XFS_ATTR_FORK); - if (error) { + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp); + if (error) return(error); - } - ASSERT(bp != NULL); leaf = bp->b_addr; - ASSERT(leaf->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); ASSERT(args->index < be16_to_cpu(leaf->hdr.count)); ASSERT(args->index >= 0); entry = &leaf->entries[ args->index ]; @@ -2633,35 +2675,28 @@ xfs_attr_leaf_flipflags(xfs_da_args_t *args) /* * Read the block containing the "old" attr */ - error = xfs_da_read_buf(args->trans, args->dp, args->blkno, -1, &bp1, - XFS_ATTR_FORK); - if (error) { - return(error); - } - ASSERT(bp1 != NULL); + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno, -1, &bp1); + if (error) + return error; /* * Read the block containing the "new" attr, if it is different */ if (args->blkno2 != args->blkno) { - error = xfs_da_read_buf(args->trans, args->dp, args->blkno2, - -1, &bp2, XFS_ATTR_FORK); - if (error) { - return(error); - } - ASSERT(bp2 != NULL); + error = xfs_attr_leaf_read(args->trans, args->dp, args->blkno2, + -1, &bp2); + if (error) + return error; } else { bp2 = bp1; } leaf1 = bp1->b_addr; - ASSERT(leaf1->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); ASSERT(args->index < be16_to_cpu(leaf1->hdr.count)); ASSERT(args->index >= 0); entry1 = &leaf1->entries[ args->index ]; leaf2 = bp2->b_addr; - ASSERT(leaf2->hdr.info.magic == cpu_to_be16(XFS_ATTR_LEAF_MAGIC)); ASSERT(args->index2 < be16_to_cpu(leaf2->hdr.count)); ASSERT(args->index2 >= 0); entry2 = &leaf2->entries[ args->index2 ]; @@ -2746,7 +2781,7 @@ xfs_attr_root_inactive(xfs_trans_t **trans, xfs_inode_t *dp) * the extents in reverse order the extent containing * block 0 must still be there. */ - error = xfs_da_read_buf(*trans, dp, 0, -1, &bp, XFS_ATTR_FORK); + error = xfs_da_node_read(*trans, dp, 0, -1, &bp, XFS_ATTR_FORK); if (error) return(error); blkno = XFS_BUF_ADDR(bp); @@ -2831,7 +2866,7 @@ xfs_attr_node_inactive( * traversal of the tree so we may deal with many blocks * before we come back to this one. */ - error = xfs_da_read_buf(*trans, dp, child_fsb, -2, &child_bp, + error = xfs_da_node_read(*trans, dp, child_fsb, -2, &child_bp, XFS_ATTR_FORK); if (error) return(error); @@ -2872,8 +2907,8 @@ xfs_attr_node_inactive( * child block number. */ if ((i+1) < count) { - error = xfs_da_read_buf(*trans, dp, 0, parent_blkno, - &bp, XFS_ATTR_FORK); + error = xfs_da_node_read(*trans, dp, 0, parent_blkno, + &bp, XFS_ATTR_FORK); if (error) return(error); child_fsb = be32_to_cpu(node->btree[i+1].before); diff --git a/fs/xfs/xfs_attr_leaf.h b/fs/xfs/xfs_attr_leaf.h index dea17722945e..77de139a58f0 100644 --- a/fs/xfs/xfs_attr_leaf.h +++ b/fs/xfs/xfs_attr_leaf.h @@ -261,4 +261,10 @@ int xfs_attr_leaf_order(struct xfs_buf *leaf1_bp, struct xfs_buf *leaf2_bp); int xfs_attr_leaf_newentsize(int namelen, int valuelen, int blocksize, int *local); +int xfs_attr_leaf_read(struct xfs_trans *tp, struct xfs_inode *dp, + xfs_dablk_t bno, xfs_daddr_t mappedbno, + struct xfs_buf **bpp); + +extern const struct xfs_buf_ops xfs_attr_leaf_buf_ops; + #endif /* __XFS_ATTR_LEAF_H__ */ diff --git a/fs/xfs/xfs_bmap.c b/fs/xfs/xfs_bmap.c index 83d0cf3df930..0e92d12765d2 100644 --- a/fs/xfs/xfs_bmap.c +++ b/fs/xfs/xfs_bmap.c @@ -2662,8 +2662,9 @@ xfs_bmap_btree_to_extents( if ((error = xfs_btree_check_lptr(cur, cbno, 1))) return error; #endif - if ((error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, - XFS_BMAP_BTREE_REF))) + error = xfs_btree_read_bufl(mp, tp, cbno, 0, &cbp, XFS_BMAP_BTREE_REF, + &xfs_bmbt_buf_ops); + if (error) return error; cblock = XFS_BUF_TO_BLOCK(cbp); if ((error = xfs_btree_check_block(cur, cblock, 0, cbp))) @@ -3123,6 +3124,7 @@ xfs_bmap_extents_to_btree( /* * Fill in the child block. */ + abp->b_ops = &xfs_bmbt_buf_ops; ablock = XFS_BUF_TO_BLOCK(abp); ablock->bb_magic = cpu_to_be32(XFS_BMAP_MAGIC); ablock->bb_level = 0; @@ -3269,6 +3271,7 @@ xfs_bmap_local_to_extents( ASSERT(args.len == 1); *firstblock = args.fsbno; bp = xfs_btree_get_bufl(args.mp, tp, args.fsbno, 0); + bp->b_ops = &xfs_bmbt_buf_ops; memcpy(bp->b_addr, ifp->if_u1.if_data, ifp->if_bytes); xfs_trans_log_buf(tp, bp, 0, ifp->if_bytes - 1); xfs_bmap_forkoff_reset(args.mp, ip, whichfork); @@ -4078,8 +4081,9 @@ xfs_bmap_read_extents( * pointer (leftmost) at each level. */ while (level-- > 0) { - if ((error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, - XFS_BMAP_BTREE_REF))) + error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, + XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops); + if (error) return error; block = XFS_BUF_TO_BLOCK(bp); XFS_WANT_CORRUPTED_GOTO( @@ -4124,7 +4128,8 @@ xfs_bmap_read_extents( */ nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib); if (nextbno != NULLFSBLOCK) - xfs_btree_reada_bufl(mp, nextbno, 1); + xfs_btree_reada_bufl(mp, nextbno, 1, + &xfs_bmbt_buf_ops); /* * Copy records into the extent records. */ @@ -4156,8 +4161,9 @@ xfs_bmap_read_extents( */ if (bno == NULLFSBLOCK) break; - if ((error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, - XFS_BMAP_BTREE_REF))) + error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, + XFS_BMAP_BTREE_REF, &xfs_bmbt_buf_ops); + if (error) return error; block = XFS_BUF_TO_BLOCK(bp); } @@ -5599,7 +5605,7 @@ xfs_getbmap( xfs_ilock(ip, XFS_IOLOCK_SHARED); if (whichfork == XFS_DATA_FORK && !(iflags & BMV_IF_DELALLOC)) { if (ip->i_delayed_blks || XFS_ISIZE(ip) > ip->i_d.di_size) { - error = xfs_flush_pages(ip, 0, -1, 0, FI_REMAPF); + error = -filemap_write_and_wait(VFS_I(ip)->i_mapping); if (error) goto out_unlock_iolock; } @@ -5868,15 +5874,16 @@ xfs_bmap_check_leaf_extents( */ while (level-- > 0) { /* See if buf is in cur first */ + bp_release = 0; bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno)); - if (bp) { - bp_release = 0; - } else { + if (!bp) { bp_release = 1; + error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp, + XFS_BMAP_BTREE_REF, + &xfs_bmbt_buf_ops); + if (error) + goto error_norelse; } - if (!bp && (error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp, - XFS_BMAP_BTREE_REF))) - goto error_norelse; block = XFS_BUF_TO_BLOCK(bp); XFS_WANT_CORRUPTED_GOTO( xfs_bmap_sanity_check(mp, bp, level), @@ -5953,15 +5960,16 @@ xfs_bmap_check_leaf_extents( if (bno == NULLFSBLOCK) break; + bp_release = 0; bp = xfs_bmap_get_bp(cur, XFS_FSB_TO_DADDR(mp, bno)); - if (bp) { - bp_release = 0; - } else { + if (!bp) { bp_release = 1; + error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp, + XFS_BMAP_BTREE_REF, + &xfs_bmbt_buf_ops); + if (error) + goto error_norelse; } - if (!bp && (error = xfs_btree_read_bufl(mp, NULL, bno, 0, &bp, - XFS_BMAP_BTREE_REF))) - goto error_norelse; block = XFS_BUF_TO_BLOCK(bp); } if (bp_release) { @@ -6052,7 +6060,9 @@ xfs_bmap_count_tree( struct xfs_btree_block *block, *nextblock; int numrecs; - if ((error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, XFS_BMAP_BTREE_REF))) + error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, XFS_BMAP_BTREE_REF, + &xfs_bmbt_buf_ops); + if (error) return error; *count += 1; block = XFS_BUF_TO_BLOCK(bp); @@ -6061,8 +6071,10 @@ xfs_bmap_count_tree( /* Not at node above leaves, count this level of nodes */ nextbno = be64_to_cpu(block->bb_u.l.bb_rightsib); while (nextbno != NULLFSBLOCK) { - if ((error = xfs_btree_read_bufl(mp, tp, nextbno, - 0, &nbp, XFS_BMAP_BTREE_REF))) + error = xfs_btree_read_bufl(mp, tp, nextbno, 0, &nbp, + XFS_BMAP_BTREE_REF, + &xfs_bmbt_buf_ops); + if (error) return error; *count += 1; nextblock = XFS_BUF_TO_BLOCK(nbp); @@ -6091,8 +6103,10 @@ xfs_bmap_count_tree( if (nextbno == NULLFSBLOCK) break; bno = nextbno; - if ((error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, - XFS_BMAP_BTREE_REF))) + error = xfs_btree_read_bufl(mp, tp, bno, 0, &bp, + XFS_BMAP_BTREE_REF, + &xfs_bmbt_buf_ops); + if (error) return error; *count += 1; block = XFS_BUF_TO_BLOCK(bp); diff --git a/fs/xfs/xfs_bmap_btree.c b/fs/xfs/xfs_bmap_btree.c index 862084a47a7e..061b45cbe614 100644 --- a/fs/xfs/xfs_bmap_btree.c +++ b/fs/xfs/xfs_bmap_btree.c @@ -36,6 +36,7 @@ #include "xfs_bmap.h" #include "xfs_error.h" #include "xfs_quota.h" +#include "xfs_trace.h" /* * Determine the extent state. @@ -707,6 +708,67 @@ xfs_bmbt_key_diff( cur->bc_rec.b.br_startoff; } +static void +xfs_bmbt_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp); + unsigned int level; + int lblock_ok; /* block passes checks */ + + /* magic number and level verification. + * + * We don't know waht fork we belong to, so just verify that the level + * is less than the maximum of the two. Later checks will be more + * precise. + */ + level = be16_to_cpu(block->bb_level); + lblock_ok = block->bb_magic == cpu_to_be32(XFS_BMAP_MAGIC) && + level < max(mp->m_bm_maxlevels[0], mp->m_bm_maxlevels[1]); + + /* numrecs verification */ + lblock_ok = lblock_ok && + be16_to_cpu(block->bb_numrecs) <= mp->m_bmap_dmxr[level != 0]; + + /* sibling pointer verification */ + lblock_ok = lblock_ok && + block->bb_u.l.bb_leftsib && + (block->bb_u.l.bb_leftsib == cpu_to_be64(NULLDFSBNO) || + XFS_FSB_SANITY_CHECK(mp, + be64_to_cpu(block->bb_u.l.bb_leftsib))) && + block->bb_u.l.bb_rightsib && + (block->bb_u.l.bb_rightsib == cpu_to_be64(NULLDFSBNO) || + XFS_FSB_SANITY_CHECK(mp, + be64_to_cpu(block->bb_u.l.bb_rightsib))); + + if (!lblock_ok) { + trace_xfs_btree_corrupt(bp, _RET_IP_); + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_bmbt_read_verify( + struct xfs_buf *bp) +{ + xfs_bmbt_verify(bp); +} + +static void +xfs_bmbt_write_verify( + struct xfs_buf *bp) +{ + xfs_bmbt_verify(bp); +} + +const struct xfs_buf_ops xfs_bmbt_buf_ops = { + .verify_read = xfs_bmbt_read_verify, + .verify_write = xfs_bmbt_write_verify, +}; + + #ifdef DEBUG STATIC int xfs_bmbt_keys_inorder( @@ -746,6 +808,7 @@ static const struct xfs_btree_ops xfs_bmbt_ops = { .init_rec_from_cur = xfs_bmbt_init_rec_from_cur, .init_ptr_from_cur = xfs_bmbt_init_ptr_from_cur, .key_diff = xfs_bmbt_key_diff, + .buf_ops = &xfs_bmbt_buf_ops, #ifdef DEBUG .keys_inorder = xfs_bmbt_keys_inorder, .recs_inorder = xfs_bmbt_recs_inorder, diff --git a/fs/xfs/xfs_bmap_btree.h b/fs/xfs/xfs_bmap_btree.h index 0e66c4ea0f85..88469ca08696 100644 --- a/fs/xfs/xfs_bmap_btree.h +++ b/fs/xfs/xfs_bmap_btree.h @@ -236,5 +236,6 @@ extern int xfs_bmbt_maxrecs(struct xfs_mount *, int blocklen, int leaf); extern struct xfs_btree_cur *xfs_bmbt_init_cursor(struct xfs_mount *, struct xfs_trans *, struct xfs_inode *, int); +extern const struct xfs_buf_ops xfs_bmbt_buf_ops; #endif /* __XFS_BMAP_BTREE_H__ */ diff --git a/fs/xfs/xfs_btree.c b/fs/xfs/xfs_btree.c index e53e317b1582..db010408d701 100644 --- a/fs/xfs/xfs_btree.c +++ b/fs/xfs/xfs_btree.c @@ -266,9 +266,13 @@ xfs_btree_dup_cursor( for (i = 0; i < new->bc_nlevels; i++) { new->bc_ptrs[i] = cur->bc_ptrs[i]; new->bc_ra[i] = cur->bc_ra[i]; - if ((bp = cur->bc_bufs[i])) { - if ((error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, - XFS_BUF_ADDR(bp), mp->m_bsize, 0, &bp))) { + bp = cur->bc_bufs[i]; + if (bp) { + error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, + XFS_BUF_ADDR(bp), mp->m_bsize, + 0, &bp, + cur->bc_ops->buf_ops); + if (error) { xfs_btree_del_cursor(new, error); *ncur = NULL; return error; @@ -609,25 +613,26 @@ xfs_btree_offsets( * Get a buffer for the block, return it read in. * Long-form addressing. */ -int /* error */ +int xfs_btree_read_bufl( - xfs_mount_t *mp, /* file system mount point */ - xfs_trans_t *tp, /* transaction pointer */ - xfs_fsblock_t fsbno, /* file system block number */ - uint lock, /* lock flags for read_buf */ - xfs_buf_t **bpp, /* buffer for fsbno */ - int refval) /* ref count value for buffer */ -{ - xfs_buf_t *bp; /* return value */ + struct xfs_mount *mp, /* file system mount point */ + struct xfs_trans *tp, /* transaction pointer */ + xfs_fsblock_t fsbno, /* file system block number */ + uint lock, /* lock flags for read_buf */ + struct xfs_buf **bpp, /* buffer for fsbno */ + int refval, /* ref count value for buffer */ + const struct xfs_buf_ops *ops) +{ + struct xfs_buf *bp; /* return value */ xfs_daddr_t d; /* real disk block address */ - int error; + int error; ASSERT(fsbno != NULLFSBLOCK); d = XFS_FSB_TO_DADDR(mp, fsbno); - if ((error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, d, - mp->m_bsize, lock, &bp))) { + error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, d, + mp->m_bsize, lock, &bp, ops); + if (error) return error; - } ASSERT(!xfs_buf_geterror(bp)); if (bp) xfs_buf_set_ref(bp, refval); @@ -642,15 +647,16 @@ xfs_btree_read_bufl( /* ARGSUSED */ void xfs_btree_reada_bufl( - xfs_mount_t *mp, /* file system mount point */ - xfs_fsblock_t fsbno, /* file system block number */ - xfs_extlen_t count) /* count of filesystem blocks */ + struct xfs_mount *mp, /* file system mount point */ + xfs_fsblock_t fsbno, /* file system block number */ + xfs_extlen_t count, /* count of filesystem blocks */ + const struct xfs_buf_ops *ops) { xfs_daddr_t d; ASSERT(fsbno != NULLFSBLOCK); d = XFS_FSB_TO_DADDR(mp, fsbno); - xfs_buf_readahead(mp->m_ddev_targp, d, mp->m_bsize * count); + xfs_buf_readahead(mp->m_ddev_targp, d, mp->m_bsize * count, ops); } /* @@ -660,17 +666,18 @@ xfs_btree_reada_bufl( /* ARGSUSED */ void xfs_btree_reada_bufs( - xfs_mount_t *mp, /* file system mount point */ - xfs_agnumber_t agno, /* allocation group number */ - xfs_agblock_t agbno, /* allocation group block number */ - xfs_extlen_t count) /* count of filesystem blocks */ + struct xfs_mount *mp, /* file system mount point */ + xfs_agnumber_t agno, /* allocation group number */ + xfs_agblock_t agbno, /* allocation group block number */ + xfs_extlen_t count, /* count of filesystem blocks */ + const struct xfs_buf_ops *ops) { xfs_daddr_t d; ASSERT(agno != NULLAGNUMBER); ASSERT(agbno != NULLAGBLOCK); d = XFS_AGB_TO_DADDR(mp, agno, agbno); - xfs_buf_readahead(mp->m_ddev_targp, d, mp->m_bsize * count); + xfs_buf_readahead(mp->m_ddev_targp, d, mp->m_bsize * count, ops); } STATIC int @@ -684,12 +691,14 @@ xfs_btree_readahead_lblock( xfs_dfsbno_t right = be64_to_cpu(block->bb_u.l.bb_rightsib); if ((lr & XFS_BTCUR_LEFTRA) && left != NULLDFSBNO) { - xfs_btree_reada_bufl(cur->bc_mp, left, 1); + xfs_btree_reada_bufl(cur->bc_mp, left, 1, + cur->bc_ops->buf_ops); rval++; } if ((lr & XFS_BTCUR_RIGHTRA) && right != NULLDFSBNO) { - xfs_btree_reada_bufl(cur->bc_mp, right, 1); + xfs_btree_reada_bufl(cur->bc_mp, right, 1, + cur->bc_ops->buf_ops); rval++; } @@ -709,13 +718,13 @@ xfs_btree_readahead_sblock( if ((lr & XFS_BTCUR_LEFTRA) && left != NULLAGBLOCK) { xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, - left, 1); + left, 1, cur->bc_ops->buf_ops); rval++; } if ((lr & XFS_BTCUR_RIGHTRA) && right != NULLAGBLOCK) { xfs_btree_reada_bufs(cur->bc_mp, cur->bc_private.a.agno, - right, 1); + right, 1, cur->bc_ops->buf_ops); rval++; } @@ -853,18 +862,22 @@ xfs_btree_set_sibling( } } -STATIC void +void xfs_btree_init_block( - struct xfs_btree_cur *cur, - int level, - int numrecs, - struct xfs_btree_block *new) /* new block */ + struct xfs_mount *mp, + struct xfs_buf *bp, + __u32 magic, + __u16 level, + __u16 numrecs, + unsigned int flags) { - new->bb_magic = cpu_to_be32(xfs_magics[cur->bc_btnum]); + struct xfs_btree_block *new = XFS_BUF_TO_BLOCK(bp); + + new->bb_magic = cpu_to_be32(magic); new->bb_level = cpu_to_be16(level); new->bb_numrecs = cpu_to_be16(numrecs); - if (cur->bc_flags & XFS_BTREE_LONG_PTRS) { + if (flags & XFS_BTREE_LONG_PTRS) { new->bb_u.l.bb_leftsib = cpu_to_be64(NULLDFSBNO); new->bb_u.l.bb_rightsib = cpu_to_be64(NULLDFSBNO); } else { @@ -873,6 +886,17 @@ xfs_btree_init_block( } } +STATIC void +xfs_btree_init_block_cur( + struct xfs_btree_cur *cur, + int level, + int numrecs, + struct xfs_buf *bp) +{ + xfs_btree_init_block(cur->bc_mp, bp, xfs_magics[cur->bc_btnum], + level, numrecs, cur->bc_flags); +} + /* * Return true if ptr is the last record in the btree and * we need to track updateÑ• to this record. The decision @@ -972,6 +996,7 @@ xfs_btree_get_buf_block( if (!*bpp) return ENOMEM; + (*bpp)->b_ops = cur->bc_ops->buf_ops; *block = XFS_BUF_TO_BLOCK(*bpp); return 0; } @@ -998,19 +1023,15 @@ xfs_btree_read_buf_block( d = xfs_btree_ptr_to_daddr(cur, ptr); error = xfs_trans_read_buf(mp, cur->bc_tp, mp->m_ddev_targp, d, - mp->m_bsize, flags, bpp); + mp->m_bsize, flags, bpp, + cur->bc_ops->buf_ops); if (error) return error; ASSERT(!xfs_buf_geterror(*bpp)); - xfs_btree_set_refs(cur, *bpp); *block = XFS_BUF_TO_BLOCK(*bpp); - - error = xfs_btree_check_block(cur, *block, level, *bpp); - if (error) - xfs_trans_brelse(cur->bc_tp, *bpp); - return error; + return 0; } /* @@ -2183,7 +2204,7 @@ xfs_btree_split( goto error0; /* Fill in the btree header for the new right block. */ - xfs_btree_init_block(cur, xfs_btree_get_level(left), 0, right); + xfs_btree_init_block_cur(cur, xfs_btree_get_level(left), 0, rbp); /* * Split the entries between the old and the new block evenly. @@ -2492,7 +2513,7 @@ xfs_btree_new_root( nptr = 2; } /* Fill in the new block's btree header and log it. */ - xfs_btree_init_block(cur, cur->bc_nlevels, 2, new); + xfs_btree_init_block_cur(cur, cur->bc_nlevels, 2, nbp); xfs_btree_log_block(cur, nbp, XFS_BB_ALL_BITS); ASSERT(!xfs_btree_ptr_is_null(cur, &lptr) && !xfs_btree_ptr_is_null(cur, &rptr)); diff --git a/fs/xfs/xfs_btree.h b/fs/xfs/xfs_btree.h index 5b240de104c0..f932897194eb 100644 --- a/fs/xfs/xfs_btree.h +++ b/fs/xfs/xfs_btree.h @@ -188,6 +188,8 @@ struct xfs_btree_ops { __int64_t (*key_diff)(struct xfs_btree_cur *cur, union xfs_btree_key *key); + const struct xfs_buf_ops *buf_ops; + #ifdef DEBUG /* check that k1 is lower than k2 */ int (*keys_inorder)(struct xfs_btree_cur *cur, @@ -355,7 +357,8 @@ xfs_btree_read_bufl( xfs_fsblock_t fsbno, /* file system block number */ uint lock, /* lock flags for read_buf */ struct xfs_buf **bpp, /* buffer for fsbno */ - int refval);/* ref count value for buffer */ + int refval, /* ref count value for buffer */ + const struct xfs_buf_ops *ops); /* * Read-ahead the block, don't wait for it, don't return a buffer. @@ -365,7 +368,8 @@ void /* error */ xfs_btree_reada_bufl( struct xfs_mount *mp, /* file system mount point */ xfs_fsblock_t fsbno, /* file system block number */ - xfs_extlen_t count); /* count of filesystem blocks */ + xfs_extlen_t count, /* count of filesystem blocks */ + const struct xfs_buf_ops *ops); /* * Read-ahead the block, don't wait for it, don't return a buffer. @@ -376,8 +380,20 @@ xfs_btree_reada_bufs( struct xfs_mount *mp, /* file system mount point */ xfs_agnumber_t agno, /* allocation group number */ xfs_agblock_t agbno, /* allocation group block number */ - xfs_extlen_t count); /* count of filesystem blocks */ + xfs_extlen_t count, /* count of filesystem blocks */ + const struct xfs_buf_ops *ops); +/* + * Initialise a new btree block header + */ +void +xfs_btree_init_block( + struct xfs_mount *mp, + struct xfs_buf *bp, + __u32 magic, + __u16 level, + __u16 numrecs, + unsigned int flags); /* * Common btree core entry points. diff --git a/fs/xfs/xfs_buf.c b/fs/xfs/xfs_buf.c index 4b0b8dd1b7b0..26673a0b20e7 100644 --- a/fs/xfs/xfs_buf.c +++ b/fs/xfs/xfs_buf.c @@ -569,7 +569,9 @@ found: */ if (bp->b_flags & XBF_STALE) { ASSERT((bp->b_flags & _XBF_DELWRI_Q) == 0); + ASSERT(bp->b_iodone == NULL); bp->b_flags &= _XBF_KMEM | _XBF_PAGES; + bp->b_ops = NULL; } trace_xfs_buf_find(bp, flags, _RET_IP_); @@ -654,7 +656,8 @@ xfs_buf_read_map( struct xfs_buftarg *target, struct xfs_buf_map *map, int nmaps, - xfs_buf_flags_t flags) + xfs_buf_flags_t flags, + const struct xfs_buf_ops *ops) { struct xfs_buf *bp; @@ -666,6 +669,7 @@ xfs_buf_read_map( if (!XFS_BUF_ISDONE(bp)) { XFS_STATS_INC(xb_get_read); + bp->b_ops = ops; _xfs_buf_read(bp, flags); } else if (flags & XBF_ASYNC) { /* @@ -691,13 +695,14 @@ void xfs_buf_readahead_map( struct xfs_buftarg *target, struct xfs_buf_map *map, - int nmaps) + int nmaps, + const struct xfs_buf_ops *ops) { if (bdi_read_congested(target->bt_bdi)) return; xfs_buf_read_map(target, map, nmaps, - XBF_TRYLOCK|XBF_ASYNC|XBF_READ_AHEAD); + XBF_TRYLOCK|XBF_ASYNC|XBF_READ_AHEAD, ops); } /* @@ -709,10 +714,10 @@ xfs_buf_read_uncached( struct xfs_buftarg *target, xfs_daddr_t daddr, size_t numblks, - int flags) + int flags, + const struct xfs_buf_ops *ops) { - xfs_buf_t *bp; - int error; + struct xfs_buf *bp; bp = xfs_buf_get_uncached(target, numblks, flags); if (!bp) @@ -723,13 +728,10 @@ xfs_buf_read_uncached( bp->b_bn = daddr; bp->b_maps[0].bm_bn = daddr; bp->b_flags |= XBF_READ; + bp->b_ops = ops; xfsbdstrat(target->bt_mount, bp); - error = xfs_buf_iowait(bp); - if (error) { - xfs_buf_relse(bp); - return NULL; - } + xfs_buf_iowait(bp); return bp; } @@ -999,27 +1001,37 @@ STATIC void xfs_buf_iodone_work( struct work_struct *work) { - xfs_buf_t *bp = + struct xfs_buf *bp = container_of(work, xfs_buf_t, b_iodone_work); + bool read = !!(bp->b_flags & XBF_READ); + + bp->b_flags &= ~(XBF_READ | XBF_WRITE | XBF_READ_AHEAD); + if (read && bp->b_ops) + bp->b_ops->verify_read(bp); if (bp->b_iodone) (*(bp->b_iodone))(bp); else if (bp->b_flags & XBF_ASYNC) xfs_buf_relse(bp); + else { + ASSERT(read && bp->b_ops); + complete(&bp->b_iowait); + } } void xfs_buf_ioend( - xfs_buf_t *bp, - int schedule) + struct xfs_buf *bp, + int schedule) { + bool read = !!(bp->b_flags & XBF_READ); + trace_xfs_buf_iodone(bp, _RET_IP_); - bp->b_flags &= ~(XBF_READ | XBF_WRITE | XBF_READ_AHEAD); if (bp->b_error == 0) bp->b_flags |= XBF_DONE; - if ((bp->b_iodone) || (bp->b_flags & XBF_ASYNC)) { + if (bp->b_iodone || (read && bp->b_ops) || (bp->b_flags & XBF_ASYNC)) { if (schedule) { INIT_WORK(&bp->b_iodone_work, xfs_buf_iodone_work); queue_work(xfslogd_workqueue, &bp->b_iodone_work); @@ -1027,6 +1039,7 @@ xfs_buf_ioend( xfs_buf_iodone_work(&bp->b_iodone_work); } } else { + bp->b_flags &= ~(XBF_READ | XBF_WRITE | XBF_READ_AHEAD); complete(&bp->b_iowait); } } @@ -1314,6 +1327,20 @@ _xfs_buf_ioapply( rw |= REQ_FUA; if (bp->b_flags & XBF_FLUSH) rw |= REQ_FLUSH; + + /* + * Run the write verifier callback function if it exists. If + * this function fails it will mark the buffer with an error and + * the IO should not be dispatched. + */ + if (bp->b_ops) { + bp->b_ops->verify_write(bp); + if (bp->b_error) { + xfs_force_shutdown(bp->b_target->bt_mount, + SHUTDOWN_CORRUPT_INCORE); + return; + } + } } else if (bp->b_flags & XBF_READ_AHEAD) { rw = READA; } else { diff --git a/fs/xfs/xfs_buf.h b/fs/xfs/xfs_buf.h index 7c0b6a0a1557..23f5642480bb 100644 --- a/fs/xfs/xfs_buf.h +++ b/fs/xfs/xfs_buf.h @@ -100,6 +100,7 @@ typedef struct xfs_buftarg { struct xfs_buf; typedef void (*xfs_buf_iodone_t)(struct xfs_buf *); + #define XB_PAGES 2 struct xfs_buf_map { @@ -110,6 +111,11 @@ struct xfs_buf_map { #define DEFINE_SINGLE_BUF_MAP(map, blkno, numblk) \ struct xfs_buf_map (map) = { .bm_bn = (blkno), .bm_len = (numblk) }; +struct xfs_buf_ops { + void (*verify_read)(struct xfs_buf *); + void (*verify_write)(struct xfs_buf *); +}; + typedef struct xfs_buf { /* * first cacheline holds all the fields needed for an uncontended cache @@ -153,13 +159,13 @@ typedef struct xfs_buf { unsigned int b_page_count; /* size of page array */ unsigned int b_offset; /* page offset in first page */ unsigned short b_error; /* error code on I/O */ + const struct xfs_buf_ops *b_ops; #ifdef XFS_BUF_LOCK_TRACKING int b_last_holder; #endif } xfs_buf_t; - /* Finding and Reading Buffers */ struct xfs_buf *_xfs_buf_find(struct xfs_buftarg *target, struct xfs_buf_map *map, int nmaps, @@ -196,9 +202,11 @@ struct xfs_buf *xfs_buf_get_map(struct xfs_buftarg *target, xfs_buf_flags_t flags); struct xfs_buf *xfs_buf_read_map(struct xfs_buftarg *target, struct xfs_buf_map *map, int nmaps, - xfs_buf_flags_t flags); + xfs_buf_flags_t flags, + const struct xfs_buf_ops *ops); void xfs_buf_readahead_map(struct xfs_buftarg *target, - struct xfs_buf_map *map, int nmaps); + struct xfs_buf_map *map, int nmaps, + const struct xfs_buf_ops *ops); static inline struct xfs_buf * xfs_buf_get( @@ -216,20 +224,22 @@ xfs_buf_read( struct xfs_buftarg *target, xfs_daddr_t blkno, size_t numblks, - xfs_buf_flags_t flags) + xfs_buf_flags_t flags, + const struct xfs_buf_ops *ops) { DEFINE_SINGLE_BUF_MAP(map, blkno, numblks); - return xfs_buf_read_map(target, &map, 1, flags); + return xfs_buf_read_map(target, &map, 1, flags, ops); } static inline void xfs_buf_readahead( struct xfs_buftarg *target, xfs_daddr_t blkno, - size_t numblks) + size_t numblks, + const struct xfs_buf_ops *ops) { DEFINE_SINGLE_BUF_MAP(map, blkno, numblks); - return xfs_buf_readahead_map(target, &map, 1); + return xfs_buf_readahead_map(target, &map, 1, ops); } struct xfs_buf *xfs_buf_get_empty(struct xfs_buftarg *target, size_t numblks); @@ -239,7 +249,8 @@ int xfs_buf_associate_memory(struct xfs_buf *bp, void *mem, size_t length); struct xfs_buf *xfs_buf_get_uncached(struct xfs_buftarg *target, size_t numblks, int flags); struct xfs_buf *xfs_buf_read_uncached(struct xfs_buftarg *target, - xfs_daddr_t daddr, size_t numblks, int flags); + xfs_daddr_t daddr, size_t numblks, int flags, + const struct xfs_buf_ops *ops); void xfs_buf_hold(struct xfs_buf *bp); /* Releasing Buffers */ diff --git a/fs/xfs/xfs_cksum.h b/fs/xfs/xfs_cksum.h new file mode 100644 index 000000000000..fad1676ad8cd --- /dev/null +++ b/fs/xfs/xfs_cksum.h @@ -0,0 +1,63 @@ +#ifndef _XFS_CKSUM_H +#define _XFS_CKSUM_H 1 + +#define XFS_CRC_SEED (~(__uint32_t)0) + +/* + * Calculate the intermediate checksum for a buffer that has the CRC field + * inside it. The offset of the 32bit crc fields is passed as the + * cksum_offset parameter. + */ +static inline __uint32_t +xfs_start_cksum(char *buffer, size_t length, unsigned long cksum_offset) +{ + __uint32_t zero = 0; + __uint32_t crc; + + /* Calculate CRC up to the checksum. */ + crc = crc32c(XFS_CRC_SEED, buffer, cksum_offset); + + /* Skip checksum field */ + crc = crc32c(crc, &zero, sizeof(__u32)); + + /* Calculate the rest of the CRC. */ + return crc32c(crc, &buffer[cksum_offset + sizeof(__be32)], + length - (cksum_offset + sizeof(__be32))); +} + +/* + * Convert the intermediate checksum to the final ondisk format. + * + * The CRC32c calculation uses LE format even on BE machines, but returns the + * result in host endian format. Hence we need to byte swap it back to LE format + * so that it is consistent on disk. + */ +static inline __le32 +xfs_end_cksum(__uint32_t crc) +{ + return ~cpu_to_le32(crc); +} + +/* + * Helper to generate the checksum for a buffer. + */ +static inline void +xfs_update_cksum(char *buffer, size_t length, unsigned long cksum_offset) +{ + __uint32_t crc = xfs_start_cksum(buffer, length, cksum_offset); + + *(__le32 *)(buffer + cksum_offset) = xfs_end_cksum(crc); +} + +/* + * Helper to verify the checksum for a buffer. + */ +static inline int +xfs_verify_cksum(char *buffer, size_t length, unsigned long cksum_offset) +{ + __uint32_t crc = xfs_start_cksum(buffer, length, cksum_offset); + + return *(__le32 *)(buffer + cksum_offset) == xfs_end_cksum(crc); +} + +#endif /* _XFS_CKSUM_H */ diff --git a/fs/xfs/xfs_da_btree.c b/fs/xfs/xfs_da_btree.c index 7bfb7dd334fc..4d7696a02418 100644 --- a/fs/xfs/xfs_da_btree.c +++ b/fs/xfs/xfs_da_btree.c @@ -91,6 +91,84 @@ STATIC int xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *save_blk); STATIC void xfs_da_state_kill_altpath(xfs_da_state_t *state); +static void +xfs_da_node_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_da_node_hdr *hdr = bp->b_addr; + int block_ok = 0; + + block_ok = hdr->info.magic == cpu_to_be16(XFS_DA_NODE_MAGIC); + block_ok = block_ok && + be16_to_cpu(hdr->level) > 0 && + be16_to_cpu(hdr->count) > 0 ; + if (!block_ok) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } + +} + +static void +xfs_da_node_write_verify( + struct xfs_buf *bp) +{ + xfs_da_node_verify(bp); +} + +/* + * leaf/node format detection on trees is sketchy, so a node read can be done on + * leaf level blocks when detection identifies the tree as a node format tree + * incorrectly. In this case, we need to swap the verifier to match the correct + * format of the block being read. + */ +static void +xfs_da_node_read_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_da_blkinfo *info = bp->b_addr; + + switch (be16_to_cpu(info->magic)) { + case XFS_DA_NODE_MAGIC: + xfs_da_node_verify(bp); + break; + case XFS_ATTR_LEAF_MAGIC: + bp->b_ops = &xfs_attr_leaf_buf_ops; + bp->b_ops->verify_read(bp); + return; + case XFS_DIR2_LEAFN_MAGIC: + bp->b_ops = &xfs_dir2_leafn_buf_ops; + bp->b_ops->verify_read(bp); + return; + default: + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, + mp, info); + xfs_buf_ioerror(bp, EFSCORRUPTED); + break; + } +} + +const struct xfs_buf_ops xfs_da_node_buf_ops = { + .verify_read = xfs_da_node_read_verify, + .verify_write = xfs_da_node_write_verify, +}; + + +int +xfs_da_node_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t bno, + xfs_daddr_t mappedbno, + struct xfs_buf **bpp, + int which_fork) +{ + return xfs_da_read_buf(tp, dp, bno, mappedbno, bpp, + which_fork, &xfs_da_node_buf_ops); +} + /*======================================================================== * Routines used for growing the Btree. *========================================================================*/ @@ -125,6 +203,7 @@ xfs_da_node_create(xfs_da_args_t *args, xfs_dablk_t blkno, int level, xfs_trans_log_buf(tp, bp, XFS_DA_LOGRANGE(node, &node->hdr, sizeof(node->hdr))); + bp->b_ops = &xfs_da_node_buf_ops; *bpp = bp; return(0); } @@ -324,6 +403,8 @@ xfs_da_root_split(xfs_da_state_t *state, xfs_da_state_blk_t *blk1, } memcpy(node, oldroot, size); xfs_trans_log_buf(tp, bp, 0, size - 1); + + bp->b_ops = blk1->bp->b_ops; blk1->bp = bp; blk1->blkno = blkno; @@ -746,7 +827,7 @@ xfs_da_root_join(xfs_da_state_t *state, xfs_da_state_blk_t *root_blk) */ child = be32_to_cpu(oldroot->btree[0].before); ASSERT(child != 0); - error = xfs_da_read_buf(args->trans, args->dp, child, -1, &bp, + error = xfs_da_node_read(args->trans, args->dp, child, -1, &bp, args->whichfork); if (error) return(error); @@ -754,7 +835,14 @@ xfs_da_root_join(xfs_da_state_t *state, xfs_da_state_blk_t *root_blk) xfs_da_blkinfo_onlychild_validate(bp->b_addr, be16_to_cpu(oldroot->hdr.level)); + /* + * This could be copying a leaf back into the root block in the case of + * there only being a single leaf block left in the tree. Hence we have + * to update the b_ops pointer as well to match the buffer type change + * that could occur. + */ memcpy(root_blk->bp->b_addr, bp->b_addr, state->blocksize); + root_blk->bp->b_ops = bp->b_ops; xfs_trans_log_buf(args->trans, root_blk->bp, 0, state->blocksize - 1); error = xfs_da_shrink_inode(args, child, bp); return(error); @@ -779,6 +867,8 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action) xfs_dablk_t blkno; struct xfs_buf *bp; + trace_xfs_da_node_toosmall(state->args); + /* * Check for the degenerate case of the block being over 50% full. * If so, it's not worth even looking to see if we might be able @@ -835,7 +925,7 @@ xfs_da_node_toosmall(xfs_da_state_t *state, int *action) blkno = be32_to_cpu(info->back); if (blkno == 0) continue; - error = xfs_da_read_buf(state->args->trans, state->args->dp, + error = xfs_da_node_read(state->args->trans, state->args->dp, blkno, -1, &bp, state->args->whichfork); if (error) return(error); @@ -900,6 +990,8 @@ xfs_da_fixhashpath(xfs_da_state_t *state, xfs_da_state_path_t *path) xfs_dahash_t lasthash=0; int level, count; + trace_xfs_da_fixhashpath(state->args); + level = path->active-1; blk = &path->blk[ level ]; switch (blk->magic) { @@ -1079,7 +1171,7 @@ xfs_da_node_lookup_int(xfs_da_state_t *state, int *result) * Read the next node down in the tree. */ blk->blkno = blkno; - error = xfs_da_read_buf(args->trans, args->dp, blkno, + error = xfs_da_node_read(args->trans, args->dp, blkno, -1, &blk->bp, args->whichfork); if (error) { blk->blkno = 0; @@ -1241,7 +1333,7 @@ xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk, new_info->forw = cpu_to_be32(old_blk->blkno); new_info->back = old_info->back; if (old_info->back) { - error = xfs_da_read_buf(args->trans, args->dp, + error = xfs_da_node_read(args->trans, args->dp, be32_to_cpu(old_info->back), -1, &bp, args->whichfork); if (error) @@ -1262,7 +1354,7 @@ xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk, new_info->forw = old_info->forw; new_info->back = cpu_to_be32(old_blk->blkno); if (old_info->forw) { - error = xfs_da_read_buf(args->trans, args->dp, + error = xfs_da_node_read(args->trans, args->dp, be32_to_cpu(old_info->forw), -1, &bp, args->whichfork); if (error) @@ -1362,7 +1454,7 @@ xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk, trace_xfs_da_unlink_back(args); save_info->back = drop_info->back; if (drop_info->back) { - error = xfs_da_read_buf(args->trans, args->dp, + error = xfs_da_node_read(args->trans, args->dp, be32_to_cpu(drop_info->back), -1, &bp, args->whichfork); if (error) @@ -1379,7 +1471,7 @@ xfs_da_blk_unlink(xfs_da_state_t *state, xfs_da_state_blk_t *drop_blk, trace_xfs_da_unlink_forward(args); save_info->forw = drop_info->forw; if (drop_info->forw) { - error = xfs_da_read_buf(args->trans, args->dp, + error = xfs_da_node_read(args->trans, args->dp, be32_to_cpu(drop_info->forw), -1, &bp, args->whichfork); if (error) @@ -1417,6 +1509,8 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path, xfs_dablk_t blkno=0; int level, error; + trace_xfs_da_path_shift(state->args); + /* * Roll up the Btree looking for the first block where our * current index is not at the edge of the block. Note that @@ -1463,8 +1557,8 @@ xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path, * Read the next child block. */ blk->blkno = blkno; - error = xfs_da_read_buf(args->trans, args->dp, blkno, -1, - &blk->bp, args->whichfork); + error = xfs_da_node_read(args->trans, args->dp, blkno, -1, + &blk->bp, args->whichfork); if (error) return(error); ASSERT(blk->bp != NULL); @@ -1727,7 +1821,8 @@ xfs_da_swap_lastblock( * Read the last block in the btree space. */ last_blkno = (xfs_dablk_t)lastoff - mp->m_dirblkfsbs; - if ((error = xfs_da_read_buf(tp, ip, last_blkno, -1, &last_buf, w))) + error = xfs_da_node_read(tp, ip, last_blkno, -1, &last_buf, w); + if (error) return error; /* * Copy the last block into the dead buffer and log it. @@ -1753,7 +1848,8 @@ xfs_da_swap_lastblock( * If the moved block has a left sibling, fix up the pointers. */ if ((sib_blkno = be32_to_cpu(dead_info->back))) { - if ((error = xfs_da_read_buf(tp, ip, sib_blkno, -1, &sib_buf, w))) + error = xfs_da_node_read(tp, ip, sib_blkno, -1, &sib_buf, w); + if (error) goto done; sib_info = sib_buf->b_addr; if (unlikely( @@ -1774,7 +1870,8 @@ xfs_da_swap_lastblock( * If the moved block has a right sibling, fix up the pointers. */ if ((sib_blkno = be32_to_cpu(dead_info->forw))) { - if ((error = xfs_da_read_buf(tp, ip, sib_blkno, -1, &sib_buf, w))) + error = xfs_da_node_read(tp, ip, sib_blkno, -1, &sib_buf, w); + if (error) goto done; sib_info = sib_buf->b_addr; if (unlikely( @@ -1797,7 +1894,8 @@ xfs_da_swap_lastblock( * Walk down the tree looking for the parent of the moved block. */ for (;;) { - if ((error = xfs_da_read_buf(tp, ip, par_blkno, -1, &par_buf, w))) + error = xfs_da_node_read(tp, ip, par_blkno, -1, &par_buf, w); + if (error) goto done; par_node = par_buf->b_addr; if (unlikely(par_node->hdr.info.magic != @@ -1847,7 +1945,8 @@ xfs_da_swap_lastblock( error = XFS_ERROR(EFSCORRUPTED); goto done; } - if ((error = xfs_da_read_buf(tp, ip, par_blkno, -1, &par_buf, w))) + error = xfs_da_node_read(tp, ip, par_blkno, -1, &par_buf, w); + if (error) goto done; par_node = par_buf->b_addr; if (unlikely( @@ -2133,7 +2232,8 @@ xfs_da_read_buf( xfs_dablk_t bno, xfs_daddr_t mappedbno, struct xfs_buf **bpp, - int whichfork) + int whichfork, + const struct xfs_buf_ops *ops) { struct xfs_buf *bp; struct xfs_buf_map map; @@ -2155,7 +2255,7 @@ xfs_da_read_buf( error = xfs_trans_read_buf_map(dp->i_mount, trans, dp->i_mount->m_ddev_targp, - mapp, nmap, 0, &bp); + mapp, nmap, 0, &bp, ops); if (error) goto out_free; @@ -2211,9 +2311,10 @@ xfs_da_reada_buf( struct xfs_trans *trans, struct xfs_inode *dp, xfs_dablk_t bno, - int whichfork) + xfs_daddr_t mappedbno, + int whichfork, + const struct xfs_buf_ops *ops) { - xfs_daddr_t mappedbno = -1; struct xfs_buf_map map; struct xfs_buf_map *mapp; int nmap; @@ -2221,7 +2322,7 @@ xfs_da_reada_buf( mapp = ↦ nmap = 1; - error = xfs_dabuf_map(trans, dp, bno, -1, whichfork, + error = xfs_dabuf_map(trans, dp, bno, mappedbno, whichfork, &mapp, &nmap); if (error) { /* mapping a hole is not an error, but we don't continue */ @@ -2231,7 +2332,7 @@ xfs_da_reada_buf( } mappedbno = mapp[0].bm_bn; - xfs_buf_readahead_map(dp->i_mount->m_ddev_targp, mapp, nmap); + xfs_buf_readahead_map(dp->i_mount->m_ddev_targp, mapp, nmap, ops); out_free: if (mapp != &map) diff --git a/fs/xfs/xfs_da_btree.h b/fs/xfs/xfs_da_btree.h index 132adafb041e..ee5170c46ae1 100644 --- a/fs/xfs/xfs_da_btree.h +++ b/fs/xfs/xfs_da_btree.h @@ -18,7 +18,6 @@ #ifndef __XFS_DA_BTREE_H__ #define __XFS_DA_BTREE_H__ -struct xfs_buf; struct xfs_bmap_free; struct xfs_inode; struct xfs_mount; @@ -214,6 +213,9 @@ int xfs_da_path_shift(xfs_da_state_t *state, xfs_da_state_path_t *path, */ int xfs_da_blk_link(xfs_da_state_t *state, xfs_da_state_blk_t *old_blk, xfs_da_state_blk_t *new_blk); +int xfs_da_node_read(struct xfs_trans *tp, struct xfs_inode *dp, + xfs_dablk_t bno, xfs_daddr_t mappedbno, + struct xfs_buf **bpp, int which_fork); /* * Utility routines. @@ -226,9 +228,11 @@ int xfs_da_get_buf(struct xfs_trans *trans, struct xfs_inode *dp, struct xfs_buf **bp, int whichfork); int xfs_da_read_buf(struct xfs_trans *trans, struct xfs_inode *dp, xfs_dablk_t bno, xfs_daddr_t mappedbno, - struct xfs_buf **bpp, int whichfork); + struct xfs_buf **bpp, int whichfork, + const struct xfs_buf_ops *ops); xfs_daddr_t xfs_da_reada_buf(struct xfs_trans *trans, struct xfs_inode *dp, - xfs_dablk_t bno, int whichfork); + xfs_dablk_t bno, xfs_daddr_t mapped_bno, + int whichfork, const struct xfs_buf_ops *ops); int xfs_da_shrink_inode(xfs_da_args_t *args, xfs_dablk_t dead_blkno, struct xfs_buf *dead_buf); diff --git a/fs/xfs/xfs_dfrag.c b/fs/xfs/xfs_dfrag.c index b9b8646e62db..d0e9c74d3d96 100644 --- a/fs/xfs/xfs_dfrag.c +++ b/fs/xfs/xfs_dfrag.c @@ -246,12 +246,10 @@ xfs_swap_extents( goto out_unlock; } - if (VN_CACHED(VFS_I(tip)) != 0) { - error = xfs_flushinval_pages(tip, 0, -1, - FI_REMAPF_LOCKED); - if (error) - goto out_unlock; - } + error = -filemap_write_and_wait(VFS_I(ip)->i_mapping); + if (error) + goto out_unlock; + truncate_pagecache_range(VFS_I(ip), 0, -1); /* Verify O_DIRECT for ftmp */ if (VN_CACHED(VFS_I(tip)) != 0) { @@ -315,8 +313,7 @@ xfs_swap_extents( * are safe. We don't really care if non-io related * fields change. */ - - xfs_tosspages(ip, 0, -1, FI_REMAPF); + truncate_pagecache_range(VFS_I(ip), 0, -1); tp = xfs_trans_alloc(mp, XFS_TRANS_SWAPEXT); if ((error = xfs_trans_reserve(tp, 0, diff --git a/fs/xfs/xfs_dir2_block.c b/fs/xfs/xfs_dir2_block.c index e93ca8f054f4..7536faaa61e7 100644 --- a/fs/xfs/xfs_dir2_block.c +++ b/fs/xfs/xfs_dir2_block.c @@ -56,6 +56,214 @@ xfs_dir_startup(void) xfs_dir_hash_dotdot = xfs_da_hashname((unsigned char *)"..", 2); } +static void +xfs_dir2_block_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_dir2_data_hdr *hdr = bp->b_addr; + int block_ok = 0; + + block_ok = hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC); + block_ok = block_ok && __xfs_dir2_data_check(NULL, bp) == 0; + + if (!block_ok) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_dir2_block_read_verify( + struct xfs_buf *bp) +{ + xfs_dir2_block_verify(bp); +} + +static void +xfs_dir2_block_write_verify( + struct xfs_buf *bp) +{ + xfs_dir2_block_verify(bp); +} + +const struct xfs_buf_ops xfs_dir2_block_buf_ops = { + .verify_read = xfs_dir2_block_read_verify, + .verify_write = xfs_dir2_block_write_verify, +}; + +static int +xfs_dir2_block_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + struct xfs_buf **bpp) +{ + struct xfs_mount *mp = dp->i_mount; + + return xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, bpp, + XFS_DATA_FORK, &xfs_dir2_block_buf_ops); +} + +static void +xfs_dir2_block_need_space( + struct xfs_dir2_data_hdr *hdr, + struct xfs_dir2_block_tail *btp, + struct xfs_dir2_leaf_entry *blp, + __be16 **tagpp, + struct xfs_dir2_data_unused **dupp, + struct xfs_dir2_data_unused **enddupp, + int *compact, + int len) +{ + struct xfs_dir2_data_free *bf; + __be16 *tagp = NULL; + struct xfs_dir2_data_unused *dup = NULL; + struct xfs_dir2_data_unused *enddup = NULL; + + *compact = 0; + bf = hdr->bestfree; + + /* + * If there are stale entries we'll use one for the leaf. + */ + if (btp->stale) { + if (be16_to_cpu(bf[0].length) >= len) { + /* + * The biggest entry enough to avoid compaction. + */ + dup = (xfs_dir2_data_unused_t *) + ((char *)hdr + be16_to_cpu(bf[0].offset)); + goto out; + } + + /* + * Will need to compact to make this work. + * Tag just before the first leaf entry. + */ + *compact = 1; + tagp = (__be16 *)blp - 1; + + /* Data object just before the first leaf entry. */ + dup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp)); + + /* + * If it's not free then the data will go where the + * leaf data starts now, if it works at all. + */ + if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) { + if (be16_to_cpu(dup->length) + (be32_to_cpu(btp->stale) - 1) * + (uint)sizeof(*blp) < len) + dup = NULL; + } else if ((be32_to_cpu(btp->stale) - 1) * (uint)sizeof(*blp) < len) + dup = NULL; + else + dup = (xfs_dir2_data_unused_t *)blp; + goto out; + } + + /* + * no stale entries, so just use free space. + * Tag just before the first leaf entry. + */ + tagp = (__be16 *)blp - 1; + + /* Data object just before the first leaf entry. */ + enddup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp)); + + /* + * If it's not free then can't do this add without cleaning up: + * the space before the first leaf entry needs to be free so it + * can be expanded to hold the pointer to the new entry. + */ + if (be16_to_cpu(enddup->freetag) == XFS_DIR2_DATA_FREE_TAG) { + /* + * Check out the biggest freespace and see if it's the same one. + */ + dup = (xfs_dir2_data_unused_t *) + ((char *)hdr + be16_to_cpu(bf[0].offset)); + if (dup != enddup) { + /* + * Not the same free entry, just check its length. + */ + if (be16_to_cpu(dup->length) < len) + dup = NULL; + goto out; + } + + /* + * It is the biggest freespace, can it hold the leaf too? + */ + if (be16_to_cpu(dup->length) < len + (uint)sizeof(*blp)) { + /* + * Yes, use the second-largest entry instead if it works. + */ + if (be16_to_cpu(bf[1].length) >= len) + dup = (xfs_dir2_data_unused_t *) + ((char *)hdr + be16_to_cpu(bf[1].offset)); + else + dup = NULL; + } + } +out: + *tagpp = tagp; + *dupp = dup; + *enddupp = enddup; +} + +/* + * compact the leaf entries. + * Leave the highest-numbered stale entry stale. + * XXX should be the one closest to mid but mid is not yet computed. + */ +static void +xfs_dir2_block_compact( + struct xfs_trans *tp, + struct xfs_buf *bp, + struct xfs_dir2_data_hdr *hdr, + struct xfs_dir2_block_tail *btp, + struct xfs_dir2_leaf_entry *blp, + int *needlog, + int *lfloghigh, + int *lfloglow) +{ + int fromidx; /* source leaf index */ + int toidx; /* target leaf index */ + int needscan = 0; + int highstale; /* high stale index */ + + fromidx = toidx = be32_to_cpu(btp->count) - 1; + highstale = *lfloghigh = -1; + for (; fromidx >= 0; fromidx--) { + if (blp[fromidx].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR)) { + if (highstale == -1) + highstale = toidx; + else { + if (*lfloghigh == -1) + *lfloghigh = toidx; + continue; + } + } + if (fromidx < toidx) + blp[toidx] = blp[fromidx]; + toidx--; + } + *lfloglow = toidx + 1 - (be32_to_cpu(btp->stale) - 1); + *lfloghigh -= be32_to_cpu(btp->stale) - 1; + be32_add_cpu(&btp->count, -(be32_to_cpu(btp->stale) - 1)); + xfs_dir2_data_make_free(tp, bp, + (xfs_dir2_data_aoff_t)((char *)blp - (char *)hdr), + (xfs_dir2_data_aoff_t)((be32_to_cpu(btp->stale) - 1) * sizeof(*blp)), + needlog, &needscan); + blp += be32_to_cpu(btp->stale) - 1; + btp->stale = cpu_to_be32(1); + /* + * If we now need to rebuild the bestfree map, do so. + * This needs to happen before the next call to use_free. + */ + if (needscan) + xfs_dir2_data_freescan(tp->t_mountp, hdr, needlog); +} + /* * Add an entry to a block directory. */ @@ -63,7 +271,6 @@ int /* error */ xfs_dir2_block_addname( xfs_da_args_t *args) /* directory op arguments */ { - xfs_dir2_data_free_t *bf; /* bestfree table in block */ xfs_dir2_data_hdr_t *hdr; /* block header */ xfs_dir2_leaf_entry_t *blp; /* block leaf entries */ struct xfs_buf *bp; /* buffer for block */ @@ -94,134 +301,44 @@ xfs_dir2_block_addname( dp = args->dp; tp = args->trans; mp = dp->i_mount; - /* - * Read the (one and only) directory block into dabuf bp. - */ - if ((error = - xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, &bp, XFS_DATA_FORK))) { + + /* Read the (one and only) directory block into bp. */ + error = xfs_dir2_block_read(tp, dp, &bp); + if (error) return error; - } - ASSERT(bp != NULL); - hdr = bp->b_addr; - /* - * Check the magic number, corrupted if wrong. - */ - if (unlikely(hdr->magic != cpu_to_be32(XFS_DIR2_BLOCK_MAGIC))) { - XFS_CORRUPTION_ERROR("xfs_dir2_block_addname", - XFS_ERRLEVEL_LOW, mp, hdr); - xfs_trans_brelse(tp, bp); - return XFS_ERROR(EFSCORRUPTED); - } + len = xfs_dir2_data_entsize(args->namelen); + /* * Set up pointers to parts of the block. */ - bf = hdr->bestfree; + hdr = bp->b_addr; btp = xfs_dir2_block_tail_p(mp, hdr); blp = xfs_dir2_block_leaf_p(btp); + /* - * No stale entries? Need space for entry and new leaf. - */ - if (!btp->stale) { - /* - * Tag just before the first leaf entry. - */ - tagp = (__be16 *)blp - 1; - /* - * Data object just before the first leaf entry. - */ - enddup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp)); - /* - * If it's not free then can't do this add without cleaning up: - * the space before the first leaf entry needs to be free so it - * can be expanded to hold the pointer to the new entry. - */ - if (be16_to_cpu(enddup->freetag) != XFS_DIR2_DATA_FREE_TAG) - dup = enddup = NULL; - /* - * Check out the biggest freespace and see if it's the same one. - */ - else { - dup = (xfs_dir2_data_unused_t *) - ((char *)hdr + be16_to_cpu(bf[0].offset)); - if (dup == enddup) { - /* - * It is the biggest freespace, is it too small - * to hold the new leaf too? - */ - if (be16_to_cpu(dup->length) < len + (uint)sizeof(*blp)) { - /* - * Yes, we use the second-largest - * entry instead if it works. - */ - if (be16_to_cpu(bf[1].length) >= len) - dup = (xfs_dir2_data_unused_t *) - ((char *)hdr + - be16_to_cpu(bf[1].offset)); - else - dup = NULL; - } - } else { - /* - * Not the same free entry, - * just check its length. - */ - if (be16_to_cpu(dup->length) < len) { - dup = NULL; - } - } - } - compact = 0; - } - /* - * If there are stale entries we'll use one for the leaf. - * Is the biggest entry enough to avoid compaction? - */ - else if (be16_to_cpu(bf[0].length) >= len) { - dup = (xfs_dir2_data_unused_t *) - ((char *)hdr + be16_to_cpu(bf[0].offset)); - compact = 0; - } - /* - * Will need to compact to make this work. + * Find out if we can reuse stale entries or whether we need extra + * space for entry and new leaf. */ - else { - /* - * Tag just before the first leaf entry. - */ - tagp = (__be16 *)blp - 1; - /* - * Data object just before the first leaf entry. - */ - dup = (xfs_dir2_data_unused_t *)((char *)hdr + be16_to_cpu(*tagp)); - /* - * If it's not free then the data will go where the - * leaf data starts now, if it works at all. - */ - if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) { - if (be16_to_cpu(dup->length) + (be32_to_cpu(btp->stale) - 1) * - (uint)sizeof(*blp) < len) - dup = NULL; - } else if ((be32_to_cpu(btp->stale) - 1) * (uint)sizeof(*blp) < len) - dup = NULL; - else - dup = (xfs_dir2_data_unused_t *)blp; - compact = 1; - } + xfs_dir2_block_need_space(hdr, btp, blp, &tagp, &dup, + &enddup, &compact, len); + /* - * If this isn't a real add, we're done with the buffer. + * Done everything we need for a space check now. */ - if (args->op_flags & XFS_DA_OP_JUSTCHECK) + if (args->op_flags & XFS_DA_OP_JUSTCHECK) { xfs_trans_brelse(tp, bp); + if (!dup) + return XFS_ERROR(ENOSPC); + return 0; + } + /* * If we don't have space for the new entry & leaf ... */ if (!dup) { - /* - * Not trying to actually do anything, or don't have - * a space reservation: return no-space. - */ - if ((args->op_flags & XFS_DA_OP_JUSTCHECK) || args->total == 0) + /* Don't have a space reservation: return no-space. */ + if (args->total == 0) return XFS_ERROR(ENOSPC); /* * Convert to the next larger format. @@ -232,65 +349,24 @@ xfs_dir2_block_addname( return error; return xfs_dir2_leaf_addname(args); } - /* - * Just checking, and it would work, so say so. - */ - if (args->op_flags & XFS_DA_OP_JUSTCHECK) - return 0; + needlog = needscan = 0; + /* * If need to compact the leaf entries, do it now. - * Leave the highest-numbered stale entry stale. - * XXX should be the one closest to mid but mid is not yet computed. - */ - if (compact) { - int fromidx; /* source leaf index */ - int toidx; /* target leaf index */ - - for (fromidx = toidx = be32_to_cpu(btp->count) - 1, - highstale = lfloghigh = -1; - fromidx >= 0; - fromidx--) { - if (blp[fromidx].address == - cpu_to_be32(XFS_DIR2_NULL_DATAPTR)) { - if (highstale == -1) - highstale = toidx; - else { - if (lfloghigh == -1) - lfloghigh = toidx; - continue; - } - } - if (fromidx < toidx) - blp[toidx] = blp[fromidx]; - toidx--; - } - lfloglow = toidx + 1 - (be32_to_cpu(btp->stale) - 1); - lfloghigh -= be32_to_cpu(btp->stale) - 1; - be32_add_cpu(&btp->count, -(be32_to_cpu(btp->stale) - 1)); - xfs_dir2_data_make_free(tp, bp, - (xfs_dir2_data_aoff_t)((char *)blp - (char *)hdr), - (xfs_dir2_data_aoff_t)((be32_to_cpu(btp->stale) - 1) * sizeof(*blp)), - &needlog, &needscan); - blp += be32_to_cpu(btp->stale) - 1; - btp->stale = cpu_to_be32(1); - /* - * If we now need to rebuild the bestfree map, do so. - * This needs to happen before the next call to use_free. - */ - if (needscan) { - xfs_dir2_data_freescan(mp, hdr, &needlog); - needscan = 0; - } - } - /* - * Set leaf logging boundaries to impossible state. - * For the no-stale case they're set explicitly. */ + if (compact) + xfs_dir2_block_compact(tp, bp, hdr, btp, blp, &needlog, + &lfloghigh, &lfloglow); else if (btp->stale) { + /* + * Set leaf logging boundaries to impossible state. + * For the no-stale case they're set explicitly. + */ lfloglow = be32_to_cpu(btp->count); lfloghigh = -1; } + /* * Find the slot that's first lower than our hash value, -1 if none. */ @@ -450,18 +526,13 @@ xfs_dir2_block_getdents( /* * If the block number in the offset is out of range, we're done. */ - if (xfs_dir2_dataptr_to_db(mp, *offset) > mp->m_dirdatablk) { + if (xfs_dir2_dataptr_to_db(mp, *offset) > mp->m_dirdatablk) return 0; - } - /* - * Can't read the block, give up, else get dabuf in bp. - */ - error = xfs_da_read_buf(NULL, dp, mp->m_dirdatablk, -1, - &bp, XFS_DATA_FORK); + + error = xfs_dir2_block_read(NULL, dp, &bp); if (error) return error; - ASSERT(bp != NULL); /* * Extract the byte offset we start at from the seek pointer. * We'll skip entries before this. @@ -637,14 +708,11 @@ xfs_dir2_block_lookup_int( dp = args->dp; tp = args->trans; mp = dp->i_mount; - /* - * Read the buffer, return error if we can't get it. - */ - if ((error = - xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, &bp, XFS_DATA_FORK))) { + + error = xfs_dir2_block_read(tp, dp, &bp); + if (error) return error; - } - ASSERT(bp != NULL); + hdr = bp->b_addr; xfs_dir2_data_check(dp, bp); btp = xfs_dir2_block_tail_p(mp, hdr); @@ -917,10 +985,10 @@ xfs_dir2_leaf_to_block( /* * Read the data block if we don't already have it, give up if it fails. */ - if (dbp == NULL && - (error = xfs_da_read_buf(tp, dp, mp->m_dirdatablk, -1, &dbp, - XFS_DATA_FORK))) { - return error; + if (!dbp) { + error = xfs_dir2_data_read(tp, dp, mp->m_dirdatablk, -1, &dbp); + if (error) + return error; } hdr = dbp->b_addr; ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC)); @@ -944,6 +1012,7 @@ xfs_dir2_leaf_to_block( /* * Start converting it to block form. */ + dbp->b_ops = &xfs_dir2_block_buf_ops; hdr->magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC); needlog = 1; needscan = 0; @@ -1073,6 +1142,7 @@ xfs_dir2_sf_to_block( kmem_free(sfp); return error; } + bp->b_ops = &xfs_dir2_block_buf_ops; hdr = bp->b_addr; hdr->magic = cpu_to_be32(XFS_DIR2_BLOCK_MAGIC); /* diff --git a/fs/xfs/xfs_dir2_data.c b/fs/xfs/xfs_dir2_data.c index 44ffd4d6bc91..ffcf1774152e 100644 --- a/fs/xfs/xfs_dir2_data.c +++ b/fs/xfs/xfs_dir2_data.c @@ -34,14 +34,13 @@ STATIC xfs_dir2_data_free_t * xfs_dir2_data_freefind(xfs_dir2_data_hdr_t *hdr, xfs_dir2_data_unused_t *dup); -#ifdef DEBUG /* * Check the consistency of the data block. * The input can also be a block-format directory. - * Pop an assert if we find anything bad. + * Return 0 is the buffer is good, otherwise an error. */ -void -xfs_dir2_data_check( +int +__xfs_dir2_data_check( struct xfs_inode *dp, /* incore inode pointer */ struct xfs_buf *bp) /* data block's buffer */ { @@ -64,18 +63,23 @@ xfs_dir2_data_check( int stale; /* count of stale leaves */ struct xfs_name name; - mp = dp->i_mount; + mp = bp->b_target->bt_mount; hdr = bp->b_addr; bf = hdr->bestfree; p = (char *)(hdr + 1); - if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) { + switch (hdr->magic) { + case cpu_to_be32(XFS_DIR2_BLOCK_MAGIC): btp = xfs_dir2_block_tail_p(mp, hdr); lep = xfs_dir2_block_leaf_p(btp); endp = (char *)lep; - } else { - ASSERT(hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC)); + break; + case cpu_to_be32(XFS_DIR2_DATA_MAGIC): endp = (char *)hdr + mp->m_dirblksize; + break; + default: + XFS_ERROR_REPORT("Bad Magic", XFS_ERRLEVEL_LOW, mp); + return EFSCORRUPTED; } count = lastfree = freeseen = 0; @@ -83,19 +87,22 @@ xfs_dir2_data_check( * Account for zero bestfree entries. */ if (!bf[0].length) { - ASSERT(!bf[0].offset); + XFS_WANT_CORRUPTED_RETURN(!bf[0].offset); freeseen |= 1 << 0; } if (!bf[1].length) { - ASSERT(!bf[1].offset); + XFS_WANT_CORRUPTED_RETURN(!bf[1].offset); freeseen |= 1 << 1; } if (!bf[2].length) { - ASSERT(!bf[2].offset); + XFS_WANT_CORRUPTED_RETURN(!bf[2].offset); freeseen |= 1 << 2; } - ASSERT(be16_to_cpu(bf[0].length) >= be16_to_cpu(bf[1].length)); - ASSERT(be16_to_cpu(bf[1].length) >= be16_to_cpu(bf[2].length)); + + XFS_WANT_CORRUPTED_RETURN(be16_to_cpu(bf[0].length) >= + be16_to_cpu(bf[1].length)); + XFS_WANT_CORRUPTED_RETURN(be16_to_cpu(bf[1].length) >= + be16_to_cpu(bf[2].length)); /* * Loop over the data/unused entries. */ @@ -107,17 +114,20 @@ xfs_dir2_data_check( * doesn't need to be there. */ if (be16_to_cpu(dup->freetag) == XFS_DIR2_DATA_FREE_TAG) { - ASSERT(lastfree == 0); - ASSERT(be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)) == - (char *)dup - (char *)hdr); + XFS_WANT_CORRUPTED_RETURN(lastfree == 0); + XFS_WANT_CORRUPTED_RETURN( + be16_to_cpu(*xfs_dir2_data_unused_tag_p(dup)) == + (char *)dup - (char *)hdr); dfp = xfs_dir2_data_freefind(hdr, dup); if (dfp) { i = (int)(dfp - bf); - ASSERT((freeseen & (1 << i)) == 0); + XFS_WANT_CORRUPTED_RETURN( + (freeseen & (1 << i)) == 0); freeseen |= 1 << i; } else { - ASSERT(be16_to_cpu(dup->length) <= - be16_to_cpu(bf[2].length)); + XFS_WANT_CORRUPTED_RETURN( + be16_to_cpu(dup->length) <= + be16_to_cpu(bf[2].length)); } p += be16_to_cpu(dup->length); lastfree = 1; @@ -130,10 +140,12 @@ xfs_dir2_data_check( * The linear search is crude but this is DEBUG code. */ dep = (xfs_dir2_data_entry_t *)p; - ASSERT(dep->namelen != 0); - ASSERT(xfs_dir_ino_validate(mp, be64_to_cpu(dep->inumber)) == 0); - ASSERT(be16_to_cpu(*xfs_dir2_data_entry_tag_p(dep)) == - (char *)dep - (char *)hdr); + XFS_WANT_CORRUPTED_RETURN(dep->namelen != 0); + XFS_WANT_CORRUPTED_RETURN( + !xfs_dir_ino_validate(mp, be64_to_cpu(dep->inumber))); + XFS_WANT_CORRUPTED_RETURN( + be16_to_cpu(*xfs_dir2_data_entry_tag_p(dep)) == + (char *)dep - (char *)hdr); count++; lastfree = 0; if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) { @@ -148,27 +160,122 @@ xfs_dir2_data_check( be32_to_cpu(lep[i].hashval) == hash) break; } - ASSERT(i < be32_to_cpu(btp->count)); + XFS_WANT_CORRUPTED_RETURN(i < be32_to_cpu(btp->count)); } p += xfs_dir2_data_entsize(dep->namelen); } /* * Need to have seen all the entries and all the bestfree slots. */ - ASSERT(freeseen == 7); + XFS_WANT_CORRUPTED_RETURN(freeseen == 7); if (hdr->magic == cpu_to_be32(XFS_DIR2_BLOCK_MAGIC)) { for (i = stale = 0; i < be32_to_cpu(btp->count); i++) { if (lep[i].address == cpu_to_be32(XFS_DIR2_NULL_DATAPTR)) stale++; if (i > 0) - ASSERT(be32_to_cpu(lep[i].hashval) >= be32_to_cpu(lep[i - 1].hashval)); + XFS_WANT_CORRUPTED_RETURN( + be32_to_cpu(lep[i].hashval) >= + be32_to_cpu(lep[i - 1].hashval)); } - ASSERT(count == be32_to_cpu(btp->count) - be32_to_cpu(btp->stale)); - ASSERT(stale == be32_to_cpu(btp->stale)); + XFS_WANT_CORRUPTED_RETURN(count == + be32_to_cpu(btp->count) - be32_to_cpu(btp->stale)); + XFS_WANT_CORRUPTED_RETURN(stale == be32_to_cpu(btp->stale)); } + return 0; +} + +static void +xfs_dir2_data_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_dir2_data_hdr *hdr = bp->b_addr; + int block_ok = 0; + + block_ok = hdr->magic == cpu_to_be32(XFS_DIR2_DATA_MAGIC); + block_ok = block_ok && __xfs_dir2_data_check(NULL, bp) == 0; + + if (!block_ok) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +/* + * Readahead of the first block of the directory when it is opened is completely + * oblivious to the format of the directory. Hence we can either get a block + * format buffer or a data format buffer on readahead. + */ +static void +xfs_dir2_data_reada_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_dir2_data_hdr *hdr = bp->b_addr; + + switch (hdr->magic) { + case cpu_to_be32(XFS_DIR2_BLOCK_MAGIC): + bp->b_ops = &xfs_dir2_block_buf_ops; + bp->b_ops->verify_read(bp); + return; + case cpu_to_be32(XFS_DIR2_DATA_MAGIC): + xfs_dir2_data_verify(bp); + return; + default: + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr); + xfs_buf_ioerror(bp, EFSCORRUPTED); + break; + } +} + +static void +xfs_dir2_data_read_verify( + struct xfs_buf *bp) +{ + xfs_dir2_data_verify(bp); +} + +static void +xfs_dir2_data_write_verify( + struct xfs_buf *bp) +{ + xfs_dir2_data_verify(bp); +} + +const struct xfs_buf_ops xfs_dir2_data_buf_ops = { + .verify_read = xfs_dir2_data_read_verify, + .verify_write = xfs_dir2_data_write_verify, +}; + +static const struct xfs_buf_ops xfs_dir2_data_reada_buf_ops = { + .verify_read = xfs_dir2_data_reada_verify, + .verify_write = xfs_dir2_data_write_verify, +}; + + +int +xfs_dir2_data_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t bno, + xfs_daddr_t mapped_bno, + struct xfs_buf **bpp) +{ + return xfs_da_read_buf(tp, dp, bno, mapped_bno, bpp, + XFS_DATA_FORK, &xfs_dir2_data_buf_ops); +} + +int +xfs_dir2_data_readahead( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t bno, + xfs_daddr_t mapped_bno) +{ + return xfs_da_reada_buf(tp, dp, bno, mapped_bno, + XFS_DATA_FORK, &xfs_dir2_data_reada_buf_ops); } -#endif /* * Given a data block and an unused entry from that block, @@ -409,10 +516,9 @@ xfs_dir2_data_init( */ error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, blkno), -1, &bp, XFS_DATA_FORK); - if (error) { + if (error) return error; - } - ASSERT(bp != NULL); + bp->b_ops = &xfs_dir2_data_buf_ops; /* * Initialize the header. diff --git a/fs/xfs/xfs_dir2_leaf.c b/fs/xfs/xfs_dir2_leaf.c index 0b296253bd01..60cd2fa4e047 100644 --- a/fs/xfs/xfs_dir2_leaf.c +++ b/fs/xfs/xfs_dir2_leaf.c @@ -48,6 +48,83 @@ static void xfs_dir2_leaf_log_bests(struct xfs_trans *tp, struct xfs_buf *bp, int first, int last); static void xfs_dir2_leaf_log_tail(struct xfs_trans *tp, struct xfs_buf *bp); +static void +xfs_dir2_leaf_verify( + struct xfs_buf *bp, + __be16 magic) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_dir2_leaf_hdr *hdr = bp->b_addr; + int block_ok = 0; + + block_ok = hdr->info.magic == magic; + if (!block_ok) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, hdr); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_dir2_leaf1_read_verify( + struct xfs_buf *bp) +{ + xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAF1_MAGIC)); +} + +static void +xfs_dir2_leaf1_write_verify( + struct xfs_buf *bp) +{ + xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAF1_MAGIC)); +} + +void +xfs_dir2_leafn_read_verify( + struct xfs_buf *bp) +{ + xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)); +} + +void +xfs_dir2_leafn_write_verify( + struct xfs_buf *bp) +{ + xfs_dir2_leaf_verify(bp, cpu_to_be16(XFS_DIR2_LEAFN_MAGIC)); +} + +static const struct xfs_buf_ops xfs_dir2_leaf1_buf_ops = { + .verify_read = xfs_dir2_leaf1_read_verify, + .verify_write = xfs_dir2_leaf1_write_verify, +}; + +const struct xfs_buf_ops xfs_dir2_leafn_buf_ops = { + .verify_read = xfs_dir2_leafn_read_verify, + .verify_write = xfs_dir2_leafn_write_verify, +}; + +static int +xfs_dir2_leaf_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t fbno, + xfs_daddr_t mappedbno, + struct xfs_buf **bpp) +{ + return xfs_da_read_buf(tp, dp, fbno, mappedbno, bpp, + XFS_DATA_FORK, &xfs_dir2_leaf1_buf_ops); +} + +int +xfs_dir2_leafn_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t fbno, + xfs_daddr_t mappedbno, + struct xfs_buf **bpp) +{ + return xfs_da_read_buf(tp, dp, fbno, mappedbno, bpp, + XFS_DATA_FORK, &xfs_dir2_leafn_buf_ops); +} /* * Convert a block form directory to a leaf form directory. @@ -125,6 +202,7 @@ xfs_dir2_block_to_leaf( /* * Fix up the block header, make it a data block. */ + dbp->b_ops = &xfs_dir2_data_buf_ops; hdr->magic = cpu_to_be32(XFS_DIR2_DATA_MAGIC); if (needscan) xfs_dir2_data_freescan(mp, hdr, &needlog); @@ -311,15 +389,11 @@ xfs_dir2_leaf_addname( dp = args->dp; tp = args->trans; mp = dp->i_mount; - /* - * Read the leaf block. - */ - error = xfs_da_read_buf(tp, dp, mp->m_dirleafblk, -1, &lbp, - XFS_DATA_FORK); - if (error) { + + error = xfs_dir2_leaf_read(tp, dp, mp->m_dirleafblk, -1, &lbp); + if (error) return error; - } - ASSERT(lbp != NULL); + /* * Look up the entry by hash value and name. * We know it's not there, our caller has already done a lookup. @@ -494,22 +568,21 @@ xfs_dir2_leaf_addname( hdr = dbp->b_addr; bestsp[use_block] = hdr->bestfree[0].length; grown = 1; - } - /* - * Already had space in some data block. - * Just read that one in. - */ - else { - if ((error = - xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, use_block), - -1, &dbp, XFS_DATA_FORK))) { + } else { + /* + * Already had space in some data block. + * Just read that one in. + */ + error = xfs_dir2_data_read(tp, dp, + xfs_dir2_db_to_da(mp, use_block), + -1, &dbp); + if (error) { xfs_trans_brelse(tp, lbp); return error; } hdr = dbp->b_addr; grown = 0; } - xfs_dir2_data_check(dp, dbp); /* * Point to the biggest freespace in our data block. */ @@ -892,10 +965,9 @@ xfs_dir2_leaf_readbuf( * Read the directory block starting at the first mapping. */ mip->curdb = xfs_dir2_da_to_db(mp, map->br_startoff); - error = xfs_da_read_buf(NULL, dp, map->br_startoff, + error = xfs_dir2_data_read(NULL, dp, map->br_startoff, map->br_blockcount >= mp->m_dirblkfsbs ? - XFS_FSB_TO_DADDR(mp, map->br_startblock) : -1, - &bp, XFS_DATA_FORK); + XFS_FSB_TO_DADDR(mp, map->br_startblock) : -1, &bp); /* * Should just skip over the data block instead of giving up. @@ -922,11 +994,11 @@ xfs_dir2_leaf_readbuf( */ if (i > mip->ra_current && map[mip->ra_index].br_blockcount >= mp->m_dirblkfsbs) { - xfs_buf_readahead(mp->m_ddev_targp, + xfs_dir2_data_readahead(NULL, dp, + map[mip->ra_index].br_startoff + mip->ra_offset, XFS_FSB_TO_DADDR(mp, map[mip->ra_index].br_startblock + - mip->ra_offset), - (int)BTOBB(mp->m_dirblksize)); + mip->ra_offset)); mip->ra_current = i; } @@ -935,10 +1007,9 @@ xfs_dir2_leaf_readbuf( * use our mapping, but this is a very rare case. */ else if (i > mip->ra_current) { - xfs_da_reada_buf(NULL, dp, + xfs_dir2_data_readahead(NULL, dp, map[mip->ra_index].br_startoff + - mip->ra_offset, - XFS_DATA_FORK); + mip->ra_offset, -1); mip->ra_current = i; } @@ -1177,15 +1248,14 @@ xfs_dir2_leaf_init( * Get the buffer for the block. */ error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, bno), -1, &bp, - XFS_DATA_FORK); - if (error) { + XFS_DATA_FORK); + if (error) return error; - } - ASSERT(bp != NULL); - leaf = bp->b_addr; + /* * Initialize the header. */ + leaf = bp->b_addr; leaf->hdr.info.magic = cpu_to_be16(magic); leaf->hdr.info.forw = 0; leaf->hdr.info.back = 0; @@ -1198,10 +1268,12 @@ xfs_dir2_leaf_init( * the block. */ if (magic == XFS_DIR2_LEAF1_MAGIC) { + bp->b_ops = &xfs_dir2_leaf1_buf_ops; ltp = xfs_dir2_leaf_tail_p(mp, leaf); ltp->bestcount = 0; xfs_dir2_leaf_log_tail(tp, bp); - } + } else + bp->b_ops = &xfs_dir2_leafn_buf_ops; *bpp = bp; return 0; } @@ -1372,13 +1444,11 @@ xfs_dir2_leaf_lookup_int( dp = args->dp; tp = args->trans; mp = dp->i_mount; - /* - * Read the leaf block into the buffer. - */ - error = xfs_da_read_buf(tp, dp, mp->m_dirleafblk, -1, &lbp, - XFS_DATA_FORK); + + error = xfs_dir2_leaf_read(tp, dp, mp->m_dirleafblk, -1, &lbp); if (error) return error; + *lbpp = lbp; leaf = lbp->b_addr; xfs_dir2_leaf_check(dp, lbp); @@ -1409,14 +1479,13 @@ xfs_dir2_leaf_lookup_int( if (newdb != curdb) { if (dbp) xfs_trans_brelse(tp, dbp); - error = xfs_da_read_buf(tp, dp, - xfs_dir2_db_to_da(mp, newdb), - -1, &dbp, XFS_DATA_FORK); + error = xfs_dir2_data_read(tp, dp, + xfs_dir2_db_to_da(mp, newdb), + -1, &dbp); if (error) { xfs_trans_brelse(tp, lbp); return error; } - xfs_dir2_data_check(dp, dbp); curdb = newdb; } /* @@ -1451,9 +1520,9 @@ xfs_dir2_leaf_lookup_int( ASSERT(cidb != -1); if (cidb != curdb) { xfs_trans_brelse(tp, dbp); - error = xfs_da_read_buf(tp, dp, - xfs_dir2_db_to_da(mp, cidb), - -1, &dbp, XFS_DATA_FORK); + error = xfs_dir2_data_read(tp, dp, + xfs_dir2_db_to_da(mp, cidb), + -1, &dbp); if (error) { xfs_trans_brelse(tp, lbp); return error; @@ -1738,10 +1807,9 @@ xfs_dir2_leaf_trim_data( /* * Read the offending data block. We need its buffer. */ - if ((error = xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, db), -1, &dbp, - XFS_DATA_FORK))) { + error = xfs_dir2_data_read(tp, dp, xfs_dir2_db_to_da(mp, db), -1, &dbp); + if (error) return error; - } leaf = lbp->b_addr; ltp = xfs_dir2_leaf_tail_p(mp, leaf); @@ -1864,10 +1932,9 @@ xfs_dir2_node_to_leaf( /* * Read the freespace block. */ - if ((error = xfs_da_read_buf(tp, dp, mp->m_dirfreeblk, -1, &fbp, - XFS_DATA_FORK))) { + error = xfs_dir2_free_read(tp, dp, mp->m_dirfreeblk, &fbp); + if (error) return error; - } free = fbp->b_addr; ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC)); ASSERT(!free->hdr.firstdb); @@ -1890,7 +1957,10 @@ xfs_dir2_node_to_leaf( xfs_dir2_leaf_compact(args, lbp); else xfs_dir2_leaf_log_header(tp, lbp); + + lbp->b_ops = &xfs_dir2_leaf1_buf_ops; leaf->hdr.info.magic = cpu_to_be16(XFS_DIR2_LEAF1_MAGIC); + /* * Set up the leaf tail from the freespace block. */ diff --git a/fs/xfs/xfs_dir2_node.c b/fs/xfs/xfs_dir2_node.c index 6c7052406605..5980f9b7fa9b 100644 --- a/fs/xfs/xfs_dir2_node.c +++ b/fs/xfs/xfs_dir2_node.c @@ -55,6 +55,74 @@ static int xfs_dir2_leafn_remove(xfs_da_args_t *args, struct xfs_buf *bp, static int xfs_dir2_node_addname_int(xfs_da_args_t *args, xfs_da_state_blk_t *fblk); +static void +xfs_dir2_free_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_dir2_free_hdr *hdr = bp->b_addr; + int block_ok = 0; + + block_ok = hdr->magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC); + if (!block_ok) { + XFS_CORRUPTION_ERROR("xfs_dir2_free_verify magic", + XFS_ERRLEVEL_LOW, mp, hdr); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_dir2_free_read_verify( + struct xfs_buf *bp) +{ + xfs_dir2_free_verify(bp); +} + +static void +xfs_dir2_free_write_verify( + struct xfs_buf *bp) +{ + xfs_dir2_free_verify(bp); +} + +static const struct xfs_buf_ops xfs_dir2_free_buf_ops = { + .verify_read = xfs_dir2_free_read_verify, + .verify_write = xfs_dir2_free_write_verify, +}; + + +static int +__xfs_dir2_free_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t fbno, + xfs_daddr_t mappedbno, + struct xfs_buf **bpp) +{ + return xfs_da_read_buf(tp, dp, fbno, mappedbno, bpp, + XFS_DATA_FORK, &xfs_dir2_free_buf_ops); +} + +int +xfs_dir2_free_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t fbno, + struct xfs_buf **bpp) +{ + return __xfs_dir2_free_read(tp, dp, fbno, -1, bpp); +} + +static int +xfs_dir2_free_try_read( + struct xfs_trans *tp, + struct xfs_inode *dp, + xfs_dablk_t fbno, + struct xfs_buf **bpp) +{ + return __xfs_dir2_free_read(tp, dp, fbno, -2, bpp); +} + /* * Log entries from a freespace block. */ @@ -131,11 +199,12 @@ xfs_dir2_leaf_to_node( /* * Get the buffer for the new freespace block. */ - if ((error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, fdb), -1, &fbp, - XFS_DATA_FORK))) { + error = xfs_da_get_buf(tp, dp, xfs_dir2_db_to_da(mp, fdb), -1, &fbp, + XFS_DATA_FORK); + if (error) return error; - } - ASSERT(fbp != NULL); + fbp->b_ops = &xfs_dir2_free_buf_ops; + free = fbp->b_addr; leaf = lbp->b_addr; ltp = xfs_dir2_leaf_tail_p(mp, leaf); @@ -157,7 +226,10 @@ xfs_dir2_leaf_to_node( *to = cpu_to_be16(off); } free->hdr.nused = cpu_to_be32(n); + + lbp->b_ops = &xfs_dir2_leafn_buf_ops; leaf->hdr.info.magic = cpu_to_be16(XFS_DIR2_LEAFN_MAGIC); + /* * Log everything. */ @@ -394,12 +466,10 @@ xfs_dir2_leafn_lookup_for_addname( */ if (curbp) xfs_trans_brelse(tp, curbp); - /* - * Read the free block. - */ - error = xfs_da_read_buf(tp, dp, + + error = xfs_dir2_free_read(tp, dp, xfs_dir2_db_to_da(mp, newfdb), - -1, &curbp, XFS_DATA_FORK); + &curbp); if (error) return error; free = curbp->b_addr; @@ -534,9 +604,9 @@ xfs_dir2_leafn_lookup_for_entry( ASSERT(state->extravalid); curbp = state->extrablk.bp; } else { - error = xfs_da_read_buf(tp, dp, + error = xfs_dir2_data_read(tp, dp, xfs_dir2_db_to_da(mp, newdb), - -1, &curbp, XFS_DATA_FORK); + -1, &curbp); if (error) return error; } @@ -568,6 +638,7 @@ xfs_dir2_leafn_lookup_for_entry( state->extrablk.index = (int)((char *)dep - (char *)curbp->b_addr); state->extrablk.magic = XFS_DIR2_DATA_MAGIC; + curbp->b_ops = &xfs_dir2_data_buf_ops; if (cmp == XFS_CMP_EXACT) return XFS_ERROR(EEXIST); } @@ -582,6 +653,7 @@ xfs_dir2_leafn_lookup_for_entry( state->extrablk.index = -1; state->extrablk.blkno = curdb; state->extrablk.magic = XFS_DIR2_DATA_MAGIC; + curbp->b_ops = &xfs_dir2_data_buf_ops; } else { /* If the curbp is not the CI match block, drop it */ if (state->extrablk.bp != curbp) @@ -825,6 +897,77 @@ xfs_dir2_leafn_rebalance( } } +static int +xfs_dir2_data_block_free( + xfs_da_args_t *args, + struct xfs_dir2_data_hdr *hdr, + struct xfs_dir2_free *free, + xfs_dir2_db_t fdb, + int findex, + struct xfs_buf *fbp, + int longest) +{ + struct xfs_trans *tp = args->trans; + int logfree = 0; + + if (!hdr) { + /* One less used entry in the free table. */ + be32_add_cpu(&free->hdr.nused, -1); + xfs_dir2_free_log_header(tp, fbp); + + /* + * If this was the last entry in the table, we can trim the + * table size back. There might be other entries at the end + * referring to non-existent data blocks, get those too. + */ + if (findex == be32_to_cpu(free->hdr.nvalid) - 1) { + int i; /* free entry index */ + + for (i = findex - 1; i >= 0; i--) { + if (free->bests[i] != cpu_to_be16(NULLDATAOFF)) + break; + } + free->hdr.nvalid = cpu_to_be32(i + 1); + logfree = 0; + } else { + /* Not the last entry, just punch it out. */ + free->bests[findex] = cpu_to_be16(NULLDATAOFF); + logfree = 1; + } + /* + * If there are no useful entries left in the block, + * get rid of the block if we can. + */ + if (!free->hdr.nused) { + int error; + + error = xfs_dir2_shrink_inode(args, fdb, fbp); + if (error == 0) { + fbp = NULL; + logfree = 0; + } else if (error != ENOSPC || args->total != 0) + return error; + /* + * It's possible to get ENOSPC if there is no + * space reservation. In this case some one + * else will eventually get rid of this block. + */ + } + } else { + /* + * Data block is not empty, just set the free entry to the new + * value. + */ + free->bests[findex] = cpu_to_be16(longest); + logfree = 1; + } + + /* Log the free entry that changed, unless we got rid of it. */ + if (logfree) + xfs_dir2_free_log_bests(tp, fbp, findex, findex); + return 0; +} + /* * Remove an entry from a node directory. * This removes the leaf entry and the data entry, @@ -908,17 +1051,16 @@ xfs_dir2_leafn_remove( xfs_dir2_db_t fdb; /* freeblock block number */ int findex; /* index in freeblock entries */ xfs_dir2_free_t *free; /* freeblock structure */ - int logfree; /* need to log free entry */ /* * Convert the data block number to a free block, * read in the free block. */ fdb = xfs_dir2_db_to_fdb(mp, db); - if ((error = xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, fdb), - -1, &fbp, XFS_DATA_FORK))) { + error = xfs_dir2_free_read(tp, dp, xfs_dir2_db_to_da(mp, fdb), + &fbp); + if (error) return error; - } free = fbp->b_addr; ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC)); ASSERT(be32_to_cpu(free->hdr.firstdb) == @@ -954,68 +1096,12 @@ xfs_dir2_leafn_remove( * If we got rid of the data block, we can eliminate that entry * in the free block. */ - if (hdr == NULL) { - /* - * One less used entry in the free table. - */ - be32_add_cpu(&free->hdr.nused, -1); - xfs_dir2_free_log_header(tp, fbp); - /* - * If this was the last entry in the table, we can - * trim the table size back. There might be other - * entries at the end referring to non-existent - * data blocks, get those too. - */ - if (findex == be32_to_cpu(free->hdr.nvalid) - 1) { - int i; /* free entry index */ - - for (i = findex - 1; - i >= 0 && - free->bests[i] == cpu_to_be16(NULLDATAOFF); - i--) - continue; - free->hdr.nvalid = cpu_to_be32(i + 1); - logfree = 0; - } - /* - * Not the last entry, just punch it out. - */ - else { - free->bests[findex] = cpu_to_be16(NULLDATAOFF); - logfree = 1; - } - /* - * If there are no useful entries left in the block, - * get rid of the block if we can. - */ - if (!free->hdr.nused) { - error = xfs_dir2_shrink_inode(args, fdb, fbp); - if (error == 0) { - fbp = NULL; - logfree = 0; - } else if (error != ENOSPC || args->total != 0) - return error; - /* - * It's possible to get ENOSPC if there is no - * space reservation. In this case some one - * else will eventually get rid of this block. - */ - } - } - /* - * Data block is not empty, just set the free entry to - * the new value. - */ - else { - free->bests[findex] = cpu_to_be16(longest); - logfree = 1; - } - /* - * Log the free entry that changed, unless we got rid of it. - */ - if (logfree) - xfs_dir2_free_log_bests(tp, fbp, findex, findex); + error = xfs_dir2_data_block_free(args, hdr, free, + fdb, findex, fbp, longest); + if (error) + return error; } + xfs_dir2_leafn_check(dp, bp); /* * Return indication of whether this leaf block is empty enough @@ -1169,12 +1255,11 @@ xfs_dir2_leafn_toosmall( /* * Read the sibling leaf block. */ - if ((error = - xfs_da_read_buf(state->args->trans, state->args->dp, blkno, - -1, &bp, XFS_DATA_FORK))) { + error = xfs_dir2_leafn_read(state->args->trans, state->args->dp, + blkno, -1, &bp); + if (error) return error; - } - ASSERT(bp != NULL); + /* * Count bytes in the two blocks combined. */ @@ -1454,14 +1539,13 @@ xfs_dir2_node_addname_int( * This should be really rare, so there's no reason * to avoid it. */ - if ((error = xfs_da_read_buf(tp, dp, - xfs_dir2_db_to_da(mp, fbno), -2, &fbp, - XFS_DATA_FORK))) { + error = xfs_dir2_free_try_read(tp, dp, + xfs_dir2_db_to_da(mp, fbno), + &fbp); + if (error) return error; - } - if (unlikely(fbp == NULL)) { + if (!fbp) continue; - } free = fbp->b_addr; ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC)); findex = 0; @@ -1520,9 +1604,10 @@ xfs_dir2_node_addname_int( * that was just allocated. */ fbno = xfs_dir2_db_to_fdb(mp, dbno); - if (unlikely(error = xfs_da_read_buf(tp, dp, - xfs_dir2_db_to_da(mp, fbno), -2, &fbp, - XFS_DATA_FORK))) + error = xfs_dir2_free_try_read(tp, dp, + xfs_dir2_db_to_da(mp, fbno), + &fbp); + if (error) return error; /* @@ -1561,12 +1646,12 @@ xfs_dir2_node_addname_int( /* * Get a buffer for the new block. */ - if ((error = xfs_da_get_buf(tp, dp, - xfs_dir2_db_to_da(mp, fbno), - -1, &fbp, XFS_DATA_FORK))) { + error = xfs_da_get_buf(tp, dp, + xfs_dir2_db_to_da(mp, fbno), + -1, &fbp, XFS_DATA_FORK); + if (error) return error; - } - ASSERT(fbp != NULL); + fbp->b_ops = &xfs_dir2_free_buf_ops; /* * Initialize the new block to be empty, and remember @@ -1630,8 +1715,8 @@ xfs_dir2_node_addname_int( /* * Read the data block in. */ - error = xfs_da_read_buf(tp, dp, xfs_dir2_db_to_da(mp, dbno), - -1, &dbp, XFS_DATA_FORK); + error = xfs_dir2_data_read(tp, dp, xfs_dir2_db_to_da(mp, dbno), + -1, &dbp); if (error) return error; hdr = dbp->b_addr; @@ -1917,18 +2002,15 @@ xfs_dir2_node_trim_free( /* * Read the freespace block. */ - if (unlikely(error = xfs_da_read_buf(tp, dp, (xfs_dablk_t)fo, -2, &bp, - XFS_DATA_FORK))) { + error = xfs_dir2_free_try_read(tp, dp, fo, &bp); + if (error) return error; - } - /* * There can be holes in freespace. If fo is a hole, there's * nothing to do. */ - if (bp == NULL) { + if (!bp) return 0; - } free = bp->b_addr; ASSERT(free->hdr.magic == cpu_to_be32(XFS_DIR2_FREE_MAGIC)); /* diff --git a/fs/xfs/xfs_dir2_priv.h b/fs/xfs/xfs_dir2_priv.h index 3523d3e15aa8..7da79f6515fd 100644 --- a/fs/xfs/xfs_dir2_priv.h +++ b/fs/xfs/xfs_dir2_priv.h @@ -30,6 +30,8 @@ extern int xfs_dir_cilookup_result(struct xfs_da_args *args, const unsigned char *name, int len); /* xfs_dir2_block.c */ +extern const struct xfs_buf_ops xfs_dir2_block_buf_ops; + extern int xfs_dir2_block_addname(struct xfs_da_args *args); extern int xfs_dir2_block_getdents(struct xfs_inode *dp, void *dirent, xfs_off_t *offset, filldir_t filldir); @@ -41,10 +43,19 @@ extern int xfs_dir2_leaf_to_block(struct xfs_da_args *args, /* xfs_dir2_data.c */ #ifdef DEBUG -extern void xfs_dir2_data_check(struct xfs_inode *dp, struct xfs_buf *bp); +#define xfs_dir2_data_check(dp,bp) __xfs_dir2_data_check(dp, bp); #else #define xfs_dir2_data_check(dp,bp) #endif + +extern const struct xfs_buf_ops xfs_dir2_data_buf_ops; + +extern int __xfs_dir2_data_check(struct xfs_inode *dp, struct xfs_buf *bp); +extern int xfs_dir2_data_read(struct xfs_trans *tp, struct xfs_inode *dp, + xfs_dablk_t bno, xfs_daddr_t mapped_bno, struct xfs_buf **bpp); +extern int xfs_dir2_data_readahead(struct xfs_trans *tp, struct xfs_inode *dp, + xfs_dablk_t bno, xfs_daddr_t mapped_bno); + extern struct xfs_dir2_data_free * xfs_dir2_data_freeinsert(struct xfs_dir2_data_hdr *hdr, struct xfs_dir2_data_unused *dup, int *loghead); @@ -66,6 +77,10 @@ extern void xfs_dir2_data_use_free(struct xfs_trans *tp, struct xfs_buf *bp, xfs_dir2_data_aoff_t len, int *needlogp, int *needscanp); /* xfs_dir2_leaf.c */ +extern const struct xfs_buf_ops xfs_dir2_leafn_buf_ops; + +extern int xfs_dir2_leafn_read(struct xfs_trans *tp, struct xfs_inode *dp, + xfs_dablk_t fbno, xfs_daddr_t mappedbno, struct xfs_buf **bpp); extern int xfs_dir2_block_to_leaf(struct xfs_da_args *args, struct xfs_buf *dbp); extern int xfs_dir2_leaf_addname(struct xfs_da_args *args); @@ -115,6 +130,8 @@ extern int xfs_dir2_node_removename(struct xfs_da_args *args); extern int xfs_dir2_node_replace(struct xfs_da_args *args); extern int xfs_dir2_node_trim_free(struct xfs_da_args *args, xfs_fileoff_t fo, int *rvalp); +extern int xfs_dir2_free_read(struct xfs_trans *tp, struct xfs_inode *dp, + xfs_dablk_t fbno, struct xfs_buf **bpp); /* xfs_dir2_sf.c */ extern xfs_ino_t xfs_dir2_sf_get_parent_ino(struct xfs_dir2_sf_hdr *sfp); diff --git a/fs/xfs/xfs_dquot.c b/fs/xfs/xfs_dquot.c index bf27fcca4843..9e1bf5294c91 100644 --- a/fs/xfs/xfs_dquot.c +++ b/fs/xfs/xfs_dquot.c @@ -248,7 +248,59 @@ xfs_qm_init_dquot_blk( xfs_trans_log_buf(tp, bp, 0, BBTOB(q->qi_dqchunklen) - 1); } +static void +xfs_dquot_buf_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_dqblk *d = (struct xfs_dqblk *)bp->b_addr; + struct xfs_disk_dquot *ddq; + xfs_dqid_t id = 0; + int i; + + /* + * On the first read of the buffer, verify that each dquot is valid. + * We don't know what the id of the dquot is supposed to be, just that + * they should be increasing monotonically within the buffer. If the + * first id is corrupt, then it will fail on the second dquot in the + * buffer so corruptions could point to the wrong dquot in this case. + */ + for (i = 0; i < mp->m_quotainfo->qi_dqperchunk; i++) { + int error; + + ddq = &d[i].dd_diskdq; + + if (i == 0) + id = be32_to_cpu(ddq->d_id); + + error = xfs_qm_dqcheck(mp, ddq, id + i, 0, XFS_QMOPT_DOWARN, + "xfs_dquot_read_verify"); + if (error) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, d); + xfs_buf_ioerror(bp, EFSCORRUPTED); + break; + } + } +} + +static void +xfs_dquot_buf_read_verify( + struct xfs_buf *bp) +{ + xfs_dquot_buf_verify(bp); +} + +void +xfs_dquot_buf_write_verify( + struct xfs_buf *bp) +{ + xfs_dquot_buf_verify(bp); +} +const struct xfs_buf_ops xfs_dquot_buf_ops = { + .verify_read = xfs_dquot_buf_read_verify, + .verify_write = xfs_dquot_buf_write_verify, +}; /* * Allocate a block and fill it with dquots. @@ -315,6 +367,7 @@ xfs_qm_dqalloc( error = xfs_buf_geterror(bp); if (error) goto error1; + bp->b_ops = &xfs_dquot_buf_ops; /* * Make a chunk of dquots out of this buffer and log @@ -359,6 +412,51 @@ xfs_qm_dqalloc( return (error); } +STATIC int +xfs_qm_dqrepair( + struct xfs_mount *mp, + struct xfs_trans *tp, + struct xfs_dquot *dqp, + xfs_dqid_t firstid, + struct xfs_buf **bpp) +{ + int error; + struct xfs_disk_dquot *ddq; + struct xfs_dqblk *d; + int i; + + /* + * Read the buffer without verification so we get the corrupted + * buffer returned to us. make sure we verify it on write, though. + */ + error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, dqp->q_blkno, + mp->m_quotainfo->qi_dqchunklen, + 0, bpp, NULL); + + if (error) { + ASSERT(*bpp == NULL); + return XFS_ERROR(error); + } + (*bpp)->b_ops = &xfs_dquot_buf_ops; + + ASSERT(xfs_buf_islocked(*bpp)); + d = (struct xfs_dqblk *)(*bpp)->b_addr; + + /* Do the actual repair of dquots in this buffer */ + for (i = 0; i < mp->m_quotainfo->qi_dqperchunk; i++) { + ddq = &d[i].dd_diskdq; + error = xfs_qm_dqcheck(mp, ddq, firstid + i, + dqp->dq_flags & XFS_DQ_ALLTYPES, + XFS_QMOPT_DQREPAIR, "xfs_qm_dqrepair"); + if (error) { + /* repair failed, we're screwed */ + xfs_trans_brelse(tp, *bpp); + return XFS_ERROR(EIO); + } + } + + return 0; +} /* * Maps a dquot to the buffer containing its on-disk version. @@ -378,7 +476,6 @@ xfs_qm_dqtobp( xfs_buf_t *bp; xfs_inode_t *quotip = XFS_DQ_TO_QIP(dqp); xfs_mount_t *mp = dqp->q_mount; - xfs_disk_dquot_t *ddq; xfs_dqid_t id = be32_to_cpu(dqp->q_core.d_id); xfs_trans_t *tp = (tpp ? *tpp : NULL); @@ -439,33 +536,24 @@ xfs_qm_dqtobp( error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, dqp->q_blkno, mp->m_quotainfo->qi_dqchunklen, - 0, &bp); - if (error || !bp) - return XFS_ERROR(error); - } - - ASSERT(xfs_buf_islocked(bp)); + 0, &bp, &xfs_dquot_buf_ops); - /* - * calculate the location of the dquot inside the buffer. - */ - ddq = bp->b_addr + dqp->q_bufoffset; + if (error == EFSCORRUPTED && (flags & XFS_QMOPT_DQREPAIR)) { + xfs_dqid_t firstid = (xfs_dqid_t)map.br_startoff * + mp->m_quotainfo->qi_dqperchunk; + ASSERT(bp == NULL); + error = xfs_qm_dqrepair(mp, tp, dqp, firstid, &bp); + } - /* - * A simple sanity check in case we got a corrupted dquot... - */ - error = xfs_qm_dqcheck(mp, ddq, id, dqp->dq_flags & XFS_DQ_ALLTYPES, - flags & (XFS_QMOPT_DQREPAIR|XFS_QMOPT_DOWARN), - "dqtobp"); - if (error) { - if (!(flags & XFS_QMOPT_DQREPAIR)) { - xfs_trans_brelse(tp, bp); - return XFS_ERROR(EIO); + if (error) { + ASSERT(bp == NULL); + return XFS_ERROR(error); } } + ASSERT(xfs_buf_islocked(bp)); *O_bpp = bp; - *O_ddpp = ddq; + *O_ddpp = bp->b_addr + dqp->q_bufoffset; return (0); } @@ -920,7 +1008,7 @@ xfs_qm_dqflush( * Get the buffer containing the on-disk dquot */ error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, dqp->q_blkno, - mp->m_quotainfo->qi_dqchunklen, 0, &bp); + mp->m_quotainfo->qi_dqchunklen, 0, &bp, NULL); if (error) goto out_unlock; diff --git a/fs/xfs/xfs_dquot.h b/fs/xfs/xfs_dquot.h index 7d20af27346d..c694a8469c4a 100644 --- a/fs/xfs/xfs_dquot.h +++ b/fs/xfs/xfs_dquot.h @@ -161,4 +161,6 @@ static inline struct xfs_dquot *xfs_qm_dqhold(struct xfs_dquot *dqp) return dqp; } +extern const struct xfs_buf_ops xfs_dquot_buf_ops; + #endif /* __XFS_DQUOT_H__ */ diff --git a/fs/xfs/xfs_export.c b/fs/xfs/xfs_export.c index 8c6d1d70278c..a83611849cee 100644 --- a/fs/xfs/xfs_export.c +++ b/fs/xfs/xfs_export.c @@ -29,6 +29,7 @@ #include "xfs_inode.h" #include "xfs_inode_item.h" #include "xfs_trace.h" +#include "xfs_icache.h" /* * Note that we only accept fileids which are long enough rather than allow diff --git a/fs/xfs/xfs_file.c b/fs/xfs/xfs_file.c index aa473fa640a2..67284edb84d7 100644 --- a/fs/xfs/xfs_file.c +++ b/fs/xfs/xfs_file.c @@ -31,6 +31,8 @@ #include "xfs_error.h" #include "xfs_vnodeops.h" #include "xfs_da_btree.h" +#include "xfs_dir2_format.h" +#include "xfs_dir2_priv.h" #include "xfs_ioctl.h" #include "xfs_trace.h" @@ -84,7 +86,7 @@ xfs_rw_ilock_demote( * valid before the operation, it will be read from disk before * being partially zeroed. */ -STATIC int +int xfs_iozero( struct xfs_inode *ip, /* inode */ loff_t pos, /* offset in file */ @@ -255,15 +257,14 @@ xfs_file_aio_read( xfs_buftarg_t *target = XFS_IS_REALTIME_INODE(ip) ? mp->m_rtdev_targp : mp->m_ddev_targp; - if ((iocb->ki_pos & target->bt_smask) || - (size & target->bt_smask)) { - if (iocb->ki_pos == i_size_read(inode)) + if ((pos & target->bt_smask) || (size & target->bt_smask)) { + if (pos == i_size_read(inode)) return 0; return -XFS_ERROR(EINVAL); } } - n = mp->m_super->s_maxbytes - iocb->ki_pos; + n = mp->m_super->s_maxbytes - pos; if (n <= 0 || size == 0) return 0; @@ -289,20 +290,21 @@ xfs_file_aio_read( xfs_rw_ilock(ip, XFS_IOLOCK_EXCL); if (inode->i_mapping->nrpages) { - ret = -xfs_flushinval_pages(ip, - (iocb->ki_pos & PAGE_CACHE_MASK), - -1, FI_REMAPF_LOCKED); + ret = -filemap_write_and_wait_range( + VFS_I(ip)->i_mapping, + pos, -1); if (ret) { xfs_rw_iunlock(ip, XFS_IOLOCK_EXCL); return ret; } + truncate_pagecache_range(VFS_I(ip), pos, -1); } xfs_rw_ilock_demote(ip, XFS_IOLOCK_EXCL); } - trace_xfs_file_read(ip, size, iocb->ki_pos, ioflags); + trace_xfs_file_read(ip, size, pos, ioflags); - ret = generic_file_aio_read(iocb, iovp, nr_segs, iocb->ki_pos); + ret = generic_file_aio_read(iocb, iovp, nr_segs, pos); if (ret > 0) XFS_STATS_ADD(xs_read_bytes, ret); @@ -670,10 +672,11 @@ xfs_file_dio_aio_write( goto out; if (mapping->nrpages) { - ret = -xfs_flushinval_pages(ip, (pos & PAGE_CACHE_MASK), -1, - FI_REMAPF_LOCKED); + ret = -filemap_write_and_wait_range(VFS_I(ip)->i_mapping, + pos, -1); if (ret) goto out; + truncate_pagecache_range(VFS_I(ip), pos, -1); } /* @@ -728,16 +731,17 @@ xfs_file_buffered_aio_write( write_retry: trace_xfs_file_buffered_write(ip, count, iocb->ki_pos, 0); ret = generic_file_buffered_write(iocb, iovp, nr_segs, - pos, &iocb->ki_pos, count, ret); + pos, &iocb->ki_pos, count, 0); + /* - * if we just got an ENOSPC, flush the inode now we aren't holding any - * page locks and retry *once* + * If we just got an ENOSPC, try to write back all dirty inodes to + * convert delalloc space to free up some of the excess reserved + * metadata space. */ if (ret == -ENOSPC && !enospc) { enospc = 1; - ret = -xfs_flush_pages(ip, 0, -1, 0, FI_NONE); - if (!ret) - goto write_retry; + xfs_flush_inodes(ip->i_mount); + goto write_retry; } current->backing_dev_info = NULL; @@ -889,7 +893,7 @@ xfs_dir_open( */ mode = xfs_ilock_map_shared(ip); if (ip->i_d.di_nextents > 0) - xfs_da_reada_buf(NULL, ip, 0, XFS_DATA_FORK); + xfs_dir2_data_readahead(NULL, ip, 0, -1); xfs_iunlock(ip, mode); return 0; } diff --git a/fs/xfs/xfs_fs.h b/fs/xfs/xfs_fs.h index c13fed8c394a..6dda3f949b04 100644 --- a/fs/xfs/xfs_fs.h +++ b/fs/xfs/xfs_fs.h @@ -233,7 +233,8 @@ typedef struct xfs_fsop_resblks { #define XFS_FSOP_GEOM_FLAGS_LOGV2 0x0100 /* log format version 2 */ #define XFS_FSOP_GEOM_FLAGS_SECTOR 0x0200 /* sector sizes >1BB */ #define XFS_FSOP_GEOM_FLAGS_ATTR2 0x0400 /* inline attributes rework */ -#define XFS_FSOP_GEOM_FLAGS_DIRV2CI 0x1000 /* ASCII only CI names */ +#define XFS_FSOP_GEOM_FLAGS_PROJID32 0x0800 /* 32-bit project IDs */ +#define XFS_FSOP_GEOM_FLAGS_DIRV2CI 0x1000 /* ASCII only CI names */ #define XFS_FSOP_GEOM_FLAGS_LAZYSB 0x4000 /* lazy superblock counters */ @@ -339,6 +340,35 @@ typedef struct xfs_error_injection { /* + * Speculative preallocation trimming. + */ +#define XFS_EOFBLOCKS_VERSION 1 +struct xfs_eofblocks { + __u32 eof_version; + __u32 eof_flags; + uid_t eof_uid; + gid_t eof_gid; + prid_t eof_prid; + __u32 pad32; + __u64 eof_min_file_size; + __u64 pad64[12]; +}; + +/* eof_flags values */ +#define XFS_EOF_FLAGS_SYNC (1 << 0) /* sync/wait mode scan */ +#define XFS_EOF_FLAGS_UID (1 << 1) /* filter by uid */ +#define XFS_EOF_FLAGS_GID (1 << 2) /* filter by gid */ +#define XFS_EOF_FLAGS_PRID (1 << 3) /* filter by project id */ +#define XFS_EOF_FLAGS_MINFILESIZE (1 << 4) /* filter by min file size */ +#define XFS_EOF_FLAGS_VALID \ + (XFS_EOF_FLAGS_SYNC | \ + XFS_EOF_FLAGS_UID | \ + XFS_EOF_FLAGS_GID | \ + XFS_EOF_FLAGS_PRID | \ + XFS_EOF_FLAGS_MINFILESIZE) + + +/* * The user-level Handle Request interface structure. */ typedef struct xfs_fsop_handlereq { @@ -456,6 +486,7 @@ typedef struct xfs_handle { /* XFS_IOC_GETBIOSIZE ---- deprecated 47 */ #define XFS_IOC_GETBMAPX _IOWR('X', 56, struct getbmap) #define XFS_IOC_ZERO_RANGE _IOW ('X', 57, struct xfs_flock64) +#define XFS_IOC_FREE_EOFBLOCKS _IOR ('X', 58, struct xfs_eofblocks) /* * ioctl commands that replace IRIX syssgi()'s diff --git a/fs/xfs/xfs_fs_subr.c b/fs/xfs/xfs_fs_subr.c deleted file mode 100644 index 652b875a9d4c..000000000000 --- a/fs/xfs/xfs_fs_subr.c +++ /dev/null @@ -1,96 +0,0 @@ -/* - * Copyright (c) 2000-2002,2005-2006 Silicon Graphics, Inc. - * 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 as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would 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 the Free Software Foundation, - * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - */ -#include "xfs.h" -#include "xfs_vnodeops.h" -#include "xfs_bmap_btree.h" -#include "xfs_inode.h" -#include "xfs_trace.h" - -/* - * note: all filemap functions return negative error codes. These - * need to be inverted before returning to the xfs core functions. - */ -void -xfs_tosspages( - xfs_inode_t *ip, - xfs_off_t first, - xfs_off_t last, - int fiopt) -{ - /* can't toss partial tail pages, so mask them out */ - last &= ~(PAGE_SIZE - 1); - truncate_inode_pages_range(VFS_I(ip)->i_mapping, first, last - 1); -} - -int -xfs_flushinval_pages( - xfs_inode_t *ip, - xfs_off_t first, - xfs_off_t last, - int fiopt) -{ - struct address_space *mapping = VFS_I(ip)->i_mapping; - int ret = 0; - - trace_xfs_pagecache_inval(ip, first, last); - - xfs_iflags_clear(ip, XFS_ITRUNCATED); - ret = filemap_write_and_wait_range(mapping, first, - last == -1 ? LLONG_MAX : last); - if (!ret) - truncate_inode_pages_range(mapping, first, last); - return -ret; -} - -int -xfs_flush_pages( - xfs_inode_t *ip, - xfs_off_t first, - xfs_off_t last, - uint64_t flags, - int fiopt) -{ - struct address_space *mapping = VFS_I(ip)->i_mapping; - int ret = 0; - int ret2; - - xfs_iflags_clear(ip, XFS_ITRUNCATED); - ret = -filemap_fdatawrite_range(mapping, first, - last == -1 ? LLONG_MAX : last); - if (flags & XBF_ASYNC) - return ret; - ret2 = xfs_wait_on_pages(ip, first, last); - if (!ret) - ret = ret2; - return ret; -} - -int -xfs_wait_on_pages( - xfs_inode_t *ip, - xfs_off_t first, - xfs_off_t last) -{ - struct address_space *mapping = VFS_I(ip)->i_mapping; - - if (mapping_tagged(mapping, PAGECACHE_TAG_WRITEBACK)) { - return -filemap_fdatawait_range(mapping, first, - last == -1 ? XFS_ISIZE(ip) - 1 : last); - } - return 0; -} diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 4beaede43277..94eaeedc5498 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -97,7 +97,9 @@ xfs_fs_geometry( (xfs_sb_version_haslazysbcount(&mp->m_sb) ? XFS_FSOP_GEOM_FLAGS_LAZYSB : 0) | (xfs_sb_version_hasattr2(&mp->m_sb) ? - XFS_FSOP_GEOM_FLAGS_ATTR2 : 0); + XFS_FSOP_GEOM_FLAGS_ATTR2 : 0) | + (xfs_sb_version_hasprojid32bit(&mp->m_sb) ? + XFS_FSOP_GEOM_FLAGS_PROJID32 : 0); geo->logsectsize = xfs_sb_version_hassector(&mp->m_sb) ? mp->m_sb.sb_logsectsize : BBSIZE; geo->rtsectsize = mp->m_sb.sb_blocksize; @@ -112,18 +114,40 @@ xfs_fs_geometry( return 0; } +static struct xfs_buf * +xfs_growfs_get_hdr_buf( + struct xfs_mount *mp, + xfs_daddr_t blkno, + size_t numblks, + int flags, + const struct xfs_buf_ops *ops) +{ + struct xfs_buf *bp; + + bp = xfs_buf_get_uncached(mp->m_ddev_targp, numblks, flags); + if (!bp) + return NULL; + + xfs_buf_zero(bp, 0, BBTOB(bp->b_length)); + bp->b_bn = blkno; + bp->b_maps[0].bm_bn = blkno; + bp->b_ops = ops; + + return bp; +} + static int xfs_growfs_data_private( xfs_mount_t *mp, /* mount point for filesystem */ xfs_growfs_data_t *in) /* growfs data input struct */ { xfs_agf_t *agf; + struct xfs_agfl *agfl; xfs_agi_t *agi; xfs_agnumber_t agno; xfs_extlen_t agsize; xfs_extlen_t tmpsize; xfs_alloc_rec_t *arec; - struct xfs_btree_block *block; xfs_buf_t *bp; int bucket; int dpct; @@ -146,9 +170,14 @@ xfs_growfs_data_private( dpct = pct - mp->m_sb.sb_imax_pct; bp = xfs_buf_read_uncached(mp->m_ddev_targp, XFS_FSB_TO_BB(mp, nb) - XFS_FSS_TO_BB(mp, 1), - XFS_FSS_TO_BB(mp, 1), 0); + XFS_FSS_TO_BB(mp, 1), 0, NULL); if (!bp) return EIO; + if (bp->b_error) { + int error = bp->b_error; + xfs_buf_relse(bp); + return error; + } xfs_buf_relse(bp); new = nb; /* use new as a temporary here */ @@ -186,17 +215,18 @@ xfs_growfs_data_private( nfree = 0; for (agno = nagcount - 1; agno >= oagcount; agno--, new -= agsize) { /* - * AG freelist header block + * AG freespace header block */ - bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0); + bp = xfs_growfs_get_hdr_buf(mp, + XFS_AG_DADDR(mp, agno, XFS_AGF_DADDR(mp)), + XFS_FSS_TO_BB(mp, 1), 0, + &xfs_agf_buf_ops); if (!bp) { error = ENOMEM; goto error0; } + agf = XFS_BUF_TO_AGF(bp); - memset(agf, 0, mp->m_sb.sb_sectsize); agf->agf_magicnum = cpu_to_be32(XFS_AGF_MAGIC); agf->agf_versionnum = cpu_to_be32(XFS_AGF_VERSION); agf->agf_seqno = cpu_to_be32(agno); @@ -223,17 +253,39 @@ xfs_growfs_data_private( goto error0; /* + * AG freelist header block + */ + bp = xfs_growfs_get_hdr_buf(mp, + XFS_AG_DADDR(mp, agno, XFS_AGFL_DADDR(mp)), + XFS_FSS_TO_BB(mp, 1), 0, + &xfs_agfl_buf_ops); + if (!bp) { + error = ENOMEM; + goto error0; + } + + agfl = XFS_BUF_TO_AGFL(bp); + for (bucket = 0; bucket < XFS_AGFL_SIZE(mp); bucket++) + agfl->agfl_bno[bucket] = cpu_to_be32(NULLAGBLOCK); + + error = xfs_bwrite(bp); + xfs_buf_relse(bp); + if (error) + goto error0; + + /* * AG inode header block */ - bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0); + bp = xfs_growfs_get_hdr_buf(mp, + XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), + XFS_FSS_TO_BB(mp, 1), 0, + &xfs_agi_buf_ops); if (!bp) { error = ENOMEM; goto error0; } + agi = XFS_BUF_TO_AGI(bp); - memset(agi, 0, mp->m_sb.sb_sectsize); agi->agi_magicnum = cpu_to_be32(XFS_AGI_MAGIC); agi->agi_versionnum = cpu_to_be32(XFS_AGI_VERSION); agi->agi_seqno = cpu_to_be32(agno); @@ -254,24 +306,22 @@ xfs_growfs_data_private( /* * BNO btree root block */ - bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AGB_TO_DADDR(mp, agno, XFS_BNO_BLOCK(mp)), - BTOBB(mp->m_sb.sb_blocksize), 0); + bp = xfs_growfs_get_hdr_buf(mp, + XFS_AGB_TO_DADDR(mp, agno, XFS_BNO_BLOCK(mp)), + BTOBB(mp->m_sb.sb_blocksize), 0, + &xfs_allocbt_buf_ops); + if (!bp) { error = ENOMEM; goto error0; } - block = XFS_BUF_TO_BLOCK(bp); - memset(block, 0, mp->m_sb.sb_blocksize); - block->bb_magic = cpu_to_be32(XFS_ABTB_MAGIC); - block->bb_level = 0; - block->bb_numrecs = cpu_to_be16(1); - block->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK); - block->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK); - arec = XFS_ALLOC_REC_ADDR(mp, block, 1); + + xfs_btree_init_block(mp, bp, XFS_ABTB_MAGIC, 0, 1, 0); + arec = XFS_ALLOC_REC_ADDR(mp, XFS_BUF_TO_BLOCK(bp), 1); arec->ar_startblock = cpu_to_be32(XFS_PREALLOC_BLOCKS(mp)); arec->ar_blockcount = cpu_to_be32( agsize - be32_to_cpu(arec->ar_startblock)); + error = xfs_bwrite(bp); xfs_buf_relse(bp); if (error) @@ -280,25 +330,22 @@ xfs_growfs_data_private( /* * CNT btree root block */ - bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AGB_TO_DADDR(mp, agno, XFS_CNT_BLOCK(mp)), - BTOBB(mp->m_sb.sb_blocksize), 0); + bp = xfs_growfs_get_hdr_buf(mp, + XFS_AGB_TO_DADDR(mp, agno, XFS_CNT_BLOCK(mp)), + BTOBB(mp->m_sb.sb_blocksize), 0, + &xfs_allocbt_buf_ops); if (!bp) { error = ENOMEM; goto error0; } - block = XFS_BUF_TO_BLOCK(bp); - memset(block, 0, mp->m_sb.sb_blocksize); - block->bb_magic = cpu_to_be32(XFS_ABTC_MAGIC); - block->bb_level = 0; - block->bb_numrecs = cpu_to_be16(1); - block->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK); - block->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK); - arec = XFS_ALLOC_REC_ADDR(mp, block, 1); + + xfs_btree_init_block(mp, bp, XFS_ABTC_MAGIC, 0, 1, 0); + arec = XFS_ALLOC_REC_ADDR(mp, XFS_BUF_TO_BLOCK(bp), 1); arec->ar_startblock = cpu_to_be32(XFS_PREALLOC_BLOCKS(mp)); arec->ar_blockcount = cpu_to_be32( agsize - be32_to_cpu(arec->ar_startblock)); nfree += be32_to_cpu(arec->ar_blockcount); + error = xfs_bwrite(bp); xfs_buf_relse(bp); if (error) @@ -307,20 +354,17 @@ xfs_growfs_data_private( /* * INO btree root block */ - bp = xfs_buf_get(mp->m_ddev_targp, - XFS_AGB_TO_DADDR(mp, agno, XFS_IBT_BLOCK(mp)), - BTOBB(mp->m_sb.sb_blocksize), 0); + bp = xfs_growfs_get_hdr_buf(mp, + XFS_AGB_TO_DADDR(mp, agno, XFS_IBT_BLOCK(mp)), + BTOBB(mp->m_sb.sb_blocksize), 0, + &xfs_inobt_buf_ops); if (!bp) { error = ENOMEM; goto error0; } - block = XFS_BUF_TO_BLOCK(bp); - memset(block, 0, mp->m_sb.sb_blocksize); - block->bb_magic = cpu_to_be32(XFS_IBT_MAGIC); - block->bb_level = 0; - block->bb_numrecs = 0; - block->bb_u.s.bb_leftsib = cpu_to_be32(NULLAGBLOCK); - block->bb_u.s.bb_rightsib = cpu_to_be32(NULLAGBLOCK); + + xfs_btree_init_block(mp, bp, XFS_IBT_MAGIC, 0, 0, 0); + error = xfs_bwrite(bp); xfs_buf_relse(bp); if (error) @@ -408,14 +452,16 @@ xfs_growfs_data_private( if (agno < oagcount) { error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, XFS_AGB_TO_DADDR(mp, agno, XFS_SB_BLOCK(mp)), - XFS_FSS_TO_BB(mp, 1), 0, &bp); + XFS_FSS_TO_BB(mp, 1), 0, &bp, + &xfs_sb_buf_ops); } else { bp = xfs_trans_get_buf(NULL, mp->m_ddev_targp, XFS_AGB_TO_DADDR(mp, agno, XFS_SB_BLOCK(mp)), XFS_FSS_TO_BB(mp, 1), 0); - if (bp) + if (bp) { + bp->b_ops = &xfs_sb_buf_ops; xfs_buf_zero(bp, 0, BBTOB(bp->b_length)); - else + } else error = ENOMEM; } @@ -426,6 +472,7 @@ xfs_growfs_data_private( break; } xfs_sb_to_disk(XFS_BUF_TO_SBP(bp), &mp->m_sb, XFS_SB_ALL_BITS); + /* * If we get an error writing out the alternate superblocks, * just issue a warning and continue. The real work is diff --git a/fs/xfs/xfs_globals.c b/fs/xfs/xfs_globals.c index 76e81cff70b9..5399ef222dd7 100644 --- a/fs/xfs/xfs_globals.c +++ b/fs/xfs/xfs_globals.c @@ -21,7 +21,8 @@ /* * Tunable XFS parameters. xfs_params is required even when CONFIG_SYSCTL=n, * other XFS code uses these values. Times are measured in centisecs (i.e. - * 100ths of a second). + * 100ths of a second) with the exception of eofb_timer, which is measured in + * seconds. */ xfs_param_t xfs_params = { /* MIN DFLT MAX */ @@ -40,4 +41,5 @@ xfs_param_t xfs_params = { .rotorstep = { 1, 1, 255 }, .inherit_nodfrg = { 0, 1, 1 }, .fstrm_timer = { 1, 30*100, 3600*100}, + .eofb_timer = { 1, 300, 3600*24}, }; diff --git a/fs/xfs/xfs_ialloc.c b/fs/xfs/xfs_ialloc.c index c5c4ef4f2bdb..a815412eab80 100644 --- a/fs/xfs/xfs_ialloc.c +++ b/fs/xfs/xfs_ialloc.c @@ -200,7 +200,8 @@ xfs_ialloc_inode_init( */ d = XFS_AGB_TO_DADDR(mp, agno, agbno + (j * blks_per_cluster)); fbuf = xfs_trans_get_buf(tp, mp->m_ddev_targp, d, - mp->m_bsize * blks_per_cluster, 0); + mp->m_bsize * blks_per_cluster, + XBF_UNMAPPED); if (!fbuf) return ENOMEM; /* @@ -210,6 +211,7 @@ xfs_ialloc_inode_init( * to log a whole cluster of inodes instead of all the * individual transactions causing a lot of log traffic. */ + fbuf->b_ops = &xfs_inode_buf_ops; xfs_buf_zero(fbuf, 0, ninodes << mp->m_sb.sb_inodelog); for (i = 0; i < ninodes; i++) { int ioffset = i << mp->m_sb.sb_inodelog; @@ -877,9 +879,9 @@ error0: * This function is designed to be called twice if it has to do an allocation * to make more free inodes. On the first call, *IO_agbp should be set to NULL. * If an inode is available without having to performn an allocation, an inode - * number is returned. In this case, *IO_agbp would be NULL. If an allocation - * needes to be done, xfs_dialloc would return the current AGI buffer in - * *IO_agbp. The caller should then commit the current transaction, allocate a + * number is returned. In this case, *IO_agbp is set to NULL. If an allocation + * needs to be done, xfs_dialloc returns the current AGI buffer in *IO_agbp. + * The caller should then commit the current transaction, allocate a * new transaction, and call xfs_dialloc() again, passing in the previous value * of *IO_agbp. IO_agbp should be held across the transactions. Since the AGI * buffer is locked across the two calls, the second call is guaranteed to have @@ -1472,6 +1474,57 @@ xfs_check_agi_unlinked( #define xfs_check_agi_unlinked(agi) #endif +static void +xfs_agi_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_agi *agi = XFS_BUF_TO_AGI(bp); + int agi_ok; + + /* + * Validate the magic number of the agi block. + */ + agi_ok = agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC) && + XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)); + + /* + * during growfs operations, the perag is not fully initialised, + * so we can't use it for any useful checking. growfs ensures we can't + * use it by using uncached buffers that don't have the perag attached + * so we can detect and avoid this problem. + */ + if (bp->b_pag) + agi_ok = agi_ok && be32_to_cpu(agi->agi_seqno) == + bp->b_pag->pag_agno; + + if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI, + XFS_RANDOM_IALLOC_READ_AGI))) { + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, agi); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } + xfs_check_agi_unlinked(agi); +} + +static void +xfs_agi_read_verify( + struct xfs_buf *bp) +{ + xfs_agi_verify(bp); +} + +static void +xfs_agi_write_verify( + struct xfs_buf *bp) +{ + xfs_agi_verify(bp); +} + +const struct xfs_buf_ops xfs_agi_buf_ops = { + .verify_read = xfs_agi_read_verify, + .verify_write = xfs_agi_write_verify, +}; + /* * Read in the allocation group header (inode allocation section) */ @@ -1482,38 +1535,18 @@ xfs_read_agi( xfs_agnumber_t agno, /* allocation group number */ struct xfs_buf **bpp) /* allocation group hdr buf */ { - struct xfs_agi *agi; /* allocation group header */ - int agi_ok; /* agi is consistent */ int error; ASSERT(agno != NULLAGNUMBER); error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, XFS_AG_DADDR(mp, agno, XFS_AGI_DADDR(mp)), - XFS_FSS_TO_BB(mp, 1), 0, bpp); + XFS_FSS_TO_BB(mp, 1), 0, bpp, &xfs_agi_buf_ops); if (error) return error; ASSERT(!xfs_buf_geterror(*bpp)); - agi = XFS_BUF_TO_AGI(*bpp); - - /* - * Validate the magic number of the agi block. - */ - agi_ok = agi->agi_magicnum == cpu_to_be32(XFS_AGI_MAGIC) && - XFS_AGI_GOOD_VERSION(be32_to_cpu(agi->agi_versionnum)) && - be32_to_cpu(agi->agi_seqno) == agno; - if (unlikely(XFS_TEST_ERROR(!agi_ok, mp, XFS_ERRTAG_IALLOC_READ_AGI, - XFS_RANDOM_IALLOC_READ_AGI))) { - XFS_CORRUPTION_ERROR("xfs_read_agi", XFS_ERRLEVEL_LOW, - mp, agi); - xfs_trans_brelse(tp, *bpp); - return XFS_ERROR(EFSCORRUPTED); - } - xfs_buf_set_ref(*bpp, XFS_AGI_REF); - - xfs_check_agi_unlinked(agi); return 0; } diff --git a/fs/xfs/xfs_ialloc.h b/fs/xfs/xfs_ialloc.h index 1fd6ea4e9c91..c8da3df271e6 100644 --- a/fs/xfs/xfs_ialloc.h +++ b/fs/xfs/xfs_ialloc.h @@ -147,7 +147,9 @@ int xfs_inobt_lookup(struct xfs_btree_cur *cur, xfs_agino_t ino, /* * Get the data from the pointed-to record. */ -extern int xfs_inobt_get_rec(struct xfs_btree_cur *cur, +int xfs_inobt_get_rec(struct xfs_btree_cur *cur, xfs_inobt_rec_incore_t *rec, int *stat); +extern const struct xfs_buf_ops xfs_agi_buf_ops; + #endif /* __XFS_IALLOC_H__ */ diff --git a/fs/xfs/xfs_ialloc_btree.c b/fs/xfs/xfs_ialloc_btree.c index 2b8b7a37aa18..bec344b36507 100644 --- a/fs/xfs/xfs_ialloc_btree.c +++ b/fs/xfs/xfs_ialloc_btree.c @@ -33,6 +33,7 @@ #include "xfs_ialloc.h" #include "xfs_alloc.h" #include "xfs_error.h" +#include "xfs_trace.h" STATIC int @@ -181,6 +182,59 @@ xfs_inobt_key_diff( cur->bc_rec.i.ir_startino; } +void +xfs_inobt_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_btree_block *block = XFS_BUF_TO_BLOCK(bp); + unsigned int level; + int sblock_ok; /* block passes checks */ + + /* magic number and level verification */ + level = be16_to_cpu(block->bb_level); + sblock_ok = block->bb_magic == cpu_to_be32(XFS_IBT_MAGIC) && + level < mp->m_in_maxlevels; + + /* numrecs verification */ + sblock_ok = sblock_ok && + be16_to_cpu(block->bb_numrecs) <= mp->m_inobt_mxr[level != 0]; + + /* sibling pointer verification */ + sblock_ok = sblock_ok && + (block->bb_u.s.bb_leftsib == cpu_to_be32(NULLAGBLOCK) || + be32_to_cpu(block->bb_u.s.bb_leftsib) < mp->m_sb.sb_agblocks) && + block->bb_u.s.bb_leftsib && + (block->bb_u.s.bb_rightsib == cpu_to_be32(NULLAGBLOCK) || + be32_to_cpu(block->bb_u.s.bb_rightsib) < mp->m_sb.sb_agblocks) && + block->bb_u.s.bb_rightsib; + + if (!sblock_ok) { + trace_xfs_btree_corrupt(bp, _RET_IP_); + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_LOW, mp, block); + xfs_buf_ioerror(bp, EFSCORRUPTED); + } +} + +static void +xfs_inobt_read_verify( + struct xfs_buf *bp) +{ + xfs_inobt_verify(bp); +} + +static void +xfs_inobt_write_verify( + struct xfs_buf *bp) +{ + xfs_inobt_verify(bp); +} + +const struct xfs_buf_ops xfs_inobt_buf_ops = { + .verify_read = xfs_inobt_read_verify, + .verify_write = xfs_inobt_write_verify, +}; + #ifdef DEBUG STATIC int xfs_inobt_keys_inorder( @@ -218,6 +272,7 @@ static const struct xfs_btree_ops xfs_inobt_ops = { .init_rec_from_cur = xfs_inobt_init_rec_from_cur, .init_ptr_from_cur = xfs_inobt_init_ptr_from_cur, .key_diff = xfs_inobt_key_diff, + .buf_ops = &xfs_inobt_buf_ops, #ifdef DEBUG .keys_inorder = xfs_inobt_keys_inorder, .recs_inorder = xfs_inobt_recs_inorder, diff --git a/fs/xfs/xfs_ialloc_btree.h b/fs/xfs/xfs_ialloc_btree.h index f782ad0c4769..25c0239a8eab 100644 --- a/fs/xfs/xfs_ialloc_btree.h +++ b/fs/xfs/xfs_ialloc_btree.h @@ -109,4 +109,6 @@ extern struct xfs_btree_cur *xfs_inobt_init_cursor(struct xfs_mount *, struct xfs_trans *, struct xfs_buf *, xfs_agnumber_t); extern int xfs_inobt_maxrecs(struct xfs_mount *, int, int); +extern const struct xfs_buf_ops xfs_inobt_buf_ops; + #endif /* __XFS_IALLOC_BTREE_H__ */ diff --git a/fs/xfs/xfs_sync.c b/fs/xfs/xfs_icache.c index 9500caf15acf..96e344e3e927 100644 --- a/fs/xfs/xfs_sync.c +++ b/fs/xfs/xfs_icache.c @@ -19,6 +19,7 @@ #include "xfs_fs.h" #include "xfs_types.h" #include "xfs_log.h" +#include "xfs_log_priv.h" #include "xfs_inum.h" #include "xfs_trans.h" #include "xfs_trans_priv.h" @@ -35,11 +36,425 @@ #include "xfs_quota.h" #include "xfs_trace.h" #include "xfs_fsops.h" +#include "xfs_icache.h" #include <linux/kthread.h> #include <linux/freezer.h> -struct workqueue_struct *xfs_syncd_wq; /* sync workqueue */ +STATIC void __xfs_inode_clear_reclaim_tag(struct xfs_mount *mp, + struct xfs_perag *pag, struct xfs_inode *ip); + +/* + * Allocate and initialise an xfs_inode. + */ +STATIC struct xfs_inode * +xfs_inode_alloc( + struct xfs_mount *mp, + xfs_ino_t ino) +{ + struct xfs_inode *ip; + + /* + * if this didn't occur in transactions, we could use + * KM_MAYFAIL and return NULL here on ENOMEM. Set the + * code up to do this anyway. + */ + ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP); + if (!ip) + return NULL; + if (inode_init_always(mp->m_super, VFS_I(ip))) { + kmem_zone_free(xfs_inode_zone, ip); + return NULL; + } + + ASSERT(atomic_read(&ip->i_pincount) == 0); + ASSERT(!spin_is_locked(&ip->i_flags_lock)); + ASSERT(!xfs_isiflocked(ip)); + ASSERT(ip->i_ino == 0); + + mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); + + /* initialise the xfs inode */ + ip->i_ino = ino; + ip->i_mount = mp; + memset(&ip->i_imap, 0, sizeof(struct xfs_imap)); + ip->i_afp = NULL; + memset(&ip->i_df, 0, sizeof(xfs_ifork_t)); + ip->i_flags = 0; + ip->i_delayed_blks = 0; + memset(&ip->i_d, 0, sizeof(xfs_icdinode_t)); + + return ip; +} + +STATIC void +xfs_inode_free_callback( + struct rcu_head *head) +{ + struct inode *inode = container_of(head, struct inode, i_rcu); + struct xfs_inode *ip = XFS_I(inode); + + kmem_zone_free(xfs_inode_zone, ip); +} + +STATIC void +xfs_inode_free( + struct xfs_inode *ip) +{ + switch (ip->i_d.di_mode & S_IFMT) { + case S_IFREG: + case S_IFDIR: + case S_IFLNK: + xfs_idestroy_fork(ip, XFS_DATA_FORK); + break; + } + + if (ip->i_afp) + xfs_idestroy_fork(ip, XFS_ATTR_FORK); + + if (ip->i_itemp) { + ASSERT(!(ip->i_itemp->ili_item.li_flags & XFS_LI_IN_AIL)); + xfs_inode_item_destroy(ip); + ip->i_itemp = NULL; + } + + /* asserts to verify all state is correct here */ + ASSERT(atomic_read(&ip->i_pincount) == 0); + ASSERT(!spin_is_locked(&ip->i_flags_lock)); + ASSERT(!xfs_isiflocked(ip)); + + /* + * Because we use RCU freeing we need to ensure the inode always + * appears to be reclaimed with an invalid inode number when in the + * free state. The ip->i_flags_lock provides the barrier against lookup + * races. + */ + spin_lock(&ip->i_flags_lock); + ip->i_flags = XFS_IRECLAIM; + ip->i_ino = 0; + spin_unlock(&ip->i_flags_lock); + + call_rcu(&VFS_I(ip)->i_rcu, xfs_inode_free_callback); +} + +/* + * Check the validity of the inode we just found it the cache + */ +static int +xfs_iget_cache_hit( + struct xfs_perag *pag, + struct xfs_inode *ip, + xfs_ino_t ino, + int flags, + int lock_flags) __releases(RCU) +{ + struct inode *inode = VFS_I(ip); + struct xfs_mount *mp = ip->i_mount; + int error; + + /* + * check for re-use of an inode within an RCU grace period due to the + * radix tree nodes not being updated yet. We monitor for this by + * setting the inode number to zero before freeing the inode structure. + * If the inode has been reallocated and set up, then the inode number + * will not match, so check for that, too. + */ + spin_lock(&ip->i_flags_lock); + if (ip->i_ino != ino) { + trace_xfs_iget_skip(ip); + XFS_STATS_INC(xs_ig_frecycle); + error = EAGAIN; + goto out_error; + } + + + /* + * If we are racing with another cache hit that is currently + * instantiating this inode or currently recycling it out of + * reclaimabe state, wait for the initialisation to complete + * before continuing. + * + * XXX(hch): eventually we should do something equivalent to + * wait_on_inode to wait for these flags to be cleared + * instead of polling for it. + */ + if (ip->i_flags & (XFS_INEW|XFS_IRECLAIM)) { + trace_xfs_iget_skip(ip); + XFS_STATS_INC(xs_ig_frecycle); + error = EAGAIN; + goto out_error; + } + + /* + * If lookup is racing with unlink return an error immediately. + */ + if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) { + error = ENOENT; + goto out_error; + } + + /* + * If IRECLAIMABLE is set, we've torn down the VFS inode already. + * Need to carefully get it back into useable state. + */ + if (ip->i_flags & XFS_IRECLAIMABLE) { + trace_xfs_iget_reclaim(ip); + + /* + * We need to set XFS_IRECLAIM to prevent xfs_reclaim_inode + * from stomping over us while we recycle the inode. We can't + * clear the radix tree reclaimable tag yet as it requires + * pag_ici_lock to be held exclusive. + */ + ip->i_flags |= XFS_IRECLAIM; + + spin_unlock(&ip->i_flags_lock); + rcu_read_unlock(); + + error = -inode_init_always(mp->m_super, inode); + if (error) { + /* + * Re-initializing the inode failed, and we are in deep + * trouble. Try to re-add it to the reclaim list. + */ + rcu_read_lock(); + spin_lock(&ip->i_flags_lock); + + ip->i_flags &= ~(XFS_INEW | XFS_IRECLAIM); + ASSERT(ip->i_flags & XFS_IRECLAIMABLE); + trace_xfs_iget_reclaim_fail(ip); + goto out_error; + } + + spin_lock(&pag->pag_ici_lock); + spin_lock(&ip->i_flags_lock); + + /* + * Clear the per-lifetime state in the inode as we are now + * effectively a new inode and need to return to the initial + * state before reuse occurs. + */ + ip->i_flags &= ~XFS_IRECLAIM_RESET_FLAGS; + ip->i_flags |= XFS_INEW; + __xfs_inode_clear_reclaim_tag(mp, pag, ip); + inode->i_state = I_NEW; + + ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock)); + mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); + + spin_unlock(&ip->i_flags_lock); + spin_unlock(&pag->pag_ici_lock); + } else { + /* If the VFS inode is being torn down, pause and try again. */ + if (!igrab(inode)) { + trace_xfs_iget_skip(ip); + error = EAGAIN; + goto out_error; + } + + /* We've got a live one. */ + spin_unlock(&ip->i_flags_lock); + rcu_read_unlock(); + trace_xfs_iget_hit(ip); + } + + if (lock_flags != 0) + xfs_ilock(ip, lock_flags); + + xfs_iflags_clear(ip, XFS_ISTALE | XFS_IDONTCACHE); + XFS_STATS_INC(xs_ig_found); + + return 0; + +out_error: + spin_unlock(&ip->i_flags_lock); + rcu_read_unlock(); + return error; +} + + +static int +xfs_iget_cache_miss( + struct xfs_mount *mp, + struct xfs_perag *pag, + xfs_trans_t *tp, + xfs_ino_t ino, + struct xfs_inode **ipp, + int flags, + int lock_flags) +{ + struct xfs_inode *ip; + int error; + xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ino); + int iflags; + + ip = xfs_inode_alloc(mp, ino); + if (!ip) + return ENOMEM; + + error = xfs_iread(mp, tp, ip, flags); + if (error) + goto out_destroy; + + trace_xfs_iget_miss(ip); + + if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) { + error = ENOENT; + goto out_destroy; + } + + /* + * Preload the radix tree so we can insert safely under the + * write spinlock. Note that we cannot sleep inside the preload + * region. Since we can be called from transaction context, don't + * recurse into the file system. + */ + if (radix_tree_preload(GFP_NOFS)) { + error = EAGAIN; + goto out_destroy; + } + + /* + * Because the inode hasn't been added to the radix-tree yet it can't + * be found by another thread, so we can do the non-sleeping lock here. + */ + if (lock_flags) { + if (!xfs_ilock_nowait(ip, lock_flags)) + BUG(); + } + + /* + * These values must be set before inserting the inode into the radix + * tree as the moment it is inserted a concurrent lookup (allowed by the + * RCU locking mechanism) can find it and that lookup must see that this + * is an inode currently under construction (i.e. that XFS_INEW is set). + * The ip->i_flags_lock that protects the XFS_INEW flag forms the + * memory barrier that ensures this detection works correctly at lookup + * time. + */ + iflags = XFS_INEW; + if (flags & XFS_IGET_DONTCACHE) + iflags |= XFS_IDONTCACHE; + ip->i_udquot = ip->i_gdquot = NULL; + xfs_iflags_set(ip, iflags); + + /* insert the new inode */ + spin_lock(&pag->pag_ici_lock); + error = radix_tree_insert(&pag->pag_ici_root, agino, ip); + if (unlikely(error)) { + WARN_ON(error != -EEXIST); + XFS_STATS_INC(xs_ig_dup); + error = EAGAIN; + goto out_preload_end; + } + spin_unlock(&pag->pag_ici_lock); + radix_tree_preload_end(); + + *ipp = ip; + return 0; + +out_preload_end: + spin_unlock(&pag->pag_ici_lock); + radix_tree_preload_end(); + if (lock_flags) + xfs_iunlock(ip, lock_flags); +out_destroy: + __destroy_inode(VFS_I(ip)); + xfs_inode_free(ip); + return error; +} + +/* + * Look up an inode by number in the given file system. + * The inode is looked up in the cache held in each AG. + * If the inode is found in the cache, initialise the vfs inode + * if necessary. + * + * If it is not in core, read it in from the file system's device, + * add it to the cache and initialise the vfs inode. + * + * The inode is locked according to the value of the lock_flags parameter. + * This flag parameter indicates how and if the inode's IO lock and inode lock + * should be taken. + * + * mp -- the mount point structure for the current file system. It points + * to the inode hash table. + * tp -- a pointer to the current transaction if there is one. This is + * simply passed through to the xfs_iread() call. + * ino -- the number of the inode desired. This is the unique identifier + * within the file system for the inode being requested. + * lock_flags -- flags indicating how to lock the inode. See the comment + * for xfs_ilock() for a list of valid values. + */ +int +xfs_iget( + xfs_mount_t *mp, + xfs_trans_t *tp, + xfs_ino_t ino, + uint flags, + uint lock_flags, + xfs_inode_t **ipp) +{ + xfs_inode_t *ip; + int error; + xfs_perag_t *pag; + xfs_agino_t agino; + + /* + * xfs_reclaim_inode() uses the ILOCK to ensure an inode + * doesn't get freed while it's being referenced during a + * radix tree traversal here. It assumes this function + * aqcuires only the ILOCK (and therefore it has no need to + * involve the IOLOCK in this synchronization). + */ + ASSERT((lock_flags & (XFS_IOLOCK_EXCL | XFS_IOLOCK_SHARED)) == 0); + + /* reject inode numbers outside existing AGs */ + if (!ino || XFS_INO_TO_AGNO(mp, ino) >= mp->m_sb.sb_agcount) + return EINVAL; + + /* get the perag structure and ensure that it's inode capable */ + pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ino)); + agino = XFS_INO_TO_AGINO(mp, ino); + +again: + error = 0; + rcu_read_lock(); + ip = radix_tree_lookup(&pag->pag_ici_root, agino); + + if (ip) { + error = xfs_iget_cache_hit(pag, ip, ino, flags, lock_flags); + if (error) + goto out_error_or_again; + } else { + rcu_read_unlock(); + XFS_STATS_INC(xs_ig_missed); + + error = xfs_iget_cache_miss(mp, pag, tp, ino, &ip, + flags, lock_flags); + if (error) + goto out_error_or_again; + } + xfs_perag_put(pag); + + *ipp = ip; + + /* + * If we have a real type for an on-disk inode, we can set ops(&unlock) + * now. If it's a new inode being created, xfs_ialloc will handle it. + */ + if (xfs_iflags_test(ip, XFS_INEW) && ip->i_d.di_mode != 0) + xfs_setup_inode(ip); + return 0; + +out_error_or_again: + if (error == EAGAIN) { + delay(1); + goto again; + } + xfs_perag_put(pag); + return error; +} /* * The inode lookup is done in batches to keep the amount of lock traffic and @@ -101,8 +516,11 @@ xfs_inode_ag_walk( struct xfs_mount *mp, struct xfs_perag *pag, int (*execute)(struct xfs_inode *ip, - struct xfs_perag *pag, int flags), - int flags) + struct xfs_perag *pag, int flags, + void *args), + int flags, + void *args, + int tag) { uint32_t first_index; int last_error = 0; @@ -121,9 +539,17 @@ restart: int i; rcu_read_lock(); - nr_found = radix_tree_gang_lookup(&pag->pag_ici_root, + + if (tag == -1) + nr_found = radix_tree_gang_lookup(&pag->pag_ici_root, (void **)batch, first_index, XFS_LOOKUP_BATCH); + else + nr_found = radix_tree_gang_lookup_tag( + &pag->pag_ici_root, + (void **) batch, first_index, + XFS_LOOKUP_BATCH, tag); + if (!nr_found) { rcu_read_unlock(); break; @@ -164,7 +590,7 @@ restart: for (i = 0; i < nr_found; i++) { if (!batch[i]) continue; - error = execute(batch[i], pag, flags); + error = execute(batch[i], pag, flags, args); IRELE(batch[i]); if (error == EAGAIN) { skipped++; @@ -189,12 +615,40 @@ restart: return last_error; } +/* + * Background scanning to trim post-EOF preallocated space. This is queued + * based on the 'background_prealloc_discard_period' tunable (5m by default). + */ +STATIC void +xfs_queue_eofblocks( + struct xfs_mount *mp) +{ + rcu_read_lock(); + if (radix_tree_tagged(&mp->m_perag_tree, XFS_ICI_EOFBLOCKS_TAG)) + queue_delayed_work(mp->m_eofblocks_workqueue, + &mp->m_eofblocks_work, + msecs_to_jiffies(xfs_eofb_secs * 1000)); + rcu_read_unlock(); +} + +void +xfs_eofblocks_worker( + struct work_struct *work) +{ + struct xfs_mount *mp = container_of(to_delayed_work(work), + struct xfs_mount, m_eofblocks_work); + xfs_icache_free_eofblocks(mp, NULL); + xfs_queue_eofblocks(mp); +} + int xfs_inode_ag_iterator( struct xfs_mount *mp, int (*execute)(struct xfs_inode *ip, - struct xfs_perag *pag, int flags), - int flags) + struct xfs_perag *pag, int flags, + void *args), + int flags, + void *args) { struct xfs_perag *pag; int error = 0; @@ -204,7 +658,7 @@ xfs_inode_ag_iterator( ag = 0; while ((pag = xfs_perag_get(mp, ag))) { ag = pag->pag_agno + 1; - error = xfs_inode_ag_walk(mp, pag, execute, flags); + error = xfs_inode_ag_walk(mp, pag, execute, flags, args, -1); xfs_perag_put(pag); if (error) { last_error = error; @@ -215,224 +669,50 @@ xfs_inode_ag_iterator( return XFS_ERROR(last_error); } -STATIC int -xfs_sync_inode_data( - struct xfs_inode *ip, - struct xfs_perag *pag, - int flags) -{ - struct inode *inode = VFS_I(ip); - struct address_space *mapping = inode->i_mapping; - int error = 0; - - if (!mapping_tagged(mapping, PAGECACHE_TAG_DIRTY)) - return 0; - - if (!xfs_ilock_nowait(ip, XFS_IOLOCK_SHARED)) { - if (flags & SYNC_TRYLOCK) - return 0; - xfs_ilock(ip, XFS_IOLOCK_SHARED); - } - - error = xfs_flush_pages(ip, 0, -1, (flags & SYNC_WAIT) ? - 0 : XBF_ASYNC, FI_NONE); - xfs_iunlock(ip, XFS_IOLOCK_SHARED); - return error; -} - -/* - * Write out pagecache data for the whole filesystem. - */ -STATIC int -xfs_sync_data( - struct xfs_mount *mp, - int flags) -{ - int error; - - ASSERT((flags & ~(SYNC_TRYLOCK|SYNC_WAIT)) == 0); - - error = xfs_inode_ag_iterator(mp, xfs_sync_inode_data, flags); - if (error) - return XFS_ERROR(error); - - xfs_log_force(mp, (flags & SYNC_WAIT) ? XFS_LOG_SYNC : 0); - return 0; -} - -STATIC int -xfs_sync_fsdata( - struct xfs_mount *mp) -{ - struct xfs_buf *bp; - int error; - - /* - * If the buffer is pinned then push on the log so we won't get stuck - * waiting in the write for someone, maybe ourselves, to flush the log. - * - * Even though we just pushed the log above, we did not have the - * superblock buffer locked at that point so it can become pinned in - * between there and here. - */ - bp = xfs_getsb(mp, 0); - if (xfs_buf_ispinned(bp)) - xfs_log_force(mp, 0); - error = xfs_bwrite(bp); - xfs_buf_relse(bp); - return error; -} - -/* - * When remounting a filesystem read-only or freezing the filesystem, we have - * two phases to execute. This first phase is syncing the data before we - * quiesce the filesystem, and the second is flushing all the inodes out after - * we've waited for all the transactions created by the first phase to - * complete. The second phase ensures that the inodes are written to their - * location on disk rather than just existing in transactions in the log. This - * means after a quiesce there is no log replay required to write the inodes to - * disk (this is the main difference between a sync and a quiesce). - */ -/* - * First stage of freeze - no writers will make progress now we are here, - * so we flush delwri and delalloc buffers here, then wait for all I/O to - * complete. Data is frozen at that point. Metadata is not frozen, - * transactions can still occur here so don't bother emptying the AIL - * because it'll just get dirty again. - */ int -xfs_quiesce_data( - struct xfs_mount *mp) -{ - int error, error2 = 0; - - /* force out the log */ - xfs_log_force(mp, XFS_LOG_SYNC); - - /* write superblock and hoover up shutdown errors */ - error = xfs_sync_fsdata(mp); - - /* mark the log as covered if needed */ - if (xfs_log_need_covered(mp)) - error2 = xfs_fs_log_dummy(mp); - - return error ? error : error2; -} - -/* - * 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 proceeding. - */ -void -xfs_quiesce_attr( - struct xfs_mount *mp) -{ - int error = 0; - - /* wait for all modifications to complete */ - while (atomic_read(&mp->m_active_trans) > 0) - delay(100); - - /* reclaim inodes to do any IO before the freeze completes */ - xfs_reclaim_inodes(mp, 0); - xfs_reclaim_inodes(mp, SYNC_WAIT); - - /* flush all pending changes from the AIL */ - xfs_ail_push_all_sync(mp->m_ail); - - /* - * Just warn here till VFS can correctly support - * read-only remount without racing. - */ - WARN_ON(atomic_read(&mp->m_active_trans) != 0); - - /* Push the superblock and write an unmount record */ - error = xfs_log_sbcount(mp); - if (error) - xfs_warn(mp, "xfs_attr_quiesce: failed to log sb changes. " - "Frozen image may not be consistent."); - xfs_log_unmount_write(mp); - - /* - * At this point we might have modified the superblock again and thus - * added an item to the AIL, thus flush it again. - */ - xfs_ail_push_all_sync(mp->m_ail); - - /* - * The superblock buffer is uncached and xfsaild_push() will lock and - * set the XBF_ASYNC flag on the buffer. We cannot do xfs_buf_iowait() - * here but a lock on the superblock buffer will block until iodone() - * has completed. - */ - xfs_buf_lock(mp->m_sb_bp); - xfs_buf_unlock(mp->m_sb_bp); -} - -static void -xfs_syncd_queue_sync( - struct xfs_mount *mp) -{ - queue_delayed_work(xfs_syncd_wq, &mp->m_sync_work, - msecs_to_jiffies(xfs_syncd_centisecs * 10)); -} - -/* - * Every sync period we need to unpin all items, reclaim inodes and sync - * disk quotas. We might need to cover the log to indicate that the - * filesystem is idle and not frozen. - */ -STATIC void -xfs_sync_worker( - struct work_struct *work) +xfs_inode_ag_iterator_tag( + struct xfs_mount *mp, + int (*execute)(struct xfs_inode *ip, + struct xfs_perag *pag, int flags, + void *args), + int flags, + void *args, + int tag) { - struct xfs_mount *mp = container_of(to_delayed_work(work), - struct xfs_mount, m_sync_work); - int error; - - /* - * We shouldn't write/force the log if we are in the mount/unmount - * process or on a read only filesystem. The workqueue still needs to be - * active in both cases, however, because it is used for inode reclaim - * during these times. Use the MS_ACTIVE flag to avoid doing anything - * during mount. Doing work during unmount is avoided by calling - * cancel_delayed_work_sync on this work queue before tearing down - * the ail and the log in xfs_log_unmount. - */ - if (!(mp->m_super->s_flags & MS_ACTIVE) && - !(mp->m_flags & XFS_MOUNT_RDONLY)) { - /* dgc: errors ignored here */ - if (mp->m_super->s_writers.frozen == SB_UNFROZEN && - xfs_log_need_covered(mp)) - error = xfs_fs_log_dummy(mp); - else - xfs_log_force(mp, 0); + struct xfs_perag *pag; + int error = 0; + int last_error = 0; + xfs_agnumber_t ag; - /* start pushing all the metadata that is currently - * dirty */ - xfs_ail_push_all(mp->m_ail); + ag = 0; + while ((pag = xfs_perag_get_tag(mp, ag, tag))) { + ag = pag->pag_agno + 1; + error = xfs_inode_ag_walk(mp, pag, execute, flags, args, tag); + xfs_perag_put(pag); + if (error) { + last_error = error; + if (error == EFSCORRUPTED) + break; + } } - - /* queue us up again */ - xfs_syncd_queue_sync(mp); + return XFS_ERROR(last_error); } /* * 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 + * on the xfs periodic sync 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( +xfs_reclaim_work_queue( struct xfs_mount *mp) { 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, + queue_delayed_work(mp->m_reclaim_workqueue, &mp->m_reclaim_work, msecs_to_jiffies(xfs_syncd_centisecs / 6 * 10)); } rcu_read_unlock(); @@ -445,7 +725,7 @@ xfs_syncd_queue_reclaim( * 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 +void xfs_reclaim_worker( struct work_struct *work) { @@ -453,65 +733,10 @@ xfs_reclaim_worker( struct xfs_mount, m_reclaim_work); xfs_reclaim_inodes(mp, SYNC_TRYLOCK); - xfs_syncd_queue_reclaim(mp); + xfs_reclaim_work_queue(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(&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) -{ - 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); - - return 0; -} - -void -xfs_syncd_stop( - struct xfs_mount *mp) -{ - cancel_delayed_work_sync(&mp->m_sync_work); - cancel_delayed_work_sync(&mp->m_reclaim_work); - cancel_work_sync(&mp->m_flush_work); -} - -void +static void __xfs_inode_set_reclaim_tag( struct xfs_perag *pag, struct xfs_inode *ip) @@ -529,7 +754,7 @@ __xfs_inode_set_reclaim_tag( spin_unlock(&ip->i_mount->m_perag_lock); /* schedule periodic background inode reclaim */ - xfs_syncd_queue_reclaim(ip->i_mount); + xfs_reclaim_work_queue(ip->i_mount); trace_xfs_perag_set_reclaim(ip->i_mount, pag->pag_agno, -1, _RET_IP_); @@ -577,7 +802,7 @@ __xfs_inode_clear_reclaim( } } -void +STATIC void __xfs_inode_clear_reclaim_tag( xfs_mount_t *mp, xfs_perag_t *pag, @@ -787,9 +1012,9 @@ out: /* * We could return EAGAIN here to make reclaim rescan the inode tree in * a short while. However, this just burns CPU time scanning the tree - * waiting for IO to complete and xfssyncd never goes back to the idle - * state. Instead, return 0 to let the next scheduled background reclaim - * attempt to reclaim the inode again. + * waiting for IO to complete and the reclaim work never goes back to + * the idle state. Instead, return 0 to let the next scheduled + * background reclaim attempt to reclaim the inode again. */ return 0; } @@ -800,7 +1025,7 @@ out: * then a shut down during filesystem unmount reclaim walk leak all the * unreclaimed inodes. */ -int +STATIC int xfs_reclaim_inodes_ag( struct xfs_mount *mp, int flags, @@ -945,7 +1170,7 @@ xfs_reclaim_inodes_nr( int nr_to_scan) { /* kick background reclaimer and push the AIL */ - xfs_syncd_queue_reclaim(mp); + xfs_reclaim_work_queue(mp); xfs_ail_push_all(mp->m_ail); xfs_reclaim_inodes_ag(mp, SYNC_TRYLOCK | SYNC_WAIT, &nr_to_scan); @@ -971,3 +1196,146 @@ xfs_reclaim_inodes_count( return reclaimable; } +STATIC int +xfs_inode_match_id( + struct xfs_inode *ip, + struct xfs_eofblocks *eofb) +{ + if (eofb->eof_flags & XFS_EOF_FLAGS_UID && + ip->i_d.di_uid != eofb->eof_uid) + return 0; + + if (eofb->eof_flags & XFS_EOF_FLAGS_GID && + ip->i_d.di_gid != eofb->eof_gid) + return 0; + + if (eofb->eof_flags & XFS_EOF_FLAGS_PRID && + xfs_get_projid(ip) != eofb->eof_prid) + return 0; + + return 1; +} + +STATIC int +xfs_inode_free_eofblocks( + struct xfs_inode *ip, + struct xfs_perag *pag, + int flags, + void *args) +{ + int ret; + struct xfs_eofblocks *eofb = args; + + if (!xfs_can_free_eofblocks(ip, false)) { + /* inode could be preallocated or append-only */ + trace_xfs_inode_free_eofblocks_invalid(ip); + xfs_inode_clear_eofblocks_tag(ip); + return 0; + } + + /* + * If the mapping is dirty the operation can block and wait for some + * time. Unless we are waiting, skip it. + */ + if (!(flags & SYNC_WAIT) && + mapping_tagged(VFS_I(ip)->i_mapping, PAGECACHE_TAG_DIRTY)) + return 0; + + if (eofb) { + if (!xfs_inode_match_id(ip, eofb)) + return 0; + + /* skip the inode if the file size is too small */ + if (eofb->eof_flags & XFS_EOF_FLAGS_MINFILESIZE && + XFS_ISIZE(ip) < eofb->eof_min_file_size) + return 0; + } + + ret = xfs_free_eofblocks(ip->i_mount, ip, true); + + /* don't revisit the inode if we're not waiting */ + if (ret == EAGAIN && !(flags & SYNC_WAIT)) + ret = 0; + + return ret; +} + +int +xfs_icache_free_eofblocks( + struct xfs_mount *mp, + struct xfs_eofblocks *eofb) +{ + int flags = SYNC_TRYLOCK; + + if (eofb && (eofb->eof_flags & XFS_EOF_FLAGS_SYNC)) + flags = SYNC_WAIT; + + return xfs_inode_ag_iterator_tag(mp, xfs_inode_free_eofblocks, flags, + eofb, XFS_ICI_EOFBLOCKS_TAG); +} + +void +xfs_inode_set_eofblocks_tag( + xfs_inode_t *ip) +{ + struct xfs_mount *mp = ip->i_mount; + struct xfs_perag *pag; + int tagged; + + pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino)); + spin_lock(&pag->pag_ici_lock); + trace_xfs_inode_set_eofblocks_tag(ip); + + tagged = radix_tree_tagged(&pag->pag_ici_root, + XFS_ICI_EOFBLOCKS_TAG); + radix_tree_tag_set(&pag->pag_ici_root, + XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino), + XFS_ICI_EOFBLOCKS_TAG); + if (!tagged) { + /* propagate the eofblocks tag up into the perag radix tree */ + spin_lock(&ip->i_mount->m_perag_lock); + radix_tree_tag_set(&ip->i_mount->m_perag_tree, + XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino), + XFS_ICI_EOFBLOCKS_TAG); + spin_unlock(&ip->i_mount->m_perag_lock); + + /* kick off background trimming */ + xfs_queue_eofblocks(ip->i_mount); + + trace_xfs_perag_set_eofblocks(ip->i_mount, pag->pag_agno, + -1, _RET_IP_); + } + + spin_unlock(&pag->pag_ici_lock); + xfs_perag_put(pag); +} + +void +xfs_inode_clear_eofblocks_tag( + xfs_inode_t *ip) +{ + struct xfs_mount *mp = ip->i_mount; + struct xfs_perag *pag; + + pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ip->i_ino)); + spin_lock(&pag->pag_ici_lock); + trace_xfs_inode_clear_eofblocks_tag(ip); + + radix_tree_tag_clear(&pag->pag_ici_root, + XFS_INO_TO_AGINO(ip->i_mount, ip->i_ino), + XFS_ICI_EOFBLOCKS_TAG); + if (!radix_tree_tagged(&pag->pag_ici_root, XFS_ICI_EOFBLOCKS_TAG)) { + /* clear the eofblocks tag from the perag radix tree */ + spin_lock(&ip->i_mount->m_perag_lock); + radix_tree_tag_clear(&ip->i_mount->m_perag_tree, + XFS_INO_TO_AGNO(ip->i_mount, ip->i_ino), + XFS_ICI_EOFBLOCKS_TAG); + spin_unlock(&ip->i_mount->m_perag_lock); + trace_xfs_perag_clear_eofblocks(ip->i_mount, pag->pag_agno, + -1, _RET_IP_); + } + + spin_unlock(&pag->pag_ici_lock); + xfs_perag_put(pag); +} + diff --git a/fs/xfs/xfs_sync.h b/fs/xfs/xfs_icache.h index 941202e7ac6e..e0f138c70a2f 100644 --- a/fs/xfs/xfs_sync.h +++ b/fs/xfs/xfs_icache.h @@ -24,28 +24,30 @@ struct xfs_perag; #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_iget(struct xfs_mount *mp, struct xfs_trans *tp, xfs_ino_t ino, + uint flags, uint lock_flags, xfs_inode_t **ipp); -int xfs_syncd_init(struct xfs_mount *mp); -void xfs_syncd_stop(struct xfs_mount *mp); - -int xfs_quiesce_data(struct xfs_mount *mp); -void xfs_quiesce_attr(struct xfs_mount *mp); - -void xfs_flush_inodes(struct xfs_inode *ip); +void xfs_reclaim_worker(struct work_struct *work); int xfs_reclaim_inodes(struct xfs_mount *mp, int mode); int xfs_reclaim_inodes_count(struct xfs_mount *mp); void xfs_reclaim_inodes_nr(struct xfs_mount *mp, int nr_to_scan); void xfs_inode_set_reclaim_tag(struct xfs_inode *ip); -void __xfs_inode_set_reclaim_tag(struct xfs_perag *pag, struct xfs_inode *ip); -void __xfs_inode_clear_reclaim_tag(struct xfs_mount *mp, struct xfs_perag *pag, - struct xfs_inode *ip); + +void xfs_inode_set_eofblocks_tag(struct xfs_inode *ip); +void xfs_inode_clear_eofblocks_tag(struct xfs_inode *ip); +int xfs_icache_free_eofblocks(struct xfs_mount *, struct xfs_eofblocks *); +void xfs_eofblocks_worker(struct work_struct *); int xfs_sync_inode_grab(struct xfs_inode *ip); int xfs_inode_ag_iterator(struct xfs_mount *mp, - int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag, int flags), - int flags); + int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag, + int flags, void *args), + int flags, void *args); +int xfs_inode_ag_iterator_tag(struct xfs_mount *mp, + int (*execute)(struct xfs_inode *ip, struct xfs_perag *pag, + int flags, void *args), + int flags, void *args, int tag); #endif diff --git a/fs/xfs/xfs_iget.c b/fs/xfs/xfs_iget.c deleted file mode 100644 index 784a803383ec..000000000000 --- a/fs/xfs/xfs_iget.c +++ /dev/null @@ -1,705 +0,0 @@ -/* - * Copyright (c) 2000-2005 Silicon Graphics, Inc. - * 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 as - * published by the Free Software Foundation. - * - * This program is distributed in the hope that it would 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 the Free Software Foundation, - * Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA - */ -#include "xfs.h" -#include "xfs_fs.h" -#include "xfs_types.h" -#include "xfs_acl.h" -#include "xfs_log.h" -#include "xfs_inum.h" -#include "xfs_trans.h" -#include "xfs_sb.h" -#include "xfs_ag.h" -#include "xfs_mount.h" -#include "xfs_bmap_btree.h" -#include "xfs_alloc_btree.h" -#include "xfs_ialloc_btree.h" -#include "xfs_dinode.h" -#include "xfs_inode.h" -#include "xfs_btree.h" -#include "xfs_ialloc.h" -#include "xfs_quota.h" -#include "xfs_utils.h" -#include "xfs_trans_priv.h" -#include "xfs_inode_item.h" -#include "xfs_bmap.h" -#include "xfs_trace.h" - - -/* - * Allocate and initialise an xfs_inode. - */ -STATIC struct xfs_inode * -xfs_inode_alloc( - struct xfs_mount *mp, - xfs_ino_t ino) -{ - struct xfs_inode *ip; - - /* - * if this didn't occur in transactions, we could use - * KM_MAYFAIL and return NULL here on ENOMEM. Set the - * code up to do this anyway. - */ - ip = kmem_zone_alloc(xfs_inode_zone, KM_SLEEP); - if (!ip) - return NULL; - if (inode_init_always(mp->m_super, VFS_I(ip))) { - kmem_zone_free(xfs_inode_zone, ip); - return NULL; - } - - ASSERT(atomic_read(&ip->i_pincount) == 0); - ASSERT(!spin_is_locked(&ip->i_flags_lock)); - ASSERT(!xfs_isiflocked(ip)); - ASSERT(ip->i_ino == 0); - - mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); - - /* initialise the xfs inode */ - ip->i_ino = ino; - ip->i_mount = mp; - memset(&ip->i_imap, 0, sizeof(struct xfs_imap)); - ip->i_afp = NULL; - memset(&ip->i_df, 0, sizeof(xfs_ifork_t)); - ip->i_flags = 0; - ip->i_delayed_blks = 0; - memset(&ip->i_d, 0, sizeof(xfs_icdinode_t)); - - return ip; -} - -STATIC void -xfs_inode_free_callback( - struct rcu_head *head) -{ - struct inode *inode = container_of(head, struct inode, i_rcu); - struct xfs_inode *ip = XFS_I(inode); - - kmem_zone_free(xfs_inode_zone, ip); -} - -void -xfs_inode_free( - struct xfs_inode *ip) -{ - switch (ip->i_d.di_mode & S_IFMT) { - case S_IFREG: - case S_IFDIR: - case S_IFLNK: - xfs_idestroy_fork(ip, XFS_DATA_FORK); - break; - } - - if (ip->i_afp) - xfs_idestroy_fork(ip, XFS_ATTR_FORK); - - if (ip->i_itemp) { - ASSERT(!(ip->i_itemp->ili_item.li_flags & XFS_LI_IN_AIL)); - xfs_inode_item_destroy(ip); - ip->i_itemp = NULL; - } - - /* asserts to verify all state is correct here */ - ASSERT(atomic_read(&ip->i_pincount) == 0); - ASSERT(!spin_is_locked(&ip->i_flags_lock)); - ASSERT(!xfs_isiflocked(ip)); - - /* - * Because we use RCU freeing we need to ensure the inode always - * appears to be reclaimed with an invalid inode number when in the - * free state. The ip->i_flags_lock provides the barrier against lookup - * races. - */ - spin_lock(&ip->i_flags_lock); - ip->i_flags = XFS_IRECLAIM; - ip->i_ino = 0; - spin_unlock(&ip->i_flags_lock); - - call_rcu(&VFS_I(ip)->i_rcu, xfs_inode_free_callback); -} - -/* - * Check the validity of the inode we just found it the cache - */ -static int -xfs_iget_cache_hit( - struct xfs_perag *pag, - struct xfs_inode *ip, - xfs_ino_t ino, - int flags, - int lock_flags) __releases(RCU) -{ - struct inode *inode = VFS_I(ip); - struct xfs_mount *mp = ip->i_mount; - int error; - - /* - * check for re-use of an inode within an RCU grace period due to the - * radix tree nodes not being updated yet. We monitor for this by - * setting the inode number to zero before freeing the inode structure. - * If the inode has been reallocated and set up, then the inode number - * will not match, so check for that, too. - */ - spin_lock(&ip->i_flags_lock); - if (ip->i_ino != ino) { - trace_xfs_iget_skip(ip); - XFS_STATS_INC(xs_ig_frecycle); - error = EAGAIN; - goto out_error; - } - - - /* - * If we are racing with another cache hit that is currently - * instantiating this inode or currently recycling it out of - * reclaimabe state, wait for the initialisation to complete - * before continuing. - * - * XXX(hch): eventually we should do something equivalent to - * wait_on_inode to wait for these flags to be cleared - * instead of polling for it. - */ - if (ip->i_flags & (XFS_INEW|XFS_IRECLAIM)) { - trace_xfs_iget_skip(ip); - XFS_STATS_INC(xs_ig_frecycle); - error = EAGAIN; - goto out_error; - } - - /* - * If lookup is racing with unlink return an error immediately. - */ - if (ip->i_d.di_mode == 0 && !(flags & XFS_IGET_CREATE)) { - error = ENOENT; - goto out_error; - } - - /* - * If IRECLAIMABLE is set, we've torn down the VFS inode already. - * Need to carefully get it back into useable state. - */ - if (ip->i_flags & XFS_IRECLAIMABLE) { - trace_xfs_iget_reclaim(ip); - - /* - * We need to set XFS_IRECLAIM to prevent xfs_reclaim_inode - * from stomping over us while we recycle the inode. We can't - * clear the radix tree reclaimable tag yet as it requires - * pag_ici_lock to be held exclusive. - */ - ip->i_flags |= XFS_IRECLAIM; - - spin_unlock(&ip->i_flags_lock); - rcu_read_unlock(); - - error = -inode_init_always(mp->m_super, inode); - if (error) { - /* - * Re-initializing the inode failed, and we are in deep - * trouble. Try to re-add it to the reclaim list. - */ - rcu_read_lock(); - spin_lock(&ip->i_flags_lock); - - ip->i_flags &= ~(XFS_INEW | XFS_IRECLAIM); - ASSERT(ip->i_flags & XFS_IRECLAIMABLE); - trace_xfs_iget_reclaim_fail(ip); - goto out_error; - } - - spin_lock(&pag->pag_ici_lock); - spin_lock(&ip->i_flags_lock); - - /* - * Clear the per-lifetime state in the inode as we are now - * effectively a new inode and need to return to the initial - * state before reuse occurs. - */ - ip->i_flags &= ~XFS_IRECLAIM_RESET_FLAGS; - ip->i_flags |= XFS_INEW; - __xfs_inode_clear_reclaim_tag(mp, pag, ip); - inode->i_state = I_NEW; - - ASSERT(!rwsem_is_locked(&ip->i_iolock.mr_lock)); - mrlock_init(&ip->i_iolock, MRLOCK_BARRIER, "xfsio", ip->i_ino); - - spin_unlock(&ip->i_flags_lock); - spin_unlock(&pag->pag_ici_lock); - } else { - /* If the VFS inode is being torn down, pause and try again. */ - if (!igrab(inode)) { - trace_xfs_iget_skip(ip); - error = EAGAIN; - goto out_error; - } - - /* We've got a live one. */ - spin_unlock(&ip->i_flags_lock); - rcu_read_unlock(); - trace_xfs_iget_hit(ip); - } - - if (lock_flags != 0) - xfs_ilock(ip, lock_flags); - - xfs_iflags_clear(ip, XFS_ISTALE | XFS_IDONTCACHE); - XFS_STATS_INC(xs_ig_found); - - return 0; - -out_error: - spin_unlock(&ip->i_flags_lock); - rcu_read_unlock(); - return error; -} - - -static int -xfs_iget_cache_miss( - struct xfs_mount *mp, - struct xfs_perag *pag, - xfs_trans_t *tp, - xfs_ino_t ino, - struct xfs_inode **ipp, - int flags, - int lock_flags) -{ - struct xfs_inode *ip; - int error; - xfs_agino_t agino = XFS_INO_TO_AGINO(mp, ino); - int iflags; - - ip = xfs_inode_alloc(mp, ino); - if (!ip) - return ENOMEM; - - error = xfs_iread(mp, tp, ip, flags); - if (error) - goto out_destroy; - - trace_xfs_iget_miss(ip); - - if ((ip->i_d.di_mode == 0) && !(flags & XFS_IGET_CREATE)) { - error = ENOENT; - goto out_destroy; - } - - /* - * Preload the radix tree so we can insert safely under the - * write spinlock. Note that we cannot sleep inside the preload - * region. Since we can be called from transaction context, don't - * recurse into the file system. - */ - if (radix_tree_preload(GFP_NOFS)) { - error = EAGAIN; - goto out_destroy; - } - - /* - * Because the inode hasn't been added to the radix-tree yet it can't - * be found by another thread, so we can do the non-sleeping lock here. - */ - if (lock_flags) { - if (!xfs_ilock_nowait(ip, lock_flags)) - BUG(); - } - - /* - * These values must be set before inserting the inode into the radix - * tree as the moment it is inserted a concurrent lookup (allowed by the - * RCU locking mechanism) can find it and that lookup must see that this - * is an inode currently under construction (i.e. that XFS_INEW is set). - * The ip->i_flags_lock that protects the XFS_INEW flag forms the - * memory barrier that ensures this detection works correctly at lookup - * time. - */ - iflags = XFS_INEW; - if (flags & XFS_IGET_DONTCACHE) - iflags |= XFS_IDONTCACHE; - ip->i_udquot = ip->i_gdquot = NULL; - xfs_iflags_set(ip, iflags); - - /* insert the new inode */ - spin_lock(&pag->pag_ici_lock); - error = radix_tree_insert(&pag->pag_ici_root, agino, ip); - if (unlikely(error)) { - WARN_ON(error != -EEXIST); - XFS_STATS_INC(xs_ig_dup); - error = EAGAIN; - goto out_preload_end; - } - spin_unlock(&pag->pag_ici_lock); - radix_tree_preload_end(); - - *ipp = ip; - return 0; - -out_preload_end: - spin_unlock(&pag->pag_ici_lock); - radix_tree_preload_end(); - if (lock_flags) - xfs_iunlock(ip, lock_flags); -out_destroy: - __destroy_inode(VFS_I(ip)); - xfs_inode_free(ip); - return error; -} - -/* - * Look up an inode by number in the given file system. - * The inode is looked up in the cache held in each AG. - * If the inode is found in the cache, initialise the vfs inode - * if necessary. - * - * If it is not in core, read it in from the file system's device, - * add it to the cache and initialise the vfs inode. - * - * The inode is locked according to the value of the lock_flags parameter. - * This flag parameter indicates how and if the inode's IO lock and inode lock - * should be taken. - * - * mp -- the mount point structure for the current file system. It points - * to the inode hash table. - * tp -- a pointer to the current transaction if there is one. This is - * simply passed through to the xfs_iread() call. - * ino -- the number of the inode desired. This is the unique identifier - * within the file system for the inode being requested. - * lock_flags -- flags indicating how to lock the inode. See the comment - * for xfs_ilock() for a list of valid values. - */ -int -xfs_iget( - xfs_mount_t *mp, - xfs_trans_t *tp, - xfs_ino_t ino, - uint flags, - uint lock_flags, - xfs_inode_t **ipp) -{ - xfs_inode_t *ip; - int error; - xfs_perag_t *pag; - xfs_agino_t agino; - - /* - * xfs_reclaim_inode() uses the ILOCK to ensure an inode - * doesn't get freed while it's being referenced during a - * radix tree traversal here. It assumes this function - * aqcuires only the ILOCK (and therefore it has no need to - * involve the IOLOCK in this synchronization). - */ - ASSERT((lock_flags & (XFS_IOLOCK_EXCL | XFS_IOLOCK_SHARED)) == 0); - - /* reject inode numbers outside existing AGs */ - if (!ino || XFS_INO_TO_AGNO(mp, ino) >= mp->m_sb.sb_agcount) - return EINVAL; - - /* get the perag structure and ensure that it's inode capable */ - pag = xfs_perag_get(mp, XFS_INO_TO_AGNO(mp, ino)); - agino = XFS_INO_TO_AGINO(mp, ino); - -again: - error = 0; - rcu_read_lock(); - ip = radix_tree_lookup(&pag->pag_ici_root, agino); - - if (ip) { - error = xfs_iget_cache_hit(pag, ip, ino, flags, lock_flags); - if (error) - goto out_error_or_again; - } else { - rcu_read_unlock(); - XFS_STATS_INC(xs_ig_missed); - - error = xfs_iget_cache_miss(mp, pag, tp, ino, &ip, - flags, lock_flags); - if (error) - goto out_error_or_again; - } - xfs_perag_put(pag); - - *ipp = ip; - - /* - * If we have a real type for an on-disk inode, we can set ops(&unlock) - * now. If it's a new inode being created, xfs_ialloc will handle it. - */ - if (xfs_iflags_test(ip, XFS_INEW) && ip->i_d.di_mode != 0) - xfs_setup_inode(ip); - return 0; - -out_error_or_again: - if (error == EAGAIN) { - delay(1); - goto again; - } - xfs_perag_put(pag); - return error; -} - -/* - * This is a wrapper routine around the xfs_ilock() routine - * used to centralize some grungy code. It is used in places - * that wish to lock the inode solely for reading the extents. - * The reason these places can't just call xfs_ilock(SHARED) - * is that the inode lock also guards to bringing in of the - * extents from disk for a file in b-tree format. If the inode - * is in b-tree format, then we need to lock the inode exclusively - * until the extents are read in. Locking it exclusively all - * the time would limit our parallelism unnecessarily, though. - * What we do instead is check to see if the extents have been - * read in yet, and only lock the inode exclusively if they - * have not. - * - * The function returns a value which should be given to the - * corresponding xfs_iunlock_map_shared(). This value is - * the mode in which the lock was actually taken. - */ -uint -xfs_ilock_map_shared( - xfs_inode_t *ip) -{ - uint lock_mode; - - if ((ip->i_d.di_format == XFS_DINODE_FMT_BTREE) && - ((ip->i_df.if_flags & XFS_IFEXTENTS) == 0)) { - lock_mode = XFS_ILOCK_EXCL; - } else { - lock_mode = XFS_ILOCK_SHARED; - } - - xfs_ilock(ip, lock_mode); - - return lock_mode; -} - -/* - * This is simply the unlock routine to go with xfs_ilock_map_shared(). - * All it does is call xfs_iunlock() with the given lock_mode. - */ -void -xfs_iunlock_map_shared( - xfs_inode_t *ip, - unsigned int lock_mode) -{ - xfs_iunlock(ip, lock_mode); -} - -/* - * The xfs inode contains 2 locks: a multi-reader lock called the - * i_iolock and a multi-reader lock called the i_lock. This routine - * allows either or both of the locks to be obtained. - * - * The 2 locks should always be ordered so that the IO lock is - * obtained first in order to prevent deadlock. - * - * ip -- the inode being locked - * lock_flags -- this parameter indicates the inode's locks - * to be locked. It can be: - * XFS_IOLOCK_SHARED, - * XFS_IOLOCK_EXCL, - * XFS_ILOCK_SHARED, - * XFS_ILOCK_EXCL, - * XFS_IOLOCK_SHARED | XFS_ILOCK_SHARED, - * XFS_IOLOCK_SHARED | XFS_ILOCK_EXCL, - * XFS_IOLOCK_EXCL | XFS_ILOCK_SHARED, - * XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL - */ -void -xfs_ilock( - xfs_inode_t *ip, - uint lock_flags) -{ - /* - * You can't set both SHARED and EXCL for the same lock, - * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, - * and XFS_ILOCK_EXCL are valid values to set in lock_flags. - */ - ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != - (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); - ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != - (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); - ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); - - if (lock_flags & XFS_IOLOCK_EXCL) - mrupdate_nested(&ip->i_iolock, XFS_IOLOCK_DEP(lock_flags)); - else if (lock_flags & XFS_IOLOCK_SHARED) - mraccess_nested(&ip->i_iolock, XFS_IOLOCK_DEP(lock_flags)); - - if (lock_flags & XFS_ILOCK_EXCL) - mrupdate_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags)); - else if (lock_flags & XFS_ILOCK_SHARED) - mraccess_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags)); - - trace_xfs_ilock(ip, lock_flags, _RET_IP_); -} - -/* - * This is just like xfs_ilock(), except that the caller - * is guaranteed not to sleep. It returns 1 if it gets - * the requested locks and 0 otherwise. If the IO lock is - * obtained but the inode lock cannot be, then the IO lock - * is dropped before returning. - * - * ip -- the inode being locked - * lock_flags -- this parameter indicates the inode's locks to be - * to be locked. See the comment for xfs_ilock() for a list - * of valid values. - */ -int -xfs_ilock_nowait( - xfs_inode_t *ip, - uint lock_flags) -{ - /* - * You can't set both SHARED and EXCL for the same lock, - * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, - * and XFS_ILOCK_EXCL are valid values to set in lock_flags. - */ - ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != - (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); - ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != - (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); - ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); - - if (lock_flags & XFS_IOLOCK_EXCL) { - if (!mrtryupdate(&ip->i_iolock)) - goto out; - } else if (lock_flags & XFS_IOLOCK_SHARED) { - if (!mrtryaccess(&ip->i_iolock)) - goto out; - } - if (lock_flags & XFS_ILOCK_EXCL) { - if (!mrtryupdate(&ip->i_lock)) - goto out_undo_iolock; - } else if (lock_flags & XFS_ILOCK_SHARED) { - if (!mrtryaccess(&ip->i_lock)) - goto out_undo_iolock; - } - trace_xfs_ilock_nowait(ip, lock_flags, _RET_IP_); - return 1; - - out_undo_iolock: - if (lock_flags & XFS_IOLOCK_EXCL) - mrunlock_excl(&ip->i_iolock); - else if (lock_flags & XFS_IOLOCK_SHARED) - mrunlock_shared(&ip->i_iolock); - out: - return 0; -} - -/* - * xfs_iunlock() is used to drop the inode locks acquired with - * xfs_ilock() and xfs_ilock_nowait(). The caller must pass - * in the flags given to xfs_ilock() or xfs_ilock_nowait() so - * that we know which locks to drop. - * - * ip -- the inode being unlocked - * lock_flags -- this parameter indicates the inode's locks to be - * to be unlocked. See the comment for xfs_ilock() for a list - * of valid values for this parameter. - * - */ -void -xfs_iunlock( - xfs_inode_t *ip, - uint lock_flags) -{ - /* - * You can't set both SHARED and EXCL for the same lock, - * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, - * and XFS_ILOCK_EXCL are valid values to set in lock_flags. - */ - ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != - (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); - ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != - (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); - ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); - ASSERT(lock_flags != 0); - - if (lock_flags & XFS_IOLOCK_EXCL) - mrunlock_excl(&ip->i_iolock); - else if (lock_flags & XFS_IOLOCK_SHARED) - mrunlock_shared(&ip->i_iolock); - - if (lock_flags & XFS_ILOCK_EXCL) - mrunlock_excl(&ip->i_lock); - else if (lock_flags & XFS_ILOCK_SHARED) - mrunlock_shared(&ip->i_lock); - - trace_xfs_iunlock(ip, lock_flags, _RET_IP_); -} - -/* - * give up write locks. the i/o lock cannot be held nested - * if it is being demoted. - */ -void -xfs_ilock_demote( - xfs_inode_t *ip, - uint lock_flags) -{ - ASSERT(lock_flags & (XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL)); - ASSERT((lock_flags & ~(XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL)) == 0); - - if (lock_flags & XFS_ILOCK_EXCL) - mrdemote(&ip->i_lock); - if (lock_flags & XFS_IOLOCK_EXCL) - mrdemote(&ip->i_iolock); - - trace_xfs_ilock_demote(ip, lock_flags, _RET_IP_); -} - -#ifdef DEBUG -int -xfs_isilocked( - xfs_inode_t *ip, - uint lock_flags) -{ - if (lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) { - if (!(lock_flags & XFS_ILOCK_SHARED)) - return !!ip->i_lock.mr_writer; - return rwsem_is_locked(&ip->i_lock.mr_lock); - } - - if (lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) { - if (!(lock_flags & XFS_IOLOCK_SHARED)) - return !!ip->i_iolock.mr_writer; - return rwsem_is_locked(&ip->i_iolock.mr_lock); - } - - ASSERT(0); - return 0; -} -#endif - -void -__xfs_iflock( - struct xfs_inode *ip) -{ - wait_queue_head_t *wq = bit_waitqueue(&ip->i_flags, __XFS_IFLOCK_BIT); - DEFINE_WAIT_BIT(wait, &ip->i_flags, __XFS_IFLOCK_BIT); - - do { - prepare_to_wait_exclusive(wq, &wait.wait, TASK_UNINTERRUPTIBLE); - if (xfs_isiflocked(ip)) - io_schedule(); - } while (!xfs_iflock_nowait(ip)); - - finish_wait(wq, &wait.wait); -} diff --git a/fs/xfs/xfs_inode.c b/fs/xfs/xfs_inode.c index 1938b41ee9f5..66282dcb821b 100644 --- a/fs/xfs/xfs_inode.c +++ b/fs/xfs/xfs_inode.c @@ -45,6 +45,7 @@ #include "xfs_filestream.h" #include "xfs_vnodeops.h" #include "xfs_trace.h" +#include "xfs_icache.h" kmem_zone_t *xfs_ifork_zone; kmem_zone_t *xfs_inode_zone; @@ -74,6 +75,256 @@ xfs_get_extsz_hint( return 0; } +/* + * This is a wrapper routine around the xfs_ilock() routine used to centralize + * some grungy code. It is used in places that wish to lock the inode solely + * for reading the extents. The reason these places can't just call + * xfs_ilock(SHARED) is that the inode lock also guards to bringing in of the + * extents from disk for a file in b-tree format. If the inode is in b-tree + * format, then we need to lock the inode exclusively until the extents are read + * in. Locking it exclusively all the time would limit our parallelism + * unnecessarily, though. What we do instead is check to see if the extents + * have been read in yet, and only lock the inode exclusively if they have not. + * + * The function returns a value which should be given to the corresponding + * xfs_iunlock_map_shared(). This value is the mode in which the lock was + * actually taken. + */ +uint +xfs_ilock_map_shared( + xfs_inode_t *ip) +{ + uint lock_mode; + + if ((ip->i_d.di_format == XFS_DINODE_FMT_BTREE) && + ((ip->i_df.if_flags & XFS_IFEXTENTS) == 0)) { + lock_mode = XFS_ILOCK_EXCL; + } else { + lock_mode = XFS_ILOCK_SHARED; + } + + xfs_ilock(ip, lock_mode); + + return lock_mode; +} + +/* + * This is simply the unlock routine to go with xfs_ilock_map_shared(). + * All it does is call xfs_iunlock() with the given lock_mode. + */ +void +xfs_iunlock_map_shared( + xfs_inode_t *ip, + unsigned int lock_mode) +{ + xfs_iunlock(ip, lock_mode); +} + +/* + * The xfs inode contains 2 locks: a multi-reader lock called the + * i_iolock and a multi-reader lock called the i_lock. This routine + * allows either or both of the locks to be obtained. + * + * The 2 locks should always be ordered so that the IO lock is + * obtained first in order to prevent deadlock. + * + * ip -- the inode being locked + * lock_flags -- this parameter indicates the inode's locks + * to be locked. It can be: + * XFS_IOLOCK_SHARED, + * XFS_IOLOCK_EXCL, + * XFS_ILOCK_SHARED, + * XFS_ILOCK_EXCL, + * XFS_IOLOCK_SHARED | XFS_ILOCK_SHARED, + * XFS_IOLOCK_SHARED | XFS_ILOCK_EXCL, + * XFS_IOLOCK_EXCL | XFS_ILOCK_SHARED, + * XFS_IOLOCK_EXCL | XFS_ILOCK_EXCL + */ +void +xfs_ilock( + xfs_inode_t *ip, + uint lock_flags) +{ + trace_xfs_ilock(ip, lock_flags, _RET_IP_); + + /* + * You can't set both SHARED and EXCL for the same lock, + * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, + * and XFS_ILOCK_EXCL are valid values to set in lock_flags. + */ + ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != + (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); + ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != + (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); + ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); + + if (lock_flags & XFS_IOLOCK_EXCL) + mrupdate_nested(&ip->i_iolock, XFS_IOLOCK_DEP(lock_flags)); + else if (lock_flags & XFS_IOLOCK_SHARED) + mraccess_nested(&ip->i_iolock, XFS_IOLOCK_DEP(lock_flags)); + + if (lock_flags & XFS_ILOCK_EXCL) + mrupdate_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags)); + else if (lock_flags & XFS_ILOCK_SHARED) + mraccess_nested(&ip->i_lock, XFS_ILOCK_DEP(lock_flags)); +} + +/* + * This is just like xfs_ilock(), except that the caller + * is guaranteed not to sleep. It returns 1 if it gets + * the requested locks and 0 otherwise. If the IO lock is + * obtained but the inode lock cannot be, then the IO lock + * is dropped before returning. + * + * ip -- the inode being locked + * lock_flags -- this parameter indicates the inode's locks to be + * to be locked. See the comment for xfs_ilock() for a list + * of valid values. + */ +int +xfs_ilock_nowait( + xfs_inode_t *ip, + uint lock_flags) +{ + trace_xfs_ilock_nowait(ip, lock_flags, _RET_IP_); + + /* + * You can't set both SHARED and EXCL for the same lock, + * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, + * and XFS_ILOCK_EXCL are valid values to set in lock_flags. + */ + ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != + (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); + ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != + (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); + ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); + + if (lock_flags & XFS_IOLOCK_EXCL) { + if (!mrtryupdate(&ip->i_iolock)) + goto out; + } else if (lock_flags & XFS_IOLOCK_SHARED) { + if (!mrtryaccess(&ip->i_iolock)) + goto out; + } + if (lock_flags & XFS_ILOCK_EXCL) { + if (!mrtryupdate(&ip->i_lock)) + goto out_undo_iolock; + } else if (lock_flags & XFS_ILOCK_SHARED) { + if (!mrtryaccess(&ip->i_lock)) + goto out_undo_iolock; + } + return 1; + + out_undo_iolock: + if (lock_flags & XFS_IOLOCK_EXCL) + mrunlock_excl(&ip->i_iolock); + else if (lock_flags & XFS_IOLOCK_SHARED) + mrunlock_shared(&ip->i_iolock); + out: + return 0; +} + +/* + * xfs_iunlock() is used to drop the inode locks acquired with + * xfs_ilock() and xfs_ilock_nowait(). The caller must pass + * in the flags given to xfs_ilock() or xfs_ilock_nowait() so + * that we know which locks to drop. + * + * ip -- the inode being unlocked + * lock_flags -- this parameter indicates the inode's locks to be + * to be unlocked. See the comment for xfs_ilock() for a list + * of valid values for this parameter. + * + */ +void +xfs_iunlock( + xfs_inode_t *ip, + uint lock_flags) +{ + /* + * You can't set both SHARED and EXCL for the same lock, + * and only XFS_IOLOCK_SHARED, XFS_IOLOCK_EXCL, XFS_ILOCK_SHARED, + * and XFS_ILOCK_EXCL are valid values to set in lock_flags. + */ + ASSERT((lock_flags & (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)) != + (XFS_IOLOCK_SHARED | XFS_IOLOCK_EXCL)); + ASSERT((lock_flags & (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)) != + (XFS_ILOCK_SHARED | XFS_ILOCK_EXCL)); + ASSERT((lock_flags & ~(XFS_LOCK_MASK | XFS_LOCK_DEP_MASK)) == 0); + ASSERT(lock_flags != 0); + + if (lock_flags & XFS_IOLOCK_EXCL) + mrunlock_excl(&ip->i_iolock); + else if (lock_flags & XFS_IOLOCK_SHARED) + mrunlock_shared(&ip->i_iolock); + + if (lock_flags & XFS_ILOCK_EXCL) + mrunlock_excl(&ip->i_lock); + else if (lock_flags & XFS_ILOCK_SHARED) + mrunlock_shared(&ip->i_lock); + + trace_xfs_iunlock(ip, lock_flags, _RET_IP_); +} + +/* + * give up write locks. the i/o lock cannot be held nested + * if it is being demoted. + */ +void +xfs_ilock_demote( + xfs_inode_t *ip, + uint lock_flags) +{ + ASSERT(lock_flags & (XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL)); + ASSERT((lock_flags & ~(XFS_IOLOCK_EXCL|XFS_ILOCK_EXCL)) == 0); + + if (lock_flags & XFS_ILOCK_EXCL) + mrdemote(&ip->i_lock); + if (lock_flags & XFS_IOLOCK_EXCL) + mrdemote(&ip->i_iolock); + + trace_xfs_ilock_demote(ip, lock_flags, _RET_IP_); +} + +#ifdef DEBUG +int +xfs_isilocked( + xfs_inode_t *ip, + uint lock_flags) +{ + if (lock_flags & (XFS_ILOCK_EXCL|XFS_ILOCK_SHARED)) { + if (!(lock_flags & XFS_ILOCK_SHARED)) + return !!ip->i_lock.mr_writer; + return rwsem_is_locked(&ip->i_lock.mr_lock); + } + + if (lock_flags & (XFS_IOLOCK_EXCL|XFS_IOLOCK_SHARED)) { + if (!(lock_flags & XFS_IOLOCK_SHARED)) + return !!ip->i_iolock.mr_writer; + return rwsem_is_locked(&ip->i_iolock.mr_lock); + } + + ASSERT(0); + return 0; +} +#endif + +void +__xfs_iflock( + struct xfs_inode *ip) +{ + wait_queue_head_t *wq = bit_waitqueue(&ip->i_flags, __XFS_IFLOCK_BIT); + DEFINE_WAIT_BIT(wait, &ip->i_flags, __XFS_IFLOCK_BIT); + + do { + prepare_to_wait_exclusive(wq, &wait.wait, TASK_UNINTERRUPTIBLE); + if (xfs_isiflocked(ip)) + io_schedule(); + } while (!xfs_iflock_nowait(ip)); + + finish_wait(wq, &wait.wait); +} + #ifdef DEBUG /* * Make sure that the extents in the given memory buffer @@ -131,6 +382,65 @@ xfs_inobp_check( } #endif +static void +xfs_inode_buf_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + int i; + int ni; + + /* + * Validate the magic number and version of every inode in the buffer + */ + ni = XFS_BB_TO_FSB(mp, bp->b_length) * mp->m_sb.sb_inopblock; + for (i = 0; i < ni; i++) { + int di_ok; + xfs_dinode_t *dip; + + dip = (struct xfs_dinode *)xfs_buf_offset(bp, + (i << mp->m_sb.sb_inodelog)); + di_ok = dip->di_magic == cpu_to_be16(XFS_DINODE_MAGIC) && + XFS_DINODE_GOOD_VERSION(dip->di_version); + if (unlikely(XFS_TEST_ERROR(!di_ok, mp, + XFS_ERRTAG_ITOBP_INOTOBP, + XFS_RANDOM_ITOBP_INOTOBP))) { + xfs_buf_ioerror(bp, EFSCORRUPTED); + XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_HIGH, + mp, dip); +#ifdef DEBUG + xfs_emerg(mp, + "bad inode magic/vsn daddr %lld #%d (magic=%x)", + (unsigned long long)bp->b_bn, i, + be16_to_cpu(dip->di_magic)); + ASSERT(0); +#endif + } + } + xfs_inobp_check(mp, bp); +} + + +static void +xfs_inode_buf_read_verify( + struct xfs_buf *bp) +{ + xfs_inode_buf_verify(bp); +} + +static void +xfs_inode_buf_write_verify( + struct xfs_buf *bp) +{ + xfs_inode_buf_verify(bp); +} + +const struct xfs_buf_ops xfs_inode_buf_ops = { + .verify_read = xfs_inode_buf_read_verify, + .verify_write = xfs_inode_buf_write_verify, +}; + + /* * This routine is called to map an inode to the buffer containing the on-disk * version of the inode. It returns a pointer to the buffer containing the @@ -145,71 +455,33 @@ xfs_imap_to_bp( struct xfs_mount *mp, struct xfs_trans *tp, struct xfs_imap *imap, - struct xfs_dinode **dipp, + struct xfs_dinode **dipp, struct xfs_buf **bpp, uint buf_flags, uint iget_flags) { struct xfs_buf *bp; int error; - int i; - int ni; buf_flags |= XBF_UNMAPPED; error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, imap->im_blkno, - (int)imap->im_len, buf_flags, &bp); + (int)imap->im_len, buf_flags, &bp, + &xfs_inode_buf_ops); if (error) { - if (error != EAGAIN) { - xfs_warn(mp, - "%s: xfs_trans_read_buf() returned error %d.", - __func__, error); - } else { + if (error == EAGAIN) { ASSERT(buf_flags & XBF_TRYLOCK); + return error; } - return error; - } - - /* - * Validate the magic number and version of every inode in the buffer - * (if DEBUG kernel) or the first inode in the buffer, otherwise. - */ -#ifdef DEBUG - ni = BBTOB(imap->im_len) >> mp->m_sb.sb_inodelog; -#else /* usual case */ - ni = 1; -#endif - for (i = 0; i < ni; i++) { - int di_ok; - xfs_dinode_t *dip; + if (error == EFSCORRUPTED && + (iget_flags & XFS_IGET_UNTRUSTED)) + return XFS_ERROR(EINVAL); - dip = (xfs_dinode_t *)xfs_buf_offset(bp, - (i << mp->m_sb.sb_inodelog)); - di_ok = dip->di_magic == cpu_to_be16(XFS_DINODE_MAGIC) && - XFS_DINODE_GOOD_VERSION(dip->di_version); - if (unlikely(XFS_TEST_ERROR(!di_ok, mp, - XFS_ERRTAG_ITOBP_INOTOBP, - XFS_RANDOM_ITOBP_INOTOBP))) { - if (iget_flags & XFS_IGET_UNTRUSTED) { - xfs_trans_brelse(tp, bp); - return XFS_ERROR(EINVAL); - } - XFS_CORRUPTION_ERROR(__func__, XFS_ERRLEVEL_HIGH, - mp, dip); -#ifdef DEBUG - xfs_emerg(mp, - "bad inode magic/vsn daddr %lld #%d (magic=%x)", - (unsigned long long)imap->im_blkno, i, - be16_to_cpu(dip->di_magic)); - ASSERT(0); -#endif - xfs_trans_brelse(tp, bp); - return XFS_ERROR(EFSCORRUPTED); - } + xfs_warn(mp, "%s: xfs_trans_read_buf() returned error %d.", + __func__, error); + return error; } - xfs_inobp_check(mp, bp); - *bpp = bp; *dipp = (struct xfs_dinode *)xfs_buf_offset(bp, imap->im_boffset); return 0; @@ -853,16 +1125,16 @@ xfs_iread_extents( * set according to the contents of the given cred structure. * * Use xfs_dialloc() to allocate the on-disk inode. If xfs_dialloc() - * has a free inode available, call xfs_iget() - * to obtain the in-core version of the allocated inode. Finally, - * fill in the inode and log its initial contents. In this case, - * ialloc_context would be set to NULL and call_again set to false. + * has a free inode available, call xfs_iget() to obtain the in-core + * version of the allocated inode. Finally, fill in the inode and + * log its initial contents. In this case, ialloc_context would be + * set to NULL. * - * If xfs_dialloc() does not have an available inode, - * it will replenish its supply by doing an allocation. Since we can - * only do one allocation within a transaction without deadlocks, we - * must commit the current transaction before returning the inode itself. - * In this case, therefore, we will set call_again to true and return. + * If xfs_dialloc() does not have an available inode, it will replenish + * its supply by doing an allocation. Since we can only do one + * allocation within a transaction without deadlocks, we must commit + * the current transaction before returning the inode itself. + * In this case, therefore, we will set ialloc_context and return. * The caller should then commit the current transaction, start a new * transaction, and call xfs_ialloc() again to actually get the inode. * @@ -1514,6 +1786,18 @@ xfs_ifree_cluster( if (!bp) return ENOMEM; + + /* + * This buffer may not have been correctly initialised as we + * didn't read it from disk. That's not important because we are + * only using to mark the buffer as stale in the log, and to + * attach stale cached inodes on it. That means it will never be + * dispatched for IO. If it is, we want to know about it, and we + * want it to fail. We can acheive this by adding a write + * verifier to the buffer. + */ + bp->b_ops = &xfs_inode_buf_ops; + /* * Walk the inodes already attached to the buffer and mark them * stale. These will all have the flush locks held, so an @@ -3661,3 +3945,40 @@ xfs_iext_irec_update_extoffs( ifp->if_u1.if_ext_irec[i].er_extoff += ext_diff; } } + +/* + * Test whether it is appropriate to check an inode for and free post EOF + * blocks. The 'force' parameter determines whether we should also consider + * regular files that are marked preallocated or append-only. + */ +bool +xfs_can_free_eofblocks(struct xfs_inode *ip, bool force) +{ + /* prealloc/delalloc exists only on regular files */ + if (!S_ISREG(ip->i_d.di_mode)) + return false; + + /* + * Zero sized files with no cached pages and delalloc blocks will not + * have speculative prealloc/delalloc blocks to remove. + */ + if (VFS_I(ip)->i_size == 0 && + VN_CACHED(VFS_I(ip)) == 0 && + ip->i_delayed_blks == 0) + return false; + + /* If we haven't read in the extent list, then don't do it now. */ + if (!(ip->i_df.if_flags & XFS_IFEXTENTS)) + return false; + + /* + * Do not free real preallocated or append-only files unless the file + * has delalloc blocks and we are forced to remove them. + */ + if (ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)) + if (!force || ip->i_delayed_blks == 0) + return false; + + return true; +} + diff --git a/fs/xfs/xfs_inode.h b/fs/xfs/xfs_inode.h index 94b32f906e79..22baf6ea4fac 100644 --- a/fs/xfs/xfs_inode.h +++ b/fs/xfs/xfs_inode.h @@ -496,11 +496,10 @@ static inline int xfs_isiflocked(struct xfs_inode *ip) (((pip)->i_mount->m_flags & XFS_MOUNT_GRPID) || \ ((pip)->i_d.di_mode & S_ISGID)) + /* - * xfs_iget.c prototypes. + * xfs_inode.c prototypes. */ -int xfs_iget(struct xfs_mount *, struct xfs_trans *, xfs_ino_t, - uint, uint, xfs_inode_t **); void xfs_ilock(xfs_inode_t *, uint); int xfs_ilock_nowait(xfs_inode_t *, uint); void xfs_iunlock(xfs_inode_t *, uint); @@ -508,11 +507,6 @@ void xfs_ilock_demote(xfs_inode_t *, uint); int xfs_isilocked(xfs_inode_t *, uint); uint xfs_ilock_map_shared(xfs_inode_t *); void xfs_iunlock_map_shared(xfs_inode_t *, uint); -void xfs_inode_free(struct xfs_inode *ip); - -/* - * xfs_inode.c prototypes. - */ int xfs_ialloc(struct xfs_trans *, xfs_inode_t *, umode_t, xfs_nlink_t, xfs_dev_t, prid_t, int, struct xfs_buf **, xfs_inode_t **); @@ -591,6 +585,7 @@ void xfs_iext_irec_compact(xfs_ifork_t *); void xfs_iext_irec_compact_pages(xfs_ifork_t *); void xfs_iext_irec_compact_full(xfs_ifork_t *); void xfs_iext_irec_update_extoffs(xfs_ifork_t *, int, int); +bool xfs_can_free_eofblocks(struct xfs_inode *, bool); #define xfs_ipincount(ip) ((unsigned int) atomic_read(&ip->i_pincount)) @@ -603,5 +598,6 @@ void xfs_inobp_check(struct xfs_mount *, struct xfs_buf *); extern struct kmem_zone *xfs_ifork_zone; extern struct kmem_zone *xfs_inode_zone; extern struct kmem_zone *xfs_ili_zone; +extern const struct xfs_buf_ops xfs_inode_buf_ops; #endif /* __XFS_INODE_H__ */ diff --git a/fs/xfs/xfs_ioctl.c b/fs/xfs/xfs_ioctl.c index c1df3c623de2..c1c3ef88a260 100644 --- a/fs/xfs/xfs_ioctl.c +++ b/fs/xfs/xfs_ioctl.c @@ -42,6 +42,7 @@ #include "xfs_inode_item.h" #include "xfs_export.h" #include "xfs_trace.h" +#include "xfs_icache.h" #include <linux/capability.h> #include <linux/dcache.h> @@ -1602,6 +1603,26 @@ xfs_file_ioctl( error = xfs_errortag_clearall(mp, 1); return -error; + case XFS_IOC_FREE_EOFBLOCKS: { + struct xfs_eofblocks eofb; + + if (copy_from_user(&eofb, arg, sizeof(eofb))) + return -XFS_ERROR(EFAULT); + + if (eofb.eof_version != XFS_EOFBLOCKS_VERSION) + return -XFS_ERROR(EINVAL); + + if (eofb.eof_flags & ~XFS_EOF_FLAGS_VALID) + return -XFS_ERROR(EINVAL); + + if (memchr_inv(&eofb.pad32, 0, sizeof(eofb.pad32)) || + memchr_inv(eofb.pad64, 0, sizeof(eofb.pad64))) + return -XFS_ERROR(EINVAL); + + error = xfs_icache_free_eofblocks(mp, &eofb); + return -error; + } + default: return -ENOTTY; } diff --git a/fs/xfs/xfs_iomap.c b/fs/xfs/xfs_iomap.c index 7f537663365b..add06b4e9a63 100644 --- a/fs/xfs/xfs_iomap.c +++ b/fs/xfs/xfs_iomap.c @@ -41,6 +41,7 @@ #include "xfs_utils.h" #include "xfs_iomap.h" #include "xfs_trace.h" +#include "xfs_icache.h" #define XFS_WRITEIO_ALIGN(mp,off) (((off) >> mp->m_writeio_log) \ @@ -373,7 +374,7 @@ xfs_iomap_write_delay( xfs_extlen_t extsz; int nimaps; xfs_bmbt_irec_t imap[XFS_WRITE_IMAPS]; - int prealloc, flushed = 0; + int prealloc; int error; ASSERT(xfs_isilocked(ip, XFS_ILOCK_EXCL)); @@ -434,31 +435,29 @@ retry: } /* - * If bmapi returned us nothing, we got either ENOSPC or EDQUOT. For - * ENOSPC, * flush all other inodes with delalloc blocks to free up - * some of the excess reserved metadata space. For both cases, retry + * If bmapi returned us nothing, we got either ENOSPC or EDQUOT. Retry * without EOF preallocation. */ if (nimaps == 0) { trace_xfs_delalloc_enospc(ip, offset, count); - if (flushed) - return XFS_ERROR(error ? error : ENOSPC); - - if (error == ENOSPC) { - xfs_iunlock(ip, XFS_ILOCK_EXCL); - xfs_flush_inodes(ip); - xfs_ilock(ip, XFS_ILOCK_EXCL); + if (prealloc) { + prealloc = 0; + error = 0; + goto retry; } - - flushed = 1; - error = 0; - prealloc = 0; - goto retry; + return XFS_ERROR(error ? error : ENOSPC); } if (!(imap[0].br_startblock || XFS_IS_REALTIME_INODE(ip))) return xfs_alert_fsblock_zero(ip, &imap[0]); + /* + * Tag the inode as speculatively preallocated so we can reclaim this + * space on demand, if necessary. + */ + if (prealloc) + xfs_inode_set_eofblocks_tag(ip); + *ret_imap = imap[0]; return 0; } diff --git a/fs/xfs/xfs_iops.c b/fs/xfs/xfs_iops.c index 4e00cf091d2c..d82efaa2ac73 100644 --- a/fs/xfs/xfs_iops.c +++ b/fs/xfs/xfs_iops.c @@ -38,6 +38,7 @@ #include "xfs_vnodeops.h" #include "xfs_inode_item.h" #include "xfs_trace.h" +#include "xfs_icache.h" #include <linux/capability.h> #include <linux/xattr.h> @@ -779,8 +780,8 @@ xfs_setattr_size( * care about here. */ if (oldsize != ip->i_d.di_size && newsize > ip->i_d.di_size) { - error = xfs_flush_pages(ip, ip->i_d.di_size, newsize, 0, - FI_NONE); + error = -filemap_write_and_wait_range(VFS_I(ip)->i_mapping, + ip->i_d.di_size, newsize); if (error) goto out_unlock; } @@ -854,6 +855,9 @@ xfs_setattr_size( * and do not wait the usual (long) time for writeout. */ xfs_iflags_set(ip, XFS_ITRUNCATED); + + /* A truncate down always removes post-EOF blocks. */ + xfs_inode_clear_eofblocks_tag(ip); } if (mask & ATTR_CTIME) { diff --git a/fs/xfs/xfs_itable.c b/fs/xfs/xfs_itable.c index 01d10a66e302..2ea7d402188d 100644 --- a/fs/xfs/xfs_itable.c +++ b/fs/xfs/xfs_itable.c @@ -34,6 +34,7 @@ #include "xfs_error.h" #include "xfs_btree.h" #include "xfs_trace.h" +#include "xfs_icache.h" STATIC int xfs_internal_inum( @@ -395,7 +396,8 @@ xfs_bulkstat( if (xfs_inobt_maskn(chunkidx, nicluster) & ~r.ir_free) xfs_btree_reada_bufs(mp, agno, - agbno, nbcluster); + agbno, nbcluster, + &xfs_inode_buf_ops); } irbp->ir_startino = r.ir_startino; irbp->ir_freecount = r.ir_freecount; diff --git a/fs/xfs/xfs_linux.h b/fs/xfs/xfs_linux.h index 828662f70d64..fe7e4df85a7b 100644 --- a/fs/xfs/xfs_linux.h +++ b/fs/xfs/xfs_linux.h @@ -44,6 +44,7 @@ #include <linux/kernel.h> #include <linux/blkdev.h> #include <linux/slab.h> +#include <linux/crc32c.h> #include <linux/module.h> #include <linux/mutex.h> #include <linux/file.h> @@ -118,6 +119,7 @@ #define xfs_rotorstep xfs_params.rotorstep.val #define xfs_inherit_nodefrag xfs_params.inherit_nodfrg.val #define xfs_fstrm_centisecs xfs_params.fstrm_timer.val +#define xfs_eofb_secs xfs_params.eofb_timer.val #define current_cpu() (raw_smp_processor_id()) #define current_pid() (current->pid) diff --git a/fs/xfs/xfs_log.c b/fs/xfs/xfs_log.c index 4dad756962d0..46bd9d52ab51 100644 --- a/fs/xfs/xfs_log.c +++ b/fs/xfs/xfs_log.c @@ -34,6 +34,8 @@ #include "xfs_dinode.h" #include "xfs_inode.h" #include "xfs_trace.h" +#include "xfs_fsops.h" +#include "xfs_cksum.h" kmem_zone_t *xfs_log_ticket_zone; @@ -458,7 +460,8 @@ xfs_log_reserve( tic->t_trans_type = t_type; *ticp = tic; - xlog_grant_push_ail(log, tic->t_unit_res * tic->t_cnt); + xlog_grant_push_ail(log, tic->t_cnt ? tic->t_unit_res * tic->t_cnt + : tic->t_unit_res); trace_xfs_log_reserve(log, tic); @@ -679,25 +682,29 @@ out: } /* - * Finish the recovery of the file system. This is separate from - * the xfs_log_mount() call, because it depends on the code in - * xfs_mountfs() to read in the root and real-time bitmap inodes - * between calling xfs_log_mount() and here. + * Finish the recovery of the file system. This is separate from the + * xfs_log_mount() call, because it depends on the code in xfs_mountfs() to read + * in the root and real-time bitmap inodes between calling xfs_log_mount() and + * here. * - * mp - ubiquitous xfs mount point structure + * If we finish recovery successfully, start the background log work. If we are + * not doing recovery, then we have a RO filesystem and we don't need to start + * it. */ int xfs_log_mount_finish(xfs_mount_t *mp) { - int error; + int error = 0; - if (!(mp->m_flags & XFS_MOUNT_NORECOVERY)) + if (!(mp->m_flags & XFS_MOUNT_NORECOVERY)) { error = xlog_recover_finish(mp->m_log); - else { - error = 0; + if (!error) + xfs_log_work_queue(mp); + } else { ASSERT(mp->m_flags & XFS_MOUNT_RDONLY); } + return error; } @@ -850,15 +857,49 @@ xfs_log_unmount_write(xfs_mount_t *mp) } /* xfs_log_unmount_write */ /* - * Deallocate log structures for unmount/relocation. + * Empty the log for unmount/freeze. + * + * To do this, we first need to shut down the background log work so it is not + * trying to cover the log as we clean up. We then need to unpin all objects in + * the log so we can then flush them out. Once they have completed their IO and + * run the callbacks removing themselves from the AIL, we can write the unmount + * record. + */ +void +xfs_log_quiesce( + struct xfs_mount *mp) +{ + cancel_delayed_work_sync(&mp->m_log->l_work); + xfs_log_force(mp, XFS_LOG_SYNC); + + /* + * The superblock buffer is uncached and while xfs_ail_push_all_sync() + * will push it, xfs_wait_buftarg() will not wait for it. Further, + * xfs_buf_iowait() cannot be used because it was pushed with the + * XBF_ASYNC flag set, so we need to use a lock/unlock pair to wait for + * the IO to complete. + */ + xfs_ail_push_all_sync(mp->m_ail); + xfs_wait_buftarg(mp->m_ddev_targp); + xfs_buf_lock(mp->m_sb_bp); + xfs_buf_unlock(mp->m_sb_bp); + + xfs_log_unmount_write(mp); +} + +/* + * Shut down and release the AIL and Log. * - * We need to stop the aild from running before we destroy - * and deallocate the log as the aild references the log. + * During unmount, we need to ensure we flush all the dirty metadata objects + * from the AIL so that the log is empty before we write the unmount record to + * the log. Once this is done, we can tear down the AIL and the log. */ void -xfs_log_unmount(xfs_mount_t *mp) +xfs_log_unmount( + struct xfs_mount *mp) { - cancel_delayed_work_sync(&mp->m_sync_work); + xfs_log_quiesce(mp); + xfs_trans_ail_destroy(mp); xlog_dealloc_log(mp->m_log); } @@ -1090,8 +1131,7 @@ xlog_iodone(xfs_buf_t *bp) * with it being freed after writing the unmount record to the * log. */ - -} /* xlog_iodone */ +} /* * Return size of each in-core log record buffer. @@ -1161,6 +1201,40 @@ done: } /* xlog_get_iclog_buffer_size */ +void +xfs_log_work_queue( + struct xfs_mount *mp) +{ + queue_delayed_work(mp->m_log_workqueue, &mp->m_log->l_work, + msecs_to_jiffies(xfs_syncd_centisecs * 10)); +} + +/* + * Every sync period we need to unpin all items in the AIL and push them to + * disk. If there is nothing dirty, then we might need to cover the log to + * indicate that the filesystem is idle. + */ +void +xfs_log_worker( + struct work_struct *work) +{ + struct xlog *log = container_of(to_delayed_work(work), + struct xlog, l_work); + struct xfs_mount *mp = log->l_mp; + + /* dgc: errors ignored - not fatal and nowhere to report them */ + if (xfs_log_need_covered(mp)) + xfs_fs_log_dummy(mp); + else + xfs_log_force(mp, 0); + + /* start pushing all the metadata that is currently dirty */ + xfs_ail_push_all(mp->m_ail); + + /* queue us up again */ + xfs_log_work_queue(mp); +} + /* * This routine initializes some of the log structure for a given mount point. * Its primary purpose is to fill in enough, so recovery can occur. However, @@ -1195,6 +1269,7 @@ xlog_alloc_log( log->l_logBBsize = num_bblks; log->l_covered_state = XLOG_STATE_COVER_IDLE; log->l_flags |= XLOG_ACTIVE_RECOVERY; + INIT_DELAYED_WORK(&log->l_work, xfs_log_worker); log->l_prev_block = -1; /* log->l_tail_lsn = 0x100000000LL; cycle = 1; current block = 0 */ @@ -1417,6 +1492,84 @@ xlog_grant_push_ail( } /* + * Stamp cycle number in every block + */ +STATIC void +xlog_pack_data( + struct xlog *log, + struct xlog_in_core *iclog, + int roundoff) +{ + int i, j, k; + int size = iclog->ic_offset + roundoff; + __be32 cycle_lsn; + xfs_caddr_t dp; + + cycle_lsn = CYCLE_LSN_DISK(iclog->ic_header.h_lsn); + + dp = iclog->ic_datap; + for (i = 0; i < BTOBB(size); i++) { + if (i >= (XLOG_HEADER_CYCLE_SIZE / BBSIZE)) + break; + iclog->ic_header.h_cycle_data[i] = *(__be32 *)dp; + *(__be32 *)dp = cycle_lsn; + dp += BBSIZE; + } + + if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) { + xlog_in_core_2_t *xhdr = iclog->ic_data; + + for ( ; i < BTOBB(size); i++) { + j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE); + k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE); + xhdr[j].hic_xheader.xh_cycle_data[k] = *(__be32 *)dp; + *(__be32 *)dp = cycle_lsn; + dp += BBSIZE; + } + + for (i = 1; i < log->l_iclog_heads; i++) + xhdr[i].hic_xheader.xh_cycle = cycle_lsn; + } +} + +/* + * Calculate the checksum for a log buffer. + * + * This is a little more complicated than it should be because the various + * headers and the actual data are non-contiguous. + */ +__le32 +xlog_cksum( + struct xlog *log, + struct xlog_rec_header *rhead, + char *dp, + int size) +{ + __uint32_t crc; + + /* first generate the crc for the record header ... */ + crc = xfs_start_cksum((char *)rhead, + sizeof(struct xlog_rec_header), + offsetof(struct xlog_rec_header, h_crc)); + + /* ... then for additional cycle data for v2 logs ... */ + if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) { + union xlog_in_core2 *xhdr = (union xlog_in_core2 *)rhead; + int i; + + for (i = 1; i < log->l_iclog_heads; i++) { + crc = crc32c(crc, &xhdr[i].hic_xheader, + sizeof(struct xlog_rec_ext_header)); + } + } + + /* ... and finally for the payload */ + crc = crc32c(crc, dp, size); + + return xfs_end_cksum(crc); +} + +/* * The bdstrat callback function for log bufs. This gives us a central * place to trap bufs in case we get hit by a log I/O error and need to * shutdown. Actually, in practice, even when we didn't get a log error, @@ -1476,7 +1629,6 @@ xlog_sync( struct xlog *log, struct xlog_in_core *iclog) { - xfs_caddr_t dptr; /* pointer to byte sized element */ xfs_buf_t *bp; int i; uint count; /* byte count of bwrite */ @@ -1485,6 +1637,7 @@ xlog_sync( int split = 0; /* split write into two regions */ int error; int v2 = xfs_sb_version_haslogv2(&log->l_mp->m_sb); + int size; XFS_STATS_INC(xs_log_writes); ASSERT(atomic_read(&iclog->ic_refcnt) == 0); @@ -1515,13 +1668,10 @@ xlog_sync( xlog_pack_data(log, iclog, roundoff); /* real byte length */ - if (v2) { - iclog->ic_header.h_len = - cpu_to_be32(iclog->ic_offset + roundoff); - } else { - iclog->ic_header.h_len = - cpu_to_be32(iclog->ic_offset); - } + size = iclog->ic_offset; + if (v2) + size += roundoff; + iclog->ic_header.h_len = cpu_to_be32(size); bp = iclog->ic_bp; XFS_BUF_SET_ADDR(bp, BLOCK_LSN(be64_to_cpu(iclog->ic_header.h_lsn))); @@ -1530,12 +1680,36 @@ xlog_sync( /* Do we need to split this write into 2 parts? */ if (XFS_BUF_ADDR(bp) + BTOBB(count) > log->l_logBBsize) { + char *dptr; + split = count - (BBTOB(log->l_logBBsize - XFS_BUF_ADDR(bp))); count = BBTOB(log->l_logBBsize - XFS_BUF_ADDR(bp)); - iclog->ic_bwritecnt = 2; /* split into 2 writes */ + iclog->ic_bwritecnt = 2; + + /* + * Bump the cycle numbers at the start of each block in the + * part of the iclog that ends up in the buffer that gets + * written to the start of the log. + * + * Watch out for the header magic number case, though. + */ + dptr = (char *)&iclog->ic_header + count; + for (i = 0; i < split; i += BBSIZE) { + __uint32_t cycle = be32_to_cpu(*(__be32 *)dptr); + if (++cycle == XLOG_HEADER_MAGIC_NUM) + cycle++; + *(__be32 *)dptr = cpu_to_be32(cycle); + + dptr += BBSIZE; + } } else { iclog->ic_bwritecnt = 1; } + + /* calculcate the checksum */ + iclog->ic_header.h_crc = xlog_cksum(log, &iclog->ic_header, + iclog->ic_datap, size); + bp->b_io_length = BTOBB(count); bp->b_fspriv = iclog; XFS_BUF_ZEROFLAGS(bp); @@ -1589,19 +1763,6 @@ xlog_sync( bp->b_flags |= XBF_SYNCIO; if (log->l_mp->m_flags & XFS_MOUNT_BARRIER) bp->b_flags |= XBF_FUA; - dptr = bp->b_addr; - /* - * Bump the cycle numbers at the start of each block - * since this part of the buffer is at the start of - * a new cycle. Watch out for the header magic number - * case, though. - */ - for (i = 0; i < split; i += BBSIZE) { - be32_add_cpu((__be32 *)dptr, 1); - if (be32_to_cpu(*(__be32 *)dptr) == XLOG_HEADER_MAGIC_NUM) - be32_add_cpu((__be32 *)dptr, 1); - dptr += BBSIZE; - } ASSERT(XFS_BUF_ADDR(bp) <= log->l_logBBsize-1); ASSERT(XFS_BUF_ADDR(bp) + BTOBB(count) <= log->l_logBBsize); @@ -1618,7 +1779,6 @@ xlog_sync( return 0; } /* xlog_sync */ - /* * Deallocate a log structure */ @@ -3713,3 +3873,4 @@ xlog_iclogs_empty( } while (iclog != log->l_iclog); return 1; } + diff --git a/fs/xfs/xfs_log.h b/fs/xfs/xfs_log.h index 748d312850e2..5caee96059df 100644 --- a/fs/xfs/xfs_log.h +++ b/fs/xfs/xfs_log.h @@ -181,5 +181,9 @@ int xfs_log_commit_cil(struct xfs_mount *mp, struct xfs_trans *tp, xfs_lsn_t *commit_lsn, int flags); bool xfs_log_item_in_current_chkpt(struct xfs_log_item *lip); +void xfs_log_work_queue(struct xfs_mount *mp); +void xfs_log_worker(struct work_struct *work); +void xfs_log_quiesce(struct xfs_mount *mp); + #endif #endif /* __XFS_LOG_H__ */ diff --git a/fs/xfs/xfs_log_priv.h b/fs/xfs/xfs_log_priv.h index 18a801d76a42..16d8d12ea3b4 100644 --- a/fs/xfs/xfs_log_priv.h +++ b/fs/xfs/xfs_log_priv.h @@ -139,7 +139,6 @@ static inline uint xlog_get_client_id(__be32 i) /* * Flags for log structure */ -#define XLOG_CHKSUM_MISMATCH 0x1 /* used only during recovery */ #define XLOG_ACTIVE_RECOVERY 0x2 /* in the middle of recovery */ #define XLOG_RECOVERY_NEEDED 0x4 /* log was recovered */ #define XLOG_IO_ERROR 0x8 /* log hit an I/O error, and being @@ -291,7 +290,7 @@ typedef struct xlog_rec_header { __be32 h_len; /* len in bytes; should be 64-bit aligned: 4 */ __be64 h_lsn; /* lsn of this LR : 8 */ __be64 h_tail_lsn; /* lsn of 1st LR w/ buffers not committed: 8 */ - __be32 h_chksum; /* may not be used; non-zero if used : 4 */ + __le32 h_crc; /* crc of log record : 4 */ __be32 h_prev_block; /* block number to previous LR : 4 */ __be32 h_num_logops; /* number of log operations in this LR : 4 */ __be32 h_cycle_data[XLOG_HEADER_CYCLE_SIZE / BBSIZE]; @@ -495,6 +494,7 @@ struct xlog { struct xfs_buf *l_xbuf; /* extra buffer for log * wrapping */ struct xfs_buftarg *l_targ; /* buftarg of log */ + struct delayed_work l_work; /* background flush work */ uint l_flags; uint l_quotaoffs_flag; /* XFS_DQ_*, for QUOTAOFFs */ struct list_head *l_buf_cancel_table; @@ -554,11 +554,9 @@ xlog_recover( extern int xlog_recover_finish( struct xlog *log); -extern void -xlog_pack_data( - struct xlog *log, - struct xlog_in_core *iclog, - int); + +extern __le32 xlog_cksum(struct xlog *log, struct xlog_rec_header *rhead, + char *dp, int size); extern kmem_zone_t *xfs_log_ticket_zone; struct xlog_ticket * diff --git a/fs/xfs/xfs_log_recover.c b/fs/xfs/xfs_log_recover.c index d308749fabf1..96fcbb85ff83 100644 --- a/fs/xfs/xfs_log_recover.c +++ b/fs/xfs/xfs_log_recover.c @@ -41,7 +41,9 @@ #include "xfs_trans_priv.h" #include "xfs_quota.h" #include "xfs_utils.h" +#include "xfs_cksum.h" #include "xfs_trace.h" +#include "xfs_icache.h" STATIC int xlog_find_zeroed( @@ -2143,7 +2145,7 @@ xlog_recover_buffer_pass2( buf_flags |= XBF_UNMAPPED; bp = xfs_buf_read(mp->m_ddev_targp, buf_f->blf_blkno, buf_f->blf_len, - buf_flags); + buf_flags, NULL); if (!bp) return XFS_ERROR(ENOMEM); error = bp->b_error; @@ -2236,7 +2238,8 @@ xlog_recover_inode_pass2( } trace_xfs_log_recover_inode_recover(log, in_f); - bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len, 0); + bp = xfs_buf_read(mp->m_ddev_targp, in_f->ilf_blkno, in_f->ilf_len, 0, + NULL); if (!bp) { error = ENOMEM; goto error; @@ -2547,7 +2550,8 @@ xlog_recover_dquot_pass2( ASSERT(dq_f->qlf_len == 1); error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, dq_f->qlf_blkno, - XFS_FSB_TO_BB(mp, dq_f->qlf_len), 0, &bp); + XFS_FSB_TO_BB(mp, dq_f->qlf_len), 0, &bp, + NULL); if (error) return error; @@ -3213,80 +3217,58 @@ xlog_recover_process_iunlinks( mp->m_dmevmask = mp_dmevmask; } - -#ifdef DEBUG -STATIC void -xlog_pack_data_checksum( - struct xlog *log, - struct xlog_in_core *iclog, - int size) -{ - int i; - __be32 *up; - uint chksum = 0; - - up = (__be32 *)iclog->ic_datap; - /* divide length by 4 to get # words */ - for (i = 0; i < (size >> 2); i++) { - chksum ^= be32_to_cpu(*up); - up++; - } - iclog->ic_header.h_chksum = cpu_to_be32(chksum); -} -#else -#define xlog_pack_data_checksum(log, iclog, size) -#endif - /* - * Stamp cycle number in every block + * Upack the log buffer data and crc check it. If the check fails, issue a + * warning if and only if the CRC in the header is non-zero. This makes the + * check an advisory warning, and the zero CRC check will prevent failure + * warnings from being emitted when upgrading the kernel from one that does not + * add CRCs by default. + * + * When filesystems are CRC enabled, this CRC mismatch becomes a fatal log + * corruption failure */ -void -xlog_pack_data( - struct xlog *log, - struct xlog_in_core *iclog, - int roundoff) +STATIC int +xlog_unpack_data_crc( + struct xlog_rec_header *rhead, + xfs_caddr_t dp, + struct xlog *log) { - int i, j, k; - int size = iclog->ic_offset + roundoff; - __be32 cycle_lsn; - xfs_caddr_t dp; - - xlog_pack_data_checksum(log, iclog, size); - - cycle_lsn = CYCLE_LSN_DISK(iclog->ic_header.h_lsn); - - dp = iclog->ic_datap; - for (i = 0; i < BTOBB(size) && - i < (XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++) { - iclog->ic_header.h_cycle_data[i] = *(__be32 *)dp; - *(__be32 *)dp = cycle_lsn; - dp += BBSIZE; - } - - if (xfs_sb_version_haslogv2(&log->l_mp->m_sb)) { - xlog_in_core_2_t *xhdr = iclog->ic_data; - - for ( ; i < BTOBB(size); i++) { - j = i / (XLOG_HEADER_CYCLE_SIZE / BBSIZE); - k = i % (XLOG_HEADER_CYCLE_SIZE / BBSIZE); - xhdr[j].hic_xheader.xh_cycle_data[k] = *(__be32 *)dp; - *(__be32 *)dp = cycle_lsn; - dp += BBSIZE; + __le32 crc; + + crc = xlog_cksum(log, rhead, dp, be32_to_cpu(rhead->h_len)); + if (crc != rhead->h_crc) { + if (rhead->h_crc || xfs_sb_version_hascrc(&log->l_mp->m_sb)) { + xfs_alert(log->l_mp, + "log record CRC mismatch: found 0x%x, expected 0x%x.\n", + le32_to_cpu(rhead->h_crc), + le32_to_cpu(crc)); + xfs_hex_dump(dp, 32); } - for (i = 1; i < log->l_iclog_heads; i++) { - xhdr[i].hic_xheader.xh_cycle = cycle_lsn; - } + /* + * If we've detected a log record corruption, then we can't + * recover past this point. Abort recovery if we are enforcing + * CRC protection by punting an error back up the stack. + */ + if (xfs_sb_version_hascrc(&log->l_mp->m_sb)) + return EFSCORRUPTED; } + + return 0; } -STATIC void +STATIC int xlog_unpack_data( struct xlog_rec_header *rhead, xfs_caddr_t dp, struct xlog *log) { int i, j, k; + int error; + + error = xlog_unpack_data_crc(rhead, dp, log); + if (error) + return error; for (i = 0; i < BTOBB(be32_to_cpu(rhead->h_len)) && i < (XLOG_HEADER_CYCLE_SIZE / BBSIZE); i++) { @@ -3303,6 +3285,8 @@ xlog_unpack_data( dp += BBSIZE; } } + + return 0; } STATIC int @@ -3434,9 +3418,13 @@ xlog_do_recovery_pass( if (error) goto bread_err2; - xlog_unpack_data(rhead, offset, log); - if ((error = xlog_recover_process_data(log, - rhash, rhead, offset, pass))) + error = xlog_unpack_data(rhead, offset, log); + if (error) + goto bread_err2; + + error = xlog_recover_process_data(log, + rhash, rhead, offset, pass); + if (error) goto bread_err2; blk_no += bblks + hblks; } @@ -3546,9 +3534,14 @@ xlog_do_recovery_pass( if (error) goto bread_err2; } - xlog_unpack_data(rhead, offset, log); - if ((error = xlog_recover_process_data(log, rhash, - rhead, offset, pass))) + + error = xlog_unpack_data(rhead, offset, log); + if (error) + goto bread_err2; + + error = xlog_recover_process_data(log, rhash, + rhead, offset, pass); + if (error) goto bread_err2; blk_no += bblks; } @@ -3573,9 +3566,13 @@ xlog_do_recovery_pass( if (error) goto bread_err2; - xlog_unpack_data(rhead, offset, log); - if ((error = xlog_recover_process_data(log, rhash, - rhead, offset, pass))) + error = xlog_unpack_data(rhead, offset, log); + if (error) + goto bread_err2; + + error = xlog_recover_process_data(log, rhash, + rhead, offset, pass); + if (error) goto bread_err2; blk_no += bblks + hblks; } @@ -3689,13 +3686,14 @@ xlog_do_recover( /* * Now that we've finished replaying all buffer and inode - * updates, re-read in the superblock. + * updates, re-read in the superblock and reverify it. */ bp = xfs_getsb(log->l_mp, 0); XFS_BUF_UNDONE(bp); ASSERT(!(XFS_BUF_ISWRITE(bp))); XFS_BUF_READ(bp); XFS_BUF_UNASYNC(bp); + bp->b_ops = &xfs_sb_buf_ops; xfsbdstrat(log->l_mp, bp); error = xfs_buf_iowait(bp); if (error) { @@ -3707,7 +3705,7 @@ xlog_do_recover( /* Convert superblock from on-disk format */ sbp = &log->l_mp->m_sb; - xfs_sb_from_disk(log->l_mp, XFS_BUF_TO_SBP(bp)); + xfs_sb_from_disk(sbp, XFS_BUF_TO_SBP(bp)); ASSERT(sbp->sb_magicnum == XFS_SB_MAGIC); ASSERT(xfs_sb_good_version(sbp)); xfs_buf_relse(bp); diff --git a/fs/xfs/xfs_mount.c b/fs/xfs/xfs_mount.c index b2bd3a0e6376..da508463ff10 100644 --- a/fs/xfs/xfs_mount.c +++ b/fs/xfs/xfs_mount.c @@ -42,6 +42,7 @@ #include "xfs_fsops.h" #include "xfs_utils.h" #include "xfs_trace.h" +#include "xfs_icache.h" #ifdef HAVE_PERCPU_SB @@ -303,9 +304,8 @@ STATIC int xfs_mount_validate_sb( xfs_mount_t *mp, xfs_sb_t *sbp, - int flags) + bool check_inprogress) { - int loud = !(flags & XFS_MFSI_QUIET); /* * If the log device and data device have the @@ -315,21 +315,18 @@ xfs_mount_validate_sb( * a volume filesystem in a non-volume manner. */ if (sbp->sb_magicnum != XFS_SB_MAGIC) { - if (loud) - xfs_warn(mp, "bad magic number"); + xfs_warn(mp, "bad magic number"); return XFS_ERROR(EWRONGFS); } if (!xfs_sb_good_version(sbp)) { - if (loud) - xfs_warn(mp, "bad version"); + xfs_warn(mp, "bad version"); return XFS_ERROR(EWRONGFS); } if (unlikely( sbp->sb_logstart == 0 && mp->m_logdev_targp == mp->m_ddev_targp)) { - if (loud) - xfs_warn(mp, + xfs_warn(mp, "filesystem is marked as having an external log; " "specify logdev on the mount command line."); return XFS_ERROR(EINVAL); @@ -337,8 +334,7 @@ xfs_mount_validate_sb( if (unlikely( sbp->sb_logstart != 0 && mp->m_logdev_targp != mp->m_ddev_targp)) { - if (loud) - xfs_warn(mp, + xfs_warn(mp, "filesystem is marked as having an internal log; " "do not specify logdev on the mount command line."); return XFS_ERROR(EINVAL); @@ -372,8 +368,7 @@ xfs_mount_validate_sb( sbp->sb_dblocks == 0 || sbp->sb_dblocks > XFS_MAX_DBLOCKS(sbp) || sbp->sb_dblocks < XFS_MIN_DBLOCKS(sbp))) { - if (loud) - XFS_CORRUPTION_ERROR("SB sanity check failed", + XFS_CORRUPTION_ERROR("SB sanity check failed", XFS_ERRLEVEL_LOW, mp, sbp); return XFS_ERROR(EFSCORRUPTED); } @@ -382,12 +377,10 @@ xfs_mount_validate_sb( * Until this is fixed only page-sized or smaller data blocks work. */ if (unlikely(sbp->sb_blocksize > PAGE_SIZE)) { - if (loud) { - xfs_warn(mp, + xfs_warn(mp, "File system with blocksize %d bytes. " "Only pagesize (%ld) or less will currently work.", sbp->sb_blocksize, PAGE_SIZE); - } return XFS_ERROR(ENOSYS); } @@ -401,23 +394,20 @@ xfs_mount_validate_sb( case 2048: break; default: - if (loud) - xfs_warn(mp, "inode size of %d bytes not supported", + xfs_warn(mp, "inode size of %d bytes not supported", sbp->sb_inodesize); return XFS_ERROR(ENOSYS); } if (xfs_sb_validate_fsb_count(sbp, sbp->sb_dblocks) || xfs_sb_validate_fsb_count(sbp, sbp->sb_rblocks)) { - if (loud) - xfs_warn(mp, + xfs_warn(mp, "file system too large to be mounted on this system."); return XFS_ERROR(EFBIG); } - if (unlikely(sbp->sb_inprogress)) { - if (loud) - xfs_warn(mp, "file system busy"); + if (check_inprogress && sbp->sb_inprogress) { + xfs_warn(mp, "Offline file system operation in progress!"); return XFS_ERROR(EFSCORRUPTED); } @@ -425,9 +415,7 @@ xfs_mount_validate_sb( * Version 1 directory format has never worked on Linux. */ if (unlikely(!xfs_sb_version_hasdirv2(sbp))) { - if (loud) - xfs_warn(mp, - "file system using version 1 directory format"); + xfs_warn(mp, "file system using version 1 directory format"); return XFS_ERROR(ENOSYS); } @@ -520,11 +508,9 @@ out_unwind: void xfs_sb_from_disk( - struct xfs_mount *mp, + struct xfs_sb *to, xfs_dsb_t *from) { - struct xfs_sb *to = &mp->m_sb; - to->sb_magicnum = be32_to_cpu(from->sb_magicnum); to->sb_blocksize = be32_to_cpu(from->sb_blocksize); to->sb_dblocks = be64_to_cpu(from->sb_dblocks); @@ -626,6 +612,72 @@ xfs_sb_to_disk( } } +static void +xfs_sb_verify( + struct xfs_buf *bp) +{ + struct xfs_mount *mp = bp->b_target->bt_mount; + struct xfs_sb sb; + int error; + + xfs_sb_from_disk(&sb, XFS_BUF_TO_SBP(bp)); + + /* + * Only check the in progress field for the primary superblock as + * mkfs.xfs doesn't clear it from secondary superblocks. + */ + error = xfs_mount_validate_sb(mp, &sb, bp->b_bn == XFS_SB_DADDR); + if (error) + xfs_buf_ioerror(bp, error); +} + +static void +xfs_sb_read_verify( + struct xfs_buf *bp) +{ + xfs_sb_verify(bp); +} + +/* + * We may be probed for a filesystem match, so we may not want to emit + * messages when the superblock buffer is not actually an XFS superblock. + * If we find an XFS superblock, the run a normal, noisy mount because we are + * really going to mount it and want to know about errors. + */ +static void +xfs_sb_quiet_read_verify( + struct xfs_buf *bp) +{ + struct xfs_sb sb; + + xfs_sb_from_disk(&sb, XFS_BUF_TO_SBP(bp)); + + if (sb.sb_magicnum == XFS_SB_MAGIC) { + /* XFS filesystem, verify noisily! */ + xfs_sb_read_verify(bp); + return; + } + /* quietly fail */ + xfs_buf_ioerror(bp, EFSCORRUPTED); +} + +static void +xfs_sb_write_verify( + struct xfs_buf *bp) +{ + xfs_sb_verify(bp); +} + +const struct xfs_buf_ops xfs_sb_buf_ops = { + .verify_read = xfs_sb_read_verify, + .verify_write = xfs_sb_write_verify, +}; + +static const struct xfs_buf_ops xfs_sb_quiet_buf_ops = { + .verify_read = xfs_sb_quiet_read_verify, + .verify_write = xfs_sb_write_verify, +}; + /* * xfs_readsb * @@ -651,26 +703,27 @@ xfs_readsb(xfs_mount_t *mp, int flags) reread: bp = xfs_buf_read_uncached(mp->m_ddev_targp, XFS_SB_DADDR, - BTOBB(sector_size), 0); + BTOBB(sector_size), 0, + loud ? &xfs_sb_buf_ops + : &xfs_sb_quiet_buf_ops); if (!bp) { if (loud) xfs_warn(mp, "SB buffer read failed"); return EIO; } - - /* - * Initialize the mount structure from the superblock. - * But first do some basic consistency checking. - */ - xfs_sb_from_disk(mp, XFS_BUF_TO_SBP(bp)); - error = xfs_mount_validate_sb(mp, &(mp->m_sb), flags); - if (error) { + if (bp->b_error) { + error = bp->b_error; if (loud) xfs_warn(mp, "SB validate failed"); goto release_buf; } /* + * Initialize the mount structure from the superblock. + */ + xfs_sb_from_disk(&mp->m_sb, XFS_BUF_TO_SBP(bp)); + + /* * We must be able to do sector-sized and sector-aligned IO. */ if (sector_size > mp->m_sb.sb_sectsize) { @@ -1001,7 +1054,7 @@ xfs_check_sizes(xfs_mount_t *mp) } bp = xfs_buf_read_uncached(mp->m_ddev_targp, d - XFS_FSS_TO_BB(mp, 1), - XFS_FSS_TO_BB(mp, 1), 0); + XFS_FSS_TO_BB(mp, 1), 0, NULL); if (!bp) { xfs_warn(mp, "last sector read failed"); return EIO; @@ -1016,7 +1069,7 @@ xfs_check_sizes(xfs_mount_t *mp) } bp = xfs_buf_read_uncached(mp->m_logdev_targp, d - XFS_FSB_TO_BB(mp, 1), - XFS_FSB_TO_BB(mp, 1), 0); + XFS_FSB_TO_BB(mp, 1), 0, NULL); if (!bp) { xfs_warn(mp, "log device read failed"); return EIO; @@ -1427,6 +1480,8 @@ xfs_unmountfs( __uint64_t resblks; int error; + cancel_delayed_work_sync(&mp->m_eofblocks_work); + xfs_qm_unmount_quotas(mp); xfs_rtunmount_inodes(mp); IRELE(mp->m_rootip); @@ -1450,21 +1505,16 @@ xfs_unmountfs( /* * And reclaim all inodes. At this point there should be no dirty - * inode, and none should be pinned or locked, but use synchronous - * reclaim just to be sure. + * inodes and none should be pinned or locked, but use synchronous + * reclaim just to be sure. We can stop background inode reclaim + * here as well if it is still running. */ + cancel_delayed_work_sync(&mp->m_reclaim_work); xfs_reclaim_inodes(mp, SYNC_WAIT); xfs_qm_unmount(mp); /* - * Flush out the log synchronously so that we know for sure - * that nothing is pinned. This is important because bflush() - * will skip pinned buffers. - */ - xfs_log_force(mp, XFS_LOG_SYNC); - - /* * Unreserve any blocks we have so that when we unmount we don't account * the reserved free space as used. This is really only necessary for * lazy superblock counting because it trusts the incore superblock @@ -1489,23 +1539,6 @@ xfs_unmountfs( xfs_warn(mp, "Unable to update superblock counters. " "Freespace may not be correct on next mount."); - /* - * At this point we might have modified the superblock again and thus - * added an item to the AIL, thus flush it again. - */ - xfs_ail_push_all_sync(mp->m_ail); - xfs_wait_buftarg(mp->m_ddev_targp); - - /* - * The superblock buffer is uncached and xfsaild_push() will lock and - * set the XBF_ASYNC flag on the buffer. We cannot do xfs_buf_iowait() - * here but a lock on the superblock buffer will block until iodone() - * has completed. - */ - xfs_buf_lock(mp->m_sb_bp); - xfs_buf_unlock(mp->m_sb_bp); - - xfs_log_unmount_write(mp); xfs_log_unmount(mp); xfs_uuid_unmount(mp); diff --git a/fs/xfs/xfs_mount.h b/fs/xfs/xfs_mount.h index deee09e534dc..bab8314507e4 100644 --- a/fs/xfs/xfs_mount.h +++ b/fs/xfs/xfs_mount.h @@ -51,8 +51,6 @@ typedef struct xfs_trans_reservations { #else /* __KERNEL__ */ -#include "xfs_sync.h" - struct xlog; struct xfs_inode; struct xfs_mru_cache; @@ -197,9 +195,9 @@ typedef struct xfs_mount { struct mutex m_icsb_mutex; /* balancer sync lock */ #endif struct xfs_mru_cache *m_filestream; /* per-mount filestream data */ - struct 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 */ + struct delayed_work m_eofblocks_work; /* background eof blocks + trimming */ __int64_t m_update_flags; /* sb flags we need to update on the next remount,rw */ struct shrinker m_inode_shrink; /* inode reclaim shrinker */ @@ -209,6 +207,9 @@ typedef struct xfs_mount { struct workqueue_struct *m_data_workqueue; struct workqueue_struct *m_unwritten_workqueue; struct workqueue_struct *m_cil_workqueue; + struct workqueue_struct *m_reclaim_workqueue; + struct workqueue_struct *m_log_workqueue; + struct workqueue_struct *m_eofblocks_workqueue; } xfs_mount_t; /* @@ -387,7 +388,9 @@ extern void xfs_set_low_space_thresholds(struct xfs_mount *); extern void xfs_mod_sb(struct xfs_trans *, __int64_t); extern int xfs_initialize_perag(struct xfs_mount *, xfs_agnumber_t, xfs_agnumber_t *); -extern void xfs_sb_from_disk(struct xfs_mount *, struct xfs_dsb *); +extern void xfs_sb_from_disk(struct xfs_sb *, struct xfs_dsb *); extern void xfs_sb_to_disk(struct xfs_dsb *, struct xfs_sb *, __int64_t); +extern const struct xfs_buf_ops xfs_sb_buf_ops; + #endif /* __XFS_MOUNT_H__ */ diff --git a/fs/xfs/xfs_qm.c b/fs/xfs/xfs_qm.c index 2e86fa0cfc0d..60eff4763156 100644 --- a/fs/xfs/xfs_qm.c +++ b/fs/xfs/xfs_qm.c @@ -40,6 +40,7 @@ #include "xfs_utils.h" #include "xfs_qm.h" #include "xfs_trace.h" +#include "xfs_icache.h" /* * The global quota manager. There is only one of these for the entire @@ -891,7 +892,8 @@ xfs_qm_dqiter_bufs( while (blkcnt--) { error = xfs_trans_read_buf(mp, NULL, mp->m_ddev_targp, XFS_FSB_TO_DADDR(mp, bno), - mp->m_quotainfo->qi_dqchunklen, 0, &bp); + mp->m_quotainfo->qi_dqchunklen, 0, &bp, + &xfs_dquot_buf_ops); if (error) break; @@ -978,7 +980,8 @@ xfs_qm_dqiterate( while (rablkcnt--) { xfs_buf_readahead(mp->m_ddev_targp, XFS_FSB_TO_DADDR(mp, rablkno), - mp->m_quotainfo->qi_dqchunklen); + mp->m_quotainfo->qi_dqchunklen, + NULL); rablkno++; } } @@ -1453,7 +1456,7 @@ xfs_qm_dqreclaim_one( int error; if (!xfs_dqlock_nowait(dqp)) - goto out_busy; + goto out_move_tail; /* * This dquot has acquired a reference in the meantime remove it from @@ -1476,7 +1479,7 @@ xfs_qm_dqreclaim_one( * getting flushed to disk, we don't want to reclaim it. */ if (!xfs_dqflock_nowait(dqp)) - goto out_busy; + goto out_unlock_move_tail; if (XFS_DQ_IS_DIRTY(dqp)) { struct xfs_buf *bp = NULL; @@ -1487,7 +1490,7 @@ xfs_qm_dqreclaim_one( if (error) { xfs_warn(mp, "%s: dquot %p flush failed", __func__, dqp); - goto out_busy; + goto out_unlock_move_tail; } xfs_buf_delwri_queue(bp, buffer_list); @@ -1496,7 +1499,7 @@ xfs_qm_dqreclaim_one( * Give the dquot another try on the freelist, as the * flushing will take some time. */ - goto out_busy; + goto out_unlock_move_tail; } xfs_dqfunlock(dqp); @@ -1515,14 +1518,13 @@ xfs_qm_dqreclaim_one( XFS_STATS_INC(xs_qm_dqreclaims); return; -out_busy: - xfs_dqunlock(dqp); - /* * Move the dquot to the tail of the list so that we don't spin on it. */ +out_unlock_move_tail: + xfs_dqunlock(dqp); +out_move_tail: list_move_tail(&dqp->q_lru, &qi->qi_lru_list); - trace_xfs_dqreclaim_busy(dqp); XFS_STATS_INC(xs_qm_dqreclaim_misses); } diff --git a/fs/xfs/xfs_qm_syscalls.c b/fs/xfs/xfs_qm_syscalls.c index 858a3b186110..5f53e75409b8 100644 --- a/fs/xfs/xfs_qm_syscalls.c +++ b/fs/xfs/xfs_qm_syscalls.c @@ -40,6 +40,7 @@ #include "xfs_utils.h" #include "xfs_qm.h" #include "xfs_trace.h" +#include "xfs_icache.h" STATIC int xfs_qm_log_quotaoff(xfs_mount_t *, xfs_qoff_logitem_t **, uint); STATIC int xfs_qm_log_quotaoff_end(xfs_mount_t *, xfs_qoff_logitem_t *, @@ -845,7 +846,8 @@ STATIC int xfs_dqrele_inode( struct xfs_inode *ip, struct xfs_perag *pag, - int flags) + int flags, + void *args) { /* skip quota inodes */ if (ip == ip->i_mount->m_quotainfo->qi_uquotaip || @@ -881,5 +883,5 @@ xfs_qm_dqrele_all_inodes( uint flags) { ASSERT(mp->m_quotainfo); - xfs_inode_ag_iterator(mp, xfs_dqrele_inode, flags); + xfs_inode_ag_iterator(mp, xfs_dqrele_inode, flags, NULL); } diff --git a/fs/xfs/xfs_rtalloc.c b/fs/xfs/xfs_rtalloc.c index ca28a4ba4b54..98dc670d3ee0 100644 --- a/fs/xfs/xfs_rtalloc.c +++ b/fs/xfs/xfs_rtalloc.c @@ -38,6 +38,7 @@ #include "xfs_utils.h" #include "xfs_trace.h" #include "xfs_buf.h" +#include "xfs_icache.h" /* @@ -869,7 +870,7 @@ xfs_rtbuf_get( ASSERT(map.br_startblock != NULLFSBLOCK); error = xfs_trans_read_buf(mp, tp, mp->m_ddev_targp, XFS_FSB_TO_DADDR(mp, map.br_startblock), - mp->m_bsize, 0, &bp); + mp->m_bsize, 0, &bp, NULL); if (error) return error; ASSERT(!xfs_buf_geterror(bp)); @@ -1872,9 +1873,14 @@ xfs_growfs_rt( */ bp = xfs_buf_read_uncached(mp->m_rtdev_targp, XFS_FSB_TO_BB(mp, nrblocks - 1), - XFS_FSB_TO_BB(mp, 1), 0); + XFS_FSB_TO_BB(mp, 1), 0, NULL); if (!bp) return EIO; + if (bp->b_error) { + error = bp->b_error; + xfs_buf_relse(bp); + return error; + } xfs_buf_relse(bp); /* @@ -2219,9 +2225,11 @@ xfs_rtmount_init( } bp = xfs_buf_read_uncached(mp->m_rtdev_targp, d - XFS_FSB_TO_BB(mp, 1), - XFS_FSB_TO_BB(mp, 1), 0); - if (!bp) { + XFS_FSB_TO_BB(mp, 1), 0, NULL); + if (!bp || bp->b_error) { xfs_warn(mp, "realtime device size check failed"); + if (bp) + xfs_buf_relse(bp); return EIO; } xfs_buf_relse(bp); diff --git a/fs/xfs/xfs_sb.h b/fs/xfs/xfs_sb.h index f429d9d5d325..a05b45175fb0 100644 --- a/fs/xfs/xfs_sb.h +++ b/fs/xfs/xfs_sb.h @@ -81,6 +81,7 @@ struct xfs_mount; #define XFS_SB_VERSION2_ATTR2BIT 0x00000008 /* Inline attr rework */ #define XFS_SB_VERSION2_PARENTBIT 0x00000010 /* parent pointers */ #define XFS_SB_VERSION2_PROJID32BIT 0x00000080 /* 32 bit project id */ +#define XFS_SB_VERSION2_CRCBIT 0x00000100 /* metadata CRCs */ #define XFS_SB_VERSION2_OKREALFBITS \ (XFS_SB_VERSION2_LAZYSBCOUNTBIT | \ @@ -503,6 +504,12 @@ static inline int xfs_sb_version_hasprojid32bit(xfs_sb_t *sbp) (sbp->sb_features2 & XFS_SB_VERSION2_PROJID32BIT); } +static inline int xfs_sb_version_hascrc(xfs_sb_t *sbp) +{ + return (xfs_sb_version_hasmorebits(sbp) && + (sbp->sb_features2 & XFS_SB_VERSION2_CRCBIT)); +} + /* * end of superblock version macros */ diff --git a/fs/xfs/xfs_super.c b/fs/xfs/xfs_super.c index 26a09bd7f975..ab8839b26272 100644 --- a/fs/xfs/xfs_super.c +++ b/fs/xfs/xfs_super.c @@ -49,7 +49,7 @@ #include "xfs_extfree_item.h" #include "xfs_mru_cache.h" #include "xfs_inode_item.h" -#include "xfs_sync.h" +#include "xfs_icache.h" #include "xfs_trace.h" #include <linux/namei.h> @@ -863,8 +863,30 @@ xfs_init_mount_workqueues( WQ_MEM_RECLAIM, 0, mp->m_fsname); if (!mp->m_cil_workqueue) goto out_destroy_unwritten; + + mp->m_reclaim_workqueue = alloc_workqueue("xfs-reclaim/%s", + WQ_NON_REENTRANT, 0, mp->m_fsname); + if (!mp->m_reclaim_workqueue) + goto out_destroy_cil; + + mp->m_log_workqueue = alloc_workqueue("xfs-log/%s", + WQ_NON_REENTRANT, 0, mp->m_fsname); + if (!mp->m_log_workqueue) + goto out_destroy_reclaim; + + mp->m_eofblocks_workqueue = alloc_workqueue("xfs-eofblocks/%s", + WQ_NON_REENTRANT, 0, mp->m_fsname); + if (!mp->m_eofblocks_workqueue) + goto out_destroy_log; + return 0; +out_destroy_log: + destroy_workqueue(mp->m_log_workqueue); +out_destroy_reclaim: + destroy_workqueue(mp->m_reclaim_workqueue); +out_destroy_cil: + destroy_workqueue(mp->m_cil_workqueue); out_destroy_unwritten: destroy_workqueue(mp->m_unwritten_workqueue); out_destroy_data_iodone_queue: @@ -877,11 +899,32 @@ STATIC void xfs_destroy_mount_workqueues( struct xfs_mount *mp) { + destroy_workqueue(mp->m_eofblocks_workqueue); + destroy_workqueue(mp->m_log_workqueue); + destroy_workqueue(mp->m_reclaim_workqueue); destroy_workqueue(mp->m_cil_workqueue); destroy_workqueue(mp->m_data_workqueue); destroy_workqueue(mp->m_unwritten_workqueue); } +/* + * Flush all dirty data to disk. Must not be called while holding an XFS_ILOCK + * or a page lock. We use sync_inodes_sb() here to ensure we block while waiting + * for IO to complete so that we effectively throttle multiple callers to the + * rate at which IO is completing. + */ +void +xfs_flush_inodes( + struct xfs_mount *mp) +{ + struct super_block *sb = mp->m_super; + + if (down_read_trylock(&sb->s_umount)) { + sync_inodes_sb(sb); + up_read(&sb->s_umount); + } +} + /* Catch misguided souls that try to use this interface on XFS */ STATIC struct inode * xfs_fs_alloc_inode( @@ -1006,9 +1049,8 @@ xfs_fs_put_super( struct xfs_mount *mp = XFS_M(sb); xfs_filestream_unmount(mp); - cancel_delayed_work_sync(&mp->m_sync_work); xfs_unmountfs(mp); - xfs_syncd_stop(mp); + xfs_freesb(mp); xfs_icsb_destroy_counters(mp); xfs_destroy_mount_workqueues(mp); @@ -1023,7 +1065,6 @@ xfs_fs_sync_fs( int wait) { struct xfs_mount *mp = XFS_M(sb); - int error; /* * Doing anything during the async pass would be counterproductive. @@ -1031,17 +1072,14 @@ xfs_fs_sync_fs( if (!wait) return 0; - error = xfs_quiesce_data(mp); - if (error) - return -error; - + xfs_log_force(mp, XFS_LOG_SYNC); if (laptop_mode) { /* * The disk must be active because we're syncing. - * We schedule xfssyncd now (now that the disk is + * We schedule log work now (now that the disk is * active) instead of later (when it might not be). */ - flush_delayed_work(&mp->m_sync_work); + flush_delayed_work(&mp->m_log->l_work); } return 0; @@ -1118,6 +1156,48 @@ xfs_restore_resvblks(struct xfs_mount *mp) xfs_reserve_blocks(mp, &resblks, NULL); } +/* + * Trigger writeback of all the dirty metadata in the file system. + * + * This ensures that the metadata is written to their location on disk rather + * than just existing in transactions in the log. This means after a quiesce + * there is no log replay required to write the inodes to disk - this is the + * primary difference between a sync and a quiesce. + * + * Note: xfs_log_quiesce() stops background log work - the callers must ensure + * it is started again when appropriate. + */ +void +xfs_quiesce_attr( + struct xfs_mount *mp) +{ + int error = 0; + + /* wait for all modifications to complete */ + while (atomic_read(&mp->m_active_trans) > 0) + delay(100); + + /* force the log to unpin objects from the now complete transactions */ + xfs_log_force(mp, XFS_LOG_SYNC); + + /* reclaim inodes to do any IO before the freeze completes */ + xfs_reclaim_inodes(mp, 0); + xfs_reclaim_inodes(mp, SYNC_WAIT); + + /* Push the superblock and write an unmount record */ + error = xfs_log_sbcount(mp); + if (error) + xfs_warn(mp, "xfs_attr_quiesce: failed to log sb changes. " + "Frozen image may not be consistent."); + /* + * Just warn here till VFS can correctly support + * read-only remount without racing. + */ + WARN_ON(atomic_read(&mp->m_active_trans) != 0); + + xfs_log_quiesce(mp); +} + STATIC int xfs_fs_remount( struct super_block *sb, @@ -1198,20 +1278,18 @@ xfs_fs_remount( * value if it is non-zero, otherwise go with the default. */ xfs_restore_resvblks(mp); + xfs_log_work_queue(mp); } /* rw -> ro */ if (!(mp->m_flags & XFS_MOUNT_RDONLY) && (*flags & MS_RDONLY)) { /* - * After we have synced the data but before we sync the - * metadata, we need to free up the reserve block pool so that - * the used block count in the superblock on disk is correct at - * the end of the remount. Stash the current reserve pool size - * so that if we get remounted rw, we can return it to the same - * size. + * Before we sync the metadata, we need to free up the reserve + * block pool so that the used block count in the superblock on + * disk is correct at the end of the remount. Stash the current + * reserve pool size so that if we get remounted rw, we can + * return it to the same size. */ - - xfs_quiesce_data(mp); xfs_save_resvblks(mp); xfs_quiesce_attr(mp); mp->m_flags |= XFS_MOUNT_RDONLY; @@ -1243,6 +1321,7 @@ xfs_fs_unfreeze( struct xfs_mount *mp = XFS_M(sb); xfs_restore_resvblks(mp); + xfs_log_work_queue(mp); return 0; } @@ -1321,6 +1400,8 @@ xfs_fs_fill_super( spin_lock_init(&mp->m_sb_lock); mutex_init(&mp->m_growlock); atomic_set(&mp->m_active_trans, 0); + INIT_DELAYED_WORK(&mp->m_reclaim_work, xfs_reclaim_worker); + INIT_DELAYED_WORK(&mp->m_eofblocks_work, xfs_eofblocks_worker); mp->m_super = sb; sb->s_fs_info = mp; @@ -1371,10 +1452,6 @@ xfs_fs_fill_super( /* * we must configure the block size in the superblock before we run the * full mount process as the mount process can lookup and cache inodes. - * For the same reason we must also initialise the syncd and register - * the inode cache shrinker so that inodes can be reclaimed during - * operations like a quotacheck that iterate all inodes in the - * filesystem. */ sb->s_magic = XFS_SB_MAGIC; sb->s_blocksize = mp->m_sb.sb_blocksize; @@ -1384,13 +1461,9 @@ xfs_fs_fill_super( sb->s_time_gran = 1; set_posix_acl_flag(sb); - error = xfs_syncd_init(mp); - if (error) - goto out_filestream_unmount; - error = xfs_mountfs(mp); if (error) - goto out_syncd_stop; + goto out_filestream_unmount; root = igrab(VFS_I(mp->m_rootip)); if (!root) { @@ -1408,8 +1481,7 @@ xfs_fs_fill_super( } return 0; - out_syncd_stop: - xfs_syncd_stop(mp); + out_filestream_unmount: xfs_filestream_unmount(mp); out_free_sb: @@ -1429,7 +1501,6 @@ out_destroy_workqueues: out_unmount: xfs_filestream_unmount(mp); xfs_unmountfs(mp); - xfs_syncd_stop(mp); goto out_free_sb; } @@ -1625,16 +1696,6 @@ STATIC int __init xfs_init_workqueues(void) { /* - * We never want to the same work item to run twice, reclaiming inodes - * or idling the log is not going to get any faster by multiple CPUs - * competing for ressources. Use the default large max_active value - * so that even lots of filesystems can perform these task in parallel. - */ - xfs_syncd_wq = alloc_workqueue("xfssyncd", WQ_NON_REENTRANT, 0); - if (!xfs_syncd_wq) - return -ENOMEM; - - /* * The allocation workqueue can be used in memory reclaim situations * (writepage path), and parallelism is only limited by the number of * AGs in all the filesystems mounted. Hence use the default large @@ -1642,20 +1703,15 @@ xfs_init_workqueues(void) */ xfs_alloc_wq = alloc_workqueue("xfsalloc", WQ_MEM_RECLAIM, 0); if (!xfs_alloc_wq) - goto out_destroy_syncd; + return -ENOMEM; return 0; - -out_destroy_syncd: - destroy_workqueue(xfs_syncd_wq); - return -ENOMEM; } STATIC void xfs_destroy_workqueues(void) { destroy_workqueue(xfs_alloc_wq); - destroy_workqueue(xfs_syncd_wq); } STATIC int __init diff --git a/fs/xfs/xfs_super.h b/fs/xfs/xfs_super.h index 9de4a920ba05..bbe3d15a7904 100644 --- a/fs/xfs/xfs_super.h +++ b/fs/xfs/xfs_super.h @@ -74,6 +74,7 @@ struct block_device; extern __uint64_t xfs_max_file_offset(unsigned int); +extern void xfs_flush_inodes(struct xfs_mount *mp); extern void xfs_blkdev_issue_flush(struct xfs_buftarg *); extern xfs_agnumber_t xfs_set_inode32(struct xfs_mount *); extern xfs_agnumber_t xfs_set_inode64(struct xfs_mount *); diff --git a/fs/xfs/xfs_sysctl.c b/fs/xfs/xfs_sysctl.c index ee2d2adaa438..2801b5ce6cdb 100644 --- a/fs/xfs/xfs_sysctl.c +++ b/fs/xfs/xfs_sysctl.c @@ -202,6 +202,15 @@ static ctl_table xfs_table[] = { .extra1 = &xfs_params.fstrm_timer.min, .extra2 = &xfs_params.fstrm_timer.max, }, + { + .procname = "speculative_prealloc_lifetime", + .data = &xfs_params.eofb_timer.val, + .maxlen = sizeof(int), + .mode = 0644, + .proc_handler = proc_dointvec_minmax, + .extra1 = &xfs_params.eofb_timer.min, + .extra2 = &xfs_params.eofb_timer.max, + }, /* please keep this the last entry */ #ifdef CONFIG_PROC_FS { diff --git a/fs/xfs/xfs_sysctl.h b/fs/xfs/xfs_sysctl.h index b9937d450f8e..bd8e157c20ef 100644 --- a/fs/xfs/xfs_sysctl.h +++ b/fs/xfs/xfs_sysctl.h @@ -47,6 +47,7 @@ typedef struct xfs_param { xfs_sysctl_val_t rotorstep; /* inode32 AG rotoring control knob */ xfs_sysctl_val_t inherit_nodfrg;/* Inherit the "nodefrag" inode flag. */ xfs_sysctl_val_t fstrm_timer; /* Filestream dir-AG assoc'n timeout. */ + xfs_sysctl_val_t eofb_timer; /* Interval between eofb scan wakeups */ } xfs_param_t; /* diff --git a/fs/xfs/xfs_trace.h b/fs/xfs/xfs_trace.h index 7d36ccf57f93..2e137d4a85ae 100644 --- a/fs/xfs/xfs_trace.h +++ b/fs/xfs/xfs_trace.h @@ -96,6 +96,8 @@ DEFINE_ATTR_LIST_EVENT(xfs_attr_list_full); DEFINE_ATTR_LIST_EVENT(xfs_attr_list_add); DEFINE_ATTR_LIST_EVENT(xfs_attr_list_wrong_blk); DEFINE_ATTR_LIST_EVENT(xfs_attr_list_notfound); +DEFINE_ATTR_LIST_EVENT(xfs_attr_leaf_list); +DEFINE_ATTR_LIST_EVENT(xfs_attr_node_list); DECLARE_EVENT_CLASS(xfs_perag_class, TP_PROTO(struct xfs_mount *mp, xfs_agnumber_t agno, int refcount, @@ -130,6 +132,8 @@ DEFINE_PERAG_REF_EVENT(xfs_perag_get_tag); DEFINE_PERAG_REF_EVENT(xfs_perag_put); DEFINE_PERAG_REF_EVENT(xfs_perag_set_reclaim); DEFINE_PERAG_REF_EVENT(xfs_perag_clear_reclaim); +DEFINE_PERAG_REF_EVENT(xfs_perag_set_eofblocks); +DEFINE_PERAG_REF_EVENT(xfs_perag_clear_eofblocks); TRACE_EVENT(xfs_attr_list_node_descend, TP_PROTO(struct xfs_attr_list_context *ctx, @@ -585,6 +589,10 @@ DEFINE_INODE_EVENT(xfs_update_time); DEFINE_INODE_EVENT(xfs_dquot_dqalloc); DEFINE_INODE_EVENT(xfs_dquot_dqdetach); +DEFINE_INODE_EVENT(xfs_inode_set_eofblocks_tag); +DEFINE_INODE_EVENT(xfs_inode_clear_eofblocks_tag); +DEFINE_INODE_EVENT(xfs_inode_free_eofblocks_invalid); + DECLARE_EVENT_CLASS(xfs_iref_class, TP_PROTO(struct xfs_inode *ip, unsigned long caller_ip), TP_ARGS(ip, caller_ip), @@ -1496,8 +1504,42 @@ DEFINE_DIR2_EVENT(xfs_dir2_node_replace); DEFINE_DIR2_EVENT(xfs_dir2_node_removename); DEFINE_DIR2_EVENT(xfs_dir2_node_to_leaf); +DECLARE_EVENT_CLASS(xfs_attr_class, + TP_PROTO(struct xfs_da_args *args), + TP_ARGS(args), + TP_STRUCT__entry( + __field(dev_t, dev) + __field(xfs_ino_t, ino) + __dynamic_array(char, name, args->namelen) + __field(int, namelen) + __field(int, valuelen) + __field(xfs_dahash_t, hashval) + __field(int, op_flags) + ), + TP_fast_assign( + __entry->dev = VFS_I(args->dp)->i_sb->s_dev; + __entry->ino = args->dp->i_ino; + if (args->namelen) + memcpy(__get_str(name), args->name, args->namelen); + __entry->namelen = args->namelen; + __entry->valuelen = args->valuelen; + __entry->hashval = args->hashval; + __entry->op_flags = args->op_flags; + ), + TP_printk("dev %d:%d ino 0x%llx name %.*s namelen %d valuelen %d " + "hashval 0x%x op_flags %s", + MAJOR(__entry->dev), MINOR(__entry->dev), + __entry->ino, + __entry->namelen, + __entry->namelen ? __get_str(name) : NULL, + __entry->namelen, + __entry->valuelen, + __entry->hashval, + __print_flags(__entry->op_flags, "|", XFS_DA_OP_FLAGS)) +) + #define DEFINE_ATTR_EVENT(name) \ -DEFINE_EVENT(xfs_da_class, name, \ +DEFINE_EVENT(xfs_attr_class, name, \ TP_PROTO(struct xfs_da_args *args), \ TP_ARGS(args)) DEFINE_ATTR_EVENT(xfs_attr_sf_add); @@ -1511,10 +1553,14 @@ DEFINE_ATTR_EVENT(xfs_attr_sf_to_leaf); DEFINE_ATTR_EVENT(xfs_attr_leaf_add); DEFINE_ATTR_EVENT(xfs_attr_leaf_add_old); DEFINE_ATTR_EVENT(xfs_attr_leaf_add_new); +DEFINE_ATTR_EVENT(xfs_attr_leaf_add_work); DEFINE_ATTR_EVENT(xfs_attr_leaf_addname); DEFINE_ATTR_EVENT(xfs_attr_leaf_create); +DEFINE_ATTR_EVENT(xfs_attr_leaf_compact); +DEFINE_ATTR_EVENT(xfs_attr_leaf_get); DEFINE_ATTR_EVENT(xfs_attr_leaf_lookup); DEFINE_ATTR_EVENT(xfs_attr_leaf_replace); +DEFINE_ATTR_EVENT(xfs_attr_leaf_remove); DEFINE_ATTR_EVENT(xfs_attr_leaf_removename); DEFINE_ATTR_EVENT(xfs_attr_leaf_split); DEFINE_ATTR_EVENT(xfs_attr_leaf_split_before); @@ -1526,12 +1572,21 @@ DEFINE_ATTR_EVENT(xfs_attr_leaf_to_sf); DEFINE_ATTR_EVENT(xfs_attr_leaf_to_node); DEFINE_ATTR_EVENT(xfs_attr_leaf_rebalance); DEFINE_ATTR_EVENT(xfs_attr_leaf_unbalance); +DEFINE_ATTR_EVENT(xfs_attr_leaf_toosmall); DEFINE_ATTR_EVENT(xfs_attr_node_addname); +DEFINE_ATTR_EVENT(xfs_attr_node_get); DEFINE_ATTR_EVENT(xfs_attr_node_lookup); DEFINE_ATTR_EVENT(xfs_attr_node_replace); DEFINE_ATTR_EVENT(xfs_attr_node_removename); +DEFINE_ATTR_EVENT(xfs_attr_fillstate); +DEFINE_ATTR_EVENT(xfs_attr_refillstate); + +DEFINE_ATTR_EVENT(xfs_attr_rmtval_get); +DEFINE_ATTR_EVENT(xfs_attr_rmtval_set); +DEFINE_ATTR_EVENT(xfs_attr_rmtval_remove); + #define DEFINE_DA_EVENT(name) \ DEFINE_EVENT(xfs_da_class, name, \ TP_PROTO(struct xfs_da_args *args), \ @@ -1550,9 +1605,12 @@ DEFINE_DA_EVENT(xfs_da_node_split); DEFINE_DA_EVENT(xfs_da_node_remove); DEFINE_DA_EVENT(xfs_da_node_rebalance); DEFINE_DA_EVENT(xfs_da_node_unbalance); +DEFINE_DA_EVENT(xfs_da_node_toosmall); DEFINE_DA_EVENT(xfs_da_swap_lastblock); DEFINE_DA_EVENT(xfs_da_grow_inode); DEFINE_DA_EVENT(xfs_da_shrink_inode); +DEFINE_DA_EVENT(xfs_da_fixhashpath); +DEFINE_DA_EVENT(xfs_da_path_shift); DECLARE_EVENT_CLASS(xfs_dir2_space_class, TP_PROTO(struct xfs_da_args *args, int idx), diff --git a/fs/xfs/xfs_trans.h b/fs/xfs/xfs_trans.h index db056544cbb5..c6c0601abd7a 100644 --- a/fs/xfs/xfs_trans.h +++ b/fs/xfs/xfs_trans.h @@ -464,10 +464,7 @@ xfs_trans_get_buf( int numblks, uint flags) { - struct xfs_buf_map map = { - .bm_bn = blkno, - .bm_len = numblks, - }; + DEFINE_SINGLE_BUF_MAP(map, blkno, numblks); return xfs_trans_get_buf_map(tp, target, &map, 1, flags); } @@ -476,7 +473,8 @@ int xfs_trans_read_buf_map(struct xfs_mount *mp, struct xfs_buftarg *target, struct xfs_buf_map *map, int nmaps, xfs_buf_flags_t flags, - struct xfs_buf **bpp); + struct xfs_buf **bpp, + const struct xfs_buf_ops *ops); static inline int xfs_trans_read_buf( @@ -486,13 +484,12 @@ xfs_trans_read_buf( xfs_daddr_t blkno, int numblks, xfs_buf_flags_t flags, - struct xfs_buf **bpp) + struct xfs_buf **bpp, + const struct xfs_buf_ops *ops) { - struct xfs_buf_map map = { - .bm_bn = blkno, - .bm_len = numblks, - }; - return xfs_trans_read_buf_map(mp, tp, target, &map, 1, flags, bpp); + DEFINE_SINGLE_BUF_MAP(map, blkno, numblks); + return xfs_trans_read_buf_map(mp, tp, target, &map, 1, + flags, bpp, ops); } struct xfs_buf *xfs_trans_getsb(xfs_trans_t *, struct xfs_mount *, int); diff --git a/fs/xfs/xfs_trans_buf.c b/fs/xfs/xfs_trans_buf.c index 6311b99c267f..4fc17d479d42 100644 --- a/fs/xfs/xfs_trans_buf.c +++ b/fs/xfs/xfs_trans_buf.c @@ -257,7 +257,8 @@ xfs_trans_read_buf_map( struct xfs_buf_map *map, int nmaps, xfs_buf_flags_t flags, - struct xfs_buf **bpp) + struct xfs_buf **bpp, + const struct xfs_buf_ops *ops) { xfs_buf_t *bp; xfs_buf_log_item_t *bip; @@ -265,7 +266,7 @@ xfs_trans_read_buf_map( *bpp = NULL; if (!tp) { - bp = xfs_buf_read_map(target, map, nmaps, flags); + bp = xfs_buf_read_map(target, map, nmaps, flags, ops); if (!bp) return (flags & XBF_TRYLOCK) ? EAGAIN : XFS_ERROR(ENOMEM); @@ -312,7 +313,9 @@ xfs_trans_read_buf_map( if (!(XFS_BUF_ISDONE(bp))) { trace_xfs_trans_read_buf_io(bp, _RET_IP_); ASSERT(!XFS_BUF_ISASYNC(bp)); + ASSERT(bp->b_iodone == NULL); XFS_BUF_READ(bp); + bp->b_ops = ops; xfsbdstrat(tp->t_mountp, bp); error = xfs_buf_iowait(bp); if (error) { @@ -349,7 +352,7 @@ xfs_trans_read_buf_map( return 0; } - bp = xfs_buf_read_map(target, map, nmaps, flags); + bp = xfs_buf_read_map(target, map, nmaps, flags, ops); if (bp == NULL) { *bpp = NULL; return (flags & XBF_TRYLOCK) ? diff --git a/fs/xfs/xfs_vnodeops.c b/fs/xfs/xfs_vnodeops.c index 2a5c637344b4..d95f565a390e 100644 --- a/fs/xfs/xfs_vnodeops.c +++ b/fs/xfs/xfs_vnodeops.c @@ -47,6 +47,7 @@ #include "xfs_filestream.h" #include "xfs_vnodeops.h" #include "xfs_trace.h" +#include "xfs_icache.h" /* * The maximum pathlen is 1024 bytes. Since the minimum file system @@ -79,7 +80,7 @@ xfs_readlink_bmap( d = XFS_FSB_TO_DADDR(mp, mval[n].br_startblock); byte_cnt = XFS_FSB_TO_B(mp, mval[n].br_blockcount); - bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), 0); + bp = xfs_buf_read(mp->m_ddev_targp, d, BTOBB(byte_cnt), 0, NULL); if (!bp) return XFS_ERROR(ENOMEM); error = bp->b_error; @@ -150,7 +151,7 @@ xfs_readlink( * when the link count isn't zero and by xfs_dm_punch_hole() when * punching a hole to EOF. */ -STATIC int +int xfs_free_eofblocks( xfs_mount_t *mp, xfs_inode_t *ip, @@ -199,7 +200,7 @@ xfs_free_eofblocks( if (need_iolock) { if (!xfs_ilock_nowait(ip, XFS_IOLOCK_EXCL)) { xfs_trans_cancel(tp, 0); - return 0; + return EAGAIN; } } @@ -237,6 +238,8 @@ xfs_free_eofblocks( } else { error = xfs_trans_commit(tp, XFS_TRANS_RELEASE_LOG_RES); + if (!error) + xfs_inode_clear_eofblocks_tag(ip); } xfs_iunlock(ip, XFS_ILOCK_EXCL); @@ -425,19 +428,18 @@ xfs_release( truncated = xfs_iflags_test_and_clear(ip, XFS_ITRUNCATED); if (truncated) { xfs_iflags_clear(ip, XFS_IDIRTY_RELEASE); - if (VN_DIRTY(VFS_I(ip)) && ip->i_delayed_blks > 0) - xfs_flush_pages(ip, 0, -1, XBF_ASYNC, FI_NONE); + if (VN_DIRTY(VFS_I(ip)) && ip->i_delayed_blks > 0) { + error = -filemap_flush(VFS_I(ip)->i_mapping); + if (error) + return error; + } } } if (ip->i_d.di_nlink == 0) return 0; - if ((S_ISREG(ip->i_d.di_mode) && - (VFS_I(ip)->i_size > 0 || - (VN_CACHED(VFS_I(ip)) > 0 || ip->i_delayed_blks > 0)) && - (ip->i_df.if_flags & XFS_IFEXTENTS)) && - (!(ip->i_d.di_flags & (XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)))) { + if (xfs_can_free_eofblocks(ip, false)) { /* * If we can't get the iolock just skip truncating the blocks @@ -464,7 +466,7 @@ xfs_release( return 0; error = xfs_free_eofblocks(mp, ip, true); - if (error) + if (error && error != EAGAIN) return error; /* delalloc blocks after truncation means it really is dirty */ @@ -513,13 +515,12 @@ xfs_inactive( goto out; if (ip->i_d.di_nlink != 0) { - if ((S_ISREG(ip->i_d.di_mode) && - (VFS_I(ip)->i_size > 0 || - (VN_CACHED(VFS_I(ip)) > 0 || ip->i_delayed_blks > 0)) && - (ip->i_df.if_flags & XFS_IFEXTENTS) && - (!(ip->i_d.di_flags & - (XFS_DIFLAG_PREALLOC | XFS_DIFLAG_APPEND)) || - ip->i_delayed_blks != 0))) { + /* + * force is true because we are evicting an inode from the + * cache. Post-eof blocks must be freed, lest we end up with + * broken free space accounting. + */ + if (xfs_can_free_eofblocks(ip, true)) { error = xfs_free_eofblocks(mp, ip, false); if (error) return VN_INACTIVE_CACHE; @@ -777,7 +778,7 @@ xfs_create( XFS_TRANS_PERM_LOG_RES, log_count); if (error == ENOSPC) { /* flush outstanding delalloc blocks and retry */ - xfs_flush_inodes(dp); + xfs_flush_inodes(mp); error = xfs_trans_reserve(tp, resblks, log_res, 0, XFS_TRANS_PERM_LOG_RES, log_count); } @@ -1957,12 +1958,11 @@ xfs_free_file_space( rounding = max_t(uint, 1 << mp->m_sb.sb_blocklog, PAGE_CACHE_SIZE); ioffset = offset & ~(rounding - 1); - - if (VN_CACHED(VFS_I(ip)) != 0) { - error = xfs_flushinval_pages(ip, ioffset, -1, FI_REMAPF_LOCKED); - if (error) - goto out_unlock_iolock; - } + error = -filemap_write_and_wait_range(VFS_I(ip)->i_mapping, + ioffset, -1); + if (error) + goto out_unlock_iolock; + truncate_pagecache_range(VFS_I(ip), ioffset, -1); /* * Need to zero the stuff we're not freeing, on disk. @@ -2095,6 +2095,73 @@ xfs_free_file_space( return error; } + +STATIC int +xfs_zero_file_space( + struct xfs_inode *ip, + xfs_off_t offset, + xfs_off_t len, + int attr_flags) +{ + struct xfs_mount *mp = ip->i_mount; + uint granularity; + xfs_off_t start_boundary; + xfs_off_t end_boundary; + int error; + + granularity = max_t(uint, 1 << mp->m_sb.sb_blocklog, PAGE_CACHE_SIZE); + + /* + * Round the range of extents we are going to convert inwards. If the + * offset is aligned, then it doesn't get changed so we zero from the + * start of the block offset points to. + */ + start_boundary = round_up(offset, granularity); + end_boundary = round_down(offset + len, granularity); + + ASSERT(start_boundary >= offset); + ASSERT(end_boundary <= offset + len); + + if (!(attr_flags & XFS_ATTR_NOLOCK)) + xfs_ilock(ip, XFS_IOLOCK_EXCL); + + if (start_boundary < end_boundary - 1) { + /* punch out the page cache over the conversion range */ + truncate_pagecache_range(VFS_I(ip), start_boundary, + end_boundary - 1); + /* convert the blocks */ + error = xfs_alloc_file_space(ip, start_boundary, + end_boundary - start_boundary - 1, + XFS_BMAPI_PREALLOC | XFS_BMAPI_CONVERT, + attr_flags); + if (error) + goto out_unlock; + + /* We've handled the interior of the range, now for the edges */ + if (start_boundary != offset) + error = xfs_iozero(ip, offset, start_boundary - offset); + if (error) + goto out_unlock; + + if (end_boundary != offset + len) + error = xfs_iozero(ip, end_boundary, + offset + len - end_boundary); + + } else { + /* + * It's either a sub-granularity range or the range spanned lies + * partially across two adjacent blocks. + */ + error = xfs_iozero(ip, offset, len); + } + +out_unlock: + if (!(attr_flags & XFS_ATTR_NOLOCK)) + xfs_iunlock(ip, XFS_IOLOCK_EXCL); + return error; + +} + /* * xfs_change_file_space() * This routine allocates or frees disk space for the given file. @@ -2120,10 +2187,8 @@ xfs_change_file_space( xfs_fsize_t fsize; int setprealloc; xfs_off_t startoffset; - xfs_off_t llen; xfs_trans_t *tp; struct iattr iattr; - int prealloc_type; if (!S_ISREG(ip->i_d.di_mode)) return XFS_ERROR(EINVAL); @@ -2141,12 +2206,30 @@ xfs_change_file_space( return XFS_ERROR(EINVAL); } - llen = bf->l_len > 0 ? bf->l_len - 1 : bf->l_len; + /* + * length of <= 0 for resv/unresv/zero is invalid. length for + * alloc/free is ignored completely and we have no idea what userspace + * might have set it to, so set it to zero to allow range + * checks to pass. + */ + switch (cmd) { + case XFS_IOC_ZERO_RANGE: + case XFS_IOC_RESVSP: + case XFS_IOC_RESVSP64: + case XFS_IOC_UNRESVSP: + case XFS_IOC_UNRESVSP64: + if (bf->l_len <= 0) + return XFS_ERROR(EINVAL); + break; + default: + bf->l_len = 0; + break; + } if (bf->l_start < 0 || bf->l_start > mp->m_super->s_maxbytes || - bf->l_start + llen < 0 || - bf->l_start + llen > mp->m_super->s_maxbytes) + bf->l_start + bf->l_len < 0 || + bf->l_start + bf->l_len >= mp->m_super->s_maxbytes) return XFS_ERROR(EINVAL); bf->l_whence = 0; @@ -2154,29 +2237,20 @@ xfs_change_file_space( startoffset = bf->l_start; fsize = XFS_ISIZE(ip); - /* - * XFS_IOC_RESVSP and XFS_IOC_UNRESVSP will reserve or unreserve - * file space. - * These calls do NOT zero the data space allocated to the file, - * nor do they change the file size. - * - * XFS_IOC_ALLOCSP and XFS_IOC_FREESP will allocate and free file - * space. - * These calls cause the new file data to be zeroed and the file - * size to be changed. - */ setprealloc = clrprealloc = 0; - prealloc_type = XFS_BMAPI_PREALLOC; - switch (cmd) { case XFS_IOC_ZERO_RANGE: - prealloc_type |= XFS_BMAPI_CONVERT; - xfs_tosspages(ip, startoffset, startoffset + bf->l_len, 0); - /* FALLTHRU */ + error = xfs_zero_file_space(ip, startoffset, bf->l_len, + attr_flags); + if (error) + return error; + setprealloc = 1; + break; + case XFS_IOC_RESVSP: case XFS_IOC_RESVSP64: error = xfs_alloc_file_space(ip, startoffset, bf->l_len, - prealloc_type, attr_flags); + XFS_BMAPI_PREALLOC, attr_flags); if (error) return error; setprealloc = 1; diff --git a/fs/xfs/xfs_vnodeops.h b/fs/xfs/xfs_vnodeops.h index 447e146b2ba6..5163022d9808 100644 --- a/fs/xfs/xfs_vnodeops.h +++ b/fs/xfs/xfs_vnodeops.h @@ -48,14 +48,9 @@ int xfs_attr_set(struct xfs_inode *dp, const unsigned char *name, int xfs_attr_remove(struct xfs_inode *dp, const unsigned char *name, int flags); int xfs_attr_list(struct xfs_inode *dp, char *buffer, int bufsize, int flags, struct attrlist_cursor_kern *cursor); -void xfs_tosspages(struct xfs_inode *inode, xfs_off_t first, - xfs_off_t last, int fiopt); -int xfs_flushinval_pages(struct xfs_inode *ip, xfs_off_t first, - xfs_off_t last, int fiopt); -int xfs_flush_pages(struct xfs_inode *ip, xfs_off_t first, - xfs_off_t last, uint64_t flags, int fiopt); -int xfs_wait_on_pages(struct xfs_inode *ip, xfs_off_t first, xfs_off_t last); +int xfs_iozero(struct xfs_inode *, loff_t, size_t); int xfs_zero_eof(struct xfs_inode *, xfs_off_t, xfs_fsize_t); +int xfs_free_eofblocks(struct xfs_mount *, struct xfs_inode *, bool); #endif /* _XFS_VNODEOPS_H */ |