From 2cb1599f9b2ecdd7a9e59feeee647eb258966839 Mon Sep 17 00:00:00 2001 From: David Chinner Date: Thu, 30 Oct 2008 17:32:23 +1100 Subject: Inode: Allow external initialisers To allow XFS to combine the XFS and linux inodes into a single structure, we need to drive inode lookup from the XFS inode cache, not the generic inode cache. This means that we need initialise a struct inode from a context outside alloc_inode() as it is no longer used by XFS. Factor and export the struct inode initialisation code from alloc_inode() to inode_init_always() as a counterpart to inode_init_once(). i.e. we have to call this init function for each inode instantiation (always), as opposed inode_init_once() which is only called on slab object instantiation (once). Signed-off-by: Dave Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- include/linux/fs.h | 1 + 1 file changed, 1 insertion(+) (limited to 'include/linux/fs.h') diff --git a/include/linux/fs.h b/include/linux/fs.h index 5b248d61430c..04abead4b021 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1881,6 +1881,7 @@ extern loff_t default_llseek(struct file *file, loff_t offset, int origin); extern loff_t vfs_llseek(struct file *file, loff_t offset, int origin); +extern struct inode * inode_init_always(struct super_block *, struct inode *); extern void inode_init_once(struct inode *); extern void iput(struct inode *); extern struct inode * igrab(struct inode *); -- cgit v1.2.3 From 8290c35f87304a6b73d4fd17b03580b4f7425de8 Mon Sep 17 00:00:00 2001 From: David Chinner Date: Thu, 30 Oct 2008 17:35:24 +1100 Subject: Inode: Allow external list initialisation To allow XFS to combine the XFS and linux inodes into a single structure, we need to drive inode lookup from the XFS inode cache, not the generic inode cache. This means that we need initialise a struct inode from a context outside alloc_inode() as it is no longer used by XFS. After inode allocation and initialisation, we need to add the inode to the superblock list, the in-use list, hash it and do some accounting. This all needs to be done with the inode_lock held and there are already several places in fs/inode.c that do this list manipulation. Factor out the common code, add a locking wrapper and export the function so ti can be called from XFS. Signed-off-by: Dave Chinner Signed-off-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- fs/inode.c | 67 +++++++++++++++++++++++++++++++++++++----------------- include/linux/fs.h | 1 + 2 files changed, 47 insertions(+), 21 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/inode.c b/fs/inode.c index e7ee99907d60..fbcf6c5e7605 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -550,6 +550,49 @@ repeat: return node ? inode : NULL; } +static unsigned long hash(struct super_block *sb, unsigned long hashval) +{ + unsigned long tmp; + + tmp = (hashval * (unsigned long)sb) ^ (GOLDEN_RATIO_PRIME + hashval) / + L1_CACHE_BYTES; + tmp = tmp ^ ((tmp ^ GOLDEN_RATIO_PRIME) >> I_HASHBITS); + return tmp & I_HASHMASK; +} + +static inline void +__inode_add_to_lists(struct super_block *sb, struct hlist_head *head, + struct inode *inode) +{ + inodes_stat.nr_inodes++; + list_add(&inode->i_list, &inode_in_use); + list_add(&inode->i_sb_list, &sb->s_inodes); + if (head) + hlist_add_head(&inode->i_hash, head); +} + +/** + * inode_add_to_lists - add a new inode to relevant lists + * @sb - superblock inode belongs to. + * @inode - inode to mark in use + * + * When an inode is allocated it needs to be accounted for, added to the in use + * list, the owning superblock and the inode hash. This needs to be done under + * the inode_lock, so export a function to do this rather than the inode lock + * itself. We calculate the hash list to add to here so it is all internal + * which requires the caller to have already set up the inode number in the + * inode to add. + */ +void inode_add_to_lists(struct super_block *sb, struct inode *inode) +{ + struct hlist_head *head = inode_hashtable + hash(sb, inode->i_ino); + + spin_lock(&inode_lock); + __inode_add_to_lists(sb, head, inode); + spin_unlock(&inode_lock); +} +EXPORT_SYMBOL_GPL(inode_add_to_lists); + /** * new_inode - obtain an inode * @sb: superblock @@ -577,9 +620,7 @@ struct inode *new_inode(struct super_block *sb) inode = alloc_inode(sb); if (inode) { spin_lock(&inode_lock); - inodes_stat.nr_inodes++; - list_add(&inode->i_list, &inode_in_use); - list_add(&inode->i_sb_list, &sb->s_inodes); + __inode_add_to_lists(sb, NULL, inode); inode->i_ino = ++last_ino; inode->i_state = 0; spin_unlock(&inode_lock); @@ -638,10 +679,7 @@ static struct inode * get_new_inode(struct super_block *sb, struct hlist_head *h if (set(inode, data)) goto set_failed; - inodes_stat.nr_inodes++; - list_add(&inode->i_list, &inode_in_use); - list_add(&inode->i_sb_list, &sb->s_inodes); - hlist_add_head(&inode->i_hash, head); + __inode_add_to_lists(sb, head, inode); inode->i_state = I_LOCK|I_NEW; spin_unlock(&inode_lock); @@ -687,10 +725,7 @@ static struct inode * get_new_inode_fast(struct super_block *sb, struct hlist_he old = find_inode_fast(sb, head, ino); if (!old) { inode->i_ino = ino; - inodes_stat.nr_inodes++; - list_add(&inode->i_list, &inode_in_use); - list_add(&inode->i_sb_list, &sb->s_inodes); - hlist_add_head(&inode->i_hash, head); + __inode_add_to_lists(sb, head, inode); inode->i_state = I_LOCK|I_NEW; spin_unlock(&inode_lock); @@ -714,16 +749,6 @@ static struct inode * get_new_inode_fast(struct super_block *sb, struct hlist_he return inode; } -static unsigned long hash(struct super_block *sb, unsigned long hashval) -{ - unsigned long tmp; - - tmp = (hashval * (unsigned long)sb) ^ (GOLDEN_RATIO_PRIME + hashval) / - L1_CACHE_BYTES; - tmp = tmp ^ ((tmp ^ GOLDEN_RATIO_PRIME) >> I_HASHBITS); - return tmp & I_HASHMASK; -} - /** * iunique - get a unique inode number * @sb: superblock diff --git a/include/linux/fs.h b/include/linux/fs.h index 04abead4b021..1deedf235d55 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1883,6 +1883,7 @@ extern loff_t vfs_llseek(struct file *file, loff_t offset, int origin); extern struct inode * inode_init_always(struct super_block *, struct inode *); extern void inode_init_once(struct inode *); +extern void inode_add_to_lists(struct super_block *, struct inode *); extern void iput(struct inode *); extern struct inode * igrab(struct inode *); extern ino_t iunique(struct super_block *, ino_t); -- cgit v1.2.3 From da9592edebceeba1b9301beafe80ec8b9c2db0ce Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 14 Nov 2008 10:39:05 +1100 Subject: CRED: Wrap task credential accesses in the filesystem subsystem Wrap access to task credentials so that they can be separated more easily from the task_struct during the introduction of COW creds. Change most current->(|e|s|fs)[ug]id to current_(|e|s|fs)[ug]id(). Change some task->e?[ug]id to task_e?[ug]id(). In some places it makes more sense to use RCU directly rather than a convenient wrapper; these will be addressed by later patches. Signed-off-by: David Howells Reviewed-by: James Morris Acked-by: Serge Hallyn Cc: Al Viro Signed-off-by: James Morris --- fs/anon_inodes.c | 4 ++-- fs/attr.c | 4 ++-- fs/binfmt_elf_fdpic.c | 8 ++++---- fs/dquot.c | 4 ++-- fs/exec.c | 18 +++++++++--------- fs/fcntl.c | 2 +- fs/inotify_user.c | 2 +- fs/ioprio.c | 4 ++-- fs/locks.c | 2 +- fs/namei.c | 10 ++++++---- fs/namespace.c | 2 +- fs/pipe.c | 4 ++-- fs/posix_acl.c | 4 ++-- fs/quota.c | 4 ++-- include/linux/fs.h | 2 +- 15 files changed, 38 insertions(+), 36 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/anon_inodes.c b/fs/anon_inodes.c index 3662dd44896b..c16d9be1b017 100644 --- a/fs/anon_inodes.c +++ b/fs/anon_inodes.c @@ -154,8 +154,8 @@ static struct inode *anon_inode_mkinode(void) */ inode->i_state = I_DIRTY; inode->i_mode = S_IRUSR | S_IWUSR; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; return inode; } diff --git a/fs/attr.c b/fs/attr.c index 7a83819f6ba2..f4360192a938 100644 --- a/fs/attr.c +++ b/fs/attr.c @@ -29,13 +29,13 @@ int inode_change_ok(struct inode *inode, struct iattr *attr) /* Make sure a caller can chown. */ if ((ia_valid & ATTR_UID) && - (current->fsuid != inode->i_uid || + (current_fsuid() != inode->i_uid || attr->ia_uid != inode->i_uid) && !capable(CAP_CHOWN)) goto error; /* Make sure caller can chgrp. */ if ((ia_valid & ATTR_GID) && - (current->fsuid != inode->i_uid || + (current_fsuid() != inode->i_uid || (!in_group_p(attr->ia_gid) && attr->ia_gid != inode->i_gid)) && !capable(CAP_CHOWN)) goto error; diff --git a/fs/binfmt_elf_fdpic.c b/fs/binfmt_elf_fdpic.c index 5b5424cb3391..488584c87512 100644 --- a/fs/binfmt_elf_fdpic.c +++ b/fs/binfmt_elf_fdpic.c @@ -623,10 +623,10 @@ static int create_elf_fdpic_tables(struct linux_binprm *bprm, NEW_AUX_ENT(AT_BASE, interp_params->elfhdr_addr); NEW_AUX_ENT(AT_FLAGS, 0); NEW_AUX_ENT(AT_ENTRY, exec_params->entry_addr); - NEW_AUX_ENT(AT_UID, (elf_addr_t) current->uid); - NEW_AUX_ENT(AT_EUID, (elf_addr_t) current->euid); - NEW_AUX_ENT(AT_GID, (elf_addr_t) current->gid); - NEW_AUX_ENT(AT_EGID, (elf_addr_t) current->egid); + NEW_AUX_ENT(AT_UID, (elf_addr_t) current_uid()); + NEW_AUX_ENT(AT_EUID, (elf_addr_t) current_euid()); + NEW_AUX_ENT(AT_GID, (elf_addr_t) current_gid()); + NEW_AUX_ENT(AT_EGID, (elf_addr_t) current_egid()); NEW_AUX_ENT(AT_SECURE, security_bprm_secureexec(bprm)); NEW_AUX_ENT(AT_EXECFN, bprm->exec); diff --git a/fs/dquot.c b/fs/dquot.c index 5e95261005b2..c237ccc8581c 100644 --- a/fs/dquot.c +++ b/fs/dquot.c @@ -874,7 +874,7 @@ static inline int need_print_warning(struct dquot *dquot) switch (dquot->dq_type) { case USRQUOTA: - return current->fsuid == dquot->dq_id; + return current_fsuid() == dquot->dq_id; case GRPQUOTA: return in_group_p(dquot->dq_id); } @@ -981,7 +981,7 @@ static void send_warning(const struct dquot *dquot, const char warntype) MINOR(dquot->dq_sb->s_dev)); if (ret) goto attr_err_out; - ret = nla_put_u64(skb, QUOTA_NL_A_CAUSED_ID, current->user->uid); + ret = nla_put_u64(skb, QUOTA_NL_A_CAUSED_ID, current_uid()); if (ret) goto attr_err_out; genlmsg_end(skb, msg_head); diff --git a/fs/exec.c b/fs/exec.c index 4e834f16d9da..604834f3b208 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -980,7 +980,7 @@ int flush_old_exec(struct linux_binprm * bprm) /* This is the point of no return */ current->sas_ss_sp = current->sas_ss_size = 0; - if (current->euid == current->uid && current->egid == current->gid) + if (current_euid() == current_uid() && current_egid() == current_gid()) set_dumpable(current->mm, 1); else set_dumpable(current->mm, suid_dumpable); @@ -1007,7 +1007,7 @@ int flush_old_exec(struct linux_binprm * bprm) */ current->mm->task_size = TASK_SIZE; - if (bprm->e_uid != current->euid || bprm->e_gid != current->egid) { + if (bprm->e_uid != current_euid() || bprm->e_gid != current_egid()) { suid_keys(current); set_dumpable(current->mm, suid_dumpable); current->pdeath_signal = 0; @@ -1047,8 +1047,8 @@ int prepare_binprm(struct linux_binprm *bprm) if (bprm->file->f_op == NULL) return -EACCES; - bprm->e_uid = current->euid; - bprm->e_gid = current->egid; + bprm->e_uid = current_euid(); + bprm->e_gid = current_egid(); if(!(bprm->file->f_path.mnt->mnt_flags & MNT_NOSUID)) { /* Set-uid? */ @@ -1096,7 +1096,7 @@ void compute_creds(struct linux_binprm *bprm) { int unsafe; - if (bprm->e_uid != current->uid) { + if (bprm->e_uid != current_uid()) { suid_keys(current); current->pdeath_signal = 0; } @@ -1424,7 +1424,7 @@ static int format_corename(char *corename, long signr) /* uid */ case 'u': rc = snprintf(out_ptr, out_end - out_ptr, - "%d", current->uid); + "%d", current_uid()); if (rc > out_end - out_ptr) goto out; out_ptr += rc; @@ -1432,7 +1432,7 @@ static int format_corename(char *corename, long signr) /* gid */ case 'g': rc = snprintf(out_ptr, out_end - out_ptr, - "%d", current->gid); + "%d", current_gid()); if (rc > out_end - out_ptr) goto out; out_ptr += rc; @@ -1709,7 +1709,7 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) struct inode * inode; struct file * file; int retval = 0; - int fsuid = current->fsuid; + int fsuid = current_fsuid(); int flag = 0; int ispipe = 0; unsigned long core_limit = current->signal->rlim[RLIMIT_CORE].rlim_cur; @@ -1815,7 +1815,7 @@ int do_coredump(long signr, int exit_code, struct pt_regs * regs) * Dont allow local users get cute and trick others to coredump * into their pre-created files: */ - if (inode->i_uid != current->fsuid) + if (inode->i_uid != current_fsuid()) goto close_fail; if (!file->f_op) goto close_fail; diff --git a/fs/fcntl.c b/fs/fcntl.c index ac4f7db9f134..bf049a805e59 100644 --- a/fs/fcntl.c +++ b/fs/fcntl.c @@ -211,7 +211,7 @@ int __f_setown(struct file *filp, struct pid *pid, enum pid_type type, if (err) return err; - f_modown(filp, pid, type, current->uid, current->euid, force); + f_modown(filp, pid, type, current_uid(), current_euid(), force); return 0; } EXPORT_SYMBOL(__f_setown); diff --git a/fs/inotify_user.c b/fs/inotify_user.c index d367e9b92862..e2425bbd871f 100644 --- a/fs/inotify_user.c +++ b/fs/inotify_user.c @@ -601,7 +601,7 @@ asmlinkage long sys_inotify_init1(int flags) goto out_put_fd; } - user = get_uid(current->user); + user = get_current_user(); if (unlikely(atomic_read(&user->inotify_devs) >= inotify_max_user_instances)) { ret = -EMFILE; diff --git a/fs/ioprio.c b/fs/ioprio.c index da3cc460d4df..68d2cd807118 100644 --- a/fs/ioprio.c +++ b/fs/ioprio.c @@ -32,8 +32,8 @@ static int set_task_ioprio(struct task_struct *task, int ioprio) int err; struct io_context *ioc; - if (task->uid != current->euid && - task->uid != current->uid && !capable(CAP_SYS_NICE)) + if (task->uid != current_euid() && + task->uid != current_uid() && !capable(CAP_SYS_NICE)) return -EPERM; err = security_task_setioprio(task, ioprio); diff --git a/fs/locks.c b/fs/locks.c index 09062e3ff104..46a2e12f7d42 100644 --- a/fs/locks.c +++ b/fs/locks.c @@ -1349,7 +1349,7 @@ int generic_setlease(struct file *filp, long arg, struct file_lock **flp) struct inode *inode = dentry->d_inode; int error, rdlease_count = 0, wrlease_count = 0; - if ((current->fsuid != inode->i_uid) && !capable(CAP_LEASE)) + if ((current_fsuid() != inode->i_uid) && !capable(CAP_LEASE)) return -EACCES; if (!S_ISREG(inode->i_mode)) return -EINVAL; diff --git a/fs/namei.c b/fs/namei.c index 09ce58e49e72..42d7b7606936 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -186,7 +186,7 @@ int generic_permission(struct inode *inode, int mask, mask &= MAY_READ | MAY_WRITE | MAY_EXEC; - if (current->fsuid == inode->i_uid) + if (current_fsuid() == inode->i_uid) mode >>= 6; else { if (IS_POSIXACL(inode) && (mode & S_IRWXG) && check_acl) { @@ -441,7 +441,7 @@ static int exec_permission_lite(struct inode *inode) if (inode->i_op && inode->i_op->permission) return -EAGAIN; - if (current->fsuid == inode->i_uid) + if (current_fsuid() == inode->i_uid) mode >>= 6; else if (in_group_p(inode->i_gid)) mode >>= 3; @@ -1334,11 +1334,13 @@ static int user_path_parent(int dfd, const char __user *path, */ static inline int check_sticky(struct inode *dir, struct inode *inode) { + uid_t fsuid = current_fsuid(); + if (!(dir->i_mode & S_ISVTX)) return 0; - if (inode->i_uid == current->fsuid) + if (inode->i_uid == fsuid) return 0; - if (dir->i_uid == current->fsuid) + if (dir->i_uid == fsuid) return 0; return !capable(CAP_FOWNER); } diff --git a/fs/namespace.c b/fs/namespace.c index cce46702d33c..d8bc2c4704a5 100644 --- a/fs/namespace.c +++ b/fs/namespace.c @@ -1176,7 +1176,7 @@ static int mount_is_safe(struct path *path) if (S_ISLNK(path->dentry->d_inode->i_mode)) return -EPERM; if (path->dentry->d_inode->i_mode & S_ISVTX) { - if (current->uid != path->dentry->d_inode->i_uid) + if (current_uid() != path->dentry->d_inode->i_uid) return -EPERM; } if (inode_permission(path->dentry->d_inode, MAY_WRITE)) diff --git a/fs/pipe.c b/fs/pipe.c index 7aea8b89baac..aaf797bd57b9 100644 --- a/fs/pipe.c +++ b/fs/pipe.c @@ -899,8 +899,8 @@ static struct inode * get_pipe_inode(void) */ inode->i_state = I_DIRTY; inode->i_mode = S_IFIFO | S_IRUSR | S_IWUSR; - inode->i_uid = current->fsuid; - inode->i_gid = current->fsgid; + inode->i_uid = current_fsuid(); + inode->i_gid = current_fsgid(); inode->i_atime = inode->i_mtime = inode->i_ctime = CURRENT_TIME; return inode; diff --git a/fs/posix_acl.c b/fs/posix_acl.c index aec931e09973..39df95a0ec25 100644 --- a/fs/posix_acl.c +++ b/fs/posix_acl.c @@ -217,11 +217,11 @@ posix_acl_permission(struct inode *inode, const struct posix_acl *acl, int want) switch(pa->e_tag) { case ACL_USER_OBJ: /* (May have been checked already) */ - if (inode->i_uid == current->fsuid) + if (inode->i_uid == current_fsuid()) goto check_perm; break; case ACL_USER: - if (pa->e_id == current->fsuid) + if (pa->e_id == current_fsuid()) goto mask; break; case ACL_GROUP_OBJ: diff --git a/fs/quota.c b/fs/quota.c index 7f4386ebc23a..b7fe44e01618 100644 --- a/fs/quota.c +++ b/fs/quota.c @@ -79,7 +79,7 @@ static int generic_quotactl_valid(struct super_block *sb, int type, int cmd, qid /* Check privileges */ if (cmd == Q_GETQUOTA) { - if (((type == USRQUOTA && current->euid != id) || + if (((type == USRQUOTA && current_euid() != id) || (type == GRPQUOTA && !in_egroup_p(id))) && !capable(CAP_SYS_ADMIN)) return -EPERM; @@ -130,7 +130,7 @@ static int xqm_quotactl_valid(struct super_block *sb, int type, int cmd, qid_t i /* Check privileges */ if (cmd == Q_XGETQUOTA) { - if (((type == XQM_USRQUOTA && current->euid != id) || + if (((type == XQM_USRQUOTA && current_euid() != id) || (type == XQM_GRPQUOTA && !in_egroup_p(id))) && !capable(CAP_SYS_ADMIN)) return -EPERM; diff --git a/include/linux/fs.h b/include/linux/fs.h index 0dcdd9458f4b..b3d404aaabed 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1193,7 +1193,7 @@ enum { #define has_fs_excl() atomic_read(¤t->fs_excl) #define is_owner_or_cap(inode) \ - ((current->fsuid == (inode)->i_uid) || capable(CAP_FOWNER)) + ((current_fsuid() == (inode)->i_uid) || capable(CAP_FOWNER)) /* not quite ready to be deprecated, but... */ extern void lock_super(struct super_block *); -- cgit v1.2.3 From 745ca2475a6ac596e3d8d37c2759c0fbe2586227 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 14 Nov 2008 10:39:22 +1100 Subject: CRED: Pass credentials through dentry_open() Pass credentials through dentry_open() so that the COW creds patch can have SELinux's flush_unauthorized_files() pass the appropriate creds back to itself when it opens its null chardev. The security_dentry_open() call also now takes a creds pointer, as does the dentry_open hook in struct security_operations. Signed-off-by: David Howells Acked-by: James Morris Signed-off-by: James Morris --- arch/powerpc/platforms/cell/spufs/inode.c | 4 ++-- arch/um/drivers/mconsole_kern.c | 3 ++- fs/autofs4/dev-ioctl.c | 3 ++- fs/ecryptfs/ecryptfs_kernel.h | 3 ++- fs/ecryptfs/kthread.c | 9 +++++---- fs/ecryptfs/main.c | 3 ++- fs/exportfs/expfs.c | 4 +++- fs/hppfs/hppfs.c | 6 ++++-- fs/nfsctl.c | 3 ++- fs/nfsd/nfs4recover.c | 3 ++- fs/nfsd/vfs.c | 3 ++- fs/open.c | 17 +++++++++++------ fs/xfs/linux-2.6/xfs_ioctl.c | 3 ++- include/linux/fs.h | 4 +++- include/linux/security.h | 7 ++++--- ipc/mqueue.c | 11 +++++++---- security/capability.c | 2 +- security/security.c | 4 ++-- security/selinux/hooks.c | 15 +++++++++------ 19 files changed, 67 insertions(+), 40 deletions(-) (limited to 'include/linux/fs.h') diff --git a/arch/powerpc/platforms/cell/spufs/inode.c b/arch/powerpc/platforms/cell/spufs/inode.c index e128ce7f0993..6296bfd9cb0b 100644 --- a/arch/powerpc/platforms/cell/spufs/inode.c +++ b/arch/powerpc/platforms/cell/spufs/inode.c @@ -323,7 +323,7 @@ static int spufs_context_open(struct dentry *dentry, struct vfsmount *mnt) goto out; } - filp = dentry_open(dentry, mnt, O_RDONLY); + filp = dentry_open(dentry, mnt, O_RDONLY, current_cred()); if (IS_ERR(filp)) { put_unused_fd(ret); ret = PTR_ERR(filp); @@ -562,7 +562,7 @@ static int spufs_gang_open(struct dentry *dentry, struct vfsmount *mnt) goto out; } - filp = dentry_open(dentry, mnt, O_RDONLY); + filp = dentry_open(dentry, mnt, O_RDONLY, current_cred()); if (IS_ERR(filp)) { put_unused_fd(ret); ret = PTR_ERR(filp); diff --git a/arch/um/drivers/mconsole_kern.c b/arch/um/drivers/mconsole_kern.c index 19d579d74d27..16d3b3789a50 100644 --- a/arch/um/drivers/mconsole_kern.c +++ b/arch/um/drivers/mconsole_kern.c @@ -159,7 +159,8 @@ void mconsole_proc(struct mc_request *req) goto out_kill; } - file = dentry_open(nd.path.dentry, nd.path.mnt, O_RDONLY); + file = dentry_open(nd.path.dentry, nd.path.mnt, O_RDONLY, + current_cred()); if (IS_ERR(file)) { mconsole_reply(req, "Failed to open file", 1, 0); goto out_kill; diff --git a/fs/autofs4/dev-ioctl.c b/fs/autofs4/dev-ioctl.c index 625abf5422e2..ec16255d27dd 100644 --- a/fs/autofs4/dev-ioctl.c +++ b/fs/autofs4/dev-ioctl.c @@ -307,7 +307,8 @@ static int autofs_dev_ioctl_open_mountpoint(const char *path, dev_t devid) goto out; } - filp = dentry_open(nd.path.dentry, nd.path.mnt, O_RDONLY); + filp = dentry_open(nd.path.dentry, nd.path.mnt, O_RDONLY, + current_cred()); if (IS_ERR(filp)) { err = PTR_ERR(filp); goto out; diff --git a/fs/ecryptfs/ecryptfs_kernel.h b/fs/ecryptfs/ecryptfs_kernel.h index 3504cf9df358..a75026d35d16 100644 --- a/fs/ecryptfs/ecryptfs_kernel.h +++ b/fs/ecryptfs/ecryptfs_kernel.h @@ -691,7 +691,8 @@ int ecryptfs_init_kthread(void); void ecryptfs_destroy_kthread(void); int ecryptfs_privileged_open(struct file **lower_file, struct dentry *lower_dentry, - struct vfsmount *lower_mnt); + struct vfsmount *lower_mnt, + const struct cred *cred); int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry); #endif /* #ifndef ECRYPTFS_KERNEL_H */ diff --git a/fs/ecryptfs/kthread.c b/fs/ecryptfs/kthread.c index c440c6b58b2d..c6d7a4d748a0 100644 --- a/fs/ecryptfs/kthread.c +++ b/fs/ecryptfs/kthread.c @@ -73,7 +73,7 @@ static int ecryptfs_threadfn(void *ignored) mntget(req->lower_mnt); (*req->lower_file) = dentry_open( req->lower_dentry, req->lower_mnt, - (O_RDWR | O_LARGEFILE)); + (O_RDWR | O_LARGEFILE), current_cred()); req->flags |= ECRYPTFS_REQ_PROCESSED; } wake_up(&req->wait); @@ -132,7 +132,8 @@ void ecryptfs_destroy_kthread(void) */ int ecryptfs_privileged_open(struct file **lower_file, struct dentry *lower_dentry, - struct vfsmount *lower_mnt) + struct vfsmount *lower_mnt, + const struct cred *cred) { struct ecryptfs_open_req *req; int rc = 0; @@ -143,7 +144,7 @@ int ecryptfs_privileged_open(struct file **lower_file, dget(lower_dentry); mntget(lower_mnt); (*lower_file) = dentry_open(lower_dentry, lower_mnt, - (O_RDWR | O_LARGEFILE)); + (O_RDWR | O_LARGEFILE), cred); if (!IS_ERR(*lower_file)) goto out; req = kmem_cache_alloc(ecryptfs_open_req_cache, GFP_KERNEL); @@ -184,7 +185,7 @@ int ecryptfs_privileged_open(struct file **lower_file, dget(lower_dentry); mntget(lower_mnt); (*lower_file) = dentry_open(lower_dentry, lower_mnt, - (O_RDONLY | O_LARGEFILE)); + (O_RDONLY | O_LARGEFILE), cred); if (IS_ERR(*lower_file)) { rc = PTR_ERR(*req->lower_file); (*lower_file) = NULL; diff --git a/fs/ecryptfs/main.c b/fs/ecryptfs/main.c index 64d2ba980df4..fd630713c5c7 100644 --- a/fs/ecryptfs/main.c +++ b/fs/ecryptfs/main.c @@ -115,6 +115,7 @@ void __ecryptfs_printk(const char *fmt, ...) */ int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry) { + const struct cred *cred = current_cred(); struct ecryptfs_inode_info *inode_info = ecryptfs_inode_to_private(ecryptfs_dentry->d_inode); int rc = 0; @@ -127,7 +128,7 @@ int ecryptfs_init_persistent_file(struct dentry *ecryptfs_dentry) lower_dentry = ecryptfs_dentry_to_lower(ecryptfs_dentry); rc = ecryptfs_privileged_open(&inode_info->lower_file, - lower_dentry, lower_mnt); + lower_dentry, lower_mnt, cred); if (rc || IS_ERR(inode_info->lower_file)) { printk(KERN_ERR "Error opening lower persistent file " "for lower_dentry [0x%p] and lower_mnt [0x%p]; " diff --git a/fs/exportfs/expfs.c b/fs/exportfs/expfs.c index 80246bad1b7f..ec1fb918200f 100644 --- a/fs/exportfs/expfs.c +++ b/fs/exportfs/expfs.c @@ -14,6 +14,7 @@ #include #include #include +#include #define dprintk(fmt, args...) do{}while(0) @@ -249,6 +250,7 @@ static int filldir_one(void * __buf, const char * name, int len, static int get_name(struct vfsmount *mnt, struct dentry *dentry, char *name, struct dentry *child) { + const struct cred *cred = current_cred(); struct inode *dir = dentry->d_inode; int error; struct file *file; @@ -263,7 +265,7 @@ static int get_name(struct vfsmount *mnt, struct dentry *dentry, /* * Open the directory ... */ - file = dentry_open(dget(dentry), mntget(mnt), O_RDONLY); + file = dentry_open(dget(dentry), mntget(mnt), O_RDONLY, cred); error = PTR_ERR(file); if (IS_ERR(file)) goto out; diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c index 2b3d1828db99..795e2c130ad7 100644 --- a/fs/hppfs/hppfs.c +++ b/fs/hppfs/hppfs.c @@ -426,6 +426,7 @@ static int file_mode(int fmode) static int hppfs_open(struct inode *inode, struct file *file) { + const struct cred *cred = current_cred(); struct hppfs_private *data; struct vfsmount *proc_mnt; struct dentry *proc_dentry; @@ -446,7 +447,7 @@ static int hppfs_open(struct inode *inode, struct file *file) /* XXX This isn't closed anywhere */ data->proc_file = dentry_open(dget(proc_dentry), mntget(proc_mnt), - file_mode(file->f_mode)); + file_mode(file->f_mode), cred); err = PTR_ERR(data->proc_file); if (IS_ERR(data->proc_file)) goto out_free1; @@ -489,6 +490,7 @@ static int hppfs_open(struct inode *inode, struct file *file) static int hppfs_dir_open(struct inode *inode, struct file *file) { + const struct cred *cred = current_cred(); struct hppfs_private *data; struct vfsmount *proc_mnt; struct dentry *proc_dentry; @@ -502,7 +504,7 @@ static int hppfs_dir_open(struct inode *inode, struct file *file) proc_dentry = HPPFS_I(inode)->proc_dentry; proc_mnt = inode->i_sb->s_fs_info; data->proc_file = dentry_open(dget(proc_dentry), mntget(proc_mnt), - file_mode(file->f_mode)); + file_mode(file->f_mode), cred); err = PTR_ERR(data->proc_file); if (IS_ERR(data->proc_file)) goto out_free; diff --git a/fs/nfsctl.c b/fs/nfsctl.c index aed8145d9087..cc4ef2642a51 100644 --- a/fs/nfsctl.c +++ b/fs/nfsctl.c @@ -41,7 +41,8 @@ static struct file *do_open(char *name, int flags) error = may_open(&nd, MAY_WRITE, FMODE_WRITE); if (!error) - return dentry_open(nd.path.dentry, nd.path.mnt, flags); + return dentry_open(nd.path.dentry, nd.path.mnt, flags, + current_cred()); path_put(&nd.path); return ERR_PTR(error); diff --git a/fs/nfsd/nfs4recover.c b/fs/nfsd/nfs4recover.c index a5e14e8695ea..632a50b4b371 100644 --- a/fs/nfsd/nfs4recover.c +++ b/fs/nfsd/nfs4recover.c @@ -226,7 +226,8 @@ nfsd4_list_rec_dir(struct dentry *dir, recdir_func *f) nfs4_save_user(&uid, &gid); - filp = dentry_open(dget(dir), mntget(rec_dir.mnt), O_RDONLY); + filp = dentry_open(dget(dir), mntget(rec_dir.mnt), O_RDONLY, + current_cred()); status = PTR_ERR(filp); if (IS_ERR(filp)) goto out; diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 890d9a68c852..b59ec5a6ed24 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -671,6 +671,7 @@ __be32 nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, int access, struct file **filp) { + const struct cred *cred = current_cred(); struct dentry *dentry; struct inode *inode; int flags = O_RDONLY|O_LARGEFILE; @@ -725,7 +726,7 @@ nfsd_open(struct svc_rqst *rqstp, struct svc_fh *fhp, int type, DQUOT_INIT(inode); } *filp = dentry_open(dget(dentry), mntget(fhp->fh_export->ex_path.mnt), - flags); + flags, cred); if (IS_ERR(*filp)) host_err = PTR_ERR(*filp); out_nfserr: diff --git a/fs/open.c b/fs/open.c index b1238e195e7e..f96eaab280a3 100644 --- a/fs/open.c +++ b/fs/open.c @@ -783,7 +783,8 @@ static inline int __get_file_write_access(struct inode *inode, static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags, struct file *f, - int (*open)(struct inode *, struct file *)) + int (*open)(struct inode *, struct file *), + const struct cred *cred) { struct inode *inode; int error; @@ -807,7 +808,7 @@ static struct file *__dentry_open(struct dentry *dentry, struct vfsmount *mnt, f->f_op = fops_get(inode->i_fop); file_move(f, &inode->i_sb->s_files); - error = security_dentry_open(f); + error = security_dentry_open(f, cred); if (error) goto cleanup_all; @@ -882,6 +883,8 @@ cleanup_file: struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry, int (*open)(struct inode *, struct file *)) { + const struct cred *cred = current_cred(); + if (IS_ERR(nd->intent.open.file)) goto out; if (IS_ERR(dentry)) @@ -889,7 +892,7 @@ struct file *lookup_instantiate_filp(struct nameidata *nd, struct dentry *dentry nd->intent.open.file = __dentry_open(dget(dentry), mntget(nd->path.mnt), nd->intent.open.flags - 1, nd->intent.open.file, - open); + open, cred); out: return nd->intent.open.file; out_err: @@ -908,6 +911,7 @@ EXPORT_SYMBOL_GPL(lookup_instantiate_filp); */ struct file *nameidata_to_filp(struct nameidata *nd, int flags) { + const struct cred *cred = current_cred(); struct file *filp; /* Pick up the filp from the open intent */ @@ -915,7 +919,7 @@ struct file *nameidata_to_filp(struct nameidata *nd, int flags) /* Has the filesystem initialised the file for us? */ if (filp->f_path.dentry == NULL) filp = __dentry_open(nd->path.dentry, nd->path.mnt, flags, filp, - NULL); + NULL, cred); else path_put(&nd->path); return filp; @@ -925,7 +929,8 @@ struct file *nameidata_to_filp(struct nameidata *nd, int flags) * dentry_open() will have done dput(dentry) and mntput(mnt) if it returns an * error. */ -struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags) +struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags, + const struct cred *cred) { int error; struct file *f; @@ -950,7 +955,7 @@ struct file *dentry_open(struct dentry *dentry, struct vfsmount *mnt, int flags) return ERR_PTR(error); } - return __dentry_open(dentry, mnt, flags, f, NULL); + return __dentry_open(dentry, mnt, flags, f, NULL, cred); } EXPORT_SYMBOL(dentry_open); diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index 67c72aec97e6..281cbd5a25cf 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -256,6 +256,7 @@ xfs_open_by_handle( struct file *parfilp, struct inode *parinode) { + const struct cred *cred = current_cred(); int error; int new_fd; int permflag; @@ -321,7 +322,7 @@ xfs_open_by_handle( mntget(parfilp->f_path.mnt); /* Create file pointer. */ - filp = dentry_open(dentry, parfilp->f_path.mnt, hreq.oflags); + filp = dentry_open(dentry, parfilp->f_path.mnt, hreq.oflags, cred); if (IS_ERR(filp)) { put_unused_fd(new_fd); return -XFS_ERROR(-PTR_ERR(filp)); diff --git a/include/linux/fs.h b/include/linux/fs.h index b3d404aaabed..3bfec1327b8d 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -315,6 +315,7 @@ struct poll_table_struct; struct kstatfs; struct vm_area_struct; struct vfsmount; +struct cred; extern void __init inode_init(void); extern void __init inode_init_early(void); @@ -1673,7 +1674,8 @@ extern int do_truncate(struct dentry *, loff_t start, unsigned int time_attrs, extern long do_sys_open(int dfd, const char __user *filename, int flags, int mode); extern struct file *filp_open(const char *, int, int); -extern struct file * dentry_open(struct dentry *, struct vfsmount *, int); +extern struct file * dentry_open(struct dentry *, struct vfsmount *, int, + const struct cred *); extern int filp_close(struct file *, fl_owner_t id); extern char * getname(const char __user *); diff --git a/include/linux/security.h b/include/linux/security.h index 9239cc11eb9c..7e9fe046a0d1 100644 --- a/include/linux/security.h +++ b/include/linux/security.h @@ -1402,7 +1402,7 @@ struct security_operations { int (*file_send_sigiotask) (struct task_struct *tsk, struct fown_struct *fown, int sig); int (*file_receive) (struct file *file); - int (*dentry_open) (struct file *file); + int (*dentry_open) (struct file *file, const struct cred *cred); int (*task_create) (unsigned long clone_flags); int (*cred_alloc_security) (struct cred *cred); @@ -1658,7 +1658,7 @@ int security_file_set_fowner(struct file *file); int security_file_send_sigiotask(struct task_struct *tsk, struct fown_struct *fown, int sig); int security_file_receive(struct file *file); -int security_dentry_open(struct file *file); +int security_dentry_open(struct file *file, const struct cred *cred); int security_task_create(unsigned long clone_flags); int security_cred_alloc(struct cred *cred); void security_cred_free(struct cred *cred); @@ -2171,7 +2171,8 @@ static inline int security_file_receive(struct file *file) return 0; } -static inline int security_dentry_open(struct file *file) +static inline int security_dentry_open(struct file *file, + const struct cred *cred) { return 0; } diff --git a/ipc/mqueue.c b/ipc/mqueue.c index 1151881ccb9a..d9393f8e4c3e 100644 --- a/ipc/mqueue.c +++ b/ipc/mqueue.c @@ -594,6 +594,7 @@ static int mq_attr_ok(struct mq_attr *attr) static struct file *do_create(struct dentry *dir, struct dentry *dentry, int oflag, mode_t mode, struct mq_attr __user *u_attr) { + const struct cred *cred = current_cred(); struct mq_attr attr; struct file *result; int ret; @@ -618,7 +619,7 @@ static struct file *do_create(struct dentry *dir, struct dentry *dentry, if (ret) goto out_drop_write; - result = dentry_open(dentry, mqueue_mnt, oflag); + result = dentry_open(dentry, mqueue_mnt, oflag, cred); /* * dentry_open() took a persistent mnt_want_write(), * so we can now drop this one. @@ -637,8 +638,10 @@ out: /* Opens existing queue */ static struct file *do_open(struct dentry *dentry, int oflag) { -static int oflag2acc[O_ACCMODE] = { MAY_READ, MAY_WRITE, - MAY_READ | MAY_WRITE }; + const struct cred *cred = current_cred(); + + static const int oflag2acc[O_ACCMODE] = { MAY_READ, MAY_WRITE, + MAY_READ | MAY_WRITE }; if ((oflag & O_ACCMODE) == (O_RDWR | O_WRONLY)) { dput(dentry); @@ -652,7 +655,7 @@ static int oflag2acc[O_ACCMODE] = { MAY_READ, MAY_WRITE, return ERR_PTR(-EACCES); } - return dentry_open(dentry, mqueue_mnt, oflag); + return dentry_open(dentry, mqueue_mnt, oflag, cred); } asmlinkage long sys_mq_open(const char __user *u_name, int oflag, mode_t mode, diff --git a/security/capability.c b/security/capability.c index 6c4b5137ca7b..fac2f61b69a9 100644 --- a/security/capability.c +++ b/security/capability.c @@ -330,7 +330,7 @@ static int cap_file_receive(struct file *file) return 0; } -static int cap_dentry_open(struct file *file) +static int cap_dentry_open(struct file *file, const struct cred *cred) { return 0; } diff --git a/security/security.c b/security/security.c index d058f7d5b10a..f40a0a04c3c2 100644 --- a/security/security.c +++ b/security/security.c @@ -606,9 +606,9 @@ int security_file_receive(struct file *file) return security_ops->file_receive(file); } -int security_dentry_open(struct file *file) +int security_dentry_open(struct file *file, const struct cred *cred) { - return security_ops->dentry_open(file); + return security_ops->dentry_open(file, cred); } int security_task_create(unsigned long clone_flags) diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c index cc6e5a3f10cc..f20cbd681ba6 100644 --- a/security/selinux/hooks.c +++ b/security/selinux/hooks.c @@ -2150,9 +2150,9 @@ extern struct vfsmount *selinuxfs_mount; extern struct dentry *selinux_null; /* Derived from fs/exec.c:flush_old_files. */ -static inline void flush_unauthorized_files(struct files_struct *files) +static inline void flush_unauthorized_files(const struct cred *cred, + struct files_struct *files) { - const struct cred *cred = current_cred(); struct avc_audit_data ad; struct file *file, *devnull = NULL; struct tty_struct *tty; @@ -2222,7 +2222,10 @@ static inline void flush_unauthorized_files(struct files_struct *files) if (devnull) { get_file(devnull); } else { - devnull = dentry_open(dget(selinux_null), mntget(selinuxfs_mount), O_RDWR); + devnull = dentry_open( + dget(selinux_null), + mntget(selinuxfs_mount), + O_RDWR, cred); if (IS_ERR(devnull)) { devnull = NULL; put_unused_fd(fd); @@ -2302,6 +2305,7 @@ static void selinux_bprm_apply_creds(struct linux_binprm *bprm, int unsafe) */ static void selinux_bprm_post_apply_creds(struct linux_binprm *bprm) { + const struct cred *cred = current_cred(); struct task_security_struct *tsec; struct rlimit *rlim, *initrlim; struct itimerval itimer; @@ -2321,7 +2325,7 @@ static void selinux_bprm_post_apply_creds(struct linux_binprm *bprm) return; /* Close files for which the new task SID is not authorized. */ - flush_unauthorized_files(current->files); + flush_unauthorized_files(cred, current->files); /* Check whether the new SID can inherit signal state from the old SID. If not, clear itimers to avoid @@ -3202,9 +3206,8 @@ static int selinux_file_receive(struct file *file) return file_has_perm(cred, file, file_to_av(file)); } -static int selinux_dentry_open(struct file *file) +static int selinux_dentry_open(struct file *file, const struct cred *cred) { - const struct cred *cred = current_cred(); struct file_security_struct *fsec; struct inode *inode; struct inode_security_struct *isec; -- cgit v1.2.3 From d76b0d9b2d87cfc95686e148767cbf7d0e22bdc0 Mon Sep 17 00:00:00 2001 From: David Howells Date: Fri, 14 Nov 2008 10:39:25 +1100 Subject: CRED: Use creds in file structs Attach creds to file structs and discard f_uid/f_gid. file_operations::open() methods (such as hppfs_open()) should use file->f_cred rather than current_cred(). At the moment file->f_cred will be current_cred() at this point. Signed-off-by: David Howells Reviewed-by: James Morris Signed-off-by: James Morris --- arch/mips/kernel/vpe.c | 4 ++-- drivers/isdn/hysdn/hysdn_procconf.c | 6 ++++-- fs/coda/file.c | 2 +- fs/file_table.c | 7 ++++--- fs/hppfs/hppfs.c | 4 ++-- include/linux/fs.h | 2 +- net/ipv4/netfilter/ipt_LOG.c | 4 ++-- net/ipv6/netfilter/ip6t_LOG.c | 4 ++-- net/netfilter/nfnetlink_log.c | 5 +++-- net/netfilter/xt_owner.c | 16 ++++++++-------- net/sched/cls_flow.c | 4 ++-- 11 files changed, 31 insertions(+), 27 deletions(-) (limited to 'include/linux/fs.h') diff --git a/arch/mips/kernel/vpe.c b/arch/mips/kernel/vpe.c index 972b2d2b8401..09786e496375 100644 --- a/arch/mips/kernel/vpe.c +++ b/arch/mips/kernel/vpe.c @@ -1085,8 +1085,8 @@ static int vpe_open(struct inode *inode, struct file *filp) v->load_addr = NULL; v->len = 0; - v->uid = filp->f_uid; - v->gid = filp->f_gid; + v->uid = filp->f_cred->fsuid; + v->gid = filp->f_cred->fsgid; #ifdef CONFIG_MIPS_APSP_KSPD /* get kspd to tell us when a syscall_exit happens */ diff --git a/drivers/isdn/hysdn/hysdn_procconf.c b/drivers/isdn/hysdn/hysdn_procconf.c index 484299b031f8..8f9f4912de32 100644 --- a/drivers/isdn/hysdn/hysdn_procconf.c +++ b/drivers/isdn/hysdn/hysdn_procconf.c @@ -246,7 +246,8 @@ hysdn_conf_open(struct inode *ino, struct file *filep) } if (card->debug_flags & (LOG_PROC_OPEN | LOG_PROC_ALL)) hysdn_addlog(card, "config open for uid=%d gid=%d mode=0x%x", - filep->f_uid, filep->f_gid, filep->f_mode); + filep->f_cred->fsuid, filep->f_cred->fsgid, + filep->f_mode); if ((filep->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_WRITE) { /* write only access -> write boot file or conf line */ @@ -331,7 +332,8 @@ hysdn_conf_close(struct inode *ino, struct file *filep) } if (card->debug_flags & (LOG_PROC_OPEN | LOG_PROC_ALL)) hysdn_addlog(card, "config close for uid=%d gid=%d mode=0x%x", - filep->f_uid, filep->f_gid, filep->f_mode); + filep->f_cred->fsuid, filep->f_cred->fsgid, + filep->f_mode); if ((filep->f_mode & (FMODE_READ | FMODE_WRITE)) == FMODE_WRITE) { /* write only access -> write boot file or conf line */ diff --git a/fs/coda/file.c b/fs/coda/file.c index 29137ff3ca67..5a8769985494 100644 --- a/fs/coda/file.c +++ b/fs/coda/file.c @@ -174,7 +174,7 @@ int coda_release(struct inode *coda_inode, struct file *coda_file) BUG_ON(!cfi || cfi->cfi_magic != CODA_MAGIC); err = venus_close(coda_inode->i_sb, coda_i2f(coda_inode), - coda_flags, coda_file->f_uid); + coda_flags, coda_file->f_cred->fsuid); host_inode = cfi->cfi_container->f_path.dentry->d_inode; cii = ITOC(coda_inode); diff --git a/fs/file_table.c b/fs/file_table.c index bc4563fe791d..0fbcacc3ea75 100644 --- a/fs/file_table.c +++ b/fs/file_table.c @@ -36,7 +36,9 @@ static struct percpu_counter nr_files __cacheline_aligned_in_smp; static inline void file_free_rcu(struct rcu_head *head) { - struct file *f = container_of(head, struct file, f_u.fu_rcuhead); + struct file *f = container_of(head, struct file, f_u.fu_rcuhead); + + put_cred(f->f_cred); kmem_cache_free(filp_cachep, f); } @@ -121,8 +123,7 @@ struct file *get_empty_filp(void) INIT_LIST_HEAD(&f->f_u.fu_list); atomic_long_set(&f->f_count, 1); rwlock_init(&f->f_owner.lock); - f->f_uid = cred->fsuid; - f->f_gid = cred->fsgid; + f->f_cred = get_cred(cred); eventpoll_init_file(f); /* f->f_version: 0 */ return f; diff --git a/fs/hppfs/hppfs.c b/fs/hppfs/hppfs.c index 795e2c130ad7..b278f7f52024 100644 --- a/fs/hppfs/hppfs.c +++ b/fs/hppfs/hppfs.c @@ -426,7 +426,7 @@ static int file_mode(int fmode) static int hppfs_open(struct inode *inode, struct file *file) { - const struct cred *cred = current_cred(); + const struct cred *cred = file->f_cred; struct hppfs_private *data; struct vfsmount *proc_mnt; struct dentry *proc_dentry; @@ -490,7 +490,7 @@ static int hppfs_open(struct inode *inode, struct file *file) static int hppfs_dir_open(struct inode *inode, struct file *file) { - const struct cred *cred = current_cred(); + const struct cred *cred = file->f_cred; struct hppfs_private *data; struct vfsmount *proc_mnt; struct dentry *proc_dentry; diff --git a/include/linux/fs.h b/include/linux/fs.h index 3bfec1327b8d..c0fb6d81d89b 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -827,7 +827,7 @@ struct file { fmode_t f_mode; loff_t f_pos; struct fown_struct f_owner; - unsigned int f_uid, f_gid; + const struct cred *f_cred; struct file_ra_state f_ra; u64 f_version; diff --git a/net/ipv4/netfilter/ipt_LOG.c b/net/ipv4/netfilter/ipt_LOG.c index fc6ce04a3e35..7b5dbe118c09 100644 --- a/net/ipv4/netfilter/ipt_LOG.c +++ b/net/ipv4/netfilter/ipt_LOG.c @@ -340,8 +340,8 @@ static void dump_packet(const struct nf_loginfo *info, read_lock_bh(&skb->sk->sk_callback_lock); if (skb->sk->sk_socket && skb->sk->sk_socket->file) printk("UID=%u GID=%u ", - skb->sk->sk_socket->file->f_uid, - skb->sk->sk_socket->file->f_gid); + skb->sk->sk_socket->file->f_cred->fsuid, + skb->sk->sk_socket->file->f_cred->fsgid); read_unlock_bh(&skb->sk->sk_callback_lock); } diff --git a/net/ipv6/netfilter/ip6t_LOG.c b/net/ipv6/netfilter/ip6t_LOG.c index caa441d09567..871d157cec4e 100644 --- a/net/ipv6/netfilter/ip6t_LOG.c +++ b/net/ipv6/netfilter/ip6t_LOG.c @@ -364,8 +364,8 @@ static void dump_packet(const struct nf_loginfo *info, read_lock_bh(&skb->sk->sk_callback_lock); if (skb->sk->sk_socket && skb->sk->sk_socket->file) printk("UID=%u GID=%u ", - skb->sk->sk_socket->file->f_uid, - skb->sk->sk_socket->file->f_gid); + skb->sk->sk_socket->file->f_cred->fsuid, + skb->sk->sk_socket->file->f_cred->fsgid); read_unlock_bh(&skb->sk->sk_callback_lock); } diff --git a/net/netfilter/nfnetlink_log.c b/net/netfilter/nfnetlink_log.c index 41e0105d3828..38f9efd90e8d 100644 --- a/net/netfilter/nfnetlink_log.c +++ b/net/netfilter/nfnetlink_log.c @@ -474,8 +474,9 @@ __build_packet_message(struct nfulnl_instance *inst, if (skb->sk) { read_lock_bh(&skb->sk->sk_callback_lock); if (skb->sk->sk_socket && skb->sk->sk_socket->file) { - __be32 uid = htonl(skb->sk->sk_socket->file->f_uid); - __be32 gid = htonl(skb->sk->sk_socket->file->f_gid); + struct file *file = skb->sk->sk_socket->file; + __be32 uid = htonl(file->f_cred->fsuid); + __be32 gid = htonl(file->f_cred->fsgid); /* need to unlock here since NLA_PUT may goto */ read_unlock_bh(&skb->sk->sk_callback_lock); NLA_PUT_BE32(inst->skb, NFULA_UID, uid); diff --git a/net/netfilter/xt_owner.c b/net/netfilter/xt_owner.c index f19ebd9b78f5..22b2a5e881ea 100644 --- a/net/netfilter/xt_owner.c +++ b/net/netfilter/xt_owner.c @@ -34,12 +34,12 @@ owner_mt_v0(const struct sk_buff *skb, const struct xt_match_param *par) return false; if (info->match & IPT_OWNER_UID) - if ((filp->f_uid != info->uid) ^ + if ((filp->f_cred->fsuid != info->uid) ^ !!(info->invert & IPT_OWNER_UID)) return false; if (info->match & IPT_OWNER_GID) - if ((filp->f_gid != info->gid) ^ + if ((filp->f_cred->fsgid != info->gid) ^ !!(info->invert & IPT_OWNER_GID)) return false; @@ -60,12 +60,12 @@ owner_mt6_v0(const struct sk_buff *skb, const struct xt_match_param *par) return false; if (info->match & IP6T_OWNER_UID) - if ((filp->f_uid != info->uid) ^ + if ((filp->f_cred->fsuid != info->uid) ^ !!(info->invert & IP6T_OWNER_UID)) return false; if (info->match & IP6T_OWNER_GID) - if ((filp->f_gid != info->gid) ^ + if ((filp->f_cred->fsgid != info->gid) ^ !!(info->invert & IP6T_OWNER_GID)) return false; @@ -93,14 +93,14 @@ owner_mt(const struct sk_buff *skb, const struct xt_match_param *par) (XT_OWNER_UID | XT_OWNER_GID)) == 0; if (info->match & XT_OWNER_UID) - if ((filp->f_uid >= info->uid_min && - filp->f_uid <= info->uid_max) ^ + if ((filp->f_cred->fsuid >= info->uid_min && + filp->f_cred->fsuid <= info->uid_max) ^ !(info->invert & XT_OWNER_UID)) return false; if (info->match & XT_OWNER_GID) - if ((filp->f_gid >= info->gid_min && - filp->f_gid <= info->gid_max) ^ + if ((filp->f_cred->fsgid >= info->gid_min && + filp->f_cred->fsgid <= info->gid_max) ^ !(info->invert & XT_OWNER_GID)) return false; diff --git a/net/sched/cls_flow.c b/net/sched/cls_flow.c index 0ebaff637e31..0ef4e3065bcd 100644 --- a/net/sched/cls_flow.c +++ b/net/sched/cls_flow.c @@ -260,14 +260,14 @@ static u32 flow_get_rtclassid(const struct sk_buff *skb) static u32 flow_get_skuid(const struct sk_buff *skb) { if (skb->sk && skb->sk->sk_socket && skb->sk->sk_socket->file) - return skb->sk->sk_socket->file->f_uid; + return skb->sk->sk_socket->file->f_cred->fsuid; return 0; } static u32 flow_get_skgid(const struct sk_buff *skb) { if (skb->sk && skb->sk->sk_socket && skb->sk->sk_socket->file) - return skb->sk->sk_socket->file->f_gid; + return skb->sk->sk_socket->file->f_cred->fsgid; return 0; } -- cgit v1.2.3 From 4d4be482a4d78ca906f45e99fd9fdb91e907f5ad Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Tue, 9 Dec 2008 04:47:33 -0500 Subject: [XFS] add a FMODE flag to make XFS invisible I/O less hacky XFS has a mode called invisble I/O that doesn't update any of the timestamps. It's used for HSM-style applications and exposed through the nasty open by handle ioctl. Instead of doing directly assignment of file operations that set an internal flag for it add a new FMODE_NOCMTIME flag that we can check in the normal file operations. (addition of the generic VFS flag has been ACKed by Al as an interims solution) Signed-off-by: Christoph Hellwig Signed-off-by: Lachlan McIlroy --- fs/xfs/linux-2.6/xfs_file.c | 145 ++++++----------------------------------- fs/xfs/linux-2.6/xfs_ioctl.c | 29 ++++++--- fs/xfs/linux-2.6/xfs_ioctl.h | 8 +-- fs/xfs/linux-2.6/xfs_ioctl32.c | 56 ++++++---------- fs/xfs/linux-2.6/xfs_iops.h | 1 - fs/xfs/xfs_vnodeops.h | 2 - include/linux/fs.h | 8 +++ 7 files changed, 71 insertions(+), 178 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/xfs/linux-2.6/xfs_file.c b/fs/xfs/linux-2.6/xfs_file.c index f999d20a429c..a0c45cc8a6b8 100644 --- a/fs/xfs/linux-2.6/xfs_file.c +++ b/fs/xfs/linux-2.6/xfs_file.c @@ -45,80 +45,44 @@ static struct vm_operations_struct xfs_file_vm_ops; -STATIC_INLINE ssize_t -__xfs_file_read( +STATIC ssize_t +xfs_file_aio_read( struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, - int ioflags, loff_t pos) { struct file *file = iocb->ki_filp; + int ioflags = IO_ISAIO; BUG_ON(iocb->ki_pos != pos); if (unlikely(file->f_flags & O_DIRECT)) ioflags |= IO_ISDIRECT; + if (file->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; return xfs_read(XFS_I(file->f_path.dentry->d_inode), iocb, iov, nr_segs, &iocb->ki_pos, ioflags); } STATIC ssize_t -xfs_file_aio_read( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_read(iocb, iov, nr_segs, IO_ISAIO, pos); -} - -STATIC ssize_t -xfs_file_aio_read_invis( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_read(iocb, iov, nr_segs, IO_ISAIO|IO_INVIS, pos); -} - -STATIC_INLINE ssize_t -__xfs_file_write( +xfs_file_aio_write( struct kiocb *iocb, const struct iovec *iov, unsigned long nr_segs, - int ioflags, loff_t pos) { - struct file *file = iocb->ki_filp; + struct file *file = iocb->ki_filp; + int ioflags = IO_ISAIO; BUG_ON(iocb->ki_pos != pos); if (unlikely(file->f_flags & O_DIRECT)) ioflags |= IO_ISDIRECT; + if (file->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; return xfs_write(XFS_I(file->f_mapping->host), iocb, iov, nr_segs, &iocb->ki_pos, ioflags); } -STATIC ssize_t -xfs_file_aio_write( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_write(iocb, iov, nr_segs, IO_ISAIO, pos); -} - -STATIC ssize_t -xfs_file_aio_write_invis( - struct kiocb *iocb, - const struct iovec *iov, - unsigned long nr_segs, - loff_t pos) -{ - return __xfs_file_write(iocb, iov, nr_segs, IO_ISAIO|IO_INVIS, pos); -} - STATIC ssize_t xfs_file_splice_read( struct file *infilp, @@ -127,20 +91,13 @@ xfs_file_splice_read( size_t len, unsigned int flags) { - return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode), - infilp, ppos, pipe, len, flags, 0); -} + int ioflags = 0; + + if (infilp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; -STATIC ssize_t -xfs_file_splice_read_invis( - struct file *infilp, - loff_t *ppos, - struct pipe_inode_info *pipe, - size_t len, - unsigned int flags) -{ return xfs_splice_read(XFS_I(infilp->f_path.dentry->d_inode), - infilp, ppos, pipe, len, flags, IO_INVIS); + infilp, ppos, pipe, len, flags, ioflags); } STATIC ssize_t @@ -151,20 +108,13 @@ xfs_file_splice_write( size_t len, unsigned int flags) { - return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode), - pipe, outfilp, ppos, len, flags, 0); -} + int ioflags = 0; + + if (outfilp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; -STATIC ssize_t -xfs_file_splice_write_invis( - struct pipe_inode_info *pipe, - struct file *outfilp, - loff_t *ppos, - size_t len, - unsigned int flags) -{ return xfs_splice_write(XFS_I(outfilp->f_path.dentry->d_inode), - pipe, outfilp, ppos, len, flags, IO_INVIS); + pipe, outfilp, ppos, len, flags, ioflags); } STATIC int @@ -275,42 +225,6 @@ xfs_file_mmap( return 0; } -STATIC long -xfs_file_ioctl( - struct file *filp, - unsigned int cmd, - unsigned long p) -{ - struct inode *inode = filp->f_path.dentry->d_inode; - - - /* NOTE: some of the ioctl's return positive #'s as a - * byte count indicating success, such as - * readlink_by_handle. So we don't "sign flip" - * like most other routines. This means true - * errors need to be returned as a negative value. - */ - return xfs_ioctl(XFS_I(inode), filp, 0, cmd, (void __user *)p); -} - -STATIC long -xfs_file_ioctl_invis( - struct file *filp, - unsigned int cmd, - unsigned long p) -{ - struct inode *inode = filp->f_path.dentry->d_inode; - - - /* NOTE: some of the ioctl's return positive #'s as a - * byte count indicating success, such as - * readlink_by_handle. So we don't "sign flip" - * like most other routines. This means true - * errors need to be returned as a negative value. - */ - return xfs_ioctl(XFS_I(inode), filp, IO_INVIS, cmd, (void __user *)p); -} - /* * mmap()d file has taken write protection fault and is being made * writable. We can set the page state up correctly for a writable @@ -346,25 +260,6 @@ const struct file_operations xfs_file_operations = { #endif }; -const struct file_operations xfs_invis_file_operations = { - .llseek = generic_file_llseek, - .read = do_sync_read, - .write = do_sync_write, - .aio_read = xfs_file_aio_read_invis, - .aio_write = xfs_file_aio_write_invis, - .splice_read = xfs_file_splice_read_invis, - .splice_write = xfs_file_splice_write_invis, - .unlocked_ioctl = xfs_file_ioctl_invis, -#ifdef CONFIG_COMPAT - .compat_ioctl = xfs_file_compat_invis_ioctl, -#endif - .mmap = xfs_file_mmap, - .open = xfs_file_open, - .release = xfs_file_release, - .fsync = xfs_file_fsync, -}; - - const struct file_operations xfs_dir_file_operations = { .open = xfs_dir_open, .read = generic_read_dir, diff --git a/fs/xfs/linux-2.6/xfs_ioctl.c b/fs/xfs/linux-2.6/xfs_ioctl.c index c8f1e632ba94..0264c8719ffd 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.c +++ b/fs/xfs/linux-2.6/xfs_ioctl.c @@ -319,10 +319,11 @@ xfs_open_by_handle( put_unused_fd(new_fd); return -XFS_ERROR(-PTR_ERR(filp)); } + if (inode->i_mode & S_IFREG) { /* invisible operation should not change atime */ filp->f_flags |= O_NOATIME; - filp->f_op = &xfs_invis_file_operations; + filp->f_mode |= FMODE_NOCMTIME; } fd_install(new_fd, filp); @@ -1328,21 +1329,31 @@ xfs_ioc_getbmapx( return 0; } -int -xfs_ioctl( - xfs_inode_t *ip, +/* + * Note: some of the ioctl's return positive numbers as a + * byte count indicating success, such as readlink_by_handle. + * So we don't "sign flip" like most other routines. This means + * true errors need to be returned as a negative value. + */ +long +xfs_file_ioctl( struct file *filp, - int ioflags, unsigned int cmd, - void __user *arg) + unsigned long p) { struct inode *inode = filp->f_path.dentry->d_inode; - xfs_mount_t *mp = ip->i_mount; + struct xfs_inode *ip = XFS_I(inode); + struct xfs_mount *mp = ip->i_mount; + void __user *arg = (void __user *)p; + int ioflags = 0; int error; - xfs_itrace_entry(XFS_I(inode)); - switch (cmd) { + if (filp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; + xfs_itrace_entry(ip); + + switch (cmd) { case XFS_IOC_ALLOCSP: case XFS_IOC_FREESP: case XFS_IOC_RESVSP: diff --git a/fs/xfs/linux-2.6/xfs_ioctl.h b/fs/xfs/linux-2.6/xfs_ioctl.h index d92131c827bc..8c16bf2d7e03 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl.h +++ b/fs/xfs/linux-2.6/xfs_ioctl.h @@ -68,13 +68,13 @@ xfs_attrmulti_attr_remove( __uint32_t flags); extern long -xfs_file_compat_ioctl( - struct file *file, +xfs_file_ioctl( + struct file *filp, unsigned int cmd, - unsigned long arg); + unsigned long p); extern long -xfs_file_compat_invis_ioctl( +xfs_file_compat_ioctl( struct file *file, unsigned int cmd, unsigned long arg); diff --git a/fs/xfs/linux-2.6/xfs_ioctl32.c b/fs/xfs/linux-2.6/xfs_ioctl32.c index b34b3d8892a2..0504cece9f66 100644 --- a/fs/xfs/linux-2.6/xfs_ioctl32.c +++ b/fs/xfs/linux-2.6/xfs_ioctl32.c @@ -599,19 +599,24 @@ out: return error; } -STATIC long -xfs_compat_ioctl( - xfs_inode_t *ip, - struct file *filp, - int ioflags, - unsigned cmd, - void __user *arg) +long +xfs_file_compat_ioctl( + struct file *filp, + unsigned cmd, + unsigned long p) { - struct inode *inode = filp->f_path.dentry->d_inode; - xfs_mount_t *mp = ip->i_mount; - int error; + struct inode *inode = filp->f_path.dentry->d_inode; + struct xfs_inode *ip = XFS_I(inode); + struct xfs_mount *mp = ip->i_mount; + void __user *arg = (void __user *)p; + int ioflags = 0; + int error; + + if (filp->f_mode & FMODE_NOCMTIME) + ioflags |= IO_INVIS; + + xfs_itrace_entry(ip); - xfs_itrace_entry(XFS_I(inode)); switch (cmd) { /* No size or alignment issues on any arch */ case XFS_IOC_DIOINFO: @@ -632,7 +637,7 @@ xfs_compat_ioctl( case XFS_IOC_GOINGDOWN: case XFS_IOC_ERROR_INJECTION: case XFS_IOC_ERROR_CLEARALL: - return xfs_ioctl(ip, filp, ioflags, cmd, arg); + return xfs_file_ioctl(filp, cmd, p); #ifndef BROKEN_X86_ALIGNMENT /* These are handled fine if no alignment issues */ case XFS_IOC_ALLOCSP: @@ -646,7 +651,7 @@ xfs_compat_ioctl( case XFS_IOC_FSGEOMETRY_V1: case XFS_IOC_FSGROWFSDATA: case XFS_IOC_FSGROWFSRT: - return xfs_ioctl(ip, filp, ioflags, cmd, arg); + return xfs_file_ioctl(filp, cmd, p); #else case XFS_IOC_ALLOCSP_32: case XFS_IOC_FREESP_32: @@ -687,7 +692,7 @@ xfs_compat_ioctl( case XFS_IOC_SETXFLAGS_32: case XFS_IOC_GETVERSION_32: cmd = _NATIVE_IOC(cmd, long); - return xfs_ioctl(ip, filp, ioflags, cmd, arg); + return xfs_file_ioctl(filp, cmd, p); case XFS_IOC_SWAPEXT: { struct xfs_swapext sxp; struct compat_xfs_swapext __user *sxu = arg; @@ -738,26 +743,3 @@ xfs_compat_ioctl( return -XFS_ERROR(ENOIOCTLCMD); } } - -long -xfs_file_compat_ioctl( - struct file *filp, - unsigned int cmd, - unsigned long p) -{ - struct inode *inode = filp->f_path.dentry->d_inode; - - return xfs_compat_ioctl(XFS_I(inode), filp, 0, cmd, (void __user *)p); -} - -long -xfs_file_compat_invis_ioctl( - struct file *filp, - unsigned int cmd, - unsigned long p) -{ - struct inode *inode = filp->f_path.dentry->d_inode; - - return xfs_compat_ioctl(XFS_I(inode), filp, IO_INVIS, cmd, - (void __user *)p); -} diff --git a/fs/xfs/linux-2.6/xfs_iops.h b/fs/xfs/linux-2.6/xfs_iops.h index 8b1a1e31dc21..ef41c92ce66e 100644 --- a/fs/xfs/linux-2.6/xfs_iops.h +++ b/fs/xfs/linux-2.6/xfs_iops.h @@ -22,7 +22,6 @@ struct xfs_inode; extern const struct file_operations xfs_file_operations; extern const struct file_operations xfs_dir_file_operations; -extern const struct file_operations xfs_invis_file_operations; extern ssize_t xfs_vn_listxattr(struct dentry *, char *data, size_t size); diff --git a/fs/xfs/xfs_vnodeops.h b/fs/xfs/xfs_vnodeops.h index 2a45b00ad32e..55d955ec4ece 100644 --- a/fs/xfs/xfs_vnodeops.h +++ b/fs/xfs/xfs_vnodeops.h @@ -53,8 +53,6 @@ int xfs_attr_set(struct xfs_inode *dp, const char *name, char *value, int xfs_attr_remove(struct xfs_inode *dp, const char *name, int flags); int xfs_attr_list(struct xfs_inode *dp, char *buffer, int bufsize, int flags, struct attrlist_cursor_kern *cursor); -int xfs_ioctl(struct xfs_inode *ip, struct file *filp, - int ioflags, unsigned int cmd, void __user *arg); ssize_t xfs_read(struct xfs_inode *ip, struct kiocb *iocb, const struct iovec *iovp, unsigned int segs, loff_t *offset, int ioflags); diff --git a/include/linux/fs.h b/include/linux/fs.h index 51bd9370d437..965b9ba3865d 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -81,6 +81,14 @@ extern int dir_notify_enable; #define FMODE_WRITE_IOCTL ((__force fmode_t)128) #define FMODE_NDELAY_NOW ((__force fmode_t)256) +/* + * Don't update ctime and mtime. + * + * Currently a special hack for the XFS open_by_handle ioctl, but we'll + * hopefully graduate it to a proper O_CMTIME flag supported by open(2) soon. + */ +#define FMODE_NOCMTIME ((__force fmode_t)2048) + #define RW_MASK 1 #define RWA_MASK 2 #define READ 0 -- cgit v1.2.3 From dded4f4d5048e64a01cf52eed4d27c8cb2600525 Mon Sep 17 00:00:00 2001 From: Jan Engelhardt Date: Mon, 1 Dec 2008 14:34:50 -0800 Subject: include: linux/fs.h: put declarations in __KERNEL__ include/linux/fs.h contains externs for a bunch of variables. That obviously belongs under ifdef __KERNEL__. Signed-off-by: Jan Engelhardt Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Al Viro --- include/linux/fs.h | 18 +++++++++--------- 1 file changed, 9 insertions(+), 9 deletions(-) (limited to 'include/linux/fs.h') diff --git a/include/linux/fs.h b/include/linux/fs.h index 001ded4845b4..c5e4c5c74034 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -21,7 +21,6 @@ /* Fixed constants first: */ #undef NR_OPEN -extern int sysctl_nr_open; #define INR_OPEN 1024 /* Initial setting for nfile rlimits */ #define BLOCK_SIZE_BITS 10 @@ -38,21 +37,13 @@ struct files_stat_struct { int nr_free_files; /* read only */ int max_files; /* tunable */ }; -extern struct files_stat_struct files_stat; -extern int get_max_files(void); struct inodes_stat_t { int nr_inodes; int nr_unused; int dummy[5]; /* padding for sysctl ABI compatibility */ }; -extern struct inodes_stat_t inodes_stat; -extern int leases_enable, lease_break_time; - -#ifdef CONFIG_DNOTIFY -extern int dir_notify_enable; -#endif #define NR_FILE 8192 /* this can well be larger on a larger system */ @@ -330,6 +321,15 @@ extern void __init inode_init(void); extern void __init inode_init_early(void); extern void __init files_init(unsigned long); +extern struct files_stat_struct files_stat; +extern int get_max_files(void); +extern int sysctl_nr_open; +extern struct inodes_stat_t inodes_stat; +extern int leases_enable, lease_break_time; +#ifdef CONFIG_DNOTIFY +extern int dir_notify_enable; +#endif + struct buffer_head; typedef int (get_block_t)(struct inode *inode, sector_t iblock, struct buffer_head *bh_result, int create); -- cgit v1.2.3 From 3fb64190aa3c23c10e6e9fd0124ac030115c99bf Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 24 Oct 2008 09:58:10 +0200 Subject: pass a struct path * to may_open No need for the nameidata in may_open - a struct path is enough. Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- fs/namei.c | 14 +++++++------- fs/nfsctl.c | 5 +++-- include/linux/fs.h | 2 +- 3 files changed, 11 insertions(+), 10 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/namei.c b/fs/namei.c index d4d0b59ed2cc..5cc0dc95a7a5 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -1487,9 +1487,9 @@ int vfs_create(struct inode *dir, struct dentry *dentry, int mode, return error; } -int may_open(struct nameidata *nd, int acc_mode, int flag) +int may_open(struct path *path, int acc_mode, int flag) { - struct dentry *dentry = nd->path.dentry; + struct dentry *dentry = path->dentry; struct inode *inode = dentry->d_inode; int error; @@ -1510,13 +1510,13 @@ int may_open(struct nameidata *nd, int acc_mode, int flag) if (S_ISFIFO(inode->i_mode) || S_ISSOCK(inode->i_mode)) { flag &= ~O_TRUNC; } else if (S_ISBLK(inode->i_mode) || S_ISCHR(inode->i_mode)) { - if (nd->path.mnt->mnt_flags & MNT_NODEV) + if (path->mnt->mnt_flags & MNT_NODEV) return -EACCES; flag &= ~O_TRUNC; } - error = vfs_permission(nd, acc_mode); + error = inode_permission(inode, acc_mode); if (error) return error; /* @@ -1551,7 +1551,7 @@ int may_open(struct nameidata *nd, int acc_mode, int flag) */ error = locks_verify_locked(inode); if (!error) - error = security_path_truncate(&nd->path, 0, + error = security_path_truncate(path, 0, ATTR_MTIME|ATTR_CTIME|ATTR_OPEN); if (!error) { DQUOT_INIT(inode); @@ -1594,7 +1594,7 @@ out_unlock: if (error) return error; /* Don't check for write permission, don't truncate */ - return may_open(nd, 0, flag & ~O_TRUNC); + return may_open(&nd->path, 0, flag & ~O_TRUNC); } /* @@ -1780,7 +1780,7 @@ ok: if (error) goto exit; } - error = may_open(&nd, acc_mode, flag); + error = may_open(&nd.path, acc_mode, flag); if (error) { if (will_write) mnt_drop_write(nd.path.mnt); diff --git a/fs/nfsctl.c b/fs/nfsctl.c index b1acbd6ab6fb..b27451909dff 100644 --- a/fs/nfsctl.c +++ b/fs/nfsctl.c @@ -38,9 +38,10 @@ static struct file *do_open(char *name, int flags) return ERR_PTR(error); if (flags == O_RDWR) - error = may_open(&nd,MAY_READ|MAY_WRITE,FMODE_READ|FMODE_WRITE); + error = may_open(&nd.path, MAY_READ|MAY_WRITE, + FMODE_READ|FMODE_WRITE); else - error = may_open(&nd, MAY_WRITE, FMODE_WRITE); + error = may_open(&nd.path, MAY_WRITE, FMODE_WRITE); if (!error) return dentry_open(nd.path.dentry, nd.path.mnt, flags, diff --git a/include/linux/fs.h b/include/linux/fs.h index c5e4c5c74034..3468df5a06e0 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1869,7 +1869,7 @@ extern void free_write_pipe(struct file *); extern struct file *do_filp_open(int dfd, const char *pathname, int open_flag, int mode); -extern int may_open(struct nameidata *, int, int); +extern int may_open(struct path *, int, int); extern int kernel_read(struct file *, unsigned long, char *, unsigned long); extern struct file * open_exec(const char *); -- cgit v1.2.3 From cb23beb55100171646e69e248fb45f10db6e99a4 Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Fri, 24 Oct 2008 09:59:29 +0200 Subject: kill vfs_permission With all the nameidata removal there's no point anymore for this helper. Of the three callers left two will go away with the next lookup series anyway. Also add proper kerneldoc to inode_permission as this is the main permission check routine now. Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- fs/exec.c | 5 +++-- fs/namei.c | 31 +++++++++++++------------------ include/linux/fs.h | 1 - 3 files changed, 16 insertions(+), 21 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/exec.c b/fs/exec.c index 02d2e120542d..dfbf7009fbe7 100644 --- a/fs/exec.c +++ b/fs/exec.c @@ -127,7 +127,8 @@ asmlinkage long sys_uselib(const char __user * library) if (nd.path.mnt->mnt_flags & MNT_NOEXEC) goto exit; - error = vfs_permission(&nd, MAY_READ | MAY_EXEC | MAY_OPEN); + error = inode_permission(nd.path.dentry->d_inode, + MAY_READ | MAY_EXEC | MAY_OPEN); if (error) goto exit; @@ -680,7 +681,7 @@ struct file *open_exec(const char *name) if (nd.path.mnt->mnt_flags & MNT_NOEXEC) goto out_path_put; - err = vfs_permission(&nd, MAY_EXEC | MAY_OPEN); + err = inode_permission(nd.path.dentry->d_inode, MAY_EXEC | MAY_OPEN); if (err) goto out_path_put; diff --git a/fs/namei.c b/fs/namei.c index 5cc0dc95a7a5..3f88e043d459 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -226,6 +226,16 @@ int generic_permission(struct inode *inode, int mask, return -EACCES; } +/** + * inode_permission - check for access rights to a given inode + * @inode: inode to check permission on + * @mask: right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC) + * + * Used to check for read/write/execute permissions on an inode. + * We use "fsuid" for this, letting us set arbitrary permissions + * for filesystem access without changing the "normal" uids which + * are used for other things. + */ int inode_permission(struct inode *inode, int mask) { int retval; @@ -263,21 +273,6 @@ int inode_permission(struct inode *inode, int mask) mask & (MAY_READ|MAY_WRITE|MAY_EXEC|MAY_APPEND)); } -/** - * vfs_permission - check for access rights to a given path - * @nd: lookup result that describes the path - * @mask: right to check for (%MAY_READ, %MAY_WRITE, %MAY_EXEC) - * - * Used to check for read/write/execute permissions on a path. - * We use "fsuid" for this, letting us set arbitrary permissions - * for filesystem access without changing the "normal" uids which - * are used for other things. - */ -int vfs_permission(struct nameidata *nd, int mask) -{ - return inode_permission(nd->path.dentry->d_inode, mask); -} - /** * file_permission - check for additional access rights to a given file * @file: file to check access rights for @@ -288,7 +283,7 @@ int vfs_permission(struct nameidata *nd, int mask) * * Note: * Do not use this function in new code. All access checks should - * be done using vfs_permission(). + * be done using inode_permission(). */ int file_permission(struct file *file, int mask) { @@ -853,7 +848,8 @@ static int __link_path_walk(const char *name, struct nameidata *nd) nd->flags |= LOOKUP_CONTINUE; err = exec_permission_lite(inode); if (err == -EAGAIN) - err = vfs_permission(nd, MAY_EXEC); + err = inode_permission(nd->path.dentry->d_inode, + MAY_EXEC); if (err) break; @@ -2882,7 +2878,6 @@ EXPORT_SYMBOL(path_lookup); EXPORT_SYMBOL(kern_path); EXPORT_SYMBOL(vfs_path_lookup); EXPORT_SYMBOL(inode_permission); -EXPORT_SYMBOL(vfs_permission); EXPORT_SYMBOL(file_permission); EXPORT_SYMBOL(unlock_rename); EXPORT_SYMBOL(vfs_create); diff --git a/include/linux/fs.h b/include/linux/fs.h index 3468df5a06e0..fd615986a41c 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1212,7 +1212,6 @@ extern void unlock_super(struct super_block *); /* * VFS helper functions.. */ -extern int vfs_permission(struct nameidata *, int); extern int vfs_create(struct inode *, struct dentry *, int, struct nameidata *); extern int vfs_mkdir(struct inode *, struct dentry *, int); extern int vfs_mknod(struct inode *, struct dentry *, int, dev_t); -- cgit v1.2.3 From 6badd79bd002788aaec27b50a74ab69ef65ab8ee Mon Sep 17 00:00:00 2001 From: Al Viro Date: Fri, 26 Dec 2008 00:57:40 -0500 Subject: kill ->dir_notify() Remove the hopelessly misguided ->dir_notify(). The only instance (cifs) has been broken by design from the very beginning; the objects it creates are never destroyed, keep references to struct file they can outlive, nothing that could possibly evict them exists on close(2) path *and* no locking whatsoever is done to prevent races with close(), should the previous, er, deficiencies someday be dealt with. Signed-off-by: Al Viro --- Documentation/filesystems/Locking | 2 - Documentation/filesystems/vfs.txt | 3 - fs/bad_inode.c | 6 -- fs/cifs/Makefile | 2 +- fs/cifs/cifsfs.c | 7 --- fs/cifs/cifsfs.h | 1 - fs/cifs/fcntl.c | 118 -------------------------------------- fs/dnotify.c | 3 - include/linux/fs.h | 1 - 9 files changed, 1 insertion(+), 142 deletions(-) delete mode 100644 fs/cifs/fcntl.c (limited to 'include/linux/fs.h') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index 23d2f4460deb..ccec55394380 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -394,7 +394,6 @@ prototypes: unsigned long (*get_unmapped_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); int (*check_flags)(int); - int (*dir_notify)(struct file *, unsigned long); }; locking rules: @@ -424,7 +423,6 @@ sendfile: no sendpage: no get_unmapped_area: no check_flags: no -dir_notify: no ->llseek() locking has moved from llseek to the individual llseek implementations. If your fs is not using generic_file_llseek, you diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt index 041cb771d505..ef19afa186a9 100644 --- a/Documentation/filesystems/vfs.txt +++ b/Documentation/filesystems/vfs.txt @@ -733,7 +733,6 @@ struct file_operations { ssize_t (*sendpage) (struct file *, struct page *, int, size_t, loff_t *, int); unsigned long (*get_unmapped_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); int (*check_flags)(int); - int (*dir_notify)(struct file *filp, unsigned long arg); int (*flock) (struct file *, int, struct file_lock *); ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, size_t, unsigned int); ssize_t (*splice_read)(struct file *, struct pipe_inode_info *, size_t, unsigned int); @@ -800,8 +799,6 @@ otherwise noted. check_flags: called by the fcntl(2) system call for F_SETFL command - dir_notify: called by the fcntl(2) system call for F_NOTIFY command - flock: called by the flock(2) system call splice_write: called by the VFS to splice data from a pipe to a file. This diff --git a/fs/bad_inode.c b/fs/bad_inode.c index 5f1538c03b1b..a05287a23f62 100644 --- a/fs/bad_inode.c +++ b/fs/bad_inode.c @@ -132,11 +132,6 @@ static int bad_file_check_flags(int flags) return -EIO; } -static int bad_file_dir_notify(struct file *file, unsigned long arg) -{ - return -EIO; -} - static int bad_file_flock(struct file *filp, int cmd, struct file_lock *fl) { return -EIO; @@ -179,7 +174,6 @@ static const struct file_operations bad_file_ops = .sendpage = bad_file_sendpage, .get_unmapped_area = bad_file_get_unmapped_area, .check_flags = bad_file_check_flags, - .dir_notify = bad_file_dir_notify, .flock = bad_file_flock, .splice_write = bad_file_splice_write, .splice_read = bad_file_splice_read, diff --git a/fs/cifs/Makefile b/fs/cifs/Makefile index 6ba43fb346fb..9948c0030e86 100644 --- a/fs/cifs/Makefile +++ b/fs/cifs/Makefile @@ -5,7 +5,7 @@ obj-$(CONFIG_CIFS) += cifs.o cifs-y := cifsfs.o cifssmb.o cifs_debug.o connect.o dir.o file.o inode.o \ link.o misc.o netmisc.o smbdes.o smbencrypt.o transport.o asn1.o \ - md4.o md5.o cifs_unicode.o nterr.o xattr.o cifsencrypt.o fcntl.o \ + md4.o md5.o cifs_unicode.o nterr.o xattr.o cifsencrypt.o \ readdir.o ioctl.o sess.o export.o cifsacl.o cifs-$(CONFIG_CIFS_UPCALL) += cifs_spnego.o diff --git a/fs/cifs/cifsfs.c b/fs/cifs/cifsfs.c index 0005a194a75c..13ea53251dcf 100644 --- a/fs/cifs/cifsfs.c +++ b/fs/cifs/cifsfs.c @@ -747,7 +747,6 @@ const struct file_operations cifs_file_ops = { #endif /* CONFIG_CIFS_POSIX */ #ifdef CONFIG_CIFS_EXPERIMENTAL - .dir_notify = cifs_dir_notify, .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; @@ -768,7 +767,6 @@ const struct file_operations cifs_file_direct_ops = { #endif /* CONFIG_CIFS_POSIX */ .llseek = cifs_llseek, #ifdef CONFIG_CIFS_EXPERIMENTAL - .dir_notify = cifs_dir_notify, .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; @@ -789,7 +787,6 @@ const struct file_operations cifs_file_nobrl_ops = { #endif /* CONFIG_CIFS_POSIX */ #ifdef CONFIG_CIFS_EXPERIMENTAL - .dir_notify = cifs_dir_notify, .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; @@ -809,7 +806,6 @@ const struct file_operations cifs_file_direct_nobrl_ops = { #endif /* CONFIG_CIFS_POSIX */ .llseek = cifs_llseek, #ifdef CONFIG_CIFS_EXPERIMENTAL - .dir_notify = cifs_dir_notify, .setlease = cifs_setlease, #endif /* CONFIG_CIFS_EXPERIMENTAL */ }; @@ -818,9 +814,6 @@ const struct file_operations cifs_dir_ops = { .readdir = cifs_readdir, .release = cifs_closedir, .read = generic_read_dir, -#ifdef CONFIG_CIFS_EXPERIMENTAL - .dir_notify = cifs_dir_notify, -#endif /* CONFIG_CIFS_EXPERIMENTAL */ .unlocked_ioctl = cifs_ioctl, .llseek = generic_file_llseek, }; diff --git a/fs/cifs/cifsfs.h b/fs/cifs/cifsfs.h index 2ce04c73d74e..7ac481841f87 100644 --- a/fs/cifs/cifsfs.h +++ b/fs/cifs/cifsfs.h @@ -76,7 +76,6 @@ extern int cifs_file_mmap(struct file * , struct vm_area_struct *); extern const struct file_operations cifs_dir_ops; extern int cifs_dir_open(struct inode *inode, struct file *file); extern int cifs_readdir(struct file *file, void *direntry, filldir_t filldir); -extern int cifs_dir_notify(struct file *, unsigned long arg); /* Functions related to dir entries */ extern struct dentry_operations cifs_dentry_ops; diff --git a/fs/cifs/fcntl.c b/fs/cifs/fcntl.c deleted file mode 100644 index 5a57581eb4b2..000000000000 --- a/fs/cifs/fcntl.c +++ /dev/null @@ -1,118 +0,0 @@ -/* - * fs/cifs/fcntl.c - * - * vfs operations that deal with the file control API - * - * Copyright (C) International Business Machines Corp., 2003,2004 - * Author(s): Steve French (sfrench@us.ibm.com) - * - * This library is free software; you can redistribute it and/or modify - * it under the terms of the GNU Lesser General Public License as published - * by the Free Software Foundation; either version 2.1 of the License, or - * (at your option) any later version. - * - * This library 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 Lesser General Public License for more details. - * - * You should have received a copy of the GNU Lesser General Public License - * along with this library; if not, write to the Free Software - * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA - */ -#include -#include -#include -#include "cifsglob.h" -#include "cifsproto.h" -#include "cifs_unicode.h" -#include "cifs_debug.h" -#include "cifsfs.h" - -static __u32 convert_to_cifs_notify_flags(unsigned long fcntl_notify_flags) -{ - __u32 cifs_ntfy_flags = 0; - - /* No way on Linux VFS to ask to monitor xattr - changes (and no stream support either */ - if (fcntl_notify_flags & DN_ACCESS) - cifs_ntfy_flags |= FILE_NOTIFY_CHANGE_LAST_ACCESS; - if (fcntl_notify_flags & DN_MODIFY) { - /* What does this mean on directories? */ - cifs_ntfy_flags |= FILE_NOTIFY_CHANGE_LAST_WRITE | - FILE_NOTIFY_CHANGE_SIZE; - } - if (fcntl_notify_flags & DN_CREATE) { - cifs_ntfy_flags |= FILE_NOTIFY_CHANGE_CREATION | - FILE_NOTIFY_CHANGE_LAST_WRITE; - } - if (fcntl_notify_flags & DN_DELETE) - cifs_ntfy_flags |= FILE_NOTIFY_CHANGE_LAST_WRITE; - if (fcntl_notify_flags & DN_RENAME) { - /* BB review this - checking various server behaviors */ - cifs_ntfy_flags |= FILE_NOTIFY_CHANGE_DIR_NAME | - FILE_NOTIFY_CHANGE_FILE_NAME; - } - if (fcntl_notify_flags & DN_ATTRIB) { - cifs_ntfy_flags |= FILE_NOTIFY_CHANGE_SECURITY | - FILE_NOTIFY_CHANGE_ATTRIBUTES; - } -/* if (fcntl_notify_flags & DN_MULTISHOT) { - cifs_ntfy_flags |= ; - } */ /* BB fixme - not sure how to handle this with CIFS yet */ - - return cifs_ntfy_flags; -} - -int cifs_dir_notify(struct file *file, unsigned long arg) -{ - int xid; - int rc = -EINVAL; - int oplock = 0; - struct cifs_sb_info *cifs_sb; - struct cifsTconInfo *pTcon; - char *full_path = NULL; - __u32 filter = FILE_NOTIFY_CHANGE_NAME | FILE_NOTIFY_CHANGE_ATTRIBUTES; - __u16 netfid; - - if (experimEnabled == 0) - return 0; - - xid = GetXid(); - cifs_sb = CIFS_SB(file->f_path.dentry->d_sb); - pTcon = cifs_sb->tcon; - - full_path = build_path_from_dentry(file->f_path.dentry); - - if (full_path == NULL) { - rc = -ENOMEM; - } else { - cFYI(1, ("dir notify on file %s Arg 0x%lx", full_path, arg)); - rc = CIFSSMBOpen(xid, pTcon, full_path, FILE_OPEN, - GENERIC_READ | SYNCHRONIZE, 0 /* create options */, - &netfid, &oplock, NULL, cifs_sb->local_nls, - cifs_sb->mnt_cifs_flags & CIFS_MOUNT_MAP_SPECIAL_CHR); - /* BB fixme - add this handle to a notify handle list */ - if (rc) { - cFYI(1, ("Could not open directory for notify")); - } else { - filter = convert_to_cifs_notify_flags(arg); - if (filter != 0) { - rc = CIFSSMBNotify(xid, pTcon, - 0 /* no subdirs */, netfid, - filter, file, arg & DN_MULTISHOT, - cifs_sb->local_nls); - } else { - rc = -EINVAL; - } - /* BB add code to close file eventually (at unmount - it would close automatically but may be a way - to do it easily when inode freed or when - notify info is cleared/changed */ - cFYI(1, ("notify rc %d", rc)); - } - } - - FreeXid(xid); - return rc; -} diff --git a/fs/dnotify.c b/fs/dnotify.c index 676073b8dda5..b0aa2cde80bd 100644 --- a/fs/dnotify.c +++ b/fs/dnotify.c @@ -115,9 +115,6 @@ int fcntl_dirnotify(int fd, struct file *filp, unsigned long arg) dn->dn_next = inode->i_dnotify; inode->i_dnotify = dn; spin_unlock(&inode->i_lock); - - if (filp->f_op && filp->f_op->dir_notify) - return filp->f_op->dir_notify(filp, arg); return 0; out_free: diff --git a/include/linux/fs.h b/include/linux/fs.h index fd615986a41c..be16ce01fb1b 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1309,7 +1309,6 @@ struct file_operations { ssize_t (*sendpage) (struct file *, struct page *, int, size_t, loff_t *, int); unsigned long (*get_unmapped_area)(struct file *, unsigned long, unsigned long, unsigned long, unsigned long); int (*check_flags)(int); - int (*dir_notify)(struct file *filp, unsigned long arg); int (*flock) (struct file *, int, struct file_lock *); ssize_t (*splice_write)(struct pipe_inode_info *, struct file *, loff_t *, size_t, unsigned int); ssize_t (*splice_read)(struct file *, loff_t *, struct pipe_inode_info *, size_t, unsigned int); -- cgit v1.2.3 From 261bca86ed4f7f391d1938167624e78da61dcc6b Mon Sep 17 00:00:00 2001 From: Al Viro Date: Tue, 30 Dec 2008 01:48:21 -0500 Subject: nfsd/create race fixes, infrastructure new helpers - insert_inode_locked() and insert_inode_locked4(). Hash new inode, making sure that there's no such inode in icache already. If there is and it does not end up unhashed (as would happen if we have nfsd trying to resolve a bogus fhandle), fail. Otherwise insert our inode into hash and succeed. In either case have i_state set to new+locked; cleanup ends up being simpler with such calling conventions. Signed-off-by: Al Viro --- fs/inode.c | 59 ++++++++++++++++++++++++++++++++++++++++++++++++++++++ include/linux/fs.h | 2 ++ 2 files changed, 61 insertions(+) (limited to 'include/linux/fs.h') diff --git a/fs/inode.c b/fs/inode.c index 098a2443196f..7de1cda92489 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -1032,6 +1032,65 @@ struct inode *iget_locked(struct super_block *sb, unsigned long ino) EXPORT_SYMBOL(iget_locked); +int insert_inode_locked(struct inode *inode) +{ + struct super_block *sb = inode->i_sb; + ino_t ino = inode->i_ino; + struct hlist_head *head = inode_hashtable + hash(sb, ino); + struct inode *old; + + inode->i_state |= I_LOCK|I_NEW; + while (1) { + spin_lock(&inode_lock); + old = find_inode_fast(sb, head, ino); + if (likely(!old)) { + hlist_add_head(&inode->i_hash, head); + spin_unlock(&inode_lock); + return 0; + } + __iget(old); + spin_unlock(&inode_lock); + wait_on_inode(old); + if (unlikely(!hlist_unhashed(&old->i_hash))) { + iput(old); + return -EBUSY; + } + iput(old); + } +} + +EXPORT_SYMBOL(insert_inode_locked); + +int insert_inode_locked4(struct inode *inode, unsigned long hashval, + int (*test)(struct inode *, void *), void *data) +{ + struct super_block *sb = inode->i_sb; + struct hlist_head *head = inode_hashtable + hash(sb, hashval); + struct inode *old; + + inode->i_state |= I_LOCK|I_NEW; + + while (1) { + spin_lock(&inode_lock); + old = find_inode(sb, head, test, data); + if (likely(!old)) { + hlist_add_head(&inode->i_hash, head); + spin_unlock(&inode_lock); + return 0; + } + __iget(old); + spin_unlock(&inode_lock); + wait_on_inode(old); + if (unlikely(!hlist_unhashed(&old->i_hash))) { + iput(old); + return -EBUSY; + } + iput(old); + } +} + +EXPORT_SYMBOL(insert_inode_locked4); + /** * __insert_inode_hash - hash an inode * @inode: unhashed inode diff --git a/include/linux/fs.h b/include/linux/fs.h index be16ce01fb1b..e2170ee21e18 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1902,6 +1902,8 @@ extern struct inode *ilookup(struct super_block *sb, unsigned long ino); extern struct inode * iget5_locked(struct super_block *, unsigned long, int (*test)(struct inode *, void *), int (*set)(struct inode *, void *), void *); extern struct inode * iget_locked(struct super_block *, unsigned long); +extern int insert_inode_locked4(struct inode *, unsigned long, int (*test)(struct inode *, void *), void *); +extern int insert_inode_locked(struct inode *); extern void unlock_new_inode(struct inode *); extern void __iget(struct inode * inode); -- cgit v1.2.3 From 87d8fe1ee6b8d2f95076142d58c440dba4e7bdc2 Mon Sep 17 00:00:00 2001 From: Theodore Ts'o Date: Sat, 3 Jan 2009 09:47:09 -0500 Subject: add releasepage hooks to block devices which can be used by file systems Implement blkdev_releasepage() to release the buffer_heads and pages after we release private data belonging to a mounted filesystem. Cc: Toshiyuki Okajima Cc: linux-fsdevel@vger.kernel.org Signed-off-by: "Theodore Ts'o" --- fs/block_dev.c | 15 +++++++++++++++ fs/super.c | 2 ++ include/linux/fs.h | 2 ++ 3 files changed, 19 insertions(+) (limited to 'include/linux/fs.h') diff --git a/fs/block_dev.c b/fs/block_dev.c index 349a26c10001..1dd07e66e98a 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -1220,6 +1220,20 @@ static long block_ioctl(struct file *file, unsigned cmd, unsigned long arg) return blkdev_ioctl(bdev, mode, cmd, arg); } +/* + * Try to release a page associated with block device when the system + * is under memory pressure. + */ +static int blkdev_releasepage(struct page *page, gfp_t wait) +{ + struct super_block *super = BDEV_I(page->mapping->host)->bdev.bd_super; + + if (super && super->s_op->bdev_try_to_free_page) + return super->s_op->bdev_try_to_free_page(super, page, wait); + + return try_to_free_buffers(page); +} + static const struct address_space_operations def_blk_aops = { .readpage = blkdev_readpage, .writepage = blkdev_writepage, @@ -1227,6 +1241,7 @@ static const struct address_space_operations def_blk_aops = { .write_begin = blkdev_write_begin, .write_end = blkdev_write_end, .writepages = generic_writepages, + .releasepage = blkdev_releasepage, .direct_IO = blkdev_direct_IO, }; diff --git a/fs/super.c b/fs/super.c index ddba069d7a99..d5fd4498548a 100644 --- a/fs/super.c +++ b/fs/super.c @@ -800,6 +800,7 @@ int get_sb_bdev(struct file_system_type *fs_type, } s->s_flags |= MS_ACTIVE; + bdev->bd_super = s; } return simple_set_mnt(mnt, s); @@ -819,6 +820,7 @@ void kill_block_super(struct super_block *sb) struct block_device *bdev = sb->s_bdev; fmode_t mode = sb->s_mode; + bdev->bd_super = 0; generic_shutdown_super(sb); sync_blockdev(bdev); close_bdev_exclusive(bdev, mode); diff --git a/include/linux/fs.h b/include/linux/fs.h index f2a3010140e3..0f54ae0f0ccd 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -565,6 +565,7 @@ struct address_space { struct block_device { dev_t bd_dev; /* not a kdev_t - it's a search key */ struct inode * bd_inode; /* will die */ + struct super_block * bd_super; int bd_openers; struct mutex bd_mutex; /* open/close mutex */ struct semaphore bd_mount_sem; @@ -1385,6 +1386,7 @@ struct super_operations { ssize_t (*quota_read)(struct super_block *, int, char *, size_t, loff_t); ssize_t (*quota_write)(struct super_block *, int, const char *, size_t, loff_t); #endif + int (*bdev_try_to_free_page)(struct super_block*, struct page*, gfp_t); }; /* -- cgit v1.2.3 From 54566b2c1594c2326a645a3551f9d989f7ba3c5e Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Sun, 4 Jan 2009 12:00:53 -0800 Subject: fs: symlink write_begin allocation context fix With the write_begin/write_end aops, page_symlink was broken because it could no longer pass a GFP_NOFS type mask into the point where the allocations happened. They are done in write_begin, which would always assume that the filesystem can be entered from reclaim. This bug could cause filesystem deadlocks. The funny thing with having a gfp_t mask there is that it doesn't really allow the caller to arbitrarily tinker with the context in which it can be called. It couldn't ever be GFP_ATOMIC, for example, because it needs to take the page lock. The only thing any callers care about is __GFP_FS anyway, so turn that into a single flag. Add a new flag for write_begin, AOP_FLAG_NOFS. Filesystems can now act on this flag in their write_begin function. Change __grab_cache_page to accept a nofs argument as well, to honour that flag (while we're there, change the name to grab_cache_page_write_begin which is more instructive and does away with random leading underscores). This is really a more flexible way to go in the end anyway -- if a filesystem happens to want any extra allocations aside from the pagecache ones in ints write_begin function, it may now use GFP_KERNEL (rather than GFP_NOFS) for common case allocations (eg. ocfs2_alloc_write_ctxt, for a random example). [kosaki.motohiro@jp.fujitsu.com: fix ubifs] [kosaki.motohiro@jp.fujitsu.com: fix fuse] Signed-off-by: Nick Piggin Reviewed-by: KOSAKI Motohiro Cc: [2.6.28.x] Signed-off-by: KOSAKI Motohiro Signed-off-by: Andrew Morton [ Cleaned up the calling convention: just pass in the AOP flags untouched to the grab_cache_page_write_begin() function. That just simplifies everybody, and may even allow future expansion of the logic. - Linus ] Signed-off-by: Linus Torvalds --- fs/affs/file.c | 2 +- fs/afs/write.c | 2 +- fs/buffer.c | 4 ++-- fs/cifs/file.c | 2 +- fs/ecryptfs/mmap.c | 2 +- fs/ext3/inode.c | 2 +- fs/ext3/namei.c | 3 +-- fs/ext4/inode.c | 4 ++-- fs/ext4/namei.c | 3 +-- fs/fuse/file.c | 4 ++-- fs/gfs2/ops_address.c | 2 +- fs/hostfs/hostfs_kern.c | 2 +- fs/jffs2/file.c | 2 +- fs/libfs.c | 2 +- fs/namei.c | 13 +++++++++---- fs/nfs/file.c | 2 +- fs/reiserfs/inode.c | 2 +- fs/smbfs/file.c | 2 +- fs/ubifs/file.c | 9 +++++---- include/linux/fs.h | 5 ++++- include/linux/pagemap.h | 3 ++- mm/filemap.c | 13 +++++++++---- 22 files changed, 49 insertions(+), 36 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/affs/file.c b/fs/affs/file.c index 1377b1240b6e..9246cb4aa018 100644 --- a/fs/affs/file.c +++ b/fs/affs/file.c @@ -628,7 +628,7 @@ static int affs_write_begin_ofs(struct file *file, struct address_space *mapping } index = pos >> PAGE_CACHE_SHIFT; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; *pagep = page; diff --git a/fs/afs/write.c b/fs/afs/write.c index d6b85dab35fc..3fb36d433621 100644 --- a/fs/afs/write.c +++ b/fs/afs/write.c @@ -144,7 +144,7 @@ int afs_write_begin(struct file *file, struct address_space *mapping, candidate->state = AFS_WBACK_PENDING; init_waitqueue_head(&candidate->waitq); - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) { kfree(candidate); return -ENOMEM; diff --git a/fs/buffer.c b/fs/buffer.c index 776ae091d3b0..a13f09b696f7 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -1996,7 +1996,7 @@ int block_write_begin(struct file *file, struct address_space *mapping, page = *pagep; if (page == NULL) { ownpage = 1; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) { status = -ENOMEM; goto out; @@ -2502,7 +2502,7 @@ int nobh_write_begin(struct file *file, struct address_space *mapping, from = pos & (PAGE_CACHE_SIZE - 1); to = from + len; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; *pagep = page; diff --git a/fs/cifs/file.c b/fs/cifs/file.c index b1e1fc6a6e6a..12bb656fbe75 100644 --- a/fs/cifs/file.c +++ b/fs/cifs/file.c @@ -2074,7 +2074,7 @@ static int cifs_write_begin(struct file *file, struct address_space *mapping, cFYI(1, ("write_begin from %lld len %d", (long long)pos, len)); - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) { rc = -ENOMEM; goto out; diff --git a/fs/ecryptfs/mmap.c b/fs/ecryptfs/mmap.c index 04d7b3fa1ac6..46cec2b69796 100644 --- a/fs/ecryptfs/mmap.c +++ b/fs/ecryptfs/mmap.c @@ -288,7 +288,7 @@ static int ecryptfs_write_begin(struct file *file, loff_t prev_page_end_size; int rc = 0; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; *pagep = page; diff --git a/fs/ext3/inode.c b/fs/ext3/inode.c index c4bdccf976b5..5fa453b49a64 100644 --- a/fs/ext3/inode.c +++ b/fs/ext3/inode.c @@ -1161,7 +1161,7 @@ static int ext3_write_begin(struct file *file, struct address_space *mapping, to = from + len; retry: - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; *pagep = page; diff --git a/fs/ext3/namei.c b/fs/ext3/namei.c index 297ea8dfac7c..1dd2abe6313e 100644 --- a/fs/ext3/namei.c +++ b/fs/ext3/namei.c @@ -2175,8 +2175,7 @@ retry: * We have a transaction open. All is sweetness. It also sets * i_size in generic_commit_write(). */ - err = __page_symlink(inode, symname, l, - mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS); + err = __page_symlink(inode, symname, l, 1); if (err) { drop_nlink(inode); unlock_new_inode(inode); diff --git a/fs/ext4/inode.c b/fs/ext4/inode.c index 7c3325e0b005..6702a49992a6 100644 --- a/fs/ext4/inode.c +++ b/fs/ext4/inode.c @@ -1346,7 +1346,7 @@ retry: goto out; } - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) { ext4_journal_stop(handle); ret = -ENOMEM; @@ -2550,7 +2550,7 @@ retry: goto out; } - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) { ext4_journal_stop(handle); ret = -ENOMEM; diff --git a/fs/ext4/namei.c b/fs/ext4/namei.c index da98a9012fa5..9fd2a5e1be4d 100644 --- a/fs/ext4/namei.c +++ b/fs/ext4/namei.c @@ -2212,8 +2212,7 @@ retry: * We have a transaction open. All is sweetness. It also sets * i_size in generic_commit_write(). */ - err = __page_symlink(inode, symname, l, - mapping_gfp_mask(inode->i_mapping) & ~__GFP_FS); + err = __page_symlink(inode, symname, l, 1); if (err) { clear_nlink(inode); unlock_new_inode(inode); diff --git a/fs/fuse/file.c b/fs/fuse/file.c index 34930a964b82..4c9ee7011265 100644 --- a/fs/fuse/file.c +++ b/fs/fuse/file.c @@ -646,7 +646,7 @@ static int fuse_write_begin(struct file *file, struct address_space *mapping, { pgoff_t index = pos >> PAGE_CACHE_SHIFT; - *pagep = __grab_cache_page(mapping, index); + *pagep = grab_cache_page_write_begin(mapping, index, flags); if (!*pagep) return -ENOMEM; return 0; @@ -779,7 +779,7 @@ static ssize_t fuse_fill_write_pages(struct fuse_req *req, break; err = -ENOMEM; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, 0); if (!page) break; diff --git a/fs/gfs2/ops_address.c b/fs/gfs2/ops_address.c index 27563816e1c5..15f710f2d4da 100644 --- a/fs/gfs2/ops_address.c +++ b/fs/gfs2/ops_address.c @@ -675,7 +675,7 @@ static int gfs2_write_begin(struct file *file, struct address_space *mapping, goto out_trans_fail; error = -ENOMEM; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); *pagep = page; if (unlikely(!page)) goto out_endtrans; diff --git a/fs/hostfs/hostfs_kern.c b/fs/hostfs/hostfs_kern.c index 3a31451ac170..5c538e0ec14b 100644 --- a/fs/hostfs/hostfs_kern.c +++ b/fs/hostfs/hostfs_kern.c @@ -501,7 +501,7 @@ int hostfs_write_begin(struct file *file, struct address_space *mapping, { pgoff_t index = pos >> PAGE_CACHE_SHIFT; - *pagep = __grab_cache_page(mapping, index); + *pagep = grab_cache_page_write_begin(mapping, index, flags); if (!*pagep) return -ENOMEM; return 0; diff --git a/fs/jffs2/file.c b/fs/jffs2/file.c index 5a98aa87c853..5edc2bf20581 100644 --- a/fs/jffs2/file.c +++ b/fs/jffs2/file.c @@ -132,7 +132,7 @@ static int jffs2_write_begin(struct file *filp, struct address_space *mapping, uint32_t pageofs = index << PAGE_CACHE_SHIFT; int ret = 0; - pg = __grab_cache_page(mapping, index); + pg = grab_cache_page_write_begin(mapping, index, flags); if (!pg) return -ENOMEM; *pagep = pg; diff --git a/fs/libfs.c b/fs/libfs.c index e960a8321902..bdaec17fa388 100644 --- a/fs/libfs.c +++ b/fs/libfs.c @@ -360,7 +360,7 @@ int simple_write_begin(struct file *file, struct address_space *mapping, index = pos >> PAGE_CACHE_SHIFT; from = pos & (PAGE_CACHE_SIZE - 1); - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; diff --git a/fs/namei.c b/fs/namei.c index dd5c9f0bf829..df2d3df4f049 100644 --- a/fs/namei.c +++ b/fs/namei.c @@ -2817,18 +2817,23 @@ void page_put_link(struct dentry *dentry, struct nameidata *nd, void *cookie) } } -int __page_symlink(struct inode *inode, const char *symname, int len, - gfp_t gfp_mask) +/* + * The nofs argument instructs pagecache_write_begin to pass AOP_FLAG_NOFS + */ +int __page_symlink(struct inode *inode, const char *symname, int len, int nofs) { struct address_space *mapping = inode->i_mapping; struct page *page; void *fsdata; int err; char *kaddr; + unsigned int flags = AOP_FLAG_UNINTERRUPTIBLE; + if (nofs) + flags |= AOP_FLAG_NOFS; retry: err = pagecache_write_begin(NULL, mapping, 0, len-1, - AOP_FLAG_UNINTERRUPTIBLE, &page, &fsdata); + flags, &page, &fsdata); if (err) goto fail; @@ -2852,7 +2857,7 @@ fail: int page_symlink(struct inode *inode, const char *symname, int len) { return __page_symlink(inode, symname, len, - mapping_gfp_mask(inode->i_mapping)); + !(mapping_gfp_mask(inode->i_mapping) & __GFP_FS)); } const struct inode_operations page_symlink_inode_operations = { diff --git a/fs/nfs/file.c b/fs/nfs/file.c index d319b49f8f06..90f292b520d2 100644 --- a/fs/nfs/file.c +++ b/fs/nfs/file.c @@ -354,7 +354,7 @@ static int nfs_write_begin(struct file *file, struct address_space *mapping, file->f_path.dentry->d_name.name, mapping->host->i_ino, len, (long long) pos); - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; *pagep = page; diff --git a/fs/reiserfs/inode.c b/fs/reiserfs/inode.c index 145c2d3e5e01..ed04f47007f8 100644 --- a/fs/reiserfs/inode.c +++ b/fs/reiserfs/inode.c @@ -2561,7 +2561,7 @@ static int reiserfs_write_begin(struct file *file, } index = pos >> PAGE_CACHE_SHIFT; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (!page) return -ENOMEM; *pagep = page; diff --git a/fs/smbfs/file.c b/fs/smbfs/file.c index e4f8d51a5553..92d5e8ffb639 100644 --- a/fs/smbfs/file.c +++ b/fs/smbfs/file.c @@ -297,7 +297,7 @@ static int smb_write_begin(struct file *file, struct address_space *mapping, struct page **pagep, void **fsdata) { pgoff_t index = pos >> PAGE_CACHE_SHIFT; - *pagep = __grab_cache_page(mapping, index); + *pagep = grab_cache_page_write_begin(mapping, index, flags); if (!*pagep) return -ENOMEM; return 0; diff --git a/fs/ubifs/file.c b/fs/ubifs/file.c index fe82d2464d46..bf37374567fa 100644 --- a/fs/ubifs/file.c +++ b/fs/ubifs/file.c @@ -219,7 +219,8 @@ static void release_existing_page_budget(struct ubifs_info *c) } static int write_begin_slow(struct address_space *mapping, - loff_t pos, unsigned len, struct page **pagep) + loff_t pos, unsigned len, struct page **pagep, + unsigned flags) { struct inode *inode = mapping->host; struct ubifs_info *c = inode->i_sb->s_fs_info; @@ -247,7 +248,7 @@ static int write_begin_slow(struct address_space *mapping, if (unlikely(err)) return err; - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (unlikely(!page)) { ubifs_release_budget(c, &req); return -ENOMEM; @@ -438,7 +439,7 @@ static int ubifs_write_begin(struct file *file, struct address_space *mapping, return -EROFS; /* Try out the fast-path part first */ - page = __grab_cache_page(mapping, index); + page = grab_cache_page_write_begin(mapping, index, flags); if (unlikely(!page)) return -ENOMEM; @@ -483,7 +484,7 @@ static int ubifs_write_begin(struct file *file, struct address_space *mapping, unlock_page(page); page_cache_release(page); - return write_begin_slow(mapping, pos, len, pagep); + return write_begin_slow(mapping, pos, len, pagep, flags); } /* diff --git a/include/linux/fs.h b/include/linux/fs.h index e2170ee21e18..f2a3010140e3 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -423,6 +423,9 @@ enum positive_aop_returns { #define AOP_FLAG_UNINTERRUPTIBLE 0x0001 /* will not do a short write */ #define AOP_FLAG_CONT_EXPAND 0x0002 /* called from cont_expand */ +#define AOP_FLAG_NOFS 0x0004 /* used by filesystem to direct + * helper code (eg buffer layer) + * to clear GFP_FS from alloc */ /* * oh the beauties of C type declarations. @@ -2035,7 +2038,7 @@ extern int page_readlink(struct dentry *, char __user *, int); extern void *page_follow_link_light(struct dentry *, struct nameidata *); extern void page_put_link(struct dentry *, struct nameidata *, void *); extern int __page_symlink(struct inode *inode, const char *symname, int len, - gfp_t gfp_mask); + int nofs); extern int page_symlink(struct inode *inode, const char *symname, int len); extern const struct inode_operations page_symlink_inode_operations; extern int generic_readlink(struct dentry *, char __user *, int); diff --git a/include/linux/pagemap.h b/include/linux/pagemap.h index 709742be02f0..01ca0856caff 100644 --- a/include/linux/pagemap.h +++ b/include/linux/pagemap.h @@ -241,7 +241,8 @@ unsigned find_get_pages_contig(struct address_space *mapping, pgoff_t start, unsigned find_get_pages_tag(struct address_space *mapping, pgoff_t *index, int tag, unsigned int nr_pages, struct page **pages); -struct page *__grab_cache_page(struct address_space *mapping, pgoff_t index); +struct page *grab_cache_page_write_begin(struct address_space *mapping, + pgoff_t index, unsigned flags); /* * Returns locked page at given index in given cache, creating it if needed. diff --git a/mm/filemap.c b/mm/filemap.c index f3e5f8944d17..f8c69273c37f 100644 --- a/mm/filemap.c +++ b/mm/filemap.c @@ -2140,19 +2140,24 @@ EXPORT_SYMBOL(generic_file_direct_write); * Find or create a page at the given pagecache position. Return the locked * page. This function is specifically for buffered writes. */ -struct page *__grab_cache_page(struct address_space *mapping, pgoff_t index) +struct page *grab_cache_page_write_begin(struct address_space *mapping, + pgoff_t index, unsigned flags) { int status; struct page *page; + gfp_t gfp_notmask = 0; + if (flags & AOP_FLAG_NOFS) + gfp_notmask = __GFP_FS; repeat: page = find_lock_page(mapping, index); if (likely(page)) return page; - page = page_cache_alloc(mapping); + page = __page_cache_alloc(mapping_gfp_mask(mapping) & ~gfp_notmask); if (!page) return NULL; - status = add_to_page_cache_lru(page, mapping, index, GFP_KERNEL); + status = add_to_page_cache_lru(page, mapping, index, + GFP_KERNEL & ~gfp_notmask); if (unlikely(status)) { page_cache_release(page); if (status == -EEXIST) @@ -2161,7 +2166,7 @@ repeat: } return page; } -EXPORT_SYMBOL(__grab_cache_page); +EXPORT_SYMBOL(grab_cache_page_write_begin); static ssize_t generic_perform_write(struct file *file, struct iov_iter *i, loff_t pos) -- cgit v1.2.3 From e9079cce201784632aed4b1a3121ee38c1ced0b6 Mon Sep 17 00:00:00 2001 From: Steven Whitehouse Date: Tue, 14 Oct 2008 14:43:29 +0100 Subject: GFS2: Support for FIEMAP ioctl This patch implements the FIEMAP ioctl for GFS2. We can use the generic code (aside from a lock order issue, solved as per Ted Tso's suggestion) for which I've introduced a new variant of the generic function. We also have one exception to deal with, namely stuffed files, so we do that "by hand", setting all the required flags. This has been tested with a modified (I could only find an old version) of Eric's test program, and appears to work correctly. This patch does not currently support FIEMAP of xattrs, but the plan is to add that feature at some future point. Signed-off-by: Steven Whitehouse Cc: Theodore Tso Cc: Eric Sandeen --- fs/gfs2/ops_inode.c | 46 ++++++++++++++++++++++++++++++++++++++++++++++ fs/ioctl.c | 44 ++++++++++++++++++++++++++++++++++---------- include/linux/fs.h | 3 +++ 3 files changed, 83 insertions(+), 10 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/gfs2/ops_inode.c b/fs/gfs2/ops_inode.c index d232991b9046..1e24b65e1d23 100644 --- a/fs/gfs2/ops_inode.c +++ b/fs/gfs2/ops_inode.c @@ -19,6 +19,7 @@ #include #include #include +#include #include #include "gfs2.h" @@ -1212,6 +1213,48 @@ static int gfs2_removexattr(struct dentry *dentry, const char *name) return gfs2_ea_remove(GFS2_I(dentry->d_inode), &er); } +static int gfs2_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, + u64 start, u64 len) +{ + struct gfs2_inode *ip = GFS2_I(inode); + struct gfs2_holder gh; + int ret; + + ret = fiemap_check_flags(fieinfo, FIEMAP_FLAG_SYNC); + if (ret) + return ret; + + mutex_lock(&inode->i_mutex); + + ret = gfs2_glock_nq_init(ip->i_gl, LM_ST_SHARED, 0, &gh); + if (ret) + goto out; + + if (gfs2_is_stuffed(ip)) { + u64 phys = ip->i_no_addr << inode->i_blkbits; + u64 size = i_size_read(inode); + u32 flags = FIEMAP_EXTENT_LAST|FIEMAP_EXTENT_NOT_ALIGNED| + FIEMAP_EXTENT_DATA_INLINE; + phys += sizeof(struct gfs2_dinode); + phys += start; + if (start + len > size) + len = size - start; + if (start < size) + ret = fiemap_fill_next_extent(fieinfo, start, phys, + len, flags); + if (ret == 1) + ret = 0; + } else { + ret = __generic_block_fiemap(inode, fieinfo, start, len, + gfs2_block_map); + } + + gfs2_glock_dq_uninit(&gh); +out: + mutex_unlock(&inode->i_mutex); + return ret; +} + const struct inode_operations gfs2_file_iops = { .permission = gfs2_permission, .setattr = gfs2_setattr, @@ -1220,6 +1263,7 @@ const struct inode_operations gfs2_file_iops = { .getxattr = gfs2_getxattr, .listxattr = gfs2_listxattr, .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, }; const struct inode_operations gfs2_dir_iops = { @@ -1239,6 +1283,7 @@ const struct inode_operations gfs2_dir_iops = { .getxattr = gfs2_getxattr, .listxattr = gfs2_listxattr, .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, }; const struct inode_operations gfs2_symlink_iops = { @@ -1251,5 +1296,6 @@ const struct inode_operations gfs2_symlink_iops = { .getxattr = gfs2_getxattr, .listxattr = gfs2_listxattr, .removexattr = gfs2_removexattr, + .fiemap = gfs2_fiemap, }; diff --git a/fs/ioctl.c b/fs/ioctl.c index 43e8b2c0664b..cc3f1aa1cf7b 100644 --- a/fs/ioctl.c +++ b/fs/ioctl.c @@ -231,7 +231,8 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg) #define blk_to_logical(inode, blk) (blk << (inode)->i_blkbits) #define logical_to_blk(inode, offset) (offset >> (inode)->i_blkbits); -/* +/** + * __generic_block_fiemap - FIEMAP for block based inodes (no locking) * @inode - the inode to map * @arg - the pointer to userspace where we copy everything to * @get_block - the fs's get_block function @@ -242,11 +243,15 @@ static int ioctl_fiemap(struct file *filp, unsigned long arg) * * If it is possible to have data blocks beyond a hole past @inode->i_size, then * please do not use this function, it will stop at the first unmapped block - * beyond i_size + * beyond i_size. + * + * If you use this function directly, you need to do your own locking. Use + * generic_block_fiemap if you want the locking done for you. */ -int generic_block_fiemap(struct inode *inode, - struct fiemap_extent_info *fieinfo, u64 start, - u64 len, get_block_t *get_block) + +int __generic_block_fiemap(struct inode *inode, + struct fiemap_extent_info *fieinfo, u64 start, + u64 len, get_block_t *get_block) { struct buffer_head tmp; unsigned int start_blk; @@ -260,9 +265,6 @@ int generic_block_fiemap(struct inode *inode, start_blk = logical_to_blk(inode, start); - /* guard against change */ - mutex_lock(&inode->i_mutex); - length = (long long)min_t(u64, len, i_size_read(inode)); map_len = length; @@ -334,14 +336,36 @@ int generic_block_fiemap(struct inode *inode, cond_resched(); } while (1); - mutex_unlock(&inode->i_mutex); - /* if ret is 1 then we just hit the end of the extent array */ if (ret == 1) ret = 0; return ret; } +EXPORT_SYMBOL(__generic_block_fiemap); + +/** + * generic_block_fiemap - FIEMAP for block based inodes + * @inode: The inode to map + * @fieinfo: The mapping information + * @start: The initial block to map + * @len: The length of the extect to attempt to map + * @get_block: The block mapping function for the fs + * + * Calls __generic_block_fiemap to map the inode, after taking + * the inode's mutex lock. + */ + +int generic_block_fiemap(struct inode *inode, + struct fiemap_extent_info *fieinfo, u64 start, + u64 len, get_block_t *get_block) +{ + int ret; + mutex_lock(&inode->i_mutex); + ret = __generic_block_fiemap(inode, fieinfo, start, len, get_block); + mutex_unlock(&inode->i_mutex); + return ret; +} EXPORT_SYMBOL(generic_block_fiemap); #endif /* CONFIG_BLOCK */ diff --git a/include/linux/fs.h b/include/linux/fs.h index f2a3010140e3..e34bc6925fdf 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -2059,6 +2059,9 @@ extern int vfs_fstat(unsigned int, struct kstat *); extern int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd, unsigned long arg); +extern int __generic_block_fiemap(struct inode *inode, + struct fiemap_extent_info *fieinfo, u64 start, + u64 len, get_block_t *get_block); extern int generic_block_fiemap(struct inode *inode, struct fiemap_extent_info *fieinfo, u64 start, u64 len, get_block_t *get_block); -- cgit v1.2.3 From 4c728ef583b3d82266584da5cb068294c09df31e Mon Sep 17 00:00:00 2001 From: Christoph Hellwig Date: Mon, 22 Dec 2008 21:11:15 +0100 Subject: add a vfs_fsync helper Fsync currently has a fdatawrite/fdatawait pair around the method call, and a mutex_lock/unlock of the inode mutex. All callers of fsync have to duplicate this, but we have a few and most of them don't quite get it right. This patch adds a new vfs_fsync that takes care of this. It's a little more complicated as usual as ->fsync might get a NULL file pointer and just a dentry from nfsd, but otherwise gets afile and we want to take the mapping and file operations from it when it is there. Notes on the fsync callers: - ecryptfs wasn't calling filemap_fdatawrite / filemap_fdatawait on the lower file - coda wasn't calling filemap_fdatawrite / filemap_fdatawait on the host file, and returning 0 when ->fsync was missing - shm wasn't calling either filemap_fdatawrite / filemap_fdatawait nor taking i_mutex. Now given that shared memory doesn't have disk backing not doing anything in fsync seems fine and I left it out of the vfs_fsync conversion for now, but in that case we might just not pass it through to the lower file at all but just call the no-op simple_sync_file directly. [and now actually export vfs_fsync] Signed-off-by: Christoph Hellwig Signed-off-by: Al Viro --- drivers/usb/gadget/file_storage.c | 18 +-------------- fs/coda/file.c | 12 ++-------- fs/ecryptfs/file.c | 15 +++--------- fs/nfsd/vfs.c | 35 +++------------------------- fs/sync.c | 48 ++++++++++++++++++++++++++++++--------- include/linux/fs.h | 2 +- mm/msync.c | 2 +- 7 files changed, 48 insertions(+), 84 deletions(-) (limited to 'include/linux/fs.h') diff --git a/drivers/usb/gadget/file_storage.c b/drivers/usb/gadget/file_storage.c index c4e62a6297d7..2e71368f45b4 100644 --- a/drivers/usb/gadget/file_storage.c +++ b/drivers/usb/gadget/file_storage.c @@ -1863,26 +1863,10 @@ static int do_write(struct fsg_dev *fsg) static int fsync_sub(struct lun *curlun) { struct file *filp = curlun->filp; - struct inode *inode; - int rc, err; if (curlun->ro || !filp) return 0; - if (!filp->f_op->fsync) - return -EINVAL; - - inode = filp->f_path.dentry->d_inode; - mutex_lock(&inode->i_mutex); - rc = filemap_fdatawrite(inode->i_mapping); - err = filp->f_op->fsync(filp, filp->f_path.dentry, 1); - if (!rc) - rc = err; - err = filemap_fdatawait(inode->i_mapping); - if (!rc) - rc = err; - mutex_unlock(&inode->i_mutex); - VLDBG(curlun, "fdatasync -> %d\n", rc); - return rc; + return vfs_fsync(filp, filp->f_path.dentry, 1); } static void fsync_all(struct fsg_dev *fsg) diff --git a/fs/coda/file.c b/fs/coda/file.c index 466303db2df6..6a347fbc998a 100644 --- a/fs/coda/file.c +++ b/fs/coda/file.c @@ -201,8 +201,7 @@ int coda_release(struct inode *coda_inode, struct file *coda_file) int coda_fsync(struct file *coda_file, struct dentry *coda_dentry, int datasync) { struct file *host_file; - struct dentry *host_dentry; - struct inode *host_inode, *coda_inode = coda_dentry->d_inode; + struct inode *coda_inode = coda_dentry->d_inode; struct coda_file_info *cfi; int err = 0; @@ -214,14 +213,7 @@ int coda_fsync(struct file *coda_file, struct dentry *coda_dentry, int datasync) BUG_ON(!cfi || cfi->cfi_magic != CODA_MAGIC); host_file = cfi->cfi_container; - if (host_file->f_op && host_file->f_op->fsync) { - host_dentry = host_file->f_path.dentry; - host_inode = host_dentry->d_inode; - mutex_lock(&host_inode->i_mutex); - err = host_file->f_op->fsync(host_file, host_dentry, datasync); - mutex_unlock(&host_inode->i_mutex); - } - + err = vfs_fsync(host_file, host_file->f_path.dentry, datasync); if ( !err && !datasync ) { lock_kernel(); err = venus_fsync(coda_inode->i_sb, coda_i2f(coda_inode)); diff --git a/fs/ecryptfs/file.c b/fs/ecryptfs/file.c index eb3dc4c7ac06..713834371229 100644 --- a/fs/ecryptfs/file.c +++ b/fs/ecryptfs/file.c @@ -275,18 +275,9 @@ static int ecryptfs_release(struct inode *inode, struct file *file) static int ecryptfs_fsync(struct file *file, struct dentry *dentry, int datasync) { - struct file *lower_file = ecryptfs_file_to_lower(file); - struct dentry *lower_dentry = ecryptfs_dentry_to_lower(dentry); - struct inode *lower_inode = lower_dentry->d_inode; - int rc = -EINVAL; - - if (lower_inode->i_fop->fsync) { - mutex_lock(&lower_inode->i_mutex); - rc = lower_inode->i_fop->fsync(lower_file, lower_dentry, - datasync); - mutex_unlock(&lower_inode->i_mutex); - } - return rc; + return vfs_fsync(ecryptfs_file_to_lower(file), + ecryptfs_dentry_to_lower(dentry), + datasync); } static int ecryptfs_fasync(int fd, struct file *file, int flag) diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 5245a3965004..44aa92aba891 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -744,45 +744,16 @@ nfsd_close(struct file *filp) fput(filp); } -/* - * Sync a file - * As this calls fsync (not fdatasync) there is no need for a write_inode - * after it. - */ -static inline int nfsd_dosync(struct file *filp, struct dentry *dp, - const struct file_operations *fop) -{ - struct inode *inode = dp->d_inode; - int (*fsync) (struct file *, struct dentry *, int); - int err; - - err = filemap_fdatawrite(inode->i_mapping); - if (err == 0 && fop && (fsync = fop->fsync)) - err = fsync(filp, dp, 0); - if (err == 0) - err = filemap_fdatawait(inode->i_mapping); - - return err; -} - - static int nfsd_sync(struct file *filp) { - int err; - struct inode *inode = filp->f_path.dentry->d_inode; - dprintk("nfsd: sync file %s\n", filp->f_path.dentry->d_name.name); - mutex_lock(&inode->i_mutex); - err=nfsd_dosync(filp, filp->f_path.dentry, filp->f_op); - mutex_unlock(&inode->i_mutex); - - return err; + return vfs_fsync(filp, filp->f_path.dentry, 0); } int -nfsd_sync_dir(struct dentry *dp) +nfsd_sync_dir(struct dentry *dentry) { - return nfsd_dosync(NULL, dp, dp->d_inode->i_fop); + return vfs_fsync(NULL, dentry, 0); } /* diff --git a/fs/sync.c b/fs/sync.c index 2967562d416f..0921d6d4b5e6 100644 --- a/fs/sync.c +++ b/fs/sync.c @@ -75,14 +75,39 @@ int file_fsync(struct file *filp, struct dentry *dentry, int datasync) return ret; } -long do_fsync(struct file *file, int datasync) +/** + * vfs_fsync - perform a fsync or fdatasync on a file + * @file: file to sync + * @dentry: dentry of @file + * @data: only perform a fdatasync operation + * + * Write back data and metadata for @file to disk. If @datasync is + * set only metadata needed to access modified file data is written. + * + * In case this function is called from nfsd @file may be %NULL and + * only @dentry is set. This can only happen when the filesystem + * implements the export_operations API. + */ +int vfs_fsync(struct file *file, struct dentry *dentry, int datasync) { - int ret; - int err; - struct address_space *mapping = file->f_mapping; + const struct file_operations *fop; + struct address_space *mapping; + int err, ret; + + /* + * Get mapping and operations from the file in case we have + * as file, or get the default values for them in case we + * don't have a struct file available. Damn nfsd.. + */ + if (file) { + mapping = file->f_mapping; + fop = file->f_op; + } else { + mapping = dentry->d_inode->i_mapping; + fop = dentry->d_inode->i_fop; + } - if (!file->f_op || !file->f_op->fsync) { - /* Why? We can still call filemap_fdatawrite */ + if (!fop || !fop->fsync) { ret = -EINVAL; goto out; } @@ -94,7 +119,7 @@ long do_fsync(struct file *file, int datasync) * livelocks in fsync_buffers_list(). */ mutex_lock(&mapping->host->i_mutex); - err = file->f_op->fsync(file, file->f_path.dentry, datasync); + err = fop->fsync(file, dentry, datasync); if (!ret) ret = err; mutex_unlock(&mapping->host->i_mutex); @@ -104,15 +129,16 @@ long do_fsync(struct file *file, int datasync) out: return ret; } +EXPORT_SYMBOL(vfs_fsync); -static long __do_fsync(unsigned int fd, int datasync) +static int do_fsync(unsigned int fd, int datasync) { struct file *file; int ret = -EBADF; file = fget(fd); if (file) { - ret = do_fsync(file, datasync); + ret = vfs_fsync(file, file->f_path.dentry, datasync); fput(file); } return ret; @@ -120,12 +146,12 @@ static long __do_fsync(unsigned int fd, int datasync) asmlinkage long sys_fsync(unsigned int fd) { - return __do_fsync(fd, 0); + return do_fsync(fd, 0); } asmlinkage long sys_fdatasync(unsigned int fd) { - return __do_fsync(fd, 1); + return do_fsync(fd, 1); } /* diff --git a/include/linux/fs.h b/include/linux/fs.h index e2170ee21e18..9ad9eac9eb0c 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1827,7 +1827,7 @@ extern int __filemap_fdatawrite_range(struct address_space *mapping, extern int filemap_fdatawrite_range(struct address_space *mapping, loff_t start, loff_t end); -extern long do_fsync(struct file *file, int datasync); +extern int vfs_fsync(struct file *file, struct dentry *dentry, int datasync); extern void sync_supers(void); extern void sync_filesystems(int wait); extern void __fsync_super(struct super_block *sb); diff --git a/mm/msync.c b/mm/msync.c index 144a7570535d..07dae08cf31c 100644 --- a/mm/msync.c +++ b/mm/msync.c @@ -82,7 +82,7 @@ asmlinkage long sys_msync(unsigned long start, size_t len, int flags) (vma->vm_flags & VM_SHARED)) { get_file(file); up_read(&mm->mmap_sem); - error = do_fsync(file, 0); + error = vfs_fsync(file, file->f_path.dentry, 0); fput(file); if (error || start >= end) goto out; -- cgit v1.2.3 From 856bf4d717feb8c55d4e2f817b71ebb70cfbc67b Mon Sep 17 00:00:00 2001 From: Nick Piggin Date: Tue, 6 Jan 2009 14:40:26 -0800 Subject: fs: sys_sync fix s_syncing livelock avoidance was breaking data integrity guarantee of sys_sync, by allowing sys_sync to skip writing or waiting for superblocks if there is a concurrent sys_sync happening. This livelock avoidance is much less important now that we don't have the get_super_to_sync() call after every sb that we sync. This was replaced by __put_super_and_need_restart. Signed-off-by: Nick Piggin Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/fs-writeback.c | 20 +------------------- include/linux/fs.h | 1 - 2 files changed, 1 insertion(+), 20 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/fs-writeback.c b/fs/fs-writeback.c index a9ee474f9691..e5eaa62fd17f 100644 --- a/fs/fs-writeback.c +++ b/fs/fs-writeback.c @@ -652,18 +652,6 @@ void sync_inodes_sb(struct super_block *sb, int wait) sync_sb_inodes(sb, &wbc); } -/* - * Rather lame livelock avoidance. - */ -static void set_sb_syncing(int val) -{ - struct super_block *sb; - spin_lock(&sb_lock); - list_for_each_entry_reverse(sb, &super_blocks, s_list) - sb->s_syncing = val; - spin_unlock(&sb_lock); -} - /** * sync_inodes - writes all inodes to disk * @wait: wait for completion @@ -690,9 +678,6 @@ static void __sync_inodes(int wait) spin_lock(&sb_lock); restart: list_for_each_entry(sb, &super_blocks, s_list) { - if (sb->s_syncing) - continue; - sb->s_syncing = 1; sb->s_count++; spin_unlock(&sb_lock); down_read(&sb->s_umount); @@ -710,13 +695,10 @@ restart: void sync_inodes(int wait) { - set_sb_syncing(0); __sync_inodes(0); - if (wait) { - set_sb_syncing(0); + if (wait) __sync_inodes(1); - } } /** diff --git a/include/linux/fs.h b/include/linux/fs.h index fb59673c60b1..d7eba77f666e 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1133,7 +1133,6 @@ struct super_block { struct rw_semaphore s_umount; struct mutex s_lock; int s_count; - int s_syncing; int s_need_sync_fs; atomic_t s_active; #ifdef CONFIG_SECURITY -- cgit v1.2.3 From efaee192063a54749c56b7383803e16fe553630e Mon Sep 17 00:00:00 2001 From: Arjan van de Ven Date: Tue, 6 Jan 2009 07:20:54 -0800 Subject: async: make the final inode deletion an asynchronous event this makes "rm -rf" on a (names cached) kernel tree go from 11.6 to 8.6 seconds on an ext3 filesystem Signed-off-by: Arjan van de Ven --- fs/inode.c | 20 +++++++++++++------- fs/super.c | 10 ++++++++++ include/linux/fs.h | 5 +++++ 3 files changed, 28 insertions(+), 7 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/inode.c b/fs/inode.c index 7a6e8c2ff7b1..0013ac1af8e7 100644 --- a/fs/inode.c +++ b/fs/inode.c @@ -22,6 +22,7 @@ #include #include #include +#include /* * This is needed for the following functions: @@ -1138,16 +1139,11 @@ EXPORT_SYMBOL(remove_inode_hash); * I_FREEING is set so that no-one will take a new reference to the inode while * it is being deleted. */ -void generic_delete_inode(struct inode *inode) +static void generic_delete_inode_async(void *data, async_cookie_t cookie) { + struct inode *inode = data; const struct super_operations *op = inode->i_sb->s_op; - list_del_init(&inode->i_list); - list_del_init(&inode->i_sb_list); - inode->i_state |= I_FREEING; - inodes_stat.nr_inodes--; - spin_unlock(&inode_lock); - security_inode_delete(inode); if (op->delete_inode) { @@ -1171,6 +1167,16 @@ void generic_delete_inode(struct inode *inode) destroy_inode(inode); } +void generic_delete_inode(struct inode *inode) +{ + list_del_init(&inode->i_list); + list_del_init(&inode->i_sb_list); + inode->i_state |= I_FREEING; + inodes_stat.nr_inodes--; + spin_unlock(&inode_lock); + async_schedule_special(generic_delete_inode_async, inode, &inode->i_sb->s_async_list); +} + EXPORT_SYMBOL(generic_delete_inode); static void generic_forget_inode(struct inode *inode) diff --git a/fs/super.c b/fs/super.c index ddba069d7a99..cb20744ec789 100644 --- a/fs/super.c +++ b/fs/super.c @@ -38,6 +38,7 @@ #include #include #include +#include #include #include "internal.h" @@ -71,6 +72,7 @@ static struct super_block *alloc_super(struct file_system_type *type) INIT_HLIST_HEAD(&s->s_anon); INIT_LIST_HEAD(&s->s_inodes); INIT_LIST_HEAD(&s->s_dentry_lru); + INIT_LIST_HEAD(&s->s_async_list); init_rwsem(&s->s_umount); mutex_init(&s->s_lock); lockdep_set_class(&s->s_umount, &type->s_umount_key); @@ -289,11 +291,18 @@ void generic_shutdown_super(struct super_block *sb) { const struct super_operations *sop = sb->s_op; + if (sb->s_root) { shrink_dcache_for_umount(sb); fsync_super(sb); lock_super(sb); sb->s_flags &= ~MS_ACTIVE; + + /* + * wait for asynchronous fs operations to finish before going further + */ + async_synchronize_full_special(&sb->s_async_list); + /* bad name - it should be evict_inodes() */ invalidate_inodes(sb); lock_kernel(); @@ -449,6 +458,7 @@ void sync_filesystems(int wait) if (sb->s_flags & MS_RDONLY) continue; sb->s_need_sync_fs = 1; + async_synchronize_full_special(&sb->s_async_list); } restart: diff --git a/include/linux/fs.h b/include/linux/fs.h index d7eba77f666e..e38a64d71eff 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1184,6 +1184,11 @@ struct super_block { * generic_show_options() */ char *s_options; + + /* + * storage for asynchronous operations + */ + struct list_head s_async_list; }; extern struct timespec current_fs_time(struct super_block *sb); -- cgit v1.2.3 From c4be0c1dc4cdc37b175579be1460f15ac6495e9a Mon Sep 17 00:00:00 2001 From: Takashi Sato Date: Fri, 9 Jan 2009 16:40:58 -0800 Subject: filesystem freeze: add error handling of write_super_lockfs/unlockfs Currently, ext3 in mainline Linux doesn't have the freeze feature which suspends write requests. So, we cannot take a backup which keeps the filesystem's consistency with the storage device's features (snapshot and replication) while it is mounted. In many case, a commercial filesystem (e.g. VxFS) has the freeze feature and it would be used to get the consistent backup. If Linux's standard filesystem ext3 has the freeze feature, we can do it without a commercial filesystem. So I have implemented the ioctls of the freeze feature. I think we can take the consistent backup with the following steps. 1. Freeze the filesystem with the freeze ioctl. 2. Separate the replication volume or create the snapshot with the storage device's feature. 3. Unfreeze the filesystem with the unfreeze ioctl. 4. Take the backup from the separated replication volume or the snapshot. This patch: VFS: Changed the type of write_super_lockfs and unlockfs from "void" to "int" so that they can return an error. Rename write_super_lockfs and unlockfs of the super block operation freeze_fs and unfreeze_fs to avoid a confusion. ext3, ext4, xfs, gfs2, jfs: Changed the type of write_super_lockfs and unlockfs from "void" to "int" so that write_super_lockfs returns an error if needed, and unlockfs always returns 0. reiserfs: Changed the type of write_super_lockfs and unlockfs from "void" to "int" so that they always return 0 (success) to keep a current behavior. Signed-off-by: Takashi Sato Signed-off-by: Masayuki Hamaguchi Cc: Cc: Cc: Christoph Hellwig Cc: Dave Kleikamp Cc: Dave Chinner Cc: Alasdair G Kergon Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- Documentation/filesystems/Locking | 8 +++---- Documentation/filesystems/vfs.txt | 8 +++---- fs/buffer.c | 8 +++---- fs/ext3/super.c | 45 +++++++++++++++++++++++++-------------- fs/ext4/super.c | 45 +++++++++++++++++++++++++++------------ fs/gfs2/ops_super.c | 16 ++++++++------ fs/jfs/super.c | 10 +++++---- fs/reiserfs/super.c | 10 +++++---- fs/xfs/linux-2.6/xfs_super.c | 8 +++---- fs/xfs/xfs_fsops.c | 11 ++++++---- fs/xfs/xfs_fsops.h | 2 +- include/linux/fs.h | 4 ++-- 12 files changed, 107 insertions(+), 68 deletions(-) (limited to 'include/linux/fs.h') diff --git a/Documentation/filesystems/Locking b/Documentation/filesystems/Locking index cfbfa15a46ba..ec6a9392a173 100644 --- a/Documentation/filesystems/Locking +++ b/Documentation/filesystems/Locking @@ -97,8 +97,8 @@ prototypes: void (*put_super) (struct super_block *); void (*write_super) (struct super_block *); int (*sync_fs)(struct super_block *sb, int wait); - void (*write_super_lockfs) (struct super_block *); - void (*unlockfs) (struct super_block *); + int (*freeze_fs) (struct super_block *); + int (*unfreeze_fs) (struct super_block *); int (*statfs) (struct dentry *, struct kstatfs *); int (*remount_fs) (struct super_block *, int *, char *); void (*clear_inode) (struct inode *); @@ -119,8 +119,8 @@ delete_inode: no put_super: yes yes no write_super: no yes read sync_fs: no no read -write_super_lockfs: ? -unlockfs: ? +freeze_fs: ? +unfreeze_fs: ? statfs: no no no remount_fs: yes yes maybe (see below) clear_inode: no diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt index ef19afa186a9..deeeed0faa8f 100644 --- a/Documentation/filesystems/vfs.txt +++ b/Documentation/filesystems/vfs.txt @@ -210,8 +210,8 @@ struct super_operations { void (*put_super) (struct super_block *); void (*write_super) (struct super_block *); int (*sync_fs)(struct super_block *sb, int wait); - void (*write_super_lockfs) (struct super_block *); - void (*unlockfs) (struct super_block *); + int (*freeze_fs) (struct super_block *); + int (*unfreeze_fs) (struct super_block *); int (*statfs) (struct dentry *, struct kstatfs *); int (*remount_fs) (struct super_block *, int *, char *); void (*clear_inode) (struct inode *); @@ -270,11 +270,11 @@ or bottom half). a superblock. The second parameter indicates whether the method should wait until the write out has been completed. Optional. - write_super_lockfs: called when VFS is locking a filesystem and + freeze_fs: called when VFS is locking a filesystem and forcing it into a consistent state. This method is currently used by the Logical Volume Manager (LVM). - unlockfs: called when VFS is unlocking a filesystem and making it writable + unfreeze_fs: called when VFS is unlocking a filesystem and making it writable again. statfs: called when the VFS needs to get filesystem statistics. This diff --git a/fs/buffer.c b/fs/buffer.c index c26da785938a..87f9e537b8c3 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -221,8 +221,8 @@ struct super_block *freeze_bdev(struct block_device *bdev) sync_blockdev(sb->s_bdev); - if (sb->s_op->write_super_lockfs) - sb->s_op->write_super_lockfs(sb); + if (sb->s_op->freeze_fs) + sb->s_op->freeze_fs(sb); } sync_blockdev(bdev); @@ -242,8 +242,8 @@ void thaw_bdev(struct block_device *bdev, struct super_block *sb) if (sb) { BUG_ON(sb->s_bdev != bdev); - if (sb->s_op->unlockfs) - sb->s_op->unlockfs(sb); + if (sb->s_op->unfreeze_fs) + sb->s_op->unfreeze_fs(sb); sb->s_frozen = SB_UNFROZEN; smp_wmb(); wake_up(&sb->s_wait_unfrozen); diff --git a/fs/ext3/super.c b/fs/ext3/super.c index 5d047a030a73..b70d90e08a3c 100644 --- a/fs/ext3/super.c +++ b/fs/ext3/super.c @@ -48,8 +48,8 @@ static int ext3_load_journal(struct super_block *, struct ext3_super_block *, unsigned long journal_devnum); static int ext3_create_journal(struct super_block *, struct ext3_super_block *, unsigned int); -static void ext3_commit_super (struct super_block * sb, - struct ext3_super_block * es, +static int ext3_commit_super(struct super_block *sb, + struct ext3_super_block *es, int sync); static void ext3_mark_recovery_complete(struct super_block * sb, struct ext3_super_block * es); @@ -60,9 +60,9 @@ static const char *ext3_decode_error(struct super_block * sb, int errno, char nbuf[16]); static int ext3_remount (struct super_block * sb, int * flags, char * data); static int ext3_statfs (struct dentry * dentry, struct kstatfs * buf); -static void ext3_unlockfs(struct super_block *sb); +static int ext3_unfreeze(struct super_block *sb); static void ext3_write_super (struct super_block * sb); -static void ext3_write_super_lockfs(struct super_block *sb); +static int ext3_freeze(struct super_block *sb); /* * Wrappers for journal_start/end. @@ -759,8 +759,8 @@ static const struct super_operations ext3_sops = { .put_super = ext3_put_super, .write_super = ext3_write_super, .sync_fs = ext3_sync_fs, - .write_super_lockfs = ext3_write_super_lockfs, - .unlockfs = ext3_unlockfs, + .freeze_fs = ext3_freeze, + .unfreeze_fs = ext3_unfreeze, .statfs = ext3_statfs, .remount_fs = ext3_remount, .clear_inode = ext3_clear_inode, @@ -2311,21 +2311,23 @@ static int ext3_create_journal(struct super_block * sb, return 0; } -static void ext3_commit_super (struct super_block * sb, - struct ext3_super_block * es, +static int ext3_commit_super(struct super_block *sb, + struct ext3_super_block *es, int sync) { struct buffer_head *sbh = EXT3_SB(sb)->s_sbh; + int error = 0; if (!sbh) - return; + return error; es->s_wtime = cpu_to_le32(get_seconds()); es->s_free_blocks_count = cpu_to_le32(ext3_count_free_blocks(sb)); es->s_free_inodes_count = cpu_to_le32(ext3_count_free_inodes(sb)); BUFFER_TRACE(sbh, "marking dirty"); mark_buffer_dirty(sbh); if (sync) - sync_dirty_buffer(sbh); + error = sync_dirty_buffer(sbh); + return error; } @@ -2439,12 +2441,14 @@ static int ext3_sync_fs(struct super_block *sb, int wait) * LVM calls this function before a (read-only) snapshot is created. This * gives us a chance to flush the journal completely and mark the fs clean. */ -static void ext3_write_super_lockfs(struct super_block *sb) +static int ext3_freeze(struct super_block *sb) { + int error = 0; + journal_t *journal; sb->s_dirt = 0; if (!(sb->s_flags & MS_RDONLY)) { - journal_t *journal = EXT3_SB(sb)->s_journal; + journal = EXT3_SB(sb)->s_journal; /* Now we set up the journal barrier. */ journal_lock_updates(journal); @@ -2453,20 +2457,28 @@ static void ext3_write_super_lockfs(struct super_block *sb) * We don't want to clear needs_recovery flag when we failed * to flush the journal. */ - if (journal_flush(journal) < 0) - return; + error = journal_flush(journal); + if (error < 0) + goto out; /* Journal blocked and flushed, clear needs_recovery flag. */ EXT3_CLEAR_INCOMPAT_FEATURE(sb, EXT3_FEATURE_INCOMPAT_RECOVER); - ext3_commit_super(sb, EXT3_SB(sb)->s_es, 1); + error = ext3_commit_super(sb, EXT3_SB(sb)->s_es, 1); + if (error) + goto out; } + return 0; + +out: + journal_unlock_updates(journal); + return error; } /* * Called by LVM after the snapshot is done. We need to reset the RECOVER * flag here, even though the filesystem is not technically dirty yet. */ -static void ext3_unlockfs(struct super_block *sb) +static int ext3_unfreeze(struct super_block *sb) { if (!(sb->s_flags & MS_RDONLY)) { lock_super(sb); @@ -2476,6 +2488,7 @@ static void ext3_unlockfs(struct super_block *sb) unlock_super(sb); journal_unlock_updates(EXT3_SB(sb)->s_journal); } + return 0; } static int ext3_remount (struct super_block * sb, int * flags, char * data) diff --git a/fs/ext4/super.c b/fs/ext4/super.c index 8f7e0be8ab1b..e5f06a5f045e 100644 --- a/fs/ext4/super.c +++ b/fs/ext4/super.c @@ -51,7 +51,7 @@ struct proc_dir_entry *ext4_proc_root; static int ext4_load_journal(struct super_block *, struct ext4_super_block *, unsigned long journal_devnum); -static void ext4_commit_super(struct super_block *sb, +static int ext4_commit_super(struct super_block *sb, struct ext4_super_block *es, int sync); static void ext4_mark_recovery_complete(struct super_block *sb, struct ext4_super_block *es); @@ -62,9 +62,9 @@ static const char *ext4_decode_error(struct super_block *sb, int errno, char nbuf[16]); static int ext4_remount(struct super_block *sb, int *flags, char *data); static int ext4_statfs(struct dentry *dentry, struct kstatfs *buf); -static void ext4_unlockfs(struct super_block *sb); +static int ext4_unfreeze(struct super_block *sb); static void ext4_write_super(struct super_block *sb); -static void ext4_write_super_lockfs(struct super_block *sb); +static int ext4_freeze(struct super_block *sb); ext4_fsblk_t ext4_block_bitmap(struct super_block *sb, @@ -978,8 +978,8 @@ static const struct super_operations ext4_sops = { .put_super = ext4_put_super, .write_super = ext4_write_super, .sync_fs = ext4_sync_fs, - .write_super_lockfs = ext4_write_super_lockfs, - .unlockfs = ext4_unlockfs, + .freeze_fs = ext4_freeze, + .unfreeze_fs = ext4_unfreeze, .statfs = ext4_statfs, .remount_fs = ext4_remount, .clear_inode = ext4_clear_inode, @@ -2888,13 +2888,14 @@ static int ext4_load_journal(struct super_block *sb, return 0; } -static void ext4_commit_super(struct super_block *sb, +static int ext4_commit_super(struct super_block *sb, struct ext4_super_block *es, int sync) { struct buffer_head *sbh = EXT4_SB(sb)->s_sbh; + int error = 0; if (!sbh) - return; + return error; if (buffer_write_io_error(sbh)) { /* * Oh, dear. A previous attempt to write the @@ -2918,14 +2919,19 @@ static void ext4_commit_super(struct super_block *sb, BUFFER_TRACE(sbh, "marking dirty"); mark_buffer_dirty(sbh); if (sync) { - sync_dirty_buffer(sbh); - if (buffer_write_io_error(sbh)) { + error = sync_dirty_buffer(sbh); + if (error) + return error; + + error = buffer_write_io_error(sbh); + if (error) { printk(KERN_ERR "EXT4-fs: I/O error while writing " "superblock for %s.\n", sb->s_id); clear_buffer_write_io_error(sbh); set_buffer_uptodate(sbh); } } + return error; } @@ -3058,12 +3064,14 @@ static int ext4_sync_fs(struct super_block *sb, int wait) * LVM calls this function before a (read-only) snapshot is created. This * gives us a chance to flush the journal completely and mark the fs clean. */ -static void ext4_write_super_lockfs(struct super_block *sb) +static int ext4_freeze(struct super_block *sb) { + int error = 0; + journal_t *journal; sb->s_dirt = 0; if (!(sb->s_flags & MS_RDONLY)) { - journal_t *journal = EXT4_SB(sb)->s_journal; + journal = EXT4_SB(sb)->s_journal; if (journal) { /* Now we set up the journal barrier. */ @@ -3073,21 +3081,29 @@ static void ext4_write_super_lockfs(struct super_block *sb) * We don't want to clear needs_recovery flag when we * failed to flush the journal. */ - if (jbd2_journal_flush(journal) < 0) - return; + error = jbd2_journal_flush(journal); + if (error < 0) + goto out; } /* Journal blocked and flushed, clear needs_recovery flag. */ EXT4_CLEAR_INCOMPAT_FEATURE(sb, EXT4_FEATURE_INCOMPAT_RECOVER); ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1); + error = ext4_commit_super(sb, EXT4_SB(sb)->s_es, 1); + if (error) + goto out; } + return 0; +out: + jbd2_journal_unlock_updates(journal); + return error; } /* * Called by LVM after the snapshot is done. We need to reset the RECOVER * flag here, even though the filesystem is not technically dirty yet. */ -static void ext4_unlockfs(struct super_block *sb) +static int ext4_unfreeze(struct super_block *sb) { if (EXT4_SB(sb)->s_journal && !(sb->s_flags & MS_RDONLY)) { lock_super(sb); @@ -3097,6 +3113,7 @@ static void ext4_unlockfs(struct super_block *sb) unlock_super(sb); jbd2_journal_unlock_updates(EXT4_SB(sb)->s_journal); } + return 0; } static int ext4_remount(struct super_block *sb, int *flags, char *data) diff --git a/fs/gfs2/ops_super.c b/fs/gfs2/ops_super.c index 777783deddcb..320323d03479 100644 --- a/fs/gfs2/ops_super.c +++ b/fs/gfs2/ops_super.c @@ -211,18 +211,18 @@ static int gfs2_sync_fs(struct super_block *sb, int wait) } /** - * gfs2_write_super_lockfs - prevent further writes to the filesystem + * gfs2_freeze - prevent further writes to the filesystem * @sb: the VFS structure for the filesystem * */ -static void gfs2_write_super_lockfs(struct super_block *sb) +static int gfs2_freeze(struct super_block *sb) { struct gfs2_sbd *sdp = sb->s_fs_info; int error; if (test_bit(SDF_SHUTDOWN, &sdp->sd_flags)) - return; + return -EINVAL; for (;;) { error = gfs2_freeze_fs(sdp); @@ -242,17 +242,19 @@ static void gfs2_write_super_lockfs(struct super_block *sb) fs_err(sdp, "retrying...\n"); msleep(1000); } + return 0; } /** - * gfs2_unlockfs - reallow writes to the filesystem + * gfs2_unfreeze - reallow writes to the filesystem * @sb: the VFS structure for the filesystem * */ -static void gfs2_unlockfs(struct super_block *sb) +static int gfs2_unfreeze(struct super_block *sb) { gfs2_unfreeze_fs(sb->s_fs_info); + return 0; } /** @@ -688,8 +690,8 @@ const struct super_operations gfs2_super_ops = { .put_super = gfs2_put_super, .write_super = gfs2_write_super, .sync_fs = gfs2_sync_fs, - .write_super_lockfs = gfs2_write_super_lockfs, - .unlockfs = gfs2_unlockfs, + .freeze_fs = gfs2_freeze, + .unfreeze_fs = gfs2_unfreeze, .statfs = gfs2_statfs, .remount_fs = gfs2_remount_fs, .clear_inode = gfs2_clear_inode, diff --git a/fs/jfs/super.c b/fs/jfs/super.c index 0dae345e481b..b37d1f78b854 100644 --- a/fs/jfs/super.c +++ b/fs/jfs/super.c @@ -543,7 +543,7 @@ out_kfree: return ret; } -static void jfs_write_super_lockfs(struct super_block *sb) +static int jfs_freeze(struct super_block *sb) { struct jfs_sb_info *sbi = JFS_SBI(sb); struct jfs_log *log = sbi->log; @@ -553,9 +553,10 @@ static void jfs_write_super_lockfs(struct super_block *sb) lmLogShutdown(log); updateSuper(sb, FM_CLEAN); } + return 0; } -static void jfs_unlockfs(struct super_block *sb) +static int jfs_unfreeze(struct super_block *sb) { struct jfs_sb_info *sbi = JFS_SBI(sb); struct jfs_log *log = sbi->log; @@ -568,6 +569,7 @@ static void jfs_unlockfs(struct super_block *sb) else txResume(sb); } + return 0; } static int jfs_get_sb(struct file_system_type *fs_type, @@ -735,8 +737,8 @@ static const struct super_operations jfs_super_operations = { .delete_inode = jfs_delete_inode, .put_super = jfs_put_super, .sync_fs = jfs_sync_fs, - .write_super_lockfs = jfs_write_super_lockfs, - .unlockfs = jfs_unlockfs, + .freeze_fs = jfs_freeze, + .unfreeze_fs = jfs_unfreeze, .statfs = jfs_statfs, .remount_fs = jfs_remount, .show_options = jfs_show_options, diff --git a/fs/reiserfs/super.c b/fs/reiserfs/super.c index c55651f1407c..f3c820b75829 100644 --- a/fs/reiserfs/super.c +++ b/fs/reiserfs/super.c @@ -83,7 +83,7 @@ static void reiserfs_write_super(struct super_block *s) reiserfs_sync_fs(s, 1); } -static void reiserfs_write_super_lockfs(struct super_block *s) +static int reiserfs_freeze(struct super_block *s) { struct reiserfs_transaction_handle th; reiserfs_write_lock(s); @@ -101,11 +101,13 @@ static void reiserfs_write_super_lockfs(struct super_block *s) } s->s_dirt = 0; reiserfs_write_unlock(s); + return 0; } -static void reiserfs_unlockfs(struct super_block *s) +static int reiserfs_unfreeze(struct super_block *s) { reiserfs_allow_writes(s); + return 0; } extern const struct in_core_key MAX_IN_CORE_KEY; @@ -613,8 +615,8 @@ static const struct super_operations reiserfs_sops = { .put_super = reiserfs_put_super, .write_super = reiserfs_write_super, .sync_fs = reiserfs_sync_fs, - .write_super_lockfs = reiserfs_write_super_lockfs, - .unlockfs = reiserfs_unlockfs, + .freeze_fs = reiserfs_freeze, + .unfreeze_fs = reiserfs_unfreeze, .statfs = reiserfs_statfs, .remount_fs = reiserfs_remount, .show_options = generic_show_options, diff --git a/fs/xfs/linux-2.6/xfs_super.c b/fs/xfs/linux-2.6/xfs_super.c index be846d606ae8..95a971080368 100644 --- a/fs/xfs/linux-2.6/xfs_super.c +++ b/fs/xfs/linux-2.6/xfs_super.c @@ -1269,14 +1269,14 @@ xfs_fs_remount( * need to take care of the metadata. Once that's done write a dummy * record to dirty the log in case of a crash while frozen. */ -STATIC void -xfs_fs_lockfs( +STATIC int +xfs_fs_freeze( struct super_block *sb) { struct xfs_mount *mp = XFS_M(sb); xfs_quiesce_attr(mp); - xfs_fs_log_dummy(mp); + return -xfs_fs_log_dummy(mp); } STATIC int @@ -1557,7 +1557,7 @@ static struct super_operations xfs_super_operations = { .put_super = xfs_fs_put_super, .write_super = xfs_fs_write_super, .sync_fs = xfs_fs_sync_super, - .write_super_lockfs = xfs_fs_lockfs, + .freeze_fs = xfs_fs_freeze, .statfs = xfs_fs_statfs, .remount_fs = xfs_fs_remount, .show_options = xfs_fs_show_options, diff --git a/fs/xfs/xfs_fsops.c b/fs/xfs/xfs_fsops.c index 852b6d32e8d0..680d0e0ec932 100644 --- a/fs/xfs/xfs_fsops.c +++ b/fs/xfs/xfs_fsops.c @@ -595,17 +595,19 @@ out: return 0; } -void +int xfs_fs_log_dummy( xfs_mount_t *mp) { xfs_trans_t *tp; xfs_inode_t *ip; + int error; tp = _xfs_trans_alloc(mp, XFS_TRANS_DUMMY1); - if (xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0)) { + error = xfs_trans_reserve(tp, 0, XFS_ICHANGE_LOG_RES(mp), 0, 0, 0); + if (error) { xfs_trans_cancel(tp, 0); - return; + return error; } ip = mp->m_rootip; @@ -615,9 +617,10 @@ xfs_fs_log_dummy( xfs_trans_ihold(tp, ip); xfs_trans_log_inode(tp, ip, XFS_ILOG_CORE); xfs_trans_set_sync(tp); - xfs_trans_commit(tp, 0); + error = xfs_trans_commit(tp, 0); xfs_iunlock(ip, XFS_ILOCK_EXCL); + return error; } int diff --git a/fs/xfs/xfs_fsops.h b/fs/xfs/xfs_fsops.h index 300d0c9d61ad..88435e0a77c9 100644 --- a/fs/xfs/xfs_fsops.h +++ b/fs/xfs/xfs_fsops.h @@ -25,6 +25,6 @@ extern int xfs_fs_counts(xfs_mount_t *mp, xfs_fsop_counts_t *cnt); extern int xfs_reserve_blocks(xfs_mount_t *mp, __uint64_t *inval, xfs_fsop_resblks_t *outval); extern int xfs_fs_goingdown(xfs_mount_t *mp, __uint32_t inflags); -extern void xfs_fs_log_dummy(xfs_mount_t *mp); +extern int xfs_fs_log_dummy(xfs_mount_t *mp); #endif /* __XFS_FSOPS_H__ */ diff --git a/include/linux/fs.h b/include/linux/fs.h index 0b87b29f4797..3e59182de9df 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -1377,8 +1377,8 @@ struct super_operations { void (*put_super) (struct super_block *); void (*write_super) (struct super_block *); int (*sync_fs)(struct super_block *sb, int wait); - void (*write_super_lockfs) (struct super_block *); - void (*unlockfs) (struct super_block *); + int (*freeze_fs) (struct super_block *); + int (*unfreeze_fs) (struct super_block *); int (*statfs) (struct dentry *, struct kstatfs *); int (*remount_fs) (struct super_block *, int *, char *); void (*clear_inode) (struct inode *); -- cgit v1.2.3 From fcccf502540e3d752d33b2d8e976034dee81f9f7 Mon Sep 17 00:00:00 2001 From: Takashi Sato Date: Fri, 9 Jan 2009 16:40:59 -0800 Subject: filesystem freeze: implement generic freeze feature The ioctls for the generic freeze feature are below. o Freeze the filesystem int ioctl(int fd, int FIFREEZE, arg) fd: The file descriptor of the mountpoint FIFREEZE: request code for the freeze arg: Ignored Return value: 0 if the operation succeeds. Otherwise, -1 o Unfreeze the filesystem int ioctl(int fd, int FITHAW, arg) fd: The file descriptor of the mountpoint FITHAW: request code for unfreeze arg: Ignored Return value: 0 if the operation succeeds. Otherwise, -1 Error number: If the filesystem has already been unfrozen, errno is set to EINVAL. [akpm@linux-foundation.org: fix CONFIG_BLOCK=n] Signed-off-by: Takashi Sato Signed-off-by: Masayuki Hamaguchi Cc: Cc: Cc: Christoph Hellwig Cc: Dave Kleikamp Cc: Dave Chinner Cc: Alasdair G Kergon Cc: Al Viro Signed-off-by: Andrew Morton Signed-off-by: Linus Torvalds --- fs/block_dev.c | 2 ++ fs/buffer.c | 74 +++++++++++++++++++++++++++++++++++++++------ fs/ioctl.c | 46 ++++++++++++++++++++++++++++ include/linux/buffer_head.h | 11 ++++++- include/linux/fs.h | 7 +++++ 5 files changed, 130 insertions(+), 10 deletions(-) (limited to 'include/linux/fs.h') diff --git a/fs/block_dev.c b/fs/block_dev.c index ac7031f12ea5..b3c1efff5e1d 100644 --- a/fs/block_dev.c +++ b/fs/block_dev.c @@ -285,6 +285,8 @@ static void init_once(void *foo) INIT_LIST_HEAD(&bdev->bd_holder_list); #endif inode_init_once(&ei->vfs_inode); + /* Initialize mutex for freeze. */ + mutex_init(&bdev->bd_fsfreeze_mutex); } static inline void __bd_forget(struct inode *inode) diff --git a/fs/buffer.c b/fs/buffer.c index 87f9e537b8c3..b6e8b8632e2f 100644 --- a/fs/buffer.c +++ b/fs/buffer.c @@ -203,10 +203,25 @@ int fsync_bdev(struct block_device *bdev) * happen on bdev until thaw_bdev() is called. * If a superblock is found on this device, we take the s_umount semaphore * on it to make sure nobody unmounts until the snapshot creation is done. + * The reference counter (bd_fsfreeze_count) guarantees that only the last + * unfreeze process can unfreeze the frozen filesystem actually when multiple + * freeze requests arrive simultaneously. It counts up in freeze_bdev() and + * count down in thaw_bdev(). When it becomes 0, thaw_bdev() will unfreeze + * actually. */ struct super_block *freeze_bdev(struct block_device *bdev) { struct super_block *sb; + int error = 0; + + mutex_lock(&bdev->bd_fsfreeze_mutex); + if (bdev->bd_fsfreeze_count > 0) { + bdev->bd_fsfreeze_count++; + sb = get_super(bdev); + mutex_unlock(&bdev->bd_fsfreeze_mutex); + return sb; + } + bdev->bd_fsfreeze_count++; down(&bdev->bd_mount_sem); sb = get_super(bdev); @@ -221,11 +236,24 @@ struct super_block *freeze_bdev(struct block_device *bdev) sync_blockdev(sb->s_bdev); - if (sb->s_op->freeze_fs) - sb->s_op->freeze_fs(sb); + if (sb->s_op->freeze_fs) { + error = sb->s_op->freeze_fs(sb); + if (error) { + printk(KERN_ERR + "VFS:Filesystem freeze failed\n"); + sb->s_frozen = SB_UNFROZEN; + drop_super(sb); + up(&bdev->bd_mount_sem); + bdev->bd_fsfreeze_count--; + mutex_unlock(&bdev->bd_fsfreeze_mutex); + return ERR_PTR(error); + } + } } sync_blockdev(bdev); + mutex_unlock(&bdev->bd_fsfreeze_mutex); + return sb; /* thaw_bdev releases s->s_umount and bd_mount_sem */ } EXPORT_SYMBOL(freeze_bdev); @@ -237,20 +265,48 @@ EXPORT_SYMBOL(freeze_bdev); * * Unlocks the filesystem and marks it writeable again after freeze_bdev(). */ -void thaw_bdev(struct block_device *bdev, struct super_block *sb) +int thaw_bdev(struct block_device *bdev, struct super_block *sb) { + int error = 0; + + mutex_lock(&bdev->bd_fsfreeze_mutex); + if (!bdev->bd_fsfreeze_count) { + mutex_unlock(&bdev->bd_fsfreeze_mutex); + return -EINVAL; + } + + bdev->bd_fsfreeze_count--; + if (bdev->bd_fsfreeze_count > 0) { + if (sb) + drop_super(sb); + mutex_unlock(&bdev->bd_fsfreeze_mutex); + return 0; + } + if (sb) { BUG_ON(sb->s_bdev != bdev); - - if (sb->s_op->unfreeze_fs) - sb->s_op->unfreeze_fs(sb); - sb->s_frozen = SB_UNFROZEN; - smp_wmb(); - wake_up(&sb->s_wait_unfrozen); + if (!(sb->s_flags & MS_RDONLY)) { + if (sb->s_op->unfreeze_fs) { + error = sb->s_op->unfreeze_fs(sb); + if (error) { + printk(KERN_ERR + "VFS:Filesystem thaw failed\n"); + sb->s_frozen = SB_FREEZE_TRANS; + bdev->bd_fsfreeze_count++; + mutex_unlock(&bdev->bd_fsfreeze_mutex); + return error; + } + } + sb->s_frozen = SB_UNFROZEN; + smp_wmb(); + wake_up(&sb->s_wait_unfrozen); + } drop_super(sb); } up(&bdev->bd_mount_sem); + mutex_unlock(&bdev->bd_fsfreeze_mutex); + return 0; } EXPORT_SYMBOL(thaw_bdev); diff --git a/fs/ioctl.c b/fs/ioctl.c index cc3f1aa1cf7b..20b0a8a24c6b 100644 --- a/fs/ioctl.c +++ b/fs/ioctl.c @@ -439,6 +439,43 @@ static int ioctl_fioasync(unsigned int fd, struct file *filp, return error; } +static int ioctl_fsfreeze(struct file *filp) +{ + struct super_block *sb = filp->f_path.dentry->d_inode->i_sb; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + /* If filesystem doesn't support freeze feature, return. */ + if (sb->s_op->freeze_fs == NULL) + return -EOPNOTSUPP; + + /* If a blockdevice-backed filesystem isn't specified, return. */ + if (sb->s_bdev == NULL) + return -EINVAL; + + /* Freeze */ + sb = freeze_bdev(sb->s_bdev); + if (IS_ERR(sb)) + return PTR_ERR(sb); + return 0; +} + +static int ioctl_fsthaw(struct file *filp) +{ + struct super_block *sb = filp->f_path.dentry->d_inode->i_sb; + + if (!capable(CAP_SYS_ADMIN)) + return -EPERM; + + /* If a blockdevice-backed filesystem isn't specified, return EINVAL. */ + if (sb->s_bdev == NULL) + return -EINVAL; + + /* Thaw */ + return thaw_bdev(sb->s_bdev, sb); +} + /* * When you add any new common ioctls to the switches above and below * please update compat_sys_ioctl() too. @@ -486,6 +523,15 @@ int do_vfs_ioctl(struct file *filp, unsigned int fd, unsigned int cmd, } else error = -ENOTTY; break; + + case FIFREEZE: + error = ioctl_fsfreeze(filp); + break; + + case FITHAW: + error = ioctl_fsthaw(filp); + break; + default: if (S_ISREG(filp->f_path.dentry->d_inode->i_mode)) error = file_ioctl(filp, cmd, arg); diff --git a/include/linux/buffer_head.h b/include/linux/buffer_head.h index 8605f8a74df9..bd7ac793be19 100644 --- a/include/linux/buffer_head.h +++ b/include/linux/buffer_head.h @@ -171,7 +171,7 @@ void __wait_on_buffer(struct buffer_head *); wait_queue_head_t *bh_waitq_head(struct buffer_head *bh); int fsync_bdev(struct block_device *); struct super_block *freeze_bdev(struct block_device *); -void thaw_bdev(struct block_device *, struct super_block *); +int thaw_bdev(struct block_device *, struct super_block *); int fsync_super(struct super_block *); int fsync_no_super(struct block_device *); struct buffer_head *__find_get_block(struct block_device *bdev, sector_t block, @@ -346,6 +346,15 @@ static inline int remove_inode_buffers(struct inode *inode) { return 1; } static inline int sync_mapping_buffers(struct address_space *mapping) { return 0; } static inline void invalidate_bdev(struct block_device *bdev) {} +static inline struct super_block *freeze_bdev(struct block_device *sb) +{ + return NULL; +} + +static inline int thaw_bdev(struct block_device *bdev, struct super_block *sb) +{ + return 0; +} #endif /* CONFIG_BLOCK */ #endif /* _LINUX_BUFFER_HEAD_H */ diff --git a/include/linux/fs.h b/include/linux/fs.h index 3e59182de9df..6022f44043f2 100644 --- a/include/linux/fs.h +++ b/include/linux/fs.h @@ -234,6 +234,8 @@ struct inodes_stat_t { #define BMAP_IOCTL 1 /* obsolete - kept for compatibility */ #define FIBMAP _IO(0x00,1) /* bmap access */ #define FIGETBSZ _IO(0x00,2) /* get the block size used for bmap */ +#define FIFREEZE _IOWR('X', 119, int) /* Freeze */ +#define FITHAW _IOWR('X', 120, int) /* Thaw */ #define FS_IOC_GETFLAGS _IOR('f', 1, long) #define FS_IOC_SETFLAGS _IOW('f', 2, long) @@ -591,6 +593,11 @@ struct block_device { * care to not mess up bd_private for that case. */ unsigned long bd_private; + + /* The counter of freeze processes */ + int bd_fsfreeze_count; + /* Mutex for freeze */ + struct mutex bd_fsfreeze_mutex; }; /* -- cgit v1.2.3