fsnotify: Generate FS_CREATE event before FS_OPEN
Jan Kara
maintainerv3 cover fsnotify: Generate FS_CREATE event before FS_OPEN
Hello,
this is a third revision of the series Neil started making sure FS_CREATE event is generated before FS_OPEN event even for filesystems using atomic_open.
Changes since v2:
* Don't generate FS_OPEN events for O_PATH fd opens
* Move generation of FS_OPEN event for atomic_open() into a more logical place
Previous version:
v2: https://lore.kernel.org/all/[email protected]/
Honza
Jan Kara
maintainerPATCH v3 1/2 fsnotify: Do not generate events for O_PATH file descriptors
Currently we will not generate FS_OPEN events for O_PATH file descriptors but we will generate FS_CLOSE events for them. This is asymmetry is confusing. Arguably no fsnotify events should be generated for O_PATH file descriptors as they cannot be used to access or modify file content, they are just convenient handles to file objects like paths. So fix the asymmetry by stopping to generate FS_CLOSE for O_PATH file descriptors.
--- include/linux/fsnotify.h | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)
include/linux/fsnotify.h +7 -1
@@ -112,7 +112,13 @@ static inline int fsnotify_file(struct file *file, __u32 mask) {const struct path *path; -if (file->f_mode &FMODE_NONOTIFY ) +/* +* FMODE_NONOTIFY are fds generated by fanotify itself which should not +* generate new events. We also don't want to generate events for +* FMODE_PATH fds (involves open & close events) as they are just +* handle creation / destruction events and not "real" file events. +*/ +if (file->f_mode & (FMODE_NONOTIFY |FMODE_PATH ))return 0 ; path = &file->f_path;
Signature · 1 line
2.35.3
Re: [PATCH 1/2] fsnotify: Do not generate events for O_PATH file descriptors
Quoted 10 lines · Jan Kara
Currently we will not generate
FS_OPENevents forO_PATHfile descriptors but we will generateFS_CLOSEevents for them. This is asymmetry is confusing. Arguably no fsnotify events should be generated forO_PATHfile descriptors as they cannot be used to access or modify file content, they are just convenient handles to file objects like paths. So fix the asymmetry by stopping to generateFS_CLOSEforO_PATHfile descriptors.Signed-off-by Jan Kara <[email protected]>
Looks good.
Quoted 26 lines
--- include/linux/fsnotify.h | 8 +++++++- 1 file changed, 7 insertions(+), 1 deletion(-)include/linux/fsnotify.h +7
-2@@ -112,7 +112,13 @@ static inline int fsnotify_file(struct file *file, __u32 mask) {const struct path *path; -if (file->f_mode &FMODE_NONOTIFY ) +/* +* FMODE_NONOTIFY are fds generated by fanotify itself which should not +* generate new events. We also don't want to generate events for +* FMODE_PATH fds (involves open & close events) as they are just +* handle creation / destruction events and not "real" file events. +*/ +if (file->f_mode & (FMODE_NONOTIFY |FMODE_PATH ))return 0 ; path = &file->f_path; --2.35.3
Re: [PATCH 1/2] fsnotify: Do not generate events for O_PATH file descriptors
In reply to Jan KaraQuoted 9 lines · Jan Kara
Currently we will not generate
FS_OPENevents forO_PATHfile descriptors but we will generateFS_CLOSEevents for them. This is asymmetry is confusing. Arguably no fsnotify events should be generated forO_PATHfile descriptors as they cannot be used to access or modify file content, they are just convenient handles to file objects like paths. So fix the asymmetry by stopping to generateFS_CLOSEforO_PATHfile descriptors.[...]
I added a Cc stable to the first patch because this seems like a bugfix per the mail discussion.
---
Applied to the vfs.fixes branch of the vfs/vfs.git tree. Patches in the vfs.fixes branch should appear in linux-next soon.
Please report any outstanding bugs that were missed during review in a new review to the original patch series allowing us to drop it.
It's encouraged to provide Acked-bys and Reviewed-bys even though the patch has now been applied. If possible patch trailers will be updated.
Note that commit hashes shown below are subject to change due to rebase, trailer updates or similar. If in doubt, please check the listed branch.
tree: https://git.kernel.org/pub/scm/linux/kernel/git/vfs/vfs.git branch: vfs.fixes
[1/2] fsnotify: Do not generate events for O_PATH file descriptors
https://git.kernel.org/vfs/vfs/c/702eb71fd650 [2/2] vfs: generate FS_CREATE before FS_OPEN when ->atomic_open used. https://git.kernel.org/vfs/vfs/c/7d1cf5e624ef
Jan Kara
maintainerRe: [PATCH 1/2] fsnotify: Do not generate events for O_PATH file descriptors
Quoted 9 lines · Christian Brauner
Quoted 8 lines · Jan Kara
Currently we will not generate
FS_OPENevents forO_PATHfile descriptors but we will generateFS_CLOSEevents for them. This is asymmetry is confusing. Arguably no fsnotify events should be generated forO_PATHfile descriptors as they cannot be used to access or modify file content, they are just convenient handles to file objects like paths. So fix the asymmetry by stopping to generateFS_CLOSEforO_PATHfile descriptors.
Quoted 4 lines
Quoted 1 line
[...]
I added a Cc stable to the first patch because this seems like a bugfix per the mail discussion.
Yes, makes sense. Thanks!
Honza
Signature · 2 lines
Jan Kara <[email protected]> SUSE Labs, CR
Jan Kara
maintainerPATCH v3 2/2 vfs: generate FS_CREATE before FS_OPEN when ->atomic_open used.
In reply to Jan KaraFrom: NeilBrown <[email protected]>
When a file is opened and created with open(..., O_CREAT) we get both the CREATE and OPEN fsnotify events and would expect them in that
order. For most filesystems we get them in that order becauseopen_last_lookups() calls fsnofify_create() and then do_open() (from path_openat()) calls vfs_open()->do_dentry_open() which calls fsnotify_open().
However when ->atomic_open is used, the
do_dentry_open() -> fsnotify_open()
call happens from finish_open() which is called from the ->atomic_open handler in lookup_open() which is called *before* open_last_lookups() calls fsnotify_create. So we get the "open" notification before "create" - which is backwards. ltp testcase inotify02 tests this and reports the inconsistency.
This patch lifts the fsnotify_open() call out of do_dentry_open() and places it higher up the call stack. There are three callers of do_dentry_open().
For vfs_open() and kernel_file_open() the fsnotify_open() is placed directly in that caller so there should be no behavioural change.
For finish_open() there are two cases:
- finish_open is used in ->atomic_open handlers. For these we add a call to fsnotify_open() at open_last_lookups() if FMODE_OPENED is set - which means do_dentry_open() has been called.
- finish_open is used in ->tmpfile() handlers. For these a similar call to fsnotify_open() is added to vfs_tmpfile()
With this patch NFSv3 is restored to its previous behaviour (before ->atomic_open support was added) of generating CREATE notifications before OPEN, and NFSv4 now has that same correct ordering that is has not had before. I haven't tested other filesystems.
--- fs/namei.c | 10 ++++++++-- fs/open.c | 22 +++++++++++++++------- 2 files changed, 23 insertions(+), 9 deletions(-)
fs/namei.c +8 -2
@@ -3572,8 +3572,12 @@ static const char *open_last_lookups(struct nameidata *nd,else inode_lock_shared (dir->d_inode); dentry =lookup_open (nd, file, op, got_write); -if (!IS_ERR (dentry) && (file->f_mode &FMODE_CREATED )) -fsnotify_create (dir->d_inode, dentry); +if (!IS_ERR (dentry)) { +if (file->f_mode &FMODE_CREATED ) +fsnotify_create (dir->d_inode, dentry); +if (file->f_mode &FMODE_OPENED ) +fsnotify_open (file); + }if (open_flag &O_CREAT )inode_unlock (dir->d_inode);else
@@ -3700,6 +3704,8 @@ int vfs_tmpfile(struct mnt_idmap *idmap, mode =vfs_prepare_mode (idmap, dir, mode, mode, mode); error = dir->i_op->tmpfile (idmap, dir, file, mode);dput (child); +if (file->f_mode &FMODE_OPENED ) +fsnotify_open (file);if (error)return error;/* Don't check for other permissions, the inode was just created */
fs/open.c +15 -7
@@ -1004,11 +1004,6 @@ static int do_dentry_open(struct file *f, } } -/* -* Once we return a file with FMODE_OPENED, __fput() will call -* fsnotify_close(), so we need fsnotify_open() here for symmetry. -*/ -fsnotify_open (f);return 0 ; cleanup_all:
@@ -1085,8 +1080,19 @@ EXPORT_SYMBOL(file_path);*/ int vfs_open (const struct path *path,struct file *file) { +int ret; + file->f_path = *path; -return do_dentry_open (file,NULL ); + ret =do_dentry_open (file,NULL ); +if (!ret) { +/* +* Once we return a file with FMODE_OPENED, __fput() will call +* fsnotify_close(), so we need fsnotify_open() here for +* symmetry. +*/ +fsnotify_open (file); + } +return ret; } struct file *dentry_open(const struct path *path, int flags,
@@ -1177,8 +1183,10 @@ struct file *kernel_file_open(const struct path *path, int flags, error =do_dentry_open (f,NULL );if (error) {fput (f); - f =ERR_PTR (error); +return ERR_PTR (error); } + +fsnotify_open (f);return f; }EXPORT_SYMBOL_GPL (kernel_file_open);
Signature · 1 line
2.35.3
Re: [PATCH 2/2] vfs: generate FS_CREATE before FS_OPEN when ->atomic_open used.
Quoted 37 lines · Jan Kara
From: NeilBrown <[email protected]>
When a file is opened and created with
open(..., O_CREAT)we get both the CREATE and OPEN fsnotify events and would expect them in that
order. For most filesystems we get them in that order becauseopen_last_lookups()callsfsnofify_create()and thendo_open()(frompath_openat()) calls vfs_open()->do_dentry_open() which callsfsnotify_open().However when ->atomic_open is used, the
do_dentry_open()->fsnotify_open()
call happens fromfinish_open()which is called from the ->atomic_open handler inlookup_open()which is called *before*open_last_lookups()callsfsnotify_create. So we get the "open" notification before "create" - which is backwards. ltp testcase inotify02 tests this and reports the inconsistency.This patch lifts the
fsnotify_open()call out ofdo_dentry_open()and places it higher up the call stack. There are three callers ofdo_dentry_open().For
vfs_open()andkernel_file_open()thefsnotify_open()is placed directly in that caller so there should be no behavioural change.For
finish_open()there are two cases:
-finish_openis used in ->atomic_open handlers. For these we add a call tofsnotify_open()atopen_last_lookups()ifFMODE_OPENEDis set - which meansdo_dentry_open()has been called.
-finish_openis used in ->tmpfile() handlers. For these a similar call tofsnotify_open()is added tovfs_tmpfile()With this patch NFSv3 is restored to its previous behaviour (before ->atomic_open support was added) of generating CREATE notifications before OPEN, and NFSv4 now has that same correct ordering that is has not had before. I haven't tested other filesystems.
Quoted 9 lines
Fixes 7c6c5249f061 ("NFS: add atomic_open for NFSv3 to handle O_TRUNC correctly.")Reported-by James Clark <[email protected]>Signed-off-by NeilBrown <[email protected]>Fixes 7b8c9d7bb457 ("fsnotify: move fsnotify_open() hook into do_dentry_open()")Tested-by James Clark <[email protected]>Signed-off-by Christian Brauner <[email protected]>Signed-off-by Jan Kara <[email protected]>
Looks good.
Quoted 85 lines
--- fs/namei.c | 10 ++++++++-- fs/open.c | 22 +++++++++++++++------- 2 files changed, 23 insertions(+), 9 deletions(-)fs/namei.c +8
-2@@ -3572,8 +3572,12 @@ static const char *open_last_lookups(struct nameidata *nd,else inode_lock_shared (dir->d_inode); dentry =lookup_open (nd, file, op, got_write); -if (!IS_ERR (dentry) && (file->f_mode &FMODE_CREATED )) -fsnotify_create (dir->d_inode, dentry); +if (!IS_ERR (dentry)) { +if (file->f_mode &FMODE_CREATED ) +fsnotify_create (dir->d_inode, dentry); +if (file->f_mode &FMODE_OPENED ) +fsnotify_open (file); + }if (open_flag &O_CREAT )inode_unlock (dir->d_inode);else @@ -3700,6 +3704,8 @@ int vfs_tmpfile(struct mnt_idmap *idmap, mode =vfs_prepare_mode (idmap, dir, mode, mode, mode); error = dir->i_op->tmpfile (idmap, dir, file, mode);dput (child); +if (file->f_mode &FMODE_OPENED ) +fsnotify_open (file);if (error)return error;/* Don't check for other permissions, the inode was just created */ fs/open.c +15
-8@@ -1004,11 +1004,6 @@ static int do_dentry_open(struct file *f, } } -/* -* Once we return a file with FMODE_OPENED, __fput() will call -* fsnotify_close(), so we need fsnotify_open() here for symmetry. -*/ -fsnotify_open (f);return 0 ; cleanup_all:@@ -1085,8 +1080,19 @@ EXPORT_SYMBOL(file_path);*/ int vfs_open (const struct path *path,struct file *file) { +int ret; + file->f_path = *path; -return do_dentry_open (file,NULL ); + ret =do_dentry_open (file,NULL ); +if (!ret) { +/* +* Once we return a file with FMODE_OPENED, __fput() will call +* fsnotify_close(), so we need fsnotify_open() here for +* symmetry. +*/ +fsnotify_open (file); + } +return ret; } struct file *dentry_open(const struct path *path, int flags,@@ -1177,8 +1183,10 @@ struct file *kernel_file_open(const struct path *path, int flags, error =do_dentry_open (f,NULL );if (error) {fput (f); - f =ERR_PTR (error); +return ERR_PTR (error); } + +fsnotify_open (f);return f; }EXPORT_SYMBOL_GPL (kernel_file_open); --2.35.3