summaryrefslogtreecommitdiff
path: root/Documentation/filesystems
diff options
context:
space:
mode:
authorLinus Torvalds <torvalds@linux-foundation.org>2013-09-19 04:22:22 +0400
committerLinus Torvalds <torvalds@linux-foundation.org>2013-09-19 04:22:22 +0400
commit3fe03debfc58aba8f5ddb32abd7343237d07714c (patch)
treef17d2fbf390810404f5d28f45ed5b379fa13c6ef /Documentation/filesystems
parent9baa5059485d6559081d27d9b1fe1714f3fccab1 (diff)
parent8061a6fa564fe0e71601632758b78d2ba737663c (diff)
downloadlinux-3fe03debfc58aba8f5ddb32abd7343237d07714c.tar.xz
Merge branch 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull vfs fixes from Al Viro: "atomic_open-related fixes (Miklos' series, with EEXIST-related parts replaced with fix in fs/namei.c:atomic_open() instead of messing with the instances) + race fix in autofs + leak on failure exit in 9p" * 'for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs: 9p: don't forget to destroy inode cache if fscache registration fails atomic_open: take care of EEXIST in no-open case with O_CREAT|O_EXCL in fs/namei.c vfs: don't set FILE_CREATED before calling ->atomic_open() nfs: set FILE_CREATED gfs2: set FILE_CREATED cifs: fix filp leak in cifs_atomic_open() vfs: improve i_op->atomic_open() documentation autofs4: close the races around autofs4_notify_daemon()
Diffstat (limited to 'Documentation/filesystems')
-rw-r--r--Documentation/filesystems/vfs.txt14
1 files changed, 7 insertions, 7 deletions
diff --git a/Documentation/filesystems/vfs.txt b/Documentation/filesystems/vfs.txt
index f93a88250a44..deb48b5fd883 100644
--- a/Documentation/filesystems/vfs.txt
+++ b/Documentation/filesystems/vfs.txt
@@ -359,11 +359,9 @@ struct inode_operations {
ssize_t (*listxattr) (struct dentry *, char *, size_t);
int (*removexattr) (struct dentry *, const char *);
void (*update_time)(struct inode *, struct timespec *, int);
- int (*atomic_open)(struct inode *, struct dentry *,
+ int (*atomic_open)(struct inode *, struct dentry *, struct file *,
+ unsigned open_flag, umode_t create_mode, int *opened);
int (*tmpfile) (struct inode *, struct dentry *, umode_t);
-} ____cacheline_aligned;
- struct file *, unsigned open_flag,
- umode_t create_mode, int *opened);
};
Again, all methods are called without any locks being held, unless
@@ -470,9 +468,11 @@ otherwise noted.
method the filesystem can look up, possibly create and open the file in
one atomic operation. If it cannot perform this (e.g. the file type
turned out to be wrong) it may signal this by returning 1 instead of
- usual 0 or -ve . This method is only called if the last
- component is negative or needs lookup. Cached positive dentries are
- still handled by f_op->open().
+ usual 0 or -ve . This method is only called if the last component is
+ negative or needs lookup. Cached positive dentries are still handled by
+ f_op->open(). If the file was created, the FILE_CREATED flag should be
+ set in "opened". In case of O_EXCL the method must only succeed if the
+ file didn't exist and hence FILE_CREATED shall always be set on success.
tmpfile: called in the end of O_TMPFILE open(). Optional, equivalent to
atomically creating, opening and unlinking a file in given directory.