diff options
| author | Trond Myklebust <trondmy@gmail.com> | 2019-11-27 17:05:51 -0500 |
|---|---|---|
| committer | J. Bruce Fields <bfields@redhat.com> | 2019-11-30 14:55:38 -0500 |
| commit | a25e3726b32c746c0098125d4c7463bb84df72bb (patch) | |
| tree | 7ae7376eae2f736843731a269c3a77f7920317d0 /fs/nfsd/vfs.c | |
| parent | 8729aaba74626c4ebce3abf1b9e96bb62d2958ca (diff) | |
nfsd: Ensure CLONE persists data and metadata changes to the target file
The NFSv4.2 CLONE operation has implicit persistence requirements on the
target file, since there is no protocol requirement that the client issue
a separate operation to persist data.
For that reason, we should call vfs_fsync_range() on the destination file
after a successful call to vfs_clone_file_range().
Fixes: ffa0160a1039 ("nfsd: implement the NFSv4.2 CLONE operation")
Signed-off-by: Trond Myklebust <trond.myklebust@hammerspace.com>
Cc: stable@vger.kernel.org # v4.5+
Signed-off-by: J. Bruce Fields <bfields@redhat.com>
Diffstat (limited to 'fs/nfsd/vfs.c')
| -rw-r--r-- | fs/nfsd/vfs.c | 8 |
1 files changed, 7 insertions, 1 deletions
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index bd0a385df3fc..cf423fea0c6f 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -525,7 +525,7 @@ __be32 nfsd4_set_nfs4_label(struct svc_rqst *rqstp, struct svc_fh *fhp, #endif __be32 nfsd4_clone_file_range(struct file *src, u64 src_pos, struct file *dst, - u64 dst_pos, u64 count) + u64 dst_pos, u64 count, bool sync) { loff_t cloned; @@ -534,6 +534,12 @@ __be32 nfsd4_clone_file_range(struct file *src, u64 src_pos, struct file *dst, return nfserrno(cloned); if (count && cloned != count) return nfserrno(-EINVAL); + if (sync) { + loff_t dst_end = count ? dst_pos + count - 1 : LLONG_MAX; + int status = vfs_fsync_range(dst, dst_pos, dst_end, 0); + if (status < 0) + return nfserrno(status); + } return 0; } |