5.10-stable review patch. If anyone has any objections, please let me know.
------------------
From: NeilBrown neilb@suse.de
[ Upstream commit 00801cd92d91e94aa04d687f9bb9a9104e7c3d46 ]
A recent patch moved ACL setting into nfsd_setattr(). Unfortunately it didn't work as nfsd_setattr() aborts early if iap->ia_valid is 0.
Remove this test, and instead avoid calling notify_change() when ia_valid is 0.
This means that nfsd_setattr() will now *always* lock the inode. Previously it didn't if only a ATTR_MODE change was requested on a symlink (see Commit 15b7a1b86d66 ("[PATCH] knfsd: fix setattr-on-symlink error return")). I don't think this change really matters.
Fixes: c0cbe70742f4 ("NFSD: add posix ACLs to struct nfsd_attrs") Signed-off-by: NeilBrown neilb@suse.de Reviewed-by: Jeff Layton jlayton@kernel.org [ cel: backported to 5.10.y, prior to idmapped mounts ] Signed-off-by: Chuck Lever chuck.lever@oracle.com Signed-off-by: Sasha Levin sashal@kernel.org --- fs/nfsd/vfs.c | 19 +++++++++---------- 1 file changed, 9 insertions(+), 10 deletions(-)
diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c index 5a7fee4ee2079..95f2e4549c034 100644 --- a/fs/nfsd/vfs.c +++ b/fs/nfsd/vfs.c @@ -299,6 +299,10 @@ commit_metadata(struct svc_fh *fhp) static void nfsd_sanitize_attrs(struct inode *inode, struct iattr *iap) { + /* Ignore mode updates on symlinks */ + if (S_ISLNK(inode->i_mode)) + iap->ia_valid &= ~ATTR_MODE; + /* sanitize the mode change */ if (iap->ia_valid & ATTR_MODE) { iap->ia_mode &= S_IALLUGO; @@ -366,7 +370,7 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, int accmode = NFSD_MAY_SATTR; umode_t ftype = 0; __be32 err; - int host_err; + int host_err = 0; bool get_write_count; bool size_change = (iap->ia_valid & ATTR_SIZE);
@@ -404,13 +408,6 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, dentry = fhp->fh_dentry; inode = d_inode(dentry);
- /* Ignore any mode updates on symlinks */ - if (S_ISLNK(inode->i_mode)) - iap->ia_valid &= ~ATTR_MODE; - - if (!iap->ia_valid) - return 0; - nfsd_sanitize_attrs(inode, iap);
if (check_guard && guardtime != inode->i_ctime.tv_sec) @@ -461,8 +458,10 @@ nfsd_setattr(struct svc_rqst *rqstp, struct svc_fh *fhp, goto out_unlock; }
- iap->ia_valid |= ATTR_CTIME; - host_err = notify_change(dentry, iap, NULL); + if (iap->ia_valid) { + iap->ia_valid |= ATTR_CTIME; + host_err = notify_change(dentry, iap, NULL); + }
out_unlock: if (attr->na_seclabel && attr->na_seclabel->len)