make nfs_atomic_open() call d_drop() on all ->open_context() errors.

Al Viro · amschuma-ntap · commit d20cb71dbf34 · 2016-06-27T08:59:08.000-04:00
In "NFSv4: Move dentry instantiation into the NFSv4-specific atomic open code"
unconditional d_drop() after the -&gt;open_context() had been removed.  It had
been correct for success cases (there -&gt;open_context() itself had been doing
dcache manipulations), but not for error ones.  Only one of those (ENOENT)
got a compensatory d_drop() added in that commit, but in fact it should've
been done for all errors.  As it is, the case of O_CREAT non-exclusive open
on a hashed negative dentry racing with e.g. symlink creation from another
client ended up with -&gt;open_context() getting an error and proceeding to
call nfs_lookup().  On a hashed dentry, which would've instantly triggered
BUG_ON() in d_materialise_unique() (or, these days, its equivalent in
d_splice_alias()).

Cc: stable@vger.kernel.org # v3.10+
Tested-by: Oleg Drokin &lt;green@linuxhacker.ru&gt;
Signed-off-by: Al Viro &lt;viro@zeniv.linux.org.uk&gt;
Signed-off-by: Trond Myklebust &lt;trond.myklebust@primarydata.com&gt;
Signed-off-by: Anna Schumaker &lt;Anna.Schumaker@Netapp.com&gt;
diff --git a/fs/nfs/dir.c b/fs/nfs/dir.c
@@ -1539,9 +1539,9 @@ int nfs_atomic_open(struct inode *dir, struct dentry *dentry,
 		err = PTR_ERR(inode);
 		trace_nfs_atomic_open_exit(dir, ctx, open_flags, err);
 		put_nfs_open_context(ctx);
+		d_drop(dentry);
 		switch (err) {
 		case -ENOENT:
-			d_drop(dentry);
 			d_add(dentry, NULL);
 			nfs_set_verifier(dentry, nfs_save_change_attribute(dir));
 			break;
