[<prev] [next>] [day] [month] [year] [list]
Message-ID: <d10d318c-7461-96fb-952b-c246811a8228@gmail.com>
Date: Thu, 27 Apr 2017 13:54:11 +0200
From: "Michael Kerrisk (man-pages)" <mtk.manpages@...il.com>
To: Linus Torvalds <torvalds@...ux-foundation.org>
Cc: mtk.manpages@...il.com, David Howells <dhowells@...hat.com>,
Al Viro <viro@...iv.linux.org.uk>,
linux-man <linux-man@...r.kernel.org>,
Linux API <linux-api@...r.kernel.org>,
Eric Sandeen <sandeen@...deen.net>,
Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
fstests <fstests@...r.kernel.org>,
linux-fsdevel <linux-fsdevel@...r.kernel.org>,
Michael Kerrisk <mtk.manpages@...il.com>, hch@...radead.org
Subject: [PATCH] statx: correct error handling of NULL pathname
The change in 1e2f82d1e9d1 to error on a NULL pathname to statx()
is inconsistent. It results in the error EINVAL for a NULL pathname.
Consistent with similar APIs (fchownat(), fstatat(), linkat()),
the error should be EFAULT.
The solution is simply to remove the EINVAL check. As I already
pointed out in https://lkml.org/lkml/2017/4/26/561, user_path_at*()
and filename_lookup() will handle the NULL pathname as per
the other APIs, to correctly produce the error EFAULT.
Signed-off-by: Michael Kerrisk <mtk.manpages@...il.com>
---
fs/stat.c | 2 --
1 file changed, 2 deletions(-)
diff --git a/fs/stat.c b/fs/stat.c
index 3d85747..a257b87 100644
--- a/fs/stat.c
+++ b/fs/stat.c
@@ -567,8 +567,6 @@ SYSCALL_DEFINE5(statx,
return -EINVAL;
if ((flags & AT_STATX_SYNC_TYPE) == AT_STATX_SYNC_TYPE)
return -EINVAL;
- if (!filename)
- return -EINVAL;
error = vfs_statx(dfd, filename, flags, &stat, mask);
if (error)
--
2.5.5
--
Michael Kerrisk
Linux man-pages maintainer; http://www.kernel.org/doc/man-pages/
Linux/UNIX System Programming Training: http://man7.org/training/
Powered by blists - more mailing lists