lkml.org 
[lkml]   [2012]   [Mar]   [21]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
Patch in this message
/
From
SubjectRe: [PATCH] do not export kernel's NULL #define to userspace
Date
On Wednesday 21 of March 2012, Arnd Bergmann wrote:
> On Wednesday 21 March 2012, Lubos Lunak wrote:
> > GCC's NULL is actually __null, which allows detecting some questionable
> > NULL usage and warn about it. Moreover each platform/compiler should have
> > its own stddef.h anyway (which is different from linux/stddef.h).
> > So there's no good reason to leak kernel's NULL to userspace and
> > override what the compiler provides.
> >
> > Signed-off-by: Luboš Luňák <l.lunak@suse.cz>
>
> Yes, this looks like a bug, but I'm not sure that the solution is
> sufficient. Have you checked the other exported header files
> for whether they use NULL after including linux/stddef.h?

I have checked that the kernel builds with this change, and I have checked
that the whole of LibreOffice compiles after the change. I'm not sure what
you mean exactly, are you asking me to manually inspect everything under
include/linux for NULL usage?

> If so, we might have to replace it with a __KERNEL_NULL constant
> or something, like we do for the stuff in linux/types.h, so we
> don't accidentally break user applications that rely on the
> header files to be self-contained.
>
> I think there is at least a NULL usage in linux/wireless.h and some
> netfilter headers.

I see. How about the attached patch then?

--
Lubos Lunak
l.lunak@suse.cz
From 9be4118ed1bc6221d4f2a73cad69de33285b0b2a Mon Sep 17 00:00:00 2001
From: Lubos Lunak <l.lunak@suse.cz>
Date: Wed, 21 Mar 2012 14:04:31 +0100
Subject: [PATCH] do not export kernel's NULL #define to userspace
MIME-Version: 1.0
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8bit

GCC's NULL is actually __null, which allows detecting some questionable
NULL usage and warn about it. Moreover each platform/compiler should have
its own stddef.h anyway (which is different from linux/stddef.h).
So there's no good reason to leak kernel's NULL to userspace and
override what the compiler provides. Include <stddef.h> in that
case to make sure linux headers are self-contained in userspace as well.

Signed-off-by: Luboš Luňák <l.lunak@suse.cz>
---
include/linux/stddef.h | 10 ++++------
1 files changed, 4 insertions(+), 6 deletions(-)

diff --git a/include/linux/stddef.h b/include/linux/stddef.h
index 6a40c76..1db853b 100644
--- a/include/linux/stddef.h
+++ b/include/linux/stddef.h
@@ -3,14 +3,10 @@

#include <linux/compiler.h>

+#ifdef __KERNEL__
+
#undef NULL
-#if defined(__cplusplus)
-#define NULL 0
-#else
#define NULL ((void *)0)
-#endif
-
-#ifdef __KERNEL__

enum {
false = 0,
@@ -23,6 +19,8 @@ enum {
#else
#define offsetof(TYPE, MEMBER) ((size_t) &((TYPE *)0)->MEMBER)
#endif
+#else /* __KERNEL__ */
+#include <stddef.h>
#endif /* __KERNEL__ */

#endif
--
1.7.3.4
\
 
 \ /
  Last update: 2012-03-21 15:11    [W:0.586 / U:0.036 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site