compat_clock_getres shouldn't return -EFAULT if res == NULL

compat_clock_getres shouldn't return -EFAULT if res == NULL

[Arun Sharma]

[-- Attachment #1: Type: text/plain, Size: 41 bytes --]


Description within the patch.

	-Arun



[-- Attachment #2: compat-clock-res.patch --]
[-- Type: text/plain, Size: 765 bytes --]

For clock_getres(clockid_t clock_id, struct timespec *res), the
specification says "If res is NULL, the clock resolution is not
returned." So this kind of call should succeed. The current implementation
returns -EFAULT.

The patch fixes the bug in compat_clock_getres().

Signed-off-by: Gordon Jin <gordon.jin@intel.com>
Signed-off-by: Arun Sharma <arun.sharma@intel.com>

--- linux-2.6.7/kernel/compat.c.orig	2004-07-25 09:02:01.000000000 -0700
+++ linux-2.6.7/kernel/compat.c	2004-07-25 09:04:51.000000000 -0700
@@ -531,7 +531,7 @@
 	err = sys_clock_getres(which_clock,
 			       (struct timespec __user *) &ts);
 	set_fs(oldfs);
-	if (!err && put_compat_timespec(&ts, tp))
+	if (!err && tp && put_compat_timespec(&ts, tp))
 		return -EFAULT; 
 	return err;
 } 

Re: compat_clock_getres shouldn't return -EFAULT if res == NULL

[David S. Miller]

On Mon, 26 Jul 2004 12:51:29 -0700
Arun Sharma <arun.sharma@intel.com> wrote:

> Description within the patch.

Looks good.

Re: compat_clock_getres shouldn't return -EFAULT if res == NULL

[Andrew Morton]

"David S. Miller" <davem@redhat.com> wrote:
>
> On Mon, 26 Jul 2004 12:51:29 -0700
> Arun Sharma <arun.sharma@intel.com> wrote:
> 
> > Description within the patch.
> 
> Looks good.

I'll queue it up.  I seem to have ~200 patches to deal with, mostly
non-trivial :(