Hi Linus,
As expected (reported by David Fries), today's linux-next build (powerpc
ppc64_defconfig) failed like this:
In file included from include/linux/tracehook.h:52,
from arch/powerpc/kernel/signal.c:13:
include/linux/security.h: In function 'security_ptrace_traceme':
include/linux/security.h:1760: error: 'parent' undeclared (first use in this function)
Caused by commit 5cd9c58fbe9ec92b45b27e131719af4f2bd9eb40 ("security: Fix
setting of PF_SUPERPRIV by __capable()"). The version of this commit
that has been in the linux-next tree from Aug 8 to Aug 15 is correct, so
I have no idea what happened.
More care needed ...
I applied the following patch from David Fries.
--
Cheers,
Stephen Rothwell [email protected]
http://www.canb.auug.org.au/~sfr/
Date: Sat, 16 Aug 2008 20:17:11 -0500
From: David Fries <[email protected]>
To: David Howells <[email protected]>
Cc: [email protected], [email protected], [email protected],
[email protected]
Subject: Fix setting of PF_SUPERPRIV by __capable()
The first two submissions this patch had parent for the argument, the third
one has child then parent and fails to compile with SECURITY disabled.
Please apply the following patch.
> +static inline int security_ptrace_traceme(struct task_struct *child)
> {
> - return cap_ptrace(parent, child, mode);
> + return cap_ptrace_traceme(parent);
> }
security_ptrace_traceme error: 'parent' undeclared
s/child/parent/
Signed-off-by: David Fries <[email protected]>
diff --git a/include/linux/security.h b/include/linux/security.h
index 2ee5ecf..80c4d00 100644
--- a/include/linux/security.h
+++ b/include/linux/security.h
@@ -1755,7 +1755,7 @@ static inline int security_ptrace_may_access(struct task_struct *child,
return cap_ptrace_may_access(child, mode);
}
-static inline int security_ptrace_traceme(struct task_struct *child)
+static inline int security_ptrace_traceme(struct task_struct *parent)
{
return cap_ptrace_traceme(parent);
}
Stephen Rothwell <[email protected]> wrote:
> As expected (reported by David Fries), today's linux-next build (powerpc
> ppc64_defconfig) failed like this:
>
> In file included from include/linux/tracehook.h:52,
> from arch/powerpc/kernel/signal.c:13:
> include/linux/security.h: In function 'security_ptrace_traceme':
> include/linux/security.h:1760: error: 'parent' undeclared (first use in this function)
>
> Caused by commit 5cd9c58fbe9ec92b45b27e131719af4f2bd9eb40 ("security: Fix
> setting of PF_SUPERPRIV by __capable()"). The version of this commit
> that has been in the linux-next tree from Aug 8 to Aug 15 is correct, so
> I have no idea what happened.
I accidentally posted an old patch which James, maybe, substituted for the
good one that he had.
David
On Mon, 18 Aug 2008, David Howells wrote:
> Stephen Rothwell <[email protected]> wrote:
>
> > As expected (reported by David Fries), today's linux-next build (powerpc
> > ppc64_defconfig) failed like this:
> >
> > In file included from include/linux/tracehook.h:52,
> > from arch/powerpc/kernel/signal.c:13:
> > include/linux/security.h: In function 'security_ptrace_traceme':
> > include/linux/security.h:1760: error: 'parent' undeclared (first use in this function)
> >
> > Caused by commit 5cd9c58fbe9ec92b45b27e131719af4f2bd9eb40 ("security: Fix
> > setting of PF_SUPERPRIV by __capable()"). The version of this commit
> > that has been in the linux-next tree from Aug 8 to Aug 15 is correct, so
> > I have no idea what happened.
>
> I accidentally posted an old patch which James, maybe, substituted for the
> good one that he had.
I took the patch you sent me, although after some IRC discussion, what
should have happened is that I should have git-cherry-pick'd the patch
from the next-creds branch (where it had been tested in linux-next for
some time) into for-linus rather than asking for an updated patch.
Sorry folks.
- James
--
James Morris
<[email protected]>