On Fri, Jan 27, 2006 at 04:16:41PM +0100, Martin Koniczek wrote:
>
> just an update to my previous post:
>
> >the following was recently observed at least on the following
> >x86 platforms:
> >
> >2.6.15-vs2.1.0.5 (gcc 3.4.4)
> >2.6.16-rc1-vs2.1.0.7 (gcc 4.0.3)
> 2.6.15-1.vs2.1.0.5 (gcc 3.4.4)
>
>
> >it does NOT happen under
> >2.6.14-vs2.1.0 (gcc 3.4.4)
> 2.6.15-vs2.1.0.2 (gcc 3.4.4)
> 2.6.15-vs2.1.0.4 (gcc 3.4.4)
>
>
> i looked into the differences between
> patch-2.6.15-vs2.1.0.4.diff
> patch-2.6.15-vs2.1.0.5.diff
>
> where interdiff produced only 672 lines:
> http://pastebin.com/525691
>
> but i did not ever work on the vserver kernel patches before,
> so right now i am just lost.
> hope this "prework" helps the real hackers,
you did excellent work, thanks a lot ...
(added you to our Hall of Fame)
the fix is simple:
--- linux-2.6.16-rc1-vs2.1.0.7.2/include/linux/vserver/cvirt_def.h 2006-01-27 20:35:10 +0100
+++ linux-2.6.16-rc1-vs2.1.0.7.3/include/linux/vserver/cvirt_def.h 2006-01-27 22:17:15 +0100
@@ -92,7 +92,7 @@ struct _vx_sock_acc {
struct _vx_cacct {
struct _vx_sock_acc sock[5][3];
atomic_t slab[8];
- atomic_t page[5][8];
+ atomic_t page[6][8];
};
#ifdef CONFIG_VSERVER_DEBUG
the patch is here:
http://vserver.13thfloor.at/Devel/PAT-2.1.1/detla-pfault-fix01.diff
the next experimental release vs2.1.0.8 will
already contain it ...
best,
Herbert
> martin aka cohan
> _______________________________________________
> Vserver mailing list
> Vserver@list.linux-vserver.org
> http://list.linux-vserver.org/mailman/listinfo/vserver
_______________________________________________
Vserver mailing list
Vserver@list.linux-vserver.org
http://list.linux-vserver.org/mailman/listinfo/vserver
Received on Fri Jan 27 21:40:41 2006