From: Frederik Deweerdt <deweerdt@free.fr>
To: Eric Sesterhenn <snakebyte@gmx.de>
Cc: linux-kernel@vger.kernel.org, suresh.b.siddha@intel.com
Subject: Re: Oops with strace_test
Date: Mon, 5 May 2008 20:00:53 +0200 [thread overview]
Message-ID: <20080505180053.GB5576@slug> (raw)
In-Reply-To: <20080505100008.GA2354@alice>
On Mon, May 05, 2008 at 12:00:08PM +0200, Eric Sesterhenn wrote:
> * Eric Sesterhenn (snakebyte@gmx.de) wrote:
> > hi,
> >
> > running the strace_test from ltp 20080229 (ltp.sf.net) gives me
> > two different oopses, so far i was not able to pinpoint to a specific
> > testcase (propably because the strace uses the rng to decided what fails
> > and what not) one oops is in iret_exc(), the other in __copy_from_user_ll()
> > The oopses dont happen with 2.6.24 so this appears to be a regression, i am starting
> > a git-bisect, but this might take some time
> >
[...]
>
> after some bisecting i found commit
> http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff_plain;h=aa283f49276e7d840a40fb01eee6de97eaa7e012;hp=61c4628b538608c1a85211ed8438136adfeb9a95
> to be guilty. After reverting this manually (didnt revert cleanly)
Hi Eric,
This appears to be caused by init_fpu() missing from the
restore_sigcontext->restore_i387->restore_fpu_checking code path.
I believe that moving the init_fpu() call from math_state_restore to
restore_fpu_checking should fix the problem?
Regards,
Frederik
Signed-off-by: Frederik Deweerdt <frederik.deweerdt@gmail.com>
diff --git a/arch/x86/kernel/traps_64.c b/arch/x86/kernel/traps_64.c
index 8069073..5b1af48 100644
--- a/arch/x86/kernel/traps_64.c
+++ b/arch/x86/kernel/traps_64.c
@@ -1142,22 +1142,6 @@ asmlinkage void math_state_restore(void)
{
struct task_struct *me = current;
- if (!used_math()) {
- local_irq_enable();
- /*
- * does a slab alloc which can sleep
- */
- if (init_fpu(me)) {
- /*
- * ran out of memory!
- */
- do_group_exit(SIGKILL);
- return;
- }
- local_irq_disable();
- }
-
- clts(); /* Allow maths ops (or we recurse) */
restore_fpu_checking(&me->thread.xstate->fxsave);
task_thread_info(me)->status |= TS_USEDFPU;
me->fpu_counter++;
diff --git a/include/asm-x86/i387.h b/include/asm-x86/i387.h
index da2adb4..bf1cabe 100644
--- a/include/asm-x86/i387.h
+++ b/include/asm-x86/i387.h
@@ -47,7 +47,20 @@ static inline void tolerant_fwait(void)
static inline int restore_fpu_checking(struct i387_fxsave_struct *fx)
{
- int err;
+ int err = -1;
+
+ if (!used_math()) {
+ local_irq_enable();
+ /*
+ * does a slab alloc which can sleep
+ */
+ if (init_fpu(current))
+ return err;
+ local_irq_disable();
+ }
+
+ clts(); /* Allow maths ops (or we recurse) */
asm volatile("1: rex64/fxrstor (%[fx])\n\t"
"2:\n"
next prev parent reply other threads:[~2008-05-05 18:01 UTC|newest]
Thread overview: 14+ messages / expand[flat|nested] mbox.gz Atom feed top
2008-04-30 12:12 Oops with strace_test Eric Sesterhenn
2008-04-30 16:19 ` Andi Kleen
2008-05-05 10:00 ` Eric Sesterhenn
2008-05-05 18:00 ` Frederik Deweerdt [this message]
2008-05-06 12:10 ` Ingo Molnar
2008-05-06 15:07 ` Eric Sesterhenn
2008-05-07 10:36 ` Eric Sesterhenn
2008-05-07 13:36 ` Thomas Gleixner
2008-05-07 17:20 ` Eric Sesterhenn
2008-05-07 17:26 ` Thomas Gleixner
2008-05-07 19:09 ` Suresh Siddha
2008-05-07 21:46 ` Eric Sesterhenn
2008-05-08 15:02 ` Thomas Gleixner
2008-05-08 18:29 ` Eric Sesterhenn
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20080505180053.GB5576@slug \
--to=deweerdt@free.fr \
--cc=linux-kernel@vger.kernel.org \
--cc=snakebyte@gmx.de \
--cc=suresh.b.siddha@intel.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox