From: Takashi Iwai <tiwai@suse.de>
To: Andy Lutomirski <luto@amacapital.net>
Cc: Denys Vlasenko <dvlasenk@redhat.com>,
Denys Vlasenko <vda.linux@googlemail.com>,
Jiri Kosina <jkosina@suse.cz>,
Linus Torvalds <torvalds@linux-foundation.org>,
Stefan Seyfried <stefan.seyfried@googlemail.com>,
X86 ML <x86@kernel.org>, LKML <linux-kernel@vger.kernel.org>,
Tejun Heo <tj@kernel.org>
Subject: Re: PANIC: double fault, error_code: 0x0 in 4.0.0-rc3-2, kvm related?
Date: Mon, 23 Mar 2015 19:56:43 +0100 [thread overview]
Message-ID: <s5hlhinlez8.wl-tiwai@suse.de> (raw)
In-Reply-To: <CALCETrWmz2JoWJptt-YxVszJX0J8m+OhQPqiXRJsE460tXbYNg@mail.gmail.com>
At Mon, 23 Mar 2015 11:38:30 -0700,
Andy Lutomirski wrote:
>
> On Mon, Mar 23, 2015 at 9:07 AM, Denys Vlasenko <dvlasenk@redhat.com> wrote:
> > On 03/23/2015 02:22 PM, Takashi Iwai wrote:
> >> At Mon, 23 Mar 2015 10:35:41 +0100,
> >> Takashi Iwai wrote:
> >>>
> >>> At Mon, 23 Mar 2015 10:02:52 +0100,
> >>> Takashi Iwai wrote:
> >>>>
> >>>> At Fri, 20 Mar 2015 19:16:53 +0100,
> >>>> Denys Vlasenko wrote:
>
> >> I'm really puzzled now. We have a few pieces of information:
> >>
> >> - git bisection pointed the commit 96b6352c1271:
> >> x86_64, entry: Remove the syscall exit audit and schedule optimizations
> >> and reverting this "fixes" the problem indeed. Even just moving two
> >> lines
> >> LOCKDEP_SYS_EXIT
> >> DISABLE_INTERRUPTS(CLBR_NONE)
> >> at the beginning of ret_from_sys_call already fixes. (Of course I
> >> can't prove the fix but it stabilizes for a day without crash while
> >> usually I hit the bug in 10 minutes in full test running.)
> >
> > The commit 96b6352c1271 moved TIF_ALLWORK_MASK check from
> > interrupt-disabled region to interrupt-enabled:
> >
> > cmpq $__NR_syscall_max,%rax
> > ja ret_from_sys_call
> > movq %r10,%rcx
> > call *sys_call_table(,%rax,8) # XXX: rip relative
> > movq %rax,RAX-ARGOFFSET(%rsp)
> > ret_from_sys_call:
> > testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET)
> > ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
> > jnz int_ret_from_sys_call_fixup /* Go the the slow path */
> > LOCKDEP_SYS_EXIT
> > DISABLE_INTERRUPTS(CLBR_NONE)
> > TRACE_IRQS_OFF
> > ...
> > ...
> > int_ret_from_sys_call_fixup:
> > FIXUP_TOP_OF_STACK %r11, -ARGOFFSET
> > jmp int_ret_from_sys_call
> > ...
> > ...
> > GLOBAL(int_ret_from_sys_call)
> > DISABLE_INTERRUPTS(CLBR_NONE)
> > TRACE_IRQS_OFF
> >
> > You reverted that by moving this insn to be after first DISABLE_INTERRUPTS(CLBR_NONE).
> >
> > I also don't see how moving that check (even if it is wrong in a more
> > benign way) can have such a drastic effect.
>
> I bet I see it. I have the advantage of having stared at KVM code and
> cursed at it more recently than you, I suspect. KVM does awful, awful
> things to CPU state, and, as an optimization, it allows kernel code to
> run with CPU state that would be totally invalid in user mode. This
> happens through a bunch of hooks, including this bit in __switch_to:
>
> /*
> * Now maybe reload the debug registers and handle I/O bitmaps
> */
> if (unlikely(task_thread_info(next_p)->flags & _TIF_WORK_CTXSW_NEXT ||
> task_thread_info(prev_p)->flags & _TIF_WORK_CTXSW_PREV))
> __switch_to_xtra(prev_p, next_p, tss);
>
> IOW, we *change* tif during context switches.
>
>
> The race looks like this:
>
> testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP)
> jnz int_ret_from_sys_call_fixup /* Go the the slow path */
>
> --- preempted here, switch to KVM guest ---
>
> KVM guest enters and screws up, say, MSR_SYSCALL_MASK. This wouldn't
> happen to be a *32-bit* KVM guest, perhaps?
>
> Now KVM schedules, calling __switch_to. __switch_to sets
> _TIF_USER_RETURN_NOTIFY. We IRET back to the syscall exit code, turn
> off interrupts, and do sysret. We are now screwed.
Thanks for enlightening! That looks like a feasible scenario.
(I tested only a 64bit KVM guest, BTW.)
> I don't know why this manifests in this particular failure, but any
> number of terrible things could happen now.
>
> FWIW, this will affect things other than KVM. For example, SIGKILL
> sent while a process is sleeping in that two-instruction window won't
> work.
>
> Takashi, can you re-send your patch so we can review it for real in
> light of this race?
The patch below worked. I'll double-check tomorrow whether this
really cures reliably.
thanks,
Takashi
diff --git a/arch/x86/kernel/entry_64.S b/arch/x86/kernel/entry_64.S
index 1d74d161687c..5340ac7f88a9 100644
--- a/arch/x86/kernel/entry_64.S
+++ b/arch/x86/kernel/entry_64.S
@@ -364,12 +364,12 @@ system_call_fastpath:
* Has incomplete stack frame and undefined top of stack.
*/
ret_from_sys_call:
- testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET)
- jnz int_ret_from_sys_call_fixup /* Go the the slow path */
-
LOCKDEP_SYS_EXIT
DISABLE_INTERRUPTS(CLBR_NONE)
TRACE_IRQS_OFF
+ testl $_TIF_ALLWORK_MASK,TI_flags+THREAD_INFO(%rsp,RIP-ARGOFFSET)
+ jnz int_ret_from_sys_call_fixup /* Go the the slow path */
+
CFI_REMEMBER_STATE
/*
* sysretq will re-enable interrupts:
next prev parent reply other threads:[~2015-03-23 18:56 UTC|newest]
Thread overview: 77+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-15 8:17 Stefan Seyfried
2015-03-18 14:16 ` Takashi Iwai
2015-03-18 15:05 ` Takashi Iwai
2015-03-18 17:43 ` Takashi Iwai
2015-03-18 17:46 ` Takashi Iwai
2015-03-18 18:03 ` Andy Lutomirski
2015-03-18 19:03 ` Stefan Seyfried
2015-03-18 19:26 ` Andy Lutomirski
2015-03-18 20:05 ` Stefan Seyfried
2015-03-18 20:51 ` Andy Lutomirski
2015-03-18 21:12 ` Stefan Seyfried
2015-03-18 21:21 ` Andy Lutomirski
2015-03-18 21:41 ` Stefan Seyfried
2015-03-18 21:49 ` Denys Vlasenko
2015-03-18 21:53 ` Stefan Seyfried
2015-03-18 20:06 ` Denys Vlasenko
2015-03-18 20:49 ` Andy Lutomirski
2015-03-18 21:06 ` Denys Vlasenko
2015-03-18 21:17 ` Andy Lutomirski
2015-03-18 21:32 ` Linus Torvalds
2015-03-18 21:42 ` Denys Vlasenko
2015-03-18 21:55 ` Andy Lutomirski
2015-03-18 22:17 ` Denys Vlasenko
2015-03-18 22:20 ` Andy Lutomirski
2015-03-18 22:27 ` Denys Vlasenko
2015-03-18 22:18 ` Linus Torvalds
2015-03-18 22:24 ` Andy Lutomirski
2015-03-18 22:22 ` Jiri Kosina
2015-03-18 22:28 ` Linus Torvalds
2015-03-18 22:29 ` Andy Lutomirski
2015-03-18 22:29 ` Andy Lutomirski
2015-03-18 22:38 ` Stefan Seyfried
2015-03-18 22:40 ` Andy Lutomirski
2015-03-18 23:22 ` Andy Lutomirski
2015-03-19 0:23 ` Stefan Seyfried
2015-03-19 0:57 ` Andy Lutomirski
2015-03-19 2:15 ` Linus Torvalds
2015-03-19 6:24 ` Stefan Seyfried
2015-03-19 10:16 ` Takashi Iwai
2015-03-19 10:58 ` Denys Vlasenko
2015-03-19 11:21 ` Takashi Iwai
2015-03-19 12:48 ` Denys Vlasenko
2015-03-19 13:47 ` Takashi Iwai
2015-03-19 14:55 ` Takashi Iwai
2015-03-19 15:22 ` Takashi Iwai
2015-03-19 15:41 ` Andy Lutomirski
2015-03-19 15:51 ` Takashi Iwai
2015-03-19 16:01 ` Andy Lutomirski
2015-03-20 18:16 ` Denys Vlasenko
2015-03-20 18:50 ` Takashi Iwai
2015-03-23 9:02 ` Takashi Iwai
2015-03-23 9:35 ` Takashi Iwai
2015-03-23 13:22 ` Takashi Iwai
2015-03-23 16:07 ` Denys Vlasenko
2015-03-23 17:18 ` Takashi Iwai
2015-03-23 17:46 ` Denys Vlasenko
2015-03-23 18:43 ` Takashi Iwai
2015-03-23 18:38 ` Andy Lutomirski
2015-03-23 18:48 ` Andy Lutomirski
2015-03-23 18:59 ` Takashi Iwai
2015-03-23 19:10 ` [PATCH] x86, entry: Check for syscall exit work with IRQs disabled Andy Lutomirski
2015-03-23 19:21 ` Denys Vlasenko
2015-03-23 19:27 ` Andy Lutomirski
2015-03-23 19:32 ` Andy Lutomirski
2015-03-24 11:17 ` Takashi Iwai
2015-03-24 20:08 ` Ingo Molnar
2015-03-25 0:35 ` Andy Lutomirski
2015-03-25 12:21 ` Ingo Molnar
2015-03-25 15:07 ` Andy Lutomirski
2015-03-25 9:13 ` [tip:x86/asm] x86/asm/entry: " tip-bot for Andy Lutomirski
2015-03-23 18:54 ` PANIC: double fault, error_code: 0x0 in 4.0.0-rc3-2, kvm related? Stefan Seyfried
2015-03-23 18:56 ` Takashi Iwai [this message]
2015-03-23 19:07 ` Denys Vlasenko
2015-03-23 19:10 ` Andy Lutomirski
2015-03-19 13:21 ` Denys Vlasenko
2015-03-18 21:49 ` Stefan Seyfried
2015-03-28 23:57 ` Maciej W. Rozycki
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=s5hlhinlez8.wl-tiwai@suse.de \
--to=tiwai@suse.de \
--cc=dvlasenk@redhat.com \
--cc=jkosina@suse.cz \
--cc=linux-kernel@vger.kernel.org \
--cc=luto@amacapital.net \
--cc=stefan.seyfried@googlemail.com \
--cc=tj@kernel.org \
--cc=torvalds@linux-foundation.org \
--cc=vda.linux@googlemail.com \
--cc=x86@kernel.org \
/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
Powered by JetHome