Re: [PATCH] perf: fix kernel panic when parsing user space CS saved in pt_regs

From: Liu ShuoX
Date: Thu Jun 05 2014 - 03:34:08 EST


On Thu 5.Jun'14 at 9:19:19 +0200, Peter Zijlstra wrote:
On Thu, Jun 05, 2014 at 10:36:10AM +0800, Liu ShuoX wrote:
From: Zhang Yanmin <yanmin.zhang@xxxxxxxxx>

We hit a kernel panic when running perf to collect some performance data.
kenel is x86_64 and user space apps are 32bit.

[ 71.965351, 1] [ Binder_2] BUG: unable to handle kernel NULL pointer dereference at 0000000000000004
[ 71.965360, 1] [ Binder_2] IP: [<ffffffff82012091>] get_segment_base+0x71/0xc0
[ 71.965367, 1] [ Binder_2] PGD 6c65f067 PUD 0
[ 71.965375, 1] [ Binder_2] Oops: 0000 [#1] PREEMPT SMP
[ 71.965413, 1] [ Binder_2] Modules linked in: ddrgx snd_merr_dpcm_wm8958 snd_intel_sst snd_soc_sst_platform snd_soc_wm8994 snd_soc_wm_hubs lm3559 imx1x5 atomisp_css2401a0_v21 libmsrlisthelper rmi4 bcm_bt_lpm videobuf_vmalloc videobuf_core fps_throttle hdmi_audio pn544(O) tngdisp bcm4335(O) cfg80211
[ 71.965420, 1] [ Binder_2] CPU: 1 PID: 304 Comm: Binder_2 Tainted: G W O 3.10.20-263902-g184bfbc-dirty #14
[ 71.965426, 1] [ Binder_2] task: ffff8800764dc300 ti: ffff88006c6e8000 task.ti: ffff88006c6e8000
[ 71.965439, 1] [ Binder_2] RIP: 0010:[<ffffffff82012091>] [<ffffff?f82012091>] get_segment_base+0x71/0xc0
^
[ 71.965<44, 1] [ Binder_2] RSP: 0018:ffff^X8007ea87b98 EFLAGS: 00010092
^ ^
[ 71.965447, 1] [ !Binder_2] RAX: 0000000000000024 RBX: 0000000000000000 RCX: 0000000000000000
^
[ 71.965450, 1] [ Binder_2] RDX: 0000000000000000 RSI: 0000000000000000 RDI: 0000000000000009
[ 71.965454, 1] [ Binder_2] RBP: ffff88007ea87ba8 R08: ffffffff83143b3c R09: ffffffff831848a8
[ 71.965458, 1] [ Binder_2] R10: 0000000000000000 R11: 00000000001bf2d8 R12: 0000000000000000
[ 71.965462, 1] [ Binder_2] R13: ffff88006c6e9fd8 R14: ffff88006c6e9f58 R15: ffff8800764dc300
[ 71.965468, 1_ [ Binder_2] FS: 0000000000000000(0000) GS:ffff88007ea80000(006b) knlGS:00000000f704add0
^

Are you suffering some serious corruption?
The log captured by pstore after rebooting, so there are some
corruption. Please ignore those.

Basically, ia32 uses sysenter to start system calls.

sysexit_from_sys_call=>trace_hardirqs_on_thunk. Before calling,
sysexit_from_sys_call already pops up pt_regs, then trace_hardirqs_on_thunk
would reuse pt_regs space. If perf NMI happens here, perf might use a bad pt_regs.

The patch fixes it by moving the calling to trace_hardirqs_on_thunk ahead of
the stack popup.

Change-Id: I6c4fc46b009ea056f2321ce5b8f54cf8769a7bdd

No idea what that is, but it needs to go.

I'll leave the actual patch to hpa, this isn't something I'm too
familiar with.
Thanks. I will send out v2. We changed the code in v2.
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/