Bug 32593

Summary:	[libva]when play an interlaced media file and seek, it will cause GPU hang
Product:	xorg	Reporter:	Hai <hai.lan>
Component:	Driver/intel	Assignee:	haihao <haihao.xiang>
Status:	CLOSED DUPLICATE	QA Contact:	Hai <hai.lan>
Severity:	normal
Priority:	medium
Version:	unspecified
Hardware:	x86 (IA32)
OS:	Linux (All)
Whiteboard:
i915 platform:		i915 features:

Description Hai 2010-12-22 17:57:10 UTC

System Environment:
--------------------------
Libdrm:		(master)2.4.23-4-gbad5242a59aa8e31cf10749e2ac69b3c66ef7da0
Mesa:		(7.10)612e26e82c60809ab9c81213f39fbe527284af9f
Xserver:	(server-1.9-branch)xorg-server-1.9.3-6-ga2c674b75d3b1a663bece6f6e2e34c3e71e6be55
Xf86_video_intel:	(master)2.13.902-5-g1ac2e04023f84dbf1f3db2ecad1cadd159aa614d
Cairo:		(master)cf0a4ed862d93b7e6d7d26942cfe98a50beccd4c
Kernel:	(drm-intel-next) 8d5203ca62539c6ab36a5bc2402c2de1de460e30
CPU: SNB

Bug detailed description:
-------------------------
when play an interlaced media file(TS1080i25_h264cab_10mbs_sally-0xe0.264) and seek, it will casue GPU hang

Following is the tail of /var/log/messages
Dec 22 16:24:01 x-hnr1 NetworkManager[2392]: <info> Activation (eth0) successful, device activated.
Dec 22 16:24:01 x-hnr1 NetworkManager[2392]: <info> Activation (eth0) Stage 5 of 5 (IP Configure Commit) complete.
Dec 22 16:24:13 x-hnr1 acpid: client connected from 2907[0:0]
Dec 22 16:24:13 x-hnr1 acpid: 1 client rule loaded
Dec 22 16:25:51 x-hnr1 kernel: EXT3-fs: barriers not enabled
Dec 22 16:25:52 x-hnr1 kernel: kjournald starting.  Commit interval 5 seconds
Dec 22 16:25:52 x-hnr1 kernel: EXT3-fs (sdb6): using internal journal
Dec 22 16:25:52 x-hnr1 kernel: EXT3-fs (sdb6): recovery complete
Dec 22 16:25:52 x-hnr1 kernel: EXT3-fs (sdb6): mounted filesystem with ordered data mode
Dec 22 16:27:11 x-hnr1 kernel: [drm:i915_hangcheck_elapsed] *ERROR* Hangcheck timer elapsed... GPU idle, missed IRQ.

Following is the mplayer stack 
[<f8406f20>] i915_do_wait_request+0x20d/0x335 [i915]
[<f840707d>] i915_gem_object_wait_rendering+0x35/0x3d [i915]
[<f840ac56>] i915_gem_do_execbuffer+0x861/0xb58 [i915]
[<f840b03f>] i915_gem_execbuffer2+0xf2/0x175 [i915]
[<f814b20c>] drm_ioctl+0x268/0x322 [drm]
[<c02c691e>] do_vfs_ioctl+0x4fa/0x53f
[<c02c69a4>] sys_ioctl+0x41/0x61
[<c020280c>] sysenter_do_call+0x12/0x22
[<ffffffff>] 0xffffffff

Reproduce steps:
-------------------------
mplayer -vo vaapi -va vaapi -fps 25 TS1080i25_h264cab_10mbs_sally-0xe0.264

Comment 1 haihao 2010-12-23 20:51:50 UTC

Is it an IRQ stall issue?

Comment 2 Hai 2010-12-27 17:05:23 UTC

Yes I think so.

Comment 3 Hai 2010-12-28 01:33:43 UTC

For Ironlake, this issue won't happen.

Comment 4 haihao 2010-12-28 18:48:35 UTC

Does kernel option nomsi or idle=poll work for you?

Comment 5 Hai 2010-12-28 19:35:27 UTC

I have tried both "pci=nomsi" and "idle=poll". This issue still happens.
The detailed environment is 
CPU:Intel(R) Core(TM) i7-QM CPU @ 2.40GHz
CPU Stepping:6
Graphics revision:08
Platform:Huron River
BIOS Version: ASNBCPT1.86C.0035.B00.1007291540
Release Date: 07/29/2010

Comment 6 haihao 2010-12-29 18:11:38 UTC

Can you reproduce this issue on another SNB machine?

Comment 7 Hai 2010-12-29 18:44:05 UTC

It's OK on another SNB( platform is sugarbay and GFX revision is 09)

Comment 8 haihao 2010-12-29 19:12:14 UTC


*** This bug has been marked as a duplicate of bug 32694 ***

Use of freedesktop.org services, including Bugzilla, is subject to our Code of Conduct. How we collect and use information is described in our Privacy Policy.