Hello,
I tested 2.6.36 on an Asus laptop with an Intel GM965 chipset. I first
noticed the oops when i tried to run Virtualbox, which links to
opengl, and i tried reproducing it with glxgears, which it
consistently does.
I also noticed an earlier email on this oops, from august 16, 2010,
that didn't get any replies.
http://kerneltrap.org/mailarchive/linux-kernel/2010/8/16/4607090
Here's output from syslog:
Oct 27 21:49:02 subgenius kernel: Oops: 0000 [#1] PREEMPT SMP
Oct 27 21:49:02 subgenius kernel: last sysfs file:
/sys/devices/platform/coretemp.1/temp1_input
Oct 27 21:49:02 subgenius kernel: Modules linked in: vboxnetflt
vboxdrv coretemp hwmon usbhid usb_storage usb_libusual uhci_hcd
ehci_hcd usbcore snd_hda_codec_realtek snd_hda_intel snd_hda_codec
snd_pcm snd_timer snd soundcore snd_page_alloc ath5k ath i915
drm_kms_helper drm fb fbdev i2c_algo_bit cfbcopyarea video backlight
output cfbimgblt cfbfillrect intel_agp agpgart
Oct 27 21:49:02 subgenius kernel:
Oct 27 21:49:02 subgenius kernel: Pid: 1846, comm: glxgears Not
tainted 2.6.36 #1 A8Le /A8Le
Oct 27 21:49:02 subgenius kernel: EIP: 0060:[<00000000>] EFLAGS: 00210202 CPU: 1
Oct 27 21:49:02 subgenius kernel: EIP is at 0x0
Oct 27 21:49:02 subgenius kernel: EAX: f722a800 EBX: 00000001 ECX:
f6888014 EDX: f6888014
Oct 27 21:49:02 subgenius kernel: ESI: 00000002 EDI: f68aafd0 EBP:
f1f5fea8 ESP: f1f5fe4c
Oct 27 21:49:02 subgenius kernel: DS: 007b ES: 007b FS: 00d8 GS: 0033 SS: 0068
Oct 27 21:49:02 subgenius kernel: Process glxgears (pid: 1846,
ti=f1f5e000 task=f6564000 task.ti=f1f5e000)
Oct 27 21:49:02 subgenius kernel: Stack:
Oct 27 21:49:02 subgenius kernel: f8365a73 f834af85 00000002 f722a800
f6888000 f802b898 00000000 f68aafc8
Oct 27 21:49:02 subgenius kernel: <0> 00000001 00000000 00000002
00000000 f1f5fea8 f722a800 00000018 f1e2c500
Oct 27 21:49:02 subgenius kernel: <0> f82302f0 bf882c08 4018644b
f8375534 00000018 f722a82c f834ac06 bf882c64
Oct 27 21:49:02 subgenius kernel: Call Trace:
Oct 27 21:49:02 subgenius kernel: [<f8365a73>] ?
intel_ring_advance+0xe/0xf [i915]
Oct 27 21:49:02 subgenius kernel: [<f834af85>] ?
i915_cmdbuffer+0x37f/0x413 [i915]
Oct 27 21:49:02 subgenius kernel: [<f82302f0>] ? drm_ioctl+0x221/0x297 [drm]
Oct 27 21:49:02 subgenius kernel: [<f834ac06>] ?
i915_cmdbuffer+0x0/0x413 [i915]
Oct 27 21:49:02 subgenius kernel: [<c106514b>] ? do_wp_page+0x67d/0x6b3
Oct 27 21:49:02 subgenius kernel: [<c11621b8>] ? pty_write+0x3d/0x43
Oct 27 21:49:02 subgenius kernel: [<c1066698>] ? handle_mm_fault+0x79c/0x858
Oct 27 21:49:02 subgenius kernel: [<c1023575>] ? __wake_up+0x29/0x39
Oct 27 21:49:02 subgenius kernel: [<f82300cf>] ? drm_ioctl+0x0/0x297 [drm]
Oct 27 21:49:02 subgenius kernel: [<c1082e42>] ? vfs_ioctl+0x16/0x2f
Oct 27 21:49:02 subgenius kernel: [<c108334c>] ? do_vfs_ioctl+0x438/0x470
Oct 27 21:49:02 subgenius kernel: [<c1018a26>] ? do_page_fault+0x31a/0x324
Oct 27 21:49:02 subgenius kernel: [<c1044820>] ? sys_futex+0xfc/0x111
Oct 27 21:49:02 subgenius kernel: [<c10833b0>] ? sys_ioctl+0x2c/0x42
Oct 27 21:49:02 subgenius kernel: [<c12590f1>] ? syscall_call+0x7/0xb
Oct 27 21:49:02 subgenius kernel: Code: Bad EIP value.
Oct 27 21:49:02 subgenius kernel: EIP: [<00000000>] 0x0 SS:ESP 0068:f1f5fe4c
Oct 27 21:49:02 subgenius kernel: CR2: 0000000000000000
Oct 27 21:49:02 subgenius kernel: ---[ end trace ccff8844e71a655b ]---
On Mon, 8 Nov 2010 16:56:30 +0800, Adrian Bono <[email protected]> wrote:
> I tested 2.6.36 on an Asus laptop with an Intel GM965 chipset. I first
> noticed the oops when i tried to run Virtualbox, which links to
> opengl, and i tried reproducing it with glxgears, which it
> consistently does.
There are two bugs at play here. The first is a broken userspace driver
causing a GPU hang. The second is that kernel hasn't noticed the hang,
tries to advance the CS ringbuffer, times out, ignores the error and
writes to beyond the end of the mapped region. There's a fix for the
latter in drm-intel-next, and it's likely that the GL bug has been
fixed in the few years since you last updated. At very least, I have a
better chance at diagnosing what went wrong with a GEM based driver.
-Chris
--
Chris Wilson, Intel Open Source Technology Centre