Bug 94848 - GPU HANG: ecode 9:0:0x85dfffff on i5-6260U
Summary: GPU HANG: ecode 9:0:0x85dfffff on i5-6260U
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: x86-64 (AMD64) Linux (All)
: medium major
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2016-04-06 17:32 UTC by Paul
Modified: 2016-09-05 10:00 UTC (History)
1 user (show)

See Also:
i915 platform: SKL
i915 features: GPU hang


Attachments
two faults (145.60 KB, text/plain)
2016-04-06 17:32 UTC, Paul
no flags Details

Description Paul 2016-04-06 17:32:33 UTC
Created attachment 122772 [details]
two faults

I know others have seen this error, but I'll add it if folks want the logs.

I'm using Ubuntu 14.04, the kernel is drm-intel-nightly, but I ran into compile issues with libdrm so everything else is the latest Ubuntu.

Some more details;
arch: x86_64
kernel: 4.6.0-rc2+ (drm-intel-nightly from 4/4/16)
Machine: NUC6I5SYH (i5-6260U)
Display: HDMI & DP

First I captured dmesg & card0/error without the drm.debug command line option, this is the f1 folder in the bz2, but then I saw the note about wanting that so the f2_dbg_cmdline folder is with "drm.debug=0x1e log_buf_len=1M"

In both of these cases the screens freeze, but the mouse continues to work, then about 10-20 seconds later it comes back to normal. There is a message in the lower right corner "KWin Window Manager, desktop effects were restarted due to a graphics restart" I've also had instances where both the screens & mouse freeze, and when this happens a reboot is needed even ssh doesn't work.

thanks,
Paul
Comment 1 Paul 2016-04-12 20:53:38 UTC
This seems to be cleared up with BIOS update 39.
Comment 2 yann 2016-09-05 10:00:35 UTC
Closing since this is BIOS update fixed it.

As reference, from these error dumps (this is similar so taking ACTHD from 1st one), hung is happening in render ring batch with active head at 0xfa870f00, with 0x7a000004 (PIPE_CONTROL) as IPEHR.

Batch extract (around 0xfa870f00):

0xfa870cf0:      0x00000000: MI_NOOP
Bad length 7 in (null), expected 6-6
0xfa870cf4:      0x7b000005: 3DPRIMITIVE: fail sequential
0xfa870cf8:      0x00000004:    vertex count
0xfa870cfc:      0x00000006:    start vertex
0xfa870d00:      0x00000000:    instance count
0xfa870d04:      0x00000001:    start instance
0xfa870d08:      0x00000000:    index bias
0xfa870d0c:      0x00000000: MI_NOOP
Bad count in PIPE_CONTROL
0xfa870d10:      0x7a000004: PIPE_CONTROL: no write, no depth stall, no RC write flush, no inst flush
0xfa870d14:      0x00000000:    destination address
0xfa870d18:      0x00000000:    immediate dword low
0xfa870d1c:      0x00000000:    immediate dword high
Bad count in PIPE_CONTROL
0xfa870d28:      0x7a000004: PIPE_CONTROL: no write, no depth stall, no RC write flush, no inst flush
0xfa870d2c:      0x00101c11:    destination address
0xfa870d30:      0x00000000:    immediate dword low
0xfa870d34:      0x00000000:    immediate dword high
0xfa870d40:      0x78230000: 3D UNKNOWN: 3d_965 opcode = 0x7823
0xfa870d44:      0x00007480: MI_NOOP


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.