Bug 73584

Summary:

[SNB]igt/gem_ctx_exec/eviction skipped

Product:

DRI

Reporter:

lu hua <huax.lu>

Component:

DRM/Intel

Assignee:

Daniel Vetter <daniel>

Status:

CLOSED FIXED

QA Contact:

Intel GFX Bugs mailing list <intel-gfx-bugs>

Severity:

major

Priority:

high

CC:

intel-gfx-bugs

Version:

unspecified

Hardware:

All

OS:

Linux (All)

Whiteboard:

i915 platform:

i915 features:

Attachments:

Description	Flags
dmesg	none
meminfo	none

Description lu hua 2014-01-14 03:43:04 UTC

System Environment:
--------------------------
Platform:  Sandybridge
Kernel:    (drm-intel-nightly)24ed7daacf91b61ea39a71daa9e37e0e80a3392d

Bug detailed description:
---------------------------
It skip on sandybridge with -nightly ,-queued and -fixes kernel.

Bisect shows: 7775fca2df815dfee18b181de6fe13df27bb9867 is the first bad commit
commit 7775fca2df815dfee18b181de6fe13df27bb9867
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Wed Jan 8 20:24:36 2014 +0100

    tests/gem_ctx_exec: fix ram requirement fumble

    We need to half available ram, not double it ...

    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=72557
    Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>

output:
IGT-Version: 1.5-g2a7dbe5 (i686) (Linux: 3.13.0-rc4_drm-intel-nightly_2c21a5_20140107+ i686)
Test requirement not met in function big_exec, file gem_ctx_exec.c:105:
Last errno: 2, No such file or directory
Test requirement: (!(intel_get_total_ram_mb() * 1024 / 2 > num_buffers * 4))
Subtest eviction: SKIP

Reproduce steps:
-------------------------
1. ./gem_ctx_exec --run-subtest eviction

Comment 1 Daniel Vetter 2014-01-14 09:51:41 UTC

Please attach debug dmesg from boot and /proc/meminfo

Comment 2 Daniel Vetter 2014-01-14 09:52:32 UTC

Also, regressions should be assigned to the patch author when a bisect is available.

Comment 3 lu hua 2014-01-16 02:02:57 UTC

Created attachment 92200 [details]
dmesg

Comment 4 lu hua 2014-01-16 02:03:15 UTC

Created attachment 92201 [details]
meminfo

Comment 5 Daniel Vetter 2014-01-17 06:39:37 UTC

Yeah, meminfo shows only 3G of memory, the tests needs about 4G to work well. The only problem was that I've screwed up the condition for that check and usually I don't test with restricted configs specifically. So this is expected behaviour.

As long as we still have one machine where this tests run successfully we have sufficient coverage. When you verify this bug, can you pls check that it still runs successfully on at least one other machine?

Comment 6 lu hua 2014-01-17 08:15:33 UTC

It works well on another SNB machine(8G memory).
Verified.Fixed.

Comment 7 Elizabeth 2017-10-06 14:40:34 UTC

Closing old verified.

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.