Bug 77869 - [BDW/BSW]Some subcases of pm_rps fail - gpu turbo busted
Summary: [BDW/BSW]Some subcases of pm_rps fail - gpu turbo busted
Status: CLOSED FIXED
Alias: None
Product: DRI
Classification: Unclassified
Component: DRM/Intel (show other bugs)
Version: unspecified
Hardware: Other All
: medium normal
Assignee: Intel GFX Bugs mailing list
QA Contact: Intel GFX Bugs mailing list
URL:
Whiteboard:
Keywords:
Depends on:
Blocks:
 
Reported: 2014-04-24 06:36 UTC by Guo Jinxian
Modified: 2017-10-06 14:38 UTC (History)
3 users (show)

See Also:
i915 platform:
i915 features:


Attachments
dmesg (121.71 KB, text/plain)
2014-04-24 06:36 UTC, Guo Jinxian
no flags Details
dmesg (83.00 KB, text/plain)
2014-05-21 01:45 UTC, Guo Jinxian
no flags Details
BDW fix (2.55 KB, patch)
2014-09-26 16:27 UTC, Rodrigo Vivi
no flags Details | Splinter Review

Description Guo Jinxian 2014-04-24 06:36:27 UTC
Created attachment 97873 [details]
dmesg

System Environment:
--------------------------
Platform: HSW
kernel:   (drm-intel-next-queued) c79057922ed6c2c6df1214e6ab4414fea1b23db2

Bug detailed description:
----------------------------
Subcases blocking and min-max-config-idle of pm_rps fail on -nightly(35dc7c8af75a3e5ae52a5469786602ce083a6368) and -next-queued(c79057922ed6c2c6df1214e6ab4414fea1b23db2)

This bug always fail

output on -next-queued kernel:
./pm_rps --run-subtest min-max-config-idle
IGT-Version: 1.6-g78e4c2b (x86_64) (Linux: 3.14.0_drm-intel-next-queued_f68918_20140424+ x86_64)
Test assertion failure function idle_check, file pm_rps.c:412:
Last errno: 22, Invalid argument
Failed assertion: freqs[CUR] == freqs[MIN]
error: 1200 == 700
Subtest min-max-config-idle: FAIL

Reproduce steps:
---------------------------- 

1. ./pm_rps --run-subtest blocking
   or
   ./pm_rps --run-subtest min-max-config-idle
Comment 1 Guo Jinxian 2014-04-28 06:01:08 UTC
Subcases below fail on BDW on latest -nightly(11ddb598492d1f97f894495eea398febb41a9eb0) too:

basic-api
min-max-config-loaded
reset
Comment 2 Daniel Vetter 2014-05-19 10:32:18 UTC
Please retest on latest -nightly. Also do the exact same tests fail on HSW as on BDW?
Comment 3 Guo Jinxian 2014-05-21 01:45:09 UTC
Created attachment 99459 [details]
dmesg

(In reply to comment #2)
> Please retest on latest -nightly. Also do the exact same tests fail on HSW
> as on BDW?
This issue unable to reproduce on HSW now. But it still able to reproduce on BDW.
./pm_rps
IGT-Version: 1.6-g737d248 (x86_64) (Linux: 3.15.0-rc3_drm-intel-nightly_0791a3_20140520+ x86_64)
Subtest basic-api: SUCCESS
Test assertion failure function idle_check, file pm_rps.c:412:
Last errno: 22, Invalid argument
Failed assertion: freqs[CUR] == freqs[MIN]
error: 400 == 100
Subtest min-max-config-idle: FAIL
Subtest min-max-config-loaded: SUCCESS
Test assertion failure function matchit, file pm_rps.c:128:
Last errno: 0, Success
Failed assertion: freqs1[CUR] == freqs2[CUR]
error: 100 == 700
Subtest reset: FAIL
Subtest blocking: SUCCESS
Comment 4 Guo Jinxian 2014-05-28 03:03:12 UTC
This bug is able to reproduce on latest -nightly(3dabfd0ec37a5079f40491f864f706a9aeb9d87b) too.
Comment 5 wendy.wang 2014-06-30 03:08:58 UTC
Jesse, any comments about this bug.
drm-intel-nightly latest kernel still can reproduce this bug.
(drm-intel-nightly/1087d4bf01e79523898c6c31615bf0c369e0039a)

root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./pm_rps
IGT-Version: 1.7-g7ef5372 (x86_64) (Linux: 3.16.0-rc2_drm-intel-nightly_1087d4_20140629+ x86_64)
Subtest basic-api: SUCCESS
Test assertion failure function idle_check, file pm_rps.c:412:
Last errno: 22, Invalid argument
Failed assertion: freqs[CUR] == freqs[MIN]
error: 400 == 100
Subtest min-max-config-idle: FAIL
Subtest min-max-config-loaded: SUCCESS
Test assertion failure function matchit, file pm_rps.c:128:
Last errno: 0, Success
Failed assertion: freqs1[CUR] == freqs2[CUR]
error: 100 == 700
Subtest reset: FAIL
Subtest blocking: SUCCESS
Comment 6 Daniel Vetter 2014-07-07 13:00:16 UTC
http://patchwork.freedesktop.org/patch/27962/

Please test this patch.
Comment 7 Guo Jinxian 2014-07-08 08:00:24 UTC
(In reply to comment #6)
> http://patchwork.freedesktop.org/patch/27962/
> 
> Please test this patch.

Sorry, the patch unable to patched, I tried to modified the code according the patch manually, the source unable to make after change. Could you tell me which commit the patch base? Thanks.
Comment 8 Guo Jinxian 2014-07-24 08:10:47 UTC
Test failed on latest -fixes(eedd10f45bdcb2a5b2afa35f845e080c3bc984f2) on HSW

[root@x-hsw27 tests]# ./pm_rps --run-subtest min-max-config-idle
IGT-Version: 1.7-g70e6ed9 (x86_64) (Linux: 3.16.0-rc6_drm-intel-fixes_eedd10_20140724+ x86_64)
Test assertion failure function idle_check, file pm_rps.c:412:
Failed assertion: freqs[CUR] == freqs[MIN]
Last errno: 22, Invalid argument
error: 350 == 200
Subtest min-max-config-idle: FAIL
Comment 9 Guo Jinxian 2014-07-25 02:45:38 UTC
Test failed while running test igt/pm_rps/reset on BDW on latest -nightly(af1aaba219fdd90ca1b30f9b8d8d19352224f170)

root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./pm_rps --run-subtest reset
IGT-Version: 1.7-gbffef22 (x86_64) (Linux: 3.16.0-rc6_drm-intel-nightly_af1aab_20140724_debug+ x86_64)
Test assertion failure function matchit, file pm_rps.c:127:
Failed assertion: freqs1[CUR] == freqs2[CUR]
error: 100 == 900
Subtest reset: FAIL
Comment 10 Guo Jinxian 2014-07-29 02:30:49 UTC
The failure is able to reproduce on BSW on latest -next-queued(eff9b57c1a91ccf309d57500ab6a365ba7be5712)

[root@x-bsw01 tests]# ./pm_rps --run-subtest min-max-config-idle
IGT-Version: 1.7-ge48c495 (x86_64) (Linux: 3.16.0-rc4_drm-intel-next-queued_eff9b5_20140728+ x86_64)
Test assertion failure function idle_check, file pm_rps.c:412:
Failed assertion: freqs[CUR] == freqs[MIN]
Last errno: 22, Invalid argument
error: 480 == 300
Subtest min-max-config-idle: FAIL
Comment 11 Guo Jinxian 2014-07-30 08:31:40 UTC
Test still failed on latest -nightly(eb2af97968f0321622cf3aeb29991ad78cfdc1ca) on BSW

[root@x-bsw01 tests]# ./pm_rps --run-subtest min-max-config-loaded
IGT-Version: 1.7-ge48c495 (x86_64) (Linux: 3.16.0-rc6_drm-intel-nightly_eb2af9_20140730+ x86_64)
Test assertion failure function loaded_check, file pm_rps.c:435:
Failed assertion: freqs[CUR] == freqs[MAX]
error: 160 == 480
Subtest min-max-config-loaded: FAIL
Test assertion failure function load_helper_stop, file pm_rps.c:258:
Failed assertion: igt_wait_helper(&lh.igt_proc) == 0
Last errno: 10, No child processes
pm_rps: igt_core.c:714: igt_fail: Assertion `!test_with_subtests || in_fixture' failed.
Aborted (core dumped)
Comment 12 Rodrigo Vivi 2014-09-24 20:05:17 UTC
Can you please vefiry latest -nightly stage of this as well?
Comment 13 Guo Jinxian 2014-09-25 07:55:35 UTC
(In reply to comment #12)
> Can you please vefiry latest -nightly stage of this as well?

Test still fails on latest -nightly(0f7cc12c94e3a3033a46ce41bed55e8b6b35561b)

root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./pm_rps --run-subtest blocking
IGT-Version: 1.7-g5f16ef6 (x86_64) (Linux: 3.17.0-rc6_drm-intel-nightly_0f7cc1_20140925+ x86_64)
Test assertion failure function idle_check, file pm_rps.c:412:
Failed assertion: freqs[CUR] == freqs[MIN]
error: 900 == 100
Subtest blocking: FAIL
root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./pm_rps --run-subtest min-max-config-idle
IGT-Version: 1.7-g5f16ef6 (x86_64) (Linux: 3.17.0-rc6_drm-intel-nightly_0f7cc1_20140925+ x86_64)
Test assertion failure function idle_check, file pm_rps.c:412:
Failed assertion: freqs[CUR] == freqs[MIN]
error: 900 == 100
Subtest min-max-config-idle: FAIL
Comment 14 Rodrigo Vivi 2014-09-26 16:27:05 UTC
Created attachment 106926 [details] [review]
BDW fix

Hi Guo, could you please test if this patch fixes the issue on BDW?
Comment 15 lu hua 2014-10-11 05:41:57 UTC
(In reply to Guo Jinxian from comment #11)
> Test still failed on latest
> -nightly(eb2af97968f0321622cf3aeb29991ad78cfdc1ca) on BSW
> 
> [root@x-bsw01 tests]# ./pm_rps --run-subtest min-max-config-loaded
> IGT-Version: 1.7-ge48c495 (x86_64) (Linux:
> 3.16.0-rc6_drm-intel-nightly_eb2af9_20140730+ x86_64)
> Test assertion failure function loaded_check, file pm_rps.c:435:
> Failed assertion: freqs[CUR] == freqs[MAX]
> error: 160 == 480
> Subtest min-max-config-loaded: FAIL
> Test assertion failure function load_helper_stop, file pm_rps.c:258:
> Failed assertion: igt_wait_helper(&lh.igt_proc) == 0
> Last errno: 10, No child processes
> pm_rps: igt_core.c:714: igt_fail: Assertion `!test_with_subtests ||
> in_fixture' failed.
> Aborted (core dumped)

report bug 84896 to track this aborted
Comment 16 Guo Jinxian 2014-11-25 07:39:51 UTC
root@x-bdw05:/GFX/Test/Intel_gpu_tools/intel-gpu-tools/tests# ./pm_rps --run-subtest blocking
IGT-Version: 1.8-gd807891 (x86_64) (Linux: 3.18.0-rc6_drm-intel-nightly_ab4b25_20141125+ x86_64)
Subtest blocking: SUCCESS (35.013s)
(In reply to Rodrigo Vivi from comment #14)
> Created attachment 106926 [details] [review] [review]
> BDW fix
> 
> Hi Guo, could you please test if this patch fixes the issue on BDW?

The patch unable to make on latest -nightly(ab4b258a6ea5f58b5cb17131aced8f9a8dd64499). Could you help to update the patch. Thanks.
Comment 17 Rodrigo Vivi 2015-01-21 23:15:46 UTC
turbo got reverted a while ago so this issue has gone.
Comment 18 Elizabeth 2017-10-06 14:38:32 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.