Bug 59836

Summary: [PNV]igt/kms_flip/absolute-wf_vblank fail
Product: DRI Reporter: lu hua <huax.lu>
Component: DRM/IntelAssignee: Todd Previte <tprevite>
Status: CLOSED FIXED QA Contact: Intel GFX Bugs mailing list <intel-gfx-bugs>
Severity: normal    
Priority: medium CC: bingx.a.yan, yangweix.shui
Version: unspecified   
Hardware: Other   
OS: All   
Whiteboard:
i915 platform: i915 features:
Attachments:
Description Flags
dmesg
none
dmesg on nightly 3477e5
none
dmesg none

Description lu hua 2013-01-25 06:58:14 UTC 
Created attachment 73631 [details]
dmesg

System Environment:
--------------------------
Arch:             i386
Platform:         Pineview
Kernel:           (drm-intel-fixes)4518f611ba21ba165ea3714055938a8984a44ff9

Bug detailed description:
-------------------------
It fails on Pineview with -queued kernel and -fixes kernel.
I can't catch a good commit.

output:
Using monotonic timestamps
running testcase: absolute-wf_vblank
could not find requested crtc 3
Beginning absolute-wf_vblank on crtc 4, connector 5
  1024x600 60 1024 1048 1184 1438 600 603 604 628 0xa 0x48 54200
.select timed out or error (ret 0)


# echo $?
1

Reproduce steps:
----------------
1. ./kms_flip --run-subtest absolute-wf_vblank
Comment 1 Daniel Vetter 2013-01-25 10:35:38 UTC 
What happens on -nightly? That contains the reworked locking and a few other patches which are only in drm-next thus far.
Comment 2 lu hua 2013-01-28 07:17:16 UTC 
It also happens on -nightly kernel.

Kernel: (drm-intel-nightly)2d039f65e28be0be6ef765f6da45df6b3c4efef2
Merge: ef13e7f 735dc0d
Comment 3 yanbing 2013-01-29 06:04:58 UTC 
The subcase "flip-vs-absolute-wf_vblank"  on pnv also exist this problem.And there are same std out.
Comment 4 Daniel Vetter 2013-01-29 10:21:43 UTC 
Hm, do you see the "select timed out" failure also on other machines or only on pnv?
Comment 5 lu hua 2013-01-30 02:16:17 UTC 
(In reply to comment #4)
> Hm, do you see the "select timed out" failure also on other machines or only
> on pnv?

Only on Pineview.
Comment 6 Daniel Vetter 2013-06-12 11:02:55 UTC 
Can you please retest with the following patch applied to latest -nightly?

https://patchwork.kernel.org/patch/2709081/
Comment 7 shui yangwei 2013-06-13 03:03:05 UTC 
(In reply to comment #6)
> Can you please retest with the following patch applied to latest -nightly?
> 
> https://patchwork.kernel.org/patch/2709081/

I applied the patch on nightly top, I find this sub case worked very well, it passed and return value is 0.
Comment 8 Daniel Vetter 2013-06-13 07:37:44 UTC 
It's a drm core patch. I'll close the bug as soon the patch has landed in
upstream drm branch and is merged into -nightly.
Comment 9 Daniel Vetter 2013-08-08 07:13:53 UTC 
Ok, the patch was finally merged to drm-fixes and so can be tested in drm-intel-nightly:

commit e91abf80a0998f326107874c88d549f94839f13c
Author: Michel Dänzer <michel.daenzer@amd.com>
Date:   Wed Jun 12 11:58:44 2013 +0200

    drm: Don't pass negative delta to ktime_sub_ns()
Comment 10 lu hua 2013-08-09 07:47:38 UTC 
Run it on latest -nightly kernel(3477e5ea598c88d21f24c00f8fcdfd7f4e837b59), the first cycle is pass, the 2nd cycle will reproduce this fail.
[root@x-pnv2 tests]# ./kms_flip --run-subtest absolute-wf_vblank
Using monotonic timestamps
running testcase: absolute-wf_vblank
no crtc with a compatible encoder (crtc_idx_mask 00000001)
Beginning absolute-wf_vblank on crtc 4, connector 5
  1024x600 60 1024 1072 1104 1200 600 603 609 625 0xa 0x48 45000
.............................................................
absolute-wf_vblank on crtc 4, connector 5: PASSED

[root@x-pnv2 tests]# ./kms_flip --run-subtest absolute-wf_vblank
Using monotonic timestamps
running testcase: absolute-wf_vblank
no crtc with a compatible encoder (crtc_idx_mask 00000001)
Beginning absolute-wf_vblank on crtc 4, connector 5
  1024x600 60 1024 1072 1104 1200 600 603 609 625 0xa 0x48 45000
.select timed out or error (ret 0)
Comment 11 lu hua 2013-08-09 07:48:23 UTC 
Created attachment 83877 [details]
dmesg on nightly 3477e5
Comment 12 Jani Nikula 2014-01-10 15:35:06 UTC 
The bug seems to have been neglected a bit. Is this still an issue on current -nightly and igt? Please attach current dmesg if it is. Thanks.
Comment 13 lu hua 2014-01-13 03:31:44 UTC 
The output info shows success, but it causes call trace.
output:
Using monotonic timestamps
Beginning absolute-wf_vblank on crtc 4, connector 5
  1024x600 60 1024 1072 1104 1200 600 603 609 625 0xa 0x48 45000
....................................................................................................................................................................................
absolute-wf_vblank on crtc 4, connector 5: PASSED

Subtest absolute-wf_vblank: SUCCESS

Call trace:
[    2.440626] WARNING: CPU: 1 PID: 1208 at drivers/gpu/drm/i915/intel_panel.c:342 intel_panel_compute_brightness+0x27/0x4b [i915]()
[    2.440635] Modules linked in: i915(+) video button drm_kms_helper drm
[    2.440640] CPU: 1 PID: 1208 Comm: udevd Not tainted 3.13.0-rc7_drm-intel-nightly_24ed7d_20140113+ #2757
[    2.440642] Hardware name: MICRO-STAR INTERNATIONAL CO., LTD MS-N014/MS-N014, BIOS EN014IMS.10B 11/30/2009
[    2.440650]  00000000 c0890ccc 00000000 c022b5df f82e47e6 00000000 f5970000 f587c000
[    2.440657]  00000293 c022b5ff 00000009 00000000 f82e47e6 f5970000 f587c000 f587d2d4
[    2.440663]  f82e483d 00000002 f831569e f8308adc f8315686 00000000 00000000 f5970000
[    2.440665] Call Trace:
[    2.440673]  [<c0890ccc>] ? dump_stack+0x3e/0x4e
[    2.440680]  [<c022b5df>] ? warn_slowpath_common+0x61/0x74
[    2.440740]  [<f82e47e6>] ? intel_panel_compute_brightness+0x27/0x4b [i915]
[    2.440746]  [<c022b5ff>] ? warn_slowpath_null+0xd/0x10
[    2.440806]  [<f82e47e6>] ? intel_panel_compute_brightness+0x27/0x4b [i915]
[    2.440867]  [<f82e483d>] ? intel_panel_actually_set_backlight+0x33/0x43 [i915]
[    2.440929]  [<f82e584a>] ? intel_panel_disable_backlight+0x58/0x6a [i915]
[    2.440986]  [<f82d5cb1>] ? intel_disable_lvds+0x35/0x110 [i915]
[    2.441011]  [<f82ca647>] ? i9xx_crtc_disable+0x50/0x20c [i915]
[    2.441011]  [<f82d0237>] ? __intel_set_mode+0xbd6/0xf82 [i915]
[    2.441011]  [<c024f907>] ? down_trylock+0x1b/0x23
[    2.441011]  [<c025727c>] ? console_trylock+0xb/0x43
[    2.441011]  [<f82d1ea1>] ? intel_set_mode+0x11/0x25 [i915]
[    2.441011]  [<f82d24b7>] ? intel_crtc_set_config+0x551/0x72a [i915]
[    2.441011]  [<f810b974>] ? drm_mode_set_config_internal+0x36/0x8c [drm]
[    2.441011]  [<f812bf55>] ? drm_fb_helper_set_par+0x46/0x7c [drm_kms_helper]
[    2.441011]  [<c04b270a>] ? fbcon_init+0x2c3/0x3a3
[    2.441011]  [<c04f60e5>] ? visual_init+0x97/0xe9
[    2.441011]  [<c04f7758>] ? do_bind_con_driver+0x16e/0x275
[    2.441011]  [<c04f796b>] ? do_take_over_console+0x10c/0x134
[    2.441011]  [<c04b1ece>] ? do_fbcon_takeover+0x49/0x88
[    2.441011]  [<c0897e62>] ? notifier_call_chain+0x20/0x42
[    2.441011]  [<c0241074>] ? __blocking_notifier_call_chain+0x34/0x4b
[    2.441011]  [<c0241094>] ? blocking_notifier_call_chain+0x9/0xc
[    2.441011]  [<c04aa838>] ? register_framebuffer+0x1f6/0x248
[    2.441011]  [<f812be8a>] ? drm_fb_helper_initial_config+0x33c/0x3c1 [drm_kms_helper]
[    2.441011]  [<f82acf19>] ? i915_driver_load+0x927/0xb64 [i915]
[    2.441011]  [<f81075a1>] ? drm_dev_register+0xb1/0x113 [drm]
[    2.441011]  [<f8108f6e>] ? drm_get_pci_dev+0xdd/0x176 [drm]
[    2.441011]  [<c049cd81>] ? pci_device_probe+0x4e/0x95
[    2.441011]  [<c050f36b>] ? driver_probe_device+0x79/0x168
[    2.441011]  [<c050f49e>] ? __driver_attach+0x44/0x5f
[    2.441011]  [<c050e177>] ? bus_for_each_dev+0x37/0x61
[    2.441011]  [<c050f03e>] ? driver_attach+0x11/0x13
[    2.441011]  [<c050f45a>] ? driver_probe_device+0x168/0x168
[    2.441011]  [<c050ed7b>] ? bus_add_driver+0xad/0x167
[    2.441011]  [<c050f8f2>] ? driver_register+0x6b/0x99
[    2.441011]  [<f8148000>] ? 0xf8147fff
[    2.441011]  [<c02003d4>] ? do_one_initcall+0x6d/0xe9
[    2.441011]  [<c0897e62>] ? notifier_call_chain+0x20/0x42
[    2.441011]  [<c024107d>] ? __blocking_notifier_call_chain+0x3d/0x4b
[    2.441011]  [<c026d0a9>] ? load_module+0x140b/0x16e5
[    2.441011]  [<c0897e35>] ? __do_page_fault+0x45d/0x45d
[    2.441011]  [<c026d3f5>] ? SyS_init_module+0x72/0x8a
[    2.441011]  [<c089991a>] ? sysenter_do_call+0x12/0x22
[    2.441011] ---[ end trace c424884943f003b3 ]---
Comment 14 lu hua 2014-01-13 03:32:05 UTC 
Created attachment 91921 [details]
dmesg
Comment 15 Todd Previte 2014-05-13 19:18:58 UTC 
Is this still an issue with current kernels? It's been a while since there was any input on this.

Looking at the code though, the call trace is from here:
   WARN_ON(panel->backlight.max == 0);

I'll see what else I can dig up, but I don't have a PNV system to investigate this on.
Comment 16 Jani Nikula 2014-05-14 08:41:19 UTC 
(In reply to comment #15)
> Looking at the code though, the call trace is from here:
>    WARN_ON(panel->backlight.max == 0);

At least this shouldn't happen on current nightly.
Comment 17 lu hua 2014-05-15 07:00:36 UTC 
It works well on latest -nightly kernel. Close it.
Comment 18 lu hua 2014-05-15 07:00:52 UTC 
Verified.Fixed.
Comment 19 Elizabeth 2017-10-06 14:47:01 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.