Bug 90809

Summary:

[SNB+]igt/gem_streaming_writes/batch-cpu doesn't exit testing

Product:

DRI

Reporter:

lu hua <huax.lu>

Component:

DRM/Intel

Assignee:

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

Status:

CLOSED FIXED

QA Contact:

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

Severity:

normal

Priority:

medium

CC:

intel-gfx-bugs

Version:

unspecified

Hardware:

All

OS:

Linux (All)

Whiteboard:

i915 platform:

i915 features:

Attachments:

Description	Flags
dmesg	none

Description lu hua 2015-06-02 05:21:52 UTC

Created attachment 116228 [details]
dmesg

==System Environment==
--------------------------
Regression: not sure, new case

Non-working platforms:  SNB+

==kernel==
--------------------------
drm-intel-nightly/4b2be843fab906467eed62265f5066418c73649c
commit 4b2be843fab906467eed62265f5066418c73649c
Author: Daniel Vetter <daniel.vetter@ffwll.ch>
Date:   Fri May 29 19:28:41 2015 +0200

    drm-intel-nightly: 2015y-05m-29d-17h-28m-04s UTC integration manifest

==Bug detailed description==
-----------------------------
It doesn't exit testing on SNB+ platforms with drm-intel-nightly and drm-intel-next-queued kernel. It fails on drm-intel-fixes kernel. 

output on -nightly kernel:
IGT-Version: 1.10-g0d16473 (x86_64) (Linux: 4.1.0-rc5_drm-intel-nightly_4b2be8_20150530+ x86_64)
^C^C^C^C^C^C^C^C^C^C^C^C^C

Call trace:
[    2.080088] WARNING: CPU: 0 PID: 1218 at drivers/gpu/drm/drm_irq.c:1159 drm_wait_one_vblank+0x3b/0x16d [drm]()
[    2.080089] vblank not available on crtc 0, ret=-22
[    2.080093] Modules linked in: i915 button video drm_kms_helper drm
[    2.080097] CPU: 0 PID: 1218 Comm: kworker/u16:4 Not tainted 4.1.0-rc5_drm-intel-nightly_4b2be8_20150530+ #718
[    2.080104] Workqueue: events_unbound async_run_entry_fn
[    2.080107]  0000000000000000 0000000000000009 ffffffff817a70bc ffff880144bbb8b8
[    2.080110]  ffffffff8103ebde 0000000000000246 ffffffffa0005a6f 0000000000000000
[    2.080112]  ffff88014953f000 0000000000000000 ffff88014964f000 0000000000000009
[    2.080113] Call Trace:
[    2.080119]  [<ffffffff817a70bc>] ? dump_stack+0x40/0x50
[    2.080123]  [<ffffffff8103ebde>] ? warn_slowpath_common+0x98/0xb0
[    2.080132]  [<ffffffffa0005a6f>] ? drm_wait_one_vblank+0x3b/0x16d [drm]
[    2.080134]  [<ffffffff8103ec3b>] ? warn_slowpath_fmt+0x45/0x4a
[    2.080142]  [<ffffffffa0005a6f>] ? drm_wait_one_vblank+0x3b/0x16d [drm]
[    2.080147]  [<ffffffff813fccdc>] ? __pm_runtime_resume+0x5b/0x6a
[    2.080219]  [<ffffffffa00d5a58>] ? intel_finish_crtc_commit+0x47/0x10b [i915]
[    2.080226]  [<ffffffffa0058b44>] ? drm_atomic_helper_commit_planes+0x16a/0x1a1 [drm_kms_helper]
[    2.080285]  [<ffffffffa00d6b5a>] ? __intel_set_mode+0x891/0x916 [i915]
[    2.080340]  [<ffffffffa00dc3f4>] ? intel_crtc_set_config+0x3ef/0x53a [i915]
[    2.080351]  [<ffffffffa00185af>] ? drm_modeset_lock+0x4e/0xa3 [drm]
[    2.080364]  [<ffffffffa000c22d>] ? drm_mode_set_config_internal+0x4e/0xd2 [drm]
[    2.080369]  [<ffffffffa005b15d>] ? restore_fbdev_mode+0xac/0xc3 [drm_kms_helper]
[    2.080373]  [<ffffffffa005c95e>] ? drm_fb_helper_restore_fbdev_mode_unlocked+0x1e/0x54 [drm_kms_helper]
[    2.080378]  [<ffffffffa005c9c2>] ? drm_fb_helper_set_par+0x2e/0x32 [drm_kms_helper]
[    2.080436]  [<ffffffffa00e8d53>] ? intel_fbdev_set_par+0x11/0x55 [i915]
[    2.080441]  [<ffffffff8137ec6e>] ? fbcon_init+0x2fd/0x406
[    2.080444]  [<ffffffff813d4777>] ? visual_init+0xaf/0x102
[    2.080447]  [<ffffffff813d5cc9>] ? do_bind_con_driver+0x19e/0x2c2
[    2.080450]  [<ffffffff813d60a1>] ? do_take_over_console+0x12c/0x15c
[    2.080453]  [<ffffffff8137e35d>] ? do_fbcon_takeover+0x53/0x97
[    2.080457]  [<ffffffff810549dc>] ? notifier_call_chain+0x35/0x59
[    2.080461]  [<ffffffff81054c23>] ? __blocking_notifier_call_chain+0x43/0x5b
[    2.080464]  [<ffffffff813863eb>] ? lock_fb_info+0x12/0x2f
[    2.080468]  [<ffffffff81387c06>] ? register_framebuffer+0x26c/0x2a2
[    2.080473]  [<ffffffffa005cc73>] ? drm_fb_helper_initial_config+0x2ad/0x34a [drm_kms_helper]
[    2.080476]  [<ffffffff81055b8f>] ? async_run_entry_fn+0x2d/0xbf
[    2.080478]  [<ffffffff8104f985>] ? process_one_work+0x1b2/0x31d
[    2.080481]  [<ffffffff8105026f>] ? worker_thread+0x265/0x351
[    2.080484]  [<ffffffff8105000a>] ? cancel_delayed_work_sync+0xa/0xa
[    2.080486]  [<ffffffff81053ee1>] ? kthread+0xce/0xd6
[    2.080490]  [<ffffffff81053e13>] ? kthread_create_on_node+0x162/0x162
[    2.080493]  [<ffffffff817acfd2>] ? ret_from_fork+0x42/0x70
[    2.080496]  [<ffffffff81053e13>] ? kthread_create_on_node+0x162/0x162
[    2.080498] ---[ end trace ee079f4240dd35e0 ]---


output on -fixes kernel:
IGT-Version: 1.10-g0d16473 (x86_64) (Linux: 4.1.0-rc5_drm-intel-fixes_b45aa8_20150531+ x86_64)
Test assertion failure function gem_execbuf, file ioctl_wrappers.c:445:
Failed assertion: result == 0
Last errno: 5, Input/output error
Stack trace:
  #0 [__igt_fail_assert+0xf1]
  #1 [gem_mmap__gtt+0x0]
  #2 [test_batch+0x3d4]
  #3 [__real_main363+0x12e]
  #4 [main+0x29]
  #5 [__libc_start_main+0xf5]
  #6 [_start+0x29]
  #7 [<unknown>+0x29]
Subtest batch-cpu failed.
**** DEBUG ****
Test requirement passed: gem_has_llc(fd)
Test assertion failure function gem_execbuf, file ioctl_wrappers.c:445:
Failed assertion: result == 0
Last errno: 5, Input/output error
****  END  ****
Subtest batch-cpu: FAIL (11.645s)

==Reproduce steps==
---------------------------- 
1. time ./gem_streaming_writes --run-subtest batch-cpu

Comment 1 Chris Wilson 2015-06-02 10:29:02 UTC

That bdw failure is still a kernel bug, but 

commit 977730084647d32b98019924b81b281bef942689
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Tue Jun 2 11:15:16 2015 +0100

    igt/gem_streaming_writes: Map the whole batch for CPU accesses
    
    The llc cpu path only partially mapped the batch buffer so confused the
    CS when attempting to execute an empty batch.
    
    Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=90809
    Signed-off-by: Chris Wilson <chris@chris-wilson.co.uk>

will prevent this test failure.

Comment 2 ye.tian 2015-06-03 02:47:07 UTC

Tested it on the SNB/IVB/BDW with the latest nightly kernel and latest igt (85ee6e7b3), this problem does not exists.
So verified it.

Comment 3 Elizabeth 2017-10-06 14:29:45 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.