[Bug 104538] New: [IGT] kms_cursor_legacy --r *2x* execution stops on next test after FAIL

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=104538

            Bug ID: 104538
           Summary: [IGT] kms_cursor_legacy --r *2x* execution stops on
                    next test after FAIL
           Product: DRI
           Version: DRI git
          Hardware: x86-64 (AMD64)
                OS: Linux (All)
            Status: NEW
          Severity: major
          Priority: medium
         Component: DRM/Intel
          Assignee: intel-gfx-bugs at lists.freedesktop.org
          Reporter: jari.tahvanainen at intel.com
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
                CC: intel-gfx-bugs at lists.freedesktop.org

Created attachment 136611
  --> https://bugs.freedesktop.org/attachment.cgi?id=136611&action=edit
--debug output with igt-commit cb43735cab7

The change on igt-commit d86d53b072b is causing the new FAIL 

(kms_cursor_legacy:18758) CRITICAL: Test assertion failure function
two_screens_flip_vs_cursor, file kms_cursor_legacy.c:1013:
(kms_cursor_legacy:18758) CRITICAL: Failed assertion: vblank_matches
(kms_cursor_legacy:18758) CRITICAL: Last errno: 16, Device or resource busy
(kms_cursor_legacy:18758) CRITICAL: During modeset at least 1 page flip needs
to match!
(kms_cursor_legacy:18758) igt-core-INFO: Stack trace:
(kms_cursor_legacy:18758) igt-core-INFO:   #0 [__igt_fail_assert+0x101]
(kms_cursor_legacy:18758) igt-core-INFO:   #1
[two_screens_flip_vs_cursor+0x8fe]
(kms_cursor_legacy:18758) igt-core-INFO:   #2 [__real_main1369+0xbee]
(kms_cursor_legacy:18758) igt-core-INFO:   #3 [main+0x27]
(kms_cursor_legacy:18758) igt-core-INFO:   #4 [__libc_start_main+0xf1]
(kms_cursor_legacy:18758) igt-core-INFO:   #5 [_start+0x2a]
(kms_cursor_legacy:18758) igt-core-INFO:   #6 [<unknown>+0x2a]
****  END  ****
Stack trace:
  #0 [__igt_fail_assert+0x101]
  #1 [two_screens_flip_vs_cursor+0x8fe]
  #2 [__real_main1369+0xbee]
  #3 [main+0x27]
  #4 [__libc_start_main+0xf1]
  #5 [_start+0x2a]
  #6 [<unknown>+0x2a]
Subtest 2x-nonblocking-modeset-vs-cursor-atomic: FAIL (1.094s)

and igt to hang on execution of the next test.
(kms_cursor_legacy:18758) igt-core-DEBUG: Starting subtest:
2x-cursor-vs-flip-atomic
...
(kms_cursor_legacy:18758) igt-kms-DEBUG: Test requirement passed: plane_idx >=
0 && plane_idx < pipe->n_planes
(kms_cursor_legacy:18758) igt-kms-DEBUG: display: B.0: plane_set_fb(124)
(kms_cursor_legacy:18758) igt-kms-DEBUG: display: B.0: plane_set_size
(1920x1200)
(kms_cursor_legacy:18758) igt-kms-DEBUG: display: B.0: fb_set_position(0,0)
(kms_cursor_legacy:18758) igt-kms-DEBUG: display: B.0: fb_set_size(1920x1200)
(kms_cursor_legacy:18758) igt-kms-DEBUG: display: commit {
(kms_cursor_legacy:18758) igt-kms-DEBUG: display:     populating plane data:
B.0, fb 124
(kms_cursor_legacy:18758) igt-kms-DEBUG: plane B.0: Setting property "SRC_X" 
(kms_cursor_legacy:18902) DEBUG: child 0: 3515 cursor updates
(kms_cursor_legacy:18903) DEBUG: child 1: 3518 cursor updates

IGT executions stops on above "step" on next test, in this case
2x-cursor-vs-flip-atomic.

On --debug logs (see attachments), the main difference seems to be
pipe(kms_cursor_legacy:18605) DEBUG: child 0: 6628 cursor updates
(kms_cursor_legacy:18606) DEBUG: child 1: 6662 cursor updates
->n_planes
vs.
(kms_cursor_legacy:10997) igt-kms-DEBUG: plane B.0: Setting property "SRC_X"
t(kms_cursor_legacy:11195) DEBUG: child 0: 8033 cursor updates
(kms_cursor_legacy:11196) DEBUG: child 1: 5988 cursor updates
o 0x0/0
with igt-commit cb43735cab7 where igt execution hang does not happen.

Latter works after FAIL case, 1st hangs on next test after FAIL, when using 
sudo ./kms_cursor_legacy --debug --r '*2x*'
on SKL with two displays attached
Connectors:
id      encoder status          type    size (mm)       modes
59      0       disconnected    HDMI-A  0x0             0
64      63      connected       HDMI-A  520x320         21
66      65      connected       DP      520x320         12

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are the QA Contact for the bug.
You are on the CC list for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20180108/56dd6266/attachment-0001.html>