[Bug 105594] [CI][ILK] igt at kms_3d - fail - Test assertion failure function __real_main29 - Failed assertion: ret == 0

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=105594

Arek Hiler <arkadiusz.hiler at intel.com> changed:

           What    |Removed                     |Added
----------------------------------------------------------------------------
            Summary|[CI] igt at kms_3d - fail -    |[CI][ILK] igt at kms_3d - fail
                   |Test assertion failure      |- Test assertion failure
                   |function __real_main29 -    |function __real_main29 -
                   |Failed assertion: ret == 0  |Failed assertion: ret == 0

--- Comment #1 from Arek Hiler <arkadiusz.hiler at intel.com> ---
The test is forcing HDMI connector with EDID exposing 3D enabled modes and then
doing a mode set using those 3D modes.

(kms_3d:1432) INFO: Testing:
(kms_3d:1432) INFO:     1920x1080i 60 1920 2008 2052 2200 1080 1084 1094 1125
0x20015 0x40 74250 (3D:SBSH)
(kms_3d:1432) igt_fb-DEBUG: igt_create_fb_with_bo_size(width=1920, height=1080,
format=XR24(0x34325258), modifier=0x0, size=0)
(kms_3d:1432) igt_fb-DEBUG: igt_create_fb_with_bo_size(handle=1, pitch=7680)
(kms_3d:1432) ioctl_wrappers-DEBUG: Test requirement passed:
igt_has_fb_modifiers(fd)
(kms_3d:1432) igt_fb-DEBUG: Test requirement passed:
cairo_surface_status(fb->cairo_surface) == CAIRO_STATUS_SUCCESS
(kms_3d:1432) CRITICAL: Test assertion failure function __real_main29, file
../tests/kms_3d.c:112:
(kms_3d:1432) CRITICAL: Failed assertion: ret == 0
(kms_3d:1432) CRITICAL: Last errno: 22, Invalid argument

The code:

for (int i = 0; i < connector->count_modes; i++) {
        int fb_id;
        struct kmstest_connector_config config;
        int crtc_mask = -1;
        int ret;

        if (!(connector->modes[i].flags & DRM_MODE_FLAG_3D_MASK))
                continue;

        /* buffer creation removed */

        ret = drmModeSetCrtc(drm_fd, config.crtc->crtc_id, fb_id, 0, 0,
                             &connector->connector_id, 1,
                             &connector->modes[i]);

        igt_assert(ret == 0);
}

So modeset fails with EINVAL.

The tested resolution is:
(kms_3d:1432) INFO:     1920x1080i 60 1920 2008 2052 2200 1080 1084 1094 1125
0x20015 0x40 74250 (3D:SBSH)

And this is not pruned for forced 3D EDID but as we can see in the logs, after
the test exit when we probe actual modes:

<7> [43.875690] [drm:drm_mode_debug_printmodeline] Modeline "1920x1080i": 60
74250 1920 2008 2052 2200 1080 1084 1094 1125 0x40 0x15
<7> [43.875695] [drm:drm_mode_prune_invalid] Not using 1920x1080i mode:
VIRTUAL_X
<7> [43.875707] [drm:drm_mode_debug_printmodeline] Modeline "1920x1080i": 0
74250 1920 2008 2052 2200 1080 1084 1094 1125 0x40 0x15
<7> [43.875712] [drm:drm_mode_prune_invalid] Not using 1920x1080i mode:
VIRTUAL_X
<7> [43.875728] [drm:drm_mode_debug_printmodeline] Modeline "1920x1080i": 60
74176 1920 2008 2052 2200 1080 1084 1094 1125 0x40 0x15
<7> [43.875732] [drm:drm_mode_prune_invalid] Not using 1920x1080i mode:
VIRTUAL_X

Seems like an issue with mode pruning for forced EDID that appears only on ILK.
Maybe other platforms are fine with this mode in the first place?

Keeping as medium as it is affecting old platform, seems to be caused by issue
in HDMI forcing code and it is for rarely used 3D.

-- 
You are receiving this mail because:
You are the assignee for the bug.
You are on the CC list for the bug.
You are the QA Contact for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20190710/2058a9b6/attachment.html>