[Bug 80079] New: [SNB/IVB/HSW/BYT]igt/gem_fence_upload some subcases fail

[bugzilla-daemon at freedesktop.org]

https://bugs.freedesktop.org/show_bug.cgi?id=80079

          Priority: medium
            Bug ID: 80079
                CC: intel-gfx-bugs at lists.freedesktop.org
          Assignee: intel-gfx-bugs at lists.freedesktop.org
           Summary: [SNB/IVB/HSW/BYT]igt/gem_fence_upload some subcases
                    fail
        QA Contact: intel-gfx-bugs at lists.freedesktop.org
          Severity: normal
    Classification: Unclassified
                OS: All
          Reporter: jinxianx.guo at intel.com
          Hardware: Other
            Status: NEW
           Version: unspecified
         Component: DRM/Intel
           Product: DRI

Created attachment 101143
  --> https://bugs.freedesktop.org/attachment.cgi?id=101143&action=edit
dmesg

==System Environment==
--------------------------
Regression: No. 
It's new case

Non-working platforms: SNB IVB HSW BYT

==kernel==
--------------------------
origin/drm-intel-nightly: b438e8793c6d18b0998d75cdf8115a5898ed21d1(fails)
    drm-intel-nightly: 2014y-06m-13d-22h-09m-12s integration manifest
origin/drm-intel-next-queued: 868d665b43473e230d560d5186535270a3d57a19(fails)
    drm/i915: Fix memory leak in intel_dsi_init() error path
origin/drm-intel-fixes: 223a6f2b975ab35d93270ea1d4fb6e0ac6b27fe6(fails)
    drm/i915/bdw: remove erroneous chv specific workarounds from bdw code

==Bug detailed description==
-----------------------------
igt/gem_fence_upload some subcases fail

Output:
[root at x-hsw26 tests]# ./gem_fence_upload
IGT-Version: 1.7-g8c1566e (x86_64) (Linux:
3.15.0-rc8_drm-intel-next-queued_868d66_20140616+ x86_64)
Upload rate for 2 linear surfaces:      15653.304MiB/s
Upload rate for 2 tiled surfaces:       16163.147MiB/s
Upload rate for 4 linear surfaces:      15617.791MiB/s
Upload rate for 4 tiled surfaces:       16181.664MiB/s
Upload rate for 8 linear surfaces:      15600.660MiB/s
Upload rate for 8 tiled surfaces:       16156.325MiB/s
Upload rate for 16 linear surfaces:     15672.455MiB/s
Upload rate for 16 tiled surfaces:      16212.921MiB/s
Upload rate for 32 linear surfaces:     15612.530MiB/s
Upload rate for 32 tiled surfaces:      16160.118MiB/s
Upload rate for 64 linear surfaces:     15633.077MiB/s
Upload rate for 64 tiled surfaces:      6755.993MiB/s
Test assertion failure function performance, file gem_fence_upload.c:108:
Last errno: 0, Success
Failed assertion: tiled[1] > 0.75 * tiled[0]
Subtest performance: FAIL
Contended upload rate for 1 threads:    1654.431MiB/s
Contended upload rate for 1 threads:    1694.556MiB/s
Contended upload rate for 2 threads:    736.089MiB/s
Contended upload rate for 2 threads:    735.497MiB/s
Contended upload rate for 4 threads:    766.770MiB/s
Contended upload rate for 4 threads:    715.844MiB/s
Contended upload rate for 8 threads:    628.011MiB/s
Contended upload rate for 8 threads:    640.349MiB/s
Contended upload rate for 16 threads:   615.857MiB/s
Contended upload rate for 16 threads:   622.288MiB/s
Contended upload rate for 32 threads:   599.360MiB/s
Contended upload rate for 32 threads:   606.371MiB/s
Contended upload rate for 64 threads:   584.327MiB/s
Contended upload rate for 64 threads:    80.699MiB/s
Test assertion failure function thread_contention, file gem_fence_upload.c:321:
Last errno: 0, Success
Failed assertion: tiled[1] > 0.75 * tiled[0]
Subtest thread-contention: FAIL
Download rate for 2 linear surfaces, 2 threads:  78.469MiB/s
Download rate for 2 tiled surfaces, 2 threads:   78.524MiB/s
Download rate for 4 linear surfaces, 4 threads: 153.073MiB/s
Download rate for 4 tiled surfaces, 4 threads:  153.146MiB/s
Download rate for 8 linear surfaces, 8 threads: 293.719MiB/s
Download rate for 8 tiled surfaces, 8 threads:  292.514MiB/s
Download rate for 16 linear surfaces, 16 threads:       293.454MiB/s
Download rate for 16 tiled surfaces, 16 threads:        292.741MiB/s
Download rate for 32 linear surfaces, 32 threads:       293.049MiB/s
Download rate for 32 tiled surfaces, 32 threads:        292.829MiB/s
Download rate for 64 linear surfaces, 64 threads:       293.125MiB/s
Download rate for 64 tiled surfaces, 64 threads:        135.928MiB/s
Subtest thread-performance-read: SUCCESS
Upload rate for 2 linear surfaces, 2 threads:   11184.900MiB/s
Upload rate for 2 tiled surfaces, 2 threads:    12052.731MiB/s
Upload rate for 4 linear surfaces, 4 threads:   7265.297MiB/s
Upload rate for 4 tiled surfaces, 4 threads:    7505.571MiB/s
Upload rate for 8 linear surfaces, 8 threads:   7456.166MiB/s
Upload rate for 8 tiled surfaces, 8 threads:    7576.654MiB/s
Upload rate for 16 linear surfaces, 16 threads: 7378.159MiB/s
Upload rate for 16 tiled surfaces, 16 threads:  7578.673MiB/s
Upload rate for 32 linear surfaces, 32 threads: 7383.054MiB/s
Upload rate for 32 tiled surfaces, 32 threads:  7504.800MiB/s
Upload rate for 64 linear surfaces, 64 threads: 7305.519MiB/s
Upload rate for 64 tiled surfaces, 64 threads:  184.805MiB/s
Test assertion failure function thread_performance, file
gem_fence_upload.c:246:
Last errno: 0, Success
Failed assertion: linear[1] > 0.75 * linear[0]
Subtest thread-performance-write: FAIL
Combined rate for 2 linear surfaces, 4 threads: 156.549MiB/s
Combined rate for 2 tiled surfaces, 4 threads:  156.109MiB/s
Combined rate for 4 linear surfaces, 8 threads: 304.679MiB/s
Combined rate for 4 tiled surfaces, 8 threads:  303.680MiB/s
Combined rate for 8 linear surfaces, 16 threads:        568.308MiB/s
Combined rate for 8 tiled surfaces, 16 threads: 562.881MiB/s
Combined rate for 16 linear surfaces, 32 threads:       572.849MiB/s
Combined rate for 16 tiled surfaces, 32 threads:        570.542MiB/s
Combined rate for 32 linear surfaces, 64 threads:       575.814MiB/s
Combined rate for 32 tiled surfaces, 64 threads:        570.089MiB/s
Combined rate for 64 linear surfaces, 128 threads:      573.400MiB/s
Combined rate for 64 tiled surfaces, 128 threads:       154.320MiB/s
Subtest thread-performance-both: SUCCESS

==Reproduce steps==
---------------------------- 
1. ./gem_fence_upload

-- 
You are receiving this mail because:
You are the QA Contact for the bug.
You are on the CC list for the bug.
You are the assignee for the bug.
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.freedesktop.org/archives/intel-gfx-bugs/attachments/20140616/2b4561f9/attachment-0001.html>