Bug 101555

(In reply to Marta Löfstedt from comment #0)
> igt@kms_pwrite_crc regressed to fail between:
> 
> http://intel-gfx-ci.fi.intel.com/archive/results/CI_IGT_test/CI_DRM_440/
> http://intel-gfx-ci.fi.intel.com/archive/results/CI_IGT_test/CI_DRM_443/
> 
> http://intel-gfx-ci.fi.intel.com/archive/results/CI_IGT_test/
> igt@kms_pwrite_crc.html
> 
> I have also reproduced this on BDW NUCi5.

Hello Marta,
Does this fail happen every time or do you have any special configuration or specific steps for it to fail??
Thank you.

(In reply to Elizabeth from comment #1)
> (In reply to Marta Löfstedt from comment #0)
> > igt@kms_pwrite_crc regressed to fail between:
> > 
> > http://intel-gfx-ci.fi.intel.com/archive/results/CI_IGT_test/CI_DRM_440/
> > http://intel-gfx-ci.fi.intel.com/archive/results/CI_IGT_test/CI_DRM_443/
> > 
> > http://intel-gfx-ci.fi.intel.com/archive/results/CI_IGT_test/
> > igt@kms_pwrite_crc.html
> > 
> > I have also reproduced this on BDW NUCi5.
> 
> Hello Marta,
> Does this fail happen every time or do you have any special configuration or
> specific steps for it to fail??
> Thank you.

Hi Elizabeth,

This fails 100% on my BDW NUCi5.

/Marta

The following test fail on IVB with latest configuration

====================================================
Test list
====================================================
igt@kms_pwrite_crc


Graphic Stack
====================================================
Component: drm
    tag: libdrm-2.4.81-24-g3095cc8
    commit: 3095cc8eaba1aa87ad38c04ae2b1eabe30f7e16c
Component: cairo
    tag: 1.15.6-2-g57b4050
    commit: 57b40507dda3f58dfc8635548d606b86dc7bcf51
Component: intel-gpu-tools
    tag: intel-gpu-tools-1.19-57-g6fcc8e8
    commit: 6fcc8e8b247661c7950b998e0b95141ffbd6b833
Component: piglit
    tag: piglit-v1
    commit: c8f4fd9eeb298a2ef0855927f22634f794ef3eff
======================================
             Hardware
======================================
platform                   : Ivy Bridge
motherboard model          : Inspiron660
motherboard id             : 0XR1GT
form factor                : Desktop
manufacturer               : DellInc.
cpu family                 : Core i5
cpu family id              : 6
cpu information            : Intel(R) Core(TM) i5-3330 CPU @ 3.00GHz
gpu card                   : Intel Corporation Xeon E3-1200 v2/3rd Gen Core processor Graphics Controller (rev 09) (prog-if 00 [VGA controller])
memory ram                 : 7.7 GB
max memory ram             : 8 GB
cpu thread                 : 4
cpu core                   : 4
cpu model                  : 58
cpu stepping               : 9
socket                     : Socket BGA1155
signature                  : Type 0, Family 6, Model 58, Stepping 9
hard drive                 : 74GiB (80GB)
current cd clock frequency : 400000 kHz
maximum cd clock frequency : 400000 kHz
displays connected         : VGA-1 HDMI-A-1

issue is still present over BDW with the following configuration 

test
=====================================
igt@kms_pwrite_crc

 Software
======================================
kernel version              : 4.12.0-drm-tip-ww28-commit-f1619bb+
architecture                : x86_64
os version                  : Ubuntu 16.10
os codename                 : yakkety
kernel driver               : i915
bios revision               : 5.6
hardware acceleration       : disabled
swap partition              : enabled on (/dev/sda3)

======================================
        Graphic drivers
======================================
modesetting                 : enabled
modesetting compiled for    : 1.18.4 X.Org Video Driver
xorg-xserver                : 1.18.4
libdrm                      : 2.4.70
libva                       : 1.7.1-2
vaapi (intel-driver)        : 1.7.1
cairo                       : 1.14.6-1build1
intel-gpu-tools             : 1.16-1

======================================
             Hardware
======================================
platform                   : Broadwell
motherboard id             : NUC5i5MYBE
form factor                : Desktop
cpu family                 : Core i5
cpu information            : Intel(R) Core(TM) i5-5300U CPU @ 2.30GHz
gpu card                   : Intel Corporation HD Graphics 5500 (rev 09) (prog-if 00 [VGA controller])
memory ram                 : 11.62 GB
max memory ram             : 16 GB
signature                  : Type 0, Family 6, Model 61, Stepping 4
hard drive                 : 111GiB (120GB)
current cd clock frequency : 337500 kHz
maximum cd clock frequency : 540000 kHz
displays connected         : DP-1 DP-2

The following test FAIL on IVB with this commit

Tests List:

igt@kms_pwrite_crc


This is my configuration:

====================================================
Graphic Stack
====================================================
Component: drm
    tag: libdrm-2.4.81-33-g3876bc2
    commit: 3876bc246a07070a6043159cd7623d4def9bbd4c

Component: cairo
    tag: 1.15.6-2-g57b4050
    commit: 57b40507dda3f58dfc8635548d606b86dc7bcf51

Component: intel-gpu-tools
    tag: intel-gpu-tools-1.19-112-g493151b
    commit: 493151b0768aa4ca535cef49cb7efa174a9c3a77

Component: piglit
    tag: piglit-v1
    commit: 973892687cf5c2f8e2dbe1d22998b82736643787

Created attachment 133437 [details]
artifacts

I added one second sleep between all flips in the tests. Attached image is from the last flip. I would classify that as caching artefacts.

The issue is not reproduced with the same solution as for BUG 101109

commit b8f55be64453ea77fc51bff6cd0d906d18ce1cd2
Author: Chris Wilson <chris@chris-wilson.co.uk>
Date:   Fri Aug 11 12:11:16 2017 +0100

    drm/i915: Split obj->cache_coherent to track r/w
    
    Another month, another story in the cache coherency saga. This time, we
    come to the realisation that i915_gem_object_is_coherent() has been
    reporting whether we can read from the target without requiring a cache
    invalidate; but we were using it in places for testing whether we could
    write into the object without requiring a cache flush. So split the
    tracking into two, one to decide before reads, one after writes.
    
    See commit e27ab73d17ef ("drm/i915: Mark CPU cache as dirty on every
    transition for CPU writes") for the previous entry in this saga.

Thanks Chris and Maarten I can no longer reproduce the issue on my BDW NUCi5.
This is also reflected on the shards now passing the test from CI_DRM_2968: 
https://intel-gfx-ci.01.org/tree/drm-tip/shards-all.html