[igt-dev] ✗ Fi.CI.IGT: failure for i915: Avoid set_domain -ENOMEM error with huge buffers

Dixit, Ashutosh ashutosh.dixit at intel.com
Tue Mar 30 05:44:11 UTC 2021


On Mon, 29 Mar 2021 22:27:46 -0700, Patchwork wrote:
>
> [1  <text/plain; utf-8 (7bit)>]
> [2  <text/html; utf-8 (quoted-printable)>]
> Project List - Patchwork
>
> Patch Details
>
>  Series:  i915: Avoid set_domain -ENOMEM error with huge buffers
>  URL:  https://patchwork.freedesktop.org/series/88561/
>  State:  failure
>  Details:  https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_5675/index.html
>
> CI Bug Log - changes from CI_DRM_9914_full -> IGTPW_5675_full
>
> Summary
>
> FAILURE
>
> Serious unknown changes coming with IGTPW_5675_full absolutely need to be
> verified manually.
>
> If you think the reported changes have nothing to do with the changes
> introduced in IGTPW_5675_full, please notify your bug team to allow them
> to document this new failure mode, which will reduce false positives in CI.
>
> External URL: https://intel-gfx-ci.01.org/tree/drm-tip/IGTPW_5675/index.html
>
> Possible new issues
>
> Here are the unknown changes that may have been introduced in IGTPW_5675_full:
>
> IGT changes
>
> Possible regressions
>
> * igt at perf@polling-parameterized:
>
>  * shard-iclb: PASS -> SKIP
>
>  * shard-tglb: PASS -> SKIP

This is false positive. This patch cannot affect this test. It can only
affect igt at prime_mmap and igt at gem_exec_params. Thanks.


More information about the igt-dev mailing list