Re: ✗ Fi.CI.BUILD: failure for Revert "tests: save GPU engine information more properly"

[Peter Senna Tschudin]

Dear List,

This is how I tested my patch before sending:
 $ cd /tmp
 $ git clone https://gitlab.freedesktop.org/drm/igt-gpu-tools.git
 $ git am --keep-cr --signoff <
0001-Revert-tests-save-GPU-engine-information-more-proper.patch
 $ git show HEAD
    commit 990f3b51d3296541a1b680c508be7cae42e97bfe (HEAD -> master)
    Author: Peter Senna Tschudin <me at petersenna.com>
    Date:   Fri Mar 22 16:28:44 2024 +0100
    Revert "tests: save GPU engine information more properly"
 $ meson build && ninja -C build tests/gem_exec_capture tests/kms_busy
 $ sudo ./build/tests/gem_exec_capture --run-subtest many-4k-zero
    ...
    Subtest many-4K-zero: SUCCESS (8.404s)
 $ sudo ./build/tests/gem_exec_capture --run-subtest many-4k-incremental
    ...
    Subtest many-4K-incremental: SUCCESS (22.667s)

However trybot could not apply my patch, claiming that "error: sha1
information is lacking or useless (lib/i915/gem_engine_topology.c).
error: could not build fake ancestor". When writing the commit
message, I was careful to get the proper short commits from git
rev-parse, but I am unsure what the actual problem is here. The commit
that the patch reverts is:

https://gitlab.freedesktop.org/drm/igt-gpu-tools/-/commit/93c5ec2105500f7083d0cb50db3fbb3bca3776bb

An important reason why I sent this to trybot is to see if trybot can
reproduce the issue 93c5ec210 was fixing. I tried to reproduce the
issue on an N3150, which is used by some Chromebooks, but I could not.
There are more details on the commit message:

How can I make my patch trybot friendly?

Here is a copy of the commit message:
----
Subject: [PATCH] Revert "tests: save GPU engine information more properly"

This reverts commit 93c5ec210 that was intended to fix:
 https://gitlab.freedesktop.org/drm/igt-gpu-tools/-/issues/147

Reverting this patch fixes the following tests:
 - gem_exec_capture at many-4k-incremental
 - gem_exec_capture at many-4k-zero

I could not reproduce issue 147, even when taking 9e635a1c5029 into
account. I tested it on an N3150 and I could not trigger the issue
trough the kms_busy test. I tested with 6.8.0-next-20240320, 5.14.18,
and 5.10.16.

Tested on an i53210M and on an N3150.

Signed-off-by: Peter Senna Tschudin <peter.senna at gmail.com>
----

Thank you,

Peter


On Sat, Mar 23, 2024 at 4:33 AM Patchwork
<patchwork at emeril.freedesktop.org> wrote:
>
> == Series Details ==
>
> Series: Revert "tests: save GPU engine information more properly"
> URL   : https://patchwork.freedesktop.org/series/131506/
> State : failure
>
> == Summary ==
>
> Error: patch https://patchwork.freedesktop.org/api/1.0/series/131506/revisions/1/mbox/ not applied
> Applying: Revert "tests: save GPU engine information more properly"
> error: sha1 information is lacking or useless (lib/i915/gem_engine_topology.c).
> error: could not build fake ancestor
> hint: Use 'git am --show-current-patch=diff' to see the failed patch
> Patch failed at 0001 Revert "tests: save GPU engine information more properly"
> When you have resolved this problem, run "git am --continue".
> If you prefer to skip this patch, run "git am --skip" instead.
> To restore the original branch and stop patching, run "git am --abort".
> Build failed, no error log produced
>
>