Pre-Merge Results

A handy overview of the emails with results sent by i915 CI.

The Sections

Test Results

• Possible new issues - issues that are not tracked by cibulog and seem to be caused by the patch series, if present the result is failure
• Known issues - issues that occurred during the run but are known and tracked, they do not affect the result for the series
• Possible fixes - issues that occurred in the base run (post-merge that serves as a baseline for comparison) but were not seen in the pre-merge run for the series
• Suppressed - content that is not affecting the end result because the machine/configuration is either too new or too unstable, they are still listed for completeness

See Understanding The Test Results for details on how to read failures/fixes.

Participating Hosts

This sections how many hosts participated in the run and the delta in comparison to the base run.

Small fluctuations are expected, as we deal with rebooting and AC-cutting physical machines - some of the host require manual intervention from time to time or may take longer than expected to boot due to automated fscks.

When assessing impact of series look for large deltas and/or patterns (e.g. half of the machines missing, or all GEN8 machines missing).

Participating hosts (52 -> 47)
------------------------------

  Additional (2): fi-byt-j1900 fi-bsw-kefka
  Missing    (7): fi-ilk-m540 fi-hsw-4200u fi-byt-squawks fi-bsw-cyan fi-ctg-p8600 fi-byt-clapper fi-bdw-samus 

Build Changes

This section lists the delta in components deployed on the machines doing the testing.

You should see only one component change at a time (e.g. post-merge kernel changed to the pre-merge) unless you are testing combined IGT and kernel changes.

Build changes
-------------

  * Linux: CI_DRM_6039 -> Patchwork_12960

  CI_DRM_6039: 941848427de77537b5709bd68ca4a4048be5b5d9 @ git://anongit.freedesktop.org/gfx-ci/linux
  IGT_4972: f052e49a43cc9704ea5f240df15dd9d3dfed68ab @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
  Patchwork_12960: 824b9e2304e344e55b1fea49f624c7b2ab0725bf @ git://anongit.freedesktop.org/gfx-ci/linux

See Reproducing The Tree on details how to read the SHA1/repos.

Understanding The Test Results

* igt@i915_suspend@debugfs-reader:
 - shard-apl:          [PASS][5] -> [DMESG-WARN][6] ([fdo#108566]) +3 similar issues
 [5]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6039/shard-apl5/igt@i915_suspend@debugfs-reader.html
 [6]: https://intel-gfx-ci.01.org/tree/drm-tip/Patchwork_12960/shard-apl5/igt@i915_suspend@debugfs-reader.html

[fdo#108566]: https://bugs.freedesktop.org/show_bug.cgi?id=108566

Each test state change (e.g. pass→fail) has few things listed:

• affected test: e.g. igt@i915_suspend@debugfs-reader
• affected machine(s): e.g. shard-apl
• state transition: [PASS][5] -> [DMESG-WARN][6]
• linked result details: e.g.: [5]: https://intel-gfx-ci.01.org/tree/drm-tip/CI_DRM_6039/shard-apl5/igt@i915_suspend@debugfs-reader.html
• associated bugs: ([fdo#108566])
• numer of occurences of same transition across multiple machines: +3 similar issues

How Serious Are Incompletes And Aborts?

Incomplete means that the machine haven’t been able to complete execution of a test (most often because of a kernel panic or a hard hang). Incompletes signal very serious issues with stability. Any tests that were supposed to follow, that are on the same testlist (shard, fast-feedback), are not going to be executed.

Aborts are as serious as incompletes but the abnormal state was detected either by IGT Runner or the test itself. They signal that no further testing is going to be done because the results would not be trustworthy do to the machine’s / kernel condition.

I’ve Got Some False Positives, What Now?

Reply to the e-mail with the results and explain why do those results are false positive in as much detail as you can.

Include the following people in the CC:

• "Lakshminarayana Vudum" <lakshminarayana.vudum@intel.com>

Reproducing The Tree

The easiest way is using “build changes” section.

Build changes
-------------

    * Linux: CI_DRM_5951 -> Patchwork_12825

  CI_DRM_5951: 08cf1213114f21fc0b1a17eac081a2db8c03311f @ git://anongit.freedesktop.org/gfx-ci/linux
  IGT_4956: 1d921615b0b706f25c856aa0eb096f274380c199 @ git://anongit.freedesktop.org/xorg/app/intel-gpu-tools
  Patchwork_12825: f3e8eb0c781a6a88cd6f67a64f4b93e132cb37fe @ git://anongit.freedesktop.org/gfx-ci/linux
  piglit_4509: fdc5a4ca11124ab8413c7988896eec4c97336694 @ git://anongit.freedesktop.org/piglit

Each component here has three things mentioned: it’s name (which is also a tag), SHA1 and git remote where given SHA1 and/or tag can be found.

$ git remote -v
ci-tags git://anongit.freedesktop.org/gfx-ci/linux (fetch)
ci-tags git://anongit.freedesktop.org/gfx-ci/linux (push)

$ git fetch --tags ci-tags

$ git tag | grep Patchwork_12825
intel/Patchwork_12825

$ git checkout intel/Patchwork_12825
Note: checking out 'intel/Patchwork_12825'.

You are in 'detached HEAD' state. You can look around, make experimental
changes and commit them, and you can discard any commits you make in this
state without impacting any branches by performing another checkout.

If you want to create a new branch to retain commits you create, you may
do so (now or later) by using -b with the checkout command again. Example:

  git checkout -b <new-branch-name>

HEAD is now at f3e8eb0c781a drm/i915: Add transcoder parameter to PSR registers macros

For the Kernel, if you want to check which upstream relase the tree is based on you can fetch Linus’ tags and use git describe to find the parent annotated tag (CI tags are not annotaded).

$ git remote add linus git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux.git

$ git fetch --tags linus

$ git describe intel/Patchwork_12825
v5.1-rc5-1414-gf3e8eb0c781a

NOTE: IGT pre-merge tagging is still WIP. For now use SHA1 on the main repository. In the meantime checkout the IGT repository to SHA1 mentioned by the IGT_ component and apply the mbox downloaded from the patchwork.