From 830fd3e7aabc451395c5363d87b7163621585fd1 Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Pablo=20Correa=20G=C3=B3mez?= Date: Sat, 15 Aug 2026 17:42:23 +0200 Subject: [PATCH] ci: do not fail kconfigcheck if the merge bot is the committer MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit This is not ideal, but while our kconfig keeps being very dynamic, we very often bump into the following situation: Author of kernel MR fixes kconfig -> the MR is fine -> 1 week passes and some other kconfig changes get merged -> we assign to the bot for merge -> CI fails and we're back at the beginning This means the authors have to get back to fixing kconfig out of us being slow. And if we are slow again, the same cycle can repeat, which has happened in the past. This is frustrating for everybody. Instead, if we are ready to merge and the merge bot is the committer, just skip kconfigcheck failures. It means that some options might have to wait for the next update to be included, but still forces authors to fix kconfig on each of their own pushes, which is a possible trade-off to reduce friction. This should be save to revert the day kconfigcheck starts changing only occassionally. Signed-off-by: Pablo Correa Gómez Part-of: --- .ci/lib/kconfig.py | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/.ci/lib/kconfig.py b/.ci/lib/kconfig.py index 0105df783..75bc160e7 100755 --- a/.ci/lib/kconfig.py +++ b/.ci/lib/kconfig.py @@ -92,6 +92,10 @@ if __name__ == "__main__": elif common.commit_message_has_string("[ci:skip-kconfigcheck]"): print("WARNING: not checking kernel configs ([ci:skip-kconfigcheck])") sys.exit(0) + elif common.all_committed_by_merge_bot(): + # This can removed once we manage to stabilize kconfigcheck + print("All commits in branch committed by merge bot, skipping check") + sys.exit(0) last_failed = check_kconfig(pkgnames)