RFR: 2454: Skara mistakenly marked a non-clean backport as clean [v3]

Erik Joelsson erikj at openjdk.org
Wed Mar 12 23:17:17 UTC 2025


On Wed, 12 Mar 2025 22:30:36 GMT, Zhao Song <zsong at openjdk.org> wrote:

>> Zhao Song has updated the pull request incrementally with two additional commits since the last revision:
>> 
>>  - fix copyright
>>  - update
>
> forge/src/main/java/org/openjdk/skara/forge/gitlab/GitLabRepository.java line 758:
> 
>> 756:                 // The diff here is only used to evaluate if a backport PR is clean or not,
>> 757:                 // assuming the diff is complete will not introduce any side effect.
>> 758:                 diffs = List.of(toDiff(metadata.parents().get(0), hash, diff, true));
> 
> I thought the diff is only limited by the count of files. But I was wrong.
> The diff is limited by 3 factors. So it's hard for us to know if it's limited or not.
> https://docs.gitlab.com/administration/diff_limits/#configure-diff-limits

I think this is fine. We could perhaps sum files and patch size and compare to the default values (200KB and 1000 files) and at least get a reasonable chance and flagging incomplete when it is.

-------------

PR Review Comment: https://git.openjdk.org/skara/pull/1708#discussion_r1992416009


More information about the skara-dev mailing list