[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Bug#1031128: release.debian.org: please suppress ppc64el and s390x CI tests for src:rust-ureq



Control: severity -1 normal
Control: tags -1 moreinfo

Hi Jonas,

On 12-02-2023 09:19, Jonas Smedegaard wrote:
Package: release.debian.org
Severity: important
            ^^^^^^^^^
Please don't do this with bugs against release.debian.org. It makes our flow worse instead of helping.

The package src:rust-urew is currently blocked from entering testing.

I suspect you typoed (twice, below as well) and you meant rust-ureq (as in the subject).

CI tests are generally not completing, as tracked as bug#1030883.

Ack. That bug has a fix pending, let's wait until its fixed and we can confirm all other architectures are OK (hence the moreinfo tag). You can easily work around this yourself by uploading a new version without testnames that start with a hyphen.

For two arches, ppc64el and s390x, CI tests now complete which reveals
that on those arches the CI infrastructure bogusly thinks that a
dependency on the virtual package
librust-rustls-0.20+dangerous-configuration-dev should be resolved as
arch-specific package which is not true for more than a week now.

Please therefore mark rust-urew as ignore CI tests for ppc64el and
s390x since those test environments are apparently broken.

Please don't use that tone. The test environments are NOT broken, it's your package that can't be installed. If anything, it's your package that's "broken" on those architectures. As I explained in bug #1030957 and you acknowledged, librust-rustls-dev depend on librust-ring-*-dev which isn't available everywhere. Nothing really broken, just regular behavior of arch:all binaries that have arch:non-all (transverse) dependencies that aren't available everywhere. As mentioned earlier, you brought this regression onto yourself.

If you decide to fix all this yourself by uploading a new version with the testname changed and the skip-not-installable restriction added back, then please close this bug, as then you don't need to wait for the fix in debci.

Paul

Attachment: OpenPGP_signature
Description: OpenPGP digital signature


Reply to: