On 2026-02-15 18:34, Lorenzo wrote:
Hello Mentors,
migration of runit [1] (unstable --> testing) is delayed and the only
reason that I see is that autopkgtest is triggered (but not run yet,
after 6 days) on riscv64.
Is that correct?
Best,
Lorenzo
[1]https://qa.debian.org/excuses.php?package=runit
Dive a little deeper into the specific autopkgtests where it says "tests
triggered" but then "tmpfail"ing and you can get information on this.
The runit autopkgtest triggered by the upload is here, it has a tmpfail:
https://ci.debian.net/packages/r/runit/testing/riscv64/68563755/
The runit-services autopkgtest triggered by the upload is here, it has a
tmpfail too:
https://ci.debian.net/packages/r/runit-services/testing/riscv64/68563756/
Both tmpfails look to be testbed failures:
autopkgtest [09:21:58]: ERROR: testbed failure: eof from the
virtualisation server
I'm not sure I have access, but I did queue a request into the system
for runit and runit-services based on the britney trigger for
runit/2.3.0-1 on riscv64. It may or may not show up, but we'll see if
it ends up in the queue (may take a while).
The system did say there may be another pending request already for it,
and in that case patience is needed if the builders are just *busy*.
It's my experience that riscv64 can be a bit slower for general
autopkgtesting overall (from Debian and Ubuntu autopkgtest experiences
with riscv64), so some of this may just be "patience required".
Thomas
Thomas
