Hi Sergio, On Fri, Oct 21, 2022 at 01:43:10PM -0400, Sergio Durigan Junior wrote: > Source: debugpy > Version: 1.6.3+ds-1 > Severity: serious > > Hi, > > debugpy is currently FTBFSing: > [...] > > The long summary is too big to post here, but you can see the full log > at: > > > https://tests.reproducible-builds.org/debian/rb-pkg/unstable/amd64/debugpy.html > [...] > I found a few upstream bugs that contain either the "Timed out waiting > for debug server to connect" or the "Address already in use" messages, > but I haven't investigated further TBH.
Sorry for the slow reply. That's interesting, because it built fine on the buildd network and using lxc. If it's "address already in use", it's probably because there are timing issues: it's running the tests on multiple CPUs and if it tries to spin up local servers on two CPUs almost simultaneously, they might end up trying to use the same port. I saw this same issue with pydevd, and worked around it by restricting it to using a single CPU; maybe I should do the same here. If it's "timed out", I'm not sure what the cause is; might it possibly be that pbuilder does not allow servers to start on local ports or for the build to connect to local ports? I had to work around this in pydevd's debian/rules, but it didn't seem to be a problem with debugpy. Maybe it is. Ho hum, something to investigate! Best wishes, Julian
