[12:53] <Laney> is there an ongoing incident with builds?
[12:54] <Laney> I'm seeing failures with no log like https://launchpad.net/ubuntu/+source/ubuntukylin-meta/0.20/+build/12391545
[13:06] <wgrant> Laney: Something's awry with lgw01. I'm manualling and investigating.
[14:45] <Laney> wgrant: righto
[18:57] <Odd_Bloke> I'm not seeing diffs for (git) MPs generated; is this a known issue?
[18:58] <Odd_Bloke> Oh, actually, I even got an OOPS for the MP I just submitted: OOPS-4ffe5668502cebd72ed22668e04934b0
[19:24] <cjwatson> Odd_Bloke: fix/workaround requested in #webops internal; you can keep an eye on it there.  will probably require a re-push once that service is restarted (though it could just be a quick create/delete of a random branch, say)
[19:27] <cjwatson> Odd_Bloke: should be OK now
[19:30] <Odd_Bloke> cjwatson: Looks good now, thanks!
[20:49] <kami> Good evening.
[20:50] <kami> What can I do if my PPA package doesn't build (one test out of hundreds fails) although pbuilder can build locally?
[20:51] <mapreri> does it do network?  IME that's one of most common cases of test failing in lp but not in pbuilder (assuming you have an old pbuilder, or explicitly enabled networking in pbulderrc)
[20:52] <mapreri> otherwise, try to debug as appropriate for your package...  is not like you can get suggestions with so few details.
[20:53] <kami> mapreri: no, it's a concurrency test where a read/write lock is acquired for a critical section in the code. Multiple threads reading/writing to some data structure.
[20:54] <mapreri> then no ideas, sorry
[20:55] <kami> mapreri: is there something special about multithreading on the builders?
[20:56] <mapreri> not really, afaik.
[20:56] <dobey> no
[20:57] <dobey> kami: please paste a link to the build with the failed test
[20:57] <kami> https://launchpad.net/~darabi/+archive/ubuntu/lisp/+build/12393406
[20:59] <kami> Searching for "Test SB-CONCURRENCY-TEST::FRLOCK.1 failed" shows the place where it fails
[21:00] <dobey> kami: i see the package is "sbcl"; does this require OpenCL to be set up and working, for the test to pass?
[21:01] <kami> dobey: no, they have no relation. sbcl is a common lisp implementation.
[21:01] <dobey> oh ok
[21:02] <kami> dobey: and the code which is being executed has probably only a dependency on libpthread (I didn't check, yet)
[21:03] <dobey> not sure why that that test would be failing exactly
[21:04] <kami> I thought there might be something related to threads, well known among LP build pros. And some kind person here would point me to it.
[21:04] <kami> :)
[21:04] <kami> Thank you for your time.
[21:04] <kami> I will go and ask in #sbcl
[21:04] <kami> Good night.