tag 424241 confirmed thanks On Wed, May 16, 2007 at 12:04:35AM +0200, Martin Zobel-Helas wrote: > Package: speedy-cgi-perl > Severity: important > Version: 2.22-6 > User: [EMAIL PROTECTED] > Usertags: qa-doublebuild
> Lucas Nussbaum has rebuilt the whole archive on i386 and your package FTBFS if > built twice in a row with the following error: > /usr/bin/make -C speedy test > make[1]: Entering directory `/build/user/speedy-cgi-perl-2.22/speedy' > PERL_DL_NONLAZY=1 SPEEDY=/build/user/speedy-cgi-perl-2.22/speedy/speedy > SPEEDY_BACKENDPROG=/build/user/speedy-cgi-perl-2.22/speedy_backend/speedy_backend > SPEEDY_MODULE=/build/user/speedy-cgi-perl-2.22/mod_speedycgi/ > SPEEDY_TIMEOUT=300 /usr/bin/perl "-MExtUtils::Command::MM" "-e" > "test_harness(0, '../blib/lib', '../blib/arch')" t/*.t > t/alarm...............semop(1): encountered an error: Identifier removed > ok > t/argv................ok > t/basic1..............ok > t/basic2..............ok > t/be_memleak..........ok > t/begin_block.........semop(1): encountered an error: Identifier removed > FAILED test 2 [...] This happens because 'make test' leaves lingering speedy_backend processes behind, and they break the second run of tests. I'm not sure yet how to fix this. The options I see are: - something like 'killall speedy_backend' after 'make test': this is potentially intrusive in case there are other speedy processes in the same host - lowering SPEEDY_TIMEOUT (currently 300 seconds) so the speedy_backend processes will exit earlier by themselves. Finding the "right" value is not trivial since the speed of the buildds varies significantly. Furthermore, I'm not sure if any delay at all is acceptable here. - dropping the tests altogether. I'd like to avoid this if possible. I'll follow up when I have decided on a solution. Cheers, -- Niko Tyni [EMAIL PROTECTED] -- To UNSUBSCRIBE, email to [EMAIL PROTECTED] with a subject of "unsubscribe". Trouble? Contact [EMAIL PROTECTED]
