Hi Mario! Welp... that's a bummer. I was sure it was a timing issue with the tests.
I'll keep poking at it on different machines to see what I'm missing. Thanks for giving it a try! Cheers! El 4 de noviembre de 2021 4:16:46 p. m. GMT-03:00, Mario Domenech Goulart <ma...@parenteses.org> escribió: >Hi Ariela, > >On Sun, 31 Oct 2021 17:18:45 -0300 Ariela Wenner <aris...@disroot.org> wrote: > >> So, here's a followup >> >> I linked pthreads and added the things you suggested to the wiki. >> >> As for the hangs, I hate to say this but it seems like the test was the >> problem. >> More precisely, it seems like the topic for the subscriber socket wasn't set >> fast enough, so when the publisher socket sent a message, the subscriber >> wasn't >> ready for it. Hang. >> >> I added some coordination in the test itself and ran the tests over 100 times >> just to be sure. No hang. Maybe I'm just being extremely unlucky though, so >> as >> always feedback is appreciated. >> >> Just pushed a new tag with the changes, fingers crossed. > >Thanks. > >I'm sad to report, though, that tests of 0.2.1 still hang here. > >I'm testing it with the following command: > > $ test-new-egg nng > https://gitlab.com/ariSun/chicken-nng/-/raw/main/nng.release-info > >I executed the command above three times and tests hanged consistently. > >I'm using strace to verify that the test process is not actually doing >anything: > >$ strace -p 14836 >strace: Process 14836 attached >restart_syscall(<... resuming interrupted poll ...> > >All the best. >Mario >-- >http://parenteses.org/mario -- Ariela Wenner
