when we build with -Dbuildtype=release optimization and run with log level = default:debug logging, the spectest execution fails in GitHub CI
https://github.com/Netatalk/netatalk/actions/runs/29505709658/job/87645788427
the failures are deterministically in the Error suite, during which the execution bails out, but I think this is a red herring and there's some kind of race condition or DSI synchronization issue that happens earlier:
Executing testset: Error_test
Error:test35: illegal volume/desktop ref - PASSED
Error:test36: no folder error (ERR_NOOBJ) - PASSED
Error:test37: no folder error ==> ERR_NOOBJ - PASSED
Error:test95: exchange files - NOT TESTED
Error:test99: test folder without access right - NOT TESTED
Error:test100: no obj cname error (AFPERR_NOOBJ) - PASSED
Error: Process completed with exit code 141.
for reference, this is the commit that works around the issue: 03903d7
when we build with
-Dbuildtype=releaseoptimization and run withlog level = default:debuglogging, the spectest execution fails in GitHub CIhttps://github.com/Netatalk/netatalk/actions/runs/29505709658/job/87645788427
the failures are deterministically in the Error suite, during which the execution bails out, but I think this is a red herring and there's some kind of race condition or DSI synchronization issue that happens earlier:
for reference, this is the commit that works around the issue: 03903d7