Hi, On Sun, Sep 21, 2025 at 09:00:00AM +0300, Alexander Lakhin wrote: > 02.07.2025 02:53, Michael Paquier wrote: > > Catching up on this thread after-the-fact, specifically looking at > > 29213636e6cd as I did the original check tweaked here for O_RDONLY. > > Agreed that a backpatch should be OK as done. The buildfarm looks OK > > currently. > > Three months later we can see a number of failures produced by that > animal on several branches, e.g. [1]: > timed out after 3600 secs
Right, I've noticed them as well of course, but did not have time to take a closer look yet. This timeout in test_shm_mq happens on 32bit hurd-i386 as well, btw. > It's not that easy to see a backtrace of the running processes on that OS, > but with some debug logging (attached), I can see that the test backend > process or shm_mq background worker just gets stuck shortly after poll(). Thanks for taking a deeper look. > So it seems to me that Hurd is not mature enough yet to test Postgres. That is a bit harsh; this issue should be looked into, but I would not say it is not mature enough to test Postgres. Michael
