(moving this to gnulib) Bruno Haible <[email protected]> writes:
> Simon Josefsson wrote: >> >>>>>> Unconnected socket test... passed >> >>>>>> Connected sockets test... failed (expecting POLLHUP after shutdown) >> >>>>>> General socket test with fork... failed (expecting POLLHUP after >> >>>>>> shutdown) >> >>>>>> FAIL: test-poll >> ... >> we'll use the latest gnulib and hopefully this problem has been resolved > > No, no one has been working on the 'poll' test for 3 years. I still see this > failure regularly on many platforms. Sigh. Have we considered disabling the part of the test that doesn't seem to work reliable, pending that someone take a look at it? I sympathise with the preference to have the test fail in order to not forget about the problem, but I think this may have passed the point where it is counter productive since it causes confusion among users (this was not the first time I've seen this reported). I wouldn't be surprised if someone just disables running 'make check' when something like this happens. Alternatively, print "BUG: Please debug this" and exit(77) to have the test be SKIP'ed? /Simon
