On Mon, 8 Jan 2024 13:23:00 -0800 Ross Vandegrift <r...@kallisti.us> said:
> On Mon, Jan 08, 2024 at 06:53:09PM +0000, Carsten Haitzler wrote: > > i'm going to guess something is missing in the env? missing hmm dbus daemon? > > something else? some env missing like xdg runtime dir ... or something? but > > eina test suite should not need all of that sho it shouldnt be timing out - > > if you just run e.g. the eina suite e.g. > > > > ./src/tests/eina/eina_suite > > > > what happens? it should complete in almost no time (like a fraction of a > > second)... > > Yea, on their own they run fine - as long as I don't pipe the output to > another program (see my original message for more of what happens in > that case) *ANY* program ??? if i pipe through grep xxx ... they exit. if i pipe through wc -l ... they exit... is it something specific to what it's being piped through somehow not handling reads right? or something else? the whole check infra should be handling reading the stdout/err log and storing the last several lines so it can dump them and then sensibly handling the read of all of this... perhaps it's some oddity in your build of check (libcheck.so i am thinking) or something it depends on (which happens to be just libc related stuff) - special build option? i have check 0.15.2 - but this is arch. > I'm running in a normal X session, so dbus should be fine. I have check > 0.15.2, which looks like the newest. I'll keep digging. we have had things fail before because of .e.g. a partial "fdo" environment - missing some XDG_ env vars. also missing "common expected services" like dbus - but in this case since even the eina suit is failing... it's lower level. but... i don't see the hangs myself. so there is some difference we have that's interjecting. i'm pretty sure it's not efl as such as well - all the tests are the same kind of things and have not changed in a long time. they don't do anything funky themselves to stdout/err (they flush but don't like fcntl them in any way or do their own select/epoll loops listening on stdin or something). they do all wait for their child forks to return as to run the checks a lot of things a`re forked off ... but why would directing to file vs to another process change this? try run the above eina test suite and pipe to something that makes it timeout... and strace it - or gdb attach to it and find out where it's sitting? it should complete in < 1 sec so launch and immediately try and strace and/or gdb attach and find out where it's at - if it is still around. is somehow a forked child not coming back that it expects to... ? -- ------------- Codito, ergo sum - "I code, therefore I am" -------------- Carsten Haitzler - ras...@rasterman.com _______________________________________________ enlightenment-devel mailing list enlightenment-devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/enlightenment-devel
