> On Sept. 11, 2018, 11:14 p.m., Benjamin Mahler wrote: > > Looks good! We should remove our libev patch file now, since (1) it's > > confusing (there's no explanation of why we have it; reading it seems to > > suggest it's needed for correctness, but it's not) and (2) makes upgrading > > libev more challenging (need to potentially adjust the patch file). Can you > > take care of that?
I'll try ti unbundle libev altogether in [MESOS-895](https://issues.apache.org/jira/browse/MESOS-895) - James ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/68660/#review208534 ----------------------------------------------------------- On Sept. 7, 2018, 4:46 p.m., James Peach wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/68660/ > ----------------------------------------------------------- > > (Updated Sept. 7, 2018, 4:46 p.m.) > > > Review request for mesos, Benjamin Bannier, Till Toenshoff, and Vinod Kone. > > > Bugs: MESOS-9212 > https://issues.apache.org/jira/browse/MESOS-9212 > > > Repository: mesos > > > Description > ------- > > libev will consume SIGCHLD signals by default, which interferes > with a number of libprocess assumptions around dealing with child > processes. The recommended way to disable this behavior is to reset > the SIGCHLD signal handler after initializing the libeve default > event loop. > > > Diffs > ----- > > 3rdparty/libprocess/src/posix/libev/libev.cpp > 173ee466e61d3a754202213b94645e125ce0421a > > > Diff: https://reviews.apache.org/r/68660/diff/2/ > > > Testing > ------- > > configure with --with-libev=/usr, make check (Fedora 28) > > > Thanks, > > James Peach > >
