----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/7781/#review13048 -----------------------------------------------------------
Ship it! Ship It! - Vinod Kone On Oct. 30, 2012, 7:39 p.m., Ben Mahler wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/7781/ > ----------------------------------------------------------- > > (Updated Oct. 30, 2012, 7:39 p.m.) > > > Review request for mesos, Benjamin Hindman and Vinod Kone. > > > Description > ------- > > This is part 1, where I'd like to make sure this is the right way function to > have. > > This way, we have a nice helper for aborting, like CHECK() or LOG(FATAL), > however these two: > 1. Print out a stack trace, which we may not want sometimes. > 2. Perform locking, and can't be used within a fork. > > Part 2, will refactor usage of: > std::cerr << "Cgroups can't be enabled." << std::endl; > abort(); > to become: > ABORT("Cgroups can't be enabled."); > > Part 3, will kill the old files: > stout/fatal.hpp > libprocess/src/fatal.hpp > libprocess/src/fatal.cpp > > > P.S. I looked at how glog implements it's style of CHECK() << "Foo" << "bar"; > It appears to call an abort immediately, but they have signal handlers that > then catch the abort and appear to wait if someone else is logging, and flush > the logs to ensure the remaining stuff sent to the check stream are flushed. > Needless to say, that's way more complexity than I'd like to have if we want > this to be standalone in stout. > > > Diffs > ----- > > third_party/libprocess/Makefile.am f32c184ca8cc8852b6c90f3c85a8f99e10ed781b > third_party/libprocess/include/stout/abort.hpp PRE-CREATION > > Diff: https://reviews.apache.org/r/7781/diff/ > > > Testing > ------- > > CentOS: make check > > > Thanks, > > Ben Mahler > >
