> On Sept. 22, 2015, 12:37 a.m., Jie Yu wrote: > > src/linux/perf.cpp, lines 516-529 > > <https://reviews.apache.org/r/37540/diff/7/?file=1077025#file1077025line516> > > > > Should we use `__sync_synchronize` instead?
Hmm, __sync_synchronize is documented as a full memory barrier, while we only need a read memory barrier here... but this isn't a very hot path, so I can replace it. - Cong ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/37540/#review99882 ----------------------------------------------------------- On Sept. 18, 2015, 10:11 p.m., Cong Wang wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/37540/ > ----------------------------------------------------------- > > (Updated Sept. 18, 2015, 10:11 p.m.) > > > Review request for mesos, Chi Zhang, Ian Downes, Jie Yu, and Vinod Kone. > > > Bugs: MESOS-2769 > https://issues.apache.org/jira/browse/MESOS-2769 > > > Repository: mesos > > > Description > ------- > > Abstract Linux kernel perf event API and provide API to collect schedule > events. > > > Diffs > ----- > > src/linux/perf.hpp d10968ca670eb516bae08385fd0ddde8e8ad83b5 > src/linux/perf.cpp f7035ddb2507a7646d88dd517d048018f695448a > src/tests/containerizer/perf_tests.cpp > ed5212ee31b8aa47339b8b8fab184bbdf85be82a > > Diff: https://reviews.apache.org/r/37540/diff/ > > > Testing > ------- > > make check > > > Thanks, > > Cong Wang > >
