> On March 11, 2015, 9:38 p.m., Ben Mahler wrote: > > src/master/http.cpp, line 471 > > <https://reviews.apache.org/r/31700/diff/2/?file=888170#file888170line471> > > > > Do you know if a move is helpful here or if the compiler is already > > optimizing this? > > > > ``` > > object.values["slaves"] = std::move(array); > > ``` > > > > Don't do it in this change, we need to measure and I'm just curious :)
I believe the compiler won't optimize this, since `array` is an lvalue and may be used after this assignment. One thing how we can try to persuade the compiler to move without writing `std::move()` is to wrap array population in a function that returns an array, this will be a rvalue then and c++11-capable compiler *should* choose the right overload, that takes a rvalue ref: http://www.cplusplus.com/reference/map/map/insert/. But everything more complex than `.reserve()` should be benchmarked, on all official supported compilers if possible. I'll play with this when I have some free cycles. - Alexander ----------------------------------------------------------- This is an automatically generated e-mail. To reply, visit: https://reviews.apache.org/r/31700/#review76133 ----------------------------------------------------------- On March 7, 2015, 1:43 a.m., Alexander Rukletsov wrote: > > ----------------------------------------------------------- > This is an automatically generated e-mail. To reply, visit: > https://reviews.apache.org/r/31700/ > ----------------------------------------------------------- > > (Updated March 7, 2015, 1:43 a.m.) > > > Review request for mesos and Ben Mahler. > > > Bugs: MESOS-2353 > https://issues.apache.org/jira/browse/MESOS-2353 > > > Repository: mesos > > > Description > ------- > > See summary. > > > Diffs > ----- > > src/common/http.cpp 0b57fb01f7769031704e5341849bf95a0197ffd9 > src/master/http.cpp b8eef69505b147d4c8a0e005dff545b9fc12a220 > > Diff: https://reviews.apache.org/r/31700/diff/ > > > Testing > ------- > > make check (OS X 10.9.5, Ubuntu 14.04) > > > Thanks, > > Alexander Rukletsov > >
