Re: r276716 - [Coverage] Do not write out coverage mappings with zero entries

2016-07-26 Thread Hans Wennborg via cfe-commits
Sure, r276801. Thanks, Hans On Tue, Jul 26, 2016 at 1:41 PM, Vedant Kumar wrote: > Thanks Justin! > > @Hans, could we take this for llvm 3.9 also? > > best, > vedant > >> On Jul 26, 2016, at 1:39 PM, Justin Bogner wrote: >> >> Vedant Kumar via cfe-commits writes: >>> Hi Justin, >>> >>> Could y

Re: r276716 - [Coverage] Do not write out coverage mappings with zero entries

2016-07-26 Thread Vedant Kumar via cfe-commits
Thanks Justin! @Hans, could we take this for llvm 3.9 also? best, vedant > On Jul 26, 2016, at 1:39 PM, Justin Bogner wrote: > > Vedant Kumar via cfe-commits writes: >> Hi Justin, >> >> Could you take a look at this code coverage fix? >> >> It skips visiting decls if they would require cont

Re: r276716 - [Coverage] Do not write out coverage mappings with zero entries

2016-07-26 Thread Justin Bogner via cfe-commits
Vedant Kumar via cfe-commits writes: > Hi Justin, > > Could you take a look at this code coverage fix? > > It skips visiting decls if they would require context from system headers to > present properly. This prevents us from writing out an empty coverage mapping > for the lambda in `assert([] { r

Re: r276716 - [Coverage] Do not write out coverage mappings with zero entries

2016-07-26 Thread Vedant Kumar via cfe-commits
Hi Justin, Could you take a look at this code coverage fix? It skips visiting decls if they would require context from system headers to present properly. This prevents us from writing out an empty coverage mapping for the lambda in `assert([] { return true; }());`. I think it would be a good id