ilya-biryukov added a comment. In https://reviews.llvm.org/D50993#1207464, @Dmitry.Kozhevnikov wrote:
> Do you think it would be OK to have pthreads-only version there, with > `std::thread` fallback on Windows? I'd like to avoid writing a Windows-only > implementation myself, as it's a bit alien for me. Yeah, we should probably do both Unix and Windows there. My guess is that the Windows implementation shouldn't be too hard, since most of the code is already there. In any way, it's fine to start with a Unix-only version and a stub in Windows and figuring out the implementation during the review. Happy to look into Windows myself, it shouldn't be too hard :-) >> 2. I wonder if we should expose the stack size as an option to clangd? > > Do you expect it would be ever used? Not frequently, but I don't see how we can figure out good defaults there. It's one of those things where the guideline is "as small as possible, as long as it does not crash clangd". An option could provide a simple workaround for cases when clangd is actually crashing. Repository: rCTE Clang Tools Extra https://reviews.llvm.org/D50993 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
