[PATCH] D66031: clangd: use -j for background index pool
Ben Jackson via Phabricator via cfe-commits
cfe-commits at lists.llvm.org
Fri Aug 9 15:16:36 PDT 2019
puremourning marked an inline comment as done.
puremourning added inline comments.
================
Comment at: clang-tools-extra/clangd/ClangdServer.cpp:152
+ std::forward<decltype(DBSF)>(DBSF),
+ Opts.AsyncThreadsCount );
+ } else {
----------------
sammccall wrote:
> puremourning wrote:
> > sammccall wrote:
> > > can we use `std::max(Opts.AsyncThreadsCount, 1)` instead?
> > >
> > > Having `-sync -background-index` use one thread seems less weird than having it use all the cores.
> > > (Or at least not more weird, and simpler in the code here)
> > Hmm. What I was thinking is more that if you specify none of sync or -j, you should get physical cores as you do now.
> >
> > But I realise that this change doesn't do that, because AsyncThreadsCount defaults slightly differently to `llvm::heavyweight_hardware_concurrency()` (it uses std::thread::hardware_concurrency)
> >
> > The difference is pretty small, so probably not material ?
> yikes, I forgot about that difference.
>
> We observed *significantly* worse performance and responsiveness when background threads was equal to the number of hardware threads rather than number of cores.
>
> If you don't mind, we should just use cores for everything: change `getDefaultAsyncThreadCount()` in TUScheduler.cpp to call llvm::heavyweight_hardware_concurrency() instead of std::thread::hardware_concurrency().
Sure thing. That makes sense.
It occurs to me that we might want to change the default value used by `BackgroundIndex` constructor, because it also can end up with `0` return from `heavyweight_hardware_concurrency`.
Worth changing that here ? I think the default is only used by the tests now though, so probably not a big issue.
Repository:
rG LLVM Github Monorepo
CHANGES SINCE LAST ACTION
https://reviews.llvm.org/D66031/new/
https://reviews.llvm.org/D66031
More information about the cfe-commits
mailing list