There are issues compiling ResourceLoadStatisticsMemoryStore in MSVC around nested lambdas. This is true even in MSVS 2019 Preview 3
Created attachment 362628 [details] Patch
Comment on attachment 362628 [details] Patch View in context: https://bugs.webkit.org/attachment.cgi?id=362628&action=review EWS seems happy, so this change looks good. > Source/WebKit/NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:369 > + RunLoop::main().dispatch([this, weakThis = WTFMove(weakThis)]() { Interesting that MSVC doesn't like this. This actually seems like better code anyway. > Source/WebKit/NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:33 > +#include "NetworkProcess.h" I wonder if this was missing because of the vagaries of our Unified Sources behavior.
(In reply to Brent Fulgham from comment #2) > Comment on attachment 362628 [details] > Patch > > View in context: > https://bugs.webkit.org/attachment.cgi?id=362628&action=review > > EWS seems happy, so this change looks good. > > > Source/WebKit/NetworkProcess/Classifier/ResourceLoadStatisticsMemoryStore.cpp:369 > > + RunLoop::main().dispatch([this, weakThis = WTFMove(weakThis)]() { > > Interesting that MSVC doesn't like this. This actually seems like better > code anyway. There are a bunch of bugs around lambdas in MSVC. Supposedly they were fixed but this was built with latest preview release of 2019 so 🤷 > > Source/WebKit/NetworkProcess/Classifier/WebResourceLoadStatisticsStore.cpp:33 > > +#include "NetworkProcess.h" > > I wonder if this was missing because of the vagaries of our Unified Sources > behavior. Most likely. I really hate Unified Sources. This https://bugs.webkit.org/show_bug.cgi?id=193073 needs to get fixed first.
Comment on attachment 362628 [details] Patch Clearing flags on attachment: 362628 Committed r241924: <https://trac.webkit.org/changeset/241924>
All reviewed patches have been landed. Closing bug.
<rdar://problem/48297768>