On Wed, 27 Nov 2024 18:00:15 +0100 Corinna Vinschen wrote: > On Nov 26 17:55, Takashi Yano wrote: > > This patch calls Sleep() in lock() in order to increase the chance > > of being unlocked in other threads. The lock(), unlock() and locked() > > are moved from sigfe.s to cygtls.h so that allows inline expansion. > > When doing the locking in C++, we should really make stacklock volatile. > That affects especially the unlock and locked methods, where you'll > never know how gcc optimizes away.
OK. I have added volatile to stacklock. Please see v3 patch. > spinning should be volatile as well, but it's up to you if you already > add that, or if we only do this when we add the SIGKILL patches. The variable spinning does not concern directly for this patch, so could you please include that in your longjmp patch? -- Takashi Yano <takashi.y...@nifty.ne.jp>
