-
-
Notifications
You must be signed in to change notification settings - Fork 747
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
The way to Unity, pt.2 #1671
The way to Unity, pt.2 #1671
Conversation
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
This comment was marked as resolved.
8ce0604
to
3aae3b5
Compare
This might be due to the fact that we're using WaitForSingleObjectEx now (just as we did previously with winpthreads), which uses quite the whole thread on infinite waits. This was removed with the pthread rewrite where we moved to STL semaphores which use atomics and WaitOnAddress and WakeByAddressAll/WakeByAddressSingle on Windows. Unfortunately, I don't see a way of resolving this as, to my knowledge, there is no other documented method of having alertable waits. |
8d2bfc2
to
b876722
Compare
My reported performance regressions are solved after the latest pushes. Both Diva X v1.02 and Dragon Ball Z Kakarot are performing as they do on current main builds now. |
cd1c7a4
to
8ddaf42
Compare
8ddaf42
to
ae55e18
Compare
I found a couple of Unity games for the test, but they failed at startup. Build e1ecfb8 CUSA05297 INSIDE
CUSA23081 Untitled Goose Game
|
This reverts commit 7ffa581.
This reverts commit 7ffa581.
A big roadblock for Unity on Windows is the use of exceptions, and to implement those, we need to do some preparations.
The idea is to use APCs—for this, we must switch our code to use alertable waits with Win32 wrappers on:
This will allow us to properly implement exceptions later on.