Thanks Alistair. That's what I've been doing. However it would be more helpful to people who don't understand it to remove the call to the real XSync. Because they'd need to modify your code to really test it.
My test so far is stubxsync.c: int XSync(void *display, int discard) { return 0; } and built with: gcc -shared -o stubxsync.so stubxsync.c Then run with: env LD_PRELOAD=./stubxsync.so compiz ......... -- You received this bug notification because you are a member of compiz packagers, which is subscribed to compiz in Ubuntu. https://bugs.launchpad.net/bugs/1049214 Title: [nvidia] XSync usage is a massive bottlenecking factor (nvidia performance regression in Compiz 0.9.8+) To manage notifications about this bug go to: https://bugs.launchpad.net/compiz/+bug/1049214/+subscriptions _______________________________________________ Mailing list: https://launchpad.net/~compiz Post to : compiz@lists.launchpad.net Unsubscribe : https://launchpad.net/~compiz More help : https://help.launchpad.net/ListHelp