Use hostcall for wait and stream GC (!85) · Merge requests · Churavy, Valentin / KernelAbstractions.jl

Churavy, Valentin requested to merge vc/hostcall into master Mar 31, 2020

Created by: vchuravy

While looking at a profile from @lcw I noticed that we spend a lot of CPU cycles on cuEventQuery.

The second change is more questionable since it might make GPU launches slower (have yet to measure costs).

Use hostcall for wait and stream GC