Am 29.11.2014 um 15:27 schrieb Paolo Bonzini: > > On 28/11/2014 21:52, Peter Lieven wrote: >> master: >> Run operation 40000000 iterations 13.612604 s, 2938K operations/s, 340ns per >> coroutine >> >> this series up to patch 6: >> Run operation 40000000 iterations 10.428382 s, 3835K operations/s, 260ns per >> coroutine >> >> this series up to patch 7: >> Run operation 40000000 iterations 9.112539 s, 4389K operations/s, 227ns per >> coroutine >> >> So this confirms the +33% Paolo sees up to Patch 5. But I have yet fully >> understood the >> +15% that this Patch gains. > No atomic operations once the release pool gets full. We're talking of > 800 clock cycles here, and one atomic operation costs 50 cycles. 100 > clock cycles out of 800 = 15% speedup (8/7 = 1.14).
Maybe its worth mentioning this (partly) in the commit message that this can give a gain of additional 15% best case. This gives a +50% for the whole series best case. Peter