https://bugs.kde.org/show_bug.cgi?id=381553
--- Comment #36 from Ivo Raisr <iv...@ivosh.net> --- Some performance comparison... Number of instructions executed either in function doRegisterAllocation() [Valgrind baseline, v2] or doRegisterAllocation_v3 [new VEX register allocator, v3]: --- memcheck on /bin/date --- v2: 123,878,741 v3: 93,544,940 --- memcheck on perf/bigcode --- v2: 871,980,968 v3: 697,984,555 These numbers are obtained from running outer callgrind on inner memcheck as per README_DEVELOPERS with "--toggle-collect=doRegisterAllocation --toggle-collect=doRegisterAllocation_v3". Then running callgrind_annotate on the callgrind output files. Also the produced code from v3 is comparable or only marginally worse than from v2. ./vg-in-place --stats=yes perf/bigcode v2: transtab: new 30,033 (318,068 -> 9,133,920; ratio 28.7) v3: transtab: new 30,028 (317,855 -> 9,133,573; ratio 28.7) ./vg-in-place --stats=yes date v2: transtab: new 2,850 (61,825 -> 1,004,273; ratio 16.2) v3: transtab: new 2,845 (61,612 -> 1,006,286; ratio 16.3) -- You are receiving this mail because: You are watching all bug changes.
