derekbruening
commented
9 years ago From bruen...@google.com on December 07, 2011 08:43:30
studying cfrac on windows:
symcache has no effect as 2nd run is no faster than 1st: % /usr/bin/time ~/drmemory/git/build_drmem_rel/bin/drmemory.exe -dr c:/src/dr/git/exports -batch -- ./cfrac.exe 41757646344123832613190542166099121 0.01user 0.01system 2:23.43elapsed 0%CPU (0avgtext+0avgdata 238336maxresident)k 0.00user 0.00system 2:25.32elapsed 0%CPU (0avgtext+0avgdata 233984maxresident)k
% for i in "-no_count_leaks" "-no_check_uninitialized" "-no_check_uninitialized -no_count_leaks" "-leaks_only -no_count_leaks -no_zero_stack" "-perturb_only" "-leaks_only -no_count_leaks -no_zero_stack -no_track_allocs"; do echo $i; /usr/bin/time ~/drmemory/git/build_drmem_rel/bin/drmemory.exe $i -quiet -dr c:/src/dr/git/exports -batch -- ./cfrac.exe 41757646344123832613190542166099121; done -no_count_leaks 0.00user 0.00system 1:18.67elapsed 0%CPU (0avgtext+0avgdata 233984maxresident)k -no_check_uninitialized 0.00user 0.01system 1:46.22elapsed 0%CPU (0avgtext+0avgdata 233984maxresident)k -no_check_uninitialized -no_count_leaks 0.00user 0.03system 0:45.65elapsed 0%CPU (0avgtext+0avgdata 234240maxresident)k -leaks_only -no_count_leaks -no_zero_stack 0.00user 0.00system 0:35.09elapsed 0%CPU (0avgtext+0avgdata 233984maxresident)k -perturb_only 0.00user 0.03system 0:10.37elapsed 0%CPU (0avgtext+0avgdata 233728maxresident)k -leaks_only -no_count_leaks -no_zero_stack -no_track_allocs 0.00user 0.01system 0:03.44elapsed 0%CPU (0avgtext+0avgdata 233984maxresident)k
-leaks_only -no_count_leaks -no_zero_stack + empty hook functions (alloc_hook, alloc_syscall_hook, and handle_alloc_post) (though I had to make handle_alloc_post have a conditional and not be entirely empty else crash b/c of some optimization) 0.00user 0.00system 0:06.41elapsed 0%CPU (0avgtext+0avgdata 234240maxresident)k
% /usr/bin/time ~/dr/git/exports/bin32/drrun.exe -quiet ./cfrac.exe 41757646344123832613190542166099121 0.01user 0.01system 0:02.50elapsed 1%CPU (0avgtext+0avgdata 238592maxresident)k
% /usr/bin/time ./cfrac.exe 41757646344123832613190542166099121 0.00user 0.01system 0:01.67elapsed 0%CPU (0avgtext+0avgdata 234496maxresident)k
I have a series of 10 or so optimizations including DRi#632 that get the 35 seconds down to about 24 seconds.
From bruen...@google.com on June 16, 2011 10:24:33
below are perf numbers for -leaks_only -no_count_leaks vs native.
-no_count_leaks disables stack zeroing, so this is the cost of malloc tracking (w/o recording callstacks): this is way too high! this is one barrier to higher performance with -no_check_uninitialized.
back when we first created Dr. Heapstat I spent some time optimizing malloc tracking, including recording callstacks, and I sure thought I had performance better than this (I was looking at spec2000 and some malloc-intensive benchmarks like roboop and cfrac). something could have regressed.
from profiling, some of this cost is from maintaining the hashtables and delay-free queue, much of which could go away w/ malloc replacement b/c the header could be used more freely: or perhaps a larger redzone could be used with more data than just the size stored directly in it. early injection would make that much easier, since today no-redzone allocs must be handled: one reason for the current design.
Original issue: http://code.google.com/p/drmemory/issues/detail?id=460