Linux Forums about Linux operating systems on the Dell Inspiron Mini.

Thread Tools Display Modes
  (#1) Old
bkero bkero is offline
Junior Member
Posts: 14
Join Date: Oct 2008
Default Optimal CFLAGS for Intel Atom processors - 11-10-2008, 01:09 AM

This weekend I've tortured my poor little Dell by running ACOVEA(Analysis of Compiler Options Via Evolutionary Algorithm) on it to see what the optimal flags are, and if they would make a noticeable performance difference. The results should be the same across all Atom-based netbooks, since they share a common platform. Here are the results.

If you're interested in the raw data, and the best-of-the-best and common compiler flags, they are located here. I'll be taking the common options from the tests, and adding them to my CFLAGS to see if I get a noticeable increase in performance.

EDIT: For those of you interested in the common options, here they are: -D__NO_MATH_INLINES -falign-functions -falign-labels -falign-loops -fbranch-target-load-optimize -fbranch-target-load-optimize2 -fbtr-bb-exclusive -fcaller-saves -fcrossjumping -fcse-follow-jumps -fcse-skip-blocks -fdelete-null-pointer-checks -fexpensive-optimizations -ffloat-store -fgcse -fgcse-after-reload -fgcse-las -fgcse-lm -fgcse-sm -finline-functions -finline-limit=600 -fmodulo-sched -fno-cprop-registers -fno-defer-pop -fno-delayed-branch -fno-function-cse -fno-if-conversion -fno-if-conversion2 -fno-inline -fno-merge-constants -fno-omit-frame-pointer -fno-thread-jumps -fno-tree-ccp -fno-tree-copyrename -fno-tree-dce -fno-tree-dominator-opts -fno-tree-dse -fno-tree-fre -fno-tree-lrs -fno-tree-sra -foptimize-sibling-calls -fpeel-loops -fpeephole2 -fprefetch-loop-arrays -fregmove -freorder-blocks -frerun-cse-after-loop -frerun-loop-opt -freschedule-modulo-scheduled-loops -fsched-interblock -fsched-spec -fschedule-insns -fschedule-insns2 -fstrength-reduce -fstrict-aliasing -fthread-jumps -ftracer -ftree-loop-im -ftree-loop-ivcanon -ftree-loop-linear -ftree-pre -ftree-vectorize -ftree-vrp -funit-at-a-time -funroll-all-loops -funroll-loops -funsafe-math-optimizations -funswitch-loops -fvariable-expansion-in-unroller -lm -lrt -maccumulate-outgoing-args -march=prescott -mfpmath=sse,387 -mieee-fp -minline-all-stringops -mno-align-stringops -mno-push-args -momit-leaf-frame-pointer -std=gnu99

I used GCC 4.1.2 for this analysis.
Reply With Quote

« Wolvix 1.1.0 Cub on USB Thumbdrive | Gentoo install instructions »
Thread Tools
Display Modes

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Copyright © 2008-2016