clang osx march=native approximately 1/3rd speed of no optimisation

Question

I have a project where speed is paramount, so was experimenting with compiler flags to try and get some free performance. I have two builds that are identical except for the additional flag march=native in build 2.

For completeness the flags are:

A) -std=c++14 -g -fno-omit-frame-pointer -O3

B) -std=c++14 -g -fno-omit-frame-pointer -O3 -march-native

Running benchmarks on these builds yields a confusing result:

A) 61s

B) 160s

What can possibly going on here?

Compilers are not magic. Sometimes they generate bad code (have bugs). — Jesper Juhl, May 29 '16 at 19:13
So this is a common phenomenon? Bugs in clang that worsen native optimised code — user3684792, May 29 '16 at 19:25
I'm not saying it is common. I'm just saying that you shouldn't blindly trust higher optimization levels or switches generating code for "your CPU" to generate better code. As always you have to test and profile. — Jesper Juhl, May 29 '16 at 19:34
ok, well these benchmarks are the result of testing and profiling. I have more granular data but it is not very enlightening. Any good resources to learn about these features further and understand why I am seeing this behaviour? — user3684792, May 29 '16 at 19:38
Looks to be a clang thing. Gcc doesn't show the same behaviour — user3684792, Jun 07 '16 at 13:30

score 3 · Accepted Answer · edited Oct 05 '18 at 02:06

Using -march=native optimizes code for your current cpu. Most of the time, it will improve the speed of execution. Sometimes it may not produce the fastest possible code because it enables certain CPU instructions.

echo | clang -E - -march=native -###

will display what clang has enabled through -march=native. The most likely culprit is CMOV, which is enabled by -march=native. You can see an explanation of why that might slow things down in the answer to this question: gcc optimization flag -O3 makes code slower than -O2.

score 0 · Answer 2 · answered Feb 27 '17 at 13:18

The parameter -march=native makes clang optimise code for your current CPU. In this case clang uses as much optimisations as possible and using it may break compatibility with other CPUs which, for example, don't have support for some instruction sets like AVX2, SSSE3 etc.

You may run

echo | clang -E - -O3 -###

and

clang -E - -march=native -O3 -###

to get the lists of features activated in the both of your cases.

clang osx march=native approximately 1/3rd speed of no optimisation

2 Answers2