Re:Optimizing compilers reload vector constants needlessly
by jeffreyrogers on Tuesday December 06, 2022 @04:51PM
(#33887593)
He sets it once like this before the loops.
__m256i c = _mm256_set1_epi32(10001);
And then the disassembly has mov eax, 10001
vpbroadcastd ymm1, eax
before each loop.
Re:Optimizing compilers reload vector constants needlessly
by pclmulqdq on Tuesday December 06, 2022 @07:31PM
(#33889134)
Yeah, if the compiler runs out of registers it will do this - it's better than spilling the constant to memory. Register allocation is one thing that compilers are still much worse at than humans, and you see it in SIMD code a lot.
Re:Optimizing compilers reload vector constants needlessly
by DannyBee on Tuesday December 06, 2022 @10:17PM
(#33890252)
Eh.
Compilers are only worse because we don't care.
Optimal register allocation + spilling is easily doable now on today's machines/algorithms.
We just don't bother because it's not truly worth it in most cases (IE we get most of the performance for a very low cost)
__m256i mask = _mm256_set1_epi8(0x0f)
If you just used the intrinsic that sets the register to a constant over and over, it often repeats the instruction.
The compilers just aren't that smart about SIMD yet.