Yeah it's easier to compile software with support for the latest vector extensions etc. if you do it from source. It is also possible to do runtime detection and switch between implementations that way, but it requires more work.
Tbh I don't think it would make much difference in practice. If you are ok with supporting only recentish CPUs you can use one of these options:
-
-march=x86-64: CMOV, CMPXCHG8B, FPU, FXSR, MMX, FXSR, SCE, SSE, SSE2
-
-march=x86-64-v2: (close to Nehalem) CMPXCHG16B, LAHF-SAHF, POPCNT, SSE3, SSE4.1, SSE4.2, SSSE3
-
-march=x86-64-v3: (close to Haswell) AVX, AVX2, BMI1, BMI2, F16C, FMA, LZCNT, MOVBE, XSAVE
-
-march=x86-64-v4: AVX512F, AVX512BW, AVX512CD, AVX512DQ, AVX512VL
v2 is definitely fine and v3 is probably acceptable by now.
In short I don't think -march
is a compelling argument for avoiding binary distribution. If it really makes a big difference either distribute multiple versions of your software using the flags above, or do runtime detection.