Benchmark prints a "neon" result (presumably SSE) on an x86 machine

[HadrienG2]

~I think the tile says it all :)~

When running "make/ninja xbenchmark" on my Haswell-based machine, a pair of "neon" rows is present in the table of timings, even though this CPU certainly has no neon instructions available.

[HadrienG2]

That... was a different bug! :face_with_head_bandage: Will teach me to write descriptions in my tickets...

The bug which I am actually observing is that when running "make/ninja xbenchmark" on my Haswell-based machine, a pair of "neon" rows is present in the table of timings, even though this CPU certainly has no neon instructions available.

It is not fixed by the patch which you have merged today.

[yunhua-deng]

I am curious about this issue too, but there is still no answer after 1 year? Here is xsimd_benchmark's output, which shows neon rows, even though it is on x86 machine.

============================
fdim
scalar float   : 0.166787ms
sse float      : 0.006347ms
sse float unr  : 0.005641ms
avx float      : 0.004584ms
avx float unr  : 0.003526ms
neon float     : 0.008462ms
neon float unr : 0.007757ms
scalar double  : 0.171724ms
sse double     : 0.016573ms
sse double unr : 0.020451ms
avx double     : 0.017278ms
avx double unr : 0.015162ms
neon double    : 0.018689ms
neon double unr: 0.014457ms
============================

[SylvainCorlay]

Yep, this is very strange. This isssue fell between the cracks as it was closed by mistake.

[serge-sans-paille]

We no longer have the issue, and I took this issue as an opportunity to cleanup becnhmark display a bit, cf. #876