Add arm feature detection for FreeBSD

[MikaelUrankar]

No description provided.

[briansmith]

@MikaelUrankar Thanks for the PR. I noticed that this looks almost exactly like the Linux version. Would it be possible to refactor this so that the Linux and FreeBSD version of the setup function use the same code, but with different values of the HWCAP_* constants and different getauxval function names?

[codecov]

Codecov Report

Merging #1007 (04a131c) into main (92e15f2) will increase coverage by 0.02%.
The diff coverage is 100.00%.

❗ Current head 04a131c differs from pull request most recent head a5dfa60. Consider uploading reports for the commit a5dfa60 to get more accurate results

@@            Coverage Diff             @@
##             main    #1007      +/-   ##
==========================================
+ Coverage   92.85%   92.87%   +0.02%     
==========================================
  Files         120      106      -14     
  Lines       18114    12590    -5524     
  Branches      195        0     -195     
==========================================
- Hits        16819    11693    -5126     
+ Misses       1260      897     -363     
+ Partials       35        0      -35     

Impacted Files	Coverage Δ
src/cpu.rs	91.78% <100.00%> (-3.88%)	⬇️
src/test.rs	79.04% <0.00%> (-9.59%)	⬇️
crypto/poly1305/poly1305.c	94.35% <0.00%> (-4.08%)	⬇️
tests/quic_tests.rs	97.14% <0.00%> (-2.86%)	⬇️
tests/digest_tests.rs	98.00% <0.00%> (-2.00%)	⬇️
tests/aead_tests.rs	97.09% <0.00%> (-1.63%)	⬇️
src/ec/suite_b/ops.rs	96.81% <0.00%> (-1.25%)	⬇️
src/digest.rs	97.43% <0.00%> (-1.01%)	⬇️
src/io/der_writer.rs	92.50% <0.00%> (-0.98%)	⬇️
... and 109 more

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update 92e15f2...a5dfa60. Read the comment docs.

[MikaelUrankar]

@MikaelUrankar Thanks for the PR. I noticed that this looks almost exactly like the Linux version. Would it be possible to refactor this so that the Linux and FreeBSD version of the setup function use the same code, but with different values of the HWCAP_* constants and different getauxval function names?

it's done, the CI is broken though, don't know why.

[MikaelUrankar]

ping

[MikaelUrankar]

That's still an issue for FreeBSD users...

[briansmith]

@MikaelUrankar Thanks for your PR. I have an alternate approach to consider in PR #1542 that allows us to avoid adding new unsafe code to ring but still get the same effect. Please take a look at that and let me know what you think.