Revert "optimize constant_time_conditional_memxor for gcc".

[briansmith]

This reverts commit 9b3ef1b3d34d09c40b999d05ca6a92c77a9345e3. It seems like the casts in the loop violate strict aliasing rules.

This may result in a performance regression when GCC is used instead of Clang, as this code path was added specifically for GCC.

[codecov[bot]]

Codecov Report

:white_check_mark: All modified and coverable lines are covered by tests. :white_check_mark: Project coverage is 96.68%. Comparing base (59f2370) to head (8ffcc7d). :warning: Report is 11 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main    #2735      +/-   ##
==========================================
+ Coverage   96.64%   96.68%   +0.04%     
==========================================
  Files         199      199              
  Lines       20311    20301      -10     
  Branches      524      524              
==========================================
  Hits        19629    19629              
+ Misses        568      558      -10     
  Partials      114      114              

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.

:rocket: New features to boost your workflow:

• :snowflake: Test Analytics: Detect flaky tests, report on failures, and find test suite problems.

[MichaIng]

It seems BoringSSL found some alternative fix:

• https://github.com/google/boringssl/commit/63c8e81
• https://github.com/google/boringssl/commit/2d3fbe0

[briansmith]

• google/boringssl@63c8e81

I think that one isn't relevant to this, right?

• google/boringssl@2d3fbe0

That is addressing a different issue in the same function; see PR #2734.

[MichaIng]

Got it, thought it might be all related to the same underlying issue, and hence potentially a way to fix the issue without performance regression.