Unittests are failed when building with optimization mode in clang

[chokobole]

Issue type

Bug

OS platform and distribution

Linux Ubuntu 22.04

Current behavior?

Some of the unittests are failed.

• //tachyon/c/math/elliptic_curves/short_weierstrass:short_weierstrass_unittests

  exec ${PAGER:-/usr/bin/less} "$0" || exit 1
  Executing tests from //tachyon/c/math/elliptic_curves/short_weierstrass:short_weierstrass_unittests
  -----------------------------------------------------------------------------
  Running main() from gmock_main.cc
  [==========] Running 36 tests from 4 test suites.
  [----------] Global test environment set-up.
  [----------] 9 tests from AffinePointTest
  [ RUN      ] AffinePointTest.Zero
  [       OK ] AffinePointTest.Zero (0 ms)
  [ RUN      ] AffinePointTest.Generator
  [       OK ] AffinePointTest.Generator (0 ms)
  [ RUN      ] AffinePointTest.Random
  [       OK ] AffinePointTest.Random (0 ms)
  [ RUN      ] AffinePointTest.Eq
  [       OK ] AffinePointTest.Eq (0 ms)
  [ RUN      ] AffinePointTest.Ne
  [       OK ] AffinePointTest.Ne (0 ms)
  [ RUN      ] AffinePointTest.Add
  [       OK ] AffinePointTest.Add (0 ms)
  [ RUN      ] AffinePointTest.Sub
  [       OK ] AffinePointTest.Sub (0 ms)
  [ RUN      ] AffinePointTest.Neg
  [       OK ] AffinePointTest.Neg (0 ms)
  [ RUN      ] AffinePointTest.Dbl
  [       OK ] AffinePointTest.Dbl (0 ms)
  [----------] 9 tests from AffinePointTest (6 ms total)
  
  [----------] 9 tests from JacobianPointTest
  [ RUN      ] JacobianPointTest.Zero
  [       OK ] JacobianPointTest.Zero (0 ms)
  [ RUN      ] JacobianPointTest.Generator
  [       OK ] JacobianPointTest.Generator (0 ms)
  [ RUN      ] JacobianPointTest.Random

• //tachyon/c/zk/plonk/halo2:halo2_unittests

  exec ${PAGER:-/usr/bin/less} "$0" || exit 1
  Executing tests from //tachyon/c/zk/plonk/halo2:halo2_unittests
  -----------------------------------------------------------------------------
  Running main() from gmock_main.cc
  [==========] Running 39 tests from 5 test suites.
  [----------] Global test environment set-up.
  [----------] 1 test from TranscriptWriterTest/0, where TypeParam = <type>
  [ RUN      ] TranscriptWriterTest/0.APIs
  [       OK ] TranscriptWriterTest/0.APIs (0 ms)
  [----------] 1 test from TranscriptWriterTest/0 (0 ms total)
  
  [----------] 1 test from TranscriptWriterTest/1, where TypeParam = <type>
  [ RUN      ] TranscriptWriterTest/1.APIs
  [       OK ] TranscriptWriterTest/1.APIs (24 ms)
  [----------] 1 test from TranscriptWriterTest/1 (24 ms total)
  
  [----------] 1 test from TranscriptWriterTest/2, where TypeParam = <type>
  [ RUN      ] TranscriptWriterTest/2.APIs
  [       OK ] TranscriptWriterTest/2.APIs (0 ms)
  [----------] 1 test from TranscriptWriterTest/2 (0 ms total)
  
  [----------] 18 tests from GWCProverTest/GWCProverTest
  [ RUN      ] GWCProverTest/GWCProverTest.Constructor/0
  [       OK ] GWCProverTest/GWCProverTest.Constructor/0 (11 ms)
  [ RUN      ] GWCProverTest/GWCProverTest.Constructor/1
  [       OK ] GWCProverTest/GWCProverTest.Constructor/1 (27 ms)
  [ RUN      ] GWCProverTest/GWCProverTest.Constructor/2
  [       OK ] GWCProverTest/GWCProverTest.Constructor/2 (7 ms)
  [ RUN      ] GWCProverTest/GWCProverTest.Getters/0

Expected Behavior?

All the unittests should be passed.

Standalone code or description to reproduce the issue

export CC=/usr/bin/clang-15
export CXX=/usr/bin/clang++-15
bazel test -c opt //...