[Clang] VectorExprEvaluator::VisitCallExpr / InterpretBuiltin - allow VPERMILPD/S variable mask intrinsics to be used in constexpr

[RKSimon]

__m128d _mm_permutevar_pd (__m128d a, __m128i b)
__m128d _mm_mask_permutevar_pd (__m128d src, __mmask8 k, __m128d a, __m128i b)
__m128d _mm_maskz_permutevar_pd (__mmask8 k, __m128d a, __m128i b)
__m256d _mm256_permutevar_pd (__m256d a, __m256i b)
__m256d _mm256_mask_permutevar_pd (__m256d src, __mmask8 k, __m256d a, __m256i b)
__m256d _mm256_maskz_permutevar_pd (__mmask8 k, __m256d a, __m256i b)
__m512d _mm512_permutevar_pd (__m512d a, __m512i b)
__m512d _mm512_mask_permutevar_pd (__m512d src, __mmask8 k, __m512d a, __m512i b)
__m512d _mm512_maskz_permutevar_pd (__mmask8 k, __m512d a, __m512i b)

__m128 _mm_permutevar_ps (__m128 a, __m128i b)
__m128 _mm_mask_permutevar_ps (__m128 src, __mmask8 k, __m128 a, __m128i b)
__m128 _mm_maskz_permutevar_ps (__mmask8 k, __m128 a, __m128i b)
__m256 _mm256_permutevar_ps (__m256 a, __m256i b)
__m256 _mm256_mask_permutevar_ps (__m256 src, __mmask8 k, __m256 a, __m256i b)
__m256 _mm256_maskz_permutevar_ps (__mmask8 k, __m256 a, __m256i b)
__m512 _mm512_permutevar_ps (__m512 a, __m512i b)
__m512 _mm512_mask_permutevar_ps (__m512 src, __mmask16 k, __m512 a, __m512i b)
__m512 _mm512_maskz_permutevar_ps (__mmask16 k, __m512 a, __m512i b)

Sibling issue to #166529 but with variable masks - unlike #167802 these shuffles are per lane

[llvmbot]

@llvm/issue-subscribers-backend-x86

Author: Simon Pilgrim (RKSimon)

``` __m128d _mm_permutevar_pd (__m128d a, __m128i b) __m128d _mm_mask_permutevar_pd (__m128d src, __mmask8 k, __m128d a, __m128i b) __m128d _mm_maskz_permutevar_pd (__mmask8 k, __m128d a, __m128i b) __m256d _mm256_permutevar_pd (__m256d a, __m256i b) __m256d _mm256_mask_permutevar_pd (__m256d src, __mmask8 k, __m256d a, __m256i b) __m256d _mm256_maskz_permutevar_pd (__mmask8 k, __m256d a, __m256i b) __m512d _mm512_permutevar_pd (__m512d a, __m512i b) __m512d _mm512_mask_permutevar_pd (__m512d src, __mmask8 k, __m512d a, __m512i b) __m512d _mm512_maskz_permutevar_pd (__mmask8 k, __m512d a, __m512i b)

__m128 _mm_permutevar_ps (__m128 a, __m128i b)
__m128 _mm_mask_permutevar_ps (__m128 src, __mmask8 k, __m128 a, __m128i b)
__m128 _mm_maskz_permutevar_ps (__mmask8 k, __m128 a, __m128i b)
__m256 _mm256_permutevar_ps (__m256 a, __m256i b)
__m256 _mm256_mask_permutevar_ps (__m256 src, __mmask8 k, __m256 a, __m256i b)
__m256 _mm256_maskz_permutevar_ps (__mmask8 k, __m256 a, __m256i b)
__m512 _mm512_permutevar_ps (__m512 a, __m512i b)
__m512 _mm512_mask_permutevar_ps (__m512 src, __mmask16 k, __m512 a, __m512i b)
__m512 _mm512_maskz_permutevar_ps (__mmask16 k, __m512 a, __m512i b)

Sibling issue to #<!-- -->166529but with variable masks - unlike #<!-- -->167802 these shuffles are per lane
</details>

[llvmbot]

Hi!

This issue may be a good introductory issue for people new to working on LLVM. If you would like to work on this issue, your first steps are:

1. Check that no other contributor is working on this issue. If someone is assigned to the issue or claimed to be working on it, ping the person. After one week without a response, the assignee may be changed.
2. Leave a comment indicating that you are working on the issue, or just create a pull request after following the steps below. Mention this issue in the description of the pull request.
3. Fix the issue locally.
4. Run the test suite locally. Remember that the subdirectories under test/ create fine-grained testing targets, so you can e.g. use make check-clang-ast to only run Clang's AST tests.
5. Create a Git commit.
6. Run git clang-format HEAD~1 to format your changes.
7. Open a pull request to the upstream repository on GitHub. Detailed instructions can be found in GitHub's documentation. Mention this issue in the description of the pull request.

If you have any further questions about this issue, don't hesitate to ask via a comment in the thread below.

[llvmbot]

@llvm/issue-subscribers-clang-frontend

Author: Simon Pilgrim (RKSimon)

``` __m128d _mm_permutevar_pd (__m128d a, __m128i b) __m128d _mm_mask_permutevar_pd (__m128d src, __mmask8 k, __m128d a, __m128i b) __m128d _mm_maskz_permutevar_pd (__mmask8 k, __m128d a, __m128i b) __m256d _mm256_permutevar_pd (__m256d a, __m256i b) __m256d _mm256_mask_permutevar_pd (__m256d src, __mmask8 k, __m256d a, __m256i b) __m256d _mm256_maskz_permutevar_pd (__mmask8 k, __m256d a, __m256i b) __m512d _mm512_permutevar_pd (__m512d a, __m512i b) __m512d _mm512_mask_permutevar_pd (__m512d src, __mmask8 k, __m512d a, __m512i b) __m512d _mm512_maskz_permutevar_pd (__mmask8 k, __m512d a, __m512i b)

__m128 _mm_permutevar_ps (__m128 a, __m128i b)
__m128 _mm_mask_permutevar_ps (__m128 src, __mmask8 k, __m128 a, __m128i b)
__m128 _mm_maskz_permutevar_ps (__mmask8 k, __m128 a, __m128i b)
__m256 _mm256_permutevar_ps (__m256 a, __m256i b)
__m256 _mm256_mask_permutevar_ps (__m256 src, __mmask8 k, __m256 a, __m256i b)
__m256 _mm256_maskz_permutevar_ps (__mmask8 k, __m256 a, __m256i b)
__m512 _mm512_permutevar_ps (__m512 a, __m512i b)
__m512 _mm512_mask_permutevar_ps (__m512 src, __mmask16 k, __m512 a, __m512i b)
__m512 _mm512_maskz_permutevar_ps (__mmask16 k, __m512 a, __m512i b)

Sibling issue to #<!-- -->166529but with variable masks - unlike #<!-- -->167802 these shuffles are per lane
</details>

[llvmbot]

@llvm/issue-subscribers-good-first-issue

Author: Simon Pilgrim (RKSimon)

``` __m128d _mm_permutevar_pd (__m128d a, __m128i b) __m128d _mm_mask_permutevar_pd (__m128d src, __mmask8 k, __m128d a, __m128i b) __m128d _mm_maskz_permutevar_pd (__mmask8 k, __m128d a, __m128i b) __m256d _mm256_permutevar_pd (__m256d a, __m256i b) __m256d _mm256_mask_permutevar_pd (__m256d src, __mmask8 k, __m256d a, __m256i b) __m256d _mm256_maskz_permutevar_pd (__mmask8 k, __m256d a, __m256i b) __m512d _mm512_permutevar_pd (__m512d a, __m512i b) __m512d _mm512_mask_permutevar_pd (__m512d src, __mmask8 k, __m512d a, __m512i b) __m512d _mm512_maskz_permutevar_pd (__mmask8 k, __m512d a, __m512i b)

__m128 _mm_permutevar_ps (__m128 a, __m128i b)
__m128 _mm_mask_permutevar_ps (__m128 src, __mmask8 k, __m128 a, __m128i b)
__m128 _mm_maskz_permutevar_ps (__mmask8 k, __m128 a, __m128i b)
__m256 _mm256_permutevar_ps (__m256 a, __m256i b)
__m256 _mm256_mask_permutevar_ps (__m256 src, __mmask8 k, __m256 a, __m256i b)
__m256 _mm256_maskz_permutevar_ps (__mmask8 k, __m256 a, __m256i b)
__m512 _mm512_permutevar_ps (__m512 a, __m512i b)
__m512 _mm512_mask_permutevar_ps (__m512 src, __mmask16 k, __m512 a, __m512i b)
__m512 _mm512_maskz_permutevar_ps (__mmask16 k, __m512 a, __m512i b)

Sibling issue to #<!-- -->166529but with variable masks - unlike #<!-- -->167802 these shuffles are per lane
</details>

[stomfaig]

I'd like to work on this, if that is okay.

[RKSimon]

@stomfaig Would be up for working on #166529 first (or both at the same time - they might share most of the same interp__builtin_ia32_shuffle_generic callback code)?

[stomfaig]

Sure.