[PATCH] D47993: [x86] fix mappings of cvttp2si x86 intrinsics to x86-specific nodes and isel patterns (PR37551)

[Sanjay Patel via Phabricator via llvm-commits]

spatel added inline comments.


================
Comment at: test/CodeGen/X86/avx512-cvttp2i.ll:7-22
 declare <16 x i32> @llvm.x86.avx512.mask.cvttps2dq.512(<16 x float>, <16 x i32>, i16, i32)
 declare <4 x i32> @llvm.x86.avx512.mask.cvttps2udq.128(<4 x float>, <4 x i32>, i8)
 declare <8 x i32> @llvm.x86.avx512.mask.cvttps2udq.256(<8 x float>, <8 x i32>, i8)
 declare <16 x i32> @llvm.x86.avx512.mask.cvttps2udq.512(<16 x float>, <16 x i32>, i16, i32)
 declare <4 x i32> @llvm.x86.avx512.mask.cvttpd2udq.256(<4 x double>, <4 x i32>, i8)
 declare <8 x i32> @llvm.x86.avx512.mask.cvttpd2udq.512(<8 x double>, <8 x i32>, i8, i32)
 declare <4 x i64> @llvm.x86.avx512.mask.cvttps2qq.256(<4 x float>, <4 x i64>, i8)
----------------
I'm going to need new glasses after this review.

There are 16 intrinsics here, but there are 17 code diffs in the intrinsics header.
Are we missing tests for:
declare <8 x i32> @llvm.x86.avx512.mask.cvttpd2dq.512(<8 x double>, <8 x i32>, i8, i32)



https://reviews.llvm.org/D47993