From 4f06dcbbb29a5b166fca80df08880b1372d6ddac Mon Sep 17 00:00:00 2001 From: Simon Pilgrim Date: Thu, 26 May 2016 10:20:25 +0000 Subject: [PATCH] [X86][F16C] Improved f16c intrinsics checks Added checks for upper elements being zero'd in scalar conversions git-svn-id: https://llvm.org/svn/llvm-project/cfe/trunk@270836 91177308-0d34-0410-b5e6-96231b3b80d8 --- test/CodeGen/f16c-builtins.c | 26 ++++++++++++++++++++------ 1 file changed, 20 insertions(+), 6 deletions(-) diff --git a/test/CodeGen/f16c-builtins.c b/test/CodeGen/f16c-builtins.c index b0f660ab0a..e99c0d00d4 100644 --- a/test/CodeGen/f16c-builtins.c +++ b/test/CodeGen/f16c-builtins.c @@ -7,36 +7,50 @@ float test_cvtsh_ss(unsigned short a) { // CHECK-LABEL: test_cvtsh_ss - // CHECK: @llvm.x86.vcvtph2ps.128 + // CHECK: insertelement <8 x i16> undef, i16 %{{.*}}, i32 0 + // CHECK: insertelement <8 x i16> %{{.*}}, i16 0, i32 1 + // CHECK: insertelement <8 x i16> %{{.*}}, i16 0, i32 2 + // CHECK: insertelement <8 x i16> %{{.*}}, i16 0, i32 3 + // CHECK: insertelement <8 x i16> %{{.*}}, i16 0, i32 4 + // CHECK: insertelement <8 x i16> %{{.*}}, i16 0, i32 5 + // CHECK: insertelement <8 x i16> %{{.*}}, i16 0, i32 6 + // CHECK: insertelement <8 x i16> %{{.*}}, i16 0, i32 7 + // CHECK: call <4 x float> @llvm.x86.vcvtph2ps.128(<8 x i16> %{{.*}}) + // CHECK: extractelement <4 x float> %{{.*}}, i32 0 return _cvtsh_ss(a); } unsigned short test_cvtss_sh(float a) { // CHECK-LABEL: test_cvtss_sh - // CHECK: @llvm.x86.vcvtps2ph.128 + // CHECK: insertelement <4 x float> undef, float %{{.*}}, i32 0 + // CHECK: insertelement <4 x float> %{{.*}}, float 0.000000e+00, i32 1 + // CHECK: insertelement <4 x float> %{{.*}}, float 0.000000e+00, i32 2 + // CHECK: insertelement <4 x float> %{{.*}}, float 0.000000e+00, i32 3 + // CHECK: call <8 x i16> @llvm.x86.vcvtps2ph.128(<4 x float> %{{.*}}, i32 0) + // CHECK: extractelement <8 x i16> %{{.*}}, i32 0 return _cvtss_sh(a, 0); } __m128 test_mm_cvtph_ps(__m128i a) { // CHECK-LABEL: test_mm_cvtph_ps - // CHECK: @llvm.x86.vcvtph2ps.128 + // CHECK: call <4 x float> @llvm.x86.vcvtph2ps.128(<8 x i16> %{{.*}}) return _mm_cvtph_ps(a); } __m256 test_mm256_cvtph_ps(__m128i a) { // CHECK-LABEL: test_mm256_cvtph_ps - // CHECK: @llvm.x86.vcvtph2ps.256 + // CHECK: call <8 x float> @llvm.x86.vcvtph2ps.256(<8 x i16> %{{.*}}) return _mm256_cvtph_ps(a); } __m128i test_mm_cvtps_ph(__m128 a) { // CHECK-LABEL: test_mm_cvtps_ph - // CHECK: @llvm.x86.vcvtps2ph.128 + // CHECK: call <8 x i16> @llvm.x86.vcvtps2ph.128(<4 x float> %{{.*}}, i32 0) return _mm_cvtps_ph(a, 0); } __m128i test_mm256_cvtps_ph(__m256 a) { // CHECK-LABEL: test_mm256_cvtps_ph - // CHECK: @llvm.x86.vcvtps2ph.256 + // CHECK: call <8 x i16> @llvm.x86.vcvtps2ph.256(<8 x float> %{{.*}}, i32 0) return _mm256_cvtps_ph(a, 0); } -- 2.40.0