1 // RUN: %clang_cc1 %s -flax-vector-conversions=none -ffreestanding -triple=x86_64-unknown-unknown \ 2 // RUN: -target-feature +amx-tf32 -target-feature +amx-transpose \ 3 // RUN: -target-feature +amx-bf16 -target-feature +avx512f \ 4 // RUN: -emit-llvm -o - -Werror -pedantic | FileCheck %s 5 6 #include <immintrin.h> 7 8 char buf[1024]; 9 #define STRIDE 32 10 11 char buf2[1024]; 12 13 void test_tile_mmultf32ps(__tile1024i a, __tile1024i b, __tile1024i c) { 14 //CHECK-LABEL: @test_tile_mmultf32ps 15 //CHECK-DAG: call x86_amx @llvm.x86.cast.vector.to.tile.v256i32(<256 x i32> {{%.*}}) 16 //CHECK-DAG: call x86_amx @llvm.x86.tmmultf32ps.internal 17 //CHECK-DAG: call <256 x i32> @llvm.x86.cast.tile.to.vector.v256i32(x86_amx {{%.*}}) 18 __tile_mmultf32ps(&c, a, b); 19 } 20 21 void test_tile_tmmultf32ps(__tile1024i a, __tile1024i b, __tile1024i c) { 22 //CHECK-LABEL: @test_tile_tmmultf32ps 23 //CHECK-DAG: call x86_amx @llvm.x86.cast.vector.to.tile.v256i32(<256 x i32> {{%.*}}) 24 //CHECK-DAG: call x86_amx @llvm.x86.ttmmultf32ps.internal 25 //CHECK-DAG: call <256 x i32> @llvm.x86.cast.tile.to.vector.v256i32(x86_amx {{%.*}}) 26 __tile_tmmultf32ps(&c, a, b); 27 } 28