blob: f1894cc14cc3930df4adef809270dfce3f3c3df4 [file] [log] [blame]
; RUN: llc -march=amdgcn -mcpu=gfx906 -verify-machineinstrs < %s | FileCheck %s --check-prefix=GCN --check-prefix=GFX906
declare i32 @llvm.amdgcn.sdot2(<2 x i16> %a, <2 x i16> %b, i32 %c, i1 %clamp)
; GCN-LABEL: {{^}}test_llvm_amdgcn_sdot2_clamp
; GFX906: v_dot2_i32_i16 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}} clamp{{$}}
define amdgpu_kernel void @test_llvm_amdgcn_sdot2_clamp(
i32 addrspace(1)* %r,
<2 x i16> addrspace(1)* %a,
<2 x i16> addrspace(1)* %b,
i32 addrspace(1)* %c) {
entry:
%a.val = load <2 x i16>, <2 x i16> addrspace(1)* %a
%b.val = load <2 x i16>, <2 x i16> addrspace(1)* %b
%c.val = load i32, i32 addrspace(1)* %c
%r.val = call i32 @llvm.amdgcn.sdot2(<2 x i16> %a.val, <2 x i16> %b.val, i32 %c.val, i1 1)
store i32 %r.val, i32 addrspace(1)* %r
ret void
}
; GCN-LABEL: {{^}}test_llvm_amdgcn_sdot2_no_clamp
; GFX906: v_dot2_i32_i16 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}{{$}}
define amdgpu_kernel void @test_llvm_amdgcn_sdot2_no_clamp(
i32 addrspace(1)* %r,
<2 x i16> addrspace(1)* %a,
<2 x i16> addrspace(1)* %b,
i32 addrspace(1)* %c) {
entry:
%a.val = load <2 x i16>, <2 x i16> addrspace(1)* %a
%b.val = load <2 x i16>, <2 x i16> addrspace(1)* %b
%c.val = load i32, i32 addrspace(1)* %c
%r.val = call i32 @llvm.amdgcn.sdot2(<2 x i16> %a.val, <2 x i16> %b.val, i32 %c.val, i1 0)
store i32 %r.val, i32 addrspace(1)* %r
ret void
}