AMDGPU: Add LLVM IR Intrinsic for v_lerp_u8

Differential Revision: http://reviews.llvm.org/D22239

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@275197 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/include/llvm/IR/IntrinsicsAMDGPU.td b/include/llvm/IR/IntrinsicsAMDGPU.td
index 119ce807a45fda28842f7ec557819b541a21fc0d..1473146cd929d1e28936c81f995c5f38ff076d13 100644 (file)
--- a/include/llvm/IR/IntrinsicsAMDGPU.td
+++ b/include/llvm/IR/IntrinsicsAMDGPU.td
@@ -384,6 +384,11 @@ def int_amdgcn_ds_swizzle :
   GCCBuiltin<"__builtin_amdgcn_ds_swizzle">,
   Intrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty], [IntrNoMem, IntrConvergent]>;
 
+// llvm.amdgcn.lerp
+def int_amdgcn_lerp :
+  GCCBuiltin<"__builtin_amdgcn_lerp">,
+  Intrinsic<[llvm_i32_ty], [llvm_i32_ty, llvm_i32_ty, llvm_i32_ty], [IntrNoMem]>;
+
 //===----------------------------------------------------------------------===//
 // CI+ Intrinsics
 //===----------------------------------------------------------------------===//

diff --git a/lib/Target/AMDGPU/SIInstructions.td b/lib/Target/AMDGPU/SIInstructions.td
index 05e0e3e504ca1777b01704360daf0e4c38e84f99..7cf5faa216d23344cce204a54e603d684c8be055 100644 (file)
--- a/lib/Target/AMDGPU/SIInstructions.td
+++ b/lib/Target/AMDGPU/SIInstructions.td
@@ -1717,6 +1717,10 @@ defm V_FMA_F32 : VOP3Inst <vop3<0x14b, 0x1cb>, "v_fma_f32",
 defm V_FMA_F64 : VOP3Inst <vop3<0x14c, 0x1cc>, "v_fma_f64",
   VOP_F64_F64_F64_F64, fma
 >;
+
+defm V_LERP_U8 : VOP3Inst <vop3<0x14d, 0x1cd>, "v_lerp_u8",
+  VOP_I32_I32_I32_I32, int_amdgcn_lerp
+>;
 } // End isCommutable = 1
 
 //def V_LERP_U8 : VOP3_U8 <0x0000014d, "v_lerp_u8", []>;

diff --git a/test/CodeGen/AMDGPU/llvm.amdgcn.lerp.ll b/test/CodeGen/AMDGPU/llvm.amdgcn.lerp.ll
new file mode 100644 (file)
index 0000000..014369b
--- /dev/null
+++ b/test/CodeGen/AMDGPU/llvm.amdgcn.lerp.ll
@@ -0,0 +1,14 @@
+; RUN: llc -march=amdgcn -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
+; RUN: llc -march=amdgcn -mcpu=fiji -verify-machineinstrs < %s | FileCheck -check-prefix=GCN %s
+
+declare i32 @llvm.amdgcn.lerp(i32, i32, i32) #0
+
+; GCN-LABEL: {{^}}v_lerp:
+; GCN: v_lerp_u8 v{{[0-9]+}}, s{{[0-9]+}}, v{{[0-9]+}}, v{{[0-9]+}}
+define void @v_lerp(i32 addrspace(1)* %out, i32 %src) nounwind {
+  %result= call i32 @llvm.amdgcn.lerp(i32 %src, i32 100, i32 100) #0
+  store i32 %result, i32 addrspace(1)* %out, align 4
+  ret void
+}
+
+attributes #0 = { nounwind readnone }