Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

AMDGPU: Add v_smfmac_f32_32x32x32_f16 for gfx950 #117205

Merged
merged 1 commit into from
Nov 21, 2024
Merged

AMDGPU: Add v_smfmac_f32_32x32x32_f16 for gfx950 #117205

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
1 change: 1 addition & 0 deletions clang/include/clang/Basic/BuiltinsAMDGPU.def
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -445,6 +445,7 @@ TARGET_BUILTIN(__builtin_amdgcn_mfma_i32_16x16x64_i8, "V4iV4iV4iV4iIiIiIi", "nc"
TARGET_BUILTIN(__builtin_amdgcn_mfma_i32_32x32x32_i8, "V16iV4iV4iV16iIiIiIi", "nc", "gfx950-insts")

TARGET_BUILTIN(__builtin_amdgcn_smfmac_f32_16x16x64_f16, "V4fV8hV16hV4fiIiIi", "nc", "gfx950-insts")
TARGET_BUILTIN(__builtin_amdgcn_smfmac_f32_32x32x32_f16, "V16fV8hV16hV16fiIiIi", "nc", "gfx950-insts")
//===----------------------------------------------------------------------===//
// GFX12+ only builtins.
//===----------------------------------------------------------------------===//
Expand Down
7 changes: 7 additions & 0 deletions clang/test/CodeGenOpenCL/builtins-amdgcn-mfma.cl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -474,4 +474,11 @@ void test_smfmac_f32_16x16x64_f16(global v4f* out, v8h a, v16h b, v4f c, int idx
*out = __builtin_amdgcn_smfmac_f32_16x16x64_f16(a, b, c, idx, 0, 0);
}

// CHECK-GFX950-LABEL: @test_smfmac_f32_32x32x32_f16
// CHECK-GFX950: call <16 x float> @llvm.amdgcn.smfmac.f32.32x32x32.f16(<8 x half> %a, <16 x half> %b, <16 x float> %c, i32 %idx, i32 0, i32 0)
void test_smfmac_f32_32x32x32_f16(global v16f* out, v8h a, v16h b, v16f c, int idx)
{
*out = __builtin_amdgcn_smfmac_f32_32x32x32_f16(a, b, c, idx, 0, 0);
}

#endif
6 changes: 6 additions & 0 deletions clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950-param.cl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -69,3 +69,9 @@ void test_smfmac_f32_16x16x64_f16(global float4* out, half8 a, half16 b, float4
*out = __builtin_amdgcn_smfmac_f32_16x16x64_f16(a, b, c, idx, d, 0); // expected-error{{argument to '__builtin_amdgcn_smfmac_f32_16x16x64_f16' must be a constant integer}}
*out = __builtin_amdgcn_smfmac_f32_16x16x64_f16(a, b, c, idx, 0, d); // expected-error{{argument to '__builtin_amdgcn_smfmac_f32_16x16x64_f16' must be a constant integer}}
}

void test_smfmac_f32_32x32x32_f16(global float16* out, half8 a, half16 b, float16 c, int idx, int d)
{
*out = __builtin_amdgcn_smfmac_f32_32x32x32_f16(a, b, c, idx, d, 0); // expected-error{{argument to '__builtin_amdgcn_smfmac_f32_32x32x32_f16' must be a constant integer}}
*out = __builtin_amdgcn_smfmac_f32_32x32x32_f16(a, b, c, idx, 0, d); // expected-error{{argument to '__builtin_amdgcn_smfmac_f32_32x32x32_f16' must be a constant integer}}
}
1 change: 1 addition & 0 deletions clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950.cl
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -35,6 +35,7 @@ void test(__global float4* out0, half8 a0, half8 b0, float4 c0,
*out4 = __builtin_amdgcn_mfma_i32_32x32x32_i8(a4, b4, c4, 0, 0, 0); // expected-error{{'__builtin_amdgcn_mfma_i32_32x32x32_i8' needs target feature gfx950-insts}}
*out5 = __builtin_amdgcn_mfma_f32_16x16x32_bf16(a5, b5, c5, 0, 0, 0); // expected-error{{'__builtin_amdgcn_mfma_f32_16x16x32_bf16' needs target feature gfx950-insts}}
*out6 = __builtin_amdgcn_smfmac_f32_16x16x64_f16(a6, b6, c6, 0, 0, 0); // expected-error{{'__builtin_amdgcn_smfmac_f32_16x16x64_f16' needs target feature gfx950-insts}}
*out7 = __builtin_amdgcn_smfmac_f32_32x32x32_f16(a7, b7, c7, 0, 0, 0); // expected-error{{'__builtin_amdgcn_smfmac_f32_32x32x32_f16' needs target feature gfx950-insts}}
*out14 = __builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4(a14, b14, c14, 0, 0, 0, d14, 0, e14); // expected-error{{'__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4' needs target feature gfx950-insts}}
*out15 = __builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4(a15, b15, c15, 0, 0, 0, d15, 0, e15); // expected-error{{'__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4' needs target feature gfx950-insts}}
}
1 change: 1 addition & 0 deletions llvm/include/llvm/IR/IntrinsicsAMDGPU.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -3153,6 +3153,7 @@ def int_amdgcn_mfma_f32_32x32x16_bf16 : AMDGPUMfmaIntrinsic<llvm_v16f32_ty, llvm
def int_amdgcn_mfma_scale_f32_16x16x128_f8f6f4 : AMDGPUMfmaScaleIntrinsic<llvm_v4f32_ty>;
def int_amdgcn_mfma_scale_f32_32x32x64_f8f6f4 : AMDGPUMfmaScaleIntrinsic<llvm_v16f32_ty>;
def int_amdgcn_smfmac_f32_16x16x64_f16 : AMDGPUMSmfmacIntrinsic<llvm_v4f32_ty, llvm_v8f16_ty, llvm_v16f16_ty>;
def int_amdgcn_smfmac_f32_32x32x32_f16 : AMDGPUMSmfmacIntrinsic<llvm_v16f32_ty, llvm_v8f16_ty, llvm_v16f16_ty>;
}

//===----------------------------------------------------------------------===//
Expand Down
4 changes: 4 additions & 0 deletions llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1091,6 +1091,7 @@ bool AMDGPUInstructionSelector::selectG_INTRINSIC(MachineInstr &I) const {
case Intrinsic::amdgcn_smfmac_f32_32x32x32_fp8_bf8:
case Intrinsic::amdgcn_smfmac_f32_32x32x32_fp8_fp8:
case Intrinsic::amdgcn_smfmac_f32_16x16x64_f16:
case Intrinsic::amdgcn_smfmac_f32_32x32x32_f16:
return selectSMFMACIntrin(I);
default:
return selectImpl(I, *CoverageInfo);
Expand Down Expand Up @@ -3486,6 +3487,9 @@ bool AMDGPUInstructionSelector::selectSMFMACIntrin(MachineInstr &MI) const {
case Intrinsic::amdgcn_smfmac_f32_16x16x64_f16:
Opc = AMDGPU::V_SMFMAC_F32_16X16X64_F16_e64;
break;
case Intrinsic::amdgcn_smfmac_f32_32x32x32_f16:
Opc = AMDGPU::V_SMFMAC_F32_32X32X32_F16_e64;
break;
default:
llvm_unreachable("unhandled smfmac intrinsic");
}
Expand Down
3 changes: 2 additions & 1 deletion llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -4805,7 +4805,8 @@ AMDGPURegisterBankInfo::getInstrMapping(const MachineInstr &MI) const {
case Intrinsic::amdgcn_smfmac_f32_32x32x32_bf8_fp8:
case Intrinsic::amdgcn_smfmac_f32_32x32x32_fp8_bf8:
case Intrinsic::amdgcn_smfmac_f32_32x32x32_fp8_fp8:
case Intrinsic::amdgcn_smfmac_f32_16x16x64_f16: {
case Intrinsic::amdgcn_smfmac_f32_16x16x64_f16:
case Intrinsic::amdgcn_smfmac_f32_32x32x32_f16: {
// vdst, srcA, srcB, srcC, idx
OpdsMapping[0] = getAGPROpMapping(MI.getOperand(0).getReg(), MRI, *TRI);
OpdsMapping[2] = getVGPROpMapping(MI.getOperand(2).getReg(), MRI, *TRI);
Expand Down
1 change: 1 addition & 0 deletions llvm/lib/Target/AMDGPU/SIInstrInfo.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2871,6 +2871,7 @@ def VOP_V16F32_I64_I64_V16F32 : VOPProfile <[v16f32, i64, i64, v16f32]>;
def VOP_V4F32_V4F16_V8F16_I32 : VOPProfile <[v4f32, v4f16, v8f16, i32]>;
def VOP_V4F32_V8F16_V16F16_I32 : VOPProfile <[v4f32, v8f16, v16f16, i32]>;
def VOP_V16F32_V4F16_V8F16_I32 : VOPProfile <[v16f32, v4f16, v8f16, i32]>;
def VOP_V16F32_V8F16_V16F16_I32 : VOPProfile <[v16f32, v8f16, v16f16, i32]>;
def VOP_V4F32_V4I16_V8I16_I32 : VOPProfile <[v4f32, v4i16, v8i16, i32]>;
def VOP_V16F32_V4I16_V8I16_I32 : VOPProfile <[v16f32, v4i16, v8i16, i32]>;
def VOP_V4I32_V2I32_V4I32_I32 : VOPProfile <[v4i32, v2i32, v4i32, i32]>;
Expand Down
3 changes: 3 additions & 0 deletions llvm/lib/Target/AMDGPU/VOP3PInstructions.td
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -630,6 +630,7 @@ def VOPProfileMAI_F32_I64_X16_VCD : VOPProfileMAI<VOP_V16F32_I64_I64_V16F32,

def VOPProfileSMFMAC_F32_16X16X32_F16 : VOPProfileSMFMAC<VOP_V4F32_V4F16_V8F16_I32, AVDst_128, AVSrc_64, AVSrc_128>;
def VOPProfileSMFMAC_F32_16X16X64_F16 : VOPProfileSMFMAC<VOP_V4F32_V8F16_V16F16_I32, AVDst_128, AVSrc_128, AVSrc_256>;
def VOPProfileSMFMAC_F32_32X32X32_F16 : VOPProfileSMFMAC<VOP_V16F32_V8F16_V16F16_I32, AVDst_512, AVSrc_128, AVSrc_256>;
def VOPProfileSMFMAC_F32_32X32X16_F16 : VOPProfileSMFMAC<VOP_V16F32_V4F16_V8F16_I32, AVDst_512, AVSrc_64, AVSrc_128>;
def VOPProfileSMFMAC_F32_16X16X32_I16 : VOPProfileSMFMAC<VOP_V4F32_V4I16_V8I16_I32, AVDst_128, AVSrc_64, AVSrc_128>;
def VOPProfileSMFMAC_F32_32X32X16_I16 : VOPProfileSMFMAC<VOP_V16F32_V4I16_V8I16_I32, AVDst_512, AVSrc_64, AVSrc_128>;
Expand Down Expand Up @@ -1045,6 +1046,7 @@ defm V_SMFMAC_F32_32X32X32_FP8_FP8 : SMFMACInst<"v_smfmac_f32_32x32x32_fp8_fp8",

let SubtargetPredicate = HasGFX950Insts in {
defm V_SMFMAC_F32_16X16X64_F16 : SMFMACInst<"v_smfmac_f32_16x16x64_f16", "F32_16X16X64_F16", int_amdgcn_smfmac_f32_16x16x64_f16>;
defm V_SMFMAC_F32_32X32X32_F16 : SMFMACInst<"v_smfmac_f32_32x32x32_f16", "F32_32X32X32_F16", int_amdgcn_smfmac_f32_32x32x32_f16>;
}

def MAIInstInfoTable : GenericTable {
Expand Down Expand Up @@ -2137,6 +2139,7 @@ defm V_SMFMAC_F32_32X32X32_FP8_BF8 : VOP3P_Real_SMFMAC <0x7e, "v_smfmac_f32_32x3
defm V_SMFMAC_F32_32X32X32_FP8_FP8 : VOP3P_Real_SMFMAC <0x7f, "v_smfmac_f32_32x32x32fp8fp8">;

defm V_SMFMAC_F32_16X16X64_F16 : VOP3P_Real_SMFMAC <0x5a, "v_smfmac_f32_16x16x64f16">;
defm V_SMFMAC_F32_32X32X32_F16 : VOP3P_Real_SMFMAC <0x5b, "v_smfmac_f32_32x32x32f16">;

defm V_PK_FMA_F32 : VOP3P_Real_vi <0x30>;
defm V_PK_MUL_F32 : VOP3P_Real_vi <0x31>;
Expand Down
9 changes: 9 additions & 0 deletions llvm/test/Analysis/UniformityAnalysis/AMDGPU/intrinsics.ll
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -341,6 +341,15 @@ define amdgpu_kernel void @smfmac_f32_16x16x64_f16(<8 x half> %arg0, <16 x half>
ret void
}

declare <16 x float> @llvm.amdgcn.smfmac.f32.32x32x32.f16(<8 x half>, <16 x half>, <16 x float>, i32, i32 immarg, i32 immarg)

; CHECK: DIVERGENT: %result = call <16 x float> @llvm.amdgcn.smfmac.f32.32x32x32.f16(<8 x half> %arg0, <16 x half> %arg1, <16 x float> %arg2, i32 %arg3, i32 immarg 0, i32 immarg 0)
define amdgpu_kernel void @smfmac_f32_32x32x32_f16(<8 x half> %arg0, <16 x half> %arg1, <16 x float> %arg2, i32 %arg3, ptr addrspace(1) %out) {
%result = call <16 x float> @llvm.amdgcn.smfmac.f32.32x32x32.f16(<8 x half> %arg0, <16 x half> %arg1, <16 x float> %arg2, i32 %arg3, i32 immarg 0, i32 immarg 0)
store <16 x float> %result, ptr addrspace(1) %out
ret void
}

declare i32 @llvm.amdgcn.ds.swizzle(i32, i32) #1
declare i32 @llvm.amdgcn.permlane16.i32(i32, i32, i32, i32, i1, i1) #1
declare i32 @llvm.amdgcn.permlanex16.i32(i32, i32, i32, i32, i1, i1) #1
Expand Down
Loading
Loading