From 450df72deb921ab3e975143c518a88e4d63f33aa Mon Sep 17 00:00:00 2001 From: Matt Arsenault Date: Fri, 5 Jan 2024 19:58:59 +0700 Subject: [PATCH] AMDGPU: Define v_mfma_f32_{16x16x128|32x32x64}_f8f6f4 instructions These use a new VOP3PX encoding for the v_mfma_scale_* instructions, which bundles the pre-scale v_mfma_ld_scale_b32. None of the modifiers are supported yet (op_sel, neg or clamp). I'm not sure the intrinsic should really expose op_sel (or any of the others). If I'm reading the documentation correctly, we should be able to just have the raw scale operands and auto-match op_sel to byte extract patterns. The op_sel syntax also seems extra horrible in this usage, especially with the usual assumed op_sel_hi=-1 behavior. The f8f6f4 intrinsics allow using different vector types, corresponding to the 3 different format widths. These can use 4, 6 or 8 x i32 vectors depending on if the format is fp4, fp6/bf6, or fp8/bf8. Verification that the used format matches the vector type will come later. This requires defining a separate pseudoinstruction for each register width combination, so 9 pseudos per opcode. This makes disassembly ambiguous, since now the opcode to use depends on the operand. Handle this by only defining the _f8_f8 variant as a real instruction, and the disassembler manually adjusts the opcode based on the format values later. The clang builtin integer operands should probably be unsigned, but all the other mfma intrinsics are using signed. --- clang/include/clang/Basic/BuiltinsAMDGPU.def | 4 +- clang/lib/CodeGen/CGBuiltin.cpp | 15 +- .../CodeGenOpenCL/builtins-amdgcn-mfma.cl | 15 + .../builtins-amdgcn-error-gfx950-param.cl | 15 + .../builtins-amdgcn-error-gfx950.cl | 23 +- llvm/docs/AMDGPUUsage.rst | 10 + llvm/include/llvm/IR/IntrinsicsAMDGPU.td | 31 + llvm/lib/Target/AMDGPU/AMDGPUGISel.td | 3 + .../AMDGPU/AMDGPUInstCombineIntrinsic.cpp | 1 + .../AMDGPU/AMDGPUInstructionSelector.cpp | 12 + .../Target/AMDGPU/AMDGPUInstructionSelector.h | 2 + llvm/lib/Target/AMDGPU/AMDGPUInstructions.td | 2 +- .../Target/AMDGPU/AMDGPURegisterBankInfo.cpp | 19 + .../Disassembler/AMDGPUDisassembler.cpp | 75 + .../AMDGPU/Disassembler/AMDGPUDisassembler.h | 1 + .../AMDGPU/MCTargetDesc/AMDGPUMCAsmInfo.cpp | 4 + llvm/lib/Target/AMDGPU/SIDefines.h | 12 + llvm/lib/Target/AMDGPU/SIISelLowering.cpp | 17 + llvm/lib/Target/AMDGPU/SIInstrFormats.td | 5 + llvm/lib/Target/AMDGPU/SIInstrInfo.h | 6 + llvm/lib/Target/AMDGPU/SIInstrInfo.td | 51 + llvm/lib/Target/AMDGPU/SIRegisterInfo.td | 3 + .../Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp | 27 + llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h | 12 + llvm/lib/Target/AMDGPU/VOP3PInstructions.td | 313 +- llvm/lib/Target/AMDGPU/VOPInstructions.td | 93 +- .../UniformityAnalysis/AMDGPU/intrinsics.ll | 20 + ....amdgcn.mfma.scale.f32.16x16x128.f8f6f4.ll | 2296 +++++++ ...m.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.ll | 6004 +++++++++++++++++ llvm/test/MC/AMDGPU/mai-gfx950.s | 616 ++ .../MC/Disassembler/AMDGPU/gfx950_mai.txt | 408 ++ .../MC/Disassembler/AMDGPU/gfx950_vop3px2.txt | 28 + llvm/test/tools/llvm-mca/AMDGPU/gfx950.s | 11 +- 33 files changed, 10121 insertions(+), 33 deletions(-) create mode 100644 llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.ll create mode 100644 llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.ll create mode 100644 llvm/test/MC/Disassembler/AMDGPU/gfx950_vop3px2.txt diff --git a/clang/include/clang/Basic/BuiltinsAMDGPU.def b/clang/include/clang/Basic/BuiltinsAMDGPU.def index 7ce8f2c1669d67..faf2e861451790 100644 --- a/clang/include/clang/Basic/BuiltinsAMDGPU.def +++ b/clang/include/clang/Basic/BuiltinsAMDGPU.def @@ -434,9 +434,11 @@ TARGET_BUILTIN(__builtin_amdgcn_cvt_sr_fp8_f32, "ifiiIi", "nc", "fp8-conversion- //===----------------------------------------------------------------------===// // GFX950 only builtins. //===----------------------------------------------------------------------===// +TARGET_BUILTIN(__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4, "V4fV8ZiV8ZiV4fIiIiIiiIii", "nc", "gfx950-insts") +TARGET_BUILTIN(__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4, "V16fV8ZiV8ZiV16fIiIiIiiIii", "nc", "gfx950-insts") + TARGET_BUILTIN(__builtin_amdgcn_mfma_f32_16x16x32_f16, "V4fV8hV8hV4fIiIiIi", "nc", "gfx950-insts") TARGET_BUILTIN(__builtin_amdgcn_mfma_f32_32x32x16_f16, "V16fV8hV8hV16fIiIiIi", "nc", "gfx950-insts") - TARGET_BUILTIN(__builtin_amdgcn_mfma_f32_32x32x16_bf16, "V16fV8yV8yV16fIiIiIi", "nc", "gfx950-insts") //===----------------------------------------------------------------------===// diff --git a/clang/lib/CodeGen/CGBuiltin.cpp b/clang/lib/CodeGen/CGBuiltin.cpp index 887e0f0e05469e..b09c21a0cfcb50 100644 --- a/clang/lib/CodeGen/CGBuiltin.cpp +++ b/clang/lib/CodeGen/CGBuiltin.cpp @@ -19729,7 +19729,20 @@ Value *CodeGenFunction::EmitAMDGPUBuiltinExpr(unsigned BuiltinID, (uint64_t)0); return Builder.CreateInsertElement(I0, A, 1); } - + case AMDGPU::BI__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4: + case AMDGPU::BI__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4: { + llvm::FixedVectorType *VT = FixedVectorType::get(Builder.getInt32Ty(), 8); + Function *F = CGM.getIntrinsic( + BuiltinID == AMDGPU::BI__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4 + ? Intrinsic::amdgcn_mfma_scale_f32_32x32x64_f8f6f4 + : Intrinsic::amdgcn_mfma_scale_f32_16x16x128_f8f6f4, + {VT, VT}); + + SmallVector Args; + for (unsigned I = 0, N = E->getNumArgs(); I != N; ++I) + Args.push_back(EmitScalarExpr(E->getArg(I))); + return Builder.CreateCall(F, Args); + } case AMDGPU::BI__builtin_amdgcn_wmma_bf16_16x16x16_bf16_w32: case AMDGPU::BI__builtin_amdgcn_wmma_bf16_16x16x16_bf16_tied_w32: case AMDGPU::BI__builtin_amdgcn_wmma_bf16_16x16x16_bf16_w64: diff --git a/clang/test/CodeGenOpenCL/builtins-amdgcn-mfma.cl b/clang/test/CodeGenOpenCL/builtins-amdgcn-mfma.cl index 841d8fcad0fee0..ea9bdcdc211623 100644 --- a/clang/test/CodeGenOpenCL/builtins-amdgcn-mfma.cl +++ b/clang/test/CodeGenOpenCL/builtins-amdgcn-mfma.cl @@ -16,6 +16,7 @@ typedef half v16h __attribute__((ext_vector_type(16))); typedef half v32h __attribute__((ext_vector_type(32))); typedef int v2i __attribute__((ext_vector_type(2))); typedef int v4i __attribute__((ext_vector_type(4))); +typedef int v8i __attribute__((ext_vector_type(8))); typedef int v16i __attribute__((ext_vector_type(16))); typedef int v32i __attribute__((ext_vector_type(32))); typedef short v2s __attribute__((ext_vector_type(2))); @@ -431,4 +432,18 @@ v16f test_mfma_f32_32x32x16_bf16(v8bf16 a, v8bf16 b, v16f c) { return __builtin_amdgcn_mfma_f32_32x32x16_bf16(a, b, c, 1, 2, 3); } +// CHECK-GFX950-LABEL: @test_mfma_scale_f32_16x16x128_f8f6f4 +// CHECK-GFX950: call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %a, <8 x i32> %b, <4 x float> %c, i32 3, i32 1, i32 2, i32 %scale_a, i32 3, i32 %scale_b) +void test_mfma_scale_f32_16x16x128_f8f6f4(global v4f* out, v8i a, v8i b, v4f c, int scale_a, int scale_b) +{ + *out = __builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4(a, b, c, 3, 1, 2, scale_a, 3, scale_b); +} + +// CHECK-GFX950-LABEL: @test_mfma_scale_f32_32x32x64_f8f6f4 +// CHECK-GFX950: call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %a, <8 x i32> %b, <16 x float> %c, i32 3, i32 1, i32 2, i32 %scale_a, i32 3, i32 %scale_b) +void test_mfma_scale_f32_32x32x64_f8f6f4(global v16f* out, v8i a, v8i b, v16f c, int scale_a, int scale_b) +{ + *out = __builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4(a, b, c, 3, 1, 2, scale_a, 3, scale_b); +} + #endif diff --git a/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950-param.cl b/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950-param.cl index 4af67763c40dd2..7f0300ec196e34 100644 --- a/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950-param.cl +++ b/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950-param.cl @@ -5,6 +5,7 @@ typedef float float4 __attribute__((ext_vector_type(4))); typedef float float16 __attribute__((ext_vector_type(16))); typedef half half8 __attribute__((ext_vector_type(8))); typedef __bf16 bfloat8 __attribute__((ext_vector_type(8))); +typedef int int8 __attribute__((ext_vector_type(8))); void test_mfma_f32_16x16x32_f16(__global float4* out, half8 a, half8 b, float4 c, int X) { @@ -26,3 +27,17 @@ void test_mfma_f32_32x32x16_bf16(__global float16* out, bfloat8 a, bfloat8 b, fl *out = __builtin_amdgcn_mfma_f32_32x32x16_bf16(a, b, c, 0, X, 0); // expected-error{{argument to '__builtin_amdgcn_mfma_f32_32x32x16_bf16' must be a constant integer}} *out = __builtin_amdgcn_mfma_f32_32x32x16_bf16(a, b, c, 0, 0, X); // expected-error{{argument to '__builtin_amdgcn_mfma_f32_32x32x16_bf16' must be a constant integer}} } + +void test_mfma_scale_f32_16x16x128_f8f6f4(__global float4* out, int8 a, int8 b, float4 c, int X, int Y) { + *out = __builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4(a, b, c, X, 0, 1, Y, 2, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4' must be a constant integer}} + *out = __builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4(a, b, c, 0, X, 1, Y, 2, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4' must be a constant integer}} + *out = __builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4(a, b, c, 0, 0, X, Y, 2, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4' must be a constant integer}} + *out = __builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4(a, b, c, 0, 0, 0, Y, X, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_16x16x128_f8f6f4' must be a constant integer}} +} + +void test_mfma_scale_f32_32x32x64_f8f6f4(__global float16* out, int8 a, int8 b, float16 c, int X, int Y) { + *out = __builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4(a, b, c, X, 0, 1, Y, 2, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4' must be a constant integer}} + *out = __builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4(a, b, c, 0, X, 1, Y, 2, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4' must be a constant integer}} + *out = __builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4(a, b, c, 0, 0, X, Y, 2, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4' must be a constant integer}} + *out = __builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4(a, b, c, 0, 0, 0, Y, X, Y); // expected-error{{argument to '__builtin_amdgcn_mfma_scale_f32_32x32x64_f8f6f4' must be a constant integer}} +} diff --git a/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950.cl b/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950.cl index e0fd2aa5c58a02..2a90bda1a5b16c 100644 --- a/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950.cl +++ b/clang/test/SemaOpenCL/builtins-amdgcn-error-gfx950.cl @@ -4,12 +4,33 @@ typedef float float4 __attribute__((ext_vector_type(4))); typedef float float16 __attribute__((ext_vector_type(16))); typedef half half8 __attribute__((ext_vector_type(8))); +typedef half half16 __attribute__((ext_vector_type(16))); typedef __bf16 bfloat8 __attribute__((ext_vector_type(8))); +typedef __bf16 bfloat16 __attribute__((ext_vector_type(16))); +typedef unsigned int uint2 __attribute__((ext_vector_type(2))); +typedef int int4 __attribute__((ext_vector_type(4))); +typedef int int8 __attribute__((ext_vector_type(8))); +typedef int int16 __attribute__((ext_vector_type(16))); void test(__global float4* out0, half8 a0, half8 b0, float4 c0, __global float16* out1, half8 a1, half8 b1, float16 c1, - __global float16* out2, bfloat8 a2, bfloat8 b2, float16 c2) { + __global float16* out2, bfloat8 a2, bfloat8 b2, float16 c2, + __global int4* out3, int4 a3, int4 b3, int4 c3, + __global int16* out4, int4 a4, int4 b4, int16 c4, + __global float4* out5, bfloat8 a5, bfloat8 b5, float4 c5, + __global float4* out6, half8 a6, half16 b6, float4 c6, + __global float16* out7, half8 a7, half16 b7, float16 c7, + __global float4* out8, bfloat8 a8, bfloat16 b8, float4 c8, + __global float16* out9, bfloat8 a9, bfloat16 b9, float16 c9, + __global int4* out10, int4 a10, int8 b10, int4 c10, + __global int16* out11, int4 a11, int8 b11, int16 c11, + __global float4* out12, int4 a12, int8 b12, float4 c12, + __global float16* out13, int4 a13, int8 b13, float16 c13, + __global float4* out14, int8 a14, int8 b14, float4 c14, int d14, int e14, + __global float16* out15, int8 a15, int8 b15, float16 c15, int d15, int e15) { *out0 = __builtin_amdgcn_mfma_f32_16x16x32_f16(a0, b0, c0, 0, 0, 0); // expected-error{{'__builtin_amdgcn_mfma_f32_16x16x32_f16' needs target feature gfx950-insts}} *out1 = __builtin_amdgcn_mfma_f32_32x32x16_f16(a1, b1, c1, 0, 0, 0); // expected-error{{'__builtin_amdgcn_mfma_f32_32x32x16_f16' needs target feature gfx950-insts}} *out2 = __builtin_amdgcn_mfma_f32_32x32x16_bf16(a2, b2, c2, 0, 0, 0); // expected-error{{'__builtin_amdgcn_mfma_f32_32x32x16_bf16' 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}} } diff --git a/llvm/docs/AMDGPUUsage.rst b/llvm/docs/AMDGPUUsage.rst index a25b6feddbeddc..161363e0dd6bcc 100644 --- a/llvm/docs/AMDGPUUsage.rst +++ b/llvm/docs/AMDGPUUsage.rst @@ -1397,6 +1397,16 @@ The AMDGPU backend implements the following LLVM IR intrinsics. used by hardware to control active lanes when used in EXEC register. For example, ballot(i1 true) return EXEC mask. + llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4 Emit `v_mfma_scale_f32_16x16x128_f8f6f4` to set the scale factor. The + last 4 operands correspond to the scale inputs. + + - 2-bit byte index to use for each lane for matrix A + - Matrix A scale values + - 2-bit byte index to use for each lane for matrix B + - Matrix B scale values + + llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4 Emit `v_mfma_scale_f32_32x32x64_f8f6f4` + ============================================== ========================================================== .. TODO:: diff --git a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td index 360af786c5160d..3a5fc86183ca0e 100644 --- a/llvm/include/llvm/IR/IntrinsicsAMDGPU.td +++ b/llvm/include/llvm/IR/IntrinsicsAMDGPU.td @@ -2968,6 +2968,35 @@ class AMDGPUMfmaIntrinsic : [IntrConvergent, IntrNoMem, ImmArg>, ImmArg>, ImmArg>]>; + +// srcA's format is determined by cbsz. srcB's format is determined by +// blgp. +// +// These should be <8 x i32> for f8 formats, <6 x i32> for f6 formats, +// and <4 x i32> for f4 formats. If the format control bits imply a +// smaller type than used, the high elements will be truncated. +// +// If the format control bits imply a larger type than used, the high +// elements are padded with undef. + +class AMDGPUMfmaScaleIntrinsic : + DefaultAttrsIntrinsic<[DestTy], + [llvm_anyvector_ty, llvm_anyvector_ty, DestTy, + llvm_i32_ty, // cbsz + llvm_i32_ty, // blgp + // llvm_i1_ty, // TODO: neg_src2 + // llvm_i1_ty, // TODO: abs_src2 + // llvm_i1_ty, // TODO: clamp + llvm_i32_ty, // op_sel (A matrix scale, 2-bits) // TODO: Make i2? + llvm_i32_ty, // v_mfma_ld_scale_b32 src0 (A matrix scale) + llvm_i32_ty, // op_sel (B matrix scale, 2-bits) // TODO: Make i2? + llvm_i32_ty // v_mfma_ld_scale_b32 src1 (B matrix scale) + ], + [IntrConvergent, IntrNoMem, + ImmArg>, ImmArg>, + ImmArg>, ImmArg> + ]>; + defset list AMDGPUMFMAIntrinsics908 = { def int_amdgcn_mfma_f32_32x32x1f32 : AMDGPUMfmaIntrinsic; def int_amdgcn_mfma_f32_16x16x1f32 : AMDGPUMfmaIntrinsic; @@ -3119,6 +3148,8 @@ def int_amdgcn_mfma_f32_16x16x32_f16 : AMDGPUMfmaIntrinsic; def int_amdgcn_mfma_f32_32x32x16_bf16 : AMDGPUMfmaIntrinsic; +def int_amdgcn_mfma_scale_f32_16x16x128_f8f6f4 : AMDGPUMfmaScaleIntrinsic; +def int_amdgcn_mfma_scale_f32_32x32x64_f8f6f4 : AMDGPUMfmaScaleIntrinsic; } //===----------------------------------------------------------------------===// diff --git a/llvm/lib/Target/AMDGPU/AMDGPUGISel.td b/llvm/lib/Target/AMDGPU/AMDGPUGISel.td index d348f489d95dd3..88fa96bd049f29 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPUGISel.td +++ b/llvm/lib/Target/AMDGPU/AMDGPUGISel.td @@ -423,3 +423,6 @@ def gi_fp_pow2_to_exponent : GICustomOperandRenderer<"renderFPPow2ToExponent">, def gi_as_hw_round_mode : GICustomOperandRenderer<"renderRoundMode">, GISDNodeXFormEquiv; + +def gi_MFMALdScaleModifierOp : GICustomOperandRenderer<"renderScaledMAIIntrinsicOperand">, + GISDNodeXFormEquiv; diff --git a/llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp b/llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp index 28d215e7b3de9f..30b9e3ea11ef4b 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp +++ b/llvm/lib/Target/AMDGPU/AMDGPUInstCombineIntrinsic.cpp @@ -1258,6 +1258,7 @@ GCNTTIImpl::instCombineIntrinsic(InstCombiner &IC, IntrinsicInst &II) const { if (isa(Src)) { return IC.replaceInstUsesWith(II, Src); } + return std::nullopt; } } if (const AMDGPU::ImageDimIntrinsicInfo *ImageDimIntr = diff --git a/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp b/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp index f2fe3befd04d55..f7cba2f13c77ac 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp +++ b/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.cpp @@ -5742,6 +5742,18 @@ void AMDGPUInstructionSelector::renderRoundMode(MachineInstrBuilder &MIB, MIB.addImm((MI.getOperand(OpIdx).getImm() + 3) % 4); } +/// Convert from 2-bit value to enum values used for op_sel* source modifiers. +void AMDGPUInstructionSelector::renderScaledMAIIntrinsicOperand( + MachineInstrBuilder &MIB, const MachineInstr &MI, int OpIdx) const { + unsigned Val = MI.getOperand(OpIdx).getImm(); + unsigned New = 0; + if (Val & 0x1) + New |= SISrcMods::OP_SEL_0; + if (Val & 0x2) + New |= SISrcMods::OP_SEL_1; + MIB.addImm(New); +} + bool AMDGPUInstructionSelector::isInlineImmediate(const APInt &Imm) const { return TII.isInlineConstant(Imm); } diff --git a/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.h b/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.h index 42343104812b66..563e40267f04b1 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.h +++ b/llvm/lib/Target/AMDGPU/AMDGPUInstructionSelector.h @@ -364,6 +364,8 @@ class AMDGPUInstructionSelector final : public InstructionSelector { void renderRoundMode(MachineInstrBuilder &MIB, const MachineInstr &MI, int OpIdx) const; + void renderScaledMAIIntrinsicOperand(MachineInstrBuilder &MIB, + const MachineInstr &MI, int OpIdx) const; bool isInlineImmediate(const APInt &Imm) const; bool isInlineImmediate(const APFloat &Imm) const; diff --git a/llvm/lib/Target/AMDGPU/AMDGPUInstructions.td b/llvm/lib/Target/AMDGPU/AMDGPUInstructions.td index 671070c70f0c41..6a5065cd4a0e8f 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPUInstructions.td +++ b/llvm/lib/Target/AMDGPU/AMDGPUInstructions.td @@ -40,7 +40,7 @@ class AMDGPUInst SoftFail = 0; + field bits<128> SoftFail = 0; // FIXME: If this is smaller than largest instruction, DecodeEmitter crashes let DecoderNamespace = Namespace; diff --git a/llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp b/llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp index b648b68f3bd2b0..7aae9194b5cd02 100644 --- a/llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp +++ b/llvm/lib/Target/AMDGPU/AMDGPURegisterBankInfo.cpp @@ -4769,6 +4769,25 @@ AMDGPURegisterBankInfo::getInstrMapping(const MachineInstr &MI) const { : getVGPROpMapping(MI.getOperand(4).getReg(), MRI, *TRI); break; } + case Intrinsic::amdgcn_mfma_scale_f32_16x16x128_f8f6f4: + case Intrinsic::amdgcn_mfma_scale_f32_32x32x64_f8f6f4: { + const SIMachineFunctionInfo *Info = MF.getInfo(); + OpdsMapping[0] = + Info->mayNeedAGPRs() + ? getAGPROpMapping(MI.getOperand(0).getReg(), MRI, *TRI) + : getVGPROpMapping(MI.getOperand(0).getReg(), MRI, *TRI); + + OpdsMapping[2] = getVGPROpMapping(MI.getOperand(2).getReg(), MRI, *TRI); + OpdsMapping[3] = getVGPROpMapping(MI.getOperand(3).getReg(), MRI, *TRI); + OpdsMapping[4] = + Info->mayNeedAGPRs() + ? getAGPROpMapping(MI.getOperand(4).getReg(), MRI, *TRI) + : getVGPROpMapping(MI.getOperand(4).getReg(), MRI, *TRI); + + OpdsMapping[8] = getVGPROpMapping(MI.getOperand(8).getReg(), MRI, *TRI); + OpdsMapping[10] = getVGPROpMapping(MI.getOperand(10).getReg(), MRI, *TRI); + break; + } case Intrinsic::amdgcn_smfmac_f32_16x16x32_f16: case Intrinsic::amdgcn_smfmac_f32_32x32x16_f16: case Intrinsic::amdgcn_smfmac_f32_16x16x32_bf16: diff --git a/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.cpp b/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.cpp index 06df08feda8fa4..f90121a86c846c 100644 --- a/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.cpp +++ b/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.cpp @@ -512,6 +512,17 @@ static inline DecoderUInt128 eat12Bytes(ArrayRef &Bytes) { return DecoderUInt128(Lo, Hi); } +static inline DecoderUInt128 eat16Bytes(ArrayRef &Bytes) { + assert(Bytes.size() >= 16); + uint64_t Lo = + support::endian::read(Bytes.data()); + Bytes = Bytes.slice(8); + uint64_t Hi = + support::endian::read(Bytes.data()); + Bytes = Bytes.slice(8); + return DecoderUInt128(Lo, Hi); +} + DecodeStatus AMDGPUDisassembler::getInstruction(MCInst &MI, uint64_t &Size, ArrayRef Bytes_, uint64_t Address, @@ -548,6 +559,15 @@ DecodeStatus AMDGPUDisassembler::getInstruction(MCInst &MI, uint64_t &Size, // Reinitialize Bytes Bytes = Bytes_.slice(0, MaxInstBytesNum); + + } else if (Bytes.size() >= 16 && + STI.hasFeature(AMDGPU::FeatureGFX950Insts)) { + DecoderUInt128 DecW = eat16Bytes(Bytes); + if (tryDecodeInst(DecoderTableGFX940128, MI, DecW, Address, CS)) + break; + + // Reinitialize Bytes + Bytes = Bytes_.slice(0, MaxInstBytesNum); } if (Bytes.size() >= 8) { @@ -759,6 +779,9 @@ DecodeStatus AMDGPUDisassembler::getInstruction(MCInst &MI, uint64_t &Size, if (MCII->get(MI.getOpcode()).TSFlags & SIInstrFlags::SDWA) convertSDWAInst(MI); + if (MCII->get(MI.getOpcode()).TSFlags & SIInstrFlags::IsMAI) + convertMAIInst(MI); + int VDstIn_Idx = AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::vdst_in); if (VDstIn_Idx != -1) { @@ -837,6 +860,58 @@ void AMDGPUDisassembler::convertSDWAInst(MCInst &MI) const { } } +/// Adjust the register values used by V_MFMA_F8F6F4_f8_f8 instructions to the +/// appropriate subregister for the used format width. +static void adjustMFMA_F8F6F4OpRegClass(const MCRegisterInfo &MRI, + MCOperand &MO, uint8_t NumRegs) { + switch (NumRegs) { + case 4: + return MO.setReg(MRI.getSubReg(MO.getReg(), AMDGPU::sub0_sub1_sub2_sub3)); + case 6: + return MO.setReg( + MRI.getSubReg(MO.getReg(), AMDGPU::sub0_sub1_sub2_sub3_sub4_sub5)); + case 8: + // No-op in cases where one operand is still f8/bf8. + return; + default: + llvm_unreachable("Unexpected size for mfma f8f6f4 operand"); + } +} + +/// f8f6f4 instructions have different pseudos depending on the used formats. In +/// the disassembler table, we only have the variants with the largest register +/// classes which assume using an fp8/bf8 format for both operands. The actual +/// register class depends on the format in blgp and cbsz operands. Adjust the +/// register classes depending on the used format. +void AMDGPUDisassembler::convertMAIInst(MCInst &MI) const { + int BlgpIdx = + AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::blgp); + if (BlgpIdx == -1) + return; + + int CbszIdx = + AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::cbsz); + + unsigned CBSZ = MI.getOperand(CbszIdx).getImm(); + unsigned BLGP = MI.getOperand(BlgpIdx).getImm(); + + const AMDGPU::MFMA_F8F6F4_Info *AdjustedRegClassOpcode = + AMDGPU::getMFMA_F8F6F4_WithFormatArgs(CBSZ, BLGP, MI.getOpcode()); + if (!AdjustedRegClassOpcode || + AdjustedRegClassOpcode->Opcode == MI.getOpcode()) + return; + + MI.setOpcode(AdjustedRegClassOpcode->Opcode); + int Src0Idx = + AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::src0); + int Src1Idx = + AMDGPU::getNamedOperandIdx(MI.getOpcode(), AMDGPU::OpName::src1); + adjustMFMA_F8F6F4OpRegClass(MRI, MI.getOperand(Src0Idx), + AdjustedRegClassOpcode->NumRegsSrcA); + adjustMFMA_F8F6F4OpRegClass(MRI, MI.getOperand(Src1Idx), + AdjustedRegClassOpcode->NumRegsSrcB); +} + struct VOPModifiers { unsigned OpSel = 0; unsigned OpSelHi = 0; diff --git a/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.h b/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.h index 694cd7a9bfd282..3e20a2ab9e66ca 100644 --- a/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.h +++ b/llvm/lib/Target/AMDGPU/Disassembler/AMDGPUDisassembler.h @@ -204,6 +204,7 @@ class AMDGPUDisassembler : public MCDisassembler { void convertVINTERPInst(MCInst &MI) const; void convertFMAanyK(MCInst &MI, int ImmLitIdx) const; void convertSDWAInst(MCInst &MI) const; + void convertMAIInst(MCInst &MI) const; void convertDPP8Inst(MCInst &MI) const; void convertMIMGInst(MCInst &MI) const; void convertVOP3DPPInst(MCInst &MI) const; diff --git a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCAsmInfo.cpp b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCAsmInfo.cpp index 3c9f6d2938075b..56ed29ede02c23 100644 --- a/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCAsmInfo.cpp +++ b/llvm/lib/Target/AMDGPU/MCTargetDesc/AMDGPUMCAsmInfo.cpp @@ -59,6 +59,10 @@ unsigned AMDGPUMCAsmInfo::getMaxInstLength(const MCSubtargetInfo *STI) const { if (STI->hasFeature(AMDGPU::FeatureNSAEncoding)) return 20; + // VOP3PX encoding. + if (STI->hasFeature(AMDGPU::FeatureGFX950Insts)) + return 16; + // 64-bit instruction with 32-bit literal. if (STI->hasFeature(AMDGPU::FeatureVOP3Literal)) return 12; diff --git a/llvm/lib/Target/AMDGPU/SIDefines.h b/llvm/lib/Target/AMDGPU/SIDefines.h index 8f297726a0df88..f812ae652b63d0 100644 --- a/llvm/lib/Target/AMDGPU/SIDefines.h +++ b/llvm/lib/Target/AMDGPU/SIDefines.h @@ -1048,6 +1048,18 @@ enum Offset_COV5 : unsigned { } // namespace ImplicitArg +namespace MFMAScaleFormats { +// Enum value used in cbsz/blgp for F8F6F4 MFMA operations to select the matrix +// format. +enum MFMAScaleFormats { + FP8_E4M3 = 0, + FP8_E5M2 = 1, + FP6_E2M3 = 2, + FP6_E3M2 = 3, + FP4_E2M1 = 4 +}; +} // namespace MFMAScaleFormats + namespace VirtRegFlag { // Virtual register flags used for various target specific handlings during // codegen. diff --git a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp index 8c8d55bbc01316..1406938592b2cb 100644 --- a/llvm/lib/Target/AMDGPU/SIISelLowering.cpp +++ b/llvm/lib/Target/AMDGPU/SIISelLowering.cpp @@ -15454,6 +15454,23 @@ void SITargetLowering::AdjustInstrPostInstrSelection(MachineInstr &MI, MRI.setRegClass(Op.getReg(), NewRC); } + if (TII->isMAI(MI)) { + // The ordinary src0, src1, src2 were legalized above. + // + // We have to also legalize the appended v_mfma_ld_scale_b32 operands, + // as a separate instruction. + int Src0Idx = AMDGPU::getNamedOperandIdx(MI.getOpcode(), + AMDGPU::OpName::scale_src0); + if (Src0Idx != -1) { + int Src1Idx = Src0Idx + 2; + assert(Src1Idx = AMDGPU::getNamedOperandIdx( + MI.getOpcode(), AMDGPU::OpName::scale_src1)); + if (TII->usesConstantBus(MRI, MI, Src0Idx) && + TII->usesConstantBus(MRI, MI, Src1Idx)) + TII->legalizeOpWithMove(MI, Src1Idx); + } + } + if (!HasAGPRs) return; diff --git a/llvm/lib/Target/AMDGPU/SIInstrFormats.td b/llvm/lib/Target/AMDGPU/SIInstrFormats.td index dd1ab2c628715d..267c9a94b90968 100644 --- a/llvm/lib/Target/AMDGPU/SIInstrFormats.td +++ b/llvm/lib/Target/AMDGPU/SIInstrFormats.td @@ -300,6 +300,11 @@ class Enc96 { int Size = 12; } +class Enc128 { + field bits<128> Inst; + int Size = 16; +} + def CPolBit { int GLC = 0; int SLC = 1; diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.h b/llvm/lib/Target/AMDGPU/SIInstrInfo.h index 1f7fff76d15210..e55418326a4bd0 100644 --- a/llvm/lib/Target/AMDGPU/SIInstrInfo.h +++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.h @@ -1115,6 +1115,12 @@ class SIInstrInfo final : public AMDGPUGenInstrInfo { const MachineOperand &MO, const MCOperandInfo &OpInfo) const; + bool usesConstantBus(const MachineRegisterInfo &MRI, const MachineInstr &MI, + int OpIdx) const { + return usesConstantBus(MRI, MI.getOperand(OpIdx), + MI.getDesc().operands()[OpIdx]); + } + /// Return true if this instruction has any modifiers. /// e.g. src[012]_mod, omod, clamp. bool hasModifiers(unsigned Opcode) const; diff --git a/llvm/lib/Target/AMDGPU/SIInstrInfo.td b/llvm/lib/Target/AMDGPU/SIInstrInfo.td index d2024cf915874d..a6496cd4a61f19 100644 --- a/llvm/lib/Target/AMDGPU/SIInstrInfo.td +++ b/llvm/lib/Target/AMDGPU/SIInstrInfo.td @@ -914,6 +914,16 @@ def fp16_zeros_high_16bits : PatLeaf<(f16 VGPR_32:$src), [{ return fp16SrcZerosHighBits(N->getOpcode()); }]>; +def MFMALdScaleXForm : SDNodeXFormgetZExtValue(); + unsigned New = 0; + if (Val & 0x1) + New |= SISrcMods::OP_SEL_0; + if (Val & 0x2) + New |= SISrcMods::OP_SEL_1; + return CurDAG->getTargetConstant(New, SDLoc(N), MVT::i32); +}]>; + def is_canonicalized : PatLeaf<(fAny srcvalue:$src), [{ const SITargetLowering &Lowering = *static_cast(getTargetLowering()); @@ -1515,6 +1525,10 @@ class PackedIntInputMods : InputMods < def PackedF16InputMods : PackedFPInputMods; def PackedI16InputMods : PackedIntInputMods; +def MFMALdScaleModifierOp : TImmLeaf(Imm); +}], MFMALdScaleXForm>; + //===----------------------------------------------------------------------===// // Complex patterns //===----------------------------------------------------------------------===// @@ -2655,6 +2669,8 @@ class VOPProfile _ArgVT, bit _EnableClamp = 0> { field string AsmVOPDX = getAsmVOPDPart.ret; field string AsmVOPDY = getAsmVOPDPart.ret; field string TieRegDPP = "$old"; + field bit IsSMFMAC = false; + field bit HasAbid = !and(IsMAI, HasSrc1); } class VOP_NO_EXT : VOPProfile { @@ -2851,6 +2867,31 @@ def VOP_V16F32_V2I32_V4I32_I32 : VOPProfile <[v16f32, v2i32, v4i32, i32]>; def VOP_V4F32_V8F16_V8F16_V4F32 : VOPProfile <[v4f32, v8f16, v8f16, v4f32]>; def VOP_V16F32_V8F16_V8F16_V16F32 : VOPProfile <[v16f32, v8f16, v8f16, v16f32]>; def VOP_V16F32_V8BF16_V8BF16_V16F32 : VOPProfile <[v16f32, v8bf16, v8bf16, v16f32]>; +def VOP_V4F32_V8I32_V8I32_V4F32 : VOPProfile <[v4f32, v8i32, v8i32, v4f32]>; + +def VOP_V4F32_V8I32_V6I32_V4F32 : VOPProfile <[v4f32, v8i32, v6i32, v4f32]>; +def VOP_V4F32_V6I32_V8I32_V4F32 : VOPProfile <[v4f32, v6i32, v8i32, v4f32]>; +def VOP_V4F32_V6I32_V6I32_V4F32 : VOPProfile <[v4f32, v6i32, v6i32, v4f32]>; + +def VOP_V4F32_V8I32_V4I32_V4F32 : VOPProfile <[v4f32, v8i32, v4i32, v4f32]>; +def VOP_V4F32_V4I32_V8I32_V4F32 : VOPProfile <[v4f32, v4i32, v8i32, v4f32]>; +def VOP_V4F32_V6I32_V4I32_V4F32 : VOPProfile <[v4f32, v6i32, v4i32, v4f32]>; +def VOP_V4F32_V4I32_V6I32_V4F32 : VOPProfile <[v4f32, v4i32, v6i32, v4f32]>; +def VOP_V4F32_V4I32_V4I32_V4F32 : VOPProfile <[v4f32, v4i32, v4i32, v4f32]>; + +def VOP_V16F32_V8I32_V8I32_V16F32 : VOPProfile <[v16f32, v8i32, v8i32, v16f32]>; +def VOP_V16F32_V8I32_V6I32_V16F32 : VOPProfile <[v16f32, v8i32, v6i32, v16f32]>; +def VOP_V16F32_V6I32_V8I32_V16F32 : VOPProfile <[v16f32, v6i32, v8i32, v16f32]>; +def VOP_V16F32_V6I32_V6I32_V16F32 : VOPProfile <[v16f32, v6i32, v6i32, v16f32]>; + +def VOP_V16F32_V8I32_V4I32_V16F32 : VOPProfile <[v16f32, v8i32, v4i32, v16f32]>; +def VOP_V16F32_V4I32_V8I32_V16F32 : VOPProfile <[v16f32, v4i32, v8i32, v16f32]>; +def VOP_V16F32_V6I32_V4I32_V16F32 : VOPProfile <[v16f32, v6i32, v4i32, v16f32]>; +def VOP_V16F32_V4I32_V6I32_V16F32 : VOPProfile <[v16f32, v4i32, v6i32, v16f32]>; +def VOP_V16F32_V4I32_V4I32_V16F32 : VOPProfile <[v16f32, v4i32, v4i32, v16f32]>; + +def VOP_V4I32_V4I32_V4I32_V4I32 : VOPProfile <[v4i32, v4i32, v4i32, v4i32]>; +def VOP_V16I32_V4I32_V4I32_V16I32 : VOPProfile <[v16i32, v4i32, v4i32, v16i32]>; class Commutable_REV { @@ -3114,6 +3155,16 @@ def getVCMPXOpFromVCMP : InstrMapping { let ValueCols = [["1"]]; } +// Map encoded mfma(_scale)?_f8f6f4 instructions depending on the +// number of registers required for the used format. +def getMFMA_F8F6F4_WithSize : GenericTable { + let FilterClass = "MFMA_F8F6F4_WithSizeTable"; + let CppTypeName = "MFMA_F8F6F4_Info"; + let Fields = [ "Opcode", "F8F8Opcode", "NumRegsSrcA", "NumRegsSrcB" ]; + let PrimaryKey = [ "NumRegsSrcA", "NumRegsSrcB", "F8F8Opcode" ]; + let PrimaryKeyName = "getMFMA_F8F6F4_InstWithNumRegs" ; +} + def FP8DstByteSelTable : GenericTable { let FilterClass = "VOP3_Pseudo"; let CppTypeName = "FP8DstByteSelInfo"; diff --git a/llvm/lib/Target/AMDGPU/SIRegisterInfo.td b/llvm/lib/Target/AMDGPU/SIRegisterInfo.td index 46b2b4a389200a..e3baeed01841ab 100644 --- a/llvm/lib/Target/AMDGPU/SIRegisterInfo.td +++ b/llvm/lib/Target/AMDGPU/SIRegisterInfo.td @@ -1346,11 +1346,14 @@ class AVSrcOperand def AVSrc_32 : AVSrcOperand; def AVSrc_64 : AVSrcOperand; def AVSrc_128 : AVSrcOperand; +def AVSrc_192 : AVSrcOperand; +def AVSrc_256 : AVSrcOperand; class AVDstOperand : AVOperand; def AVDst_128 : AVDstOperand; +def AVDst_256 : AVDstOperand; def AVDst_512 : AVDstOperand; class AVLdStOperand diff --git a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp index 501d00b1f308d9..c7e4659b15d299 100644 --- a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp +++ b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.cpp @@ -415,6 +415,8 @@ struct FP8DstByteSelInfo { #define GET_WMMAOpcode2AddrMappingTable_IMPL #define GET_WMMAOpcode3AddrMappingTable_DECL #define GET_WMMAOpcode3AddrMappingTable_IMPL +#define GET_getMFMA_F8F6F4_WithSize_DECL +#define GET_getMFMA_F8F6F4_WithSize_IMPL #include "AMDGPUGenSearchableTables.inc" int getMTBUFBaseOpcode(unsigned Opc) { @@ -523,6 +525,30 @@ bool getMAIIsGFX940XDL(unsigned Opc) { return Info ? Info->is_gfx940_xdl : false; } +static uint8_t mfmaScaleF8F6F4FormatToNumRegs(unsigned EncodingVal) { + switch (EncodingVal) { + case MFMAScaleFormats::FP6_E2M3: + case MFMAScaleFormats::FP6_E3M2: + return 6; + case MFMAScaleFormats::FP4_E2M1: + return 4; + case MFMAScaleFormats::FP8_E4M3: + case MFMAScaleFormats::FP8_E5M2: + default: + return 8; + } + + llvm_unreachable("covered switch over mfma scale formats"); +} + +const MFMA_F8F6F4_Info *getMFMA_F8F6F4_WithFormatArgs(unsigned CBSZ, + unsigned BLGP, + unsigned F8F8Opcode) { + uint8_t SrcANumRegs = mfmaScaleF8F6F4FormatToNumRegs(CBSZ); + uint8_t SrcBNumRegs = mfmaScaleF8F6F4FormatToNumRegs(BLGP); + return getMFMA_F8F6F4_InstWithNumRegs(SrcANumRegs, SrcBNumRegs, F8F8Opcode); +} + unsigned getVOPDEncodingFamily(const MCSubtargetInfo &ST) { if (ST.hasFeature(AMDGPU::FeatureGFX12Insts)) return SIEncodingFamily::GFX12; @@ -2912,6 +2938,7 @@ const AlwaysUniform *lookupAlwaysUniform(unsigned Intr); #define GET_Gfx9BufferFormat_IMPL #define GET_Gfx10BufferFormat_IMPL #define GET_Gfx11PlusBufferFormat_IMPL + #include "AMDGPUGenSearchableTables.inc" } // end anonymous namespace diff --git a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h index 88a6d75b72c7d0..b0581711961b13 100644 --- a/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h +++ b/llvm/lib/Target/AMDGPU/Utils/AMDGPUBaseInfo.h @@ -581,6 +581,18 @@ unsigned getVOPDEncodingFamily(const MCSubtargetInfo &ST); LLVM_READONLY CanBeVOPD getCanBeVOPD(unsigned Opc); +struct MFMA_F8F6F4_Info { + unsigned Opcode; + unsigned F8F8Opcode; + uint8_t NumRegsSrcA; + uint8_t NumRegsSrcB; +}; + +LLVM_READONLY +const MFMA_F8F6F4_Info *getMFMA_F8F6F4_WithFormatArgs(unsigned CBSZ, + unsigned BLGP, + unsigned F8F8Opcode); + LLVM_READONLY const GcnBufferFormatInfo *getGcnBufferFormatInfo(uint8_t BitsPerComp, uint8_t NumComponents, diff --git a/llvm/lib/Target/AMDGPU/VOP3PInstructions.td b/llvm/lib/Target/AMDGPU/VOP3PInstructions.td index 876d4e1acf5964..a6ec1dba23aad2 100644 --- a/llvm/lib/Target/AMDGPU/VOP3PInstructions.td +++ b/llvm/lib/Target/AMDGPU/VOP3PInstructions.td @@ -18,6 +18,7 @@ class VOP3P_Profile, VOP3P_LD_SCALE> { @@ -542,19 +543,23 @@ def VOPProfileAccWrite : VOP3P_Profile { } class VOPProfileMAI + RegisterOperand SrcARC = AVSrc_32, RegisterOperand SrcBRC = SrcARC> : VOP3P_Profile { + bit HasAbid = true; let DstRC = _DstRC; - let Src0RC64 = SrcABRC; - let Src1RC64 = SrcABRC; + let Src0RC64 = SrcARC; + let Src1RC64 = SrcBRC; let Src2RC64 = _SrcRC; let HasOpSel = 0; let HasClamp = 0; let HasIntClamp = 0; let HasOMod = 0; let HasModifiers = 0; - let AsmVOP3Base = "$vdst, $src0, $src1, $src2$cbsz$abid$blgp"; - let Ins64 = (ins Src0RC64:$src0, Src1RC64:$src1, Src2RC64:$src2, CBSZ:$cbsz, ABID:$abid, blgp:$blgp); + let AsmVOP3Base = "$vdst, $src0, $src1, $src2$cbsz"#!if(HasAbid,"$abid","")#"$blgp"; + let Ins64 = !con( + (ins Src0RC64:$src0, Src1RC64:$src1, Src2RC64:$src2, CBSZ:$cbsz), + !if(HasAbid, (ins ABID:$abid), (ins)), + (ins blgp:$blgp)); let InsVOP3Base = Ins64; // Dst and SrcC cannot partially overlap if SrcC/Dst is bigger than 4 VGPRs. // We then create two versions of the instruction: with tied dst and src2 @@ -572,6 +577,7 @@ class VOPProfileSMFMAC; @@ -639,14 +645,112 @@ def VOPProfileMAI_F32_V8F16_X16_VCD : VOPProfileMAI; def VOPProfileMAI_F32_V8BF16_X16_VCD : VOPProfileMAI; + +let HasAbid = false in { +// For f32_16x16x128_f8f6f4 - f8 x f8 case +def VOPProfileMAI_F32_V8I32_V8I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V8I32_V8I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f8 x f6 case +def VOPProfileMAI_F32_V8I32_V6I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V8I32_V6I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f6 x f8 case +def VOPProfileMAI_F32_V6I32_V8I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V6I32_V8I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f6 x f6 case +def VOPProfileMAI_F32_V6I32_V6I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V6I32_V6I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f6 x f4 case +def VOPProfileMAI_F32_V6I32_V4I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V6I32_V4I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f4 x f6 case +def VOPProfileMAI_F32_V4I32_V6I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V4I32_V6I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f8 x f4 case +def VOPProfileMAI_F32_V8I32_V4I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V8I32_V4I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f4 x f8 case +def VOPProfileMAI_F32_V4I32_V8I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V4I32_V8I32_X128_VCD : VOPProfileMAI; + +// For f32_16x16x128_f8f6f4 - f4 x f4 case +def VOPProfileMAI_F32_V4I32_V4I32_X128 : VOPProfileMAI; +def VOPProfileMAI_F32_V4I32_V4I32_X128_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f8 x f8 case +def VOPProfileMAI_F32_V8I32_V8I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V8I32_V8I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f8 x f6 case +def VOPProfileMAI_F32_V8I32_V6I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V8I32_V6I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f8 x f4 case +def VOPProfileMAI_F32_V8I32_V4I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V8I32_V4I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f4 x f8 case +def VOPProfileMAI_F32_V4I32_V8I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V4I32_V8I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f6 x f8 case +def VOPProfileMAI_F32_V6I32_V8I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V6I32_V8I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f6 x f6 case +def VOPProfileMAI_F32_V6I32_V6I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V6I32_V6I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f6 x f4 case +def VOPProfileMAI_F32_V6I32_V4I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V6I32_V4I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f4 x f6 case +def VOPProfileMAI_F32_V4I32_V6I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V4I32_V6I32_X512_VCD : VOPProfileMAI; + +// For f32_32x32x64_f8f6f4 - f4 x f4 case +def VOPProfileMAI_F32_V4I32_V4I32_X512 : VOPProfileMAI; +def VOPProfileMAI_F32_V4I32_V4I32_X512_VCD : VOPProfileMAI; +} + + class MFMATable { bit IsMac = is_mac; string FMAOp = Name; } -class MAIFrag : PatFrag < - (ops node:$src0, node:$src1, node:$src2, node:$cbsz, node:$abid, node:$blgp), - (Op $src0, $src1, $src2, $cbsz, $abid, $blgp), +class MFMA_F8F6F4_WithSizeTable { + Instruction F8F8Opcode = F8F8Variant; + Instruction Opcode = ThisVariant; + bits<8> NumRegsSrcA = A; + bits<8> NumRegsSrcB = B; +} + +class MFMA_F8F6F4_WithSizeTable_Helper : + MFMA_F8F6F4_WithSizeTable(NAME), + !cast(F8F8Op)> { +} + +// Currently assumes scaled instructions never have abid +class MAIFrag : PatFrag < + !if(Scaled, (ops node:$src0, node:$src1, node:$src2, node:$cbsz, node:$blgp, + node:$scale_src0_opsel, node:$scale_src0, + node:$scale_src1_opsel, node:$scale_src1), + !con((ops node:$src0, node:$src1, node:$src2, node:$cbsz), + !if(HasAbid, (ops node:$abid), (ops)), + (ops node:$blgp))), + !if(Scaled, (Op $src0, $src1, $src2, $cbsz, $blgp, $scale_src0_opsel, $scale_src0, $scale_src1_opsel, $scale_src1), + !if(HasAbid, (Op $src0, $src1, $src2, $cbsz, $abid, $blgp), + (Op $src0, $src1, $src2, $cbsz, $blgp))), pred >; @@ -666,11 +770,15 @@ defvar MayNotNeedAGPRs_gisel = [{ return !MF.getInfo()->mayNeedAGPRs(); }]; -class AgprMAIFrag : MAIFrag { +class AgprMAIFrag : + MAIFrag { let GISelPredicateCode = MayNeedAGPRs_gisel; } -class VgprMAIFrag : MAIFrag { +class VgprMAIFrag : + MAIFrag { let GISelPredicateCode = MayNotNeedAGPRs_gisel; } @@ -681,27 +789,51 @@ let isAsCheapAsAMove = 1, isReMaterializable = 1 in { } // End isMoveImm = 1 } // End isAsCheapAsAMove = 1, isReMaterializable = 1 -class MAIInst - : VOP3InstBase { +class MAIInst + : VOP3InstBase { let SubtargetPredicate = HasMAIInsts; Instruction Opcode = !cast(NAME); bit is_dgemm = 0; bit is_gfx940_xdl = 0; + let PseudoInstr = NAME; // FIXME: Why is this not the default } -multiclass MAIInst { +// FIXME: Intrinsic should probably not have op_sel operands, we can +// pattern match byte select patterns into op_sel. +// FIXME: Missing neg and clamp modifiers +// +// FIXME: Usual syntax for op_sel is quite hostile here. +class ScaledMAIInst : + MAIInst { + // Append operands from V_MFMA_LD_SCALE_B32, but we need to rename them. + let InOperandList = !con(BaseInst.InOperandList, + (ins VSrc_b32:$scale_src0, + VSrc_b32:$scale_src1, + op_sel0:$scale_src0_opsel, + op_sel_hi0:$scale_src1_opsel)); + let AsmOperands = + "$vdst, $src0, $src1, $src2, $scale_src0, $scale_src1" + "$scale_src0_opsel$scale_src1_opsel$cbsz$blgp"; + + let FixedSize = 1; + let Size = 16; +} + +multiclass MAIInst { defvar NoDstOverlap = !cast("VOPProfileMAI_" # P).NoDstOverlap; let isConvergent = 1, mayRaiseFPException = 0, ReadsModeReg = 1 in { // FP32 denorm mode is respected, rounding mode is not. Exceptions are not supported. let Constraints = !if(NoDstOverlap, "@earlyclobber $vdst", "") in { def _e64 : MAIInst("VOPProfileMAI_" # P), - !if(!or(NoDstOverlap, !eq(node, null_frag)), null_frag, AgprMAIFrag)>, + !if(!or(NoDstOverlap, !eq(node, null_frag)), null_frag, AgprMAIFrag), Scaled>, MFMATable<0, NAME # "_e64">; let OtherPredicates = [isGFX90APlus], Mnemonic = OpName in def _vgprcd_e64 : MAIInst("VOPProfileMAI_" # P # "_VCD"), - !if(!or(NoDstOverlap, !eq(node, null_frag)), null_frag, VgprMAIFrag)>, + !if(!or(NoDstOverlap, !eq(node, null_frag)), null_frag, VgprMAIFrag), Scaled>, MFMATable<0, NAME # "_vgprcd_e64">; } @@ -710,18 +842,77 @@ multiclass MAIInst { isConvertibleToThreeAddress = NoDstOverlap, Mnemonic = OpName in { def "_mac_e64" : MAIInst("VOPProfileMAI_" # P), - !if(!eq(node, null_frag), null_frag, AgprMAIFrag)>, + !if(!eq(node, null_frag), null_frag, AgprMAIFrag), Scaled>, MFMATable<1, NAME # "_e64">; let OtherPredicates = [isGFX90APlus] in def _mac_vgprcd_e64 : MAIInst("VOPProfileMAI_" # P # "_VCD"), - !if(!eq(node, null_frag), null_frag, VgprMAIFrag)>, + !if(!eq(node, null_frag), null_frag, VgprMAIFrag), Scaled>, MFMATable<1, NAME # "_vgprcd_e64">; } } } // End isConvergent = 1, mayRaiseFPException = 0, ReadsModeReg = 1 } +// Provide a wrapper around MAIInst that provides the appended operands from V_MFMA_LD_SCALE_B32 +multiclass ScaledMAIInst_mc { + defvar VariantSuffix = !subst(!toupper(OpName), "", NAME); // Drop the main opcode name prefix to get the "_fN_fM" suffix. + defvar UnscaledOpName = UnscaledOpName_#VariantSuffix; + + defvar HasAbid = false; + + defvar NoDstOverlap = !cast(!cast(UnscaledOpName#"_e64").Pfl).NoDstOverlap; + + def _e64 : ScaledMAIInst(UnscaledOpName#"_e64"), !if(NoDstOverlap, null_frag, AgprMAIFrag)>, + MFMATable<0, NAME # "_e64">; + + def _vgprcd_e64 : ScaledMAIInst(UnscaledOpName#"_vgprcd_e64"), !if(NoDstOverlap, null_frag, VgprMAIFrag)>, + MFMATable<0, NAME # "_vgprcd_e64">; + + if NoDstOverlap then { + let Constraints = !if(NoDstOverlap, "$vdst = $src2", ""), + isConvertibleToThreeAddress = NoDstOverlap, + Mnemonic = UnscaledOpName_ in { + def _mac_e64 : ScaledMAIInst(UnscaledOpName # "_mac_e64"), AgprMAIFrag>, + MFMATable<1, NAME # "_e64">; + + def _mac_vgprcd_e64 : ScaledMAIInst(UnscaledOpName # "_mac_vgprcd_e64"), VgprMAIFrag>, + MFMATable<1, NAME # "_vgprcd_e64">; + } + } +} + +// Each of SrcA and SrcB can be encoded using 3 different sizes, so +// define 9 permutations of register classes. +multiclass MAIInst_SrcFormats_mc { + defvar HasAbid = false; + defm _f8_f8 : MAIInst; + defm _f8_f6 : MAIInst; + defm _f6_f8 : MAIInst; + defm _f6_f6 : MAIInst; + defm _f8_f4 : MAIInst; + defm _f4_f8 : MAIInst; + defm _f6_f4 : MAIInst; + defm _f4_f6 : MAIInst; + defm _f4_f4 : MAIInst; +} + +multiclass MAIInst_SrcFormats_Scaled_mc { + defm _f8_f8 : ScaledMAIInst_mc; + defm _f8_f6 : ScaledMAIInst_mc; + defm _f6_f8 : ScaledMAIInst_mc; + defm _f6_f6 : ScaledMAIInst_mc; + defm _f8_f4 : ScaledMAIInst_mc; + defm _f4_f8 : ScaledMAIInst_mc; + defm _f6_f4 : ScaledMAIInst_mc; + defm _f4_f6 : ScaledMAIInst_mc; + defm _f4_f4 : ScaledMAIInst_mc; +} + defm V_MFMA_F32_4X4X1F32 : MAIInst<"v_mfma_f32_4x4x1f32", "F32_F32_X4", int_amdgcn_mfma_f32_4x4x1f32>; defm V_MFMA_F32_16X16X1F32 : MAIInst<"v_mfma_f32_16x16x1f32", "F32_F32_X16", int_amdgcn_mfma_f32_16x16x1f32>; defm V_MFMA_F32_16X16X4F32 : MAIInst<"v_mfma_f32_16x16x4f32", "F32_F32_X4", int_amdgcn_mfma_f32_16x16x4f32>; @@ -753,6 +944,20 @@ let SubtargetPredicate = HasGFX950Insts, is_gfx940_xdl = 1 in { defm V_MFMA_F32_16X16X32_F16 : MAIInst<"v_mfma_f32_16x16x32f16", "F32_V8F16_X32", int_amdgcn_mfma_f32_16x16x32_f16>; defm V_MFMA_F32_32X32X16_F16 : MAIInst<"v_mfma_f32_32x32x16f16", "F32_V8F16_X16", int_amdgcn_mfma_f32_32x32x16_f16>; defm V_MFMA_F32_32X32X16_BF16 : MAIInst<"v_mfma_f32_32x32x16bf16", "F32_V8BF16_X16", int_amdgcn_mfma_f32_32x32x16_bf16>; + +defm V_MFMA_F32_16X16X128_F8F6F4 : MAIInst_SrcFormats_mc<"v_mfma_f32_16x16x128f8f6f4", + "_X128">; +defm V_MFMA_F32_32X32X64_F8F6F4 : MAIInst_SrcFormats_mc<"v_mfma_f32_32x32x64f8f6f4", + "_X512">; + +defm V_MFMA_SCALE_F32_16X16X128_F8F6F4 : MAIInst_SrcFormats_Scaled_mc< + "v_mfma_scale_f32_16x16x128_f8f6f4", "V_MFMA_F32_16X16X128_F8F6F4", + int_amdgcn_mfma_scale_f32_16x16x128_f8f6f4>; + +defm V_MFMA_SCALE_F32_32X32X64_F8F6F4 : MAIInst_SrcFormats_Scaled_mc< + "v_mfma_scale_f32_32x32x64_f8f6f4", + "V_MFMA_F32_32X32X64_F8F6F4", + int_amdgcn_mfma_scale_f32_32x32x64_f8f6f4>; } let SubtargetPredicate = HasGFX950Insts in { @@ -869,7 +1074,6 @@ def VOP_V4F32_V16F16_V16F16_V4F32 : VOPProfile <[v4f32, v16f16, v16f16, v4f32]>; def VOP_V4F32_V16I16_V16I16_V4F32 : VOPProfile <[v4f32, v16i16, v16i16, v4f32]>; def VOP_V8F16_V16F16_V16F16_V8F16 : VOPProfile <[v8f16, v16f16, v16f16, v8f16]>; def VOP_V8I16_V16I16_V16I16_V8I16 : VOPProfile <[v8i16, v16i16, v16i16, v8i16]>; -def VOP_V4I32_V4I32_V4I32_V4I32 : VOPProfile <[v4i32, v4i32, v4i32, v4i32]>; def VOP_V4I32_V2I32_V2I32_V4I32 : VOPProfile <[v4i32, v2i32, v2i32, v4i32]>; @@ -1676,6 +1880,26 @@ multiclass VOP3P_Real_MFMA_gfx940 op, string Name = !cast(N } } +multiclass VOP3P_Real_MFMA_F8F6F4_gfx940 op, string Name = !cast(NAME#"_e64").Mnemonic, + VOP3_Pseudo PS_ACD = !cast(NAME # "_e64"), + VOP3_Pseudo PS_VCD = !cast(NAME # "_vgprcd" # "_e64")> { + + defvar F8F8Name = !substr(NAME, 0, !sub(!size(NAME), !size("_fN_fM")))#"_f8_f8"; + + let AssemblerPredicate = isGFX940Plus, + DecoderNamespace = "GFX940", + AsmString = Name # PS_ACD.AsmOperands, + Constraints = "" in { + def _gfx940_acd : VOP3P_Real, + VOP3Pe_MAI , + MFMA_F8F6F4_WithSizeTable_Helper; + + def _gfx940_vcd : VOP3P_Real, + VOP3Pe_MAI , + MFMA_F8F6F4_WithSizeTable_Helper; + } // End AssemblerPredicate = isGFX940Plus, DecoderNamespace = "GFX940" +} + multiclass VOP3P_Real_MFMA_gfx950 op, string Name = !cast(NAME#"_e64").Mnemonic, VOP3_Pseudo PS_ACD = !cast(NAME # "_e64"), VOP3_Pseudo PS_VCD = !cast(NAME # "_vgprcd" # "_e64")> { @@ -1686,6 +1910,55 @@ multiclass VOP3P_Real_MFMA_gfx950 op, string Name = !cast(N } +multiclass VOP3P_Real_MFMA_F8F6F4_gfx950_mc op, string Name> { + defm _f8_f8 : VOP3P_Real_MFMA_F8F6F4_gfx940; + + let isAsmParserOnly = true in { // Disable ambiguous disassembly. + defm _f8_f6 : VOP3P_Real_MFMA_F8F6F4_gfx940; + defm _f6_f8 : VOP3P_Real_MFMA_F8F6F4_gfx940; + defm _f8_f4 : VOP3P_Real_MFMA_F8F6F4_gfx940; + defm _f4_f8 : VOP3P_Real_MFMA_F8F6F4_gfx940; + defm _f6_f6 : VOP3P_Real_MFMA_F8F6F4_gfx940; + defm _f6_f4 : VOP3P_Real_MFMA_F8F6F4_gfx940; + defm _f4_f6 : VOP3P_Real_MFMA_F8F6F4_gfx940; + defm _f4_f4 : VOP3P_Real_MFMA_F8F6F4_gfx940; + } +} + +multiclass VOP3PX_Real_ScaledMFMA op> { + defvar PS_ACD = !cast(NAME # "_e64"); + defvar PS_VCD = !cast(NAME # "_vgprcd" # "_e64"); + defvar Name = PS_ACD.Mnemonic; + defvar F8F8Name = !substr(NAME, 0, !sub(!size(NAME), !size("_fN_fM")))#"_f8_f8"; + + let SubtargetPredicate = HasGFX950Insts, + DecoderNamespace = "GFX940", + AsmString = Name # PS_ACD.AsmOperands, Constraints = "" in { + def _gfx940_acd : VOP3P_Real, + VOP3PXe , + MFMA_F8F6F4_WithSizeTable_Helper; + + def _gfx940_vcd : VOP3P_Real, + VOP3PXe , + MFMA_F8F6F4_WithSizeTable_Helper; + } +} + +multiclass VOP3PX_Real_ScaledMFMA_F8F6F4_mc op> { + defm _f8_f8 : VOP3PX_Real_ScaledMFMA; + + let isAsmParserOnly = 1 in { // Disable ambiguous disassembly. + defm _f8_f6 : VOP3PX_Real_ScaledMFMA; + defm _f6_f8 : VOP3PX_Real_ScaledMFMA; + defm _f8_f4 : VOP3PX_Real_ScaledMFMA; + defm _f4_f8 : VOP3PX_Real_ScaledMFMA; + defm _f6_f6 : VOP3PX_Real_ScaledMFMA; + defm _f6_f4 : VOP3PX_Real_ScaledMFMA; + defm _f4_f6 : VOP3PX_Real_ScaledMFMA; + defm _f4_f4 : VOP3PX_Real_ScaledMFMA; + } +} + multiclass VOP3P_Real_MFMA_vi op> { def _vi : VOP3P_Real(NAME#"_e64"), SIEncodingFamily.VI>, VOP3Pe_MAI (NAME#"_e64").Pfl, ?> { @@ -1797,6 +2070,10 @@ defm V_MFMA_F32_32X32X16_F16 : VOP3P_Real_MFMA_gfx950 <0x55, "v_mfma_f32_32x defm V_MFMA_F32_32X32X16_BF16 : VOP3P_Real_MFMA_gfx950 <0x37, "v_mfma_f32_32x32x16_bf16">; defm V_MFMA_LD_SCALE_B32 : VOP3P_Real_vi <0x2c>; +defm V_MFMA_F32_16X16X128_F8F6F4 : VOP3P_Real_MFMA_F8F6F4_gfx950_mc <0x2d, "v_mfma_f32_16x16x128_f8f6f4">; +defm V_MFMA_SCALE_F32_16X16X128_F8F6F4 : VOP3PX_Real_ScaledMFMA_F8F6F4_mc <0x2d>; +defm V_MFMA_F32_32X32X64_F8F6F4 : VOP3P_Real_MFMA_F8F6F4_gfx950_mc <0x2e, "v_mfma_f32_32x32x64_f8f6f4">; +defm V_MFMA_SCALE_F32_32X32X64_F8F6F4 : VOP3PX_Real_ScaledMFMA_F8F6F4_mc <0x2e>; defm V_MFMA_I32_32X32X16I8 : VOP3P_Real_MFMA_gfx940 <0x56, "v_mfma_i32_32x32x16_i8">; defm V_MFMA_I32_16X16X32I8 : VOP3P_Real_MFMA_gfx940 <0x57, "v_mfma_i32_16x16x32_i8">; diff --git a/llvm/lib/Target/AMDGPU/VOPInstructions.td b/llvm/lib/Target/AMDGPU/VOPInstructions.td index a6e6adac04e5a9..aa9758219db914 100644 --- a/llvm/lib/Target/AMDGPU/VOPInstructions.td +++ b/llvm/lib/Target/AMDGPU/VOPInstructions.td @@ -448,7 +448,7 @@ class VOP3Pe op, VOPProfile P> : Enc64 { let Inst{63} = !if(P.HasSrc2Mods, src2_modifiers{0}, 0); // neg (lo) } -class VOP3Pe_MAI op, VOPProfile P, bit acc_cd = 0> : Enc64 { +class VOP3Pe_MAI_Base { bits<8> vdst; bits<10> src0; bits<10> src1; @@ -456,11 +456,13 @@ class VOP3Pe_MAI op, VOPProfile P, bit acc_cd = 0> : Enc64 { bits<3> blgp; bits<3> cbsz; bits<4> abid; +} +class VOP3Pe_MAI op, VOPProfile P, bit acc_cd = 0> : Enc64, VOP3Pe_MAI_Base { let Inst{7-0} = vdst; let Inst{10-8} = !if(P.HasSrc1, cbsz, 0); - let Inst{14-11} = !if(P.HasSrc1, abid, 0); + let Inst{14-11} = !if(P.HasAbid, abid, 0); let Inst{15} = acc_cd; @@ -506,6 +508,59 @@ class VOP3Pe_SMFMAC op> : Enc64 { let Inst{63-61} = blgp; } +class VOP3PXe op, VOPProfile MFMAPfl, bit acc_cd = 0> : Enc128, VOP3Pe_MAI_Base { + bits<9> scale_src0; + bits<9> scale_src1; + + bits<2> scale_src0_opsel; + bits<2> scale_src1_opsel; + + // Inst{7-0} = unused + // Inst{10-8} = neg_hi; + // Inst{13-11} = op_sel + let Inst{11} = scale_src0_opsel{0}; + let Inst{12} = scale_src1_opsel{0}; + // Inst{13} = unused op_sel + // Inst{14} = unused op_sel_hi2 + + let Inst{31-16} = 0b1101001110101100; + let Inst{40-32} = scale_src0; + let Inst{49-41} = scale_src1; + // Inst{50-58} = unused + // Inst{60-59} = op_sel_hi; + let Inst{59} = scale_src0_opsel{1}; + let Inst{60} = scale_src1_opsel{1}; + // Inst{63-61} = neg; + + // The high half of the encoding is the unscaled mfma op. + // + // FIXME: Defining the encoding in terms of the base instruction + // seems to not work, results in all 0 encoding, so replicate all + // the fields from VOP3Pe_MAI, shifted up by 64 + // + // defvar Hi = VOP3Pe_MAI; + // let Inst{127-64} = Hi.Inst; + + let Inst{71-64} = vdst; + let Inst{74-72} = !if(MFMAPfl.HasSrc1, cbsz, 0); + + // abid must be 1 to use a scale. + let Inst{78-75} = 0b0001; // abid + + let Inst{79} = acc_cd; + + let Inst{86-80} = op; + let Inst{95-87} = 0x1a7; //encoding + let Inst{104-96} = !if(MFMAPfl.HasSrc0, src0{8-0}, 0); + let Inst{113-105} = !if(MFMAPfl.HasSrc1, src1{8-0}, 0); + let Inst{122-114} = !if(MFMAPfl.HasSrc2, src2, 0); + + let Inst{123} = !if(MFMAPfl.HasSrc0, src0{9}, 0); // acc(0) + let Inst{124} = !if(MFMAPfl.HasSrc1, src1{9}, 0); // acc(1) + + let Inst{127-125} = !if(MFMAPfl.HasSrc1, blgp, 0); +} + class VOP3Pe_gfx10 op, VOPProfile P> : VOP3Pe { let Inst{31-23} = 0x198; //encoding } @@ -1343,15 +1398,39 @@ class getVOP3ClampPat { } class getVOP3MAIPat { - list ret = !if(!eq(P.Src0VT, P.Src1VT), + list mfma_with_abid = [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, + timm:$cbsz, timm:$abid, timm:$blgp))]; + list mfma_no_abid = [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, + timm:$cbsz, timm:$blgp))]; + + list ret = !if(!not(P.IsSMFMAC), // mfma - [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, - timm:$cbsz, timm:$abid, timm:$blgp))], + !if(P.HasAbid, mfma_with_abid, mfma_no_abid), + // smfmac [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, i32:$idx, timm:$cbsz, timm:$abid))]); } +class getVOP3MAIScaledPat { + list ret = !if(!not(P.IsSMFMAC), + // mfma + [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, + timm:$cbsz, timm:$blgp, + MFMALdScaleModifierOp:$scale_src0_opsel, + i32:$scale_src0, + MFMALdScaleModifierOp:$scale_src1_opsel, + i32:$scale_src1 + ))], + // smfmac + [(set P.DstVT:$vdst, (node P.Src0VT:$src0, P.Src1VT:$src1, P.Src2VT:$src2, i32:$idx, + timm:$cbsz, timm:$abid, + MFMALdScaleModifierOp:$scale_src0_opsel, + i32:$scale_src0, + MFMALdScaleModifierOp:$scale_src1_opsel, + i32:$scale_src1))]); +} + class VOP3Features { bit HasClamp = Clamp; bit HasOpSel = OpSel; @@ -1415,7 +1494,7 @@ multiclass VOP3Inst_Pseudo_Wrapper patter def _e64 : VOP3_Pseudo; } -class VOP3InstBase : +class VOP3InstBase : VOP3_Pseudo.ret, !if (P.IsMAI, - getVOP3MAIPat.ret, + !if(MAIScaled, getVOP3MAIScaledPat.ret, getVOP3MAIPat.ret), getVOP3Pat.ret))))), 0, P.HasOpSel> { diff --git a/llvm/test/Analysis/UniformityAnalysis/AMDGPU/intrinsics.ll b/llvm/test/Analysis/UniformityAnalysis/AMDGPU/intrinsics.ll index 00a3aaf77f9003..9769d8b1d910d2 100644 --- a/llvm/test/Analysis/UniformityAnalysis/AMDGPU/intrinsics.ll +++ b/llvm/test/Analysis/UniformityAnalysis/AMDGPU/intrinsics.ll @@ -285,6 +285,26 @@ define amdgpu_kernel void @mfma_f32_32x32x16_bf16(<8 x bfloat> %arg0, <8 x bfloa ret void } +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32>, <8 x i32>, <4 x float>, i32 immarg, i32 immarg, + i32 immarg, i32, i32 immarg, i32) + +; CHECK: DIVERGENT: %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 immarg 0, i32 immarg 0, i32 immarg 0, i32 %arg3, i32 immarg 0, i32 %arg4) +define amdgpu_kernel void @mfma_scale_f32_16x16x128_f8f6f4(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %arg3, i32 %arg4, ptr addrspace(1) %out) { + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 immarg 0, i32 immarg 0, i32 immarg 0, i32 %arg3, i32 immarg 0, i32 %arg4) + store <4 x float> %result, ptr addrspace(1) %out + ret void +} + +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4(<8 x i32>, <8 x i32>, <16 x float>, i32 immarg, i32 immarg, + i32 immarg, i32, i32 immarg, i32) + +; CHECK: DIVERGENT: %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 immarg 0, i32 immarg 0, i32 immarg 0, i32 %arg3, i32 immarg 0, i32 %arg4) +define amdgpu_kernel void @mfma_f32_scale_32x32x64_f8f6f4(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %arg3, i32 %arg4, ptr addrspace(1) %out) { + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 immarg 0, i32 immarg 0, i32 immarg 0, i32 %arg3, i32 immarg 0, i32 %arg4) + 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 diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.ll new file mode 100644 index 00000000000000..9658f8381bff21 --- /dev/null +++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.ll @@ -0,0 +1,2296 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx950 -global-isel=0 < %s | FileCheck -check-prefixes=GCN,SDAG %s +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx950 -global-isel=1 < %s | FileCheck -check-prefixes=GCN,GISEL %s + +; 0 = fp8 +; 1 = bf8 +; 2 = fp6 +; 3 = bf6 +; 4 = fp4 + +; -------------------------------------------------------------------- +; Different format signatures +; -------------------------------------------------------------------- + +; fp8 x fp8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp0(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_1_1__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_1_1__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 1, i32 %scale0, i32 1, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_2_2__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_2_2__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 2, i32 %scale0, i32 2, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_3_3__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_3_3__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 3, i32 %scale0, i32 3, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_3__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_3__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 3, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_3_0__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_3_0__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 3, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_2_3__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_2_3__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 2, i32 %scale0, i32 3, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_3_2__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_3_2__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 3, i32 %scale0, i32 2, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp0__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp8 x bf8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp1__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp8 x fp6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp2(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp2__constant_scale_0_0(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp8 x bf6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp3(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp3__constant_scale_0_0(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp8 x fp4 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp4(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp4__constant_scale_0_0(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz0__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf8 x fp8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp0(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] cbsz:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp0__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf8 x bf8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] cbsz:1 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp1__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf8 x fp6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp2(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] cbsz:1 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp2__constant_scale_0(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp2__constant_scale_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf8 x bf6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp3(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] cbsz:1 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp3__constant_scale_0_0(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf8 x fp4 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp4(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] cbsz:1 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp4__constant_scale_0_0(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz1__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 1, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp6 x fp8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp0(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] cbsz:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp0__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp6 x bf8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp1(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] cbsz:2 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp1__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp6 x fp6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp2(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp2__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp6 x bf6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp3(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] cbsz:2 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp3__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + + +; bf6 x fp8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp0(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] cbsz:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp0__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf6 x bf8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp1(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] cbsz:3 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp1__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf6 x fp6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp2(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] cbsz:3 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp2__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf6 x fp4 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp4(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:9], a[0:3], v14, v15 op_sel_hi:[0,0,0] cbsz:3 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp4__constant_scale_0_0(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:9], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; bf6 x bf6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp3(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] cbsz:3 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp3__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz3__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 3, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp6 x fp4 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp4(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:9], a[0:3], v14, v15 op_sel_hi:[0,0,0] cbsz:2 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp4__constant_scale_0_0(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz2__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:9], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp4 x fp8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp0(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] cbsz:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp0__constant_scale_0_0(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp4 x bf8 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp1(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:11], a[0:3], v16, v17 op_sel_hi:[0,0,0] cbsz:4 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp1__constant_scale_0_0(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:11], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp4 x fp6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp2(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:9], a[0:3], v14, v15 op_sel_hi:[0,0,0] cbsz:4 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp2__constant_scale_0_0(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:9], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp4 x bf6 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp3(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:9], a[0:3], v14, v15 op_sel_hi:[0,0,0] cbsz:4 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp3__constant_scale_0_0(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:9], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; fp4 x fp4 +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp4(<4 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v8 +; GCN-NEXT: v_accvgpr_write_b32 a1, v9 +; GCN-NEXT: v_accvgpr_write_b32 a2, v10 +; GCN-NEXT: v_accvgpr_write_b32 a3, v11 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:7], a[0:3], v12, v13 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v4i32(<4 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp4__constant_scale_0_0(<4 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__cbsz4__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v8 +; GCN-NEXT: v_accvgpr_write_b32 a1, v9 +; GCN-NEXT: v_accvgpr_write_b32 a2, v10 +; GCN-NEXT: v_accvgpr_write_b32 a3, v11 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:3], v[4:7], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v4i32(<4 x i32> %arg0, <4 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; -------------------------------------------------------------------- +; Different input parameter classes +; -------------------------------------------------------------------- + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__sgpr_scaleA__sgpr_scaleB(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 inreg %scale0, i32 inreg %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__sgpr_scaleA__sgpr_scaleB: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_mov_b32_e32 v16, s1 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], s0, v16 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__sgpr_scaleA__vgpr_scaleB(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 inreg %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__sgpr_scaleA__vgpr_scaleB: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], s0, v20 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__vgpr_scaleA__sgpr_scaleB(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 inreg %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__vgpr_scaleA__sgpr_scaleB: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, s0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgprs(<8 x i32> inreg %arg0, <8 x i32> inreg %arg1, <4 x float> inreg %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgprs: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v20, s28 +; SDAG-NEXT: v_mov_b32_e32 v23, v1 +; SDAG-NEXT: v_mov_b32_e32 v22, v0 +; SDAG-NEXT: v_mov_b32_e32 v21, s29 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v20 +; SDAG-NEXT: v_mov_b32_e32 v4, s20 +; SDAG-NEXT: v_mov_b32_e32 v5, s21 +; SDAG-NEXT: v_mov_b32_e32 v6, s22 +; SDAG-NEXT: v_mov_b32_e32 v7, s23 +; SDAG-NEXT: v_mov_b32_e32 v8, s24 +; SDAG-NEXT: v_mov_b32_e32 v9, s25 +; SDAG-NEXT: v_mov_b32_e32 v10, s26 +; SDAG-NEXT: v_mov_b32_e32 v11, s27 +; SDAG-NEXT: v_mov_b32_e32 v12, s0 +; SDAG-NEXT: v_mov_b32_e32 v13, s1 +; SDAG-NEXT: v_mov_b32_e32 v14, s2 +; SDAG-NEXT: v_mov_b32_e32 v15, s3 +; SDAG-NEXT: v_mov_b32_e32 v16, s16 +; SDAG-NEXT: v_mov_b32_e32 v17, s17 +; SDAG-NEXT: v_mov_b32_e32 v18, s18 +; SDAG-NEXT: v_mov_b32_e32 v19, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[12:19], v[4:11], a[0:3], v2, v3 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgprs: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: v_mov_b32_e32 v20, s28 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b32_e32 v22, v0 +; GISEL-NEXT: v_mov_b32_e32 v23, v1 +; GISEL-NEXT: v_mov_b32_e32 v21, s29 +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[20:21] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v20 +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[18:19] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[22:23] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[24:25] +; GISEL-NEXT: v_mov_b64_e32 v[18:19], s[26:27] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v23 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[4:11], v[12:19], a[0:3], v2, v3 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_vgpr__sgpr_vgpr(<8 x i32> inreg %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 inreg %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_vgpr__sgpr_vgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v8 +; SDAG-NEXT: v_mov_b32_e32 v14, s0 +; SDAG-NEXT: v_mov_b32_e32 v15, s1 +; SDAG-NEXT: v_mov_b32_e32 v16, s2 +; SDAG-NEXT: v_mov_b32_e32 v17, s3 +; SDAG-NEXT: v_mov_b32_e32 v18, s16 +; SDAG-NEXT: v_mov_b32_e32 v19, s17 +; SDAG-NEXT: v_mov_b32_e32 v20, s18 +; SDAG-NEXT: v_mov_b32_e32 v21, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v11 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[14:21], v[0:7], a[0:3], s20, v12 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_vgpr__sgpr_vgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b64_e32 v[20:21], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v8 +; GISEL-NEXT: v_mov_b64_e32 v[18:19], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v11 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[14:21], v[0:7], a[0:3], s20, v12 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_vgpr__vgpr_sgpr(<8 x i32> inreg %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 inreg %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_vgpr__vgpr_sgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v8 +; SDAG-NEXT: v_mov_b32_e32 v14, s0 +; SDAG-NEXT: v_mov_b32_e32 v15, s1 +; SDAG-NEXT: v_mov_b32_e32 v16, s2 +; SDAG-NEXT: v_mov_b32_e32 v17, s3 +; SDAG-NEXT: v_mov_b32_e32 v18, s16 +; SDAG-NEXT: v_mov_b32_e32 v19, s17 +; SDAG-NEXT: v_mov_b32_e32 v20, s18 +; SDAG-NEXT: v_mov_b32_e32 v21, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v11 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[14:21], v[0:7], a[0:3], v12, s20 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_vgpr__vgpr_sgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b64_e32 v[20:21], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v8 +; GISEL-NEXT: v_mov_b64_e32 v[18:19], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v11 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[14:21], v[0:7], a[0:3], v12, s20 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0_vgpr_sgpr_vgpr__vgpr_sgpr(<8 x i32> %arg0, <8 x i32> inreg %arg1, <4 x float> %arg2, i32 %scale0, i32 inreg %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_vgpr_sgpr_vgpr__vgpr_sgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v8 +; SDAG-NEXT: v_mov_b32_e32 v14, s0 +; SDAG-NEXT: v_mov_b32_e32 v15, s1 +; SDAG-NEXT: v_mov_b32_e32 v16, s2 +; SDAG-NEXT: v_mov_b32_e32 v17, s3 +; SDAG-NEXT: v_mov_b32_e32 v18, s16 +; SDAG-NEXT: v_mov_b32_e32 v19, s17 +; SDAG-NEXT: v_mov_b32_e32 v20, s18 +; SDAG-NEXT: v_mov_b32_e32 v21, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v11 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[14:21], a[0:3], v12, s20 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_vgpr_sgpr_vgpr__vgpr_sgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b64_e32 v[20:21], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v8 +; GISEL-NEXT: v_mov_b64_e32 v[18:19], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v11 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[14:21], a[0:3], v12, s20 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0_vgpr_vgpr_sgpr__vgpr_sgpr(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> inreg %arg2, i32 %scale0, i32 inreg %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_vgpr_vgpr_sgpr__vgpr_sgpr: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, s0 +; GCN-NEXT: v_accvgpr_write_b32 a1, s1 +; GCN-NEXT: v_accvgpr_write_b32 a2, s2 +; GCN-NEXT: v_accvgpr_write_b32 a3, s3 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v16, s16 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_sgpr__vgpr_sgpr(<8 x i32> inreg %arg0, <8 x i32> %arg1, <4 x float> inreg %arg2, i32 %scale0, i32 inreg %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_sgpr__vgpr_sgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v10, s0 +; SDAG-NEXT: v_mov_b32_e32 v11, s1 +; SDAG-NEXT: v_mov_b32_e32 v12, s2 +; SDAG-NEXT: v_mov_b32_e32 v13, s3 +; SDAG-NEXT: v_mov_b32_e32 v14, s16 +; SDAG-NEXT: v_mov_b32_e32 v15, s17 +; SDAG-NEXT: v_mov_b32_e32 v16, s18 +; SDAG-NEXT: v_mov_b32_e32 v17, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a0, s20 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s21 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s22 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[10:17], v[0:7], a[0:3], v8, s24 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0_sgpr_vgpr_sgpr__vgpr_sgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s20 +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s21 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s22 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s23 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[10:17], v[0:7], a[0:3], v8, s24 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_inlineimm__scaleB_inlineimm(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_inlineimm__scaleB_inlineimm: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 33, -2 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 2, i32 33, i32 2, i32 -2) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_kimm__scaleB_inlineimm(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_kimm__scaleB_inlineimm: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: s_movk_i32 s0, 0x41 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], s0, -2 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_kimm__scaleB_inlineimm: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_mov_b32_e32 v16, 0x41 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v16, -2 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 2, i32 65, i32 2, i32 -2) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_kimm__scaleB_kimm(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_kimm__scaleB_kimm: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: s_movk_i32 s0, 0x41 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_mov_b32_e32 v16, 0x4d +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], s0, v16 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4_0_0__scaleA_kimm__scaleB_kimm: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_mov_b32_e32 v16, 0x41 +; GISEL-NEXT: v_mov_b32_e32 v17, 0x4d +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v16, v17 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 2, i32 65, i32 2, i32 77) + ret <4 x float> %result +} + +define amdgpu_kernel void @test_mfma_scale_f32_16x16x128_f8f6f4__vgprcd(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1, ptr addrspace(1) %ptr) #0 { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4__vgprcd: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; SDAG-NEXT: v_mov_b32_e32 v16, 0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: v_mov_b32_e32 v4, s12 +; SDAG-NEXT: v_mov_b32_e32 v5, s13 +; SDAG-NEXT: v_mov_b32_e32 v6, s14 +; SDAG-NEXT: v_mov_b32_e32 v7, s15 +; SDAG-NEXT: s_load_dwordx8 s[8:15], s[4:5], 0x40 +; SDAG-NEXT: v_mov_b32_e32 v8, s16 +; SDAG-NEXT: v_mov_b32_e32 v9, s17 +; SDAG-NEXT: v_mov_b32_e32 v10, s18 +; SDAG-NEXT: v_mov_b32_e32 v11, s19 +; SDAG-NEXT: v_mov_b32_e32 v12, s20 +; SDAG-NEXT: v_mov_b32_e32 v13, s21 +; SDAG-NEXT: v_mov_b32_e32 v14, s22 +; SDAG-NEXT: v_mov_b32_e32 v15, s23 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s8 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s11 +; SDAG-NEXT: v_mov_b32_e32 v17, s13 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], s12, v17 op_sel_hi:[0,0,0] blgp:2 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: global_store_dwordx4 v16, a[0:3], s[14:15] +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4__vgprcd: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx8 s[24:31], s[4:5], 0x40 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[8:9] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[16:17] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s24 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[18:19] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[22:23] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s25 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s26 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s27 +; GISEL-NEXT: v_mov_b32_e32 v16, s29 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], s28, v16 op_sel_hi:[0,0,0] blgp:2 +; GISEL-NEXT: v_mov_b32_e32 v0, 0 +; GISEL-NEXT: s_nop 2 +; GISEL-NEXT: global_store_dwordx4 v0, a[0:3], s[30:31] +; GISEL-NEXT: s_endpgm + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 2, i32 3, i32 %scale0, i32 1, i32 %scale1) + store <4 x float> %result, ptr addrspace(1) %ptr, align 16 + ret void +} + +define amdgpu_kernel void @test_mfma_scale_f32_16x16x128_f8f6f4__vgprcd___scaleA_kimm__scaleB__inlineimm(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, ptr addrspace(1) %ptr) #0 { +; SDAG-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4__vgprcd___scaleA_kimm__scaleB__inlineimm: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x40 +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; SDAG-NEXT: s_movk_i32 s6, 0x41 +; SDAG-NEXT: s_load_dwordx2 s[4:5], s[4:5], 0x50 +; SDAG-NEXT: v_mov_b32_e32 v16, 0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s0 +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: v_mov_b32_e32 v4, s12 +; SDAG-NEXT: v_mov_b32_e32 v5, s13 +; SDAG-NEXT: v_mov_b32_e32 v6, s14 +; SDAG-NEXT: v_mov_b32_e32 v7, s15 +; SDAG-NEXT: v_mov_b32_e32 v8, s16 +; SDAG-NEXT: v_mov_b32_e32 v9, s17 +; SDAG-NEXT: v_mov_b32_e32 v10, s18 +; SDAG-NEXT: v_mov_b32_e32 v11, s19 +; SDAG-NEXT: v_mov_b32_e32 v12, s20 +; SDAG-NEXT: v_mov_b32_e32 v13, s21 +; SDAG-NEXT: v_mov_b32_e32 v14, s22 +; SDAG-NEXT: v_mov_b32_e32 v15, s23 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s1 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s2 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s3 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], s6, -2 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: global_store_dwordx4 v16, a[0:3], s[4:5] +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4__vgprcd___scaleA_kimm__scaleB__inlineimm: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x40 +; GISEL-NEXT: s_load_dwordx2 s[6:7], s[4:5], 0x50 +; GISEL-NEXT: v_mov_b32_e32 v16, 0x41 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[8:9] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[16:17] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s0 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[18:19] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[22:23] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s1 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s2 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s3 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v16, -2 op_sel_hi:[0,0,0] +; GISEL-NEXT: v_mov_b32_e32 v0, 0 +; GISEL-NEXT: s_nop 2 +; GISEL-NEXT: global_store_dwordx4 v0, a[0:3], s[6:7] +; GISEL-NEXT: s_endpgm + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 3, i32 65, i32 1, i32 -2) + store <4 x float> %result, ptr addrspace(1) %ptr, align 16 + ret void +} + +; This should be optimized to avoid the scale +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_0_0_a(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_0_0_a: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +; This should be optimized to avoid the scale, with non-0 op_sel arguments. +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_0_0_b(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_0_0_b: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 3, i32 0, i32 1, i32 0) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_0_1(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_0_1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 1 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 0, i32 0, i32 1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_1_0_a(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___constant_scale_1_0_a: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 1, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 0, i32 0, i32 0, i32 1, i32 0, i32 0) + ret <4 x float> %result +} + +; -------------------------------------------------------------------- +; Incorrect signature for format cases (IR vector too large) +; -------------------------------------------------------------------- + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp6: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp8: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] cbsz:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v8i32_fp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp8: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] cbsz:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v6i32_fp4(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp8__v6i32_fp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <4 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v6i32_fp4__v8i32_fp8(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v6i32_fp4__v8i32_fp8: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:5], v[6:13], a[0:3], v18, v19 op_sel_hi:[0,0,0] cbsz:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <4 x float> %result +} + +define <4 x float> @test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_16x16x128_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <4 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <4 x float> %result +} + +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v8i32(<8 x i32>, <8 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v6i32(<6 x i32>, <6 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v4i32(<4 x i32>, <4 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v6i32(<4 x i32>, <6 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v4i32.v8i32(<4 x i32>, <8 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v4i32(<6 x i32>, <4 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v6i32.v8i32(<6 x i32>, <8 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v4i32(<8 x i32>, <4 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 +declare <4 x float> @llvm.amdgcn.mfma.scale.f32.16x16x128.f8f6f4.v8i32.v6i32(<8 x i32>, <6 x i32>, <4 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #1 + +attributes #0 = { "amdgpu-flat-work-group-size"="512,512" } +attributes #1 = { convergent nocallback nofree nosync nounwind willreturn memory(none) } diff --git a/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.ll b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.ll new file mode 100644 index 00000000000000..67d887d87dd973 --- /dev/null +++ b/llvm/test/CodeGen/AMDGPU/llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.ll @@ -0,0 +1,6004 @@ +; NOTE: Assertions have been autogenerated by utils/update_llc_test_checks.py UTC_ARGS: --version 5 +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx950 -global-isel=0 < %s | FileCheck -check-prefixes=GCN,SDAG %s +; RUN: llc -mtriple=amdgcn-amd-amdhsa -mcpu=gfx950 -global-isel=1 < %s | FileCheck -check-prefixes=GCN,GISEL %s + +; 0 = fp8 +; 1 = bf8 +; 2 = fp6 +; 3 = bf6 +; 4 = fp4 + +; -------------------------------------------------------------------- +; Different format signatures +; -------------------------------------------------------------------- + +; fp8 x fp8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp0(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp0: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp0: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_1_1__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_1_1__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_1_1__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 1, i32 %scale0, i32 1, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_2_2__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_2_2__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_2_2__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 2, i32 %scale0, i32 2, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_3_3__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_3_3__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_3_3__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 3, i32 %scale0, i32 3, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_3__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_3__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_3__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 3, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_3_0__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_3_0__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_3_0__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 3, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_2_3__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_2_3__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_2_3__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 2, i32 %scale0, i32 3, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_3_2__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_3_2__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_3_2__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 3, i32 %scale0, i32 2, i32 %scale1) + ret <16 x float> %result +} + +; This should be optimized to avoid the scale +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp0__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp8 x bf8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] blgp:1 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] blgp:1 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp1__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp8 x fp6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp2(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp2__constant_scale_0_0(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp8 x bf6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp3(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp3__constant_scale_0_0(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp8 x fp4 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp4(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp4__constant_scale_0_0(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz0__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf8 x fp8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp0(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp0: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] cbsz:1 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp0: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] cbsz:1 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp0__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf8 x bf8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp1: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] cbsz:1 blgp:1 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp1: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] cbsz:1 blgp:1 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp1__constant_scale_0_0(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf8 x fp6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp2(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] cbsz:1 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp2__constant_scale_0(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp2__constant_scale_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf8 x bf6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp3(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] cbsz:1 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp3__constant_scale_0_0(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf8 x fp4 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp4(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] cbsz:1 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp4__constant_scale_0_0(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz1__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:1 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v4i32(<8 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 1, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp6 x fp8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp0(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] cbsz:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp0__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp6 x bf8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp1(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] cbsz:2 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp1__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp6 x fp6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp2(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp2__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp6 x bf6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp3(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] cbsz:2 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp3__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + + +; bf6 x fp8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp0(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] cbsz:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp0__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf6 x bf8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp1(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] cbsz:3 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp1__constant_scale_0_0(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf6 x fp6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp2(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] cbsz:3 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp2__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf6 x fp4 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp4(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:9], a[0:15], v26, v27 op_sel_hi:[0,0,0] cbsz:3 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp4__constant_scale_0_0(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:9], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; bf6 x bf6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp3(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] cbsz:3 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp3__constant_scale_0_0(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz3__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:3 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 3, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp6 x fp4 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp4(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:9], a[0:15], v26, v27 op_sel_hi:[0,0,0] cbsz:2 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp4__constant_scale_0_0(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz2__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:9], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v4i32(<6 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp4 x fp8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp0(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] cbsz:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp0__constant_scale_0_0(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp0__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp4 x bf8 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp1(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:11], a[0:15], v28, v29 op_sel_hi:[0,0,0] cbsz:4 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 1, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp1__constant_scale_0_0(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp1__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v12 +; GCN-NEXT: v_accvgpr_write_b32 a1, v13 +; GCN-NEXT: v_accvgpr_write_b32 a2, v14 +; GCN-NEXT: v_accvgpr_write_b32 a3, v15 +; GCN-NEXT: v_accvgpr_write_b32 a4, v16 +; GCN-NEXT: v_accvgpr_write_b32 a5, v17 +; GCN-NEXT: v_accvgpr_write_b32 a6, v18 +; GCN-NEXT: v_accvgpr_write_b32 a7, v19 +; GCN-NEXT: v_accvgpr_write_b32 a8, v20 +; GCN-NEXT: v_accvgpr_write_b32 a9, v21 +; GCN-NEXT: v_accvgpr_write_b32 a10, v22 +; GCN-NEXT: v_accvgpr_write_b32 a11, v23 +; GCN-NEXT: v_accvgpr_write_b32 a12, v24 +; GCN-NEXT: v_accvgpr_write_b32 a13, v25 +; GCN-NEXT: v_accvgpr_write_b32 a14, v26 +; GCN-NEXT: v_accvgpr_write_b32 a15, v27 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:11], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:1 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v8i32(<4 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 1, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp4 x fp6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp2(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp2: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:9], a[0:15], v26, v27 op_sel_hi:[0,0,0] cbsz:4 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp2__constant_scale_0_0(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp2__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:9], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp4 x bf6 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp3(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp3: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:9], a[0:15], v26, v27 op_sel_hi:[0,0,0] cbsz:4 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 3, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp3__constant_scale_0_0(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp3__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v10 +; GCN-NEXT: v_accvgpr_write_b32 a1, v11 +; GCN-NEXT: v_accvgpr_write_b32 a2, v12 +; GCN-NEXT: v_accvgpr_write_b32 a3, v13 +; GCN-NEXT: v_accvgpr_write_b32 a4, v14 +; GCN-NEXT: v_accvgpr_write_b32 a5, v15 +; GCN-NEXT: v_accvgpr_write_b32 a6, v16 +; GCN-NEXT: v_accvgpr_write_b32 a7, v17 +; GCN-NEXT: v_accvgpr_write_b32 a8, v18 +; GCN-NEXT: v_accvgpr_write_b32 a9, v19 +; GCN-NEXT: v_accvgpr_write_b32 a10, v20 +; GCN-NEXT: v_accvgpr_write_b32 a11, v21 +; GCN-NEXT: v_accvgpr_write_b32 a12, v22 +; GCN-NEXT: v_accvgpr_write_b32 a13, v23 +; GCN-NEXT: v_accvgpr_write_b32 a14, v24 +; GCN-NEXT: v_accvgpr_write_b32 a15, v25 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:9], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:3 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v6i32(<4 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 3, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; fp4 x fp4 +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp4(<4 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v8 +; GCN-NEXT: v_accvgpr_write_b32 a1, v9 +; GCN-NEXT: v_accvgpr_write_b32 a2, v10 +; GCN-NEXT: v_accvgpr_write_b32 a3, v11 +; GCN-NEXT: v_accvgpr_write_b32 a4, v12 +; GCN-NEXT: v_accvgpr_write_b32 a5, v13 +; GCN-NEXT: v_accvgpr_write_b32 a6, v14 +; GCN-NEXT: v_accvgpr_write_b32 a7, v15 +; GCN-NEXT: v_accvgpr_write_b32 a8, v16 +; GCN-NEXT: v_accvgpr_write_b32 a9, v17 +; GCN-NEXT: v_accvgpr_write_b32 a10, v18 +; GCN-NEXT: v_accvgpr_write_b32 a11, v19 +; GCN-NEXT: v_accvgpr_write_b32 a12, v20 +; GCN-NEXT: v_accvgpr_write_b32 a13, v21 +; GCN-NEXT: v_accvgpr_write_b32 a14, v22 +; GCN-NEXT: v_accvgpr_write_b32 a15, v23 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:7], a[0:15], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v4i32(<4 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp4__constant_scale_0_0(<4 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__cbsz4__blgp4__constant_scale_0_0: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: v_accvgpr_write_b32 a0, v8 +; GCN-NEXT: v_accvgpr_write_b32 a1, v9 +; GCN-NEXT: v_accvgpr_write_b32 a2, v10 +; GCN-NEXT: v_accvgpr_write_b32 a3, v11 +; GCN-NEXT: v_accvgpr_write_b32 a4, v12 +; GCN-NEXT: v_accvgpr_write_b32 a5, v13 +; GCN-NEXT: v_accvgpr_write_b32 a6, v14 +; GCN-NEXT: v_accvgpr_write_b32 a7, v15 +; GCN-NEXT: v_accvgpr_write_b32 a8, v16 +; GCN-NEXT: v_accvgpr_write_b32 a9, v17 +; GCN-NEXT: v_accvgpr_write_b32 a10, v18 +; GCN-NEXT: v_accvgpr_write_b32 a11, v19 +; GCN-NEXT: v_accvgpr_write_b32 a12, v20 +; GCN-NEXT: v_accvgpr_write_b32 a13, v21 +; GCN-NEXT: v_accvgpr_write_b32 a14, v22 +; GCN-NEXT: v_accvgpr_write_b32 a15, v23 +; GCN-NEXT: s_nop 1 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:3], v[4:7], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v4i32(<4 x i32> %arg0, <4 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +; -------------------------------------------------------------------- +; Different input parameter classes +; -------------------------------------------------------------------- + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__sgpr_scaleA__sgpr_scaleB(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 inreg %scale0, i32 inreg %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__sgpr_scaleA__sgpr_scaleB: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: v_mov_b32_e32 v16, s1 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, v16 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__sgpr_scaleA__vgpr_scaleB(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 inreg %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__sgpr_scaleA__vgpr_scaleB: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, v31 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__vgpr_scaleA__sgpr_scaleB(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 inreg %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__vgpr_scaleA__sgpr_scaleB: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, s0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgprs(<8 x i32> inreg %arg0, <8 x i32> inreg %arg1, <16 x float> inreg %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgprs: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v16, s28 +; SDAG-NEXT: v_mov_b32_e32 v31, v13 +; SDAG-NEXT: v_mov_b32_e32 v30, v12 +; SDAG-NEXT: v_mov_b32_e32 v29, v11 +; SDAG-NEXT: v_mov_b32_e32 v28, v10 +; SDAG-NEXT: v_mov_b32_e32 v27, v9 +; SDAG-NEXT: v_mov_b32_e32 v26, v8 +; SDAG-NEXT: v_mov_b32_e32 v25, v7 +; SDAG-NEXT: v_mov_b32_e32 v24, v6 +; SDAG-NEXT: v_mov_b32_e32 v23, v5 +; SDAG-NEXT: v_mov_b32_e32 v22, v4 +; SDAG-NEXT: v_mov_b32_e32 v21, v3 +; SDAG-NEXT: v_mov_b32_e32 v20, v2 +; SDAG-NEXT: v_mov_b32_e32 v19, v1 +; SDAG-NEXT: v_mov_b32_e32 v18, v0 +; SDAG-NEXT: v_mov_b32_e32 v17, s29 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_mov_b32_e32 v0, s20 +; SDAG-NEXT: v_mov_b32_e32 v1, s21 +; SDAG-NEXT: v_mov_b32_e32 v2, s22 +; SDAG-NEXT: v_mov_b32_e32 v3, s23 +; SDAG-NEXT: v_mov_b32_e32 v4, s24 +; SDAG-NEXT: v_mov_b32_e32 v5, s25 +; SDAG-NEXT: v_mov_b32_e32 v6, s26 +; SDAG-NEXT: v_mov_b32_e32 v7, s27 +; SDAG-NEXT: v_mov_b32_e32 v32, s0 +; SDAG-NEXT: v_mov_b32_e32 v33, s1 +; SDAG-NEXT: v_mov_b32_e32 v34, s2 +; SDAG-NEXT: v_mov_b32_e32 v35, s3 +; SDAG-NEXT: v_mov_b32_e32 v36, s16 +; SDAG-NEXT: v_mov_b32_e32 v37, s17 +; SDAG-NEXT: v_mov_b32_e32 v38, s18 +; SDAG-NEXT: v_mov_b32_e32 v39, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: v_accvgpr_write_b32 a15, v31 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[32:39], v[0:7], a[0:15], v14, v15 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgprs: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: v_mov_b32_e32 v16, s28 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b32_e32 v18, v0 +; GISEL-NEXT: v_mov_b32_e32 v19, v1 +; GISEL-NEXT: v_mov_b32_e32 v20, v2 +; GISEL-NEXT: v_mov_b32_e32 v21, v3 +; GISEL-NEXT: v_mov_b32_e32 v22, v4 +; GISEL-NEXT: v_mov_b32_e32 v23, v5 +; GISEL-NEXT: v_mov_b32_e32 v24, v6 +; GISEL-NEXT: v_mov_b32_e32 v25, v7 +; GISEL-NEXT: v_mov_b32_e32 v26, v8 +; GISEL-NEXT: v_mov_b32_e32 v27, v9 +; GISEL-NEXT: v_mov_b32_e32 v28, v10 +; GISEL-NEXT: v_mov_b32_e32 v29, v11 +; GISEL-NEXT: v_mov_b32_e32 v30, v12 +; GISEL-NEXT: v_mov_b32_e32 v31, v13 +; GISEL-NEXT: v_mov_b32_e32 v17, s29 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[38:39], s[26:27] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[18:19] +; GISEL-NEXT: v_mov_b64_e32 v[36:37], s[24:25] +; GISEL-NEXT: v_mov_b64_e32 v[34:35], s[22:23] +; GISEL-NEXT: v_mov_b64_e32 v[32:33], s[20:21] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: v_accvgpr_write_b32 a15, v31 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[32:39], a[0:15], v14, v15 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_vgpr__sgpr_vgpr(<8 x i32> inreg %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 inreg %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_vgpr__sgpr_vgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v8 +; SDAG-NEXT: v_mov_b32_e32 v26, s0 +; SDAG-NEXT: v_mov_b32_e32 v27, s1 +; SDAG-NEXT: v_mov_b32_e32 v28, s2 +; SDAG-NEXT: v_mov_b32_e32 v29, s3 +; SDAG-NEXT: v_mov_b32_e32 v30, s16 +; SDAG-NEXT: v_mov_b32_e32 v31, s17 +; SDAG-NEXT: v_mov_b32_e32 v32, s18 +; SDAG-NEXT: v_mov_b32_e32 v33, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v11 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v12 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v13 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v14 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v15 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a15, v23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[26:33], v[0:7], a[0:15], s20, v24 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_vgpr__sgpr_vgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b64_e32 v[32:33], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v8 +; GISEL-NEXT: v_mov_b64_e32 v[30:31], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[28:29], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[26:27], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v11 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v12 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v13 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v14 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v15 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a15, v23 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[26:33], v[0:7], a[0:15], s20, v24 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_vgpr__vgpr_sgpr(<8 x i32> inreg %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 inreg %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_vgpr__vgpr_sgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v8 +; SDAG-NEXT: v_mov_b32_e32 v26, s0 +; SDAG-NEXT: v_mov_b32_e32 v27, s1 +; SDAG-NEXT: v_mov_b32_e32 v28, s2 +; SDAG-NEXT: v_mov_b32_e32 v29, s3 +; SDAG-NEXT: v_mov_b32_e32 v30, s16 +; SDAG-NEXT: v_mov_b32_e32 v31, s17 +; SDAG-NEXT: v_mov_b32_e32 v32, s18 +; SDAG-NEXT: v_mov_b32_e32 v33, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v11 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v12 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v13 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v14 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v15 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a15, v23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[26:33], v[0:7], a[0:15], v24, s20 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_vgpr__vgpr_sgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b64_e32 v[32:33], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v8 +; GISEL-NEXT: v_mov_b64_e32 v[30:31], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[28:29], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[26:27], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v11 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v12 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v13 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v14 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v15 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a15, v23 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[26:33], v[0:7], a[0:15], v24, s20 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0_vgpr_sgpr_vgpr__vgpr_sgpr(<8 x i32> %arg0, <8 x i32> inreg %arg1, <16 x float> %arg2, i32 %scale0, i32 inreg %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_vgpr_sgpr_vgpr__vgpr_sgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, v8 +; SDAG-NEXT: v_mov_b32_e32 v26, s0 +; SDAG-NEXT: v_mov_b32_e32 v27, s1 +; SDAG-NEXT: v_mov_b32_e32 v28, s2 +; SDAG-NEXT: v_mov_b32_e32 v29, s3 +; SDAG-NEXT: v_mov_b32_e32 v30, s16 +; SDAG-NEXT: v_mov_b32_e32 v31, s17 +; SDAG-NEXT: v_mov_b32_e32 v32, s18 +; SDAG-NEXT: v_mov_b32_e32 v33, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v11 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v12 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v13 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v14 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v15 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a15, v23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[26:33], a[0:15], v24, s20 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_vgpr_sgpr_vgpr__vgpr_sgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b64_e32 v[32:33], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v8 +; GISEL-NEXT: v_mov_b64_e32 v[30:31], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[28:29], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[26:27], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v11 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v12 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v13 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v14 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v15 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a15, v23 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[26:33], a[0:15], v24, s20 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0_vgpr_vgpr_sgpr__vgpr_sgpr(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> inreg %arg2, i32 %scale0, i32 inreg %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_vgpr_vgpr_sgpr__vgpr_sgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s0 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s1 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s2 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s3 +; SDAG-NEXT: v_accvgpr_write_b32 a4, s16 +; SDAG-NEXT: v_accvgpr_write_b32 a5, s17 +; SDAG-NEXT: v_accvgpr_write_b32 a6, s18 +; SDAG-NEXT: v_accvgpr_write_b32 a7, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a8, s20 +; SDAG-NEXT: v_accvgpr_write_b32 a9, s21 +; SDAG-NEXT: v_accvgpr_write_b32 a10, s22 +; SDAG-NEXT: v_accvgpr_write_b32 a11, s23 +; SDAG-NEXT: v_accvgpr_write_b32 a12, s24 +; SDAG-NEXT: v_accvgpr_write_b32 a13, s25 +; SDAG-NEXT: v_accvgpr_write_b32 a14, s26 +; SDAG-NEXT: v_accvgpr_write_b32 a15, s27 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v16, s28 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_vgpr_vgpr_sgpr__vgpr_sgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_accvgpr_write_b32 a0, s12 +; GISEL-NEXT: v_accvgpr_write_b32 a1, s13 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s14 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s15 +; GISEL-NEXT: v_accvgpr_write_b32 a4, s16 +; GISEL-NEXT: v_accvgpr_write_b32 a5, s17 +; GISEL-NEXT: v_accvgpr_write_b32 a6, s18 +; GISEL-NEXT: v_accvgpr_write_b32 a7, s19 +; GISEL-NEXT: v_accvgpr_write_b32 a8, s20 +; GISEL-NEXT: v_accvgpr_write_b32 a9, s21 +; GISEL-NEXT: v_accvgpr_write_b32 a10, s22 +; GISEL-NEXT: v_accvgpr_write_b32 a11, s23 +; GISEL-NEXT: v_accvgpr_write_b32 a12, s24 +; GISEL-NEXT: v_accvgpr_write_b32 a13, s25 +; GISEL-NEXT: v_accvgpr_write_b32 a14, s26 +; GISEL-NEXT: v_accvgpr_write_b32 a15, s27 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v16, s28 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_sgpr__vgpr_sgpr(<8 x i32> inreg %arg0, <8 x i32> %arg1, <16 x float> inreg %arg2, i32 %scale0, i32 inreg %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_sgpr__vgpr_sgpr: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v16, s20 +; SDAG-NEXT: v_mov_b32_e32 v31, v13 +; SDAG-NEXT: v_mov_b32_e32 v30, v12 +; SDAG-NEXT: v_mov_b32_e32 v29, v11 +; SDAG-NEXT: v_mov_b32_e32 v28, v10 +; SDAG-NEXT: v_mov_b32_e32 v27, v9 +; SDAG-NEXT: v_mov_b32_e32 v26, v8 +; SDAG-NEXT: v_mov_b32_e32 v17, s21 +; SDAG-NEXT: v_mov_b32_e32 v18, s22 +; SDAG-NEXT: v_mov_b32_e32 v19, s23 +; SDAG-NEXT: v_mov_b32_e32 v20, s24 +; SDAG-NEXT: v_mov_b32_e32 v21, s25 +; SDAG-NEXT: v_mov_b32_e32 v22, s26 +; SDAG-NEXT: v_mov_b32_e32 v23, s27 +; SDAG-NEXT: v_mov_b32_e32 v24, s28 +; SDAG-NEXT: v_mov_b32_e32 v25, s29 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_mov_b32_e32 v32, s0 +; SDAG-NEXT: v_mov_b32_e32 v33, s1 +; SDAG-NEXT: v_mov_b32_e32 v34, s2 +; SDAG-NEXT: v_mov_b32_e32 v35, s3 +; SDAG-NEXT: v_mov_b32_e32 v36, s16 +; SDAG-NEXT: v_mov_b32_e32 v37, s17 +; SDAG-NEXT: v_mov_b32_e32 v38, s18 +; SDAG-NEXT: v_mov_b32_e32 v39, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: v_accvgpr_write_b32 a15, v31 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[32:39], v[0:7], a[0:15], v14, v15 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0_sgpr_vgpr_sgpr__vgpr_sgpr: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: v_mov_b32_e32 v16, s20 +; GISEL-NEXT: s_mov_b32 s12, s0 +; GISEL-NEXT: s_mov_b32 s13, s1 +; GISEL-NEXT: s_mov_b32 s14, s2 +; GISEL-NEXT: s_mov_b32 s15, s3 +; GISEL-NEXT: v_mov_b32_e32 v26, v8 +; GISEL-NEXT: v_mov_b32_e32 v27, v9 +; GISEL-NEXT: v_mov_b32_e32 v28, v10 +; GISEL-NEXT: v_mov_b32_e32 v29, v11 +; GISEL-NEXT: v_mov_b32_e32 v30, v12 +; GISEL-NEXT: v_mov_b32_e32 v31, v13 +; GISEL-NEXT: v_mov_b32_e32 v17, s21 +; GISEL-NEXT: v_mov_b32_e32 v18, s22 +; GISEL-NEXT: v_mov_b32_e32 v19, s23 +; GISEL-NEXT: v_mov_b32_e32 v20, s24 +; GISEL-NEXT: v_mov_b32_e32 v21, s25 +; GISEL-NEXT: v_mov_b32_e32 v22, s26 +; GISEL-NEXT: v_mov_b32_e32 v23, s27 +; GISEL-NEXT: v_mov_b32_e32 v24, s28 +; GISEL-NEXT: v_mov_b32_e32 v25, s29 +; GISEL-NEXT: v_mov_b64_e32 v[38:39], s[18:19] +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_mov_b64_e32 v[36:37], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[34:35], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[32:33], s[12:13] +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: v_accvgpr_write_b32 a15, v31 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[32:39], v[0:7], a[0:15], v14, v15 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_inlineimm__scaleB_inlineimm(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_inlineimm__scaleB_inlineimm: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 33, -2 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 2, i32 33, i32 2, i32 -2) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_kimm__scaleB_inlineimm(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_kimm__scaleB_inlineimm: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: s_movk_i32 s0, 0x41 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, -2 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_kimm__scaleB_inlineimm: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: v_mov_b32_e32 v31, 0x41 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, -2 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 2, i32 65, i32 2, i32 -2) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_kimm__scaleB_kimm(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_kimm__scaleB_kimm: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: s_movk_i32 s0, 0x41 +; SDAG-NEXT: v_mov_b32_e32 v31, 0x4d +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, v31 op_sel_hi:[0,0,0] +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__scaleA_kimm__scaleB_kimm: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: v_mov_b32_e32 v31, 0x41 +; GISEL-NEXT: v_mov_b32_e32 v32, 0x4d +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 2, i32 65, i32 2, i32 77) + ret <16 x float> %result +} + +define amdgpu_kernel void @test_mfma_scale_f32_32x32x64_f8f6f4__vgprcd(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1, ptr addrspace(1) %ptr) #0 { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4__vgprcd: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; SDAG-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x40 +; SDAG-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x80 +; SDAG-NEXT: v_mov_b32_e32 v16, 0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s36 +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: v_mov_b32_e32 v4, s12 +; SDAG-NEXT: v_mov_b32_e32 v5, s13 +; SDAG-NEXT: v_mov_b32_e32 v6, s14 +; SDAG-NEXT: v_mov_b32_e32 v7, s15 +; SDAG-NEXT: v_mov_b32_e32 v8, s16 +; SDAG-NEXT: v_mov_b32_e32 v9, s17 +; SDAG-NEXT: v_mov_b32_e32 v10, s18 +; SDAG-NEXT: v_mov_b32_e32 v11, s19 +; SDAG-NEXT: v_mov_b32_e32 v12, s20 +; SDAG-NEXT: v_mov_b32_e32 v13, s21 +; SDAG-NEXT: v_mov_b32_e32 v14, s22 +; SDAG-NEXT: v_mov_b32_e32 v15, s23 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s37 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s38 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s39 +; SDAG-NEXT: v_accvgpr_write_b32 a4, s40 +; SDAG-NEXT: v_accvgpr_write_b32 a5, s41 +; SDAG-NEXT: v_accvgpr_write_b32 a6, s42 +; SDAG-NEXT: v_accvgpr_write_b32 a7, s43 +; SDAG-NEXT: v_accvgpr_write_b32 a8, s44 +; SDAG-NEXT: v_accvgpr_write_b32 a9, s45 +; SDAG-NEXT: v_accvgpr_write_b32 a10, s46 +; SDAG-NEXT: v_accvgpr_write_b32 a11, s47 +; SDAG-NEXT: v_accvgpr_write_b32 a12, s48 +; SDAG-NEXT: v_accvgpr_write_b32 a13, s49 +; SDAG-NEXT: v_accvgpr_write_b32 a14, s50 +; SDAG-NEXT: v_accvgpr_write_b32 a15, s51 +; SDAG-NEXT: v_mov_b32_e32 v17, s1 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, v17 op_sel_hi:[0,0,0] blgp:2 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: global_store_dwordx4 v16, a[12:15], s[2:3] offset:48 +; SDAG-NEXT: global_store_dwordx4 v16, a[8:11], s[2:3] offset:32 +; SDAG-NEXT: global_store_dwordx4 v16, a[4:7], s[2:3] offset:16 +; SDAG-NEXT: global_store_dwordx4 v16, a[0:3], s[2:3] +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4__vgprcd: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x40 +; GISEL-NEXT: s_load_dwordx4 s[0:3], s[4:5], 0x80 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[8:9] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[16:17] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s36 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[18:19] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[22:23] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s37 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s38 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s39 +; GISEL-NEXT: v_accvgpr_write_b32 a4, s40 +; GISEL-NEXT: v_accvgpr_write_b32 a5, s41 +; GISEL-NEXT: v_accvgpr_write_b32 a6, s42 +; GISEL-NEXT: v_accvgpr_write_b32 a7, s43 +; GISEL-NEXT: v_accvgpr_write_b32 a8, s44 +; GISEL-NEXT: v_accvgpr_write_b32 a9, s45 +; GISEL-NEXT: v_accvgpr_write_b32 a10, s46 +; GISEL-NEXT: v_accvgpr_write_b32 a11, s47 +; GISEL-NEXT: v_accvgpr_write_b32 a12, s48 +; GISEL-NEXT: v_accvgpr_write_b32 a13, s49 +; GISEL-NEXT: v_accvgpr_write_b32 a14, s50 +; GISEL-NEXT: v_accvgpr_write_b32 a15, s51 +; GISEL-NEXT: v_mov_b32_e32 v16, s1 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, v16 op_sel_hi:[0,0,0] blgp:2 +; GISEL-NEXT: v_mov_b32_e32 v0, 0 +; GISEL-NEXT: s_nop 2 +; GISEL-NEXT: global_store_dwordx4 v0, a[0:3], s[2:3] +; GISEL-NEXT: global_store_dwordx4 v0, a[4:7], s[2:3] offset:16 +; GISEL-NEXT: global_store_dwordx4 v0, a[8:11], s[2:3] offset:32 +; GISEL-NEXT: global_store_dwordx4 v0, a[12:15], s[2:3] offset:48 +; GISEL-NEXT: s_endpgm + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 2, i32 3, i32 %scale0, i32 1, i32 %scale1) + store <16 x float> %result, ptr addrspace(1) %ptr, align 64 + ret void +} + +define amdgpu_kernel void @test_mfma_scale_f32_32x32x64_f8f6f4__vgprcd___scaleA_kimm__scaleB__inlineimm(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, ptr addrspace(1) %ptr) #0 { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4__vgprcd___scaleA_kimm__scaleB__inlineimm: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x40 +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; SDAG-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x80 +; SDAG-NEXT: s_movk_i32 s2, 0x41 +; SDAG-NEXT: v_mov_b32_e32 v16, 0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s36 +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: v_mov_b32_e32 v4, s12 +; SDAG-NEXT: v_mov_b32_e32 v5, s13 +; SDAG-NEXT: v_mov_b32_e32 v6, s14 +; SDAG-NEXT: v_mov_b32_e32 v7, s15 +; SDAG-NEXT: v_mov_b32_e32 v8, s16 +; SDAG-NEXT: v_mov_b32_e32 v9, s17 +; SDAG-NEXT: v_mov_b32_e32 v10, s18 +; SDAG-NEXT: v_mov_b32_e32 v11, s19 +; SDAG-NEXT: v_mov_b32_e32 v12, s20 +; SDAG-NEXT: v_mov_b32_e32 v13, s21 +; SDAG-NEXT: v_mov_b32_e32 v14, s22 +; SDAG-NEXT: v_mov_b32_e32 v15, s23 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s37 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s38 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s39 +; SDAG-NEXT: v_accvgpr_write_b32 a4, s40 +; SDAG-NEXT: v_accvgpr_write_b32 a5, s41 +; SDAG-NEXT: v_accvgpr_write_b32 a6, s42 +; SDAG-NEXT: v_accvgpr_write_b32 a7, s43 +; SDAG-NEXT: v_accvgpr_write_b32 a8, s44 +; SDAG-NEXT: v_accvgpr_write_b32 a9, s45 +; SDAG-NEXT: v_accvgpr_write_b32 a10, s46 +; SDAG-NEXT: v_accvgpr_write_b32 a11, s47 +; SDAG-NEXT: v_accvgpr_write_b32 a12, s48 +; SDAG-NEXT: v_accvgpr_write_b32 a13, s49 +; SDAG-NEXT: v_accvgpr_write_b32 a14, s50 +; SDAG-NEXT: v_accvgpr_write_b32 a15, s51 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s2, -2 op_sel_hi:[0,0,0] blgp:2 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: global_store_dwordx4 v16, a[12:15], s[0:1] offset:48 +; SDAG-NEXT: global_store_dwordx4 v16, a[8:11], s[0:1] offset:32 +; SDAG-NEXT: global_store_dwordx4 v16, a[4:7], s[0:1] offset:16 +; SDAG-NEXT: global_store_dwordx4 v16, a[0:3], s[0:1] +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4__vgprcd___scaleA_kimm__scaleB__inlineimm: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x40 +; GISEL-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x80 +; GISEL-NEXT: v_mov_b32_e32 v16, 0x41 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[8:9] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[16:17] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s36 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[18:19] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[22:23] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s37 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s38 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s39 +; GISEL-NEXT: v_accvgpr_write_b32 a4, s40 +; GISEL-NEXT: v_accvgpr_write_b32 a5, s41 +; GISEL-NEXT: v_accvgpr_write_b32 a6, s42 +; GISEL-NEXT: v_accvgpr_write_b32 a7, s43 +; GISEL-NEXT: v_accvgpr_write_b32 a8, s44 +; GISEL-NEXT: v_accvgpr_write_b32 a9, s45 +; GISEL-NEXT: v_accvgpr_write_b32 a10, s46 +; GISEL-NEXT: v_accvgpr_write_b32 a11, s47 +; GISEL-NEXT: v_accvgpr_write_b32 a12, s48 +; GISEL-NEXT: v_accvgpr_write_b32 a13, s49 +; GISEL-NEXT: v_accvgpr_write_b32 a14, s50 +; GISEL-NEXT: v_accvgpr_write_b32 a15, s51 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v16, -2 op_sel_hi:[0,0,0] blgp:2 +; GISEL-NEXT: v_mov_b32_e32 v0, 0 +; GISEL-NEXT: s_nop 2 +; GISEL-NEXT: global_store_dwordx4 v0, a[0:3], s[0:1] +; GISEL-NEXT: global_store_dwordx4 v0, a[4:7], s[0:1] offset:16 +; GISEL-NEXT: global_store_dwordx4 v0, a[8:11], s[0:1] offset:32 +; GISEL-NEXT: global_store_dwordx4 v0, a[12:15], s[0:1] offset:48 +; GISEL-NEXT: s_endpgm + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 2, i32 3, i32 65, i32 1, i32 -2) + store <16 x float> %result, ptr addrspace(1) %ptr, align 64 + ret void +} + +define amdgpu_kernel void @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__nonmac(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) #1 { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__nonmac: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx16 s[12:27], s[4:5], 0x0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: v_mov_b32_e32 v4, s16 +; SDAG-NEXT: v_mov_b32_e32 v5, s17 +; SDAG-NEXT: v_mov_b32_e32 v6, s18 +; SDAG-NEXT: v_mov_b32_e32 v7, s19 +; SDAG-NEXT: v_mov_b32_e32 v8, s20 +; SDAG-NEXT: v_mov_b32_e32 v9, s21 +; SDAG-NEXT: v_mov_b32_e32 v10, s22 +; SDAG-NEXT: v_mov_b32_e32 v11, s23 +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; SDAG-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x80 +; SDAG-NEXT: v_mov_b32_e32 v12, s24 +; SDAG-NEXT: v_mov_b32_e32 v13, s25 +; SDAG-NEXT: v_mov_b32_e32 v14, s26 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s8 +; SDAG-NEXT: v_mov_b32_e32 v15, s27 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s11 +; SDAG-NEXT: v_accvgpr_write_b32 a4, s12 +; SDAG-NEXT: v_accvgpr_write_b32 a5, s13 +; SDAG-NEXT: v_accvgpr_write_b32 a6, s14 +; SDAG-NEXT: v_accvgpr_write_b32 a7, s15 +; SDAG-NEXT: v_accvgpr_write_b32 a8, s16 +; SDAG-NEXT: v_accvgpr_write_b32 a9, s17 +; SDAG-NEXT: v_accvgpr_write_b32 a10, s18 +; SDAG-NEXT: v_accvgpr_write_b32 a11, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a12, s20 +; SDAG-NEXT: v_accvgpr_write_b32 a13, s21 +; SDAG-NEXT: v_accvgpr_write_b32 a14, s22 +; SDAG-NEXT: v_accvgpr_write_b32 a15, s23 +; SDAG-NEXT: v_mov_b32_e32 v16, s1 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, v16 op_sel_hi:[0,0,0] +; SDAG-NEXT: v_mov_b32_e32 v0, s20 +; SDAG-NEXT: v_mov_b32_e32 v1, s21 +; SDAG-NEXT: v_mov_b32_e32 v2, s22 +; SDAG-NEXT: v_mov_b32_e32 v3, s23 +; SDAG-NEXT: v_mov_b64_e32 v[4:5], 48 +; SDAG-NEXT: global_store_dwordx4 v[4:5], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[6:7], 32 +; SDAG-NEXT: v_mov_b64_e32 v[8:9], 16 +; SDAG-NEXT: v_mov_b32_e32 v0, s16 +; SDAG-NEXT: v_mov_b32_e32 v1, s17 +; SDAG-NEXT: v_mov_b32_e32 v2, s18 +; SDAG-NEXT: v_mov_b32_e32 v3, s19 +; SDAG-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[10:11], 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[6:7], a[8:11], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[4:5], a[12:15], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[10:11], a[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[8:9], a[4:7], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__nonmac: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; GISEL-NEXT: s_load_dwordx2 s[0:1], s[4:5], 0x80 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[36:37] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[44:45] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s8 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[38:39] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[40:41] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[42:43] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[46:47] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[48:49] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[50:51] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s11 +; GISEL-NEXT: v_accvgpr_write_b32 a4, s12 +; GISEL-NEXT: v_accvgpr_write_b32 a5, s13 +; GISEL-NEXT: v_accvgpr_write_b32 a6, s14 +; GISEL-NEXT: v_accvgpr_write_b32 a7, s15 +; GISEL-NEXT: v_accvgpr_write_b32 a8, s16 +; GISEL-NEXT: v_accvgpr_write_b32 a9, s17 +; GISEL-NEXT: v_accvgpr_write_b32 a10, s18 +; GISEL-NEXT: v_accvgpr_write_b32 a11, s19 +; GISEL-NEXT: v_accvgpr_write_b32 a12, s20 +; GISEL-NEXT: v_accvgpr_write_b32 a13, s21 +; GISEL-NEXT: v_accvgpr_write_b32 a14, s22 +; GISEL-NEXT: v_accvgpr_write_b32 a15, s23 +; GISEL-NEXT: v_mov_b32_e32 v16, s1 +; GISEL-NEXT: s_nop 1 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], s0, v16 op_sel_hi:[0,0,0] +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[8:9] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], 0 +; GISEL-NEXT: global_store_dwordx4 v[4:5], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[6:7], 16 +; GISEL-NEXT: v_mov_b64_e32 v[8:9], 32 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[14:15] +; GISEL-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[10:11], 48 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[18:19] +; GISEL-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[22:23] +; GISEL-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[4:5], a[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[6:7], a[4:7], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[8:9], a[8:11], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[10:11], a[12:15], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_endpgm + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 %scale0, i32 0, i32 %scale1) + store volatile <16 x float> %arg2, ptr addrspace(1) null, align 64 + store volatile <16 x float> %result, ptr addrspace(1) null, align 64 + ret void +} + +define amdgpu_kernel void @test_mfma_scale_f32_32x32x64_f8f6f4_25_42__nonmac(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) #1 { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_25_42__nonmac: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx16 s[12:27], s[4:5], 0x0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: v_mov_b32_e32 v4, s16 +; SDAG-NEXT: v_mov_b32_e32 v5, s17 +; SDAG-NEXT: v_mov_b32_e32 v6, s18 +; SDAG-NEXT: v_mov_b32_e32 v7, s19 +; SDAG-NEXT: v_mov_b32_e32 v8, s20 +; SDAG-NEXT: v_mov_b32_e32 v9, s21 +; SDAG-NEXT: v_mov_b32_e32 v10, s22 +; SDAG-NEXT: v_mov_b32_e32 v11, s23 +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; SDAG-NEXT: v_mov_b32_e32 v12, s24 +; SDAG-NEXT: v_mov_b32_e32 v13, s25 +; SDAG-NEXT: v_mov_b32_e32 v14, s26 +; SDAG-NEXT: v_mov_b32_e32 v15, s27 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s8 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s11 +; SDAG-NEXT: v_accvgpr_write_b32 a4, s12 +; SDAG-NEXT: v_accvgpr_write_b32 a5, s13 +; SDAG-NEXT: v_accvgpr_write_b32 a6, s14 +; SDAG-NEXT: v_accvgpr_write_b32 a7, s15 +; SDAG-NEXT: v_accvgpr_write_b32 a8, s16 +; SDAG-NEXT: v_accvgpr_write_b32 a9, s17 +; SDAG-NEXT: v_accvgpr_write_b32 a10, s18 +; SDAG-NEXT: v_accvgpr_write_b32 a11, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a12, s20 +; SDAG-NEXT: v_accvgpr_write_b32 a13, s21 +; SDAG-NEXT: v_accvgpr_write_b32 a14, s22 +; SDAG-NEXT: v_accvgpr_write_b32 a15, s23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 25, 42 op_sel_hi:[0,0,0] blgp:2 +; SDAG-NEXT: v_mov_b32_e32 v0, s20 +; SDAG-NEXT: v_mov_b32_e32 v1, s21 +; SDAG-NEXT: v_mov_b32_e32 v2, s22 +; SDAG-NEXT: v_mov_b32_e32 v3, s23 +; SDAG-NEXT: v_mov_b64_e32 v[4:5], 48 +; SDAG-NEXT: global_store_dwordx4 v[4:5], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[6:7], 32 +; SDAG-NEXT: v_mov_b64_e32 v[8:9], 16 +; SDAG-NEXT: v_mov_b32_e32 v0, s16 +; SDAG-NEXT: v_mov_b32_e32 v1, s17 +; SDAG-NEXT: v_mov_b32_e32 v2, s18 +; SDAG-NEXT: v_mov_b32_e32 v3, s19 +; SDAG-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[10:11], 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[6:7], a[8:11], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[4:5], a[12:15], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[10:11], a[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[8:9], a[4:7], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_25_42__nonmac: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[36:37] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[44:45] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s8 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[38:39] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[40:41] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[42:43] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[46:47] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[48:49] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[50:51] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s11 +; GISEL-NEXT: v_accvgpr_write_b32 a4, s12 +; GISEL-NEXT: v_accvgpr_write_b32 a5, s13 +; GISEL-NEXT: v_accvgpr_write_b32 a6, s14 +; GISEL-NEXT: v_accvgpr_write_b32 a7, s15 +; GISEL-NEXT: v_accvgpr_write_b32 a8, s16 +; GISEL-NEXT: v_accvgpr_write_b32 a9, s17 +; GISEL-NEXT: v_accvgpr_write_b32 a10, s18 +; GISEL-NEXT: v_accvgpr_write_b32 a11, s19 +; GISEL-NEXT: v_accvgpr_write_b32 a12, s20 +; GISEL-NEXT: v_accvgpr_write_b32 a13, s21 +; GISEL-NEXT: v_accvgpr_write_b32 a14, s22 +; GISEL-NEXT: v_accvgpr_write_b32 a15, s23 +; GISEL-NEXT: v_mov_b64_e32 v[18:19], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[8:9] +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 25, 42 op_sel_hi:[0,0,0] blgp:2 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], 0 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], 16 +; GISEL-NEXT: global_store_dwordx4 v[4:5], v[16:19], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[8:9], 32 +; GISEL-NEXT: v_mov_b64_e32 v[10:11], 48 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[18:19] +; GISEL-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[22:23] +; GISEL-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[4:5], a[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[6:7], a[4:7], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[8:9], a[8:11], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[10:11], a[12:15], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_endpgm + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 2, i32 0, i32 25, i32 0, i32 42) + store volatile <16 x float> %arg2, ptr addrspace(1) null, align 64 + store volatile <16 x float> %result, ptr addrspace(1) null, align 64 + ret void +} + +define amdgpu_kernel void @test_mfma_scale_f32_32x32x64_f8f6f4_0_0__vgprcd_nonmac(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) #0 { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__vgprcd_nonmac: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx16 s[12:27], s[4:5], 0x0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: v_mov_b32_e32 v4, s16 +; SDAG-NEXT: v_mov_b32_e32 v5, s17 +; SDAG-NEXT: v_mov_b32_e32 v6, s18 +; SDAG-NEXT: v_mov_b32_e32 v7, s19 +; SDAG-NEXT: v_mov_b32_e32 v8, s20 +; SDAG-NEXT: v_mov_b32_e32 v9, s21 +; SDAG-NEXT: v_mov_b32_e32 v10, s22 +; SDAG-NEXT: v_mov_b32_e32 v11, s23 +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; SDAG-NEXT: v_mov_b32_e32 v12, s24 +; SDAG-NEXT: v_mov_b32_e32 v13, s25 +; SDAG-NEXT: v_mov_b32_e32 v14, s26 +; SDAG-NEXT: v_mov_b32_e32 v15, s27 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s8 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s11 +; SDAG-NEXT: v_accvgpr_write_b32 a4, s12 +; SDAG-NEXT: v_accvgpr_write_b32 a5, s13 +; SDAG-NEXT: v_accvgpr_write_b32 a6, s14 +; SDAG-NEXT: v_accvgpr_write_b32 a7, s15 +; SDAG-NEXT: v_accvgpr_write_b32 a8, s16 +; SDAG-NEXT: v_accvgpr_write_b32 a9, s17 +; SDAG-NEXT: v_accvgpr_write_b32 a10, s18 +; SDAG-NEXT: v_accvgpr_write_b32 a11, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a12, s20 +; SDAG-NEXT: v_accvgpr_write_b32 a13, s21 +; SDAG-NEXT: v_accvgpr_write_b32 a14, s22 +; SDAG-NEXT: v_accvgpr_write_b32 a15, s23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] blgp:2 +; SDAG-NEXT: v_mov_b32_e32 v0, s20 +; SDAG-NEXT: v_mov_b32_e32 v1, s21 +; SDAG-NEXT: v_mov_b32_e32 v2, s22 +; SDAG-NEXT: v_mov_b32_e32 v3, s23 +; SDAG-NEXT: v_mov_b64_e32 v[4:5], 48 +; SDAG-NEXT: global_store_dwordx4 v[4:5], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[6:7], 32 +; SDAG-NEXT: v_mov_b64_e32 v[8:9], 16 +; SDAG-NEXT: v_mov_b32_e32 v0, s16 +; SDAG-NEXT: v_mov_b32_e32 v1, s17 +; SDAG-NEXT: v_mov_b32_e32 v2, s18 +; SDAG-NEXT: v_mov_b32_e32 v3, s19 +; SDAG-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[10:11], 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[6:7], a[8:11], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[4:5], a[12:15], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[10:11], a[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[8:9], a[4:7], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_0_0__vgprcd_nonmac: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[36:37] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[44:45] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s8 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[38:39] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[40:41] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[42:43] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[46:47] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[48:49] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[50:51] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s11 +; GISEL-NEXT: v_accvgpr_write_b32 a4, s12 +; GISEL-NEXT: v_accvgpr_write_b32 a5, s13 +; GISEL-NEXT: v_accvgpr_write_b32 a6, s14 +; GISEL-NEXT: v_accvgpr_write_b32 a7, s15 +; GISEL-NEXT: v_accvgpr_write_b32 a8, s16 +; GISEL-NEXT: v_accvgpr_write_b32 a9, s17 +; GISEL-NEXT: v_accvgpr_write_b32 a10, s18 +; GISEL-NEXT: v_accvgpr_write_b32 a11, s19 +; GISEL-NEXT: v_accvgpr_write_b32 a12, s20 +; GISEL-NEXT: v_accvgpr_write_b32 a13, s21 +; GISEL-NEXT: v_accvgpr_write_b32 a14, s22 +; GISEL-NEXT: v_accvgpr_write_b32 a15, s23 +; GISEL-NEXT: v_mov_b64_e32 v[18:19], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[8:9] +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] blgp:2 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], 0 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], 16 +; GISEL-NEXT: global_store_dwordx4 v[4:5], v[16:19], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[8:9], 32 +; GISEL-NEXT: v_mov_b64_e32 v[10:11], 48 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[18:19] +; GISEL-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[22:23] +; GISEL-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[4:5], a[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[6:7], a[4:7], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[8:9], a[8:11], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[10:11], a[12:15], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_endpgm + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 2, i32 0, i32 0, i32 0, i32 0) + store volatile <16 x float> %arg2, ptr addrspace(1) null, align 64 + store volatile <16 x float> %result, ptr addrspace(1) null, align 64 + ret void +} + +define amdgpu_kernel void @test_mfma_scale_f32_32x32x64_f8f6f4_25_42__vgprcd_nonmac(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) #0 { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_25_42__vgprcd_nonmac: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_load_dwordx16 s[12:27], s[4:5], 0x0 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: v_mov_b32_e32 v4, s16 +; SDAG-NEXT: v_mov_b32_e32 v5, s17 +; SDAG-NEXT: v_mov_b32_e32 v6, s18 +; SDAG-NEXT: v_mov_b32_e32 v7, s19 +; SDAG-NEXT: v_mov_b32_e32 v8, s20 +; SDAG-NEXT: v_mov_b32_e32 v9, s21 +; SDAG-NEXT: v_mov_b32_e32 v10, s22 +; SDAG-NEXT: v_mov_b32_e32 v11, s23 +; SDAG-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; SDAG-NEXT: v_mov_b32_e32 v12, s24 +; SDAG-NEXT: v_mov_b32_e32 v13, s25 +; SDAG-NEXT: v_mov_b32_e32 v14, s26 +; SDAG-NEXT: v_mov_b32_e32 v15, s27 +; SDAG-NEXT: s_waitcnt lgkmcnt(0) +; SDAG-NEXT: v_accvgpr_write_b32 a0, s8 +; SDAG-NEXT: v_accvgpr_write_b32 a1, s9 +; SDAG-NEXT: v_accvgpr_write_b32 a2, s10 +; SDAG-NEXT: v_accvgpr_write_b32 a3, s11 +; SDAG-NEXT: v_accvgpr_write_b32 a4, s12 +; SDAG-NEXT: v_accvgpr_write_b32 a5, s13 +; SDAG-NEXT: v_accvgpr_write_b32 a6, s14 +; SDAG-NEXT: v_accvgpr_write_b32 a7, s15 +; SDAG-NEXT: v_accvgpr_write_b32 a8, s16 +; SDAG-NEXT: v_accvgpr_write_b32 a9, s17 +; SDAG-NEXT: v_accvgpr_write_b32 a10, s18 +; SDAG-NEXT: v_accvgpr_write_b32 a11, s19 +; SDAG-NEXT: v_accvgpr_write_b32 a12, s20 +; SDAG-NEXT: v_accvgpr_write_b32 a13, s21 +; SDAG-NEXT: v_accvgpr_write_b32 a14, s22 +; SDAG-NEXT: v_accvgpr_write_b32 a15, s23 +; SDAG-NEXT: s_nop 1 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 25, 42 op_sel_hi:[0,0,0] blgp:2 +; SDAG-NEXT: v_mov_b32_e32 v0, s20 +; SDAG-NEXT: v_mov_b32_e32 v1, s21 +; SDAG-NEXT: v_mov_b32_e32 v2, s22 +; SDAG-NEXT: v_mov_b32_e32 v3, s23 +; SDAG-NEXT: v_mov_b64_e32 v[4:5], 48 +; SDAG-NEXT: global_store_dwordx4 v[4:5], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[6:7], 32 +; SDAG-NEXT: v_mov_b64_e32 v[8:9], 16 +; SDAG-NEXT: v_mov_b32_e32 v0, s16 +; SDAG-NEXT: v_mov_b32_e32 v1, s17 +; SDAG-NEXT: v_mov_b32_e32 v2, s18 +; SDAG-NEXT: v_mov_b32_e32 v3, s19 +; SDAG-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: v_mov_b64_e32 v[10:11], 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s12 +; SDAG-NEXT: v_mov_b32_e32 v1, s13 +; SDAG-NEXT: v_mov_b32_e32 v2, s14 +; SDAG-NEXT: v_mov_b32_e32 v3, s15 +; SDAG-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mov_b32_e32 v0, s8 +; SDAG-NEXT: v_mov_b32_e32 v1, s9 +; SDAG-NEXT: v_mov_b32_e32 v2, s10 +; SDAG-NEXT: v_mov_b32_e32 v3, s11 +; SDAG-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[6:7], a[8:11], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[4:5], a[12:15], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[10:11], a[0:3], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: global_store_dwordx4 v[8:9], a[4:7], off sc0 sc1 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_endpgm +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4_25_42__vgprcd_nonmac: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_load_dwordx16 s[36:51], s[4:5], 0x0 +; GISEL-NEXT: s_load_dwordx16 s[8:23], s[4:5], 0x40 +; GISEL-NEXT: s_waitcnt lgkmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[36:37] +; GISEL-NEXT: v_mov_b64_e32 v[8:9], s[44:45] +; GISEL-NEXT: v_accvgpr_write_b32 a0, s8 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[38:39] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], s[40:41] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], s[42:43] +; GISEL-NEXT: v_mov_b64_e32 v[10:11], s[46:47] +; GISEL-NEXT: v_mov_b64_e32 v[12:13], s[48:49] +; GISEL-NEXT: v_mov_b64_e32 v[14:15], s[50:51] +; GISEL-NEXT: v_accvgpr_write_b32 a1, s9 +; GISEL-NEXT: v_accvgpr_write_b32 a2, s10 +; GISEL-NEXT: v_accvgpr_write_b32 a3, s11 +; GISEL-NEXT: v_accvgpr_write_b32 a4, s12 +; GISEL-NEXT: v_accvgpr_write_b32 a5, s13 +; GISEL-NEXT: v_accvgpr_write_b32 a6, s14 +; GISEL-NEXT: v_accvgpr_write_b32 a7, s15 +; GISEL-NEXT: v_accvgpr_write_b32 a8, s16 +; GISEL-NEXT: v_accvgpr_write_b32 a9, s17 +; GISEL-NEXT: v_accvgpr_write_b32 a10, s18 +; GISEL-NEXT: v_accvgpr_write_b32 a11, s19 +; GISEL-NEXT: v_accvgpr_write_b32 a12, s20 +; GISEL-NEXT: v_accvgpr_write_b32 a13, s21 +; GISEL-NEXT: v_accvgpr_write_b32 a14, s22 +; GISEL-NEXT: v_accvgpr_write_b32 a15, s23 +; GISEL-NEXT: v_mov_b64_e32 v[18:19], s[10:11] +; GISEL-NEXT: v_mov_b64_e32 v[16:17], s[8:9] +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 25, 42 op_sel_hi:[0,0,0] blgp:2 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[12:13] +; GISEL-NEXT: v_mov_b64_e32 v[4:5], 0 +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[14:15] +; GISEL-NEXT: v_mov_b64_e32 v[6:7], 16 +; GISEL-NEXT: global_store_dwordx4 v[4:5], v[16:19], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[6:7], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: v_mov_b64_e32 v[8:9], 32 +; GISEL-NEXT: v_mov_b64_e32 v[10:11], 48 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[16:17] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[18:19] +; GISEL-NEXT: global_store_dwordx4 v[8:9], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mov_b64_e32 v[0:1], s[20:21] +; GISEL-NEXT: v_mov_b64_e32 v[2:3], s[22:23] +; GISEL-NEXT: global_store_dwordx4 v[10:11], v[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[4:5], a[0:3], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[6:7], a[4:7], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[8:9], a[8:11], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: global_store_dwordx4 v[10:11], a[12:15], off sc0 sc1 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_endpgm + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 2, i32 0, i32 25, i32 0, i32 42) + store volatile <16 x float> %arg2, ptr addrspace(1) null, align 64 + store volatile <16 x float> %result, ptr addrspace(1) null, align 64 + ret void +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_0_0_a(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_0_0_a: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_0_0_b(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_0_0_b: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 3, i32 0, i32 1, i32 0) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_0_1(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_0_1: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 1 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 0, i32 0, i32 1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_1_0_a(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___constant_scale_1_0_a: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 1, 0 op_sel_hi:[0,0,0] +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 0, i32 0, i32 0, i32 1, i32 0, i32 0) + ret <16 x float> %result +} + +; -------------------------------------------------------------------- +; Incorrect signature for format cases (IR vector too large) +; -------------------------------------------------------------------- + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp6: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] blgp:2 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp6: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] blgp:2 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp8: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] cbsz:2 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp8: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] cbsz:2 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp6__v8i32_fp6__0_scale: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:2 blgp:2 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 2, ; cbsz + i32 2, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp4: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] blgp:4 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v8i32_fp4: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] blgp:4 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp8(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp8: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] cbsz:4 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp8: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] cbsz:4 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v6i32_fp4(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp8__v6i32_fp4: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32> %arg0, <6 x i32> %arg1, <16 x float> %arg2, + i32 0, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v6i32_fp4__v8i32_fp8(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v6i32_fp4__v8i32_fp8: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword v31, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v14 +; GCN-NEXT: v_accvgpr_write_b32 a1, v15 +; GCN-NEXT: v_accvgpr_write_b32 a2, v16 +; GCN-NEXT: v_accvgpr_write_b32 a3, v17 +; GCN-NEXT: v_accvgpr_write_b32 a4, v18 +; GCN-NEXT: v_accvgpr_write_b32 a5, v19 +; GCN-NEXT: v_accvgpr_write_b32 a6, v20 +; GCN-NEXT: v_accvgpr_write_b32 a7, v21 +; GCN-NEXT: v_accvgpr_write_b32 a8, v22 +; GCN-NEXT: v_accvgpr_write_b32 a9, v23 +; GCN-NEXT: v_accvgpr_write_b32 a10, v24 +; GCN-NEXT: v_accvgpr_write_b32 a11, v25 +; GCN-NEXT: v_accvgpr_write_b32 a12, v26 +; GCN-NEXT: v_accvgpr_write_b32 a13, v27 +; GCN-NEXT: v_accvgpr_write_b32 a14, v28 +; GCN-NEXT: v_accvgpr_write_b32 a15, v29 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:5], v[6:13], a[0:15], v30, v31 op_sel_hi:[0,0,0] cbsz:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 0, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, i32 %scale0, i32 %scale1) { +; SDAG-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4: +; SDAG: ; %bb.0: +; SDAG-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; SDAG-NEXT: scratch_load_dword a15, off, s32 +; SDAG-NEXT: scratch_load_dword v31, off, s32 offset:8 +; SDAG-NEXT: scratch_load_dword v32, off, s32 offset:4 +; SDAG-NEXT: v_accvgpr_write_b32 a0, v16 +; SDAG-NEXT: v_accvgpr_write_b32 a1, v17 +; SDAG-NEXT: v_accvgpr_write_b32 a2, v18 +; SDAG-NEXT: v_accvgpr_write_b32 a3, v19 +; SDAG-NEXT: v_accvgpr_write_b32 a4, v20 +; SDAG-NEXT: v_accvgpr_write_b32 a5, v21 +; SDAG-NEXT: v_accvgpr_write_b32 a6, v22 +; SDAG-NEXT: v_accvgpr_write_b32 a7, v23 +; SDAG-NEXT: v_accvgpr_write_b32 a8, v24 +; SDAG-NEXT: v_accvgpr_write_b32 a9, v25 +; SDAG-NEXT: v_accvgpr_write_b32 a10, v26 +; SDAG-NEXT: v_accvgpr_write_b32 a11, v27 +; SDAG-NEXT: v_accvgpr_write_b32 a12, v28 +; SDAG-NEXT: v_accvgpr_write_b32 a13, v29 +; SDAG-NEXT: v_accvgpr_write_b32 a14, v30 +; SDAG-NEXT: s_waitcnt vmcnt(0) +; SDAG-NEXT: s_nop 0 +; SDAG-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v32, v31 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; SDAG-NEXT: s_nop 3 +; SDAG-NEXT: v_accvgpr_read_b32 v0, a0 +; SDAG-NEXT: v_accvgpr_read_b32 v1, a1 +; SDAG-NEXT: v_accvgpr_read_b32 v2, a2 +; SDAG-NEXT: v_accvgpr_read_b32 v3, a3 +; SDAG-NEXT: v_accvgpr_read_b32 v4, a4 +; SDAG-NEXT: v_accvgpr_read_b32 v5, a5 +; SDAG-NEXT: v_accvgpr_read_b32 v6, a6 +; SDAG-NEXT: v_accvgpr_read_b32 v7, a7 +; SDAG-NEXT: v_accvgpr_read_b32 v8, a8 +; SDAG-NEXT: v_accvgpr_read_b32 v9, a9 +; SDAG-NEXT: v_accvgpr_read_b32 v10, a10 +; SDAG-NEXT: v_accvgpr_read_b32 v11, a11 +; SDAG-NEXT: v_accvgpr_read_b32 v12, a12 +; SDAG-NEXT: v_accvgpr_read_b32 v13, a13 +; SDAG-NEXT: v_accvgpr_read_b32 v14, a14 +; SDAG-NEXT: v_accvgpr_read_b32 v15, a15 +; SDAG-NEXT: s_setpc_b64 s[30:31] +; +; GISEL-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4: +; GISEL: ; %bb.0: +; GISEL-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GISEL-NEXT: scratch_load_dword a15, off, s32 +; GISEL-NEXT: scratch_load_dword v31, off, s32 offset:4 +; GISEL-NEXT: scratch_load_dword v32, off, s32 offset:8 +; GISEL-NEXT: v_accvgpr_write_b32 a0, v16 +; GISEL-NEXT: v_accvgpr_write_b32 a1, v17 +; GISEL-NEXT: v_accvgpr_write_b32 a2, v18 +; GISEL-NEXT: v_accvgpr_write_b32 a3, v19 +; GISEL-NEXT: v_accvgpr_write_b32 a4, v20 +; GISEL-NEXT: v_accvgpr_write_b32 a5, v21 +; GISEL-NEXT: v_accvgpr_write_b32 a6, v22 +; GISEL-NEXT: v_accvgpr_write_b32 a7, v23 +; GISEL-NEXT: v_accvgpr_write_b32 a8, v24 +; GISEL-NEXT: v_accvgpr_write_b32 a9, v25 +; GISEL-NEXT: v_accvgpr_write_b32 a10, v26 +; GISEL-NEXT: v_accvgpr_write_b32 a11, v27 +; GISEL-NEXT: v_accvgpr_write_b32 a12, v28 +; GISEL-NEXT: v_accvgpr_write_b32 a13, v29 +; GISEL-NEXT: v_accvgpr_write_b32 a14, v30 +; GISEL-NEXT: s_waitcnt vmcnt(0) +; GISEL-NEXT: s_nop 0 +; GISEL-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], v31, v32 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GISEL-NEXT: s_nop 3 +; GISEL-NEXT: v_accvgpr_read_b32 v0, a0 +; GISEL-NEXT: v_accvgpr_read_b32 v1, a1 +; GISEL-NEXT: v_accvgpr_read_b32 v2, a2 +; GISEL-NEXT: v_accvgpr_read_b32 v3, a3 +; GISEL-NEXT: v_accvgpr_read_b32 v4, a4 +; GISEL-NEXT: v_accvgpr_read_b32 v5, a5 +; GISEL-NEXT: v_accvgpr_read_b32 v6, a6 +; GISEL-NEXT: v_accvgpr_read_b32 v7, a7 +; GISEL-NEXT: v_accvgpr_read_b32 v8, a8 +; GISEL-NEXT: v_accvgpr_read_b32 v9, a9 +; GISEL-NEXT: v_accvgpr_read_b32 v10, a10 +; GISEL-NEXT: v_accvgpr_read_b32 v11, a11 +; GISEL-NEXT: v_accvgpr_read_b32 v12, a12 +; GISEL-NEXT: v_accvgpr_read_b32 v13, a13 +; GISEL-NEXT: v_accvgpr_read_b32 v14, a14 +; GISEL-NEXT: v_accvgpr_read_b32 v15, a15 +; GISEL-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 %scale0, i32 0, i32 %scale1) + ret <16 x float> %result +} + +define <16 x float> @test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2) { +; GCN-LABEL: test_mfma_scale_f32_32x32x64_f8f6f4___v8i32_fp4__v8i32_fp4__0_scale: +; GCN: ; %bb.0: +; GCN-NEXT: s_waitcnt vmcnt(0) expcnt(0) lgkmcnt(0) +; GCN-NEXT: scratch_load_dword a15, off, s32 +; GCN-NEXT: v_accvgpr_write_b32 a0, v16 +; GCN-NEXT: v_accvgpr_write_b32 a1, v17 +; GCN-NEXT: v_accvgpr_write_b32 a2, v18 +; GCN-NEXT: v_accvgpr_write_b32 a3, v19 +; GCN-NEXT: v_accvgpr_write_b32 a4, v20 +; GCN-NEXT: v_accvgpr_write_b32 a5, v21 +; GCN-NEXT: v_accvgpr_write_b32 a6, v22 +; GCN-NEXT: v_accvgpr_write_b32 a7, v23 +; GCN-NEXT: v_accvgpr_write_b32 a8, v24 +; GCN-NEXT: v_accvgpr_write_b32 a9, v25 +; GCN-NEXT: v_accvgpr_write_b32 a10, v26 +; GCN-NEXT: v_accvgpr_write_b32 a11, v27 +; GCN-NEXT: v_accvgpr_write_b32 a12, v28 +; GCN-NEXT: v_accvgpr_write_b32 a13, v29 +; GCN-NEXT: v_accvgpr_write_b32 a14, v30 +; GCN-NEXT: s_waitcnt vmcnt(0) +; GCN-NEXT: s_nop 0 +; GCN-NEXT: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[0:7], v[8:15], a[0:15], 0, 0 op_sel_hi:[0,0,0] cbsz:4 blgp:4 +; GCN-NEXT: s_nop 3 +; GCN-NEXT: v_accvgpr_read_b32 v0, a0 +; GCN-NEXT: v_accvgpr_read_b32 v1, a1 +; GCN-NEXT: v_accvgpr_read_b32 v2, a2 +; GCN-NEXT: v_accvgpr_read_b32 v3, a3 +; GCN-NEXT: v_accvgpr_read_b32 v4, a4 +; GCN-NEXT: v_accvgpr_read_b32 v5, a5 +; GCN-NEXT: v_accvgpr_read_b32 v6, a6 +; GCN-NEXT: v_accvgpr_read_b32 v7, a7 +; GCN-NEXT: v_accvgpr_read_b32 v8, a8 +; GCN-NEXT: v_accvgpr_read_b32 v9, a9 +; GCN-NEXT: v_accvgpr_read_b32 v10, a10 +; GCN-NEXT: v_accvgpr_read_b32 v11, a11 +; GCN-NEXT: v_accvgpr_read_b32 v12, a12 +; GCN-NEXT: v_accvgpr_read_b32 v13, a13 +; GCN-NEXT: v_accvgpr_read_b32 v14, a14 +; GCN-NEXT: v_accvgpr_read_b32 v15, a15 +; GCN-NEXT: s_setpc_b64 s[30:31] + %result = call <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32> %arg0, <8 x i32> %arg1, <16 x float> %arg2, + i32 4, ; cbsz + i32 4, ; blgp + i32 0, i32 0, i32 0, i32 0) + ret <16 x float> %result +} + +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v8i32(<8 x i32>, <8 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v6i32(<6 x i32>, <6 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v4i32(<4 x i32>, <4 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v6i32(<4 x i32>, <6 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v4i32.v8i32(<4 x i32>, <8 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v4i32(<6 x i32>, <4 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v6i32.v8i32(<6 x i32>, <8 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v4i32(<8 x i32>, <4 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 +declare <16 x float> @llvm.amdgcn.mfma.scale.f32.32x32x64.f8f6f4.v8i32.v6i32(<8 x i32>, <6 x i32>, <16 x float>, i32 immarg, i32 immarg, i32 immarg, i32, i32 immarg, i32) #2 + +attributes #0 = { "amdgpu-flat-work-group-size"="512,512" } +attributes #1 = { "amdgpu-flat-work-group-size"="128,128" } +attributes #2 = { convergent nocallback nofree nosync nounwind willreturn memory(none) } diff --git a/llvm/test/MC/AMDGPU/mai-gfx950.s b/llvm/test/MC/AMDGPU/mai-gfx950.s index a692693638c692..07d930b0b64dfb 100644 --- a/llvm/test/MC/AMDGPU/mai-gfx950.s +++ b/llvm/test/MC/AMDGPU/mai-gfx950.s @@ -2,6 +2,13 @@ // RUN: not llvm-mc -triple=amdgcn -mcpu=gfx940 %s 2>&1 | FileCheck -check-prefix=ERR %s // RUN: not llvm-mc -triple=amdgcn -mcpu=gfx90a %s 2>&1 | FileCheck -check-prefix=ERR %s +// cbsz = SrcA +// blgp = SrcB + +// 0 = fp8. 1 = bf8 +// 2 = fp6, 3 = bf6 +// 4 = fp4 + //===----------------------------------------------------------------------===// // MFMA opcodes. //===----------------------------------------------------------------------===// @@ -275,3 +282,612 @@ v_mfma_ld_scale_b32 v1, v1 op_sel:[1,0] op_sel_hi:[0,1] // ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU v_mfma_ld_scale_b32 v1, v1 op_sel:[0,1] op_sel_hi:[1,0] + +//===----------------------------------------------------------------------===// +// v_mfma_f32_16x16x128_f8f6f4 +//===----------------------------------------------------------------------===// + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] ; encoding: [0x00,0x00,0xad,0xd3,0x04,0x09,0x02,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] blgp:1 ; encoding: [0x00,0x00,0xad,0xd3,0x04,0x09,0x02,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] blgp:1 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[4:11], v[0:3] cbsz:3 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x09,0x02,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[4:11], v[0:3] cbsz:3 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] ; encoding: [0x00,0x00,0xad,0xd3,0x04,0x09,0x02,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[4:11], v[0:3] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x09,0x02,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[4:11], v[0:3] cbsz:3 blgp:1 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] ; encoding: [0x00,0x80,0xad,0xd3,0x04,0x09,0x02,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] blgp:1 ; encoding: [0x00,0x80,0xad,0xd3,0x04,0x09,0x02,0x3c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] blgp:1 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[4:11], a[0:3] cbsz:3 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x09,0x02,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[4:11], a[0:3] cbsz:3 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] ; encoding: [0x00,0x80,0xad,0xd3,0x04,0x09,0x02,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:9], a[0:3] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xad,0xd3,0x04,0x09,0x02,0x7c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:9], a[0:3] cbsz:1 blgp:3 + +//===----------------------------------------------------------------------===// +// v_mfma_f32_32x32x64_f8f6f4 +//===----------------------------------------------------------------------===// + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] ; encoding: [0x00,0x00,0xae,0xd3,0x04,0x09,0x02,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] blgp:1 ; encoding: [0x00,0x00,0xae,0xd3,0x04,0x09,0x02,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] blgp:1 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[4:11], v[0:15] cbsz:3 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x09,0x02,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[4:11], v[0:15] cbsz:3 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] ; encoding: [0x00,0x00,0xae,0xd3,0x04,0x09,0x02,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[4:11], v[0:15] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x09,0x02,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[4:11], v[0:15] cbsz:3 blgp:1 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] ; encoding: [0x00,0x80,0xae,0xd3,0x04,0x09,0x02,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] blgp:1 ; encoding: [0x00,0x80,0xae,0xd3,0x04,0x09,0x02,0x3c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] blgp:1 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[4:11], a[0:15] cbsz:3 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x09,0x02,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[4:11], a[0:15] cbsz:3 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] ; encoding: [0x00,0x80,0xae,0xd3,0x04,0x09,0x02,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:9], a[0:15] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xae,0xd3,0x04,0x09,0x02,0x7c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:9], a[0:15] cbsz:1 blgp:3 + +//===----------------------------------------------------------------------===// +// v_mfma_scale_f32_16x16x128_f8f6f4 +//===----------------------------------------------------------------------===// +// FIXME: Test op_sel, neg, clamp + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[4:11], v[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[4:11], v[12:19], a[20:23], v24, v25 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], a[4:11], a[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], a[4:11], a[12:19], v[20:23], v24, v25 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], a[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x14] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], a[12:19], v[20:23], v24, v25 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], a[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x0c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], a[4:11], v[12:19], v[20:23], v24, v25 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[50:53], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x32,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[50:53], v[4:11], v[12:19], v[20:23], v24, v25 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v44, s24 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x2c,0x31,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v44, s24 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s24, s24 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x30,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s24, s24 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s24, v44 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x58,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s24, v44 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], m0, m0 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x7c,0xf8,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], m0, m0 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], vcc_lo, v2 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x6a,0x04,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], vcc_lo, v2 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], 9, v2 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x89,0x04,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], 9, v2 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v2, 9 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x02,0x13,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v2, 9 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s20, 9 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x14,0x12,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s20, 9 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], 33, 9 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0xa1,0x12,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], 33, 9 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 cbsz:3 blgp:1 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x08,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel:[0,1,0] cbsz:3 blgp:1 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel:[0,1,0] op_sel_hi:[0,1,0] cbsz:3 blgp:1 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x44] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23], v24, v25 op_sel:[0,1,0] op_sel_hi:[0,1,0] blgp:2 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel:[0,1,0] op_sel_hi:[0,1,0] + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel:[0,1,0] op_sel_hi:[0,1,0] cbsz:3 + +//===----------------------------------------------------------------------===// +// v_mfma_scale_f32_32x32x64_f8f6f4 +//===----------------------------------------------------------------------===// +// FIXME: Test op_sel, neg, clamp + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[16:23], v[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[16:23], v[24:31], a[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], a[16:23], a[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x1c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], a[16:23], a[24:31], v[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], a[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x14] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], a[24:31], v[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], v[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x0c] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], v[24:31], a[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[50:65], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x32,0x08,0xae,0xd3,0x10,0x31,0x82,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[50:65], v[16:23], v[24:31], v[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x24] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 cbsz:3 blgp:1 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 cbsz:2 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x04] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x44] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:29], v[32:47], v48, v49 blgp:2 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x18,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x64] +// ERR: :[[@LINE+1]]:{{[0-9]+}}: error: instruction not supported on this GPU +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel:[0,1,0] op_sel_hi:[0,1,0] cbsz:2 blgp:3 + +//===----------------------------------------------------------------------===// +// v_mfma_f32_16x16x128_f8f6f4 with appropriate register widths +//===----------------------------------------------------------------------===// + +// bf8 x fp4 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23] cbsz:1 blgp:4 ; encoding: [0x00,0x01,0xad,0xd3,0x04,0x19,0x52,0x84] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23] cbsz:1 blgp:4 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23] cbsz:1 blgp:4 ; encoding: [0x00,0x01,0xad,0xd3,0x04,0x19,0x52,0x84] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23] cbsz:1 blgp:4 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:15], a[20:23] cbsz:1 blgp:4 ; encoding: [0x00,0x81,0xad,0xd3,0x04,0x19,0x52,0x9c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:15], a[20:23] cbsz:1 blgp:4 + + +// fp4 x bf8 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:19], v[20:23] cbsz:4 blgp:1 ; encoding: [0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x24] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:19], v[20:23] cbsz:4 blgp:1 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:19], a[20:23] cbsz:4 blgp:1 ; encoding: [0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x3c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:19], a[20:23] cbsz:4 blgp:1 + + +// bf6 x bf8 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x24] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23] cbsz:3 blgp:1 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:19], a[20:23] cbsz:3 blgp:1 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x3c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:19], a[20:23] cbsz:3 blgp:1 + + +// bf8 x bf6 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23] cbsz:1 blgp:3 ; encoding: [0x00,0x01,0xad,0xd3,0x04,0x19,0x52,0x64] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23] cbsz:1 blgp:3 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:17], a[20:23] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xad,0xd3,0x04,0x19,0x52,0x7c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:17], a[20:23] cbsz:1 blgp:3 + + +// bf6 x bf6 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:3 blgp:3 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x64] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:3 blgp:3 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:3 blgp:3 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x7c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:3 blgp:3 + +// bf6 x fp6 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:3 blgp:2 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x44] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:3 blgp:2 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:3 blgp:2 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x5c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:3 blgp:2 + + +// fp6 x bf6 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:2 blgp:3 ; encoding: [0x00,0x02,0xad,0xd3,0x04,0x19,0x52,0x64] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:2 blgp:3 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:2 blgp:3 ; encoding: [0x00,0x82,0xad,0xd3,0x04,0x19,0x52,0x7c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:2 blgp:3 + + +// fp6 x fp4 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:15], v[20:23] cbsz:2 blgp:4 ; encoding: [0x00,0x02,0xad,0xd3,0x04,0x19,0x52,0x84] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:15], v[20:23] cbsz:2 blgp:4 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:15], a[20:23] cbsz:2 blgp:4 ; encoding: [0x00,0x82,0xad,0xd3,0x04,0x19,0x52,0x9c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:15], a[20:23] cbsz:2 blgp:4 + + +// fp4 x fp6 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:17], v[20:23] cbsz:4 blgp:2 ; encoding: [0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x44] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:17], v[20:23] cbsz:4 blgp:2 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:17], a[20:23] cbsz:4 blgp:2 ; encoding: [0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x5c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:17], a[20:23] cbsz:4 blgp:2 + + +// fp4 x fp4 +// GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:15], v[20:23] cbsz:4 blgp:4 ; encoding: [0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x84] +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:15], v[20:23] cbsz:4 blgp:4 + +// GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:15], a[20:23] cbsz:4 blgp:4 ; encoding: [0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x9c] +v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:15], a[20:23] cbsz:4 blgp:4 + +//===----------------------------------------------------------------------===// +// v_mfma_f32_32x32x64_f8f6f4 with appropriate register widths +//===----------------------------------------------------------------------===// + +// bf8 x fp4 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[12:15], v[16:31] cbsz:1 blgp:4 ; encoding: [0x00,0x01,0xae,0xd3,0x04,0x19,0x42,0x84] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[12:15], v[16:31] cbsz:1 blgp:4 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[12:15], a[16:31] cbsz:1 blgp:4 ; encoding: [0x00,0x81,0xae,0xd3,0x04,0x19,0x42,0x9c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[12:15], a[16:31] cbsz:1 blgp:4 + + +// fp4 x bf8 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[8:15], v[16:31] cbsz:4 blgp:1 ; encoding: [0x00,0x04,0xae,0xd3,0x04,0x11,0x42,0x24] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[8:15], v[16:31] cbsz:4 blgp:1 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[8:15], a[16:31] cbsz:4 blgp:1 ; encoding: [0x00,0x84,0xae,0xd3,0x04,0x11,0x42,0x3c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[8:15], a[16:31] cbsz:4 blgp:1 + + +// bf6 x bf8 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:19], v[16:31] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x24] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:19], v[16:31] cbsz:3 blgp:1 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:19], a[16:31] cbsz:3 blgp:1 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x3c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:19], a[16:31] cbsz:3 blgp:1 + +// bf8 x bf6 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[12:17], v[16:31] cbsz:1 blgp:3 ; encoding: [0x00,0x01,0xae,0xd3,0x04,0x19,0x42,0x64] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[12:17], v[16:31] cbsz:1 blgp:3 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[12:17], a[16:31] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xae,0xd3,0x04,0x19,0x42,0x7c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[12:17], a[16:31] cbsz:1 blgp:3 + + +// bf6 x bf6 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:3 blgp:3 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x64] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:3 blgp:3 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:3 blgp:3 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x7c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:3 blgp:3 + + +// bf6 x fp6 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:3 blgp:2 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x44] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:3 blgp:2 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:3 blgp:2 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x5c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:3 blgp:2 + + +// fp6 x bf6 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:2 blgp:3 ; encoding: [0x00,0x02,0xae,0xd3,0x04,0x19,0x42,0x64] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:2 blgp:3 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:2 blgp:3 ; encoding: [0x00,0x82,0xae,0xd3,0x04,0x19,0x42,0x7c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:2 blgp:3 + + +// fp6 x fp4 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:15], v[16:31] cbsz:2 blgp:4 ; encoding: [0x00,0x02,0xae,0xd3,0x04,0x19,0x42,0x84] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:15], v[16:31] cbsz:2 blgp:4 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:15], a[16:31] cbsz:2 blgp:4 ; encoding: [0x00,0x82,0xae,0xd3,0x04,0x19,0x42,0x9c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:15], a[16:31] cbsz:2 blgp:4 + + +// fp4 x fp6 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[12:17], v[16:31] cbsz:4 blgp:2 ; encoding: [0x00,0x04,0xae,0xd3,0x04,0x19,0x42,0x44] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[12:17], v[16:31] cbsz:4 blgp:2 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[12:17], a[16:31] cbsz:4 blgp:2 ; encoding: [0x00,0x84,0xae,0xd3,0x04,0x19,0x42,0x5c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[12:17], a[16:31] cbsz:4 blgp:2 + + +// fp4 x fp4 +// GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[12:15], v[16:31] cbsz:4 blgp:4 ; encoding: [0x00,0x04,0xae,0xd3,0x04,0x19,0x42,0x84] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[12:15], v[16:31] cbsz:4 blgp:4 + +// GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[12:15], a[16:31] cbsz:4 blgp:4 ; encoding: [0x00,0x84,0xae,0xd3,0x04,0x19,0x42,0x9c] +v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[12:15], a[16:31] cbsz:4 blgp:4 + +//===----------------------------------------------------------------------===// +// v_mfma_scale_f32_16x16x128_f8f6f4 corrected widths +//===----------------------------------------------------------------------===// + +// fp8 x fp8 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 cbsz:0 blgp:0 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x1c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 cbsz:0 blgp:0 + + +// bf8 x fp8 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x04] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 cbsz:1 blgp:0 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x1c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 cbsz:1 blgp:0 + +// fp8 x bf8 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x24] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 cbsz:0 blgp:1 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x3c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 cbsz:0 blgp:1 + + +// bf8 x fp4 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x84] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23], v24, v25 cbsz:1 blgp:4 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:15], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x9c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:15], a[20:23], v24, v25 cbsz:1 blgp:4 + +// fp4 x bf8 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x24] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:19], v[20:23], v24, v25 cbsz:4 blgp:1 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x3c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:19], a[20:23], v24, v25 cbsz:4 blgp:1 + + +// bf6 x bf8 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 cbsz:3 blgp:1 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x3c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:19], a[20:23], v24, v25 cbsz:3 blgp:1 + + +// bf8 x bf6 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x64] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23], v24, v25 cbsz:1 blgp:3 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x7c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:17], a[20:23], v24, v25 cbsz:1 blgp:3 + + +// bf6 x bf6 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x64] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 cbsz:3 blgp:3 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x7c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 cbsz:3 blgp:3 + + +// bf6 x fp6 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x44] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 cbsz:3 blgp:2 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x5c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 cbsz:3 blgp:2 + + +// fp6 x bf6 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0a,0xad,0xd3,0x04,0x19,0x52,0x64 +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 cbsz:2 blgp:3 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8a,0xad,0xd3,0x04,0x19,0x52,0x7c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 cbsz:2 blgp:3 + + +// fp6 x fp4 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:15], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0a,0xad,0xd3,0x04,0x19,0x52,0x84] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:15], v[20:23], v24, v25 cbsz:2 blgp:4 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:15], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8a,0xad,0xd3,0x04,0x19,0x52,0x9c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:15], a[20:23], v24, v25 cbsz:2 blgp:4 + +// fp4 x fp6 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x44] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:17], v[20:23], v24, v25 cbsz:4 blgp:2 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x5c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:17], a[20:23], v24, v25 cbsz:4 blgp:2 + +// fp4 x fp4 +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:15], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x84] +v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:15], v[20:23], v24, v25 cbsz:4 blgp:4 + +// GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:15], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x9c] +v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:15], a[20:23], v24, v25 cbsz:4 blgp:4 + + +//===----------------------------------------------------------------------===// +// v_mfma_scale_f32_32x32x64_f8f6f4 corrected widths +//===----------------------------------------------------------------------===// + +// fp8 x fp8 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x04] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 cbsz:0 blgp:0 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x1c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 cbsz:0 blgp:0 + +// bf8 x fp8 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x04] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 cbsz:1 blgp:0 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x1c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 cbsz:1 blgp:0 + + +// fp8 x bf8 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x24] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 cbsz:0 blgp:1 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x3c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 cbsz:0 blgp:1 + +// bf8 x fp4 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:27], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x84] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:27], v[32:47], v48, v49 cbsz:1 blgp:4 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:27], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x9c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:27], a[32:47], v48, v49 cbsz:1 blgp:4 + + +// fp4 x bf8 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x24] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:31], v[32:47], v48, v49 cbsz:4 blgp:1 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x3c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:31], a[32:47], v48, v49 cbsz:4 blgp:1 + + +// bf6 x bf8 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x24] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 cbsz:3 blgp:1 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x3c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:31], a[32:47], v48, v49 cbsz:3 blgp:1 + + +// bf8 x bf6 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x64] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:29], v[32:47], v48, v49 cbsz:1 blgp:3 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x7c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:29], a[32:47], v48, v49 cbsz:1 blgp:3 + +// bf6 x bf6 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x64] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 cbsz:3 blgp:3 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x7c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 cbsz:3 blgp:3 + +// bf6 x fp6 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x44] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 cbsz:3 blgp:2 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x5c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 cbsz:3 blgp:2 + + +// fp6 x bf6 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x64] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 cbsz:2 blgp:3 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8a,0xae,0xd3,0x10,0x31,0x82,0x7c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 cbsz:2 blgp:3 + + +// fp6 x fp4 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:27], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x84] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:27], v[32:47], v48, v49 cbsz:2 blgp:4 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:27], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8a,0xae,0xd3,0x10,0x31,0x82,0x9c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:27], a[32:47], v48, v49 cbsz:2 blgp:4 + + +// fp4 x fp6 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x44] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:29], v[32:47], v48, v49 cbsz:4 blgp:2 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x5c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:29], a[32:47], v48, v49 cbsz:4 blgp:2 + + +// fp4 x fp4 +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:27], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x84] +v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:27], v[32:47], v48, v49 cbsz:4 blgp:4 + +// GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:27], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x9c] +v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:27], a[32:47], v48, v49 cbsz:4 blgp:4 + diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx950_mai.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx950_mai.txt index 1fa48fca80fb45..b35789cbf500f6 100644 --- a/llvm/test/MC/Disassembler/AMDGPU/gfx950_mai.txt +++ b/llvm/test/MC/Disassembler/AMDGPU/gfx950_mai.txt @@ -159,3 +159,411 @@ # GFX950: v_mfma_ld_scale_b32 vcc_lo, vcc_lo ; encoding: [0x00,0x40,0xac,0xd3,0x6a,0xd4,0x00,0x18] 0x00,0x40,0xac,0xd3,0x6a,0xd4,0x00,0x18 + +# GFX950: s_nop 0 ; encoding: [0x00,0x00,0x80,0xbf] +0x00,0x00,0x80,0xbf + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:15], a[20:23] cbsz:1 blgp:4 ; encoding: [0x00,0x81,0xad,0xd3,0x04,0x19,0x52,0x9c] +0x00,0x81,0xad,0xd3,0x04,0x19,0x52,0x9c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:17], a[20:23] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xad,0xd3,0x04,0x19,0x52,0x7c] +0x00,0x81,0xad,0xd3,0x04,0x19,0x52,0x7c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] ; encoding: [0x00,0x80,0xad,0xd3,0x04,0x09,0x02,0x1c] +0x00,0x80,0xad,0xd3,0x04,0x09,0x02,0x1c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:11], a[0:3] blgp:1 ; encoding: [0x00,0x80,0xad,0xd3,0x04,0x09,0x02,0x3c] +0x00,0x80,0xad,0xd3,0x04,0x09,0x02,0x3c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[4:9], a[0:3] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xad,0xd3,0x04,0x09,0x02,0x7c] +0x00,0x81,0xad,0xd3,0x04,0x09,0x02,0x7c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:15], a[20:23] cbsz:4 blgp:4 ; encoding: [0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x9c] +0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x9c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:17], a[20:23] cbsz:4 blgp:2 ; encoding: [0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x5c] +0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x5c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:19], a[20:23] cbsz:4 blgp:1 ; encoding: [0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x3c] +0x00,0x84,0xad,0xd3,0x04,0x19,0x52,0x3c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:15], a[20:23] cbsz:2 blgp:4 ; encoding: [0x00,0x82,0xad,0xd3,0x04,0x19,0x52,0x9c] +0x00,0x82,0xad,0xd3,0x04,0x19,0x52,0x9c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:2 blgp:3 ; encoding: [0x00,0x82,0xad,0xd3,0x04,0x19,0x52,0x7c] +0x00,0x82,0xad,0xd3,0x04,0x19,0x52,0x7c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:3 blgp:2 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x5c] +0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x5c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23] cbsz:3 blgp:3 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x7c] +0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x7c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:19], a[20:23] cbsz:3 blgp:1 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x3c] +0x00,0x83,0xad,0xd3,0x04,0x19,0x52,0x3c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[4:11], a[0:3] cbsz:3 ; encoding: [0x00,0x83,0xad,0xd3,0x04,0x09,0x02,0x1c] +0x00,0x83,0xad,0xd3,0x04,0x09,0x02,0x1c + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23] cbsz:1 blgp:4 ; encoding: [0x00,0x01,0xad,0xd3,0x04,0x19,0x52,0x84] +0x00,0x01,0xad,0xd3,0x04,0x19,0x52,0x84 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23] cbsz:1 blgp:3 ; encoding: [0x00,0x01,0xad,0xd3,0x04,0x19,0x52,0x64] +0x00,0x01,0xad,0xd3,0x04,0x19,0x52,0x64 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] ; encoding: [0x00,0x00,0xad,0xd3,0x04,0x09,0x02,0x04] +0x00,0x00,0xad,0xd3,0x04,0x09,0x02,0x04 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] blgp:1 ; encoding: [0x00,0x00,0xad,0xd3,0x04,0x09,0x02,0x24] +0x00,0x00,0xad,0xd3,0x04,0x09,0x02,0x24 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:15], v[20:23] cbsz:4 blgp:4 ; encoding: [0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x84] +0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x84 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:17], v[20:23] cbsz:4 blgp:2 ; encoding: [0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x44] +0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x44 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:19], v[20:23] cbsz:4 blgp:1 ; encoding: [0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x24] +0x00,0x04,0xad,0xd3,0x04,0x19,0x52,0x24 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:15], v[20:23] cbsz:2 blgp:4 ; encoding: [0x00,0x02,0xad,0xd3,0x04,0x19,0x52,0x84] +0x00,0x02,0xad,0xd3,0x04,0x19,0x52,0x84 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:2 blgp:3 ; encoding: [0x00,0x02,0xad,0xd3,0x04,0x19,0x52,0x64] +0x00,0x02,0xad,0xd3,0x04,0x19,0x52,0x64 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:3 blgp:2 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x44] +0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x44 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23] cbsz:3 blgp:3 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x64] +0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x64 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x24] +0x00,0x03,0xad,0xd3,0x04,0x19,0x52,0x24 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[4:11], v[0:3] cbsz:3 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x09,0x02,0x04] +0x00,0x03,0xad,0xd3,0x04,0x09,0x02,0x04 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[4:11], v[0:3] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xad,0xd3,0x04,0x09,0x02,0x24] +0x00,0x03,0xad,0xd3,0x04,0x09,0x02,0x24 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[12:15], a[16:31] cbsz:1 blgp:4 ; encoding: [0x00,0x81,0xae,0xd3,0x04,0x19,0x42,0x9c] +0x00,0x81,0xae,0xd3,0x04,0x19,0x42,0x9c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[12:17], a[16:31] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xae,0xd3,0x04,0x19,0x42,0x7c] +0x00,0x81,0xae,0xd3,0x04,0x19,0x42,0x7c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] ; encoding: [0x00,0x80,0xae,0xd3,0x04,0x09,0x02,0x1c] +0x00,0x80,0xae,0xd3,0x04,0x09,0x02,0x1c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:11], a[0:15] blgp:1 ; encoding: [0x00,0x80,0xae,0xd3,0x04,0x09,0x02,0x3c] +0x00,0x80,0xae,0xd3,0x04,0x09,0x02,0x3c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:11], a[4:9], a[0:15] cbsz:1 blgp:3 ; encoding: [0x00,0x81,0xae,0xd3,0x04,0x09,0x02,0x7c] +0x00,0x81,0xae,0xd3,0x04,0x09,0x02,0x7c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[12:15], a[16:31] cbsz:4 blgp:4 ; encoding: [0x00,0x84,0xae,0xd3,0x04,0x19,0x42,0x9c] +0x00,0x84,0xae,0xd3,0x04,0x19,0x42,0x9c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[12:17], a[16:31] cbsz:4 blgp:2 ; encoding: [0x00,0x84,0xae,0xd3,0x04,0x19,0x42,0x5c] +0x00,0x84,0xae,0xd3,0x04,0x19,0x42,0x5c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:7], a[8:15], a[16:31] cbsz:4 blgp:1 ; encoding: [0x00,0x84,0xae,0xd3,0x04,0x11,0x42,0x3c] +0x00,0x84,0xae,0xd3,0x04,0x11,0x42,0x3c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:15], a[16:31] cbsz:2 blgp:4 ; encoding: [0x00,0x82,0xae,0xd3,0x04,0x19,0x42,0x9c] +0x00,0x82,0xae,0xd3,0x04,0x19,0x42,0x9c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:2 blgp:3 ; encoding: [0x00,0x82,0xae,0xd3,0x04,0x19,0x42,0x7c] +0x00,0x82,0xae,0xd3,0x04,0x19,0x42,0x7c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:3 blgp:2 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x5c] +0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x5c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:17], a[16:31] cbsz:3 blgp:3 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x7c] +0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x7c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[12:19], a[16:31] cbsz:3 blgp:1 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x3c] +0x00,0x83,0xae,0xd3,0x04,0x19,0x42,0x3c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 a[0:15], a[4:9], a[4:11], a[0:15] cbsz:3 ; encoding: [0x00,0x83,0xae,0xd3,0x04,0x09,0x02,0x1c] +0x00,0x83,0xae,0xd3,0x04,0x09,0x02,0x1c + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[12:15], v[16:31] cbsz:1 blgp:4 ; encoding: [0x00,0x01,0xae,0xd3,0x04,0x19,0x42,0x84] +0x00,0x01,0xae,0xd3,0x04,0x19,0x42,0x84 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[12:17], v[16:31] cbsz:1 blgp:3 ; encoding: [0x00,0x01,0xae,0xd3,0x04,0x19,0x42,0x64] +0x00,0x01,0xae,0xd3,0x04,0x19,0x42,0x64 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] ; encoding: [0x00,0x00,0xae,0xd3,0x04,0x09,0x02,0x04] +0x00,0x00,0xae,0xd3,0x04,0x09,0x02,0x04 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] blgp:1 ; encoding: [0x00,0x00,0xae,0xd3,0x04,0x09,0x02,0x24] +0x00,0x00,0xae,0xd3,0x04,0x09,0x02,0x24 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[12:15], v[16:31] cbsz:4 blgp:4 ; encoding: [0x00,0x04,0xae,0xd3,0x04,0x19,0x42,0x84] +0x00,0x04,0xae,0xd3,0x04,0x19,0x42,0x84 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[12:17], v[16:31] cbsz:4 blgp:2 ; encoding: [0x00,0x04,0xae,0xd3,0x04,0x19,0x42,0x44] +0x00,0x04,0xae,0xd3,0x04,0x19,0x42,0x44 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:7], v[8:15], v[16:31] cbsz:4 blgp:1 ; encoding: [0x00,0x04,0xae,0xd3,0x04,0x11,0x42,0x24] +0x00,0x04,0xae,0xd3,0x04,0x11,0x42,0x24 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:15], v[16:31] cbsz:2 blgp:4 ; encoding: [0x00,0x02,0xae,0xd3,0x04,0x19,0x42,0x84] +0x00,0x02,0xae,0xd3,0x04,0x19,0x42,0x84 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:2 blgp:3 ; encoding: [0x00,0x02,0xae,0xd3,0x04,0x19,0x42,0x64] +0x00,0x02,0xae,0xd3,0x04,0x19,0x42,0x64 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:3 blgp:2 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x44] +0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x44 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:17], v[16:31] cbsz:3 blgp:3 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x64] +0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x64 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[12:19], v[16:31] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x24] +0x00,0x03,0xae,0xd3,0x04,0x19,0x42,0x24 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[4:11], v[0:15] cbsz:3 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x09,0x02,0x04] +0x00,0x03,0xae,0xd3,0x04,0x09,0x02,0x04 + +# GFX950: v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:9], v[4:11], v[0:15] cbsz:3 blgp:1 ; encoding: [0x00,0x03,0xae,0xd3,0x04,0x09,0x02,0x24] +0x00,0x03,0xae,0xd3,0x04,0x09,0x02,0x24 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:15], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x9c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x9c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x7c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x7c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x1c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x1c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x3c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x3c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:11], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x1c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x89,0xad,0xd3,0x04,0x19,0x52,0x1c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:15], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x9c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x9c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x5c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x5c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:7], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x3c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8c,0xad,0xd3,0x04,0x19,0x52,0x3c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:15], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8a,0xad,0xd3,0x04,0x19,0x52,0x9c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8a,0xad,0xd3,0x04,0x19,0x52,0x9c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8a,0xad,0xd3,0x04,0x19,0x52,0x7c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8a,0xad,0xd3,0x04,0x19,0x52,0x7c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x5c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x5c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:17], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x7c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x7c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], a[4:9], a[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x3c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x8b,0xad,0xd3,0x04,0x19,0x52,0x3c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[4:11], v[12:19], a[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x88,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], a[4:11], a[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x1c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x1c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], a[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x0c] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x0c + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], a[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x14] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x14 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:15], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x84] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x84 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x44] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x44 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x64] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x64 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], 33, 9 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0xa1,0x12,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0xa1,0x12,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], 9, v2 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x89,0x04,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x89,0x04,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], m0, m0 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x7c,0xf8,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x7c,0xf8,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s20, 9 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x14,0x12,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x14,0x12,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s24, s24 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x30,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x30,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], s24, v44 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x58,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x58,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v2, 9 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x02,0x13,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x02,0x13,0x01,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x24] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x24 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x09,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], v44, s24 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x2c,0x31,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x2c,0x31,0x00,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[12:19], v[20:23], vcc_lo, v2 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x6a,0x04,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x6a,0x04,0x02,0x00,0x00,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:15], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x84] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x84 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x44] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x44 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:7], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x24] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0c,0xad,0xd3,0x04,0x19,0x52,0x24 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:15], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0a,0xad,0xd3,0x04,0x19,0x52,0x84] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0a,0xad,0xd3,0x04,0x19,0x52,0x84 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0a,0xad,0xd3,0x04,0x19,0x52,0x64] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0a,0xad,0xd3,0x04,0x19,0x52,0x64 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x44] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x44 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:17], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x64] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x64 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x08,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x08,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[0:3], v[4:9], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x18,0x00,0x0b,0xad,0xd3,0x04,0x19,0x52,0x24 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 v[50:53], v[4:11], v[12:19], v[20:23], v24, v25 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x32,0x08,0xad,0xd3,0x04,0x19,0x52,0x04] +0x00,0x00,0xac,0xd3,0x18,0x33,0x02,0x00,0x32,0x08,0xad,0xd3,0x04,0x19,0x52,0x04 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:27], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x9c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x9c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x5c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x5c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:19], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x3c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8c,0xae,0xd3,0x10,0x31,0x82,0x3c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:27], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8a,0xae,0xd3,0x10,0x31,0x82,0x9c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8a,0xae,0xd3,0x10,0x31,0x82,0x9c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8a,0xae,0xd3,0x10,0x31,0x82,0x7c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8a,0xae,0xd3,0x10,0x31,0x82,0x7c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x5c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x5c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x7c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x7c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:21], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x3c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x8b,0xae,0xd3,0x10,0x31,0x82,0x3c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:27], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x9c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x9c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:29], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x7c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x7c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x1c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x1c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x3c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x3c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], a[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x1c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x89,0xae,0xd3,0x10,0x31,0x82,0x1c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], a[16:23], v[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x0c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x0c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 a[0:15], v[16:23], v[24:31], a[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x04] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x88,0xae,0xd3,0x10,0x31,0x82,0x04 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], a[16:23], a[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x1c] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x1c + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:27], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x84] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x84 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x44] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x44 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:19], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:4 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x24] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0c,0xae,0xd3,0x10,0x31,0x82,0x24 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:27], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x84] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x84 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x64] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x64 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x18,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x64] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x18,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x64 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x44] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x44 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x64] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x64 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x04] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0a,0xae,0xd3,0x10,0x31,0x82,0x04 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:21], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:3 blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x24] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x0b,0xae,0xd3,0x10,0x31,0x82,0x24 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], a[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x14] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x14 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:27], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:4 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x84] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x84 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] blgp:2 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x44] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x44 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:29], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 blgp:3 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x64] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x64 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x04] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x04 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] blgp:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x24] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x08,0xae,0xd3,0x10,0x31,0x82,0x24 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[0:15], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] cbsz:1 ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x04] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x00,0x09,0xae,0xd3,0x10,0x31,0x82,0x04 + +# GFX950: v_mfma_scale_f32_32x32x64_f8f6f4 v[50:65], v[16:23], v[24:31], v[32:47], v48, v49 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x32,0x08,0xae,0xd3,0x10,0x31,0x82,0x04] +0x00,0x00,0xac,0xd3,0x30,0x63,0x02,0x00,0x32,0x08,0xae,0xd3,0x10,0x31,0x82,0x04 diff --git a/llvm/test/MC/Disassembler/AMDGPU/gfx950_vop3px2.txt b/llvm/test/MC/Disassembler/AMDGPU/gfx950_vop3px2.txt new file mode 100644 index 00000000000000..3227b619de4fdd --- /dev/null +++ b/llvm/test/MC/Disassembler/AMDGPU/gfx950_vop3px2.txt @@ -0,0 +1,28 @@ +# RUN: llvm-mc -triple=amdgcn -mcpu=gfx950 -show-encoding -disassemble %s | FileCheck -check-prefix=GFX950 %s + +# Check behavior of truncated instruction just to the ld_scale + +# GFX950: v_mfma_ld_scale_b32 v20, v21 op_sel_hi:[0,0] ; encoding: [0x00,0x40,0xac,0xd3,0x14,0x2b,0x02,0x00] +0x00,0x00,0xac,0xd3,0x14,0x2b,0x02,0x00 + +# GFX950: v_mfma_ld_scale_b32 v20, v21 op_sel_hi:[0,0] ; encoding: [0x00,0x40,0xac,0xd3,0x14,0x2b,0x02,0x00] +0x00,0x00,0xac,0xd3,0x14,0x2b,0x02,0x00 + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3] ; encoding: [0x00,0x80,0xad,0xd3,0x00,0x11,0x02,0x04] +0x00,0x00,0xac,0xd3,0x14,0x2b,0x02,0x00,0x00,0x80,0xad,0xd3,0x00,0x11,0x02,0x04 + +# GFX950: v_mfma_ld_scale_b32 v20, v21 op_sel_hi:[0,0] ; encoding: [0x00,0x40,0xac,0xd3,0x14,0x2b,0x02,0x00] +0x00,0x00,0xac,0xd3,0x14,0x2b,0x02,0x00 + +# GFX950: s_nop 0 ; encoding: [0x00,0x00,0x80,0xbf] +0x00,0x00,0x80,0xbf + +# GFX950: v_mfma_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3] ; encoding: [0x00,0x80,0xad,0xd3,0x00,0x11,0x02,0x04] +0x00,0x80,0xad,0xd3,0x00,0x11,0x02,0x04 + + +# GFX950: v_mfma_ld_scale_b32 v20, v21 op_sel_hi:[0,0] ; encoding: [0x00,0x40,0xac,0xd3,0x14,0x2b,0x02,0x00] +0x00,0x00,0xac,0xd3,0x14,0x2b,0x02,0x00 + +# GFX950: v_mfma_scale_f32_16x16x128_f8f6f4 a[0:3], v[0:7], v[8:15], a[0:3], v20, v21 op_sel_hi:[0,0,0] ; encoding: [0x00,0x00,0xac,0xd3,0x14,0x2b,0x02,0x00,0x00,0x88,0xad,0xd3,0x00,0x11,0x02,0x04] +0x00,0x00,0xac,0xd3,0x14,0x2b,0x02,0x00,0x00,0x88,0xad,0xd3,0x00,0x11,0x02,0x04 diff --git a/llvm/test/tools/llvm-mca/AMDGPU/gfx950.s b/llvm/test/tools/llvm-mca/AMDGPU/gfx950.s index e601de8d706b44..3ab3319c21ecdf 100644 --- a/llvm/test/tools/llvm-mca/AMDGPU/gfx950.s +++ b/llvm/test/tools/llvm-mca/AMDGPU/gfx950.s @@ -1,11 +1,13 @@ # RUN: llvm-mca -mtriple=amdgcn -mcpu=gfx950 --timeline --iterations=1 --timeline-max-cycles=0 < %s | FileCheck %s # CHECK: Iterations: 1 -# CHECK: Instructions: 7 -# CHECK: Total Cycles: 42 -# CHECK: Total uOps: 7 +# CHECK: Instructions: 9 +# CHECK: Total Cycles: 44 +# CHECK: Total uOps: 9 v_mfma_ld_scale_b32 v0, v0 +v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] +v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] v_mfma_f32_16x16x32_f16 a[0:3], a[0:3], a[0:3], a[0:3] blgp:1 v_mfma_f32_16x16x32_f16 a[0:3], v[0:3], v[0:3], a[4:7] @@ -14,9 +16,10 @@ v_mfma_f32_32x32x16_f16 a[0:15], a[0:3], a[0:3], a[0:15] blgp:2 v_mfma_f32_32x32x16_bf16 v[0:15], v[0:3], v[0:3], v[0:15] v_mfma_f32_32x32x16_bf16 a[0:15], a[0:3], a[0:3], a[0:15] blgp:2 - # CHECK: [0] [1] [2] [3] [4] [5] [6] Instructions: # CHECK-NEXT: - - - - 1.00 - - v_mfma_ld_scale_b32 v0, v0 +# CHECK-NEXT: - - - - 1.00 - - v_mfma_f32_16x16x128_f8f6f4 v[0:3], v[4:11], v[4:11], v[0:3] +# CHECK-NEXT: - - - - 1.00 - - v_mfma_f32_32x32x64_f8f6f4 v[0:15], v[4:11], v[4:11], v[0:15] # CHECK-NEXT: - - - - - - 4.00 v_mfma_f32_16x16x32_f16 a[0:3], a[0:3], a[0:3], a[0:3] blgp:1 # CHECK-NEXT: - - - - - - 4.00 v_mfma_f32_16x16x32_f16 a[0:3], v[0:3], v[0:3], a[4:7] # CHECK-NEXT: - - - - - - 8.00 v_mfma_f32_32x32x16_f16 v[0:15], v[0:3], v[0:3], v[0:15]