Skip to content

[func][spirv] Add necessary patch, pass-pipeline, and test case for n… #725

New issue

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

Sign up for GitHub

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

Already on GitHub? Sign in to your account

Merged
merged 1 commit into from
Apr 18, 2024
Merged

[func][spirv] Add necessary patch, pass-pipeline, and test case for n… #725

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

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
63 changes: 63 additions & 0 deletions build_tools/patches/0008-Add-memref.extract_aligned_pointer_as_index-to-spirv.patch
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,63 @@
Upstream PR in progress: https://github.com/llvm/llvm-project/pull/86750

diff --git a/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp b/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp
index 57d8e894a24b..3fc68c65de05 100644
--- a/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp
+++ b/mlir/lib/Conversion/MemRefToSPIRV/MemRefToSPIRV.cpp
@@ -306,6 +306,18 @@ public:
}
};

+/// Converts memref.extract_aligned_pointer_as_index to spirv.ConvertPtrToU.
+class ExtractAlignedPointerAsIndexOpPattern
+ : public OpConversionPattern<memref::ExtractAlignedPointerAsIndexOp> {
+public:
+ using OpConversionPattern::OpConversionPattern;
+
+ LogicalResult
+ matchAndRewrite(memref::ExtractAlignedPointerAsIndexOp extractOp,
+ OpAdaptor adaptor,
+ ConversionPatternRewriter &rewriter) const override;
+};
+
} // namespace

//===----------------------------------------------------------------------===//
@@ -905,6 +917,20 @@ LogicalResult ReinterpretCastPattern::matchAndRewrite(
return success();
}

+//===----------------------------------------------------------------------===//
+// ExtractAlignedPointerAsIndexOp
+//===----------------------------------------------------------------------===//
+
+LogicalResult ExtractAlignedPointerAsIndexOpPattern::matchAndRewrite(
+ memref::ExtractAlignedPointerAsIndexOp extractOp, OpAdaptor adaptor,
+ ConversionPatternRewriter &rewriter) const {
+ auto &typeConverter = *getTypeConverter<SPIRVTypeConverter>();
+ Type indexType = typeConverter.getIndexType();
+ rewriter.replaceOpWithNewOp<spirv::ConvertPtrToUOp>(extractOp, indexType,
+ adaptor.getSource());
+ return success();
+}
+
//===----------------------------------------------------------------------===//
// Pattern population
//===----------------------------------------------------------------------===//
@@ -912,10 +938,11 @@ LogicalResult ReinterpretCastPattern::matchAndRewrite(
namespace mlir {
void populateMemRefToSPIRVPatterns(SPIRVTypeConverter &typeConverter,
RewritePatternSet &patterns) {
- patterns.add<AllocaOpPattern, AllocOpPattern, AtomicRMWOpPattern,
- DeallocOpPattern, IntLoadOpPattern, IntStoreOpPattern,
- LoadOpPattern, MemorySpaceCastOpPattern, StoreOpPattern,
- ReinterpretCastPattern, CastPattern>(typeConverter,
- patterns.getContext());
+ patterns
+ .add<AllocaOpPattern, AllocOpPattern, AtomicRMWOpPattern,
+ DeallocOpPattern, IntLoadOpPattern, IntStoreOpPattern, LoadOpPattern,
+ MemorySpaceCastOpPattern, StoreOpPattern, ReinterpretCastPattern,
+ CastPattern, ExtractAlignedPointerAsIndexOpPattern>(
+ typeConverter, patterns.getContext());
}
} // namespace mlir
28 changes: 28 additions & 0 deletions test/Integration/Dialect/Func/func-to-llvm.pp
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,28 @@
// gpu dialect with intel intrinsic functions (func dialect) to
// llvm dialect (for host code) and
// spirv dialect (for device code) lowering pipeline.
// Ready for imex runner starting from GPU dialect.
builtin.module(
gpu.module(convert-func-to-spirv)
gpu.module(convert-vector-to-spirv)
imex-convert-gpu-to-spirv{enable-vc-intrinsic=true}
spirv.module(spirv-lower-abi-attrs
spirv-update-vce)
func.func(llvm-request-c-wrappers)
serialize-spirv
convert-vector-to-scf
convert-gpu-to-gpux
convert-scf-to-cf
convert-cf-to-llvm
convert-vector-to-llvm
convert-index-to-llvm
convert-arith-to-llvm
convert-func-to-llvm
convert-math-to-llvm
convert-gpux-to-llvm
convert-index-to-llvm
expand-strided-metadata
lower-affine
finalize-memref-to-llvm
reconcile-unrealized-casts)
// End
132 changes: 132 additions & 0 deletions test/Integration/Dialect/Func/gemm-with-func-and-intel-intrinsic.mlir
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,132 @@
// RUN: %python_executable %imex_runner --requires=l0-runtime -i %s --pass-pipeline-file=%p/func-to-llvm.pp \
// RUN: --runner imex-cpu-runner -e main \
// RUN: --entry-point-result=void \
// RUN: --shared-libs=%irunner_utils,%mlir_runner_utils,%mlir_c_runner_utils,%levelzero_runtime --filecheck
// RUN: %python_executable %imex_runner --requires=sycl-runtime -i %s --pass-pipeline-file=%p/func-to-llvm.pp \
// RUN: --runner imex-cpu-runner -e main \
// RUN: --entry-point-result=void \
// RUN: --shared-libs=%irunner_utils,%mlir_runner_utils,%mlir_c_runner_utils,%sycl_runtime --filecheck

module @gemm attributes {gpu.container_module,
spirv.target_env = #spirv.target_env<#spirv.vce<v1.4, [Addresses, Float16Buffer, Int64, Int16, Int8, Kernel, Linkage, Vector16, GenericPointer, Groups, Float16, Float64, AtomicFloat32AddEXT, ExpectAssumeKHR, SubgroupDispatch, VectorComputeINTEL, VectorAnyINTEL], [SPV_EXT_shader_atomic_float_add, SPV_KHR_expect_assume, SPV_INTEL_vector_compute]>, api=OpenCL, #spirv.resource_limits<>>} {
memref.global "private" constant @__constant_8x16xf16 : memref<8x16xf16> = dense<5.000000e-01>
memref.global "private" constant @__constant_16x16xf16 : memref<16x16xf16> = dense<1.099610e+00>
func.func @test(%arg0: memref<8x16xf16>, %arg1: memref<16x16xf16>) -> memref<8x16xf32> attributes {llvm.emit_c_interface} {
%c1 = arith.constant 1 : index
%memref = gpu.alloc host_shared () : memref<8x16xf16>
memref.copy %arg0, %memref : memref<8x16xf16> to memref<8x16xf16>
%memref_0 = gpu.alloc host_shared () : memref<16x16xf16>
memref.copy %arg1, %memref_0 : memref<16x16xf16> to memref<16x16xf16>
%memref_1 = gpu.alloc host_shared () : memref<8x16xf32>
gpu.launch_func @test_kernel::@test_kernel blocks in (%c1, %c1, %c1) threads in (%c1, %c1, %c1) args(%memref : memref<8x16xf16>, %memref_0 : memref<16x16xf16>, %memref_1 : memref<8x16xf32>)
gpu.dealloc %memref : memref<8x16xf16>
gpu.dealloc %memref_0 : memref<16x16xf16>
return %memref_1 : memref<8x16xf32>
}

gpu.module @test_kernel {
func.func private @llvm.genx.raw.sends2.noresult.i1.v8i32.v64i64(i8, i8, i1, i8, i8, i8, i32, i32, vector<8xi32>, vector<64xi64>) attributes{
linkage_attributes=#spirv.linkage_attributes<
linkage_name="llvm.genx.raw.sends2.noresult.i1.v8i32.v64i64",
linkage_type=<Import>
>,
VectorComputeFunctionINTEL}
func.func private @llvm.genx.dpas.nosrc0.v128f32.v128i32.v64i32(vector<128xi32>, vector<64xi32>, i32) -> vector<128xf32> attributes{
linkage_attributes=#spirv.linkage_attributes<
linkage_name="llvm.genx.dpas.nosrc0.v128f32.v128i32.v64i32",
linkage_type=<Import>
>,
VectorComputeFunctionINTEL}
func.func private @llvm.genx.raw.send2.v128i32.i1.v8i32(i8, i8, i1, i8, i8, i8, i32, i32, vector<8xi32>, vector<128xi32>) -> vector<128xi32> attributes{
linkage_attributes=#spirv.linkage_attributes<
linkage_name="llvm.genx.raw.send2.v128i32.i1.v8i32",
linkage_type=<Import>
>,
VectorComputeFunctionINTEL}
func.func private @llvm.genx.raw.send2.v32i64.i1.v8i32(i8, i8, i1, i8, i8, i8, i32, i32, vector<8xi32>, vector<32xi64>) -> vector<32xi64> attributes{
linkage_attributes=#spirv.linkage_attributes<
linkage_name="llvm.genx.raw.send2.v32i64.i1.v8i32",
linkage_type=<Import>
>,
VectorComputeFunctionINTEL}
gpu.func @test_kernel(%arg0: memref<8x16xf16>, %arg1: memref<16x16xf16>, %arg2: memref<8x16xf32>) kernel attributes {VectorComputeFunctionINTEL,spirv.entry_point_abi = #spirv.entry_point_abi<>} {
%reinterpret_cast = memref.reinterpret_cast %arg0 to offset: [0], sizes: [128], strides: [1] : memref<8x16xf16> to memref<128xf16>
%cst = arith.constant dense<0> : vector<4xi64>
%intptr = memref.extract_aligned_pointer_as_index %reinterpret_cast : memref<128xf16> -> index
%0 = arith.index_castui %intptr : index to i64
%1 = vector.insert %0, %cst[0] : i64 into vector<4xi64>
%2 = vector.bitcast %1 : vector<4xi64> to vector<8xi32>
%cst_0 = arith.constant dense<0> : vector<4xi64>
%cst_to_non_cst = arith.addi %cst_0, %cst_0 : vector<4xi64>
%intptr_1 = memref.extract_aligned_pointer_as_index %arg1 : memref<16x16xf16> -> index
%3 = arith.index_castui %intptr_1 : index to i64
%4 = vector.insert %3, %cst_to_non_cst[0] : i64 into vector<4xi64>
%5 = vector.bitcast %4 : vector<4xi64> to vector<8xi32>
%c31_i32 = arith.constant 31 : i32
%c15_i32 = arith.constant 15 : i32
%c31_i32_2 = arith.constant 31 : i32
%6 = vector.insert %c31_i32, %5 [2] : i32 into vector<8xi32>
%7 = vector.insert %c15_i32, %6 [3] : i32 into vector<8xi32>
%8 = vector.insert %c31_i32_2, %7 [4] : i32 into vector<8xi32>
%c0_i32 = arith.constant 0 : i32
%c0_i32_3 = arith.constant 0 : i32
%9 = vector.insert %c0_i32, %8 [5] : i32 into vector<8xi32>
%10 = vector.insert %c0_i32_3, %9 [6] : i32 into vector<8xi32>
%c3855_i32 = arith.constant 3855 : i32
%11 = vector.insert %c3855_i32, %10 [7] : i32 into vector<8xi32>
%reinterpret_cast_4 = memref.reinterpret_cast %arg2 to offset: [0], sizes: [128], strides: [1] : memref<8x16xf32> to memref<128xf32>
%cst_5_t = arith.constant dense<0> : vector<4xi64>
%cst_5 = arith.addi %cst_5_t, %cst_5_t : vector<4xi64>
%intptr_6 = memref.extract_aligned_pointer_as_index %reinterpret_cast_4 : memref<128xf32> -> index
%12 = arith.index_castui %intptr_6 : index to i64
%13 = vector.insert %12, %cst_5 [0] : i64 into vector<4xi64>
%14 = vector.bitcast %13 : vector<4xi64> to vector<8xi32>
%c0_i8 = arith.constant 0 : i8
%c0_i8_7 = arith.constant 0 : i8
%true = arith.constant true
%c1_i8 = arith.constant 1 : i8
%c4_i8 = arith.constant 4 : i8
%c15_i8 = arith.constant 15 : i8
%c0_i32_8 = arith.constant 0 : i32
%c42133376_i32 = arith.constant 42133376 : i32
%cst_9 = arith.constant dense<0> : vector<32xi64>
%15 = func.call @llvm.genx.raw.send2.v32i64.i1.v8i32(%c0_i8, %c0_i8_7, %true, %c1_i8, %c4_i8, %c15_i8, %c0_i32_8, %c42133376_i32, %2, %cst_9) : (i8, i8, i1, i8, i8, i8, i32, i32, vector<8xi32>, vector<32xi64>) -> vector<32xi64>
%16 = vector.bitcast %15 : vector<32xi64> to vector<128xf16>
%c0_i8_10 = arith.constant 0 : i8
%c0_i8_11 = arith.constant 0 : i8
%true_12 = arith.constant true
%c1_i8_13 = arith.constant 1 : i8
%c8_i8 = arith.constant 8 : i8
%c15_i8_14 = arith.constant 15 : i8
%c0_i32_15 = arith.constant 0 : i32
%c42074755_i32 = arith.constant 42074755 : i32
%cst_16 = arith.constant dense<0> : vector<128xi32>
%17 = func.call @llvm.genx.raw.send2.v128i32.i1.v8i32(%c0_i8_10, %c0_i8_11, %true_12, %c1_i8_13, %c8_i8, %c15_i8_14, %c0_i32_15, %c42074755_i32, %11, %cst_16) : (i8, i8, i1, i8, i8, i8, i32, i32, vector<8xi32>, vector<128xi32>) -> vector<128xi32>
%18 = vector.bitcast %16 : vector<128xf16> to vector<64xi32>
%c134744586_i32 = arith.constant 134744586 : i32
%19 = func.call @llvm.genx.dpas.nosrc0.v128f32.v128i32.v64i32(%17, %18, %c134744586_i32) : (vector<128xi32>, vector<64xi32>, i32) -> vector<128xf32>
%c0_i8_17 = arith.constant 0 : i8
%c0_i8_18 = arith.constant 0 : i8
%true_19 = arith.constant true
%c1_i8_20 = arith.constant 1 : i8
%c8_i8_21 = arith.constant 8 : i8
%c15_i8_22 = arith.constant 15 : i8
%c0_i32_23 = arith.constant 0 : i32
%c33748868_i32 = arith.constant 33748868 : i32
%20 = vector.bitcast %19 : vector<128xf32> to vector<64xi64>
func.call @llvm.genx.raw.sends2.noresult.i1.v8i32.v64i64(%c0_i8_17, %c0_i8_18, %true_19, %c1_i8_20, %c8_i8_21, %c15_i8_22, %c0_i32_23, %c33748868_i32, %14, %20) : (i8, i8, i1, i8, i8, i8, i32, i32, vector<8xi32>, vector<64xi64>) -> ()
gpu.return
}
}
func.func @main() attributes {llvm.emit_c_interface} {
%0 = memref.get_global @__constant_8x16xf16 : memref<8x16xf16>
%1 = memref.get_global @__constant_16x16xf16 : memref<16x16xf16>
%2 = call @test(%0, %1) : (memref<8x16xf16>, memref<16x16xf16>) -> memref<8x16xf32>
%cast = memref.cast %2 : memref<8x16xf32> to memref<*xf32>
call @printMemrefF32(%cast) : (memref<*xf32>) -> ()
// CHECK: Unranked Memref base@ = {{(0x)?[-9a-f]*}}
// CHECK-COUNT-128: 8.79688
return
}
func.func private @printMemrefF32(memref<*xf32>) attributes {llvm.emit_c_interface}
}
Loading