gaspersic/triton
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Fix

Browse Source
This commit is contained in:
Jokeren
2022-12-06 13:24:50 -08:00
parent 74f3d7a80f
commit bab7338965
2 changed files with 10 additions and 47 deletions
Download Patch File Download Diff File Expand all files Collapse all files

6
lib/Conversion/TritonGPUToLLVM/TritonGPUToLLVM.cpp
View File

@@ -4105,10 +4105,12 @@ struct InsertSliceAsyncOpConversion
}
v = bitcast(v, IntegerType::get(getContext(), byteWidth));
// Write shared memory if predicate is true
auto *valOperand = ptxBuilder.newOperand(v, "r");
auto &st = *ptxBuilder.create<PTXInstr>("st");
PTXBuilder ptxStoreBuilder;
auto *valOperand = ptxStoreBuilder.newOperand(v, "r");
auto &st = *ptxStoreBuilder.create<PTXInstr>("st");
st.shared().o("b" + std::to_string(bitWidth));
st(dstOperand, valOperand).predicate(pred);
ptxStoreBuilder.launch(rewriter, loc, void_ty(getContext()));
}
}
copyAsyncOp(dstOperand, srcOperand, copySize, srcSize);

51
test/Conversion/tritongpu_to_llvm.mlir
View File

@@ -504,11 +504,13 @@ module attributes {"triton_gpu.num-warps" = 4 : i32} {
%other_tensor = tt.splat %other : (f32) -> tensor<16x64xf32, #AL>
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: st.shared.b128 [ ${{.*}} + 0 ]
// CHECK-SAME: cp.async.cg.shared.global [ ${{.*}} + 0 ], [ ${{.*}} + 0 ], 0x10, 0x10
// CHECK-SAME: @${{.*}} st.shared.b128 [ ${{.*}} + 0 ], ${{.*}}
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: st.shared.b128 [ ${{.*}} + 16 ]
// CHECK-SAME: cp.async.cg.shared.global [ ${{.*}} + 16 ], [ ${{.*}} + 0 ], 0x10, 0x10
// CHECK-SAME: cp.async.cg.shared.global [ ${{.*}} + 0 ], [ ${{.*}} + 0 ], 0x10, ${{.*}}
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: @${{.*}} st.shared.b128 [ ${{.*}} + 16 ], ${{.*}}
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: cp.async.cg.shared.global [ ${{.*}} + 16 ], [ ${{.*}} + 0 ], 0x10, ${{.*}}
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: cp.async.commit_group
%a = triton_gpu.insert_slice_async %a_ptr, %tensor, %index, %true_tensor, %other_tensor {axis = 0 : i32, cache = 1 : i32, evict = 1 : i32, isVolatile = false} : tensor<16x64x!tt.ptr<f32>, #AL> -> tensor<2x16x64xf32, #A>
@@ -518,47 +520,6 @@ module attributes {"triton_gpu.num-warps" = 4 : i32} {
// -----
#block0 = #triton_gpu.blocked<{sizePerThread = [1], threadsPerWarp = [4], warpsPerCTA = [4], order = [0]}>
#block1 = #triton_gpu.blocked<{sizePerThread = [1], threadsPerWarp = [8], warpsPerCTA = [4], order = [0]}>
#block2 = #triton_gpu.blocked<{sizePerThread = [1, 1], threadsPerWarp = [4, 1], warpsPerCTA = [4, 1], order = [1, 0]}>
#block3 = #triton_gpu.blocked<{sizePerThread = [1, 1], threadsPerWarp = [1, 8], warpsPerCTA = [1, 4], order = [1, 0]}>
#slice2d1 = #triton_gpu.slice<{dim = 1, parent=#block2}>
#slice3d0 = #triton_gpu.slice<{dim = 0, parent=#block3}>
#AL = #triton_gpu.blocked<{sizePerThread = [1, 8], threadsPerWarp = [4, 8], warpsPerCTA = [4, 1], order = [1, 0]}>
#A = #triton_gpu.shared<{vec = 8, perPhase = 1, maxPhase = 4, order = [1, 0]}>
module attributes {"triton_gpu.num-warps" = 4 : i32} {
// CHECK-LABEL: basic_insert_slice_async_v4
func @basic_insert_slice_async_v4(%arg0: !tt.ptr<f32> {tt.divisibility = 8 : i32}) {
%off0_ = tt.make_range {end = 16 : i32, start = 0 : i32} : tensor<16xi32, #slice2d1>
%off1_ = tt.make_range {end = 32 : i32, start = 0 : i32} : tensor<64xi32, #slice3d0>
%off0 = tt.expand_dims %off0_ {axis = 1 : i32} : (tensor<16xi32, #slice2d1>) -> tensor<16x1xi32, #block2>
%off1 = tt.expand_dims %off1_ {axis = 0 : i32} : (tensor<64xi32, #slice3d0>) -> tensor<1x64xi32, #block3>
%broadcast_off0_scalar = tt.broadcast %off0 : (tensor<16x1xi32, #block2>) -> tensor<16x64xi32, #block2>
%cst_scalar = arith.constant 64 : i32
%cst = tt.splat %cst_scalar : (i32) -> tensor<16x64xi32, #block2>
%broadcast_off0_ = arith.muli %broadcast_off0_scalar, %cst : tensor<16x64xi32, #block2>
%broadcast_off1_ = tt.broadcast %off1 : (tensor<1x64xi32, #block3>) -> tensor<16x64xi32, #block3>
%broadcast_off0 = triton_gpu.convert_layout %broadcast_off0_ : (tensor<16x64xi32, #block2>) -> tensor<16x64xi32, #AL>
%broadcast_off1 = triton_gpu.convert_layout %broadcast_off1_ : (tensor<16x64xi32, #block3>) -> tensor<16x64xi32, #AL>
%off = arith.addi %broadcast_off0, %broadcast_off1 : tensor<16x64xi32, #AL>
%a_init = tt.splat %arg0 : (!tt.ptr<f32>) -> tensor<16x64x!tt.ptr<f32>, #AL>
%a_ptr = tt.addptr %a_init, %off : tensor<16x64x!tt.ptr<f32>, #AL>
%tensor = triton_gpu.alloc_tensor : tensor<2x16x64xf32, #A>
%index = arith.constant 1 : i32
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: cp.async.cg.shared.global [ ${{.*}} + 0 ], [ ${{.*}} + 0 ], 0x10, 0x10
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: cp.async.cg.shared.global [ ${{.*}} + 16 ], [ ${{.*}} + 0 ], 0x10, 0x10
// CHECK: llvm.inline_asm has_side_effects asm_dialect = att
// CHECK-SAME: cp.async.commit_group
%a = triton_gpu.insert_slice_async %a_ptr, %tensor, %index {axis = 0 : i32, cache = 1 : i32, evict = 1 : i32, isVolatile = false} : tensor<16x64x!tt.ptr<f32>, #AL> -> tensor<2x16x64xf32, #A>
return
}
}
// -----
#block0 = #triton_gpu.blocked<{sizePerThread = [1], threadsPerWarp = [4], warpsPerCTA = [4], order = [0]}>
#block1 = #triton_gpu.blocked<{sizePerThread = [1], threadsPerWarp = [8], warpsPerCTA = [4], order = [0]}>
#block2 = #triton_gpu.blocked<{sizePerThread = [1, 1], threadsPerWarp = [4, 1], warpsPerCTA = [4, 1], order = [1, 0]}>