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

[OPTIMIZER] Fixed memory coalescing (#847)

Browse Source
This commit is contained in:
Philippe Tillet
2022-11-07 06:22:18 -08:00
committed by GitHub
parent b6f15e214b
commit 976cf12af1
7 changed files with 25 additions and 29 deletions
Download Patch File Download Diff File Expand all files Collapse all files

4
test/TritonGPU/coalesce.mlir
View File

@@ -9,8 +9,8 @@
module attributes {"triton_gpu.num-warps" = 4 : i32} {
// CHECK: [[row_layout:#.*]] = #triton_gpu.blocked<{sizePerThread = [1, 4], threadsPerWarp = [8, 4], warpsPerCTA = [1, 4], order = [1, 0]}>
// CHECK: [[col_layout:#.*]] = #triton_gpu.blocked<{sizePerThread = [4, 1], threadsPerWarp = [4, 8], warpsPerCTA = [4, 1], order = [0, 1]}>
// CHECK: [[row_layout:#.*]] = #triton_gpu.blocked<{sizePerThread = [1, 4], threadsPerWarp = [2, 16], warpsPerCTA = [4, 1], order = [1, 0]}>
// CHECK: [[col_layout:#.*]] = #triton_gpu.blocked<{sizePerThread = [4, 1], threadsPerWarp = [16, 2], warpsPerCTA = [1, 4], order = [0, 1]}>
// CHECK: [[load_ptr:%.*]] = triton_gpu.convert_layout {{.*}} -> tensor<64x64x!tt.ptr<f32>, [[row_layout]]>
// CHECK: [[load_mask:%.*]] = triton_gpu.convert_layout {{.*}} -> tensor<64x64xi1, [[row_layout]]>
// CHECK: [[load_other:%.*]] = triton_gpu.convert_layout {{.*}} -> tensor<64x64xf32, [[row_layout]]>