[Mlir-commits] [mlir] 30c1866 - [mlir][sparse][gpu] enable SpGEMM on GPU for libgen path

[Aart Bik]

Author: Aart Bik
Date: 2023-08-15T17:16:37-07:00
New Revision: 30c1866dec58ea10345f90b9ca2a8902a2f76e50

URL: https://github.com/llvm/llvm-project/commit/30c1866dec58ea10345f90b9ca2a8902a2f76e50
DIFF: https://github.com/llvm/llvm-project/commit/30c1866dec58ea10345f90b9ca2a8902a2f76e50.diff

LOG: [mlir][sparse][gpu] enable SpGEMM on GPU for libgen path

Direct IR supports pack, but libgen parth did not until
this was added in https://reviews.llvm.org/D158012

Reviewed By: Peiming

Differential Revision: https://reviews.llvm.org/D158020

Added: 
    

Modified: 
    mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-gemm-lib.mlir
    mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-sampled-matmul-lib.mlir

Removed: 
    


################################################################################
diff  --git a/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-gemm-lib.mlir b/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-gemm-lib.mlir
index a39fdd8dc0ac6a..a310e59d53038f 100644
--- a/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-gemm-lib.mlir
+++ b/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-gemm-lib.mlir
@@ -1,13 +1,23 @@
 //
 // NOTE: this test requires gpu-sm80
 //
+// with RT lib:
+//
+// RUN: mlir-opt %s \
+// RUN:   --sparse-compiler="enable-runtime-library=true enable-gpu-libgen gpu-triple=nvptx64-nvidia-cuda gpu-chip=sm_80 gpu-features=+ptx71"  \
+// RUN: | mlir-cpu-runner \
+// RUN:   --shared-libs=%mlir_cuda_runtime \
+// RUN:   --shared-libs=%mlir_c_runner_utils \
+// RUN:   --e main --entry-point-result=void \
+// RUN: | FileCheck %s
+//
 // without RT lib:
 //
 // RUN: mlir-opt %s \
 // RUN:   --sparse-compiler="enable-runtime-library=false enable-gpu-libgen gpu-triple=nvptx64-nvidia-cuda gpu-chip=sm_80 gpu-features=+ptx71"  \
 // RUN: | mlir-cpu-runner \
 // RUN:   --shared-libs=%mlir_cuda_runtime \
-// RUN:   --shared-libs=%mlir_runner_utils \
+// RUN:   --shared-libs=%mlir_c_runner_utils \
 // RUN:   --e main --entry-point-result=void \
 // RUN: | FileCheck %s
 

diff  --git a/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-sampled-matmul-lib.mlir b/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-sampled-matmul-lib.mlir
index 3332e3c4b17a30..934bd837420c1b 100644
--- a/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-sampled-matmul-lib.mlir
+++ b/mlir/test/Integration/Dialect/SparseTensor/GPU/CUDA/sparse-sampled-matmul-lib.mlir
@@ -23,7 +23,7 @@
 // RUN:  %{compile} enable-runtime-library=false gpu-data-transfer-strategy=pinned-dma" | %{run}
 //  Tracker #64316
 // RUNNOT: %{compile} enable-runtime-library=false gpu-data-transfer-strategy=zero-copy" | %{run}
-// 
+//
 
 !Filename = !llvm.ptr<i8>