From 914eff5a8ca9a05cf14ca758c282e8ca8987967d Mon Sep 17 00:00:00 2001
From: Aart Bik <ajcbik@google.com>
Date: Fri, 18 Nov 2022 16:57:02 -0800
Subject: [PATCH] [mlir][sparse][vector] ensure loop peeling to remove vector
 masks works

Reviewed By: Peiming

Differential Revision: https://reviews.llvm.org/D138343
---
 .../Dialect/SparseTensor/sparse_vector_peeled.mlir | 63 ++++++++++++++++++++++
 1 file changed, 63 insertions(+)
 create mode 100644 mlir/test/Dialect/SparseTensor/sparse_vector_peeled.mlir

diff --git a/mlir/test/Dialect/SparseTensor/sparse_vector_peeled.mlir b/mlir/test/Dialect/SparseTensor/sparse_vector_peeled.mlir
new file mode 100644
index 0000000..d379965
--- /dev/null
+++ b/mlir/test/Dialect/SparseTensor/sparse_vector_peeled.mlir
@@ -0,0 +1,63 @@
+// RUN: mlir-opt %s --sparsification -cse -sparse-vectorization="vl=16" -scf-for-loop-peeling -canonicalize -cse | \
+// RUN:   FileCheck %s
+
+#SparseVector = #sparse_tensor.encoding<{
+  dimLevelType = [ "compressed" ],
+  pointerBitWidth = 32,
+  indexBitWidth = 32
+}>
+
+#trait_mul_s = {
+  indexing_maps = [
+    affine_map<(i) -> (i)>,  // a
+    affine_map<(i) -> (i)>,  // b
+    affine_map<(i) -> (i)>   // x (out)
+  ],
+  iterator_types = ["parallel"],
+  doc = "x(i) = a(i) * b(i)"
+}
+
+// CHECK-DAG:   #[[$map0:.*]] = affine_map<()[s0, s1] -> (s0 + ((-s0 + s1) floordiv 16) * 16)>
+// CHECK-DAG:   #[[$map1:.*]] = affine_map<(d0)[s0] -> (-d0 + s0)>
+// CHECK-LABEL: func @mul_s
+// CHECK-DAG:   %[[c0:.*]] = arith.constant 0 : index
+// CHECK-DAG:   %[[c1:.*]] = arith.constant 1 : index
+// CHECK-DAG:   %[[c16:.*]] = arith.constant 16 : index
+// CHECK:       %[[p:.*]] = memref.load %{{.*}}[%[[c0]]] : memref<?xi32>
+// CHECK:       %[[a:.*]] = arith.extui %[[p]] : i32 to i64
+// CHECK:       %[[q:.*]] = arith.index_cast %[[a]] : i64 to index
+// CHECK:       %[[r:.*]] = memref.load %{{.*}}[%[[c1]]] : memref<?xi32>
+// CHECK:       %[[b:.*]] = arith.extui %[[r]] : i32 to i64
+// CHECK:       %[[s:.*]] = arith.index_cast %[[b]] : i64 to index
+// CHECK:       %[[boundary:.*]] = affine.apply #[[$map0]]()[%[[q]], %[[s]]]
+// CHECK:       scf.for %[[i:.*]] = %[[q]] to %[[boundary]] step %[[c16]] {
+// CHECK:         %[[mask:.*]] = vector.constant_mask [16] : vector<16xi1>
+// CHECK:         %[[li:.*]] = vector.load %{{.*}}[%[[i]]] : memref<?xi32>, vector<16xi32>
+// CHECK:         %[[zi:.*]] = arith.extui %[[li]] : vector<16xi32> to vector<16xi64>
+// CHECK:         %[[la:.*]] = vector.load %{{.*}}[%[[i]]] : memref<?xf32>, vector<16xf32>
+// CHECK:         %[[lb:.*]] = vector.gather %{{.*}}[%[[c0]]] [%[[zi]]], %[[mask]], %{{.*}} : memref<1024xf32>, vector<16xi64>, vector<16xi1>, vector<16xf32> into vector<16xf32>
+// CHECK:         %[[m:.*]] = arith.mulf %[[la]], %[[lb]] : vector<16xf32>
+// CHECK:         vector.scatter %{{.*}}[%[[c0]]] [%[[zi]]], %[[mask]], %[[m]] : memref<1024xf32>, vector<16xi64>, vector<16xi1>, vector<16xf32>
+// CHECK:       }
+// CHECK:       scf.for %[[i2:.*]] = %[[boundary]] to %[[s]] step %[[c16]] {
+// CHECK:         %[[sub:.*]] = affine.apply #[[$map1]](%[[i2]])[%[[s]]]
+// CHECK:         %[[mask2:.*]] = vector.create_mask %[[sub]] : vector<16xi1>
+// CHECK:         %[[li2:.*]] = vector.maskedload %{{.*}}[%[[i2]]], %[[mask2]], %{{.*}} : memref<?xi32>, vector<16xi1>, vector<16xi32> into vector<16xi32>
+// CHECK:         %[[zi2:.*]] = arith.extui %[[li2]] : vector<16xi32> to vector<16xi64>
+// CHECK:         %[[la2:.*]] = vector.maskedload %{{.*}}[%[[i2]]], %[[mask2]], %{{.*}} : memref<?xf32>, vector<16xi1>, vector<16xf32> into vector<16xf32>
+// CHECK:         %[[lb2:.*]] = vector.gather %{{.*}}[%[[c0]]] [%[[zi2]]], %[[mask2]], %{{.*}} : memref<1024xf32>, vector<16xi64>, vector<16xi1>, vector<16xf32> into vector<16xf32>
+// CHECK:         %[[m2:.*]] = arith.mulf %[[la2]], %[[lb2]] : vector<16xf32>
+// CHECK:         vector.scatter %{{.*}}[%[[c0]]] [%[[zi2]]], %[[mask2]], %[[m2]] : memref<1024xf32>, vector<16xi64>, vector<16xi1>, vector<16xf32>
+// CHECK:       }
+// CHECK:       return
+//
+func.func @mul_s(%arga: tensor<1024xf32, #SparseVector>, %argb: tensor<1024xf32>, %argx: tensor<1024xf32>) -> tensor<1024xf32> {
+  %0 = linalg.generic #trait_mul_s
+    ins(%arga, %argb: tensor<1024xf32, #SparseVector>, tensor<1024xf32>)
+    outs(%argx: tensor<1024xf32>) {
+      ^bb(%a: f32, %b: f32, %x: f32):
+        %0 = arith.mulf %a, %b : f32
+        linalg.yield %0 : f32
+  } -> tensor<1024xf32>
+  return %0 : tensor<1024xf32>
+}
-- 
2.7.4