[mlir][sparse] use straightline and loop to insert into tensor

[aartbik]

Summary

This exposed a missing type conversion for codegen

[aartbik]

fixed header

[aartbik]

mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_insert_1d.mlir

63

Just FYI at the moment. We break here because

(1) bufferization copy
(2) SCF.for does not allow 1:N type conversions yet

[Peiming]

mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_insert_1d.mlir

68

You forget a sparse_tensor.load %7 here

[aartbik]

mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_insert_1d.mlir

68

I sure did!

[aartbik]

load + hasInserts

[aartbik]

rebased with main, added NOE usage

(note that this of course still needs the 1:N conversion)

[aartbik]

rebased with main, functional now

mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_insert_1d.mlir

 // RUN: mlir-opt %s --sparse-compiler=enable-runtime-library=false | \
-// RUN: mlir-cpu-runner \
-// RUN:  -e entry -entry-point-result=void  \
-// RUN:  -shared-libs=%mlir_lib_dir/libmlir_c_runner_utils%shlibext | \
+// RUN: mlir-cpu-runner -e entry -entry-point-result=void          | \
 // RUN: FileCheck %s
 
-// Insertion example using pure codegen (no sparse runtime support lib).
+// Insertion example using pure codegen (no support lib).
 
 #SparseVector = #sparse_tensor.encoding<{ dimLevelType = [ "compressed" ] }>
 
 #trait_mul_s = {
   indexing_maps = [
     affine_map<(i) -> (i)>   // x (out)
   ],
   iterator_types = ["parallel"],
module {
   }
 
   func.func @entry() {
     %f1    = arith.constant 1.0 : f32
     %f2    = arith.constant 2.0 : f32
     %c0    = arith.constant 0 : index
     %c1    = arith.constant 1 : index
     %c3    = arith.constant 3 : index
+    %c8    = arith.constant 8 : index
     %c1023 = arith.constant 1023 : index
 
-    // Build the sparse vector from code.
+    // Build the sparse vector from straightline code.
     %0 = bufferization.alloc_tensor() : tensor<1024xf32, #SparseVector>
     %1 = sparse_tensor.insert %f1 into %0[%c0] : tensor<1024xf32, #SparseVector>
     %2 = sparse_tensor.insert %f2 into %1[%c1] : tensor<1024xf32, #SparseVector>
     %3 = sparse_tensor.insert %f1 into %2[%c3] : tensor<1024xf32, #SparseVector>
     %4 = sparse_tensor.insert %f2 into %3[%c1023] : tensor<1024xf32, #SparseVector>
     %5 = sparse_tensor.load %4 hasInserts : tensor<1024xf32, #SparseVector>
 
     // CHECK:      ( 0, 4, 99, 99, 99, 99, 99, 99 )
     // CHECK-NEXT: ( 0, 1, 3, 1023, 99, 99, 99, 99 )
     // CHECK-NEXT: ( 1, 2, 1, 2, 99, 99, 99, 99 )
     call @dump(%5) : (tensor<1024xf32, #SparseVector>) -> ()
 
+    // Build another sparse vector in a loop.
+    %6 = bufferization.alloc_tensor() : tensor<1024xf32, #SparseVector>
+    %7 = scf.for %i = %c0 to %c8 step %c1 iter_args(%vin = %6) -> tensor<1024xf32, #SparseVector> {
+      %ii = arith.muli %i, %c3 : index
+      %vout = sparse_tensor.insert %f1 into %vin[%ii] : tensor<1024xf32, #SparseVector>
+      scf.yield %vout : tensor<1024xf32, #SparseVector>
+    }
+
+    // CHECK:      
+    // CHECK-NEXT: 
+    // CHECK-NEXT: 
+    call @dump(%7) : (tensor<1024xf32, #SparseVector>) -> ()
+
+    // Free resources.
     bufferization.dealloc_tensor %5 : tensor<1024xf32, #SparseVector>
+    bufferization.dealloc_tensor %7 : tensor<1024xf32, #SparseVector>
     return
   }
 }