diff --git a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_conversion_sparse2dense.mlir b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_conversion_sparse2dense.mlir
--- a/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_conversion_sparse2dense.mlir
+++ b/mlir/test/Integration/Dialect/SparseTensor/CPU/sparse_conversion_sparse2dense.mlir
@@ -132,18 +132,21 @@
     %s2345 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x3x4xf64, #Tensor5>
     %s2346 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x3x4xf64, #Tensor6>
 
+    // NOTE: We commented out most cases with the dynamic-sized output tensor because finishing
+    // all of them are currently taking too long, and they are not covering too many new code
+    // pathes.
     %sp344 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<?x3x4xf64, #Tensor4>
-    %sp345 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<?x3x4xf64, #Tensor5>
-    %sp346 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<?x3x4xf64, #Tensor6>
-    %s2p44 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x4xf64, #Tensor4>
+    // %sp345 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<?x3x4xf64, #Tensor5>
+    // %sp346 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<?x3x4xf64, #Tensor6>
+    // %s2p44 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x4xf64, #Tensor4>
     %s2p45 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x4xf64, #Tensor5>
-    %s2p46 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x4xf64, #Tensor6>
-    %s23p4 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x3x?xf64, #Tensor4>
-    %s23p5 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x3x?xf64, #Tensor5>
+    // %s2p46 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x4xf64, #Tensor6>
+    // %s23p4 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x3x?xf64, #Tensor4>
+    // %s23p5 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x3x?xf64, #Tensor5>
     %s23p6 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x3x?xf64, #Tensor6>
-    %s2pp4 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x?xf64, #Tensor4>
-    %s2pp5 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x?xf64, #Tensor5>
-    %s2pp6 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x?xf64, #Tensor6>
+    // %s2pp4 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x?xf64, #Tensor4>
+    // %s2pp5 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x?xf64, #Tensor5>
+    // %s2pp6 = sparse_tensor.convert %src : tensor<2x3x4xf64> to tensor<2x?x?xf64, #Tensor6>
 
     //
     // Convert sparse tensor back to dense.
@@ -156,32 +159,32 @@
     %d2346 = sparse_tensor.convert %s2346 : tensor<2x3x4xf64, #Tensor6> to tensor<2x3x4xf64>
 
     %dp344 = sparse_tensor.convert %sp344 : tensor<?x3x4xf64, #Tensor4> to tensor<?x3x4xf64>
-    %dp345 = sparse_tensor.convert %sp345 : tensor<?x3x4xf64, #Tensor5> to tensor<?x3x4xf64>
-    %dp346 = sparse_tensor.convert %sp346 : tensor<?x3x4xf64, #Tensor6> to tensor<?x3x4xf64>
-    %d2p44 = sparse_tensor.convert %s2p44 : tensor<2x?x4xf64, #Tensor4> to tensor<2x?x4xf64>
+    // %dp345 = sparse_tensor.convert %sp345 : tensor<?x3x4xf64, #Tensor5> to tensor<?x3x4xf64>
+    // %dp346 = sparse_tensor.convert %sp346 : tensor<?x3x4xf64, #Tensor6> to tensor<?x3x4xf64>
+    // %d2p44 = sparse_tensor.convert %s2p44 : tensor<2x?x4xf64, #Tensor4> to tensor<2x?x4xf64>
     %d2p45 = sparse_tensor.convert %s2p45 : tensor<2x?x4xf64, #Tensor5> to tensor<2x?x4xf64>
-    %d2p46 = sparse_tensor.convert %s2p46 : tensor<2x?x4xf64, #Tensor6> to tensor<2x?x4xf64>
-    %d23p4 = sparse_tensor.convert %s23p4 : tensor<2x3x?xf64, #Tensor4> to tensor<2x3x?xf64>
-    %d23p5 = sparse_tensor.convert %s23p5 : tensor<2x3x?xf64, #Tensor5> to tensor<2x3x?xf64>
+    // %d2p46 = sparse_tensor.convert %s2p46 : tensor<2x?x4xf64, #Tensor6> to tensor<2x?x4xf64>
+    // %d23p4 = sparse_tensor.convert %s23p4 : tensor<2x3x?xf64, #Tensor4> to tensor<2x3x?xf64>
+    // %d23p5 = sparse_tensor.convert %s23p5 : tensor<2x3x?xf64, #Tensor5> to tensor<2x3x?xf64>
     %d23p6 = sparse_tensor.convert %s23p6 : tensor<2x3x?xf64, #Tensor6> to tensor<2x3x?xf64>
-    %d2pp4 = sparse_tensor.convert %s2pp4 : tensor<2x?x?xf64, #Tensor4> to tensor<2x?x?xf64>
-    %d2pp5 = sparse_tensor.convert %s2pp5 : tensor<2x?x?xf64, #Tensor5> to tensor<2x?x?xf64>
-    %d2pp6 = sparse_tensor.convert %s2pp6 : tensor<2x?x?xf64, #Tensor6> to tensor<2x?x?xf64>
+    // %d2pp4 = sparse_tensor.convert %s2pp4 : tensor<2x?x?xf64, #Tensor4> to tensor<2x?x?xf64>
+    // %d2pp5 = sparse_tensor.convert %s2pp5 : tensor<2x?x?xf64, #Tensor5> to tensor<2x?x?xf64>
+    // %d2pp6 = sparse_tensor.convert %s2pp6 : tensor<2x?x?xf64, #Tensor6> to tensor<2x?x?xf64>
 
     %dp3p4 = sparse_tensor.convert %sp344 : tensor<?x3x4xf64, #Tensor4> to tensor<?x3x?xf64>
-    %dp3p5 = sparse_tensor.convert %sp345 : tensor<?x3x4xf64, #Tensor5> to tensor<?x3x?xf64>
-    %dp3p6 = sparse_tensor.convert %sp346 : tensor<?x3x4xf64, #Tensor6> to tensor<?x3x?xf64>
-    %dpp44 = sparse_tensor.convert %s2p44 : tensor<2x?x4xf64, #Tensor4> to tensor<?x?x4xf64>
+    // %dp3p5 = sparse_tensor.convert %sp345 : tensor<?x3x4xf64, #Tensor5> to tensor<?x3x?xf64>
+    // %dp3p6 = sparse_tensor.convert %sp346 : tensor<?x3x4xf64, #Tensor6> to tensor<?x3x?xf64>
+    // %dpp44 = sparse_tensor.convert %s2p44 : tensor<2x?x4xf64, #Tensor4> to tensor<?x?x4xf64>
     %dpp45 = sparse_tensor.convert %s2p45 : tensor<2x?x4xf64, #Tensor5> to tensor<?x?x4xf64>
-    %dpp46 = sparse_tensor.convert %s2p46 : tensor<2x?x4xf64, #Tensor6> to tensor<?x?x4xf64>
-    %dppp4 = sparse_tensor.convert %s2pp4 : tensor<2x?x?xf64, #Tensor4> to tensor<?x?x?xf64>
-    %dppp5 = sparse_tensor.convert %s2pp5 : tensor<2x?x?xf64, #Tensor5> to tensor<?x?x?xf64>
-    %dppp6 = sparse_tensor.convert %s2pp6 : tensor<2x?x?xf64, #Tensor6> to tensor<?x?x?xf64>
+    // %dpp46 = sparse_tensor.convert %s2p46 : tensor<2x?x4xf64, #Tensor6> to tensor<?x?x4xf64>
+    // %dppp4 = sparse_tensor.convert %s2pp4 : tensor<2x?x?xf64, #Tensor4> to tensor<?x?x?xf64>
+    // %dppp5 = sparse_tensor.convert %s2pp5 : tensor<2x?x?xf64, #Tensor5> to tensor<?x?x?xf64>
+    // %dppp6 = sparse_tensor.convert %s2pp6 : tensor<2x?x?xf64, #Tensor6> to tensor<?x?x?xf64>
 
     //
     // Check round-trip equality.  And release dense tensors.
     //
-    // CHECK-COUNT-28: ( ( ( 1, 2, 3, 4 ), ( 5, 6, 7, 8 ), ( 9, 10, 11, 12 ) ), ( ( 13, 14, 15, 16 ), ( 17, 18, 19, 20 ), ( 21, 22, 23, 24 ) ) )
+    // CHECK-COUNT-12: ( ( ( 1, 2, 3, 4 ), ( 5, 6, 7, 8 ), ( 9, 10, 11, 12 ) ), ( ( 13, 14, 15, 16 ), ( 17, 18, 19, 20 ), ( 21, 22, 23, 24 ) ) )
     call @dump_234(%src) : (tensor<2x3x4xf64>) -> ()
     call @dump_234(%d2341) : (tensor<2x3x4xf64>) -> ()
     call @dump_234(%d2342) : (tensor<2x3x4xf64>) -> ()
@@ -190,26 +193,26 @@
     call @dump_234(%d2345) : (tensor<2x3x4xf64>) -> ()
     call @dump_234(%d2346) : (tensor<2x3x4xf64>) -> ()
     call @dump_p34(%dp344) : (tensor<?x3x4xf64>) -> ()
-    call @dump_p34(%dp345) : (tensor<?x3x4xf64>) -> ()
-    call @dump_p34(%dp346) : (tensor<?x3x4xf64>) -> ()
-    call @dump_2p4(%d2p44) : (tensor<2x?x4xf64>) -> ()
+    // call @dump_p34(%dp345) : (tensor<?x3x4xf64>) -> ()
+    // call @dump_p34(%dp346) : (tensor<?x3x4xf64>) -> ()
+    // call @dump_2p4(%d2p44) : (tensor<2x?x4xf64>) -> ()
     call @dump_2p4(%d2p45) : (tensor<2x?x4xf64>) -> ()
-    call @dump_2p4(%d2p46) : (tensor<2x?x4xf64>) -> ()
-    call @dump_23p(%d23p4) : (tensor<2x3x?xf64>) -> ()
-    call @dump_23p(%d23p5) : (tensor<2x3x?xf64>) -> ()
+    // call @dump_2p4(%d2p46) : (tensor<2x?x4xf64>) -> ()
+    // call @dump_23p(%d23p4) : (tensor<2x3x?xf64>) -> ()
+    // call @dump_23p(%d23p5) : (tensor<2x3x?xf64>) -> ()
     call @dump_23p(%d23p6) : (tensor<2x3x?xf64>) -> ()
-    call @dump_2pp(%d2pp4) : (tensor<2x?x?xf64>) -> ()
-    call @dump_2pp(%d2pp5) : (tensor<2x?x?xf64>) -> ()
-    call @dump_2pp(%d2pp6) : (tensor<2x?x?xf64>) -> ()
+    // call @dump_2pp(%d2pp4) : (tensor<2x?x?xf64>) -> ()
+    // call @dump_2pp(%d2pp5) : (tensor<2x?x?xf64>) -> ()
+    // call @dump_2pp(%d2pp6) : (tensor<2x?x?xf64>) -> ()
     call @dump_p3p(%dp3p4) : (tensor<?x3x?xf64>) -> ()
-    call @dump_p3p(%dp3p5) : (tensor<?x3x?xf64>) -> ()
-    call @dump_p3p(%dp3p6) : (tensor<?x3x?xf64>) -> ()
-    call @dump_pp4(%dpp44) : (tensor<?x?x4xf64>) -> ()
+    // call @dump_p3p(%dp3p5) : (tensor<?x3x?xf64>) -> ()
+    // call @dump_p3p(%dp3p6) : (tensor<?x3x?xf64>) -> ()
+    // call @dump_pp4(%dpp44) : (tensor<?x?x4xf64>) -> ()
     call @dump_pp4(%dpp45) : (tensor<?x?x4xf64>) -> ()
-    call @dump_pp4(%dpp46) : (tensor<?x?x4xf64>) -> ()
-    call @dump_ppp(%dppp4) : (tensor<?x?x?xf64>) -> ()
-    call @dump_ppp(%dppp5) : (tensor<?x?x?xf64>) -> ()
-    call @dump_ppp(%dppp6) : (tensor<?x?x?xf64>) -> ()
+    // call @dump_pp4(%dpp46) : (tensor<?x?x4xf64>) -> ()
+    // call @dump_ppp(%dppp4) : (tensor<?x?x?xf64>) -> ()
+    // call @dump_ppp(%dppp5) : (tensor<?x?x?xf64>) -> ()
+    // call @dump_ppp(%dppp6) : (tensor<?x?x?xf64>) -> ()
 
     //
     // Release sparse tensors.
@@ -221,18 +224,17 @@
     bufferization.dealloc_tensor %s2345 : tensor<2x3x4xf64, #Tensor5>
     bufferization.dealloc_tensor %s2346 : tensor<2x3x4xf64, #Tensor6>
     bufferization.dealloc_tensor %sp344 : tensor<?x3x4xf64, #Tensor4>
-    bufferization.dealloc_tensor %sp345 : tensor<?x3x4xf64, #Tensor5>
-    bufferization.dealloc_tensor %sp346 : tensor<?x3x4xf64, #Tensor6>
-    bufferization.dealloc_tensor %s2p44 : tensor<2x?x4xf64, #Tensor4>
+    // bufferization.dealloc_tensor %sp345 : tensor<?x3x4xf64, #Tensor5>
+    // bufferization.dealloc_tensor %sp346 : tensor<?x3x4xf64, #Tensor6>
+    // bufferization.dealloc_tensor %s2p44 : tensor<2x?x4xf64, #Tensor4>
     bufferization.dealloc_tensor %s2p45 : tensor<2x?x4xf64, #Tensor5>
-    bufferization.dealloc_tensor %s2p46 : tensor<2x?x4xf64, #Tensor6>
-    bufferization.dealloc_tensor %s23p4 : tensor<2x3x?xf64, #Tensor4>
-    bufferization.dealloc_tensor %s23p5 : tensor<2x3x?xf64, #Tensor5>
+    // bufferization.dealloc_tensor %s2p46 : tensor<2x?x4xf64, #Tensor6>
+    // bufferization.dealloc_tensor %s23p4 : tensor<2x3x?xf64, #Tensor4>
+    // bufferization.dealloc_tensor %s23p5 : tensor<2x3x?xf64, #Tensor5>
     bufferization.dealloc_tensor %s23p6 : tensor<2x3x?xf64, #Tensor6>
-    bufferization.dealloc_tensor %s2pp4 : tensor<2x?x?xf64, #Tensor4>
-    bufferization.dealloc_tensor %s2pp5 : tensor<2x?x?xf64, #Tensor5>
-    bufferization.dealloc_tensor %s2pp6 : tensor<2x?x?xf64, #Tensor6>
-
+    //bufferization.dealloc_tensor %s2pp4 : tensor<2x?x?xf64, #Tensor4>
+    // bufferization.dealloc_tensor %s2pp5 : tensor<2x?x?xf64, #Tensor5>
+    //bufferization.dealloc_tensor %s2pp6 : tensor<2x?x?xf64, #Tensor6>
     return
   }
 }