fixed spspmm for cpu

Author: rusty1s

csrc/cpu/spspmm_cpu.cpp

@@ -48,80 +48,56 @@ spspmm_cpu(torch::Tensor rowptrA, torch::Tensor colA,
   auto rowptrB_data = rowptrB.data_ptr<int64_t>();
   auto colB_data = colB.data_ptr<int64_t>();
 
-  // Pass 1: Compute CSR row pointer.
   auto rowptrC = torch::empty_like(rowptrA);
   auto rowptrC_data = rowptrC.data_ptr<int64_t>();
   rowptrC_data[0] = 0;
 
-  std::vector<int64_t> mask(K, -1);
-  int64_t nnz = 0, row_nnz, rowA_start, rowA_end, rowB_start, rowB_end, cA, cB;
-  for (auto n = 0; n < rowptrA.numel() - 1; n++) {
-    row_nnz = 0;
-
-    for (auto eA = rowptrA_data[n]; eA < rowptrA_data[n + 1]; eA++) {
-      cA = colA_data[eA];
-      for (auto eB = rowptrB_data[cA]; eB < rowptrB_data[cA + 1]; eB++) {
-        cB = colB_data[eB];
-        if (mask[cB] != n) {
-          mask[cB] = n;
-          row_nnz++;
-        }
-      }
-    }
-
-    nnz += row_nnz;
-    rowptrC_data[n + 1] = nnz;
-  }
-
-  // Pass 2: Compute CSR entries.
-  auto colC = torch::empty(nnz, rowptrC.options());
-  auto colC_data = colC.data_ptr<int64_t>();
-
+  torch::Tensor colC;
   torch::optional<torch::Tensor> optional_valueC = torch::nullopt;
-  if (optional_valueA.has_value())
-    optional_valueC = torch::empty(nnz, optional_valueA.value().options());
 
   AT_DISPATCH_ALL_TYPES(scalar_type, "spspmm", [&] {
-    AT_DISPATCH_HAS_VALUE(optional_valueC, [&] {
-      scalar_t *valA_data = nullptr, *valB_data = nullptr, *valC_data = nullptr;
+    AT_DISPATCH_HAS_VALUE(optional_valueA, [&] {
+      scalar_t *valA_data = nullptr, *valB_data = nullptr;
       if (HAS_VALUE) {
         valA_data = optional_valueA.value().data_ptr<scalar_t>();
         valB_data = optional_valueB.value().data_ptr<scalar_t>();
-        valC_data = optional_valueC.value().data_ptr<scalar_t>();
       }
-      scalar_t valA;
 
-      rowA_start = 0, nnz = 0;
-      std::vector<scalar_t> vals(K, 0);
-      for (auto n = 1; n < rowptrA.numel(); n++) {
-        rowA_end = rowptrA_data[n];
+      int64_t nnz = 0, cA, cB;
+      std::vector<scalar_t> tmp_vals(K, 0);
+      std::vector<int64_t> cols;
+      std::vector<scalar_t> vals;
 
-        for (auto eA = rowA_start; eA < rowA_end; eA++) {
+      for (auto rA = 0; rA < rowptrA.numel() - 1; rA++) {
+        for (auto eA = rowptrA_data[rA]; eA < rowptrA_data[rA + 1]; eA++) {
           cA = colA_data[eA];
-          if (HAS_VALUE)
-            valA = valA_data[eA];
-
-          rowB_start = rowptrB_data[cA], rowB_end = rowptrB_data[cA + 1];
-          for (auto eB = rowB_start; eB < rowB_end; eB++) {
+          for (auto eB = rowptrB_data[cA]; eB < rowptrB_data[cA + 1]; eB++) {
             cB = colB_data[eB];
+
             if (HAS_VALUE)
-              vals[cB] += valA * valB_data[eB];
+              tmp_vals[cB] += valA_data[eA] * valB_data[eB];
             else
-              vals[cB] += 1;
+              tmp_vals[cB]++;
           }
         }
 
         for (auto k = 0; k < K; k++) {
-          if (vals[k] != 0) {
-            colC_data[nnz] = k;
+          if (tmp_vals[k] != 0) {
+            cols.push_back(k);
             if (HAS_VALUE)
-              valC_data[nnz] = vals[k];
+              vals.push_back(tmp_vals[k]);
             nnz++;
           }
-          vals[k] = (scalar_t)0;
+          tmp_vals[k] = (scalar_t)0;
         }
+        rowptrC_data[rA + 1] = nnz;
+      }
 
-        rowA_start = rowA_end;
+      colC = torch::from_blob(cols.data(), {nnz}, colA.options()).clone();
+      if (HAS_VALUE) {
+        optional_valueC = torch::from_blob(vals.data(), {nnz},
+                                           optional_valueA.value().options());
+        optional_valueC = optional_valueC.value().clone();
       }
     });
   });