Remove move from Tile. (#4749)

This will help the memory tracking effort by removing all move
constructors/assign operator for Tile, ResultTile, etc.

[SC-41591]

---
TYPE: NO_HISTORY
DESC: Remove move from Tile.

Author: KiterLuc

test/src/unit-Reader.cc

@@ -278,38 +278,27 @@ TEST_CASE_METHOD(
       result_space_tiles);
   CHECK(result_space_tiles.size() == 6);
 
-  // Result tiles for fragment #1
-  ResultTile result_tile_1_0_1(1, 0, *fragments[0]);
-  ResultTile result_tile_1_2_1(1, 2, *fragments[0]);
-
-  // Result tiles for fragment #2
-  ResultTile result_tile_1_0_2(2, 0, *fragments[1]);
-
-  // Result tiles for fragment #3
-  ResultTile result_tile_2_0_3(3, 0, *fragments[2]);
-  ResultTile result_tile_3_0_3(3, 2, *fragments[2]);
-
   // Initialize result_space_tiles
   ResultSpaceTile<int32_t> rst_1_0;
   rst_1_0.set_start_coords({3, 1});
   rst_1_0.append_frag_domain(2, ds2);
   rst_1_0.append_frag_domain(1, ds1);
-  rst_1_0.set_result_tile(1, result_tile_1_0_1);
-  rst_1_0.set_result_tile(2, result_tile_1_0_2);
+  rst_1_0.set_result_tile(1, 0, *fragments[0]);
+  rst_1_0.set_result_tile(2, 0, *fragments[1]);
   ResultSpaceTile<int32_t> rst_1_2;
   rst_1_2.set_start_coords({3, 11});
   rst_1_2.append_frag_domain(1, ds1);
-  rst_1_2.set_result_tile(1, result_tile_1_2_1);
+  rst_1_2.set_result_tile(1, 2, *fragments[0]);
   ResultSpaceTile<int32_t> rst_2_0;
   rst_2_0.set_start_coords({5, 1});
   rst_2_0.append_frag_domain(3, ds3);
-  rst_2_0.set_result_tile(3, result_tile_2_0_3);
+  rst_2_0.set_result_tile(3, 0, *fragments[2]);
   ResultSpaceTile<int32_t> rst_2_2;
   rst_2_2.set_start_coords({5, 11});
   ResultSpaceTile<int32_t> rst_3_0;
   rst_3_0.set_start_coords({7, 1});
   rst_3_0.append_frag_domain(3, ds3);
-  rst_3_0.set_result_tile(3, result_tile_3_0_3);
+  rst_3_0.set_result_tile(3, 2, *fragments[2]);
   ResultSpaceTile<int32_t> rst_3_2;
   rst_3_2.set_start_coords({7, 11});
 

test/src/unit-filter-pipeline.cc

@@ -103,15 +103,14 @@ WriterTile make_offsets_tile(std::vector<uint64_t>& offsets) {
 }
 
 Tile create_tile_for_unfiltering(uint64_t nelts, WriterTile& tile) {
-  Tile ret(
+  return {
       tile.format_version(),
       tile.type(),
       tile.cell_size(),
       0,
       tile.cell_size() * nelts,
       tile.filtered_buffer().data(),
-      tile.filtered_buffer().size());
-  return ret;
+      tile.filtered_buffer().size()};
 }
 
 void run_reverse(
@@ -1221,19 +1220,19 @@ TEST_CASE("Filter: Test encryption", "[filter][encryption]") {
     key[0]++;
     filter->set_key(key);
 
-    unfiltered_tile = create_tile_for_unfiltering(nelts, tile);
-    run_reverse(config, tp, unfiltered_tile, pipeline, false);
+    auto unfiltered_tile2 = create_tile_for_unfiltering(nelts, tile);
+    run_reverse(config, tp, unfiltered_tile2, pipeline, false);
 
     // Fix key and check success.
-    unfiltered_tile = create_tile_for_unfiltering(nelts, tile);
+    auto unfiltered_tile3 = create_tile_for_unfiltering(nelts, tile);
     key[0]--;
     filter->set_key(key);
-    run_reverse(config, tp, unfiltered_tile, pipeline);
+    run_reverse(config, tp, unfiltered_tile3, pipeline);
 
     for (uint64_t i = 0; i < nelts; i++) {
       uint64_t elt = 0;
       CHECK_NOTHROW(
-          unfiltered_tile.read(&elt, i * sizeof(uint64_t), sizeof(uint64_t)));
+          unfiltered_tile3.read(&elt, i * sizeof(uint64_t), sizeof(uint64_t)));
       CHECK(elt == i);
     }
   }

test/src/unit-sparse-unordered-with-dups-reader.cc

@@ -1527,7 +1527,7 @@ TEST_CASE_METHOD(
   auto&& [array, fragments] = open_default_array_1d_with_fragments(capacity);
 
   // Make a vector of tiles.
-  std::vector<UnorderedWithDupsResultTile<uint64_t>> rt;
+  std::list<UnorderedWithDupsResultTile<uint64_t>> rt;
   for (uint64_t t = 0; t < num_tiles; t++) {
     rt.emplace_back(0, t, *fragments[0]);
 
@@ -1540,8 +1540,9 @@ TEST_CASE_METHOD(
 
   // Create the result_tiles pointer vector.
   std::vector<ResultTile*> result_tiles(rt.size());
-  for (uint64_t i = 0; i < rt.size(); i++) {
-    result_tiles[i] = &rt[i];
+  uint64_t i = 0;
+  for (auto& t : rt) {
+    result_tiles[i++] = &t;
   }
 
   // Create a Query buffer.
@@ -1743,7 +1744,7 @@ TEST_CASE_METHOD(
   auto&& [array, fragments] = open_default_array_1d_with_fragments(capacity);
 
   // Make a vector of tiles.
-  std::vector<UnorderedWithDupsResultTile<uint64_t>> rt;
+  std::list<UnorderedWithDupsResultTile<uint64_t>> rt;
   for (uint64_t t = 0; t < num_tiles; t++) {
     rt.emplace_back(0, t, *fragments[0]);
 
@@ -1756,8 +1757,9 @@ TEST_CASE_METHOD(
 
   // Create the result_tiles pointer vector.
   std::vector<ResultTile*> result_tiles(rt.size());
-  for (uint64_t i = 0; i < rt.size(); i++) {
-    result_tiles[i] = &rt[i];
+  uint64_t i = 0;
+  for (auto& t : rt) {
+    result_tiles[i++] = &t;
   }
 
   // Call the function.

tiledb/sm/array/array.cc

@@ -1478,10 +1478,8 @@ void Array::do_load_metadata() {
   throw_if_not_ok(
       parallel_for(&resources_.compute_tp(), 0, metadata_num, [&](size_t m) {
         const auto& uri = array_metadata_to_load[m].uri_;
-
-        auto&& tile =
+        metadata_tiles[m] =
             GenericTileIO::load(resources_, uri, 0, *encryption_key());
-        metadata_tiles[m] = tdb::make_shared<Tile>(HERE(), std::move(tile));
 
         return Status::Ok();
       }));

tiledb/sm/array/array_directory.cc

@@ -96,12 +96,12 @@ shared_ptr<ArraySchema> ArrayDirectory::load_array_schema_from_uri(
   auto timer_se =
       resources.stats().start_timer("sm_load_array_schema_from_uri");
 
-  auto&& tile = GenericTileIO::load(resources, schema_uri, 0, encryption_key);
+  auto tile = GenericTileIO::load(resources, schema_uri, 0, encryption_key);
 
-  resources.stats().add_counter("read_array_schema_size", tile.size());
+  resources.stats().add_counter("read_array_schema_size", tile->size());
 
   // Deserialize
-  Deserializer deserializer(tile.data(), tile.size());
+  Deserializer deserializer(tile->data(), tile->size());
   auto memory_tracker = resources.create_memory_tracker();
   memory_tracker->set_type(MemoryTrackerType::ARRAY_LOAD);
   return ArraySchema::deserialize(deserializer, schema_uri, memory_tracker);
@@ -1323,18 +1323,18 @@ shared_ptr<const Enumeration> ArrayDirectory::load_enumeration(
                       .join_path(constants::array_enumerations_dir_name)
                       .join_path(enumeration_path);
 
-  auto&& tile = GenericTileIO::load(resources_, enmr_uri, 0, encryption_key);
-  resources_.get().stats().add_counter("read_enumeration_size", tile.size());
+  auto tile = GenericTileIO::load(resources_, enmr_uri, 0, encryption_key);
+  resources_.get().stats().add_counter("read_enumeration_size", tile->size());
 
-  if (!memory_tracker->take_memory(tile.size(), MemoryType::ENUMERATION)) {
+  if (!memory_tracker->take_memory(tile->size(), MemoryType::ENUMERATION)) {
     throw ArrayDirectoryException(
         "Error loading enumeration; Insufficient memory budget; Needed " +
-        std::to_string(tile.size()) + " but only had " +
+        std::to_string(tile->size()) + " but only had " +
         std::to_string(memory_tracker->get_memory_available()) +
         " from budget " + std::to_string(memory_tracker->get_memory_budget()));
   }
 
-  Deserializer deserializer(tile.data(), tile.size());
+  Deserializer deserializer(tile->data(), tile->size());
   return Enumeration::deserialize(deserializer);
 }
 

tiledb/sm/fragment/fragment_info.cc

@@ -952,7 +952,7 @@ Status FragmentInfo::load_and_replace(
   return Status::Ok();
 }
 
-tuple<Tile, std::vector<std::pair<std::string, uint64_t>>>
+tuple<shared_ptr<Tile>, std::vector<std::pair<std::string, uint64_t>>>
 load_consolidated_fragment_meta(
     ContextResources& resources, const URI& uri, const EncryptionKey& enc_key) {
   auto timer_se =
@@ -963,12 +963,12 @@ load_consolidated_fragment_meta(
     throw StatusException(Status_FragmentInfoError(
         "Cannot load consolidated fragment metadata; URI is empty."));
 
-  auto&& tile = GenericTileIO::load(resources, uri, 0, enc_key);
+  auto tile = GenericTileIO::load(resources, uri, 0, enc_key);
 
-  resources.stats().add_counter("consolidated_frag_meta_size", tile.size());
+  resources.stats().add_counter("consolidated_frag_meta_size", tile->size());
 
   uint32_t fragment_num;
-  Deserializer deserializer(tile.data(), tile.size());
+  Deserializer deserializer(tile->data(), tile->size());
   fragment_num = deserializer.read<uint32_t>();
 
   uint64_t name_size, offset;
@@ -983,7 +983,7 @@ load_consolidated_fragment_meta(
     ret.emplace_back(name, offset);
   }
 
-  return {std::move(tile), std::move(ret)};
+  return {tile, std::move(ret)};
 }
 
 std::tuple<
@@ -1021,8 +1021,7 @@ FragmentInfo::load_array_schemas_and_fragment_metadata(
       parallel_for(&resources.compute_tp(), 0, meta_uris.size(), [&](size_t i) {
         auto&& [tile_opt, offsets] =
             load_consolidated_fragment_meta(resources, meta_uris[i], enc_key);
-        fragment_metadata_tiles[i] =
-            make_shared<Tile>(HERE(), std::move(tile_opt));
+        fragment_metadata_tiles[i] = tile_opt;
         offsets_vectors[i] = std::move(offsets);
         return Status::Ok();
       }));