Eliminate literal parameters to uvector::set_element_async and device…

…_scalar::set_value (#7563)

After rapidsai/rmm#725 is merged, this PR updates cuspatial to eliminate passing literal values to `device_scalar::set_value`.

Note there are still similar issues with uses of `cudf::scalar` that need to be addressed. But I wanted to get this open before the end of the week. This can be merged after RMM 725 is merged to fix the build, so I think the `scalar` issues should be fixed separately.

Authors:
  - Mark Harris (@harrism)

Approvers:
  - Devavret Makkar (@devavret)
  - Paul Taylor (@trxcllnt)
  - Mike Wilson (@hyperbolic2346)

URL: #7563

cpp/src/io/json/reader_impl.cu

@@ -158,7 +158,7 @@ std::unique_ptr<table> create_json_keys_info_table(const parse_options_view &opt
   auto const info_table_mdv = mutable_table_device_view::create(info_table->mutable_view(), stream);
 
   // Reset the key counter - now used for indexing
-  key_counter.set_value(0, stream);
+  key_counter.set_value_zero(stream);
   // Fill the allocated columns
   cudf::io::json::gpu::collect_keys_info(
     options, data, row_offsets, key_counter.data(), {*info_table_mdv}, stream);

cpp/src/join/hash_join.cu

@@ -287,7 +287,7 @@ std::pair<rmm::device_vector<size_type>, rmm::device_vector<size_type>> probe_jo
 
     constexpr int block_size{DEFAULT_JOIN_BLOCK_SIZE};
     detail::grid_1d config(probe_table.num_rows(), block_size);
-    write_index.set_value(0, stream);
+    write_index.set_value_zero(stream);
 
     row_hash hash_probe{probe_table};
     row_equality equality{probe_table, build_table, compare_nulls == null_equality::EQUAL};

cpp/src/join/hash_join.cuh

@@ -120,7 +120,7 @@ size_type estimate_join_output_size(table_device_view build_table,
   do {
     sample_probe_num_rows = std::min(sample_probe_num_rows, probe_table_num_rows);
 
-    size_estimate.set_value(0, stream);
+    size_estimate.set_value_zero(stream);
 
     row_hash hash_probe{probe_table};
     row_equality equality{probe_table, build_table, compare_nulls == null_equality::EQUAL};

cpp/src/join/nested_loop_join.cuh

@@ -89,7 +89,7 @@ size_type estimate_nested_loop_join_output_size(table_device_view left,
   int num_sms{-1};
   CUDA_TRY(cudaDeviceGetAttribute(&num_sms, cudaDevAttrMultiProcessorCount, dev_id));
 
-  size_estimate.set_value(0, stream);
+  size_estimate.set_value_zero(stream);
 
   row_equality equality{left, right, compare_nulls == null_equality::EQUAL};
   // Determine number of output rows without actually building the output to simply
@@ -163,7 +163,7 @@ get_base_nested_loop_join_indices(table_view const& left,
 
     constexpr int block_size{DEFAULT_JOIN_BLOCK_SIZE};
     detail::grid_1d config(left_table->num_rows(), block_size);
-    write_index.set_value(0);
+    write_index.set_value_zero(stream);
 
     row_equality equality{*left_table, *right_table, compare_nulls == null_equality::EQUAL};
     const auto& join_output_l =
@@ -182,7 +182,7 @@ get_base_nested_loop_join_indices(table_view const& left,
 
     CHECK_CUDA(stream.value());
 
-    join_size              = write_index.value();
+    join_size              = write_index.value(stream);
     current_estimated_size = estimated_size;
     estimated_size *= 2;
   } while ((current_estimated_size < join_size));

cpp/tests/jit/jit-cache-multiprocess-test.cpp

@@ -49,8 +49,10 @@ TEST_F(JitCacheMultiProcessTest, MultiProcessTest)
     // Brand new cache object that has nothing in in-memory cache
     cudf::jit::cudfJitCache cache;
 
-    input->set_value(4);
-    output->set_value(1);
+    auto const in{4};
+    auto const out{1};
+    input->set_value(in);
+    output->set_value(out);
 
     // make program
     auto program = cache.getProgram("FileCacheTestProg3", program3_source);