Apply suggestions from code review

Co-authored-by: Yunsong Wang <[email protected]>

Author: sleeepyjack

include/cuco/detail/open_addressing/open_addressing_impl.cuh

@@ -1130,17 +1130,6 @@ class open_addressing_impl {
     counter.reset(stream.get());
 
     int32_t constexpr block_size = cuco::detail::default_block_size();
-    // int32_t grid_size =
-    //   detail::max_occupancy_grid_size(block_size,
-    //                                   detail::retrieve<IsOuter,
-    //                                                    block_size,
-    //                                                    InputProbeIt,
-    //                                                    OutputProbeIt,
-    //                                                    OutputMatchIt,
-    //                                                    typename counter_type::value_type,
-    //                                                    Ref>);
-    // grid_size *= 64;  // oversubscription factor
-    // // TODO shrink grid if n is very small
 
     auto constexpr grid_stride = 1;
     auto const grid_size       = cuco::detail::grid_size(n, cg_size, grid_stride, block_size);

include/cuco/detail/open_addressing/open_addressing_ref_impl.cuh

@@ -1253,22 +1253,6 @@ class open_addressing_ref_impl {
                                ? sentinel_writers.size()
                                : active_flushing_tile.size() - sentinel_writers.size();
             }
-            //             if (finished and not found_any_match) {
-            // #if defined(CUCO_HAS_CG_INVOKE_ONE)
-            //               cg::invoke_one(probing_tile, [&]() {
-            //                 probe_buffers[flushing_tile_id][num_matches] = probe;
-            //                 probe_buffers[flushing_tile_id][num_matches] =
-            //                 this->empty_slot_sentinel();
-            //               });
-            // #else
-            //               if (probing_tile.thread_rank() == 0) {
-            //                 probe_buffers[flushing_tile_id][num_matches] = probe;
-            //                 probe_buffers[flushing_tile_id][num_matches] =
-            //                 this->empty_slot_sentinel();
-            //               }
-            // #endif
-            //               num_matches++;  // not really a match but a sentinel in the buffer
-            //             }
           }
 
           // if the buffer has not enough empty slots for the next iteration

include/cuco/detail/static_multiset/static_multiset.inl

@@ -277,7 +277,6 @@ void static_multiset<Key, Extent, Scope, KeyEqual, ProbingScheme, Allocator, Sto
   impl_->find_async(first, last, output_begin, ref(op::find), stream);
 }
 
-// TODO docs
 template <class Key,
           class Extent,
           cuda::thread_scope Scope,
@@ -298,7 +297,6 @@ static_multiset<Key, Extent, Scope, KeyEqual, ProbingScheme, Allocator, Storage>
     first, last, output_probe, output_match, this->ref(op::retrieve), stream);
 }
 
-// TODO docs
 template <class Key,
           class Extent,
           cuda::thread_scope Scope,

include/cuco/detail/static_multiset/static_multiset_ref.inl

@@ -554,7 +554,7 @@ class operator_impl<
    * slot contents to `output_match`, respectively. The output order is unspecified.
    *
    * Behavior is undefined if the size of the output range exceeds the number of retrieved slots.
-   * Use `count()` to determine the size of the output range.
+   * Use `count_outer()` to determine the size of the output range.
    *
    * If a key `k` has no matches in the container, then `{key, empty_slot_sentinel}` will be added
    * to the output sequence.