doc reorder mr, stream to stream, mr (#9308)

Doc reorder mr, stream to stream, mr

Authors:
  - Karthikeyan (https://github.com/karthikeyann)

Approvers:
  - David Wendt (https://github.com/davidwendt)
  - MithunR (https://github.com/mythrocks)

URL: #9308

cpp/include/cudf/detail/copy_if_else.cuh

@@ -152,8 +152,8 @@ __launch_bounds__(block_size) __global__
  * @param filter      Function of type `FilterFn` which determines for index `i` where to get the
  *                    corresponding output value from
  * @param out_type    `cudf::data_type` of the returned column
- * @param mr          Device memory resource used to allocate the returned column's device memory
  * @param stream      CUDA stream used for device memory operations and kernel launches.
+ * @param mr          Device memory resource used to allocate the returned column's device memory
  * @return            A new column that contains the values from either `lhs` or `rhs` as determined
  *                    by `filter[i]`
  */

cpp/include/cudf/detail/gather.cuh

@@ -640,8 +640,8 @@ void gather_bitmask(table_view const& source,
  * use `DONT_CHECK` when they are certain that the gather_map contains only valid indices for
  * better performance. In case there are out-of-bound indices in the gather map, the behavior
  * is undefined. Defaults to `DONT_CHECK`.
- * @param[in] mr Device memory resource used to allocate the returned table's device memory
  * @param[in] stream CUDA stream used for device memory operations and kernel launches.
+ * @param[in] mr Device memory resource used to allocate the returned table's device memory
  * @return cudf::table Result of the gather
  */
 template <typename MapIterator>

cpp/include/cudf/detail/gather.hpp

@@ -56,8 +56,8 @@ enum class negative_index_policy : bool { ALLOWED, NOT_ALLOWED };
  * the behavior is undefined.
  * @param[in] negative_index_policy Interpret each negative index `i` in the
  * gathermap as the positive index `i+num_source_rows`.
- * @param[in] mr Device memory resource used to allocate the returned table's device memory
  * @param[in] stream CUDA stream used for device memory operations and kernel launches.
+ * @param[in] mr Device memory resource used to allocate the returned table's device memory
  * @return cudf::table Result of the gather
  */
 std::unique_ptr<table> gather(

cpp/include/cudf/detail/reduction_functions.hpp

@@ -33,8 +33,8 @@ namespace reduction {
  *
  * @param col input column to compute sum
  * @param output_dtype data type of return type and typecast elements of input column
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return Sum as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> sum(
@@ -52,8 +52,8 @@ std::unique_ptr<scalar> sum(
  *
  * @param col input column to compute minimum.
  * @param output_dtype data type of return type and typecast elements of input column
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return Minimum element as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> min(
@@ -71,8 +71,8 @@ std::unique_ptr<scalar> min(
  *
  * @param col input column to compute maximum.
  * @param output_dtype data type of return type and typecast elements of input column
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return Maximum element as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> max(
@@ -91,8 +91,8 @@ std::unique_ptr<scalar> max(
  *
  * @param col input column to compute any_of.
  * @param output_dtype data type of return type and typecast elements of input column
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return bool scalar if any of elements is true when typecasted to bool
  */
 std::unique_ptr<scalar> any(
@@ -111,8 +111,8 @@ std::unique_ptr<scalar> any(
  *
  * @param col input column to compute all_of.
  * @param output_dtype data type of return type and typecast elements of input column
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return bool scalar if all of elements is true when typecasted to bool
  */
 std::unique_ptr<scalar> all(
@@ -131,8 +131,8 @@ std::unique_ptr<scalar> all(
  *
  * @param col input column to compute product.
  * @param output_dtype data type of return type and typecast elements of input column
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return Product as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> product(
@@ -151,8 +151,8 @@ std::unique_ptr<scalar> product(
  *
  * @param col input column to compute sum of squares.
  * @param output_dtype data type of return type and typecast elements of input column
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return Sum of squares as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> sum_of_squares(
@@ -171,8 +171,8 @@ std::unique_ptr<scalar> sum_of_squares(
  *
  * @param col input column to compute mean.
  * @param output_dtype data type of return type and typecast elements of input column.
- * @param mr Device memory resource used to allocate the returned scalar's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory.
  * @return Mean as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> mean(
@@ -191,8 +191,8 @@ std::unique_ptr<scalar> mean(
  *
  * @param col input column to compute variance.
  * @param output_dtype data type of return type and typecast elements of input column.
- * @param mr Device memory resource used to allocate the returned scalar's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory.
  * @return Variance as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> variance(
@@ -212,8 +212,8 @@ std::unique_ptr<scalar> variance(
  *
  * @param col input column to compute standard deviation.
  * @param output_dtype data type of return type and typecast elements of input column.
- * @param mr Device memory resource used to allocate the returned scalar's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory.
  * @return Standard deviation as scalar of type `output_dtype`.
  */
 std::unique_ptr<scalar> standard_deviation(
@@ -243,8 +243,8 @@ std::unique_ptr<scalar> standard_deviation(
  * @param col input column to get nth element from.
  * @param n index of element to get
  * @param null_handling Indicates if null values will be counted while indexing.
- * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned scalar's device memory
  * @return nth element as scalar
  */
 std::unique_ptr<scalar> nth_element(

cpp/include/cudf/dictionary/detail/concatenate.hpp

@@ -32,8 +32,8 @@ namespace detail {
  * @throw cudf::logic_error if dictionary column keys are not all the same type.
  *
  * @param columns Vector of dictionary columns to concatenate.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New column with concatenated results.
  */
 std::unique_ptr<column> concatenate(

cpp/include/cudf/dictionary/detail/update_keys.hpp

@@ -98,8 +98,8 @@ std::vector<std::unique_ptr<column>> match_dictionaries(
  * Any null rows are left unchanged.
  *
  * @param input Vector of cudf::table_views that include dictionary columns to be matched.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New dictionary columns and updated cudf::table_views.
  */
 std::pair<std::vector<std::unique_ptr<column>>, std::vector<table_view>> match_dictionaries(

cpp/include/cudf/strings/detail/concatenate.hpp

@@ -37,8 +37,8 @@ namespace detail {
  * ```
  *
  * @param columns List of string columns to concatenate.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New column with concatenated results.
  */
 std::unique_ptr<column> concatenate(

cpp/include/cudf/strings/detail/copy_if_else.cuh

@@ -47,8 +47,8 @@ namespace detail {
  * @param rhs_begin Strings of second set of data. Used when filter_fn returns false.
  * @param filter_fn Called to determine which iterator (lhs or rhs) to retrieve an entry for a
  * specific row.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New strings column.
  */
 template <typename StringPairIterLeft, typename StringPairIterRight, typename Filter>

cpp/include/cudf/strings/detail/copy_range.cuh

@@ -93,8 +93,8 @@ namespace detail {
  * @param target_begin The starting index of the target range (inclusive)
  * @param target_end The index of the last element in the target range
  * (exclusive)
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return std::unique_ptr<column> The result target column
  */
 template <typename SourceValueIterator, typename SourceValidityIterator>

cpp/include/cudf/strings/detail/fill.hpp

@@ -37,8 +37,8 @@ namespace detail {
  * @param begin First row index to include the new string.
  * @param end Last row index (exclusive).
  * @param value String to use when filling the range.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New strings column.
  */
 std::unique_ptr<column> fill(

cpp/include/cudf/strings/detail/gather.cuh

@@ -213,8 +213,8 @@ __global__ void gather_chars_fn_char_parallel(StringIterator strings_begin,
  * @param map_end End of index iterator.
  * @param offsets The offset values to be associated with the output chars column.
  * @param chars_bytes The total number of bytes for the output chars column.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New chars column fit for a strings column.
  */
 template <typename StringIterator, typename MapIterator>
@@ -278,8 +278,8 @@ std::unique_ptr<cudf::column> gather_chars(StringIterator strings_begin,
  * @param strings Strings instance for this operation.
  * @param begin Start of index iterator.
  * @param end End of index iterator.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New strings column containing the gathered strings.
  */
 template <bool NullifyOutOfBounds, typename MapIterator>
@@ -361,8 +361,8 @@ std::unique_ptr<cudf::column> gather(
  * @param begin Start of index iterator.
  * @param end End of index iterator.
  * @param nullify_out_of_bounds If true, indices outside the column's range are nullified.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New strings column containing the gathered strings.
  */
 template <typename MapIterator>

cpp/include/cudf/strings/detail/merge.cuh

@@ -39,8 +39,8 @@ namespace detail {
  * @param lhs First column.
  * @param rhs Second column.
  * @param row_order Indexes for each column.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New strings column.
  */
 template <typename index_type, typename row_order_iterator>

cpp/include/cudf/strings/detail/scatter.cuh

@@ -48,8 +48,8 @@ namespace detail {
  * @param scatter_map Iterator of indices into the output column.
  * @param target The set of columns into which values from the source column
  *        are to be scattered.
- * @param mr Device memory resource used to allocate the returned column's device memory
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory
  * @return New strings column.
  */
 template <typename SourceIterator, typename MapIterator>

cpp/include/cudf/strings/detail/utilities.cuh

@@ -41,8 +41,8 @@ namespace detail {
  * @tparam Iterator Used as input to scan to set the offset values.
  * @param begin The beginning of the input sequence
  * @param end The end of the input sequence
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return offsets child column for strings column
  */
 template <typename InputIterator>
@@ -74,8 +74,8 @@ std::unique_ptr<column> make_offsets_child_column(
  * @tparam Iter Iterator type that returns string_view instances
  * @param strings_begin Iterator to the beginning of the string_view sequence
  * @param num_strings The number of string_view instances in the sequence
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return Child offsets column
  */
 template <typename Iter>
@@ -131,8 +131,8 @@ __device__ inline char* copy_string(char* buffer, const string_view& d_string)
  *        chars memory.
  * @param exec_size Number of rows for executing the `size_and_exec_fn` function.
  * @param strings_count Number of strings.
- * @param mr Device memory resource used to allocate the returned columns' device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned columns' device memory.
  * @return offsets child column and chars child column for a strings column
  */
 template <typename SizeAndExecuteFunction>
@@ -190,8 +190,8 @@ auto make_strings_children(
  *        After that, the d_offsets and d_chars are set and this is called again to fill in the
  *        chars memory.
  * @param strings_count Number of strings.
- * @param mr Device memory resource used to allocate the returned columns' device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned columns' device memory.
  * @return offsets child column and chars child column for a strings column
  */
 template <typename SizeAndExecuteFunction>

cpp/include/nvtext/detail/tokenize.hpp

@@ -30,8 +30,8 @@ namespace detail {
  * @param strings Strings column tokenize.
  * @param delimiter UTF-8 characters used to separate each string into tokens.
  *                  The default of empty string will separate tokens using whitespace.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New strings columns of tokens.
  */
 std::unique_ptr<cudf::column> tokenize(
@@ -46,8 +46,8 @@ std::unique_ptr<cudf::column> tokenize(
  *
  * @param strings Strings column to tokenize.
  * @param delimiters Strings used to separate individual strings into tokens.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New strings columns of tokens.
  */
 std::unique_ptr<cudf::column> tokenize(
@@ -63,8 +63,8 @@ std::unique_ptr<cudf::column> tokenize(
  * @param strings Strings column to use for this operation.
  * @param delimiter Strings used to separate each string into tokens.
  *                  The default of empty string will separate tokens using whitespace.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New INT32 column of token counts.
  */
 std::unique_ptr<cudf::column> count_tokens(
@@ -79,8 +79,8 @@ std::unique_ptr<cudf::column> count_tokens(
  *
  * @param strings Strings column to use for this operation.
  * @param delimiters Strings used to separate each string into tokens.
- * @param mr Device memory resource used to allocate the returned column's device memory.
  * @param stream CUDA stream used for device memory operations and kernel launches.
+ * @param mr Device memory resource used to allocate the returned column's device memory.
  * @return New INT32 column of token counts.
  */
 std::unique_ptr<cudf::column> count_tokens(