Merge branch 'prerelease_doc_update' into 'master'

Update info about hipCUB

See merge request amd/rocPRIM!116

doc/Doxyfile

@@ -782,7 +782,7 @@ INPUT                  = mainpage.dox \
                          intrinsicsmodule.dox \
                          hipcub.dox \
                          glossary.dox \
-                         ../library/rocprim/include
+                         ../rocprim/include/rocprim
 
 # This tag can be used to specify the character encoding of the source files
 # that doxygen parses. Internally doxygen uses the UTF-8 encoding. Doxygen uses

doc/hipcub.dox

@@ -25,18 +25,20 @@
  *     Not-supported features and differences:
  *
  *     * Functions, classes and macros which are not in the public API are not supported.
- *     * `BlockHistogram` block primitive.
- *     * From device-wide primitives only `DeviceRadixSort`, `DeviceReduce`, and `DeviceScan` are supported. It's planned to
- *     add support for all device-wide primitives except `DeviceSpmv`.
- *     * Fancy iterators.
+ *     * Device-wide primitives can't be called from kernels (dynamic parallelism is not supported in HIP/HC).
+ *     * From device-wide primitives `DeviceRadixSort`, `DeviceReduce`, `DeviceScan`, `DeviceHistogram` (limited support),
+ *     `DeviceSegmentedRadixSort`, and `DeviceSegmentedReduce` are supported. It's planned to add support for all
+ *     device-wide primitives except `DeviceSpmv`.
+ *     * Fancy iterators: `CacheModifiedInputIterator`, `CacheModifiedOutputIterator`, `TexObjInputIterator`, and `TexRefInputIterator`
+ *     are not supported. Support for `TexObjInputIterator` is planned.
  *     * Thread I/O:
  *          * `CacheLoadModifier`, `CacheStoreModifier` cache modifiers are not supported.
  *          * `ThreadLoad`, `ThreadStore` functions are not supported.
  *     * Storage management and debug functions:
  *          * `CachingDeviceAllocator` class is not supported.
  *          * `Debug`, `PtxVersion`, `SmVersion` functions and `CubDebug`, `CubDebugExit`, `_CubLog` macros are not supported.
  *     * Intrinsics:
- *          * `ThreadExit`, `ThreadTrap`, `PRMT` - not supported.
+ *          * `ThreadExit`, `ThreadTrap` - not supported.
  *          * Warp thread masks (when used) are 64-bit unsigned integers.
  *          * `member_mask` input argument is ignored in `WARP_*` functions.
  *          * Arguments `first_lane`, `last_lane`, and `member_mask` are ignored in `Shuffle*` functions.

rocprim/include/rocprim/block/block_histogram.hpp

@@ -34,6 +34,9 @@
 
 BEGIN_ROCPRIM_NAMESPACE
 
+/// \addtogroup blockmodule
+/// @{
+
 enum class block_histogram_algorithm
 {
     using_atomic,
@@ -139,4 +142,7 @@ class block_histogram
 
 END_ROCPRIM_NAMESPACE
 
+/// @}
+// end of group blockmodule
+
 #endif // ROCPRIM_BLOCK_BLOCK_HISTOGRAM_HPP_

rocprim/include/rocprim/device/device_select_hc.hpp

@@ -37,6 +37,9 @@ BEGIN_ROCPRIM_NAMESPACE
 /// \addtogroup devicemodule_hc
 /// @{
 
+namespace detail
+{
+
 #define ROCPRIM_DETAIL_HC_SYNC(name, size, start) \
     { \
         if(debug_synchronous) \
@@ -49,6 +52,8 @@ BEGIN_ROCPRIM_NAMESPACE
         } \
     }
 
+} // end detail namespace
+
 template<
     class InputIterator,
     class FlagIterator,

rocprim/include/rocprim/device/device_select_hip.hpp

@@ -103,8 +103,6 @@ void flag_unique_kernel(InputIterator input,
     );
 }
 
-} // end detail namespace
-
 #define ROCPRIM_DETAIL_HIP_SYNC_AND_RETURN_ON_ERROR(name, size, start) \
     { \
         if(error != hipSuccess) return error; \
@@ -119,6 +117,8 @@ void flag_unique_kernel(InputIterator input,
         } \
     }
 
+} // end detail namespace
+
 template<
     class InputIterator,
     class FlagIterator,