Commit 76f98b74 authored by Jakub Klinkovský's avatar Jakub Klinkovský
Browse files

Fixed fallback condition in reductionOnCudaDevice

parent 28938623

src/TNL/Containers/Algorithms/Reduction_impl.h

+4 −4
Original line number Diff line number Diff line
@@ -56,13 +56,14 @@ reductionOnCudaDevice( Operation& operation, 
    * objects stored on the device might contain pointers into the device memory,

    * in which case reduction on host might fail.

    */

   constexpr bool can_copy_to_host = std::is_fundamental< RealType >::value || std::is_pointer< RealType >::value;

   constexpr bool can_reduce_all_on_host = std::is_fundamental< RealType >::value || std::is_pointer< RealType >::value;

   constexpr bool can_reduce_later_on_host = std::is_fundamental< ResultType >::value || std::is_pointer< ResultType >::value;



   /***

    * First check if the input array(s) is/are large enough for the reduction on GPU.

    * Otherwise copy it/them to host and reduce on CPU.

    */

   if( can_copy_to_host && size <= minGPUReductionDataSize )

   if( can_reduce_all_on_host && size <= minGPUReductionDataSize )

   {

      RealType hostArray1[ minGPUReductionDataSize ];

      RealType hostArray2[ minGPUReductionDataSize ];
@@ -99,7 +100,7 @@ reductionOnCudaDevice( Operation& operation, 
      timer.start();

   #endif



   if( can_copy_to_host ) {

   if( can_reduce_later_on_host ) {

      /***

       * Transfer the reduced data from device to host.

       */
@@ -133,7 +134,6 @@ reductionOnCudaDevice( Operation& operation, 
       */

      LaterReductionOperation laterReductionOperation;

      while( reducedSize > 1 ) {

         // TODO: copy the intermediate result somewhere else, in-place reduction probably does not work

         reducedSize = CudaReductionKernelLauncher( laterReductionOperation,

                                                    reducedSize,

                                                    deviceAux1,