Hi all,
I’m trying to compute prefix sum using CUB library but I got bunch of errors, does anyone know how to run this?
my errors :
: error: class “cub::BlockScan<int, 1024, cub::BLOCK_SCAN_RAKING, 1, 1, 200>” has no member “SmemStorage”
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(250): here
inclusive.cu(82): error: identifier “BlockLoadVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(250): here
inclusive.cu(89): error: no instance of overloaded function “cub::BlockScan<T, BLOCK_DIM_X, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::ExclusiveSum [with T=int, BLOCK_DIM_X=1024, ALGORITHM=cub::BLOCK_SCAN_RAKING, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=200]” matches the argument list
argument types are: (, int [1], int [1], int)
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(250): here
inclusive.cu(95): error: identifier “BlockStoreVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=1024, ITEMS_PER_THREAD=1]”
(250): here
inclusive.cu(78): error: class “cub::BlockScan<int, 512, cub::BLOCK_SCAN_RAKING, 1, 1, 200>” has no member “SmemStorage”
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(251): here
inclusive.cu(82): error: identifier “BlockLoadVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(251): here
inclusive.cu(89): error: no instance of overloaded function “cub::BlockScan<T, BLOCK_DIM_X, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::ExclusiveSum [with T=int, BLOCK_DIM_X=512, ALGORITHM=cub::BLOCK_SCAN_RAKING, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=200]” matches the argument list
argument types are: (, int [2], int [2], int)
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(251): here
inclusive.cu(95): error: identifier “BlockStoreVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=512, ITEMS_PER_THREAD=2]”
(251): here
inclusive.cu(78): error: class “cub::BlockScan<int, 256, cub::BLOCK_SCAN_RAKING, 1, 1, 200>” has no member “SmemStorage”
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(252): here
inclusive.cu(82): error: identifier “BlockLoadVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(252): here
inclusive.cu(89): error: no instance of overloaded function “cub::BlockScan<T, BLOCK_DIM_X, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::ExclusiveSum [with T=int, BLOCK_DIM_X=256, ALGORITHM=cub::BLOCK_SCAN_RAKING, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=200]” matches the argument list
argument types are: (, int [4], int [4], int)
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(252): here
inclusive.cu(95): error: identifier “BlockStoreVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=256, ITEMS_PER_THREAD=4]”
(252): here
inclusive.cu(78): error: class “cub::BlockScan<int, 128, cub::BLOCK_SCAN_RAKING, 1, 1, 200>” has no member “SmemStorage”
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(253): here
inclusive.cu(82): error: identifier “BlockLoadVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(253): here
inclusive.cu(89): error: no instance of overloaded function “cub::BlockScan<T, BLOCK_DIM_X, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::ExclusiveSum [with T=int, BLOCK_DIM_X=128, ALGORITHM=cub::BLOCK_SCAN_RAKING, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=200]” matches the argument list
argument types are: (, int [8], int [8], int)
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(253): here
inclusive.cu(95): error: identifier “BlockStoreVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=128, ITEMS_PER_THREAD=8]”
(253): here
inclusive.cu(78): error: class “cub::BlockScan<int, 64, cub::BLOCK_SCAN_RAKING, 1, 1, 200>” has no member “SmemStorage”
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(254): here
inclusive.cu(82): error: identifier “BlockLoadVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(254): here
inclusive.cu(89): error: no instance of overloaded function “cub::BlockScan<T, BLOCK_DIM_X, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::ExclusiveSum [with T=int, BLOCK_DIM_X=64, ALGORITHM=cub::BLOCK_SCAN_RAKING, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=200]” matches the argument list
argument types are: (, int [16], int [16], int)
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(254): here
inclusive.cu(95): error: identifier “BlockStoreVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=64, ITEMS_PER_THREAD=16]”
(254): here
inclusive.cu(78): error: class “cub::BlockScan<int, 32, cub::BLOCK_SCAN_RAKING, 1, 1, 200>” has no member “SmemStorage”
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(255): here
inclusive.cu(82): error: identifier “BlockLoadVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(255): here
inclusive.cu(89): error: no instance of overloaded function “cub::BlockScan<T, BLOCK_DIM_X, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::ExclusiveSum [with T=int, BLOCK_DIM_X=32, ALGORITHM=cub::BLOCK_SCAN_RAKING, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=200]” matches the argument list
argument types are: (, int [32], int [32], int)
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(255): here
inclusive.cu(95): error: identifier “BlockStoreVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=32, ITEMS_PER_THREAD=32]”
(255): here
inclusive.cu(78): error: class “cub::BlockScan<int, 16, cub::BLOCK_SCAN_RAKING, 1, 1, 200>” has no member “SmemStorage”
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(256): here
inclusive.cu(82): error: identifier “BlockLoadVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(256): here
inclusive.cu(89): error: no instance of overloaded function “cub::BlockScan<T, BLOCK_DIM_X, ALGORITHM, BLOCK_DIM_Y, BLOCK_DIM_Z, PTX_ARCH>::ExclusiveSum [with T=int, BLOCK_DIM_X=16, ALGORITHM=cub::BLOCK_SCAN_RAKING, BLOCK_DIM_Y=1, BLOCK_DIM_Z=1, PTX_ARCH=200]” matches the argument list
argument types are: (, int [64], int [64], int)
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(256): here
inclusive.cu(95): error: identifier “BlockStoreVectorized” is undefined
detected during:
instantiation of “void BlockPrefixSumKernel<BLOCK_THREADS,ITEMS_PER_THREAD>(int *, int *, clock_t *) [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(183): here
instantiation of “void Test<BLOCK_THREADS,ITEMS_PER_THREAD>() [with BLOCK_THREADS=16, ITEMS_PER_THREAD=64]”
(256): here
28 errors detected in the compilation of “/tmp/tmpxft_0000d7bc_00000000-8_inclusive.cpp1.ii”.