- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
HI,
I was converting spmv_compare.cu from repo cub and dpct gave SIGSEGV.
dpct output:
>dpct --cuda-include-path="C:\nvcc\include" --extra-arg="-I../" --out-root=dpct spmv_compare.cu NOTE: Could not auto-detect compilation database for file 'spmv_compare.cu' in 'C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental' or any parent directory. Processing: C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:49: C:\Users\...\Downloads\cuda code\cub-1.8.0\test\test_util.h:395:36: warning: result of comparison of constant 4294967295 with expression of type 'volatile unsigned short' is always true [-Wtautological-constant-out-of-range-compare] ((bits >= 0xFC01) && (bits <= 0xFFFFFFFF))); ~~~~ ^ ~~~~~~~~~~ In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:41: C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\sparse_matrix.h:360:31: warning: 'sscanf' is deprecated: This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_WARNINGS. See online help for details. [-Wdeprecated-declarations] int nparsed = sscanf(line, "%d %d %d", &num_rows, &num_cols, &num_nonzeros); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:906:9: note: in instantiation of function template specialization 'RunTests<double, int>' requested here RunTests<double, int>(rcm_relabel, alpha, beta, mtx_filename, grid2d, grid3d, wheel, dense, timing_iterations, args); ^ C:\Program Files (x86)\Windows Kits\10\include\10.0.18362.0\ucrt\stdio.h:2262:20: note: 'sscanf' has been explicitly marked deprecated here _Check_return_ _CRT_INSECURE_DEPRECATE(sscanf_s) ^ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\vcruntime.h:255:55: note: expanded from macro '_CRT_INSECURE_DEPRECATE' #define _CRT_INSECURE_DEPRECATE(_Replacement) _CRT_DEPRECATE_TEXT( \ ^ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\vcruntime.h:245:47: note: expanded from macro '_CRT_DEPRECATE_TEXT' #define _CRT_DEPRECATE_TEXT(_Text) __declspec(deprecated(_Text)) ^ In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:41: C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\sparse_matrix.h:399:25: warning: 'sscanf' is deprecated: This function or variable may be unsafe. Consider using sscanf_s instead. To disable deprecation, use _CRT_SECURE_NO_WARNINGS. See online help for details. [-Wdeprecated-declarations] if (sscanf(line, "%lf", &val) != 1) ^ C:\Program Files (x86)\Windows Kits\10\include\10.0.18362.0\ucrt\stdio.h:2262:20: note: 'sscanf' has been explicitly marked deprecated here _Check_return_ _CRT_INSECURE_DEPRECATE(sscanf_s) ^ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\vcruntime.h:255:55: note: expanded from macro '_CRT_INSECURE_DEPRECATE' #define _CRT_INSECURE_DEPRECATE(_Replacement) _CRT_DEPRECATE_TEXT( \ ^ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\vcruntime.h:245:47: note: expanded from macro '_CRT_DEPRECATE_TEXT' #define _CRT_DEPRECATE_TEXT(_Text) __declspec(deprecated(_Text)) ^ In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:33: In file included from C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\map:8: C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\xtree:1971:23: error: no matching function for call to object of type 'const std::_Tree_comp_alloc<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::key_compare' (aka 'const OrderByLow<int>') if (_DEBUG_LT_PRED(this->_Getcomp(), this->_Key(_Pnode), _Keyval)) ^~~~~~~~~~~~~~~~ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\xutility:1093:38: note: expanded from macro '_DEBUG_LT_PRED' #define _DEBUG_LT_PRED(pred, x, y) pred(x, y) ^~~~ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\xtree:2011:50: note: in instantiation of function template specialization 'std::_Tree<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::_Eqrange<int>' requested here _Paircc _Ans(static_cast<const _Tree *>(this)->_Eqrange(_Keyval)); ^ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\xtree:1533:11: note: in instantiation of function template specialization 'std::_Tree<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::_Eqrange<int>' requested here return (_Eqrange(_Keyval)); ^ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\xtree:1401:20: note: in instantiation of member function 'std::_Tree<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::equal_range' requested here _Pairii _Where = equal_range(_Keyval); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\sparse_matrix.h:1161:27: note: in instantiation of member function 'std::_Tree<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::erase' requested here unlabeled.erase(vertex); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\sparse_matrix.h:1222:5: note: in instantiation of function template specialization 'RcmRelabel<double, int>' requested here RcmRelabel(matrix, relabel_indices); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:656:9: note: in instantiation of function template specialization 'RcmRelabel<double, int>' requested here RcmRelabel(csr_matrix, !g_quiet); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:830:5: note: in instantiation of function template specialization 'RunTest<double, int>' requested here RunTest( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:906:9: note: in instantiation of function template specialization 'RunTests<double, int>' requested here RunTests<double, int>(rcm_relabel, alpha, beta, mtx_filename, grid2d, grid3d, wheel, dense, timing_iterations, args); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\sparse_matrix.h:1080:10: note: candidate function not viable: 'this' argument has type 'const std::_Tree_comp_alloc<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::key_compare' (aka 'const OrderByLow<int>'), but method is not marked const bool operator()(const OffsetT &a, const OffsetT &b) ^ In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:33: In file included from C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\map:8: C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\xtree:1992:23: error: no matching function for call to object of type 'const std::_Tree_comp_alloc<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::key_compare' (aka 'const OrderByLow<int>') if (_DEBUG_LT_PRED(this->_Getcomp(), _Keyval, this->_Key(_Pnode))) ^~~~~~~~~~~~~~~~ C:\Program Files (x86)\Microsoft Visual Studio\2017\Community\VC\Tools\MSVC\14.16.27023\include\xutility:1093:38: note: expanded from macro '_DEBUG_LT_PRED' #define _DEBUG_LT_PRED(pred, x, y) pred(x, y) ^~~~ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\sparse_matrix.h:1080:10: note: candidate function not viable: 'this' argument has type 'const std::_Tree_comp_alloc<std::_Tset_traits<int, OrderByLow<int>, std::allocator<int>, false> >::key_compare' (aka 'const OrderByLow<int>'), but method is not marked const bool operator()(const OffsetT &a, const OffsetT &b) ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:598:13: warning: format specifies type 'int' but the argument has type 'unsigned long long' [-Wformat] sizeof(ValueT) * 8, ^~~~~~~~~~~~~~~~~~ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:719:5: note: in instantiation of function template specialization 'DisplayPerf<double, int>' requested here DisplayPerf(device_giga_bandwidth, avg_millis, csr_matrix); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:830:5: note: in instantiation of function template specialization 'RunTest<double, int>' requested here RunTest( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:906:9: note: in instantiation of function template specialization 'RunTests<double, int>' requested here RunTests<double, int>(rcm_relabel, alpha, beta, mtx_filename, grid2d, grid3d, wheel, dense, timing_iterations, args); ^ In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:46: In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\device_spmv.cuh:41: In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\dispatch\dispatch_spmv_orig.cuh:41: In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\agent\agent_segment_fixup.cuh:39: C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:172:37: warning: expression does not compute the number of elements in this array; element type is 'cub::KeyValuePair<int, double>', not 'DeviceWord' (aka 'unsigned long long') [-Wsizeof-array-div] TOTAL_WORDS = sizeof(items) / sizeof(DeviceWord), ~~~~~ ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:765:13: note: in instantiation of function template specialization 'cub::InternalLoadDirectBlockedVectorized<cub::LOAD_DEFAULT, cub::KeyValuePair<int, double>, 4>' requested here InternalLoadDirectBlockedVectorized<MODIFIER>(linear_tid, block_itr.ptr, items); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:1134:48: note: in instantiation of function template specialization 'cub::BlockLoad<cub::KeyValuePair<int, double>, 128, 4, cub::BLOCK_LOAD_VECTORIZE, 1, 1, 0>::LoadInternal<cub::BLOCK_LOAD_VECTORIZE, 0>::Load<cub::LOAD_DEFAULT, cub::KeyValuePair<int, double>, int>' requested here InternalLoad(temp_storage, linear_tid).Load(block_itr, items); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\agent\agent_segment_fixup.cuh:284:53: note: in instantiation of function template specialization 'cub::BlockLoad<cub::KeyValuePair<int, double>, 128, 4, cub::BLOCK_LOAD_VECTORIZE, 1, 1, 0>::Load<cub::CacheModifiedInputIterator<cub::LOAD_DEFAULT, cub::KeyValuePair<int, double>, int> >' requested here BlockLoadPairs(temp_storage.load_pairs).Load(d_pairs_in + tile_offset, pairs); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\agent\agent_segment_fixup.cuh:361:13: note: in instantiation of function template specialization 'cub::AgentSegmentFixup<cub::DispatchSpmv<double, int>::PtxSegmentFixupPolicy, cub::KeyValuePair<int, double> *, double *, cub::Equality, cub::Sum, int>::ConsumeTile<false>' requested here ConsumeTile<false>(num_remaining, tile_idx, tile_offset, tile_state, Int2Type<USE_ATOMIC_FIXUP>()); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\dispatch\dispatch_spmv_orig.cuh:220:97: note: in instantiation of member function 'cub::AgentSegmentFixup<cub::DispatchSpmv<double, int>::PtxSegmentFixupPolicy, cub::KeyValuePair<int, double> *, double *, cub::Equality, cub::Sum, int>::ConsumeRange' requested here AgentSegmentFixupT(temp_storage, d_pairs_in, d_aggregates_out, cub::Equality(), cub::Sum()).ConsumeRange( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\dispatch\dispatch_spmv_orig.cuh:820:17: note: (skipping 1 context in backtrace; use -ftemplate-backtrace-limit=0 to see all) DeviceSegmentFixupKernel<PtxSegmentFixupPolicy, KeyValuePairT*, ValueT*, OffsetT, ScanTileStateT>, ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\device_spmv.cuh:158:43: note: in instantiation of member function 'cub::DispatchSpmv<double, int>::Dispatch' requested here return DispatchSpmv<ValueT, int>::Dispatch( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:526:30: note: in instantiation of function template specialization 'cub::DeviceSpmv::CsrMV<double>' requested here CubDebugExit(DeviceSpmv::CsrMV( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:725:22: note: in instantiation of function template specialization 'TestGpuMergeCsrmv<double, int>' requested here avg_millis = TestGpuMergeCsrmv(vector_y_in, vector_y_out, params, timing_iterations); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:830:5: note: in instantiation of function template specialization 'RunTest<double, int>' requested here RunTest( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:906:9: note: in instantiation of function template specialization 'RunTests<double, int>' requested here RunTests<double, int>(rcm_relabel, alpha, beta, mtx_filename, grid2d, grid3d, wheel, dense, timing_iterations, args); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:165:14: note: array 'items' declared here T (&items)[ITEMS_PER_THREAD]) ///< [out] Data to load ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:172:37: note: place parentheses around the 'sizeof(DeviceWord)' expression to silence this warning TOTAL_WORDS = sizeof(items) / sizeof(DeviceWord), ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:598:13: warning: format specifies type 'int' but the argument has type 'unsigned long long' [-Wformat] sizeof(ValueT) * 8, ^~~~~~~~~~~~~~~~~~ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:719:5: note: in instantiation of function template specialization 'DisplayPerf<float, int>' requested here DisplayPerf(device_giga_bandwidth, avg_millis, csr_matrix); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:830:5: note: in instantiation of function template specialization 'RunTest<float, int>' requested here RunTest( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:910:9: note: in instantiation of function template specialization 'RunTests<float, int>' requested here RunTests<float, int>(rcm_relabel, alpha, beta, mtx_filename, grid2d, grid3d, wheel, dense, timing_iterations, args); ^ In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:46: In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\device_spmv.cuh:41: In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\dispatch\dispatch_spmv_orig.cuh:41: In file included from C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\agent\agent_segment_fixup.cuh:39: C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:172:37: warning: expression does not compute the number of elements in this array; element type is 'cub::KeyValuePair<int, float>', not 'DeviceWord' (aka 'unsigned int') [-Wsizeof-array-div] TOTAL_WORDS = sizeof(items) / sizeof(DeviceWord), ~~~~~ ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:765:13: note: in instantiation of function template specialization 'cub::InternalLoadDirectBlockedVectorized<cub::LOAD_DEFAULT, cub::KeyValuePair<int, float>, 4>' requested here InternalLoadDirectBlockedVectorized<MODIFIER>(linear_tid, block_itr.ptr, items); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:1134:48: note: in instantiation of function template specialization 'cub::BlockLoad<cub::KeyValuePair<int, float>, 128, 4, cub::BLOCK_LOAD_VECTORIZE, 1, 1, 0>::LoadInternal<cub::BLOCK_LOAD_VECTORIZE, 0>::Load<cub::LOAD_DEFAULT, cub::KeyValuePair<int, float>, int>' requested here InternalLoad(temp_storage, linear_tid).Load(block_itr, items); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\agent\agent_segment_fixup.cuh:284:53: note: in instantiation of function template specialization 'cub::BlockLoad<cub::KeyValuePair<int, float>, 128, 4, cub::BLOCK_LOAD_VECTORIZE, 1, 1, 0>::Load<cub::CacheModifiedInputIterator<cub::LOAD_DEFAULT, cub::KeyValuePair<int, float>, int> >' requested here BlockLoadPairs(temp_storage.load_pairs).Load(d_pairs_in + tile_offset, pairs); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\agent\agent_segment_fixup.cuh:361:13: note: in instantiation of function template specialization 'cub::AgentSegmentFixup<cub::DispatchSpmv<float, int>::PtxSegmentFixupPolicy, cub::KeyValuePair<int, float> *, float *, cub::Equality, cub::Sum, int>::ConsumeTile<false>' requested here ConsumeTile<false>(num_remaining, tile_idx, tile_offset, tile_state, Int2Type<USE_ATOMIC_FIXUP>()); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\dispatch\dispatch_spmv_orig.cuh:220:97: note: in instantiation of member function 'cub::AgentSegmentFixup<cub::DispatchSpmv<float, int>::PtxSegmentFixupPolicy, cub::KeyValuePair<int, float> *, float *, cub::Equality, cub::Sum, int>::ConsumeRange' requested here AgentSegmentFixupT(temp_storage, d_pairs_in, d_aggregates_out, cub::Equality(), cub::Sum()).ConsumeRange( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\dispatch\dispatch_spmv_orig.cuh:820:17: note: (skipping 1 context in backtrace; use -ftemplate-backtrace-limit=0 to see all) DeviceSegmentFixupKernel<PtxSegmentFixupPolicy, KeyValuePairT*, ValueT*, OffsetT, ScanTileStateT>, ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\device\device_spmv.cuh:158:43: note: in instantiation of member function 'cub::DispatchSpmv<float, int>::Dispatch' requested here return DispatchSpmv<ValueT, int>::Dispatch( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:526:30: note: in instantiation of function template specialization 'cub::DeviceSpmv::CsrMV<float>' requested here CubDebugExit(DeviceSpmv::CsrMV( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:725:22: note: in instantiation of function template specialization 'TestGpuMergeCsrmv<float, int>' requested here avg_millis = TestGpuMergeCsrmv(vector_y_in, vector_y_out, params, timing_iterations); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:830:5: note: in instantiation of function template specialization 'RunTest<float, int>' requested here RunTest( ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\experimental\spmv_compare.cu:910:9: note: in instantiation of function template specialization 'RunTests<float, int>' requested here RunTests<float, int>(rcm_relabel, alpha, beta, mtx_filename, grid2d, grid3d, wheel, dense, timing_iterations, args); ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:165:14: note: array 'items' declared here T (&items)[ITEMS_PER_THREAD]) ///< [out] Data to load ^ C:\Users\...\Downloads\cuda code\cub-1.8.0\cub\block\block_load.cuh:172:37: note: place parentheses around the 'sizeof(DeviceWord)' expression to silence this warning TOTAL_WORDS = sizeof(items) / sizeof(DeviceWord), ^ Meet signal:SIGSEGV Intel(R) DPC++ Compatibility Tool tries to give analysis reports and terminates...
Regards,
Gagan
- Tags:
- Bug
Link Copied
3 Replies
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
Hi Gagan,
We are also getting the same errors while trying to migrate the cuda file provided by you. So we are escalating this to the concerned team.
Thanks for reporting the issue.
Warm Regards,
Abhishek
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
This issue is fixed in Beta08. Please confirm.
- Mark as New
- Bookmark
- Subscribe
- Mute
- Subscribe to RSS Feed
- Permalink
- Report Inappropriate Content
This issue has been resolved and we will no longer respond to this thread. If you require additional assistance from Intel, please start a new thread. Any further interaction in this thread will be considered community only
![](/skins/images/C4F7392A5C68ACA60290C9576F107044/responsive_peak/images/icon_anonymous_message.png)
Reply
Topic Options
- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page