In file included from /home/miki/del/ygg-brute/src/cuda/engine.cu:27:
/home/miki/del/ygg-brute/src/generic/scalar_mul.h:36:74: warning: backslash-newline at end of file
36 | #define load_fold_table(DST, SRC) \
|
/home/miki/del/ygg-brute/src/cuda/engine.cu(54): warning: variable "cuda::<unnamed>::l_3" was declared but never referenced
/home/miki/del/ygg-brute/src/cuda/engine.cu(60): warning: variable "cuda::<unnamed>::l_3_lb" was declared but never referenced
/home/miki/del/ygg-brute/src/cuda/engine.cu(96): warning: function "cuda::<unnamed>::dbg_point(const char *, const AffineNielsPoint &)" was declared but never referenced
/home/miki/del/ygg-brute/src/cuda/engine.cu(86): warning: function "cuda::<unnamed>::dbg_point(const char *, const EdwardsPoint &)" was declared but never referenced
ptxas info : 0 bytes gmem, 1024 bytes cmem[3]
ptxas info : Compiling entry function 'compute_address_kernel_v1' for 'sm_75'
ptxas info : Function properties for compute_address_kernel_v1
32 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 94 registers, 376 bytes cmem[0]
ptxas info : Compiling entry function 'batch_invert_kernel_2' for 'sm_75'
ptxas info : Function properties for batch_invert_kernel_2
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 64 registers, 376 bytes cmem[0]
ptxas info : Compiling entry function 'batch_invert_kernel_1' for 'sm_75'
ptxas info : Function properties for batch_invert_kernel_1
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 72 registers, 368 bytes cmem[0]
ptxas info : Compiling entry function 'batch_invert_kernel_0' for 'sm_75'
ptxas info : Function properties for batch_invert_kernel_0
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 62 registers, 376 bytes cmem[0]
ptxas info : Compiling entry function 'compute_uv_kernel_1' for 'sm_75'
ptxas info : Function properties for compute_uv_kernel_1
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 64 registers, 64 bytes smem, 424 bytes cmem[0]
ptxas info : Compiling entry function 'compute_uv_kernel_0' for 'sm_75'
ptxas info : Function properties for compute_uv_kernel_0
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 72 registers, 32 bytes smem, 392 bytes cmem[0]
ptxas info : Compiling entry function 'compute_midstate_kernel_1' for 'sm_75'
ptxas info : Function properties for compute_midstate_kernel_1
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 117 registers, 384 bytes cmem[0]
ptxas info : Compiling entry function 'compute_midstate_kernel_0' for 'sm_75'
ptxas info : Function properties for compute_midstate_kernel_0
64 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 124 registers, 24576 bytes smem, 408 bytes cmem[0]
ptxas info : Compiling entry function 'precompute_addends_kernel' for 'sm_75'
ptxas info : Function properties for precompute_addends_kernel
32 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 123 registers, 24576 bytes smem, 400 bytes cmem[0]
In file included from /home/miki/del/ygg-brute/src/cuda/engine.cu:27:
/home/miki/del/ygg-brute/src/generic/scalar_mul.h:36:74: warning: backslash-newline at end of file
36 | #define load_fold_table(DST, SRC) \
|
/home/miki/del/ygg-brute/src/cuda/engine.cu: In member function ‘virtual void cuda::_GLOBAL__N__41_tmpxft_0001dac7_00000000_7_engine_cpp1_ii_d855fbfe::CudaEngine::print_info()’:
/home/miki/del/ygg-brute/src/cuda/engine.cu:420:8: warning: format ‘%d’ expects argument of type ‘int’, but argument 2 has type ‘std::vector<cudaDeviceProp>::size_type’ {aka ‘long unsigned int’} [-Wformat=]
420 | printf("CUDA: %d devices\n", props.size());
| ^~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~
| |
| std::vector<cudaDeviceProp>::size_type {aka long unsigned int}
/home/miki/del/ygg-brute/src/cuda/engine.cu:424:8: warning: format ‘%d’ expects argument of type ‘int’, but argument 2 has type ‘size_t’ {aka ‘long unsigned int’} [-Wformat=]
424 | printf("Device #%d (%s)\n", i, prop.name);
| ^ ~~~~~~~~~~~~~~ ~
| |
| size_t {aka long unsigned int}
In file included from /usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/ext/string_conversions.h:41,
from /usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/bits/basic_string.h:6545,
from /usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/string:55,
from /home/miki/del/ygg-brute/src/opencl/../engine.hpp:10,
from /home/miki/del/ygg-brute/src/opencl/engine.hpp:9,
from /home/miki/del/ygg-brute/src/opencl/engine.cpp:7:
/usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/cstdlib:75:15: fatal error: stdlib.h: No such file or directory
75 | #include_next <stdlib.h>
| ^~~~~~~~~~
compilation terminated.
gmake[2]: *** [CMakeFiles/opencl-engine.dir/build.make:82: CMakeFiles/opencl-engine.dir/src/opencl/engine.cpp.o] Error 1
gmake[1]: *** [CMakeFiles/Makefile2:131: CMakeFiles/opencl-engine.dir/all] Error 2
gmake: *** [Makefile:114: all] Error 2
```
In file included from /home/miki/del/ygg-brute/src/cuda/engine.cu:27:
/home/miki/del/ygg-brute/src/generic/scalar_mul.h:36:74: warning: backslash-newline at end of file
36 | #define load_fold_table(DST, SRC) \
|
/home/miki/del/ygg-brute/src/cuda/engine.cu(54): warning: variable "cuda::<unnamed>::l_3" was declared but never referenced
/home/miki/del/ygg-brute/src/cuda/engine.cu(60): warning: variable "cuda::<unnamed>::l_3_lb" was declared but never referenced
/home/miki/del/ygg-brute/src/cuda/engine.cu(96): warning: function "cuda::<unnamed>::dbg_point(const char *, const AffineNielsPoint &)" was declared but never referenced
/home/miki/del/ygg-brute/src/cuda/engine.cu(86): warning: function "cuda::<unnamed>::dbg_point(const char *, const EdwardsPoint &)" was declared but never referenced
ptxas info : 0 bytes gmem, 1024 bytes cmem[3]
ptxas info : Compiling entry function 'compute_address_kernel_v1' for 'sm_75'
ptxas info : Function properties for compute_address_kernel_v1
32 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 94 registers, 376 bytes cmem[0]
ptxas info : Compiling entry function 'batch_invert_kernel_2' for 'sm_75'
ptxas info : Function properties for batch_invert_kernel_2
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 64 registers, 376 bytes cmem[0]
ptxas info : Compiling entry function 'batch_invert_kernel_1' for 'sm_75'
ptxas info : Function properties for batch_invert_kernel_1
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 72 registers, 368 bytes cmem[0]
ptxas info : Compiling entry function 'batch_invert_kernel_0' for 'sm_75'
ptxas info : Function properties for batch_invert_kernel_0
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 62 registers, 376 bytes cmem[0]
ptxas info : Compiling entry function 'compute_uv_kernel_1' for 'sm_75'
ptxas info : Function properties for compute_uv_kernel_1
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 64 registers, 64 bytes smem, 424 bytes cmem[0]
ptxas info : Compiling entry function 'compute_uv_kernel_0' for 'sm_75'
ptxas info : Function properties for compute_uv_kernel_0
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 72 registers, 32 bytes smem, 392 bytes cmem[0]
ptxas info : Compiling entry function 'compute_midstate_kernel_1' for 'sm_75'
ptxas info : Function properties for compute_midstate_kernel_1
0 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 117 registers, 384 bytes cmem[0]
ptxas info : Compiling entry function 'compute_midstate_kernel_0' for 'sm_75'
ptxas info : Function properties for compute_midstate_kernel_0
64 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 124 registers, 24576 bytes smem, 408 bytes cmem[0]
ptxas info : Compiling entry function 'precompute_addends_kernel' for 'sm_75'
ptxas info : Function properties for precompute_addends_kernel
32 bytes stack frame, 0 bytes spill stores, 0 bytes spill loads
ptxas info : Used 123 registers, 24576 bytes smem, 400 bytes cmem[0]
In file included from /home/miki/del/ygg-brute/src/cuda/engine.cu:27:
/home/miki/del/ygg-brute/src/generic/scalar_mul.h:36:74: warning: backslash-newline at end of file
36 | #define load_fold_table(DST, SRC) \
|
/home/miki/del/ygg-brute/src/cuda/engine.cu: In member function ‘virtual void cuda::_GLOBAL__N__41_tmpxft_0001dac7_00000000_7_engine_cpp1_ii_d855fbfe::CudaEngine::print_info()’:
/home/miki/del/ygg-brute/src/cuda/engine.cu:420:8: warning: format ‘%d’ expects argument of type ‘int’, but argument 2 has type ‘std::vector<cudaDeviceProp>::size_type’ {aka ‘long unsigned int’} [-Wformat=]
420 | printf("CUDA: %d devices\n", props.size());
| ^~~~~~~~~~~~~~~~~~~~ ~~~~~~~~~~~~
| |
| std::vector<cudaDeviceProp>::size_type {aka long unsigned int}
/home/miki/del/ygg-brute/src/cuda/engine.cu:424:8: warning: format ‘%d’ expects argument of type ‘int’, but argument 2 has type ‘size_t’ {aka ‘long unsigned int’} [-Wformat=]
424 | printf("Device #%d (%s)\n", i, prop.name);
| ^ ~~~~~~~~~~~~~~ ~
| |
| size_t {aka long unsigned int}
In file included from /usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/ext/string_conversions.h:41,
from /usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/bits/basic_string.h:6545,
from /usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/string:55,
from /home/miki/del/ygg-brute/src/opencl/../engine.hpp:10,
from /home/miki/del/ygg-brute/src/opencl/engine.hpp:9,
from /home/miki/del/ygg-brute/src/opencl/engine.cpp:7:
/usr/lib/gcc/x86_64-pc-linux-gnu/10.3.0/include/g++-v10/cstdlib:75:15: fatal error: stdlib.h: No such file or directory
75 | #include_next <stdlib.h>
| ^~~~~~~~~~
compilation terminated.
gmake[2]: *** [CMakeFiles/opencl-engine.dir/build.make:82: CMakeFiles/opencl-engine.dir/src/opencl/engine.cpp.o] Error 1
gmake[1]: *** [CMakeFiles/Makefile2:131: CMakeFiles/opencl-engine.dir/all] Error 2
gmake: *** [Makefile:114: all] Error 2
```
Could you please provide verbose cmake log (obtained by adding --verbose flag to the second build command, like cmake --build build --verbose)? Also, a quick internet search suggests that it might have something to do with -isystem gcc flag.
Could you please provide verbose cmake log (obtained by adding `--verbose` flag to the second build command, like `cmake --build build --verbose`)? Also, a quick internet search suggests that it might have something to do with `-isystem` gcc flag.
Thank you for the information! The problem seems to be that OpenCL C++ headers usually are in a separate package (on Gentoo it's dev-libs/clhpp). Better cmake error message and a note in README are added. Could you try the solution and tell if it works?
Thank you for the information! The problem seems to be that OpenCL C++ headers usually are in a separate package (on Gentoo it's `dev-libs/clhpp`). Better cmake error message and a note in README are added. Could you try the solution and tell if it works?
clhpp package does not contain CL/cl.hpp, required by ygg-brute.
After commit #dfba56d260, ygg-brute reauire CL/opencl.hpp. But dev-libs/clhpp does not contain this file too, and ygg-brute fails on configure stage (cmake -H. -Bbuild -DCMAKE_CUDA_ARCHITECTURES=75).
$ equery f dev-libs/clhpp
* Searching for clhpp in dev-libs ...
* Contents of dev-libs/clhpp-2.0.12:
/usr
/usr/include
/usr/include/CL
/usr/include/CL/cl2.hpp
/usr/share
/usr/share/doc
/usr/share/doc/clhpp-2.0.12
/usr/share/doc/clhpp-2.0.12/README.md.bz2
With patch in commit #dfba56d260, ygg-brute can not be configuread and compiled.
May you add tag, like "v0.1", to #ec13c30940 commit?
clhpp package does not contain CL/cl.hpp, required by ygg-brute.
After commit #dfba56d260, ygg-brute reauire CL/opencl.hpp. But dev-libs/clhpp does not contain this file too, and ygg-brute fails on configure stage (```cmake -H. -Bbuild -DCMAKE_CUDA_ARCHITECTURES=75```).
```
$ equery f dev-libs/clhpp
* Searching for clhpp in dev-libs ...
* Contents of dev-libs/clhpp-2.0.12:
/usr
/usr/include
/usr/include/CL
/usr/include/CL/cl2.hpp
/usr/share
/usr/share/doc
/usr/share/doc/clhpp-2.0.12
/usr/share/doc/clhpp-2.0.12/README.md.bz2
```
With patch in commit #dfba56d260, ygg-brute can not be configuread and compiled.
May you add tag, like "v0.1", to #ec13c30940 commit?
Switched from CL/opencl.hpp to CL/cl2.hpp, added info about backend selection (it's possible to disable cuda or opencl at build time) to README, added 0.1 tag to #c13c30940.
Switched from `CL/opencl.hpp` to `CL/cl2.hpp`, added info about backend selection (it's possible to disable cuda or opencl at build time) to README, added 0.1 tag to #c13c30940.
GeForce GTX 1650 Gentoo gcc 9.3.0/10.3.0 nvidia-cuda-sdk 11.2.0
The same error with gcc-9.3.0 and gcc-10.3.0.
Cnake log attached.
Could you please provide verbose cmake log (obtained by adding
--verbose
flag to the second build command, likecmake --build build --verbose
)? Also, a quick internet search suggests that it might have something to do with-isystem
gcc flag.If I add
include_directories("/opt/cuda/targets/x86_64-linux/include")
to CMakeLists.txt, then compilation finished successful.Thank you for the information! The problem seems to be that OpenCL C++ headers usually are in a separate package (on Gentoo it's
dev-libs/clhpp
). Better cmake error message and a note in README are added. Could you try the solution and tell if it works?clhpp package does not contain CL/cl.hpp, required by ygg-brute.
After commit #dfba56d260, ygg-brute reauire CL/opencl.hpp. But dev-libs/clhpp does not contain this file too, and ygg-brute fails on configure stage (
cmake -H. -Bbuild -DCMAKE_CUDA_ARCHITECTURES=75
).With patch in commit #dfba56d260, ygg-brute can not be configuread and compiled.
May you add tag, like "v0.1", to #ec13c30940 commit?
Switched from
CL/opencl.hpp
toCL/cl2.hpp
, added info about backend selection (it's possible to disable cuda or opencl at build time) to README, added 0.1 tag to #c13c30940.