Dlib: Fail to compile dlib with cuda 10.1
Hi,
Tried to compile dlib 19.16 with cuda 10.1, and got the following error during running cmake:
-- Found CUDA: /usr/local/cuda (found suitable version "10.1", minimum required is "7.5")
-- Looking for cuDNN install...
-- Found cuDNN: /usr/local/cuda/lib64/libcudnn.so
-- Building a CUDA test project to see if your compiler is compatible with CUDA...
-- * CUDA was found but your compiler failed to compile a simple CUDA program so dlib isn't going to use CUDA.
-- The output of the failed CUDA test compile is shown below:
-- Change Dir: /home/ron/DeepLearning/dlib-19.16/build/dlib/cuda_test_build
Run Build Command:"/usr/bin/gmake"
[ 50%] Building NVCC (Device) object CMakeFiles/cuda_test.dir/cuda_test_generated_cuda_test.cu.o
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘dlib::matrix
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1774:36: error: cannot call member function ‘long int dlib::matrix
~^~~~
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘const dlib::matrix
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/../optimization/optimization_line_search.h:209:23: required from here
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:14: error: cannot call member function ‘long int dlib::matrix
DLIB_CASSERT(r < m->nr() && c < m->nc(),
^ ~~
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:30: error: cannot call member function ‘long int dlib::matrix
~^~~
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:893: error: cannot call member function ‘long int dlib::matrix
^
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:929: error: cannot call member function ‘long int dlib::matrix
DLIB_CASSERT(r < m->nr() && c < m->nc(),
^
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘dlib::matrix
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1774:36: error: cannot call member function ‘long int dlib::matrix
~^~
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘const dlib::matrix
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:14: error: cannot call member function ‘long int dlib::matrix
^ ~~ /home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:30: error: cannot call member function ‘long int dlib::matrix
DLIB_CASSERT(r < m->nr() && c < m->nc(),
~^~~~~
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:893: error: cannot call member function ‘long int dlib::matrix
^
/home/ron/DeepLearning/dlib-19.16/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:929: error: cannot call member function ‘long int dlib::matrix
DLIB_CASSERT(r < m->nr() && c < m->nc(),
^
CMake Error at cuda_test_generated_cuda_test.cu.o.cmake:275 (message):
Error generating file
/home/ron/DeepLearning/dlib-19.16/build/dlib/cuda_test_build/CMakeFiles/cuda_test.dir//./cuda_test_generated_cuda_test.cu.o
gmake[2]: [CMakeFiles/cuda_test.dir/build.make:65: CMakeFiles/cuda_test.dir/cuda_test_generated_cuda_test.cu.o] Error 1
gmake[1]: [CMakeFiles/Makefile2:72: CMakeFiles/cuda_test.dir/all] Error 2
gmake: [Makefile:84: all] Error 2
*
Any comment?
Thanks.
Ron.
daisyron
All 23 comments
same error...
yuaowang
on 2 Mar 2019
I just tried cuda 10.1 and the only issue I found was on some versions of cmake it doesn't find cusolver. I just pushed a fix for that, but I didn't see any other issues. There is probably just something wrong with your cuda install unrelated to dlib.
Make sure to follow the official cuda install steps on nvidia's website. A lot of people complain about cuda being hard to install, but I never have any issues installing cuda on any platforms. However, I never install it any other way than via the official install instructions. I've seen lots of crazy stuff on the internet for hacking cuda into something in a non-supported way. Maybe that's the issue here.
davisking
on 2 Mar 2019
Both of you need to fill out the issue template. I don't even know what systems you are trying to build on. Maybe you are trying to use cuda on an OS NVIDIA doesn't support. I have no idea.
davisking
on 2 Mar 2019
Hi,
Actually it seems that my cuda installation is OK - I can compile and run cuda programs I have without any issue.
The problem I described occurs in the cmake stage (cmake version 3.12.1), where he tries to compile a "simple" cuda program (with includes from dlib) - does it relate to the issue you found?
My system configuration:
dlib version:
19.16
where from:
official download (from web site)
version: (Fedora 29, 'lsb_release -a'):
LSB Version: :core-4.1-amd64:core-4.1-noarch
Distributor ID: Fedora
Description: Fedora release 29 (Twenty Nine)
Release: 29
Codename: TwentyNine
Compiler:
g++ (GCC) 8.3.1 20190223 (Red Hat 8.3.1-2)
CMAKE:
cmake version 3.12.1
Thanks,
Ron.
daisyron
on 2 Mar 2019
Update: Problem repeated also by taking the repository head (git clone).
daisyron
on 2 Mar 2019
Hi, thank you for your help.
Here is my system:
- Distributor ID: Ubuntu
- Description: Ubuntu 18.04.2 LTS
- Release: 18.04
Codename: bionic
CMake version: 3.10.2
dlib version: 19.16.0
cuda version: cuda-10.1
g++ (Ubuntu 7.3.0-27ubuntu1~18.04) 7.3.0
Cheers,
Yuao
yuaowang
on 2 Mar 2019
Go try and build the dlib/cmake_utils/test_for_cuda test project directly
and see what it says.
davisking
on 2 Mar 2019
Same results :-(
Actually the issue is related to initialization of 2x2 matrix at optimization_line_search.h:
matrix<double,2,2> m;
matrix<double,2,1> v;
const double aa2 = x2*x2;
const double aa1 = x1*x1;
m = aa2, -aa1,
-aa2*x2, aa1*x1;
v = f_x1 - f0 - d0*x1,
f_x2 - f0 - d0*x2;
There is an issue initializing the m matrix...
Ron.
daisyron
on 2 Mar 2019
Expected Behavior
I expect dlib to compile successfully using CUDA 10.1
Current Behavior
Cmake configuration fails with CUDA was found but your compiler failed to compile a simple CUDA program so dlib isn't going to use CUDA.
The error is the same as OP's, but I'm pasting it here for completeness.
-- The C compiler identification is GNU 8.3.1
-- The CXX compiler identification is GNU 8.3.1
-- Check for working C compiler: /usr/bin/cc
-- Check for working C compiler: /usr/bin/cc -- works
-- Detecting C compiler ABI info
-- Detecting C compiler ABI info - done
-- Detecting C compile features
-- Detecting C compile features - done
-- Check for working CXX compiler: /usr/bin/c++
-- Check for working CXX compiler: /usr/bin/c++ -- works
-- Detecting CXX compiler ABI info
-- Detecting CXX compiler ABI info - done
-- Detecting CXX compile features
-- Detecting CXX compile features - done
-- Using CMake version: 3.12.1
-- Compiling dlib version: 19.16.99
-- Looking for pthread.h
-- Looking for pthread.h - found
-- Looking for pthread_create
-- Looking for pthread_create - not found
-- Looking for pthread_create in pthreads
-- Looking for pthread_create in pthreads - not found
-- Looking for pthread_create in pthread
-- Looking for pthread_create in pthread - found
-- Found Threads: TRUE
-- Looking for XOpenDisplay in /usr/lib64/libX11.so;/usr/lib64/libXext.so
-- Looking for XOpenDisplay in /usr/lib64/libX11.so;/usr/lib64/libXext.so - found
-- Looking for gethostbyname
-- Looking for gethostbyname - found
-- Looking for connect
-- Looking for connect - found
-- Looking for remove
-- Looking for remove - found
-- Looking for shmat
-- Looking for shmat - found
-- Looking for IceConnectionNumber in ICE
-- Looking for IceConnectionNumber in ICE - found
-- Found X11: /usr/lib64/libX11.so
-- Looking for png_create_read_struct
-- Looking for png_create_read_struct - found
-- Searching for BLAS and LAPACK
-- Searching for BLAS and LAPACK
-- Found PkgConfig: /usr/bin/pkg-config (found version "1.5.3")
-- Checking for module 'cblas'
-- Package 'cblas', required by 'virtual:world', not found
-- Checking for module 'lapack'
-- Package 'lapack', required by 'virtual:world', not found
-- Looking for sys/types.h
-- Looking for sys/types.h - found
-- Looking for stdint.h
-- Looking for stdint.h - found
-- Looking for stddef.h
-- Looking for stddef.h - found
-- Check size of void*
-- Check size of void* - done
-- Found OpenBLAS library
-- Looking for sgetrf_single
-- Looking for sgetrf_single - found
-- Using OpenBLAS's built in LAPACK
-- Looking for cblas_ddot
-- Looking for cblas_ddot - found
-- Looking for sgesv
-- Looking for sgesv - not found
-- Looking for sgesv_
-- Looking for sgesv_ - not found
-- Found CUDA: /usr/local/cuda (found suitable version "10.1", minimum required is "7.5")
-- Looking for cuDNN install...
-- Found cuDNN: /usr/lib64/libcudnn.so
-- Building a CUDA test project to see if your compiler is compatible with CUDA...
-- *****************************************************************************************************************
-- *** CUDA was found but your compiler failed to compile a simple CUDA program so dlib isn't going to use CUDA.
-- *** The output of the failed CUDA test compile is shown below:
-- *** Change Dir: /home/alex/Build/dlib/dlib_git/build/dlib/cuda_test_build
***
*** Run Build Command:"/usr/bin/gmake"
*** [ 50%] Building NVCC (Device) object CMakeFiles/cuda_test.dir/cuda_test_generated_cuda_test.cu.o
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::literal_assign_helper::~literal_assign_helper() [with T = double; long int num_rows = 2; long int num_cols = 2; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’:
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/../optimization/optimization_line_search.h:209:11: required from here
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1774:36: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nr() const [with T = double; long int num_rows = 2; long int num_cols = 2; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(!has_been_used || r == m->nr(),
*** ~~~~~^~~~~~~~
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘const dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::literal_assign_helper& dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::literal_assign_helper::operator,(const T&) const [with T = double; long int num_rows = 2; long int num_cols = 2; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’:
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/../optimization/optimization_line_search.h:209:23: required from here
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:14: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nr() const [with T = double; long int num_rows = 2; long int num_cols = 2; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ^ ~~~~
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:30: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nc() const [with T = double; long int num_rows = 2; long int num_cols = 2; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ~~~~~^~~~~~~
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:890: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nr() const [with T = double; long int num_rows = 2; long int num_cols = 2; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ^
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:926: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nc() const [with T = double; long int num_rows = 2; long int num_cols = 2; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ^
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::literal_assign_helper::~literal_assign_helper() [with T = double; long int num_rows = 2; long int num_cols = 1; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’:
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/../optimization/optimization_line_search.h:211:32: required from here
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1774:36: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nr() const [with T = double; long int num_rows = 2; long int num_cols = 1; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(!has_been_used || r == m->nr(),
*** ~~~~~^~~~~~~~
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h: In instantiation of ‘const dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::literal_assign_helper& dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::literal_assign_helper::operator,(const T&) const [with T = double; long int num_rows = 2; long int num_cols = 1; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’:
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/../optimization/optimization_line_search.h:211:61: required from here
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:14: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nr() const [with T = double; long int num_rows = 2; long int num_cols = 1; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ^ ~~~~
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:30: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nc() const [with T = double; long int num_rows = 2; long int num_cols = 1; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ~~~~~^~~~~~~
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:890: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nr() const [with T = double; long int num_rows = 2; long int num_cols = 1; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ^
*** /home/alex/Build/dlib/dlib_git/dlib/cmake_utils/test_for_cuda/../../cuda/../matrix/matrix.h:1783:926: error: cannot call member function ‘long int dlib::matrix<T, num_rows, num_cols, mem_manager, layout>::nc() const [with T = double; long int num_rows = 2; long int num_cols = 1; mem_manager = dlib::memory_manager_stateless_kernel_1<char>; layout = dlib::row_major_layout]’ without object
*** DLIB_CASSERT(r < m->nr() && c < m->nc(),
*** ^
*** CMake Error at cuda_test_generated_cuda_test.cu.o.cmake:275 (message):
*** Error generating file
*** /home/alex/Build/dlib/dlib_git/build/dlib/cuda_test_build/CMakeFiles/cuda_test.dir//./cuda_test_generated_cuda_test.cu.o
***
***
*** gmake[2]: *** [CMakeFiles/cuda_test.dir/build.make:65: CMakeFiles/cuda_test.dir/cuda_test_generated_cuda_test.cu.o] Error 1
*** gmake[1]: *** [CMakeFiles/Makefile2:72: CMakeFiles/cuda_test.dir/all] Error 2
*** gmake: *** [Makefile:84: all] Error 2
***
-- *****************************************************************************************************************
-- Disabling CUDA support for dlib. DLIB WILL NOT USE CUDA
-- C++11 activated.
-- Configuring done
-- Generating done
-- Build files have been written to: /home/alex/Build/dlib/dlib_git/build
Steps to Reproduce
- Git clone dlib and cd into the project.
mkdir build && cd buildcmake ..
Version:
Git HEAD 70f9a9f11fb2ae82a057e5f4b3bae9238087e418Where did you get dlib:
Github.Platform:
$ lsb_release -a
LSB Version: :core-4.1-amd64:core-4.1-noarch:cxx-4.1-amd64:cxx-4.1-noarch:desktop-4.1-amd64:desktop-4.1-noarch:languages-4.1-amd64:languages-4.1-noarch:printing-4.1-amd64:printing-4.1-noarch
Distributor ID: Fedora
Description: Fedora release 29 (Twenty Nine)
Release: 29
Codename: TwentyNine
$ uname -a
Linux REDACTED 4.19.12-301.fc29.x86_64 #1 SMP Mon Dec 24 01:58:57 UTC 2018 x86_64 x86_64 x86_64 GNU/Linux
- Compiler:
gcc (GCC) 8.3.1 20190223 (Red Hat 8.3.1-2)
alexmarkley
on 2 Mar 2019
The following patch allowed the test program to compile successfully:
diff --git a/dlib/matrix/matrix.h b/dlib/matrix/matrix.h
index 224affb6..cdfbe891 100644
--- a/dlib/matrix/matrix.h
+++ b/dlib/matrix/matrix.h
@@ -1771,7 +1771,7 @@ namespace dlib
explicit literal_assign_helper(matrix* m_): m(m_), r(0), c(0),has_been_used(false) {next();}
~literal_assign_helper() noexcept(false)
{
- DLIB_CASSERT(!has_been_used || r == m->nr(),
+ DLIB_CASSERT(!has_been_used || r == (*m).nr(),
"You have used the matrix comma based assignment incorrectly by failing to\n"
"supply a full set of values for every element of a matrix object.\n");
}
@@ -1780,14 +1780,14 @@ namespace dlib
const T& val
) const
{
- DLIB_CASSERT(r < m->nr() && c < m->nc(),
+ DLIB_CASSERT(r < (*m).nr() && c < (*m).nc(),
"You have used the matrix comma based assignment incorrectly by attempting to\n" <<
"supply more values than there are elements in the matrix object being assigned to.\n\n" <<
"Did you forget to call set_size()?"
<< "\n\t r: " << r
<< "\n\t c: " << c
- << "\n\t m->nr(): " << m->nr()
- << "\n\t m->nc(): " << m->nc());
+ << "\n\t m->nr(): " << (*m).nr()
+ << "\n\t m->nc(): " << (*m).nc());
(*m)(r,c) = val;
next();
has_been_used = true;
@@ -1802,7 +1802,7 @@ namespace dlib
) const
{
++c;
- if (c == m->nc())
+ if (c == (*m).nc())
{
c = 0;
++r;
I don't know if this is some GCC 8 weirdness or what. We'll see if the patch works or if this just makes everything worse.
alexmarkley
on 2 Mar 2019
It worked. The patch I created above allowed me to successfully compile and install Dlib. I've also tested it with my workload and it appears to be working.
alexmarkley
on 2 Mar 2019
Tried it and it worked for me as well :-)
daisyron
on 2 Mar 2019
Sorry to spam the issue, but I was able to reproduce the same problem on another system:
$ lsb_release -a
No LSB modules are available.
Distributor ID: Ubuntu
Description: Ubuntu 18.04.2 LTS
Release: 18.04
Codename: bionic
$ uname -a
Linux REDACTED 4.15.0-45-generic #48-Ubuntu SMP Tue Jan 29 16:28:13 UTC 2019 x86_64 x86_64 x86_64 GNU/Linux
$ gcc --version
gcc (Ubuntu 7.3.0-27ubuntu1~18.04) 7.3.0
Copyright (C) 2017 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
My patch worked for this system as well.
alexmarkley
on 2 Mar 2019
WOW, it worked for my system! Cheers.
yuaowang
on 3 Mar 2019
@alexmarkley wow, that's a weird compiler bug. Thanks for figuring it out :) Do you want to submit a PR for it?
davisking
on 3 Mar 2019
Solved my problem too. Thanks!
qc2105
on 3 Mar 2019
I just tried cuda 10.1 and the only issue I found was on some versions of cmake it doesn't find cusolver. I just pushed a fix for that, but I didn't see any other issues. There is probably just something wrong with your cuda install unrelated to dlib.
This solved my recent dlib setup problem (since last Thursday).
Just to for the record my platform:
Distributor ID: Ubuntu
Description: Ubuntu 16.04.4 LTS
Release: 16.04
Code name: xenial
CMake version: 3.5.1
dlib version: 19.16.99
cuda version: cuda-10.1
g++ (Ubuntu 5.4.0-6ubuntu1~16.04) 5.4.0
pavelpe
on 3 Mar 2019
Thanks everyone for reporting this. I just pushed the fix into master on github. I've also been able to reproduce this myself, although only on Ubuntu 18.04. Oddly never able to reproduce it on Ubuntu 16.04. But in any case, it should be fixed now.
davisking
on 3 Mar 2019
just wanna to help, cmake version 3.14 rc2 solved this for me, it seems cmake 3.12 doesn't recognize cuda 10.1
Christianwidjaya
on 4 Mar 2019
Warning: this issue has been inactive for 30 days and will be automatically closed on 2019-04-08 if there is no further activity.
If you are waiting for a response but haven't received one it's possible your question is somehow inappropriate. E.g. it is off topic, you didn't follow the issue submission instructions, or your question is easily answerable by reading the FAQ, dlib's official compilation instructions, dlib's API documentation, or a Google search.
dlib-issue-bot
on 3 Apr 2019
Warning: this issue has been inactive for 37 days and will be automatically closed on 2019-04-18 if there is no further activity.
If you are waiting for a response but haven't received one it's possible your question is somehow inappropriate. E.g. it is off topic, you didn't follow the issue submission instructions, or your question is easily answerable by reading the FAQ, dlib's official compilation instructions, dlib's API documentation, or a Google search.
dlib-issue-bot
on 10 Apr 2019
Notice: this issue has been closed because it has been inactive for 45 days. You may reopen this issue if it has been closed in error.
dlib-issue-bot
on 18 Apr 2019
just wanna to help, cmake version 3.14 rc2 solved this for me, it seems cmake 3.12 doesn't recognize cuda 10.1
Try
Using CMake version: 3.16.8
But problem not solved :(
pnzr00t
on 21 Jun 2020
Related issues
joeking11829
·
5Comments
lvella
·
4Comments
great-thoughts
·
5Comments
yourmailhacked
·
3Comments
abhisheksoni27
·
4Comments
Most helpful comment
The following patch allowed the test program to compile successfully:
I don't know if this is some GCC 8 weirdness or what. We'll see if the patch works or if this just makes everything worse.