Menu
▾
▴
viennacl-support — Main public mailing list for ViennaCL
You can subscribe to this list here.
| 2010 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(1) |
Jun
(8) |
Jul
(16) |
Aug
(6) |
Sep
|
Oct
|
Nov
|
Dec
(5) |
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2011 |
Jan
(4) |
Feb
(3) |
Mar
(5) |
Apr
|
May
(24) |
Jun
|
Jul
(5) |
Aug
(17) |
Sep
|
Oct
(6) |
Nov
(9) |
Dec
(8) |
| 2012 |
Jan
(5) |
Feb
(14) |
Mar
(25) |
Apr
(7) |
May
(15) |
Jun
(12) |
Jul
(22) |
Aug
(4) |
Sep
(10) |
Oct
(10) |
Nov
(19) |
Dec
(17) |
| 2013 |
Jan
(8) |
Feb
(10) |
Mar
(16) |
Apr
(3) |
May
(16) |
Jun
(26) |
Jul
|
Aug
(9) |
Sep
|
Oct
(8) |
Nov
(17) |
Dec
(2) |
| 2014 |
Jan
(37) |
Feb
(15) |
Mar
(6) |
Apr
(9) |
May
(11) |
Jun
(11) |
Jul
(9) |
Aug
(9) |
Sep
(19) |
Oct
(4) |
Nov
(22) |
Dec
(21) |
| 2015 |
Jan
|
Feb
(7) |
Mar
(2) |
Apr
(17) |
May
(22) |
Jun
(11) |
Jul
(11) |
Aug
(6) |
Sep
(7) |
Oct
|
Nov
(5) |
Dec
|
| 2016 |
Jan
(1) |
Feb
(3) |
Mar
(4) |
Apr
(8) |
May
(8) |
Jun
(11) |
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(6) |
| 2017 |
Jan
|
Feb
(1) |
Mar
(2) |
Apr
(19) |
May
|
Jun
(7) |
Jul
(7) |
Aug
(2) |
Sep
(6) |
Oct
|
Nov
(3) |
Dec
|
| 2018 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
(4) |
Oct
|
Nov
(2) |
Dec
|
| 2019 |
Jan
(2) |
Feb
|
Mar
(1) |
Apr
|
May
|
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
(3) |
| 2020 |
Jan
|
Feb
(4) |
Mar
|
Apr
|
May
|
Jun
|
Jul
(2) |
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
| 2021 |
Jan
|
Feb
|
Mar
|
Apr
|
May
|
Jun
(1) |
Jul
(2) |
Aug
|
Sep
(2) |
Oct
|
Nov
(31) |
Dec
(4) |
| 2024 |
Jan
|
Feb
|
Mar
|
Apr
|
May
(2) |
Jun
|
Jul
|
Aug
|
Sep
|
Oct
|
Nov
|
Dec
|
|
From: Karl R. <ru...@iu...> - 2024-05-09 12:03:50
|
Hello, an unknown OpenCL error is an indication that maybe your OpenCL installation isn't working correctly. Can you please verify that OpenCL works fine (i.e. outside of ViennaCL)? I'm not sure whether the problem has been resolved already - the stackoverflow page suggests so. Best regards, Karli On 5/8/24 17:41, Francisco Ossian wrote: > Hello, > > I hope this email finds you well. I am writing to seek assistance > regarding challenges I have encountered while integrating ViennaCL into > my project and configuring it to work with OpenCL. > > Initially, I was puzzled by a significant performance gap between GPU > and CPU executions when performing matrix multiplication operations > using the Eigen and ViennaCL libraries in C++. However, upon further > investigation, I discovered that ViennaCL needs to be explicitly > instructed to utilize OpenCL for GPU computations. > > In an attempt to address this issue, I added the necessary configuration > flags to my premake script, including defining VIENNACL_WITH_OPENCL, > linking to OpenCL libraries, and specifying library directories. > However, despite these efforts, I encountered compilation errors that I > have been unable to resolve.The error message I received reads as follows: > > terminate called after throwing an instance of > 'viennacl::ocl::unknown_error' > what(): ViennaCL: FATAL ERROR: ViennaCL encountered an unknown > OpenCL error. In some cases, this might be due to an invalid global work > size, but it can also be due to several compilation errors. > If you think that this is a bug in ViennaCL, please report it at > vie...@li... > <mailto:vie...@li...> and supply at least the > following information: > * Operating System > * Which OpenCL implementation (AMD, NVIDIA, etc.) > * ViennaCL version > > Despite my best efforts, I have been unable to extract relevant > information from the documentation to properly configure ViennaCL for my > project. I have reviewed example CMake configurations provided with > ViennaCL but have struggled to apply them to my premake setup. > > At this juncture, I find myself at an impasse and am reaching out to > seek guidance and assistance from the community. If you have any > insights, suggestions, or solutions regarding the correct configuration > of ViennaCL with OpenCL, or if you can offer assistance in resolving the > compilation errors I have encountered, I would greatly appreciate your > ideas. > > link to my question in stack: > https://stackoverflow.com/questions/78437721/performance-discrepancy-between-gpu-and-cpu-for-matrix-multiplication-eigen-vs <https://stackoverflow.com/questions/78437721/performance-discrepancy-between-gpu-and-cpu-for-matrix-multiplication-eigen-vs> > > Thank you very much for your time and assistance. I look forward to > hearing from you soon. > > Best regards, > > -- > Francisco Ossian Lopes Neto > > > _______________________________________________ > ViennaCL-support mailing list > Vie...@li... > https://lists.sourceforge.net/lists/listinfo/viennacl-support |
|
From: Francisco O. <oss...@gm...> - 2024-05-08 15:41:37
|
Hello, I hope this email finds you well. I am writing to seek assistance regarding challenges I have encountered while integrating ViennaCL into my project and configuring it to work with OpenCL. Initially, I was puzzled by a significant performance gap between GPU and CPU executions when performing matrix multiplication operations using the Eigen and ViennaCL libraries in C++. However, upon further investigation, I discovered that ViennaCL needs to be explicitly instructed to utilize OpenCL for GPU computations. In an attempt to address this issue, I added the necessary configuration flags to my premake script, including defining VIENNACL_WITH_OPENCL, linking to OpenCL libraries, and specifying library directories. However, despite these efforts, I encountered compilation errors that I have been unable to resolve.The error message I received reads as follows: terminate called after throwing an instance of 'viennacl::ocl::unknown_error' what(): ViennaCL: FATAL ERROR: ViennaCL encountered an unknown OpenCL error. In some cases, this might be due to an invalid global work size, but it can also be due to several compilation errors. If you think that this is a bug in ViennaCL, please report it at vie...@li... and supply at least the following information: * Operating System * Which OpenCL implementation (AMD, NVIDIA, etc.) * ViennaCL version Despite my best efforts, I have been unable to extract relevant information from the documentation to properly configure ViennaCL for my project. I have reviewed example CMake configurations provided with ViennaCL but have struggled to apply them to my premake setup. At this juncture, I find myself at an impasse and am reaching out to seek guidance and assistance from the community. If you have any insights, suggestions, or solutions regarding the correct configuration of ViennaCL with OpenCL, or if you can offer assistance in resolving the compilation errors I have encountered, I would greatly appreciate your ideas. link to my question in stack: https://stackoverflow.com/questions/78437721/performance-discrepancy-between-gpu-and-cpu-for-matrix-multiplication-eigen-vs Thank you very much for your time and assistance. I look forward to hearing from you soon. Best regards, -- Francisco Ossian Lopes Neto |
|
From: Karl R. <ru...@iu...> - 2021-12-02 14:06:35
|
Hey, because you can concatenate kernels in various ways through C++, there is no single list of kernels. For using QR, please refer to the following example: https://github.com/viennacl/viennacl-dev/blob/master/examples/tutorial/qr.cpp I don't know how well the matrix-matrix products will run on the board. Best regards, Karli On 12/2/21 14:57, SANDRA CATALAN PALLARES wrote: > Hi, > > Great! Then I can use it as it is. One more question, is there any > documentation with all the available kernels? I found some information > online, but I think I have seen more functions in use in the examples. > The thing is I want to compute a QR (dgeqrf from Lapack) on the GPU, but > I could not find the right function for now. > > Many thanks. > > Regards, > Sandra > > El jue, 2 dic 2021 a las 14:51, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > Hi Sandra, > > ViennaCL is a header-only library. So there isn't anything that > needs to > be installed separately. You can just use the GCC compilation line that > we identified as working and continue with that. The CMake environment > is just for building examples, but is not a requirement per se. > > Best regards, > Karli > > > On 12/2/21 14:38, SANDRA CATALAN PALLARES wrote: > > Hi Karl, > > > > I am back these days to this work. Could you confirm to me if I need > > anything else to work with the library? I am not sure if the > > installation is properly done after the emails we exchanged. > > > > Thanks again for your help. > > > > Regards, > > Sandra > > > > El vie, 26 nov 2021 a las 14:25, SANDRA CATALAN PALLARES > > (<sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>>) escribió: > > > > Hi, > > > > Ok. Then, is there anything left in order to make use of the > > library? I am a bit lost right now. > > > > Regards, > > Sandra > > > > El vie, 26 nov 2021 a las 13:58, Karl Rupp > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>) escribió: > > > > Hello, > > > > > > > They are the same compiler. Is that what you expected? > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ c++ > > --version > > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > > Copyright (C) 2013 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. > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > --version > > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > > Copyright (C) 2013 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. > > > > > > Okay, that's good. I was expecting this, yes. Better safe > than > > sorry ;-) > > > > Best regards, > > Karli > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... > <mailto:sca...@uc...> <mailto:sca...@uc... > <mailto:sca...@uc...>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> |
|
From: SANDRA C. P. <sca...@uc...> - 2021-12-02 13:57:30
|
Hi, Great! Then I can use it as it is. One more question, is there any documentation with all the available kernels? I found some information online, but I think I have seen more functions in use in the examples. The thing is I want to compute a QR (dgeqrf from Lapack) on the GPU, but I could not find the right function for now. Many thanks. Regards, Sandra El jue, 2 dic 2021 a las 14:51, Karl Rupp (<ru...@iu...>) escribió: > Hi Sandra, > > ViennaCL is a header-only library. So there isn't anything that needs to > be installed separately. You can just use the GCC compilation line that > we identified as working and continue with that. The CMake environment > is just for building examples, but is not a requirement per se. > > Best regards, > Karli > > > On 12/2/21 14:38, SANDRA CATALAN PALLARES wrote: > > Hi Karl, > > > > I am back these days to this work. Could you confirm to me if I need > > anything else to work with the library? I am not sure if the > > installation is properly done after the emails we exchanged. > > > > Thanks again for your help. > > > > Regards, > > Sandra > > > > El vie, 26 nov 2021 a las 14:25, SANDRA CATALAN PALLARES > > (<sca...@uc... <mailto:sca...@uc...>>) escribió: > > > > Hi, > > > > Ok. Then, is there anything left in order to make use of the > > library? I am a bit lost right now. > > > > Regards, > > Sandra > > > > El vie, 26 nov 2021 a las 13:58, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > Hello, > > > > > > > They are the same compiler. Is that what you expected? > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ c++ > > --version > > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > > Copyright (C) 2013 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. > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > --version > > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > > Copyright (C) 2013 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. > > > > > > Okay, that's good. I was expecting this, yes. Better safe than > > sorry ;-) > > > > Best regards, > > Karli > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-12-02 13:51:28
|
Hi Sandra, ViennaCL is a header-only library. So there isn't anything that needs to be installed separately. You can just use the GCC compilation line that we identified as working and continue with that. The CMake environment is just for building examples, but is not a requirement per se. Best regards, Karli On 12/2/21 14:38, SANDRA CATALAN PALLARES wrote: > Hi Karl, > > I am back these days to this work. Could you confirm to me if I need > anything else to work with the library? I am not sure if the > installation is properly done after the emails we exchanged. > > Thanks again for your help. > > Regards, > Sandra > > El vie, 26 nov 2021 a las 14:25, SANDRA CATALAN PALLARES > (<sca...@uc... <mailto:sca...@uc...>>) escribió: > > Hi, > > Ok. Then, is there anything left in order to make use of the > library? I am a bit lost right now. > > Regards, > Sandra > > El vie, 26 nov 2021 a las 13:58, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > Hello, > > > > They are the same compiler. Is that what you expected? > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ c++ > --version > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > Copyright (C) 2013 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. > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > --version > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > Copyright (C) 2013 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. > > > Okay, that's good. I was expecting this, yes. Better safe than > sorry ;-) > > Best regards, > Karli > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> |
|
From: SANDRA C. P. <sca...@uc...> - 2021-12-02 13:38:27
|
Hi Karl, I am back these days to this work. Could you confirm to me if I need anything else to work with the library? I am not sure if the installation is properly done after the emails we exchanged. Thanks again for your help. Regards, Sandra El vie, 26 nov 2021 a las 14:25, SANDRA CATALAN PALLARES (<sca...@uc...>) escribió: > Hi, > > Ok. Then, is there anything left in order to make use of the library? I am > a bit lost right now. > > Regards, > Sandra > > El vie, 26 nov 2021 a las 13:58, Karl Rupp (<ru...@iu...>) > escribió: > >> Hello, >> >> >> > They are the same compiler. Is that what you expected? >> > >> > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ c++ --version >> > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 >> > Copyright (C) 2013 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. >> > >> > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ --version >> > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 >> > Copyright (C) 2013 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. >> >> >> Okay, that's good. I was expecting this, yes. Better safe than sorry ;-) >> >> Best regards, >> Karli >> > > > -- > > *Sandra Catalán Pallarés* > Sección Dept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-26 13:25:53
|
Hi, Ok. Then, is there anything left in order to make use of the library? I am a bit lost right now. Regards, Sandra El vie, 26 nov 2021 a las 13:58, Karl Rupp (<ru...@iu...>) escribió: > Hello, > > > > They are the same compiler. Is that what you expected? > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ c++ --version > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > Copyright (C) 2013 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. > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ --version > > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > > Copyright (C) 2013 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. > > > Okay, that's good. I was expecting this, yes. Better safe than sorry ;-) > > Best regards, > Karli > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-26 12:58:29
|
Hello, > They are the same compiler. Is that what you expected? > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ c++ --version > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > Copyright (C) 2013 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. > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ --version > g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 > Copyright (C) 2013 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. Okay, that's good. I was expecting this, yes. Better safe than sorry ;-) Best regards, Karli |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-26 11:24:35
|
Hi, They are the same compiler. Is that what you expected? jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ c++ --version g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 Copyright (C) 2013 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. jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ --version g++-4.8.real (Ubuntu/Linaro 4.8.2-19ubuntu1) 4.8.2 Copyright (C) 2013 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. Regards, Sandra El vie, 26 nov 2021 a las 12:13, Karl Rupp (<ru...@iu...>) escribió: > Interesting, thanks! This is really strange... > > What do you get for > $> c++ --version > and > $> g++ --version > > Is it the same compiler? Also, you can try to incrementally transform > the CMake compilation line to the g++ compilation line and see where > things break. I don't know how familiar you are with compilers, so this > might be quite tricky. > > In any case, you can experiment with ViennaCL by using the g++ > compilation command. Since ViennaCL is header-only, there isn't anything > else that you need to copy around. > > Best regards, > Karli > > > > > On 11/26/21 10:55, SANDRA CATALAN PALLARES wrote: > > Hi, > > > > I just run the two commands you sent me. This is the output: > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > example/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > g++-4.8.real: error: example/tutorial/blas1.cpp: No such file or > directory > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > examples/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out > > Copying a few scalars... > > Manipulating a few scalars... > > operator += > > operator *= > > operator -= > > operator /= > > operator + > > multiple operators > > mixed operations > > CPU scalar s3: 42 > > GPU scalar vcl_s3: 42 > > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > > > Regards, > > Sandra > > > > El vie, 26 nov 2021 a las 10:44, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > Thanks! This is the line used by CMake: > > > > /usr/bin/c++ -I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > -I/usr/local/include > > > -I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/libviennacl/include > > > > -DVIENNACL_WITH_OPENCL -o CMakeFiles/blas1.dir/blas1.cpp.o -c > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/examples/tutorial/blas1.cpp > > > > As you say, it doesn't yield a proper executable. Just to verify: > Does > > > > $> g++ example/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > $> ./a.out > > > > in the main ViennaCL folder work if you build and execute manually? > > > > Best regards, > > Karli > > > > > > > > > > On 11/26/21 10:34, SANDRA CATALAN PALLARES wrote: > > > Hi Karli, > > > > > > No problem. Here is the new output. > > > > > > Regards, > > > Sandra > > > > > > El vie, 26 nov 2021 a las 9:59, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... <mailto:ru...@iu...>>>) > > escribió: > > > > > > Dear Sandra, > > > > > > sorry for the late reply! > > > > > > So CMake only relinked the executable. The linker command > looks > > > fine. So > > > we need to check the compilation as well. Please run > > > $> make clean > > > $> make blas1 > > > and re-send the output. This time it should also include the > > > compilation > > > command. > > > > > > Thanks and best regards, > > > Karli > > > > > > On 11/25/21 11:09, SANDRA CATALAN PALLARES wrote: > > > > Hi, > > > > > > > > I just wanted to ask you if the output I sent was the > > right one, > > > meaning > > > > the installation is correct this time. > > > > > > > > Regards, > > > > Sandra > > > > > > > > El vie, 19 nov 2021 a las 14:28, SANDRA CATALAN PALLARES > > > > (<sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>>) escribió: > > > > > > > > The output is attached. > > > > > > > > Best regards, > > > > Sandra > > > > > > > > El vie, 19 nov 2021 a las 14:03, Karl Rupp > > > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>) escribió: > > > > > > > > Ok, so now we really need to actual compiler call > from > > > CMake. In > > > > build/: > > > > > > > > $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On > > > > $> make blas1 > > > > (send output) > > > > > > > > Thanks and best regards, > > > > Karli > > > > > > > > > > > > On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: > > > > > Hi, > > > > > > > > > > The example that you mention is not in the new > > examples > > > > folder. However, > > > > > I tried with blas1 and it does not work: > > > > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ > > > > > examples/tutorial/blas1 > > > > > terminate called after throwing an instance of > > > > > 'viennacl::ocl::unknown_error' > > > > > what(): ViennaCL: FATAL ERROR: ViennaCL > > encountered an > > > > unknown > > > > > OpenCL error. Most likely your OpenCL SDK or > > driver is not > > > > installed > > > > > properly. In some cases, this error is due to > > an invalid > > > > global work > > > > > size or several kernel compilation errors. > > > > > If you think that this is a bug in ViennaCL, > please > > > report it at > > > > > vie...@li... > > <mailto:vie...@li...> > > > <mailto:vie...@li... > > <mailto:vie...@li...>> > > > > <mailto:vie...@li... > > <mailto:vie...@li...> > > > <mailto:vie...@li... > > <mailto:vie...@li...>>> > > > > > <mailto:vie...@li... > > <mailto:vie...@li...> > > > <mailto:vie...@li... > > <mailto:vie...@li...>> > > > > <mailto:vie...@li... > > <mailto:vie...@li...> > > > <mailto:vie...@li... > > <mailto:vie...@li...>>>> and supply at > > > > least the > > > > > following information: > > > > > * Operating System > > > > > * Which OpenCL implementation (AMD, NVIDIA, > etc.) > > > > > * ViennaCL version > > > > > Many thanks in advance! > > > > > Aborted > > > > > > > > > > Best regards, > > > > > Sandra > > > > > > > > > > El vie, 19 nov 2021 a las 13:27, Karl Rupp > > > > (<ru...@iu... > > <mailto:ru...@iu...> <mailto:ru...@iu... > > <mailto:ru...@iu...>> > > > <mailto:ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>>> > > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>>) escribió: > > > > > > > > > > Can you try to build in a subfolder? Maybe > > there's > > > some > > > > issue with the > > > > > build inside the project folder. Also, it > > would be > > > good > > > > to start with a > > > > > fresh ViennaCL folder just to rule out any > > > side-effects. > > > > > > > > > > $> cd /path/to/ViennaCL > > > > > $> mkdir build > > > > > $> cd build > > > > > $> cmake .. > > > > > > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > $> make > > > > > > > > > > ENABLE_OPENCL is set by default. We can > > skip ublas > > > for this. > > > > > Then run via > > > > > > > > > > $> examples/tutorial/blas1-opencl > > > > > > > > > > inside the build folder. Does this work? > > > > > > > > > > Best regards, > > > > > Karli > > > > > > > > > > > > > > > > > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES > > wrote: > > > > > > And this is the output of the cmake: > > > > > > > > > > > > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > > > ENABLE_OPENCL > > > > > > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > > > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > > > > > > > ENABLE_UBLAS . > > > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > > > > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > > > > > -- The C compiler identification is GNU > > 4.8.2 > > > > > > -- The CXX compiler identification is > > GNU 4.8.2 > > > > > > -- 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 > > > > > > -- Boost version: 1.54.0 > > > > > > -- 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 > > > > > > -- Boost version: 1.54.0 > > > > > > -- Found the following Boost libraries: > > > > > > -- chrono > > > > > > -- date_time > > > > > > -- serialization > > > > > > -- system > > > > > > -- thread > > > > > > -- atomic > > > > > > -- Found OPENCL: > > > > > > > > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > > -- Configuring done > > > > > > -- Generating done > > > > > > CMake Warning: > > > > > > Manually-specified variables were not > > used > > > by the > > > > project: > > > > > > > > > > > > OPENCLROOT > > > > > > > > > > > > > > > > > > -- Build files have been written to: > > > > > > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > > > > > > > > > El vie, 19 nov 2021 a las 10:36, SANDRA > > CATALAN > > > PALLARES > > > > > > (<sca...@uc... > > <mailto:sca...@uc...> <mailto:sca...@uc... > > <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>> > > > > > <mailto:sca...@uc... > > <mailto:sca...@uc...> <mailto:sca...@uc... > > <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>>>>) escribió: > > > > > > > > > > > > This is the last compilation command > > I used > > > > > > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > > > > > > > > > > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > > > > ENABLE_OPENCL > > > > > > > > > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > > > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > > ENABLE_UBLAS . > > > > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > > > > > / > > > > > > / > > > > > > Best regards, > > > > > > Sandra > > > > > > > > > > > > > > > > > > -- > > > > > > *Sandra Catalán Pallarés* > > > SecciónDept. Arquitectura de Computadores y Automática > > > Facultad de Ciencias Físicas - Desp. 223 > > > Universidad Complutense de Madrid > > > 28040 - Madrid (Spain) > > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-26 11:13:17
|
Interesting, thanks! This is really strange... What do you get for $> c++ --version and $> g++ --version Is it the same compiler? Also, you can try to incrementally transform the CMake compilation line to the g++ compilation line and see where things break. I don't know how familiar you are with compilers, so this might be quite tricky. In any case, you can experiment with ViennaCL by using the g++ compilation command. Since ViennaCL is header-only, there isn't anything else that you need to copy around. Best regards, Karli On 11/26/21 10:55, SANDRA CATALAN PALLARES wrote: > Hi, > > I just run the two commands you sent me. This is the output: > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > example/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > g++-4.8.real: error: example/tutorial/blas1.cpp: No such file or directory > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > examples/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out > Copying a few scalars... > Manipulating a few scalars... > operator += > operator *= > operator -= > operator /= > operator + > multiple operators > mixed operations > CPU scalar s3: 42 > GPU scalar vcl_s3: 42 > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > Regards, > Sandra > > El vie, 26 nov 2021 a las 10:44, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > Thanks! This is the line used by CMake: > > /usr/bin/c++ -I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > -I/usr/local/include > -I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/libviennacl/include > > -DVIENNACL_WITH_OPENCL -o CMakeFiles/blas1.dir/blas1.cpp.o -c > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/examples/tutorial/blas1.cpp > > As you say, it doesn't yield a proper executable. Just to verify: Does > > $> g++ example/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > $> ./a.out > > in the main ViennaCL folder work if you build and execute manually? > > Best regards, > Karli > > > > > On 11/26/21 10:34, SANDRA CATALAN PALLARES wrote: > > Hi Karli, > > > > No problem. Here is the new output. > > > > Regards, > > Sandra > > > > El vie, 26 nov 2021 a las 9:59, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>>>) > escribió: > > > > Dear Sandra, > > > > sorry for the late reply! > > > > So CMake only relinked the executable. The linker command looks > > fine. So > > we need to check the compilation as well. Please run > > $> make clean > > $> make blas1 > > and re-send the output. This time it should also include the > > compilation > > command. > > > > Thanks and best regards, > > Karli > > > > On 11/25/21 11:09, SANDRA CATALAN PALLARES wrote: > > > Hi, > > > > > > I just wanted to ask you if the output I sent was the > right one, > > meaning > > > the installation is correct this time. > > > > > > Regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 14:28, SANDRA CATALAN PALLARES > > > (<sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>>>) escribió: > > > > > > The output is attached. > > > > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 14:03, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>>) escribió: > > > > > > Ok, so now we really need to actual compiler call from > > CMake. In > > > build/: > > > > > > $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On > > > $> make blas1 > > > (send output) > > > > > > Thanks and best regards, > > > Karli > > > > > > > > > On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: > > > > Hi, > > > > > > > > The example that you mention is not in the new > examples > > > folder. However, > > > > I tried with blas1 and it does not work: > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ > > > > examples/tutorial/blas1 > > > > terminate called after throwing an instance of > > > > 'viennacl::ocl::unknown_error' > > > > what(): ViennaCL: FATAL ERROR: ViennaCL > encountered an > > > unknown > > > > OpenCL error. Most likely your OpenCL SDK or > driver is not > > > installed > > > > properly. In some cases, this error is due to > an invalid > > > global work > > > > size or several kernel compilation errors. > > > > If you think that this is a bug in ViennaCL, please > > report it at > > > > vie...@li... > <mailto:vie...@li...> > > <mailto:vie...@li... > <mailto:vie...@li...>> > > > <mailto:vie...@li... > <mailto:vie...@li...> > > <mailto:vie...@li... > <mailto:vie...@li...>>> > > > > <mailto:vie...@li... > <mailto:vie...@li...> > > <mailto:vie...@li... > <mailto:vie...@li...>> > > > <mailto:vie...@li... > <mailto:vie...@li...> > > <mailto:vie...@li... > <mailto:vie...@li...>>>> and supply at > > > least the > > > > following information: > > > > * Operating System > > > > * Which OpenCL implementation (AMD, NVIDIA, etc.) > > > > * ViennaCL version > > > > Many thanks in advance! > > > > Aborted > > > > > > > > Best regards, > > > > Sandra > > > > > > > > El vie, 19 nov 2021 a las 13:27, Karl Rupp > > > (<ru...@iu... > <mailto:ru...@iu...> <mailto:ru...@iu... > <mailto:ru...@iu...>> > > <mailto:ru...@iu... <mailto:ru...@iu...> > <mailto:ru...@iu... <mailto:ru...@iu...>>> > > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>>>) escribió: > > > > > > > > Can you try to build in a subfolder? Maybe > there's > > some > > > issue with the > > > > build inside the project folder. Also, it > would be > > good > > > to start with a > > > > fresh ViennaCL folder just to rule out any > > side-effects. > > > > > > > > $> cd /path/to/ViennaCL > > > > $> mkdir build > > > > $> cd build > > > > $> cmake .. > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > $> make > > > > > > > > ENABLE_OPENCL is set by default. We can > skip ublas > > for this. > > > > Then run via > > > > > > > > $> examples/tutorial/blas1-opencl > > > > > > > > inside the build folder. Does this work? > > > > > > > > Best regards, > > > > Karli > > > > > > > > > > > > > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES > wrote: > > > > > And this is the output of the cmake: > > > > > > > > > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > > ENABLE_OPENCL > > > > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > > > > > ENABLE_UBLAS . > > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > > > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > > > > -- The C compiler identification is GNU > 4.8.2 > > > > > -- The CXX compiler identification is > GNU 4.8.2 > > > > > -- 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 > > > > > -- Boost version: 1.54.0 > > > > > -- 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 > > > > > -- Boost version: 1.54.0 > > > > > -- Found the following Boost libraries: > > > > > -- chrono > > > > > -- date_time > > > > > -- serialization > > > > > -- system > > > > > -- thread > > > > > -- atomic > > > > > -- Found OPENCL: > > > > > > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > -- Configuring done > > > > > -- Generating done > > > > > CMake Warning: > > > > > Manually-specified variables were not > used > > by the > > > project: > > > > > > > > > > OPENCLROOT > > > > > > > > > > > > > > > -- Build files have been written to: > > > > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > > > > > > > El vie, 19 nov 2021 a las 10:36, SANDRA > CATALAN > > PALLARES > > > > > (<sca...@uc... > <mailto:sca...@uc...> <mailto:sca...@uc... > <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>>> > > > > <mailto:sca...@uc... > <mailto:sca...@uc...> <mailto:sca...@uc... > <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>>>>>) escribió: > > > > > > > > > > This is the last compilation command > I used > > > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > > > > > > > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > > > ENABLE_OPENCL > > > > > > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > ENABLE_UBLAS . > > > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > > > > / > > > > > / > > > > > Best regards, > > > > > Sandra > > > > > > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-26 09:55:27
|
Hi, I just run the two commands you sent me. This is the output: jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ example/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ g++-4.8.real: error: example/tutorial/blas1.cpp: No such file or directory jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ examples/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out Copying a few scalars... Manipulating a few scalars... operator += operator *= operator -= operator /= operator + multiple operators mixed operations CPU scalar s3: 42 GPU scalar vcl_s3: 42 !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! Regards, Sandra El vie, 26 nov 2021 a las 10:44, Karl Rupp (<ru...@iu...>) escribió: > Thanks! This is the line used by CMake: > > /usr/bin/c++ -I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > -I/usr/local/include > -I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/libviennacl/include > -DVIENNACL_WITH_OPENCL -o CMakeFiles/blas1.dir/blas1.cpp.o -c > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/examples/tutorial/blas1.cpp > > As you say, it doesn't yield a proper executable. Just to verify: Does > > $> g++ example/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > $> ./a.out > > in the main ViennaCL folder work if you build and execute manually? > > Best regards, > Karli > > > > > On 11/26/21 10:34, SANDRA CATALAN PALLARES wrote: > > Hi Karli, > > > > No problem. Here is the new output. > > > > Regards, > > Sandra > > > > El vie, 26 nov 2021 a las 9:59, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > Dear Sandra, > > > > sorry for the late reply! > > > > So CMake only relinked the executable. The linker command looks > > fine. So > > we need to check the compilation as well. Please run > > $> make clean > > $> make blas1 > > and re-send the output. This time it should also include the > > compilation > > command. > > > > Thanks and best regards, > > Karli > > > > On 11/25/21 11:09, SANDRA CATALAN PALLARES wrote: > > > Hi, > > > > > > I just wanted to ask you if the output I sent was the right one, > > meaning > > > the installation is correct this time. > > > > > > Regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 14:28, SANDRA CATALAN PALLARES > > > (<sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>) escribió: > > > > > > The output is attached. > > > > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 14:03, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>) escribió: > > > > > > Ok, so now we really need to actual compiler call from > > CMake. In > > > build/: > > > > > > $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On > > > $> make blas1 > > > (send output) > > > > > > Thanks and best regards, > > > Karli > > > > > > > > > On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: > > > > Hi, > > > > > > > > The example that you mention is not in the new examples > > > folder. However, > > > > I tried with blas1 and it does not work: > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ > > > > examples/tutorial/blas1 > > > > terminate called after throwing an instance of > > > > 'viennacl::ocl::unknown_error' > > > > what(): ViennaCL: FATAL ERROR: ViennaCL > encountered an > > > unknown > > > > OpenCL error. Most likely your OpenCL SDK or driver is > not > > > installed > > > > properly. In some cases, this error is due to an > invalid > > > global work > > > > size or several kernel compilation errors. > > > > If you think that this is a bug in ViennaCL, please > > report it at > > > > vie...@li... > > <mailto:vie...@li...> > > > <mailto:vie...@li... > > <mailto:vie...@li...>> > > > > <mailto:vie...@li... > > <mailto:vie...@li...> > > > <mailto:vie...@li... > > <mailto:vie...@li...>>> and supply at > > > least the > > > > following information: > > > > * Operating System > > > > * Which OpenCL implementation (AMD, NVIDIA, etc.) > > > > * ViennaCL version > > > > Many thanks in advance! > > > > Aborted > > > > > > > > Best regards, > > > > Sandra > > > > > > > > El vie, 19 nov 2021 a las 13:27, Karl Rupp > > > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>) escribió: > > > > > > > > Can you try to build in a subfolder? Maybe there's > > some > > > issue with the > > > > build inside the project folder. Also, it would be > > good > > > to start with a > > > > fresh ViennaCL folder just to rule out any > > side-effects. > > > > > > > > $> cd /path/to/ViennaCL > > > > $> mkdir build > > > > $> cd build > > > > $> cmake .. > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > $> make > > > > > > > > ENABLE_OPENCL is set by default. We can skip ublas > > for this. > > > > Then run via > > > > > > > > $> examples/tutorial/blas1-opencl > > > > > > > > inside the build folder. Does this work? > > > > > > > > Best regards, > > > > Karli > > > > > > > > > > > > > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: > > > > > And this is the output of the cmake: > > > > > > > > > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > > ENABLE_OPENCL > > > > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > > > > > ENABLE_UBLAS . > > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > > > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > > > > -- The C compiler identification is GNU 4.8.2 > > > > > -- The CXX compiler identification is GNU 4.8.2 > > > > > -- 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 > > > > > -- Boost version: 1.54.0 > > > > > -- 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 > > > > > -- Boost version: 1.54.0 > > > > > -- Found the following Boost libraries: > > > > > -- chrono > > > > > -- date_time > > > > > -- serialization > > > > > -- system > > > > > -- thread > > > > > -- atomic > > > > > -- Found OPENCL: > > > > > > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > -- Configuring done > > > > > -- Generating done > > > > > CMake Warning: > > > > > Manually-specified variables were not used > > by the > > > project: > > > > > > > > > > OPENCLROOT > > > > > > > > > > > > > > > -- Build files have been written to: > > > > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > > > > > > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN > > PALLARES > > > > > (<sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>>>) escribió: > > > > > > > > > > This is the last compilation command I used > > > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > > > > > > > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > > > ENABLE_OPENCL > > > > > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > ENABLE_UBLAS . > > > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > > > > / > > > > > / > > > > > Best regards, > > > > > Sandra > > > > > > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-26 09:44:01
|
Thanks! This is the line used by CMake:
/usr/bin/c++ -I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1
-I/usr/local/include
-I/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/libviennacl/include
-DVIENNACL_WITH_OPENCL -o CMakeFiles/blas1.dir/blas1.cpp.o -c
/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1/examples/tutorial/blas1.cpp
As you say, it doesn't yield a proper executable. Just to verify: Does
$> g++ example/tutorial/blas1.cpp -I. -DVIENNACL_WITH_OPENCL -lOpenCL
-L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/
$> ./a.out
in the main ViennaCL folder work if you build and execute manually?
Best regards,
Karli
On 11/26/21 10:34, SANDRA CATALAN PALLARES wrote:
> Hi Karli,
>
> No problem. Here is the new output.
>
> Regards,
> Sandra
>
> El vie, 26 nov 2021 a las 9:59, Karl Rupp (<ru...@iu...
> <mailto:ru...@iu...>>) escribió:
>
> Dear Sandra,
>
> sorry for the late reply!
>
> So CMake only relinked the executable. The linker command looks
> fine. So
> we need to check the compilation as well. Please run
> $> make clean
> $> make blas1
> and re-send the output. This time it should also include the
> compilation
> command.
>
> Thanks and best regards,
> Karli
>
> On 11/25/21 11:09, SANDRA CATALAN PALLARES wrote:
> > Hi,
> >
> > I just wanted to ask you if the output I sent was the right one,
> meaning
> > the installation is correct this time.
> >
> > Regards,
> > Sandra
> >
> > El vie, 19 nov 2021 a las 14:28, SANDRA CATALAN PALLARES
> > (<sca...@uc... <mailto:sca...@uc...>
> <mailto:sca...@uc... <mailto:sca...@uc...>>>) escribió:
> >
> > The output is attached.
> >
> > Best regards,
> > Sandra
> >
> > El vie, 19 nov 2021 a las 14:03, Karl Rupp
> (<ru...@iu... <mailto:ru...@iu...>
> > <mailto:ru...@iu...
> <mailto:ru...@iu...>>>) escribió:
> >
> > Ok, so now we really need to actual compiler call from
> CMake. In
> > build/:
> >
> > $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On
> > $> make blas1
> > (send output)
> >
> > Thanks and best regards,
> > Karli
> >
> >
> > On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote:
> > > Hi,
> > >
> > > The example that you mention is not in the new examples
> > folder. However,
> > > I tried with blas1 and it does not work:
> > >
> > >
> jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$
> > > examples/tutorial/blas1
> > > terminate called after throwing an instance of
> > > 'viennacl::ocl::unknown_error'
> > > what(): ViennaCL: FATAL ERROR: ViennaCL encountered an
> > unknown
> > > OpenCL error. Most likely your OpenCL SDK or driver is not
> > installed
> > > properly. In some cases, this error is due to an invalid
> > global work
> > > size or several kernel compilation errors.
> > > If you think that this is a bug in ViennaCL, please
> report it at
> > > vie...@li...
> <mailto:vie...@li...>
> > <mailto:vie...@li...
> <mailto:vie...@li...>>
> > > <mailto:vie...@li...
> <mailto:vie...@li...>
> > <mailto:vie...@li...
> <mailto:vie...@li...>>> and supply at
> > least the
> > > following information:
> > > * Operating System
> > > * Which OpenCL implementation (AMD, NVIDIA, etc.)
> > > * ViennaCL version
> > > Many thanks in advance!
> > > Aborted
> > >
> > > Best regards,
> > > Sandra
> > >
> > > El vie, 19 nov 2021 a las 13:27, Karl Rupp
> > (<ru...@iu... <mailto:ru...@iu...>
> <mailto:ru...@iu... <mailto:ru...@iu...>>
> > > <mailto:ru...@iu...
> <mailto:ru...@iu...>
> > <mailto:ru...@iu...
> <mailto:ru...@iu...>>>>) escribió:
> > >
> > > Can you try to build in a subfolder? Maybe there's
> some
> > issue with the
> > > build inside the project folder. Also, it would be
> good
> > to start with a
> > > fresh ViennaCL folder just to rule out any
> side-effects.
> > >
> > > $> cd /path/to/ViennaCL
> > > $> mkdir build
> > > $> cd build
> > > $> cmake ..
> > >
> >
> -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so
> > > $> make
> > >
> > > ENABLE_OPENCL is set by default. We can skip ublas
> for this.
> > > Then run via
> > >
> > > $> examples/tutorial/blas1-opencl
> > >
> > > inside the build folder. Does this work?
> > >
> > > Best regards,
> > > Karli
> > >
> > >
> > >
> > > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote:
> > > > And this is the output of the cmake:
> > > >
> > > >
> /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake
> > > ENABLE_OPENCL
> > > >
> -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl
> > > >
> > >
> >
> -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so
> > >
> > > > ENABLE_UBLAS .
> -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"
> > > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON
> > > > -- The C compiler identification is GNU 4.8.2
> > > > -- The CXX compiler identification is GNU 4.8.2
> > > > -- 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
> > > > -- Boost version: 1.54.0
> > > > -- 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
> > > > -- Boost version: 1.54.0
> > > > -- Found the following Boost libraries:
> > > > -- chrono
> > > > -- date_time
> > > > -- serialization
> > > > -- system
> > > > -- thread
> > > > -- atomic
> > > > -- Found OPENCL:
> > >
> /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so
> > > > -- Configuring done
> > > > -- Generating done
> > > > CMake Warning:
> > > > Manually-specified variables were not used
> by the
> > project:
> > > >
> > > > OPENCLROOT
> > > >
> > > >
> > > > -- Build files have been written to:
> > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1
> > > >
> > > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN
> PALLARES
> > > > (<sca...@uc... <mailto:sca...@uc...>
> <mailto:sca...@uc... <mailto:sca...@uc...>>
> > <mailto:sca...@uc... <mailto:sca...@uc...>
> <mailto:sca...@uc... <mailto:sca...@uc...>>>
> > > <mailto:sca...@uc... <mailto:sca...@uc...>
> <mailto:sca...@uc... <mailto:sca...@uc...>>
> > <mailto:sca...@uc... <mailto:sca...@uc...>
> <mailto:sca...@uc... <mailto:sca...@uc...>>>>>) escribió:
> > > >
> > > > This is the last compilation command I used
> > > >
> > > >
> jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$
> > > >
> > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake
> > > > ENABLE_OPENCL
> > > >
> > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl
> > > >
> > >
> >
> -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so
> > > > ENABLE_UBLAS .
> > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/
> > > > /
> > > > /
> > > > Best regards,
> > > > Sandra
> > > >
>
>
>
> --
>
> *Sandra Catalán Pallarés*
> SecciónDept. Arquitectura de Computadores y Automática
> Facultad de Ciencias Físicas - Desp. 223
> Universidad Complutense de Madrid
> 28040 - Madrid (Spain)
> (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...>
|
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-26 09:34:53
|
Hi Karli, No problem. Here is the new output. Regards, Sandra El vie, 26 nov 2021 a las 9:59, Karl Rupp (<ru...@iu...>) escribió: > Dear Sandra, > > sorry for the late reply! > > So CMake only relinked the executable. The linker command looks fine. So > we need to check the compilation as well. Please run > $> make clean > $> make blas1 > and re-send the output. This time it should also include the compilation > command. > > Thanks and best regards, > Karli > > On 11/25/21 11:09, SANDRA CATALAN PALLARES wrote: > > Hi, > > > > I just wanted to ask you if the output I sent was the right one, meaning > > the installation is correct this time. > > > > Regards, > > Sandra > > > > El vie, 19 nov 2021 a las 14:28, SANDRA CATALAN PALLARES > > (<sca...@uc... <mailto:sca...@uc...>>) escribió: > > > > The output is attached. > > > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 14:03, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > Ok, so now we really need to actual compiler call from CMake. In > > build/: > > > > $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On > > $> make blas1 > > (send output) > > > > Thanks and best regards, > > Karli > > > > > > On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: > > > Hi, > > > > > > The example that you mention is not in the new examples > > folder. However, > > > I tried with blas1 and it does not work: > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ > > > examples/tutorial/blas1 > > > terminate called after throwing an instance of > > > 'viennacl::ocl::unknown_error' > > > what(): ViennaCL: FATAL ERROR: ViennaCL encountered an > > unknown > > > OpenCL error. Most likely your OpenCL SDK or driver is not > > installed > > > properly. In some cases, this error is due to an invalid > > global work > > > size or several kernel compilation errors. > > > If you think that this is a bug in ViennaCL, please report it > at > > > vie...@li... > > <mailto:vie...@li...> > > > <mailto:vie...@li... > > <mailto:vie...@li...>> and supply at > > least the > > > following information: > > > * Operating System > > > * Which OpenCL implementation (AMD, NVIDIA, etc.) > > > * ViennaCL version > > > Many thanks in advance! > > > Aborted > > > > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 13:27, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>) escribió: > > > > > > Can you try to build in a subfolder? Maybe there's some > > issue with the > > > build inside the project folder. Also, it would be good > > to start with a > > > fresh ViennaCL folder just to rule out any side-effects. > > > > > > $> cd /path/to/ViennaCL > > > $> mkdir build > > > $> cd build > > > $> cmake .. > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > $> make > > > > > > ENABLE_OPENCL is set by default. We can skip ublas for > this. > > > Then run via > > > > > > $> examples/tutorial/blas1-opencl > > > > > > inside the build folder. Does this work? > > > > > > Best regards, > > > Karli > > > > > > > > > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: > > > > And this is the output of the cmake: > > > > > > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > ENABLE_OPENCL > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > > > -- The C compiler identification is GNU 4.8.2 > > > > -- The CXX compiler identification is GNU 4.8.2 > > > > -- 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 > > > > -- Boost version: 1.54.0 > > > > -- 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 > > > > -- Boost version: 1.54.0 > > > > -- Found the following Boost libraries: > > > > -- chrono > > > > -- date_time > > > > -- serialization > > > > -- system > > > > -- thread > > > > -- atomic > > > > -- Found OPENCL: > > > > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > -- Configuring done > > > > -- Generating done > > > > CMake Warning: > > > > Manually-specified variables were not used by the > > project: > > > > > > > > OPENCLROOT > > > > > > > > > > > > -- Build files have been written to: > > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > > > > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN > PALLARES > > > > (<sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>>) escribió: > > > > > > > > This is the last compilation command I used > > > > > > > > jbelloch@odroid > :~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > > > > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > > ENABLE_OPENCL > > > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > ENABLE_UBLAS . > > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > > > / > > > > / > > > > Best regards, > > > > Sandra > > > > > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-26 08:59:50
|
Dear Sandra, sorry for the late reply! So CMake only relinked the executable. The linker command looks fine. So we need to check the compilation as well. Please run $> make clean $> make blas1 and re-send the output. This time it should also include the compilation command. Thanks and best regards, Karli On 11/25/21 11:09, SANDRA CATALAN PALLARES wrote: > Hi, > > I just wanted to ask you if the output I sent was the right one, meaning > the installation is correct this time. > > Regards, > Sandra > > El vie, 19 nov 2021 a las 14:28, SANDRA CATALAN PALLARES > (<sca...@uc... <mailto:sca...@uc...>>) escribió: > > The output is attached. > > Best regards, > Sandra > > El vie, 19 nov 2021 a las 14:03, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > Ok, so now we really need to actual compiler call from CMake. In > build/: > > $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On > $> make blas1 > (send output) > > Thanks and best regards, > Karli > > > On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: > > Hi, > > > > The example that you mention is not in the new examples > folder. However, > > I tried with blas1 and it does not work: > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ > > examples/tutorial/blas1 > > terminate called after throwing an instance of > > 'viennacl::ocl::unknown_error' > > what(): ViennaCL: FATAL ERROR: ViennaCL encountered an > unknown > > OpenCL error. Most likely your OpenCL SDK or driver is not > installed > > properly. In some cases, this error is due to an invalid > global work > > size or several kernel compilation errors. > > If you think that this is a bug in ViennaCL, please report it at > > vie...@li... > <mailto:vie...@li...> > > <mailto:vie...@li... > <mailto:vie...@li...>> and supply at > least the > > following information: > > * Operating System > > * Which OpenCL implementation (AMD, NVIDIA, etc.) > > * ViennaCL version > > Many thanks in advance! > > Aborted > > > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 13:27, Karl Rupp > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>) escribió: > > > > Can you try to build in a subfolder? Maybe there's some > issue with the > > build inside the project folder. Also, it would be good > to start with a > > fresh ViennaCL folder just to rule out any side-effects. > > > > $> cd /path/to/ViennaCL > > $> mkdir build > > $> cd build > > $> cmake .. > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > $> make > > > > ENABLE_OPENCL is set by default. We can skip ublas for this. > > Then run via > > > > $> examples/tutorial/blas1-opencl > > > > inside the build folder. Does this work? > > > > Best regards, > > Karli > > > > > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: > > > And this is the output of the cmake: > > > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > ENABLE_OPENCL > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > > -- The C compiler identification is GNU 4.8.2 > > > -- The CXX compiler identification is GNU 4.8.2 > > > -- 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 > > > -- Boost version: 1.54.0 > > > -- 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 > > > -- Boost version: 1.54.0 > > > -- Found the following Boost libraries: > > > -- chrono > > > -- date_time > > > -- serialization > > > -- system > > > -- thread > > > -- atomic > > > -- Found OPENCL: > > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > -- Configuring done > > > -- Generating done > > > CMake Warning: > > > Manually-specified variables were not used by the > project: > > > > > > OPENCLROOT > > > > > > > > > -- Build files have been written to: > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN PALLARES > > > (<sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>>>) escribió: > > > > > > This is the last compilation command I used > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > ENABLE_OPENCL > > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > ENABLE_UBLAS . > -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > > / > > > / > > > Best regards, > > > Sandra > > > |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-25 10:09:48
|
Hi, I just wanted to ask you if the output I sent was the right one, meaning the installation is correct this time. Regards, Sandra El vie, 19 nov 2021 a las 14:28, SANDRA CATALAN PALLARES (<sca...@uc...>) escribió: > The output is attached. > > Best regards, > Sandra > > El vie, 19 nov 2021 a las 14:03, Karl Rupp (<ru...@iu...>) > escribió: > >> Ok, so now we really need to actual compiler call from CMake. In build/: >> >> $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On >> $> make blas1 >> (send output) >> >> Thanks and best regards, >> Karli >> >> >> On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: >> > Hi, >> > >> > The example that you mention is not in the new examples folder. >> However, >> > I tried with blas1 and it does not work: >> > >> > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ >> > examples/tutorial/blas1 >> > terminate called after throwing an instance of >> > 'viennacl::ocl::unknown_error' >> > what(): ViennaCL: FATAL ERROR: ViennaCL encountered an unknown >> > OpenCL error. Most likely your OpenCL SDK or driver is not installed >> > properly. In some cases, this error is due to an invalid global work >> > size or several kernel compilation errors. >> > If you think that this is a bug in ViennaCL, please report it at >> > vie...@li... >> > <mailto:vie...@li...> and supply at least >> the >> > following information: >> > * Operating System >> > * Which OpenCL implementation (AMD, NVIDIA, etc.) >> > * ViennaCL version >> > Many thanks in advance! >> > Aborted >> > >> > Best regards, >> > Sandra >> > >> > El vie, 19 nov 2021 a las 13:27, Karl Rupp (<ru...@iu... >> > <mailto:ru...@iu...>>) escribió: >> > >> > Can you try to build in a subfolder? Maybe there's some issue with >> the >> > build inside the project folder. Also, it would be good to start >> with a >> > fresh ViennaCL folder just to rule out any side-effects. >> > >> > $> cd /path/to/ViennaCL >> > $> mkdir build >> > $> cd build >> > $> cmake .. >> > >> -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so >> > $> make >> > >> > ENABLE_OPENCL is set by default. We can skip ublas for this. >> > Then run via >> > >> > $> examples/tutorial/blas1-opencl >> > >> > inside the build folder. Does this work? >> > >> > Best regards, >> > Karli >> > >> > >> > >> > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: >> > > And this is the output of the cmake: >> > > >> > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake >> > ENABLE_OPENCL >> > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl >> > > >> > >> -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so >> > >> > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" >> > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON >> > > -- The C compiler identification is GNU 4.8.2 >> > > -- The CXX compiler identification is GNU 4.8.2 >> > > -- 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 >> > > -- Boost version: 1.54.0 >> > > -- 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 >> > > -- Boost version: 1.54.0 >> > > -- Found the following Boost libraries: >> > > -- chrono >> > > -- date_time >> > > -- serialization >> > > -- system >> > > -- thread >> > > -- atomic >> > > -- Found OPENCL: >> > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so >> > > -- Configuring done >> > > -- Generating done >> > > CMake Warning: >> > > Manually-specified variables were not used by the project: >> > > >> > > OPENCLROOT >> > > >> > > >> > > -- Build files have been written to: >> > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 >> > > >> > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN PALLARES >> > > (<sca...@uc... <mailto:sca...@uc...> >> > <mailto:sca...@uc... <mailto:sca...@uc...>>>) escribió: >> > > >> > > This is the last compilation command I used >> > > >> > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ >> > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake >> > > ENABLE_OPENCL >> > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl >> > > >> > >> -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so >> > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ >> > > / >> > > / >> > > Best regards, >> > > Sandra >> > > >> > > El vie, 19 nov 2021 a las 10:00, Karl Rupp >> > (<ru...@iu... <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>>>) escribió: >> > > >> > > Okay, so CMake must be doing something odd in the build >> > stage. >> > > Can you please send a compilation command when building >> with >> > > CMake? You >> > > may have to set CMAKE_VERBOSE_MAKEFILE to 'On'. >> > > >> > > Best regards, >> > > Karli >> > > >> > > >> > > >> > > >> > > On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote: >> > > > It works! >> > > > >> > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ >> g++ >> > > > examples/tutorial/blas1.cpp -I. -lOpenCL >> > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ >> > > > jbelloch@odroid >> :~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ >> > ./a.out >> > > > Copying a few scalars... >> > > > Manipulating a few scalars... >> > > > operator += >> > > > operator *= >> > > > operator -= >> > > > operator /= >> > > > operator + >> > > > multiple operators >> > > > mixed operations >> > > > CPU scalar s3: 42 >> > > > GPU scalar vcl_s3: 42 >> > > > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! >> > > > >> > > > Best regards, >> > > > Sandra >> > > > >> > > > El vie, 19 nov 2021 a las 8:49, Karl Rupp >> > > (<ru...@iu... <mailto:ru...@iu...> >> > <mailto:ru...@iu... <mailto:ru...@iu...>> >> > > > <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>>>>) escribió: >> > > > >> > > > Indeed! >> > > > >> > > > Line 58 causes the error: >> > > > >> > > >> > >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp >> > < >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp >> > >> > > >> > < >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp >> < >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp >> >> >> > > > >> > > >> > < >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp >> < >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp> >> < >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp >> < >> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp >> >>> >> > > > >> > > > However, these lines are *exactly* the same as in >> the >> > > minimal example; >> > > > but the minimal example succeeds. Also, this is >> > the very >> > > first call to >> > > > OpenCL, so there can't be any prior undetected >> errors. >> > > > >> > > > What happens if you compile manually? In the main >> > > ViennaCL folder: >> > > > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL >> > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ >> > > > >> > > > and then >> > > > $> ./a.out >> > > > >> > > > Best regards, >> > > > Karli >> > > > >> > > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: >> > > > > The mistery continues: >> > > > > >> > > > > jbelloch@odroid >> :~/EQU_FILTERING_sandra/test_opencl$ >> > > ./a.out >> > > > > # Platforms found: 1 >> > > > > 0 >> > > > > >> > > > > Best regards, >> > > > > Sandra >> > > > > >> > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp >> > > (<ru...@iu... <mailto:ru...@iu...> >> > <mailto:ru...@iu... <mailto:ru...@iu...>> >> > > > <mailto:ru...@iu... >> > <mailto:ru...@iu...> <mailto:ru...@iu... >> > <mailto:ru...@iu...>>> >> > > > > <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>> <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>>>>>) >> > > > escribió: >> > > > > >> > > > > >> > > > > Good morning, >> > > > > >> > > > > oh - now I'm puzzled. :-/ >> > > > > >> > > > > Can you please add a >> > > > > std::cout << err << std::endl; >> > > > > after >> > > > > std::cout << "# Platforms found: " << >> > > num_platforms << >> > > > std::endl; >> > > > > >> > > > > Which value do you get? Maybe >> clGetPlatformIDs >> > > returns a nonzero >> > > > > error code? >> > > > > >> > > > > Best regards, >> > > > > Karli >> > > > > >> > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES >> > wrote: >> > > > > > Hi, >> > > > > > >> > > > > > The example is working :) >> > > > > > >> > > > > > >> > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ >> > > > test.cpp >> > > > > > -lOpenCL >> > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ >> > > > > > >> > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ >> ./a.out >> > > > > > # Platforms found: 1 >> > > > > > >> > > > > > Cheers, >> > > > > > Sandra >> > > > > > >> > > > > > El jue, 18 nov 2021 a las 19:48, Karl >> Rupp >> > > > > (<ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>> <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>>> >> > > > <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>> <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>>>> >> > > > > > <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... <mailto: >> ru...@iu...>> >> > > > <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>>> <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... <mailto: >> ru...@iu...>> >> > > > <mailto:ru...@iu... >> > <mailto:ru...@iu...> >> > > <mailto:ru...@iu... >> > <mailto:ru...@iu...>>>>>>) >> > > > > escribió: >> > > > > > >> > > > > > thanks! This indicates that there >> is no >> > > OpenCL platform >> > > > > detected on >> > > > > > your >> > > > > > system. Basically this is the >> simplest >> > > query of all. >> > > > > > >> > > > > > Can you try to compile and run the >> > > following minimal >> > > > example? >> > > > > > >> > > > > > >> > > > > > #include <iostream> >> > > > > > #include <string> >> > > > > > #include <vector> >> > > > > > #include <cmath> >> > > > > > #include <stdexcept> >> > > > > > >> > > > > > #ifdef __APPLE__ >> > > > > > #include <OpenCL/cl.h> >> > > > > > #else >> > > > > > #include <CL/cl.h> >> > > > > > #endif >> > > > > > >> > > > > > int main() >> > > > > > { >> > > > > > cl_uint num_platforms; >> > > > > > cl_platform_id platform_ids[42]; >> > > > > > cl_int err = >> clGetPlatformIDs(42, >> > > platform_ids, >> > > > > &num_platforms); >> > > > > > std::cout << "# Platforms found: >> > " << >> > > num_platforms << >> > > > > std::endl; >> > > > > > >> > > > > > return 0; >> > > > > > } >> > > > > > >> > > > > > >> > > > > > >> > > > > > $> g++ test.cpp -lOpenCL >> > > > > > >> > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ >> > > > > > >> > > > > > should produce the executable a.out. >> > > > > > >> > > > > > It should return at least one >> platform. >> > > > > > >> > > > > > Thanks and best regards, >> > > > > > Karli >> > > > > > >> > > > > > >> > > > > > >> > > > > > >> > > > > > >> > > > > >> > > > > >> > > > > >> > > > > -- >> > > > > >> > > > > *Sandra Catalán Pallarés* >> > > > > SecciónDept. Arquitectura de Computadores y >> > Automática >> > > > > Facultad de Ciencias Físicas - Desp. 223 >> > > > > Universidad Complutense de Madrid >> > > > > 28040 - Madrid (Spain) >> > > > > (+34) 91 394 *4394 - *sca...@uc... >> > <mailto:sca...@uc...> >> > > <mailto:sca...@uc... <mailto:sca...@uc...>> >> > <mailto:sca...@uc... <mailto:sca...@uc...> >> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> >> > > > <mailto:sca...@uc... <mailto:sca...@uc...> >> > <mailto:sca...@uc... <mailto:sca...@uc...>> >> > > <mailto:sca...@uc... <mailto:sca...@uc...> >> > <mailto:sca...@uc... <mailto:sca...@uc...>>>> >> > > > >> > > > >> > > > >> > > > -- >> > > > >> > > > *Sandra Catalán Pallarés* >> > > > SecciónDept. Arquitectura de Computadores y Automática >> > > > Facultad de Ciencias Físicas - Desp. 223 >> > > > Universidad Complutense de Madrid >> > > > 28040 - Madrid (Spain) >> > > > (+34) 91 394 *4394 - *sca...@uc... >> > <mailto:sca...@uc...> >> > > <mailto:sca...@uc... <mailto:sca...@uc...>> >> > <mailto:sca...@uc... <mailto:sca...@uc...> >> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> >> > > >> > > >> > > >> > > -- >> > > >> > > *Sandra Catalán Pallarés* >> > > SecciónDept. Arquitectura de Computadores y Automática >> > > Facultad de Ciencias Físicas - Desp. 223 >> > > Universidad Complutense de Madrid >> > > 28040 - Madrid (Spain) >> > > (+34) 91 394 *4394 - *sca...@uc... >> > <mailto:sca...@uc...> <mailto:sca...@uc... >> > <mailto:sca...@uc...>> >> > > >> > > >> > > >> > > -- >> > > >> > > *Sandra Catalán Pallarés* >> > > SecciónDept. Arquitectura de Computadores y Automática >> > > Facultad de Ciencias Físicas - Desp. 223 >> > > Universidad Complutense de Madrid >> > > 28040 - Madrid (Spain) >> > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> >> > <mailto:sca...@uc... <mailto:sca...@uc...>> >> > >> > >> > >> > -- >> > >> > *Sandra Catalán Pallarés* >> > SecciónDept. Arquitectura de Computadores y Automática >> > Facultad de Ciencias Físicas - Desp. 223 >> > Universidad Complutense de Madrid >> > 28040 - Madrid (Spain) >> > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> >> > > > -- > > *Sandra Catalán Pallarés* > Sección Dept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-19 13:29:04
|
The output is attached. Best regards, Sandra El vie, 19 nov 2021 a las 14:03, Karl Rupp (<ru...@iu...>) escribió: > Ok, so now we really need to actual compiler call from CMake. In build/: > > $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On > $> make blas1 > (send output) > > Thanks and best regards, > Karli > > > On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: > > Hi, > > > > The example that you mention is not in the new examples folder. However, > > I tried with blas1 and it does not work: > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ > > examples/tutorial/blas1 > > terminate called after throwing an instance of > > 'viennacl::ocl::unknown_error' > > what(): ViennaCL: FATAL ERROR: ViennaCL encountered an unknown > > OpenCL error. Most likely your OpenCL SDK or driver is not installed > > properly. In some cases, this error is due to an invalid global work > > size or several kernel compilation errors. > > If you think that this is a bug in ViennaCL, please report it at > > vie...@li... > > <mailto:vie...@li...> and supply at least the > > following information: > > * Operating System > > * Which OpenCL implementation (AMD, NVIDIA, etc.) > > * ViennaCL version > > Many thanks in advance! > > Aborted > > > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 13:27, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > Can you try to build in a subfolder? Maybe there's some issue with > the > > build inside the project folder. Also, it would be good to start > with a > > fresh ViennaCL folder just to rule out any side-effects. > > > > $> cd /path/to/ViennaCL > > $> mkdir build > > $> cd build > > $> cmake .. > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > $> make > > > > ENABLE_OPENCL is set by default. We can skip ublas for this. > > Then run via > > > > $> examples/tutorial/blas1-opencl > > > > inside the build folder. Does this work? > > > > Best regards, > > Karli > > > > > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: > > > And this is the output of the cmake: > > > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > ENABLE_OPENCL > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > > -- The C compiler identification is GNU 4.8.2 > > > -- The CXX compiler identification is GNU 4.8.2 > > > -- 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 > > > -- Boost version: 1.54.0 > > > -- 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 > > > -- Boost version: 1.54.0 > > > -- Found the following Boost libraries: > > > -- chrono > > > -- date_time > > > -- serialization > > > -- system > > > -- thread > > > -- atomic > > > -- Found OPENCL: > > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > -- Configuring done > > > -- Generating done > > > CMake Warning: > > > Manually-specified variables were not used by the project: > > > > > > OPENCLROOT > > > > > > > > > -- Build files have been written to: > > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN PALLARES > > > (<sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>) escribió: > > > > > > This is the last compilation command I used > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > > ENABLE_OPENCL > > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > > / > > > / > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 10:00, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>) escribió: > > > > > > Okay, so CMake must be doing something odd in the build > > stage. > > > Can you please send a compilation command when building > with > > > CMake? You > > > may have to set CMAKE_VERBOSE_MAKEFILE to 'On'. > > > > > > Best regards, > > > Karli > > > > > > > > > > > > > > > On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote: > > > > It works! > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > g++ > > > > examples/tutorial/blas1.cpp -I. -lOpenCL > > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > ./a.out > > > > Copying a few scalars... > > > > Manipulating a few scalars... > > > > operator += > > > > operator *= > > > > operator -= > > > > operator /= > > > > operator + > > > > multiple operators > > > > mixed operations > > > > CPU scalar s3: 42 > > > > GPU scalar vcl_s3: 42 > > > > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > > > > > > > Best regards, > > > > Sandra > > > > > > > > El vie, 19 nov 2021 a las 8:49, Karl Rupp > > > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>) escribió: > > > > > > > > Indeed! > > > > > > > > Line 58 causes the error: > > > > > > > > > > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > > > > > > > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > >> > > > > > > > > > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp> > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > >>> > > > > > > > > However, these lines are *exactly* the same as in > the > > > minimal example; > > > > but the minimal example succeeds. Also, this is > > the very > > > first call to > > > > OpenCL, so there can't be any prior undetected > errors. > > > > > > > > What happens if you compile manually? In the main > > > ViennaCL folder: > > > > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL > > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > > and then > > > > $> ./a.out > > > > > > > > Best regards, > > > > Karli > > > > > > > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > > > > > The mistery continues: > > > > > > > > > > jbelloch@odroid > :~/EQU_FILTERING_sandra/test_opencl$ > > > ./a.out > > > > > # Platforms found: 1 > > > > > 0 > > > > > > > > > > Best regards, > > > > > Sandra > > > > > > > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp > > > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> <mailto:ru...@iu... > > <mailto:ru...@iu...>>> > > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>> <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>>) > > > > escribió: > > > > > > > > > > > > > > > Good morning, > > > > > > > > > > oh - now I'm puzzled. :-/ > > > > > > > > > > Can you please add a > > > > > std::cout << err << std::endl; > > > > > after > > > > > std::cout << "# Platforms found: " << > > > num_platforms << > > > > std::endl; > > > > > > > > > > Which value do you get? Maybe > clGetPlatformIDs > > > returns a nonzero > > > > > error code? > > > > > > > > > > Best regards, > > > > > Karli > > > > > > > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES > > wrote: > > > > > > Hi, > > > > > > > > > > > > The example is working :) > > > > > > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ > > > > test.cpp > > > > > > -lOpenCL > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ > ./a.out > > > > > > # Platforms found: 1 > > > > > > > > > > > > Cheers, > > > > > > Sandra > > > > > > > > > > > > El jue, 18 nov 2021 a las 19:48, Karl > Rupp > > > > > (<ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>> <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>> <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>> > > > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... <mailto: > ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>> <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... <mailto: > ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>>>) > > > > > escribió: > > > > > > > > > > > > thanks! This indicates that there is > no > > > OpenCL platform > > > > > detected on > > > > > > your > > > > > > system. Basically this is the > simplest > > > query of all. > > > > > > > > > > > > Can you try to compile and run the > > > following minimal > > > > example? > > > > > > > > > > > > > > > > > > #include <iostream> > > > > > > #include <string> > > > > > > #include <vector> > > > > > > #include <cmath> > > > > > > #include <stdexcept> > > > > > > > > > > > > #ifdef __APPLE__ > > > > > > #include <OpenCL/cl.h> > > > > > > #else > > > > > > #include <CL/cl.h> > > > > > > #endif > > > > > > > > > > > > int main() > > > > > > { > > > > > > cl_uint num_platforms; > > > > > > cl_platform_id platform_ids[42]; > > > > > > cl_int err = clGetPlatformIDs(42, > > > platform_ids, > > > > > &num_platforms); > > > > > > std::cout << "# Platforms found: > > " << > > > num_platforms << > > > > > std::endl; > > > > > > > > > > > > return 0; > > > > > > } > > > > > > > > > > > > > > > > > > > > > > > > $> g++ test.cpp -lOpenCL > > > > > > > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > > > > > > should produce the executable a.out. > > > > > > > > > > > > It should return at least one > platform. > > > > > > > > > > > > Thanks and best regards, > > > > > > Karli > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > > *Sandra Catalán Pallarés* > > > > > SecciónDept. Arquitectura de Computadores y > > Automática > > > > > Facultad de Ciencias Físicas - Desp. 223 > > > > > Universidad Complutense de Madrid > > > > > 28040 - Madrid (Spain) > > > > > (+34) 91 394 *4394 - *sca...@uc... > > <mailto:sca...@uc...> > > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>>> > > > > > > > > > > > > > > > > -- > > > > > > > > *Sandra Catalán Pallarés* > > > > SecciónDept. Arquitectura de Computadores y Automática > > > > Facultad de Ciencias Físicas - Desp. 223 > > > > Universidad Complutense de Madrid > > > > 28040 - Madrid (Spain) > > > > (+34) 91 394 *4394 - *sca...@uc... > > <mailto:sca...@uc...> > > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > > > > > > > > > -- > > > > > > *Sandra Catalán Pallarés* > > > SecciónDept. Arquitectura de Computadores y Automática > > > Facultad de Ciencias Físicas - Desp. 223 > > > Universidad Complutense de Madrid > > > 28040 - Madrid (Spain) > > > (+34) 91 394 *4394 - *sca...@uc... > > <mailto:sca...@uc...> <mailto:sca...@uc... > > <mailto:sca...@uc...>> > > > > > > > > > > > > -- > > > > > > *Sandra Catalán Pallarés* > > > SecciónDept. Arquitectura de Computadores y Automática > > > Facultad de Ciencias Físicas - Desp. 223 > > > Universidad Complutense de Madrid > > > 28040 - Madrid (Spain) > > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-19 13:03:18
|
Ok, so now we really need to actual compiler call from CMake. In build/: $> cmake .. -DCMAKE_VERBOSE_MAKEFILE=On $> make blas1 (send output) Thanks and best regards, Karli On 11/19/21 13:59, SANDRA CATALAN PALLARES wrote: > Hi, > > The example that you mention is not in the new examples folder. However, > I tried with blas1 and it does not work: > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ > examples/tutorial/blas1 > terminate called after throwing an instance of > 'viennacl::ocl::unknown_error' > what(): ViennaCL: FATAL ERROR: ViennaCL encountered an unknown > OpenCL error. Most likely your OpenCL SDK or driver is not installed > properly. In some cases, this error is due to an invalid global work > size or several kernel compilation errors. > If you think that this is a bug in ViennaCL, please report it at > vie...@li... > <mailto:vie...@li...> and supply at least the > following information: > * Operating System > * Which OpenCL implementation (AMD, NVIDIA, etc.) > * ViennaCL version > Many thanks in advance! > Aborted > > Best regards, > Sandra > > El vie, 19 nov 2021 a las 13:27, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > Can you try to build in a subfolder? Maybe there's some issue with the > build inside the project folder. Also, it would be good to start with a > fresh ViennaCL folder just to rule out any side-effects. > > $> cd /path/to/ViennaCL > $> mkdir build > $> cd build > $> cmake .. > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > $> make > > ENABLE_OPENCL is set by default. We can skip ublas for this. > Then run via > > $> examples/tutorial/blas1-opencl > > inside the build folder. Does this work? > > Best regards, > Karli > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: > > And this is the output of the cmake: > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > ENABLE_OPENCL > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > -- The C compiler identification is GNU 4.8.2 > > -- The CXX compiler identification is GNU 4.8.2 > > -- 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 > > -- Boost version: 1.54.0 > > -- 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 > > -- Boost version: 1.54.0 > > -- Found the following Boost libraries: > > -- chrono > > -- date_time > > -- serialization > > -- system > > -- thread > > -- atomic > > -- Found OPENCL: > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > -- Configuring done > > -- Generating done > > CMake Warning: > > Manually-specified variables were not used by the project: > > > > OPENCLROOT > > > > > > -- Build files have been written to: > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN PALLARES > > (<sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>>) escribió: > > > > This is the last compilation command I used > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > ENABLE_OPENCL > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > / > > / > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 10:00, Karl Rupp > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>) escribió: > > > > Okay, so CMake must be doing something odd in the build > stage. > > Can you please send a compilation command when building with > > CMake? You > > may have to set CMAKE_VERBOSE_MAKEFILE to 'On'. > > > > Best regards, > > Karli > > > > > > > > > > On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote: > > > It works! > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > > examples/tutorial/blas1.cpp -I. -lOpenCL > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > ./a.out > > > Copying a few scalars... > > > Manipulating a few scalars... > > > operator += > > > operator *= > > > operator -= > > > operator /= > > > operator + > > > multiple operators > > > mixed operations > > > CPU scalar s3: 42 > > > GPU scalar vcl_s3: 42 > > > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > > > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 8:49, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>>) escribió: > > > > > > Indeed! > > > > > > Line 58 causes the error: > > > > > > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp> > > > <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp>> > > > > > > <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp> <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp>>> > > > > > > However, these lines are *exactly* the same as in the > > minimal example; > > > but the minimal example succeeds. Also, this is > the very > > first call to > > > OpenCL, so there can't be any prior undetected errors. > > > > > > What happens if you compile manually? In the main > > ViennaCL folder: > > > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > and then > > > $> ./a.out > > > > > > Best regards, > > > Karli > > > > > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > > > > The mistery continues: > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ > > ./a.out > > > > # Platforms found: 1 > > > > 0 > > > > > > > > Best regards, > > > > Sandra > > > > > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> <mailto:ru...@iu... > <mailto:ru...@iu...>>> > > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>> <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>>>) > > > escribió: > > > > > > > > > > > > Good morning, > > > > > > > > oh - now I'm puzzled. :-/ > > > > > > > > Can you please add a > > > > std::cout << err << std::endl; > > > > after > > > > std::cout << "# Platforms found: " << > > num_platforms << > > > std::endl; > > > > > > > > Which value do you get? Maybe clGetPlatformIDs > > returns a nonzero > > > > error code? > > > > > > > > Best regards, > > > > Karli > > > > > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES > wrote: > > > > > Hi, > > > > > > > > > > The example is working :) > > > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ > > > test.cpp > > > > > -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > > > > # Platforms found: 1 > > > > > > > > > > Cheers, > > > > > Sandra > > > > > > > > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp > > > > (<ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>> <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>> <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>> > > > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>> <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>>>>) > > > > escribió: > > > > > > > > > > thanks! This indicates that there is no > > OpenCL platform > > > > detected on > > > > > your > > > > > system. Basically this is the simplest > > query of all. > > > > > > > > > > Can you try to compile and run the > > following minimal > > > example? > > > > > > > > > > > > > > > #include <iostream> > > > > > #include <string> > > > > > #include <vector> > > > > > #include <cmath> > > > > > #include <stdexcept> > > > > > > > > > > #ifdef __APPLE__ > > > > > #include <OpenCL/cl.h> > > > > > #else > > > > > #include <CL/cl.h> > > > > > #endif > > > > > > > > > > int main() > > > > > { > > > > > cl_uint num_platforms; > > > > > cl_platform_id platform_ids[42]; > > > > > cl_int err = clGetPlatformIDs(42, > > platform_ids, > > > > &num_platforms); > > > > > std::cout << "# Platforms found: > " << > > num_platforms << > > > > std::endl; > > > > > > > > > > return 0; > > > > > } > > > > > > > > > > > > > > > > > > > > $> g++ test.cpp -lOpenCL > > > > > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > > > > should produce the executable a.out. > > > > > > > > > > It should return at least one platform. > > > > > > > > > > Thanks and best regards, > > > > > Karli > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > *Sandra Catalán Pallarés* > > > > SecciónDept. Arquitectura de Computadores y > Automática > > > > Facultad de Ciencias Físicas - Desp. 223 > > > > Universidad Complutense de Madrid > > > > 28040 - Madrid (Spain) > > > > (+34) 91 394 *4394 - *sca...@uc... > <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>>> > > > > > > > > > > > > -- > > > > > > *Sandra Catalán Pallarés* > > > SecciónDept. Arquitectura de Computadores y Automática > > > Facultad de Ciencias Físicas - Desp. 223 > > > Universidad Complutense de Madrid > > > 28040 - Madrid (Spain) > > > (+34) 91 394 *4394 - *sca...@uc... > <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... > <mailto:sca...@uc...> <mailto:sca...@uc... > <mailto:sca...@uc...>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-19 13:00:30
|
Hi, The example that you mention is not in the new examples folder. However, I tried with blas1 and it does not work: jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1/build$ examples/tutorial/blas1 terminate called after throwing an instance of 'viennacl::ocl::unknown_error' what(): ViennaCL: FATAL ERROR: ViennaCL encountered an unknown OpenCL error. Most likely your OpenCL SDK or driver is not installed properly. In some cases, this error is due to an invalid global work size or several kernel compilation errors. If you think that this is a bug in ViennaCL, please report it at vie...@li... and supply at least the following information: * Operating System * Which OpenCL implementation (AMD, NVIDIA, etc.) * ViennaCL version Many thanks in advance! Aborted Best regards, Sandra El vie, 19 nov 2021 a las 13:27, Karl Rupp (<ru...@iu...>) escribió: > Can you try to build in a subfolder? Maybe there's some issue with the > build inside the project folder. Also, it would be good to start with a > fresh ViennaCL folder just to rule out any side-effects. > > $> cd /path/to/ViennaCL > $> mkdir build > $> cd build > $> cmake .. > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > $> make > > ENABLE_OPENCL is set by default. We can skip ublas for this. > Then run via > > $> examples/tutorial/blas1-opencl > > inside the build folder. Does this work? > > Best regards, > Karli > > > > On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: > > And this is the output of the cmake: > > > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake ENABLE_OPENCL > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > > -- The C compiler identification is GNU 4.8.2 > > -- The CXX compiler identification is GNU 4.8.2 > > -- 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 > > -- Boost version: 1.54.0 > > -- 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 > > -- Boost version: 1.54.0 > > -- Found the following Boost libraries: > > -- chrono > > -- date_time > > -- serialization > > -- system > > -- thread > > -- atomic > > -- Found OPENCL: > /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > -- Configuring done > > -- Generating done > > CMake Warning: > > Manually-specified variables were not used by the project: > > > > OPENCLROOT > > > > > > -- Build files have been written to: > > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN PALLARES > > (<sca...@uc... <mailto:sca...@uc...>>) escribió: > > > > This is the last compilation command I used > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > > ENABLE_OPENCL > > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > > > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > > / > > / > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 10:00, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > Okay, so CMake must be doing something odd in the build stage. > > Can you please send a compilation command when building with > > CMake? You > > may have to set CMAKE_VERBOSE_MAKEFILE to 'On'. > > > > Best regards, > > Karli > > > > > > > > > > On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote: > > > It works! > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > > examples/tutorial/blas1.cpp -I. -lOpenCL > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > ./a.out > > > Copying a few scalars... > > > Manipulating a few scalars... > > > operator += > > > operator *= > > > operator -= > > > operator /= > > > operator + > > > multiple operators > > > mixed operations > > > CPU scalar s3: 42 > > > GPU scalar vcl_s3: 42 > > > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > > > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 8:49, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>) escribió: > > > > > > Indeed! > > > > > > Line 58 causes the error: > > > > > > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > > > > > > > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > >> > > > > > > However, these lines are *exactly* the same as in the > > minimal example; > > > but the minimal example succeeds. Also, this is the very > > first call to > > > OpenCL, so there can't be any prior undetected errors. > > > > > > What happens if you compile manually? In the main > > ViennaCL folder: > > > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > and then > > > $> ./a.out > > > > > > Best regards, > > > Karli > > > > > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > > > > The mistery continues: > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ > > ./a.out > > > > # Platforms found: 1 > > > > 0 > > > > > > > > Best regards, > > > > Sandra > > > > > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > > > <mailto:ru...@iu... <mailto: > ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>) > > > escribió: > > > > > > > > > > > > Good morning, > > > > > > > > oh - now I'm puzzled. :-/ > > > > > > > > Can you please add a > > > > std::cout << err << std::endl; > > > > after > > > > std::cout << "# Platforms found: " << > > num_platforms << > > > std::endl; > > > > > > > > Which value do you get? Maybe clGetPlatformIDs > > returns a nonzero > > > > error code? > > > > > > > > Best regards, > > > > Karli > > > > > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote: > > > > > Hi, > > > > > > > > > > The example is working :) > > > > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ > > > test.cpp > > > > > -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > > > > # Platforms found: 1 > > > > > > > > > > Cheers, > > > > > Sandra > > > > > > > > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp > > > > (<ru...@iu... > > <mailto:ru...@iu...> <mailto:ru...@iu... > > <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> <mailto:ru...@iu... > > <mailto:ru...@iu...>>> > > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>> <mailto:ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>>) > > > > escribió: > > > > > > > > > > thanks! This indicates that there is no > > OpenCL platform > > > > detected on > > > > > your > > > > > system. Basically this is the simplest > > query of all. > > > > > > > > > > Can you try to compile and run the > > following minimal > > > example? > > > > > > > > > > > > > > > #include <iostream> > > > > > #include <string> > > > > > #include <vector> > > > > > #include <cmath> > > > > > #include <stdexcept> > > > > > > > > > > #ifdef __APPLE__ > > > > > #include <OpenCL/cl.h> > > > > > #else > > > > > #include <CL/cl.h> > > > > > #endif > > > > > > > > > > int main() > > > > > { > > > > > cl_uint num_platforms; > > > > > cl_platform_id platform_ids[42]; > > > > > cl_int err = clGetPlatformIDs(42, > > platform_ids, > > > > &num_platforms); > > > > > std::cout << "# Platforms found: " << > > num_platforms << > > > > std::endl; > > > > > > > > > > return 0; > > > > > } > > > > > > > > > > > > > > > > > > > > $> g++ test.cpp -lOpenCL > > > > > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > > > > should produce the executable a.out. > > > > > > > > > > It should return at least one platform. > > > > > > > > > > Thanks and best regards, > > > > > Karli > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > *Sandra Catalán Pallarés* > > > > SecciónDept. Arquitectura de Computadores y Automática > > > > Facultad de Ciencias Físicas - Desp. 223 > > > > Universidad Complutense de Madrid > > > > 28040 - Madrid (Spain) > > > > (+34) 91 394 *4394 - *sca...@uc... > > <mailto:sca...@uc...> <mailto:sca...@uc... > > <mailto:sca...@uc...>> > > > <mailto:sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > > > > > > > > > -- > > > > > > *Sandra Catalán Pallarés* > > > SecciónDept. Arquitectura de Computadores y Automática > > > Facultad de Ciencias Físicas - Desp. 223 > > > Universidad Complutense de Madrid > > > 28040 - Madrid (Spain) > > > (+34) 91 394 *4394 - *sca...@uc... > > <mailto:sca...@uc...> <mailto:sca...@uc... > > <mailto:sca...@uc...>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-19 12:27:52
|
Can you try to build in a subfolder? Maybe there's some issue with the build inside the project folder. Also, it would be good to start with a fresh ViennaCL folder just to rule out any side-effects. $> cd /path/to/ViennaCL $> mkdir build $> cd build $> cmake .. -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so $> make ENABLE_OPENCL is set by default. We can skip ublas for this. Then run via $> examples/tutorial/blas1-opencl inside the build folder. Does this work? Best regards, Karli On 11/19/21 13:04, SANDRA CATALAN PALLARES wrote: > And this is the output of the cmake: > > /home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake ENABLE_OPENCL > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL" > -DCMAKE_VERBOSE_MAKEFILE:BOOL=ON > -- The C compiler identification is GNU 4.8.2 > -- The CXX compiler identification is GNU 4.8.2 > -- 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 > -- Boost version: 1.54.0 > -- 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 > -- Boost version: 1.54.0 > -- Found the following Boost libraries: > -- chrono > -- date_time > -- serialization > -- system > -- thread > -- atomic > -- Found OPENCL: /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > -- Configuring done > -- Generating done > CMake Warning: > Manually-specified variables were not used by the project: > > OPENCLROOT > > > -- Build files have been written to: > /home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1 > > El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN PALLARES > (<sca...@uc... <mailto:sca...@uc...>>) escribió: > > This is the last compilation command I used > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ > //home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake > ENABLE_OPENCL > -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl > -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so > ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"/ > / > / > Best regards, > Sandra > > El vie, 19 nov 2021 a las 10:00, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > Okay, so CMake must be doing something odd in the build stage. > Can you please send a compilation command when building with > CMake? You > may have to set CMAKE_VERBOSE_MAKEFILE to 'On'. > > Best regards, > Karli > > > > > On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote: > > It works! > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > examples/tutorial/blas1.cpp -I. -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out > > Copying a few scalars... > > Manipulating a few scalars... > > operator += > > operator *= > > operator -= > > operator /= > > operator + > > multiple operators > > mixed operations > > CPU scalar s3: 42 > > GPU scalar vcl_s3: 42 > > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 8:49, Karl Rupp > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>) escribió: > > > > Indeed! > > > > Line 58 causes the error: > > > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp> > > > <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp>> > > > > However, these lines are *exactly* the same as in the > minimal example; > > but the minimal example succeeds. Also, this is the very > first call to > > OpenCL, so there can't be any prior undetected errors. > > > > What happens if you compile manually? In the main > ViennaCL folder: > > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > and then > > $> ./a.out > > > > Best regards, > > Karli > > > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > > > The mistery continues: > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ > ./a.out > > > # Platforms found: 1 > > > 0 > > > > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> <mailto:ru...@iu... > <mailto:ru...@iu...>>>>) > > escribió: > > > > > > > > > Good morning, > > > > > > oh - now I'm puzzled. :-/ > > > > > > Can you please add a > > > std::cout << err << std::endl; > > > after > > > std::cout << "# Platforms found: " << > num_platforms << > > std::endl; > > > > > > Which value do you get? Maybe clGetPlatformIDs > returns a nonzero > > > error code? > > > > > > Best regards, > > > Karli > > > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote: > > > > Hi, > > > > > > > > The example is working :) > > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ > > test.cpp > > > > -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > > > # Platforms found: 1 > > > > > > > > Cheers, > > > > Sandra > > > > > > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp > > > (<ru...@iu... > <mailto:ru...@iu...> <mailto:ru...@iu... > <mailto:ru...@iu...>> > > <mailto:ru...@iu... > <mailto:ru...@iu...> <mailto:ru...@iu... > <mailto:ru...@iu...>>> > > > > <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>> <mailto:ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... > <mailto:ru...@iu...>>>>>) > > > escribió: > > > > > > > > thanks! This indicates that there is no > OpenCL platform > > > detected on > > > > your > > > > system. Basically this is the simplest > query of all. > > > > > > > > Can you try to compile and run the > following minimal > > example? > > > > > > > > > > > > #include <iostream> > > > > #include <string> > > > > #include <vector> > > > > #include <cmath> > > > > #include <stdexcept> > > > > > > > > #ifdef __APPLE__ > > > > #include <OpenCL/cl.h> > > > > #else > > > > #include <CL/cl.h> > > > > #endif > > > > > > > > int main() > > > > { > > > > cl_uint num_platforms; > > > > cl_platform_id platform_ids[42]; > > > > cl_int err = clGetPlatformIDs(42, > platform_ids, > > > &num_platforms); > > > > std::cout << "# Platforms found: " << > num_platforms << > > > std::endl; > > > > > > > > return 0; > > > > } > > > > > > > > > > > > > > > > $> g++ test.cpp -lOpenCL > > > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > > should produce the executable a.out. > > > > > > > > It should return at least one platform. > > > > > > > > Thanks and best regards, > > > > Karli > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > *Sandra Catalán Pallarés* > > > SecciónDept. Arquitectura de Computadores y Automática > > > Facultad de Ciencias Físicas - Desp. 223 > > > Universidad Complutense de Madrid > > > 28040 - Madrid (Spain) > > > (+34) 91 394 *4394 - *sca...@uc... > <mailto:sca...@uc...> <mailto:sca...@uc... > <mailto:sca...@uc...>> > > <mailto:sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... > <mailto:sca...@uc...> <mailto:sca...@uc... > <mailto:sca...@uc...>> > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-19 12:04:56
|
And this is the output of the cmake:
/home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake ENABLE_OPENCL
-DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl
-DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so
ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"
-DCMAKE_VERBOSE_MAKEFILE:BOOL=ON
-- The C compiler identification is GNU 4.8.2
-- The CXX compiler identification is GNU 4.8.2
-- 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
-- Boost version: 1.54.0
-- 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
-- Boost version: 1.54.0
-- Found the following Boost libraries:
-- chrono
-- date_time
-- serialization
-- system
-- thread
-- atomic
-- Found OPENCL: /home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so
-- Configuring done
-- Generating done
CMake Warning:
Manually-specified variables were not used by the project:
OPENCLROOT
-- Build files have been written to:
/home/jbelloch/EQU_FILTERING_sandra/ViennaCL-1.7.1
El vie, 19 nov 2021 a las 10:36, SANDRA CATALAN PALLARES (<sca...@uc...>)
escribió:
> This is the last compilation command I used
>
> jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ */home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake
> ENABLE_OPENCL -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl
> -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so
> ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"*
>
> Best regards,
> Sandra
>
> El vie, 19 nov 2021 a las 10:00, Karl Rupp (<ru...@iu...>)
> escribió:
>
>> Okay, so CMake must be doing something odd in the build stage.
>> Can you please send a compilation command when building with CMake? You
>> may have to set CMAKE_VERBOSE_MAKEFILE to 'On'.
>>
>> Best regards,
>> Karli
>>
>>
>>
>>
>> On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote:
>> > It works!
>> >
>> > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++
>> > examples/tutorial/blas1.cpp -I. -lOpenCL
>> > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/
>> > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out
>> > Copying a few scalars...
>> > Manipulating a few scalars...
>> > operator +=
>> > operator *=
>> > operator -=
>> > operator /=
>> > operator +
>> > multiple operators
>> > mixed operations
>> > CPU scalar s3: 42
>> > GPU scalar vcl_s3: 42
>> > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!!
>> >
>> > Best regards,
>> > Sandra
>> >
>> > El vie, 19 nov 2021 a las 8:49, Karl Rupp (<ru...@iu...
>> > <mailto:ru...@iu...>>) escribió:
>> >
>> > Indeed!
>> >
>> > Line 58 causes the error:
>> >
>> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp
>> > <
>> https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp
>> >
>> >
>> > However, these lines are *exactly* the same as in the minimal
>> example;
>> > but the minimal example succeeds. Also, this is the very first call
>> to
>> > OpenCL, so there can't be any prior undetected errors.
>> >
>> > What happens if you compile manually? In the main ViennaCL folder:
>> > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL
>> > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/
>> >
>> > and then
>> > $> ./a.out
>> >
>> > Best regards,
>> > Karli
>> >
>> > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote:
>> > > The mistery continues:
>> > >
>> > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out
>> > > # Platforms found: 1
>> > > 0
>> > >
>> > > Best regards,
>> > > Sandra
>> > >
>> > > El vie, 19 nov 2021 a las 8:33, Karl Rupp (<
>> ru...@iu...
>> > <mailto:ru...@iu...>
>> > > <mailto:ru...@iu... <mailto:ru...@iu...>>>)
>> > escribió:
>> > >
>> > >
>> > > Good morning,
>> > >
>> > > oh - now I'm puzzled. :-/
>> > >
>> > > Can you please add a
>> > > std::cout << err << std::endl;
>> > > after
>> > > std::cout << "# Platforms found: " << num_platforms <<
>> > std::endl;
>> > >
>> > > Which value do you get? Maybe clGetPlatformIDs returns a
>> nonzero
>> > > error code?
>> > >
>> > > Best regards,
>> > > Karli
>> > >
>> > > On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote:
>> > > > Hi,
>> > > >
>> > > > The example is working :)
>> > > >
>> > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++
>> > test.cpp
>> > > > -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/
>> > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$
>> ./a.out
>> > > > # Platforms found: 1
>> > > >
>> > > > Cheers,
>> > > > Sandra
>> > > >
>> > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp
>> > > (<ru...@iu... <mailto:ru...@iu...>
>> > <mailto:ru...@iu... <mailto:ru...@iu...>>
>> > > > <mailto:ru...@iu...
>> > <mailto:ru...@iu...> <mailto:ru...@iu...
>> > <mailto:ru...@iu...>>>>)
>> > > escribió:
>> > > >
>> > > > thanks! This indicates that there is no OpenCL
>> platform
>> > > detected on
>> > > > your
>> > > > system. Basically this is the simplest query of all.
>> > > >
>> > > > Can you try to compile and run the following minimal
>> > example?
>> > > >
>> > > >
>> > > > #include <iostream>
>> > > > #include <string>
>> > > > #include <vector>
>> > > > #include <cmath>
>> > > > #include <stdexcept>
>> > > >
>> > > > #ifdef __APPLE__
>> > > > #include <OpenCL/cl.h>
>> > > > #else
>> > > > #include <CL/cl.h>
>> > > > #endif
>> > > >
>> > > > int main()
>> > > > {
>> > > > cl_uint num_platforms;
>> > > > cl_platform_id platform_ids[42];
>> > > > cl_int err = clGetPlatformIDs(42, platform_ids,
>> > > &num_platforms);
>> > > > std::cout << "# Platforms found: " <<
>> num_platforms <<
>> > > std::endl;
>> > > >
>> > > > return 0;
>> > > > }
>> > > >
>> > > >
>> > > >
>> > > > $> g++ test.cpp -lOpenCL
>> > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/
>> > > >
>> > > > should produce the executable a.out.
>> > > >
>> > > > It should return at least one platform.
>> > > >
>> > > > Thanks and best regards,
>> > > > Karli
>> > > >
>> > > >
>> > > >
>> > > >
>> > > >
>> > >
>> > >
>> > >
>> > > --
>> > >
>> > > *Sandra Catalán Pallarés*
>> > > SecciónDept. Arquitectura de Computadores y Automática
>> > > Facultad de Ciencias Físicas - Desp. 223
>> > > Universidad Complutense de Madrid
>> > > 28040 - Madrid (Spain)
>> > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...>
>> > <mailto:sca...@uc... <mailto:sca...@uc...>>
>> >
>> >
>> >
>> > --
>> >
>> > *Sandra Catalán Pallarés*
>> > SecciónDept. Arquitectura de Computadores y Automática
>> > Facultad de Ciencias Físicas - Desp. 223
>> > Universidad Complutense de Madrid
>> > 28040 - Madrid (Spain)
>> > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...>
>>
>
>
> --
>
> *Sandra Catalán Pallarés*
> Sección Dept. Arquitectura de Computadores y Automática
> Facultad de Ciencias Físicas - Desp. 223
> Universidad Complutense de Madrid
> 28040 - Madrid (Spain)
> (+34) 91 394 *4394 - *sca...@uc...
>
--
*Sandra Catalán Pallarés*
Sección Dept. Arquitectura de Computadores y Automática
Facultad de Ciencias Físicas - Desp. 223
Universidad Complutense de Madrid
28040 - Madrid (Spain)
(+34) 91 394 *4394 - *sca...@uc...
|
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-19 09:36:54
|
This is the last compilation command I used jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ */home/jbelloch/EQU_FILTERING_sandra/cmake-3.10.0/bin/cmake ENABLE_OPENCL -DOPENCLROOT=/home/jbelloch/EQU_FILTERING_sandra/opencl -DOPENCL_LIBRARY=/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/libOpenCL.so ENABLE_UBLAS . -DCMAKE_CXX_FLAGS="-DVIENNACL_DEBUG_ALL"* Best regards, Sandra El vie, 19 nov 2021 a las 10:00, Karl Rupp (<ru...@iu...>) escribió: > Okay, so CMake must be doing something odd in the build stage. > Can you please send a compilation command when building with CMake? You > may have to set CMAKE_VERBOSE_MAKEFILE to 'On'. > > Best regards, > Karli > > > > > On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote: > > It works! > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > > examples/tutorial/blas1.cpp -I. -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out > > Copying a few scalars... > > Manipulating a few scalars... > > operator += > > operator *= > > operator -= > > operator /= > > operator + > > multiple operators > > mixed operations > > CPU scalar s3: 42 > > GPU scalar vcl_s3: 42 > > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 8:49, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > Indeed! > > > > Line 58 causes the error: > > > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > > < > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > > > > > > However, these lines are *exactly* the same as in the minimal > example; > > but the minimal example succeeds. Also, this is the very first call > to > > OpenCL, so there can't be any prior undetected errors. > > > > What happens if you compile manually? In the main ViennaCL folder: > > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > and then > > $> ./a.out > > > > Best regards, > > Karli > > > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > > > The mistery continues: > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > > # Platforms found: 1 > > > 0 > > > > > > Best regards, > > > Sandra > > > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...> > > > <mailto:ru...@iu... <mailto:ru...@iu...>>>) > > escribió: > > > > > > > > > Good morning, > > > > > > oh - now I'm puzzled. :-/ > > > > > > Can you please add a > > > std::cout << err << std::endl; > > > after > > > std::cout << "# Platforms found: " << num_platforms << > > std::endl; > > > > > > Which value do you get? Maybe clGetPlatformIDs returns a > nonzero > > > error code? > > > > > > Best regards, > > > Karli > > > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote: > > > > Hi, > > > > > > > > The example is working :) > > > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ > > test.cpp > > > > -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ > ./a.out > > > > # Platforms found: 1 > > > > > > > > Cheers, > > > > Sandra > > > > > > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp > > > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > > <mailto:ru...@iu... > > <mailto:ru...@iu...> <mailto:ru...@iu... > > <mailto:ru...@iu...>>>>) > > > escribió: > > > > > > > > thanks! This indicates that there is no OpenCL platform > > > detected on > > > > your > > > > system. Basically this is the simplest query of all. > > > > > > > > Can you try to compile and run the following minimal > > example? > > > > > > > > > > > > #include <iostream> > > > > #include <string> > > > > #include <vector> > > > > #include <cmath> > > > > #include <stdexcept> > > > > > > > > #ifdef __APPLE__ > > > > #include <OpenCL/cl.h> > > > > #else > > > > #include <CL/cl.h> > > > > #endif > > > > > > > > int main() > > > > { > > > > cl_uint num_platforms; > > > > cl_platform_id platform_ids[42]; > > > > cl_int err = clGetPlatformIDs(42, platform_ids, > > > &num_platforms); > > > > std::cout << "# Platforms found: " << > num_platforms << > > > std::endl; > > > > > > > > return 0; > > > > } > > > > > > > > > > > > > > > > $> g++ test.cpp -lOpenCL > > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > > > should produce the executable a.out. > > > > > > > > It should return at least one platform. > > > > > > > > Thanks and best regards, > > > > Karli > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > *Sandra Catalán Pallarés* > > > SecciónDept. Arquitectura de Computadores y Automática > > > Facultad de Ciencias Físicas - Desp. 223 > > > Universidad Complutense de Madrid > > > 28040 - Madrid (Spain) > > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-19 09:00:38
|
Okay, so CMake must be doing something odd in the build stage. Can you please send a compilation command when building with CMake? You may have to set CMAKE_VERBOSE_MAKEFILE to 'On'. Best regards, Karli On 11/19/21 09:12, SANDRA CATALAN PALLARES wrote: > It works! > > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ > examples/tutorial/blas1.cpp -I. -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out > Copying a few scalars... > Manipulating a few scalars... > operator += > operator *= > operator -= > operator /= > operator + > multiple operators > mixed operations > CPU scalar s3: 42 > GPU scalar vcl_s3: 42 > !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! > > Best regards, > Sandra > > El vie, 19 nov 2021 a las 8:49, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > Indeed! > > Line 58 causes the error: > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > <https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp> > > However, these lines are *exactly* the same as in the minimal example; > but the minimal example succeeds. Also, this is the very first call to > OpenCL, so there can't be any prior undetected errors. > > What happens if you compile manually? In the main ViennaCL folder: > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > and then > $> ./a.out > > Best regards, > Karli > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > > The mistery continues: > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > # Platforms found: 1 > > 0 > > > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>>>) > escribió: > > > > > > Good morning, > > > > oh - now I'm puzzled. :-/ > > > > Can you please add a > > std::cout << err << std::endl; > > after > > std::cout << "# Platforms found: " << num_platforms << > std::endl; > > > > Which value do you get? Maybe clGetPlatformIDs returns a nonzero > > error code? > > > > Best regards, > > Karli > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote: > > > Hi, > > > > > > The example is working :) > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ > test.cpp > > > -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > > # Platforms found: 1 > > > > > > Cheers, > > > Sandra > > > > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > <mailto:ru...@iu... <mailto:ru...@iu...>> > > > <mailto:ru...@iu... > <mailto:ru...@iu...> <mailto:ru...@iu... > <mailto:ru...@iu...>>>>) > > escribió: > > > > > > thanks! This indicates that there is no OpenCL platform > > detected on > > > your > > > system. Basically this is the simplest query of all. > > > > > > Can you try to compile and run the following minimal > example? > > > > > > > > > #include <iostream> > > > #include <string> > > > #include <vector> > > > #include <cmath> > > > #include <stdexcept> > > > > > > #ifdef __APPLE__ > > > #include <OpenCL/cl.h> > > > #else > > > #include <CL/cl.h> > > > #endif > > > > > > int main() > > > { > > > cl_uint num_platforms; > > > cl_platform_id platform_ids[42]; > > > cl_int err = clGetPlatformIDs(42, platform_ids, > > &num_platforms); > > > std::cout << "# Platforms found: " << num_platforms << > > std::endl; > > > > > > return 0; > > > } > > > > > > > > > > > > $> g++ test.cpp -lOpenCL > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > should produce the executable a.out. > > > > > > It should return at least one platform. > > > > > > Thanks and best regards, > > > Karli > > > > > > > > > > > > > > > > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > <mailto:sca...@uc... <mailto:sca...@uc...>> > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-19 08:12:44
|
It works! jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ g++ examples/tutorial/blas1.cpp -I. -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ jbelloch@odroid:~/EQU_FILTERING_sandra/ViennaCL-1.7.1$ ./a.out Copying a few scalars... Manipulating a few scalars... operator += operator *= operator -= operator /= operator + multiple operators mixed operations CPU scalar s3: 42 GPU scalar vcl_s3: 42 !!!! TUTORIAL COMPLETED SUCCESSFULLY !!!! Best regards, Sandra El vie, 19 nov 2021 a las 8:49, Karl Rupp (<ru...@iu...>) escribió: > Indeed! > > Line 58 causes the error: > > https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp > > However, these lines are *exactly* the same as in the minimal example; > but the minimal example succeeds. Also, this is the very first call to > OpenCL, so there can't be any prior undetected errors. > > What happens if you compile manually? In the main ViennaCL folder: > $> g++ example/tutorial/blas1.cpp -I. -lOpenCL > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > and then > $> ./a.out > > Best regards, > Karli > > On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > > The mistery continues: > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > # Platforms found: 1 > > 0 > > > > Best regards, > > Sandra > > > > El vie, 19 nov 2021 a las 8:33, Karl Rupp (<ru...@iu... > > <mailto:ru...@iu...>>) escribió: > > > > > > Good morning, > > > > oh - now I'm puzzled. :-/ > > > > Can you please add a > > std::cout << err << std::endl; > > after > > std::cout << "# Platforms found: " << num_platforms << std::endl; > > > > Which value do you get? Maybe clGetPlatformIDs returns a nonzero > > error code? > > > > Best regards, > > Karli > > > > On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote: > > > Hi, > > > > > > The example is working :) > > > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ test.cpp > > > -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > > # Platforms found: 1 > > > > > > Cheers, > > > Sandra > > > > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp > > (<ru...@iu... <mailto:ru...@iu...> > > > <mailto:ru...@iu... <mailto:ru...@iu...>>>) > > escribió: > > > > > > thanks! This indicates that there is no OpenCL platform > > detected on > > > your > > > system. Basically this is the simplest query of all. > > > > > > Can you try to compile and run the following minimal example? > > > > > > > > > #include <iostream> > > > #include <string> > > > #include <vector> > > > #include <cmath> > > > #include <stdexcept> > > > > > > #ifdef __APPLE__ > > > #include <OpenCL/cl.h> > > > #else > > > #include <CL/cl.h> > > > #endif > > > > > > int main() > > > { > > > cl_uint num_platforms; > > > cl_platform_id platform_ids[42]; > > > cl_int err = clGetPlatformIDs(42, platform_ids, > > &num_platforms); > > > std::cout << "# Platforms found: " << num_platforms << > > std::endl; > > > > > > return 0; > > > } > > > > > > > > > > > > $> g++ test.cpp -lOpenCL > > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > > > should produce the executable a.out. > > > > > > It should return at least one platform. > > > > > > Thanks and best regards, > > > Karli > > > > > > > > > > > > > > > > > > > > > > > -- > > > > *Sandra Catalán Pallarés* > > SecciónDept. Arquitectura de Computadores y Automática > > Facultad de Ciencias Físicas - Desp. 223 > > Universidad Complutense de Madrid > > 28040 - Madrid (Spain) > > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> > -- *Sandra Catalán Pallarés* Sección Dept. Arquitectura de Computadores y Automática Facultad de Ciencias Físicas - Desp. 223 Universidad Complutense de Madrid 28040 - Madrid (Spain) (+34) 91 394 *4394 - *sca...@uc... |
|
From: Karl R. <ru...@iu...> - 2021-11-19 07:49:11
|
Indeed! Line 58 causes the error: https://github.com/viennacl/viennacl-dev/blob/master/viennacl/ocl/platform.hpp However, these lines are *exactly* the same as in the minimal example; but the minimal example succeeds. Also, this is the very first call to OpenCL, so there can't be any prior undetected errors. What happens if you compile manually? In the main ViennaCL folder: $> g++ example/tutorial/blas1.cpp -I. -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ and then $> ./a.out Best regards, Karli On 11/19/21 08:37, SANDRA CATALAN PALLARES wrote: > The mistery continues: > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > # Platforms found: 1 > 0 > > Best regards, > Sandra > > El vie, 19 nov 2021 a las 8:33, Karl Rupp (<ru...@iu... > <mailto:ru...@iu...>>) escribió: > > > Good morning, > > oh - now I'm puzzled. :-/ > > Can you please add a > std::cout << err << std::endl; > after > std::cout << "# Platforms found: " << num_platforms << std::endl; > > Which value do you get? Maybe clGetPlatformIDs returns a nonzero > error code? > > Best regards, > Karli > > On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote: > > Hi, > > > > The example is working :) > > > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ test.cpp > > -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out > > # Platforms found: 1 > > > > Cheers, > > Sandra > > > > El jue, 18 nov 2021 a las 19:48, Karl Rupp > (<ru...@iu... <mailto:ru...@iu...> > > <mailto:ru...@iu... <mailto:ru...@iu...>>>) > escribió: > > > > thanks! This indicates that there is no OpenCL platform > detected on > > your > > system. Basically this is the simplest query of all. > > > > Can you try to compile and run the following minimal example? > > > > > > #include <iostream> > > #include <string> > > #include <vector> > > #include <cmath> > > #include <stdexcept> > > > > #ifdef __APPLE__ > > #include <OpenCL/cl.h> > > #else > > #include <CL/cl.h> > > #endif > > > > int main() > > { > > cl_uint num_platforms; > > cl_platform_id platform_ids[42]; > > cl_int err = clGetPlatformIDs(42, platform_ids, > &num_platforms); > > std::cout << "# Platforms found: " << num_platforms << > std::endl; > > > > return 0; > > } > > > > > > > > $> g++ test.cpp -lOpenCL > > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/ > > > > should produce the executable a.out. > > > > It should return at least one platform. > > > > Thanks and best regards, > > Karli > > > > > > > > > > > > > > -- > > *Sandra Catalán Pallarés* > SecciónDept. Arquitectura de Computadores y Automática > Facultad de Ciencias Físicas - Desp. 223 > Universidad Complutense de Madrid > 28040 - Madrid (Spain) > (+34) 91 394 *4394 - *sca...@uc... <mailto:sca...@uc...> |
|
From: SANDRA C. P. <sca...@uc...> - 2021-11-19 07:37:30
|
The mistery continues:
jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out
# Platforms found: 1
0
Best regards,
Sandra
El vie, 19 nov 2021 a las 8:33, Karl Rupp (<ru...@iu...>)
escribió:
>
> Good morning,
>
> oh - now I'm puzzled. :-/
>
> Can you please add a
> std::cout << err << std::endl;
> after
> std::cout << "# Platforms found: " << num_platforms << std::endl;
>
> Which value do you get? Maybe clGetPlatformIDs returns a nonzero error
> code?
>
> Best regards,
> Karli
>
> On 11/19/21 08:29, SANDRA CATALAN PALLARES wrote:
> > Hi,
> >
> > The example is working :)
> >
> > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ g++ test.cpp
> > -lOpenCL -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/
> > jbelloch@odroid:~/EQU_FILTERING_sandra/test_opencl$ ./a.out
> > # Platforms found: 1
> >
> > Cheers,
> > Sandra
> >
> > El jue, 18 nov 2021 a las 19:48, Karl Rupp (<ru...@iu...
> > <mailto:ru...@iu...>>) escribió:
> >
> > thanks! This indicates that there is no OpenCL platform detected on
> > your
> > system. Basically this is the simplest query of all.
> >
> > Can you try to compile and run the following minimal example?
> >
> >
> > #include <iostream>
> > #include <string>
> > #include <vector>
> > #include <cmath>
> > #include <stdexcept>
> >
> > #ifdef __APPLE__
> > #include <OpenCL/cl.h>
> > #else
> > #include <CL/cl.h>
> > #endif
> >
> > int main()
> > {
> > cl_uint num_platforms;
> > cl_platform_id platform_ids[42];
> > cl_int err = clGetPlatformIDs(42, platform_ids, &num_platforms);
> > std::cout << "# Platforms found: " << num_platforms << std::endl;
> >
> > return 0;
> > }
> >
> >
> >
> > $> g++ test.cpp -lOpenCL
> > -L/home/jbelloch/EQU_FILTERING_sandra/opencl/lib/
> >
> > should produce the executable a.out.
> >
> > It should return at least one platform.
> >
> > Thanks and best regards,
> > Karli
> >
> >
> >
> >
> >
>
--
*Sandra Catalán Pallarés*
Sección Dept. Arquitectura de Computadores y Automática
Facultad de Ciencias Físicas - Desp. 223
Universidad Complutense de Madrid
28040 - Madrid (Spain)
(+34) 91 394 *4394 - *sca...@uc...
|
7 messages has been excluded from this view by a project administrator.