November 2021 - petsc-dev - mailman3.cels.anl.gov

Re: [petsc-dev] I am getting this error ...
by Mark Adams 04 Nov '21

04 Nov '21

It gets a lot further. On Thu, Nov 4, 2021 at 5:32 PM Mark Adams <mfadams(a)lbl.gov> wrote: > OK, sorry I missed the CPPFLAGS. It is running now. > Thanks, > > On Thu, Nov 4, 2021 at 4:43 PM Satish Balay <balay(a)mcs.anl.gov> wrote: > >> Multiple e-mail threads on the same issue (: >> >> As suggested in my earlier thread - add -mp=gpu to both CPPFLAGS and >> CFLAGS [or LDFLAGS] >> >> Satish >> >> ------- >> Executing: cc -o /tmp/petsc-Vvs8_T/config.setCompilers/conftest -g >> -DLANDAU_DIM=2 -DLANDAU_MAX_SPECIES=10 -DLANDAU_MAX_Q=4 >> /tmp/petsc-Vvs8_T/config.setCompilers/conftest.o >> Possible ERROR while running linker: >> stderr: >> nvc-Warning-The -gpu option has no effect unless a language-specific >> option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, >> -cuda) >> >> >> On Thu, 4 Nov 2021, Mark Adams wrote: >> >> > It is CPPFLAGS. I seem to get the same behavior. >> > >> > FWIW, I did get this response from NERSc but I don't know how to >> interpret >> > it. >> > >> > He seems to be saying that I don't need -mp=gpu for the device compiler >> > (nvcc). He seems to think that I am adding -gpu. >> > >> > nvcc -- The device compiler does not need any of those flags because it >> > already knows that it's being fed cuda code. The warning you're seeing >> is >> > coming from nvc (which is the host / CPU side compiler) if you're in the >> > PrgEnv-nvidia environment. You should not need to add -mp=gpu and -cuda, >> > please just add the -cuda flag (to your host code) not to the device >> code. >> > >> > I will try to talk with this guy again. >> > >> > Thanks, >> > >> > >> > On Thu, Nov 4, 2021 at 4:11 PM Barry Smith <bsmith(a)petsc.dev> wrote: >> > >> > > >> > > Yes, you need to use the CPPFLAGS which maybe called CPPCFLAGS I am >> not >> > > sure >> > > >> > > >> > > On Nov 4, 2021, at 3:23 PM, Mark Adams <mfadams(a)lbl.gov> wrote: >> > > >> > > Ah, CCFLAGS does not seem to work. >> > > >> > > On Thu, Nov 4, 2021 at 3:07 PM Barry Smith <bsmith(a)petsc.dev> wrote: >> > > >> > >> >> > >> You have to pass in the flag to turn off the bitching about -gpu >> to the >> > >> C preprocessor, not the C compiler. >> > >> >> > >> >> > >> stderr: >> > >> nvc-Warning-The -gpu option has no effect unless a language-specific >> > >> option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, >> -stdpar, >> > >> -cuda) >> > >> Source: >> > >> #include "confdefs.h" >> > >> #include "conffix.h" >> > >> #include <stdlib.h> >> > >> >> > >> >> > >> >> > >> > On Nov 4, 2021, at 2:49 PM, Mark Adams <mfadams(a)lbl.gov> wrote: >> > >> > >> > >> > on Perlmutter with nvhpc: >> > >> > >> > >> > Defined make macro "CPP" to "cc --use cpp32" >> > >> > Preprocessing source: >> > >> > #include "confdefs.h" >> > >> > #include "conffix.h" >> > >> > #include <stdlib.h> >> > >> > >> > >> > Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers >> > >> /tmp/petsc-jV9U1b/config.setCompilers/conftest.c >> > >> > Possible ERROR while running preprocessor: exit code 1 >> > >> > stderr: >> > >> > nvc-Error-Unknown switch: --use >> > >> > Source: >> > >> > <configure.log> >> > >> >> > >> <configure.log> >> > > >> > > >> > > >> > >> >>

1 0

Re: [petsc-dev] I am getting this error ...
by Mark Adams 04 Nov '21

04 Nov '21

OK, sorry I missed the CPPFLAGS. It is running now. Thanks, On Thu, Nov 4, 2021 at 4:43 PM Satish Balay <balay(a)mcs.anl.gov> wrote: > Multiple e-mail threads on the same issue (: > > As suggested in my earlier thread - add -mp=gpu to both CPPFLAGS and > CFLAGS [or LDFLAGS] > > Satish > > ------- > Executing: cc -o /tmp/petsc-Vvs8_T/config.setCompilers/conftest -g > -DLANDAU_DIM=2 -DLANDAU_MAX_SPECIES=10 -DLANDAU_MAX_Q=4 > /tmp/petsc-Vvs8_T/config.setCompilers/conftest.o > Possible ERROR while running linker: > stderr: > nvc-Warning-The -gpu option has no effect unless a language-specific > option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, > -cuda) > > > On Thu, 4 Nov 2021, Mark Adams wrote: > > > It is CPPFLAGS. I seem to get the same behavior. > > > > FWIW, I did get this response from NERSc but I don't know how to > interpret > > it. > > > > He seems to be saying that I don't need -mp=gpu for the device compiler > > (nvcc). He seems to think that I am adding -gpu. > > > > nvcc -- The device compiler does not need any of those flags because it > > already knows that it's being fed cuda code. The warning you're seeing is > > coming from nvc (which is the host / CPU side compiler) if you're in the > > PrgEnv-nvidia environment. You should not need to add -mp=gpu and -cuda, > > please just add the -cuda flag (to your host code) not to the device > code. > > > > I will try to talk with this guy again. > > > > Thanks, > > > > > > On Thu, Nov 4, 2021 at 4:11 PM Barry Smith <bsmith(a)petsc.dev> wrote: > > > > > > > > Yes, you need to use the CPPFLAGS which maybe called CPPCFLAGS I am > not > > > sure > > > > > > > > > On Nov 4, 2021, at 3:23 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > > > > > > Ah, CCFLAGS does not seem to work. > > > > > > On Thu, Nov 4, 2021 at 3:07 PM Barry Smith <bsmith(a)petsc.dev> wrote: > > > > > >> > > >> You have to pass in the flag to turn off the bitching about -gpu to > the > > >> C preprocessor, not the C compiler. > > >> > > >> > > >> stderr: > > >> nvc-Warning-The -gpu option has no effect unless a language-specific > > >> option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, > -stdpar, > > >> -cuda) > > >> Source: > > >> #include "confdefs.h" > > >> #include "conffix.h" > > >> #include <stdlib.h> > > >> > > >> > > >> > > >> > On Nov 4, 2021, at 2:49 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > > >> > > > >> > on Perlmutter with nvhpc: > > >> > > > >> > Defined make macro "CPP" to "cc --use cpp32" > > >> > Preprocessing source: > > >> > #include "confdefs.h" > > >> > #include "conffix.h" > > >> > #include <stdlib.h> > > >> > > > >> > Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers > > >> /tmp/petsc-jV9U1b/config.setCompilers/conftest.c > > >> > Possible ERROR while running preprocessor: exit code 1 > > >> > stderr: > > >> > nvc-Error-Unknown switch: --use > > >> > Source: > > >> > <configure.log> > > >> > > >> <configure.log> > > > > > > > > > > > > >

1 0

Re: [petsc-dev] I am getting this error ...
by Satish Balay 04 Nov '21

04 Nov '21

Multiple e-mail threads on the same issue (: As suggested in my earlier thread - add -mp=gpu to both CPPFLAGS and CFLAGS [or LDFLAGS] Satish ------- Executing: cc -o /tmp/petsc-Vvs8_T/config.setCompilers/conftest -g -DLANDAU_DIM=2 -DLANDAU_MAX_SPECIES=10 -DLANDAU_MAX_Q=4 /tmp/petsc-Vvs8_T/config.setCompilers/conftest.o Possible ERROR while running linker: stderr: nvc-Warning-The -gpu option has no effect unless a language-specific option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, -cuda) On Thu, 4 Nov 2021, Mark Adams wrote: > It is CPPFLAGS. I seem to get the same behavior. > > FWIW, I did get this response from NERSc but I don't know how to interpret > it. > > He seems to be saying that I don't need -mp=gpu for the device compiler > (nvcc). He seems to think that I am adding -gpu. > > nvcc -- The device compiler does not need any of those flags because it > already knows that it's being fed cuda code. The warning you're seeing is > coming from nvc (which is the host / CPU side compiler) if you're in the > PrgEnv-nvidia environment. You should not need to add -mp=gpu and -cuda, > please just add the -cuda flag (to your host code) not to the device code. > > I will try to talk with this guy again. > > Thanks, > > > On Thu, Nov 4, 2021 at 4:11 PM Barry Smith <bsmith(a)petsc.dev> wrote: > > > > > Yes, you need to use the CPPFLAGS which maybe called CPPCFLAGS I am not > > sure > > > > > > On Nov 4, 2021, at 3:23 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > > > > Ah, CCFLAGS does not seem to work. > > > > On Thu, Nov 4, 2021 at 3:07 PM Barry Smith <bsmith(a)petsc.dev> wrote: > > > >> > >> You have to pass in the flag to turn off the bitching about -gpu to the > >> C preprocessor, not the C compiler. > >> > >> > >> stderr: > >> nvc-Warning-The -gpu option has no effect unless a language-specific > >> option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, > >> -cuda) > >> Source: > >> #include "confdefs.h" > >> #include "conffix.h" > >> #include <stdlib.h> > >> > >> > >> > >> > On Nov 4, 2021, at 2:49 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > >> > > >> > on Perlmutter with nvhpc: > >> > > >> > Defined make macro "CPP" to "cc --use cpp32" > >> > Preprocessing source: > >> > #include "confdefs.h" > >> > #include "conffix.h" > >> > #include <stdlib.h> > >> > > >> > Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers > >> /tmp/petsc-jV9U1b/config.setCompilers/conftest.c > >> > Possible ERROR while running preprocessor: exit code 1 > >> > stderr: > >> > nvc-Error-Unknown switch: --use > >> > Source: > >> > <configure.log> > >> > >> <configure.log> > > > > > > >

1 0

Re: [petsc-dev] I am getting this error ...
by Mark Adams 04 Nov '21

04 Nov '21

It is CPPFLAGS. I seem to get the same behavior. FWIW, I did get this response from NERSc but I don't know how to interpret it. He seems to be saying that I don't need -mp=gpu for the device compiler (nvcc). He seems to think that I am adding -gpu. nvcc -- The device compiler does not need any of those flags because it already knows that it's being fed cuda code. The warning you're seeing is coming from nvc (which is the host / CPU side compiler) if you're in the PrgEnv-nvidia environment. You should not need to add -mp=gpu and -cuda, please just add the -cuda flag (to your host code) not to the device code. I will try to talk with this guy again. Thanks, On Thu, Nov 4, 2021 at 4:11 PM Barry Smith <bsmith(a)petsc.dev> wrote: > > Yes, you need to use the CPPFLAGS which maybe called CPPCFLAGS I am not > sure > > > On Nov 4, 2021, at 3:23 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > > Ah, CCFLAGS does not seem to work. > > On Thu, Nov 4, 2021 at 3:07 PM Barry Smith <bsmith(a)petsc.dev> wrote: > >> >> You have to pass in the flag to turn off the bitching about -gpu to the >> C preprocessor, not the C compiler. >> >> >> stderr: >> nvc-Warning-The -gpu option has no effect unless a language-specific >> option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, >> -cuda) >> Source: >> #include "confdefs.h" >> #include "conffix.h" >> #include <stdlib.h> >> >> >> >> > On Nov 4, 2021, at 2:49 PM, Mark Adams <mfadams(a)lbl.gov> wrote: >> > >> > on Perlmutter with nvhpc: >> > >> > Defined make macro "CPP" to "cc --use cpp32" >> > Preprocessing source: >> > #include "confdefs.h" >> > #include "conffix.h" >> > #include <stdlib.h> >> > >> > Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers >> /tmp/petsc-jV9U1b/config.setCompilers/conftest.c >> > Possible ERROR while running preprocessor: exit code 1 >> > stderr: >> > nvc-Error-Unknown switch: --use >> > Source: >> > <configure.log> >> >> <configure.log> > > >

1 0

Re: [petsc-dev] I am getting this error ...
by Barry Smith 04 Nov '21

04 Nov '21

Yes, you need to use the CPPFLAGS which maybe called CPPCFLAGS I am not sure > On Nov 4, 2021, at 3:23 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > > Ah, CCFLAGS does not seem to work. > > On Thu, Nov 4, 2021 at 3:07 PM Barry Smith <bsmith(a)petsc.dev <mailto:[email protected]>> wrote: > > You have to pass in the flag to turn off the bitching about -gpu to the C preprocessor, not the C compiler. > > > stderr: > nvc-Warning-The -gpu option has no effect unless a language-specific option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, -cuda) > Source: > #include "confdefs.h" > #include "conffix.h" > #include <stdlib.h> > > > > > On Nov 4, 2021, at 2:49 PM, Mark Adams <mfadams(a)lbl.gov <mailto:[email protected]>> wrote: > > > > on Perlmutter with nvhpc: > > > > Defined make macro "CPP" to "cc --use cpp32" > > Preprocessing source: > > #include "confdefs.h" > > #include "conffix.h" > > #include <stdlib.h> > > > > Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers /tmp/petsc-jV9U1b/config.setCompilers/conftest.c > > Possible ERROR while running preprocessor: exit code 1 > > stderr: > > nvc-Error-Unknown switch: --use > > Source: > > <configure.log> > > <configure.log>

1 0

Re: [petsc-dev] I am getting this error ...
by Mark Adams 04 Nov '21

04 Nov '21

Ah, CCFLAGS does not seem to work. On Thu, Nov 4, 2021 at 3:07 PM Barry Smith <bsmith(a)petsc.dev> wrote: > > You have to pass in the flag to turn off the bitching about -gpu to the > C preprocessor, not the C compiler. > > > stderr: > nvc-Warning-The -gpu option has no effect unless a language-specific > option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, > -cuda) > Source: > #include "confdefs.h" > #include "conffix.h" > #include <stdlib.h> > > > > > On Nov 4, 2021, at 2:49 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > > > > on Perlmutter with nvhpc: > > > > Defined make macro "CPP" to "cc --use cpp32" > > Preprocessing source: > > #include "confdefs.h" > > #include "conffix.h" > > #include <stdlib.h> > > > > Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers > /tmp/petsc-jV9U1b/config.setCompilers/conftest.c > > Possible ERROR while running preprocessor: exit code 1 > > stderr: > > nvc-Error-Unknown switch: --use > > Source: > > <configure.log> > >

1 0

Re: [petsc-dev] I am getting this error ...
by Barry Smith 04 Nov '21

04 Nov '21

You have to pass in the flag to turn off the bitching about -gpu to the C preprocessor, not the C compiler. stderr: nvc-Warning-The -gpu option has no effect unless a language-specific option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, -cuda) Source: #include "confdefs.h" #include "conffix.h" #include <stdlib.h> > On Nov 4, 2021, at 2:49 PM, Mark Adams <mfadams(a)lbl.gov> wrote: > > on Perlmutter with nvhpc: > > Defined make macro "CPP" to "cc --use cpp32" > Preprocessing source: > #include "confdefs.h" > #include "conffix.h" > #include <stdlib.h> > > Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers /tmp/petsc-jV9U1b/config.setCompilers/conftest.c > Possible ERROR while running preprocessor: exit code 1 > stderr: > nvc-Error-Unknown switch: --use > Source: > <configure.log>

1 0

I am getting this error ...
by Mark Adams 04 Nov '21

04 Nov '21

on Perlmutter with nvhpc: Defined make macro "CPP" to "cc --use cpp32" Preprocessing source: #include "confdefs.h" #include "conffix.h" #include <stdlib.h> Executing: cc --use cpp32 -I/tmp/petsc-jV9U1b/config.setCompilers /tmp/petsc-jV9U1b/config.setCompilers/conftest.c Possible ERROR while running preprocessor: exit code 1 stderr: nvc-Error-Unknown switch: --use Source:

1 0

Re: [petsc-dev] invocation of nvcc
by Satish Balay 04 Nov '21

04 Nov '21

Then you can try: CPPFLAGS=-mp-gpu CFLAGS=-mp-gpu Satish On Thu, 4 Nov 2021, Mark Adams wrote: > But I can get rid of that with: cc -E -mp-gpu > I expect that is what they will say. > > On Thu, Nov 4, 2021 at 1:40 PM Satish Balay <balay(a)mcs.anl.gov> wrote: > > > I think we went through this issue before. > > > > nvc is the 'c' compiler. And for some reason its giving 'cuda' warnings. > > > > I think you might have switched progenv last time [and avoided this > > compiler] > > > > You can try using this compiler manually - on simple code - and then seek > > help from the admins on how to avoid these warnings.. > > > > Satish > > > > - create test.c with: > > > > #include <stdlib.h> > > > > compile (or preporcess): > > > > cc -E test.c > > > > - You should get: > > > > stderr: > > nvc-Warning-The -gpu option has no effect unless a language-specific > > option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, > > -cuda) > > > > Now provide this info to the machine admins [or nvidia folk] - and ask how > > to get rid of this message. > > > > Satish > > > > > > On Thu, 4 Nov 2021, Mark Adams wrote: > > > > > Correction it is nvc: > > > > > > nvc-Warning-The -gpu option has no effect unless a language-specific > > option > > > to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, > > -cuda) > > > > > > And I add -mp=gpu to CUDAFLAGS > > > > > > > > > On Thu, Nov 4, 2021 at 1:09 PM Satish Balay <balay(a)mcs.anl.gov> wrote: > > > > > > > > > > > On Thu, 4 Nov 2021, Mark Adams wrote: > > > > > > > > > Does anyone know if PETSc calls nvcc and hence can add flags to the > > > > > invocation? nvcc wants a flag like -mp=gpu but I don't know if we do > > that > > > > > or a compiler wrapper in the environment. > > > > > > > > $ ./configure --help |grep CUDA > > > > <snip> > > > > --CUDAC=<prog> > > > > Specify the CUDA compiler > > > > --CUDAFLAGS=<string> > > > > Specify the CUDA compiler options > > > > <snip> > > > > > > > > Satish > > > > > > > > > > > >

1 0

Re: [petsc-dev] invocation of nvcc
by Mark Adams 04 Nov '21

04 Nov '21

But I can get rid of that with: cc -E -mp-gpu I expect that is what they will say. On Thu, Nov 4, 2021 at 1:40 PM Satish Balay <balay(a)mcs.anl.gov> wrote: > I think we went through this issue before. > > nvc is the 'c' compiler. And for some reason its giving 'cuda' warnings. > > I think you might have switched progenv last time [and avoided this > compiler] > > You can try using this compiler manually - on simple code - and then seek > help from the admins on how to avoid these warnings.. > > Satish > > - create test.c with: > > #include <stdlib.h> > > compile (or preporcess): > > cc -E test.c > > - You should get: > > stderr: > nvc-Warning-The -gpu option has no effect unless a language-specific > option to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, > -cuda) > > Now provide this info to the machine admins [or nvidia folk] - and ask how > to get rid of this message. > > Satish > > > On Thu, 4 Nov 2021, Mark Adams wrote: > > > Correction it is nvc: > > > > nvc-Warning-The -gpu option has no effect unless a language-specific > option > > to enable GPU code generation is used (e.g.: -acc, -mp=gpu, -stdpar, > -cuda) > > > > And I add -mp=gpu to CUDAFLAGS > > > > > > On Thu, Nov 4, 2021 at 1:09 PM Satish Balay <balay(a)mcs.anl.gov> wrote: > > > > > > > > On Thu, 4 Nov 2021, Mark Adams wrote: > > > > > > > Does anyone know if PETSc calls nvcc and hence can add flags to the > > > > invocation? nvcc wants a flag like -mp=gpu but I don't know if we do > that > > > > or a compiler wrapper in the environment. > > > > > > $ ./configure --help |grep CUDA > > > <snip> > > > --CUDAC=<prog> > > > Specify the CUDA compiler > > > --CUDAFLAGS=<string> > > > Specify the CUDA compiler options > > > <snip> > > > > > > Satish > > > > > > >

1 0