[petsc-users] Unable to configure PETSc with CUDA: Problem with thrust directory

Satish Balay balay at mcs.anl.gov
Fri Aug 29 10:30:20 CDT 2014 

Ops - I want' checking clearly..

> /usr/lib64/nvidia: file not recognized: Is a directory

Where is this coming from?

LPATH=/usr/lib64/nvidia:/usr/lib/nvidia
CUDA_SEARCH_PATH=/usr/lib64/nvidia
GLPATH=/usr/lib64/nvidia
LIBRARIES=/usr/lib64/nvidia
GLLINK=-L/usr/lib64/nvidia

Perhaps all these env variables are incorrectly set - thus messing up nvcc?

Satish


On Fri, 29 Aug 2014, Satish Balay wrote:

> #define PETSC_VERSION_GIT "v3.5.1-1951-g15ac067"
> 
> Hm - you are using recent petsc master - and it stills broken...
> 
> Perhaps Karl will check..
> 
> Perhaps we should process --with-cuda-arch in setCompilers.py - and not
> cuda.py.. [but then - there is a dependency on precision..]
> 
> Satish
> 
> 
> On Fri, 29 Aug 2014, Ashwin Srinath wrote:
> 
> > Hello,
> > 
> > My original message is awaiting approval as attaching configure.log made it
> > too large. I hope it's okay for me to repeat here:
> > 
> > I'm having a similar issue with CUDA 6.0. PETSc configured just fine with
> > version 5.0 and 5.5, and I'm using an almost identical configure command
> > (only changing --with-cuda-dir). I get the following error:
> > 
> > CUDA version error: PETSC currently requires CUDA version 4.0 or higher -
> > when compiling with CUDA
> > 
> > I suspect the problem might be with cuda.py checking for my CUDA version,
> > but I can't be sure.
> > 
> > Any suggestions?
> > 
> > 
> > Thanks,
> > Ashwin
> > 
> > 
> > On Fri, Aug 29, 2014 at 10:31 AM, Matthew Knepley <knepley at gmail.com> wrote:
> > 
> > > On Fri, Aug 29, 2014 at 9:35 AM, Dominic Meiser <dmeiser at txcorp.com>
> > > wrote:
> > >
> > >>  On 08/29/2014 08:08 AM, Matthew Knepley wrote:
> > >>
> > >>  On Fri, Aug 29, 2014 at 8:35 AM, Christophe Ortiz <
> > >> christophe.ortiz at ciemat.es> wrote:
> > >>
> > >>> Sorry about that.
> > >>>
> > >>>  Please find in attachment the configure.log file. Hope it will help
> > >>> you find the error.
> > >>>
> > >>
> > >>  Here is the problem:
> > >>
> > >>  Possible ERROR while running preprocessor: nvcc warning : The
> > >> 'compute_10' and 'sm_10' architectures are deprecated, and may be removed
> > >> in a future release.
> > >> error message = {nvcc warning : The 'compute_10' and 'sm_10'
> > >> architectures are deprecated, and may be removed in a future release.
> > >> }
> > >>
> > >>  meaning the nvcc compiler does not "work" as provided. Here is a
> > >> workaround:
> > >>
> > >>       --with-cudac='nvcc --arch=sm_35'
> > >>
> > >>  Dominic, I think that thrust.py should depend on cuda.py. Do you know
> > >> why it does not?
> > >>
> > >> In principle you are right, thrust.py should depend on cuda.py.
> > >>
> > >> However, in my opinion, thrust.py should go away as a separate package
> > >> altogether. Thrust is shipped as part of any recent version of the cuda
> > >> toolkit (I forget since which version, Paul might know) and it's always
> > >> installed in $CUDA_TOOLKIT_ROOT/include/thrust. Thus we can automatically
> > >> deduct the thrust location from the cuda location. Thrust should be
> > >> considered part of cuda.
> > >>
> > >
> > > That does not mean the configure modules should be organized that way.
> > > Thrust is a package included in CUDA, thus
> > > it should have an independent module, but depend on the CUDA module.
> > >
> > >    Matt
> > >
> > >
> > >>  Cheers,
> > >> Dominic
> > >>
> > >>
> > >>    Thanks,
> > >>
> > >>        Matt
> > >>
> > >>
> > >>
> > >>>  Christophe
> > >>>
> > >>> CIEMAT
> > >>> Laboratorio Nacional de Fusión por Confinamiento Magnético
> > >>> Unidad de Materiales
> > >>> Edificio 2 - Planta 0 - Despacho 28m
> > >>> Avenida Complutense 40,
> > >>> 28040 Madrid, Spain
> > >>> Tel: +34 91496 2582 <%2B34%2091496%202582>
> > >>> Fax: +34 91346 6442 <%2B34%2091346%206442>
> > >>>
> > >>> --
> > >>> Q
> > >>> Por favor, piense en el medio ambiente antes de imprimir este mensaje.
> > >>> Please consider the environment before printing this email.
> > >>>
> > >>>
> > >>> On Fri, Aug 29, 2014 at 3:28 PM, Matthew Knepley <knepley at gmail.com>
> > >>> wrote:
> > >>>
> > >>>> Always attach configure.log or we cannot see what went wrong.
> > >>>>
> > >>>>      Matt
> > >>>>
> > >>>>
> > >>>> On Fri, Aug 29, 2014 at 2:35 AM, Christophe Ortiz <
> > >>>> christophe.ortiz at ciemat.es> wrote:
> > >>>>
> > >>>>>   Hi Dominic and Paul,
> > >>>>>
> > >>>>>  Thanks for your answers. Unfortunately, it did not work. Same error
> > >>>>> message related to the thrust directory:
> > >>>>>
> > >>>>>
> > >>>>>  ./configure --with-x=1 --with-mpi=0 --with-cc=icc --with-cxx=icpc
> > >>>>> --with-clanguage=cxx --with-fc=ifort --with-cuda=1
> > >>>>> --with-cuda-dir=/usr/local/cuda-6.0 --with-cuda-arch=sm_35 --with-thrust=1
> > >>>>> --with-thrust-dir=/usr/local/cuda-6.0/include --with-cusp=1
> > >>>>> --with-cusp-dir=/usr/local/cuda-6.0/include --with-debugging=1
> > >>>>> --with-scalar-type=real --with-precision=double --download-fblaslapack
> > >>>>>
> > >>>>> ===============================================================================
> > >>>>>              Configuring PETSc to compile on your system
> > >>>>>
> > >>>>>
> > >>>>> ===============================================================================
> > >>>>> TESTING: checkInclude from
> > >>>>> config.headers(config/BuildSystem/config/headers.py:86)
> > >>>>>
> > >>>>>  *******************************************************************************
> > >>>>>          UNABLE to CONFIGURE with GIVEN OPTIONS    (see configure.log
> > >>>>> for details):
> > >>>>>
> > >>>>> -------------------------------------------------------------------------------
> > >>>>> --with-thrust-dir=/usr/local/cuda-6.0/include did not work
> > >>>>>
> > >>>>> *******************************************************************************
> > >>>>>
> > >>>>>
> > >>>>>  I do not see what else to do...I have a standard CUDA installation
> > >>>>> that came with thrust.
> > >>>>> I am using ubuntu 12.04, I do not know if that could help.
> > >>>>> I tried with PETSc 3.4.1 and 3.5.1. Same problem.
> > >>>>>
> > >>>>>  Christophe
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>> ------------------------------
> > >>>>>>
> > >>>>>> Message: 3
> > >>>>>> Date: Thu, 28 Aug 2014 08:38:12 -0600
> > >>>>>> From: Dominic Meiser <dmeiser at txcorp.com>
> > >>>>>> To: petsc-users at mcs.anl.gov
> > >>>>>> Subject: Re: [petsc-users] Unable to configure PETSc with CUDA:
> > >>>>>>         Problem with thrust directory
> > >>>>>> Message-ID: <53FF3ED4.1090002 at txcorp.com>
> > >>>>>> Content-Type: text/plain; charset=UTF-8; format=flowed
> > >>>>>>
> > >>>>>> Can you try configuring with
> > >>>>>>
> > >>>>>> --with-thrust-dir=/usr/local/cuda-6.0/include
> > >>>>>>
> > >>>>>> I.e. omit the 'thrust' at the end of the path. Probably you'll need
> > >>>>>> the
> > >>>>>> same fix with cusp:
> > >>>>>>
> > >>>>>> --with-cusp-dir=/usr/local/cuda-6.0/include
> > >>>>>>
> > >>>>>> Cheers,
> > >>>>>> Dominic
> > >>>>>>
> > >>>>>
> > >>>>>
> > >>>>>
> > >>>>>>
> > >>>>>> On 08/28/2014 08:13 AM, Christophe Ortiz wrote:
> > >>>>>> > Hi all,
> > >>>>>> >
> > >>>>>> > I have implemented a problem of coupled diffusion equations in 1D
> > >>>>>> > using PETSc. I was able to configure PETSC with and without MPI. The
> > >>>>>> > code runs well in both cases.
> > >>>>>> >
> > >>>>>> > Now I would like to test CUDA but I am unable to configure it. I get
> > >>>>>> > an error message related to thrust directory.
> > >>>>>> >
> > >>>>>> > The options I use to configure are the following:
> > >>>>>> >
> > >>>>>> > --with-x=1 --with-mpi=0 --with-cc=icc --with-cxx=icpc
> > >>>>>> > --with-clanguage=cxx --with-fc=ifort --with-cuda=1
> > >>>>>> > --with-cuda-dir=/usr/local/cuda-6.0 --with-cuda-arch=sm_35
> > >>>>>> > --with-cudac=/usr/local/cuda/bin/nvcc --with-thrust=1
> > >>>>>> > --with-thrust-dir=/usr/local/cuda-6.0/include/thrust --with-cusp=1
> > >>>>>> > --with-cusp-dir=/usr/local/cuda-6.0/include/cusp
> > >>>>>> > --download-txpetscgpu=1 --with-debugging=1 --with-scalar-type=real
> > >>>>>> > --with-precision=double --download-f-blas-lapack
> > >>>>>> >
> > >>>>>> > The error message I get is:
> > >>>>>> >
> > >>>>>> >
> > >>>>>> ===============================================================================
> > >>>>>> >              Configuring PETSc to compile on your system
> > >>>>>> >
> > >>>>>> ===============================================================================
> > >>>>>> > TESTING: checkInclude from
> > >>>>>> > config.headers(config/BuildSystem/config/headers.py:86)
> > >>>>>> >
> > >>>>>> *******************************************************************************
> > >>>>>> >          UNABLE to CONFIGURE with GIVEN OPTIONS    (see
> > >>>>>> configure.log
> > >>>>>> > for details):
> > >>>>>> >
> > >>>>>> -------------------------------------------------------------------------------
> > >>>>>> > --with-thrust-dir=/usr/local/cuda-6.0/include/thrust did not work
> > >>>>>> >
> > >>>>>> *******************************************************************************
> > >>>>>> >
> > >>>>>> >
> > >>>>>> > Did I miss something ?
> > >>>>>> > My GPU is a GeForce Titan Black and I am using CUDA 6.0.
> > >>>>>> >
> > >>>>>> > Many thanks in advance for your help.
> > >>>>>> > Christophe
> > >>>>>> >
> > >>>>>>
> > >>>>>>
> > >>>>>> --
> > >>>>>> Dominic Meiser
> > >>>>>> Tech-X Corporation
> > >>>>>> 5621 Arapahoe Avenue
> > >>>>>> Boulder, CO 80303
> > >>>>>> USA
> > >>>>>> Telephone: 303-996-2036
> > >>>>>> Fax: 303-448-7756
> > >>>>>> www.txcorp.com
> > >>>>>>
> > >>>>>>
> > >>>>>>
> > >>>>>> ------------------------------
> > >>>>>>
> > >>>>>> Message: 4
> > >>>>>> Date: Thu, 28 Aug 2014 08:30:40 -0600
> > >>>>>> From: Paul Mullowney <paulmullowney at gmail.com>
> > >>>>>> To: Dominic Meiser <dmeiser at txcorp.com>
> > >>>>>> Cc: petsc-users at mcs.anl.gov
> > >>>>>> Subject: Re: [petsc-users] Unable to configure PETSc with CUDA:
> > >>>>>>         Problem with thrust directory
> > >>>>>> Message-ID:
> > >>>>>>         <CAMJ8fwroFUxG-rpbMpaJEw=
> > >>>>>> 0aQdyNm9NMxMzsxWVGGPWZqhGCQ at mail.gmail.com>
> > >>>>>> Content-Type: text/plain; charset="utf-8"
> > >>>>>>
> > >>>>>> Also, get rid of --download-txpetscgpu=1. That doesn't exist anymore.
> > >>>>>> -Paul
> > >>>>>>
> > >>>>>>
> > >>>>>> On Thu, Aug 28, 2014 at 8:38 AM, Dominic Meiser <dmeiser at txcorp.com>
> > >>>>>> wrote:
> > >>>>>>
> > >>>>>> > Can you try configuring with
> > >>>>>> >
> > >>>>>> > --with-thrust-dir=/usr/local/cuda-6.0/include
> > >>>>>> >
> > >>>>>> > I.e. omit the 'thrust' at the end of the path. Probably you'll need
> > >>>>>> the
> > >>>>>> > same fix with cusp:
> > >>>>>> >
> > >>>>>> > --with-cusp-dir=/usr/local/cuda-6.0/include
> > >>>>>> >
> > >>>>>> > Cheers,
> > >>>>>> > Dominic
> > >>>>>> >
> > >>>>>> >
> > >>>>>> > On 08/28/2014 08:13 AM, Christophe Ortiz wrote:
> > >>>>>> >
> > >>>>>> >> Hi all,
> > >>>>>> >>
> > >>>>>> >> I have implemented a problem of coupled diffusion equations in 1D
> > >>>>>> using
> > >>>>>> >> PETSc. I was able to configure PETSC with and without MPI. The
> > >>>>>> code runs
> > >>>>>> >> well in both cases.
> > >>>>>> >>
> > >>>>>> >> Now I would like to test CUDA but I am unable to configure it. I
> > >>>>>> get an
> > >>>>>> >> error message related to thrust directory.
> > >>>>>> >>
> > >>>>>> >> The options I use to configure are the following:
> > >>>>>> >>
> > >>>>>> >> --with-x=1 --with-mpi=0 --with-cc=icc --with-cxx=icpc
> > >>>>>> >> --with-clanguage=cxx --with-fc=ifort --with-cuda=1
> > >>>>>> >> --with-cuda-dir=/usr/local/cuda-6.0 --with-cuda-arch=sm_35
> > >>>>>> >> --with-cudac=/usr/local/cuda/bin/nvcc --with-thrust=1
> > >>>>>> >> --with-thrust-dir=/usr/local/cuda-6.0/include/thrust --with-cusp=1
> > >>>>>> >> --with-cusp-dir=/usr/local/cuda-6.0/include/cusp
> > >>>>>> --download-txpetscgpu=1
> > >>>>>> >> --with-debugging=1 --with-scalar-type=real --with-precision=double
> > >>>>>> >> --download-f-blas-lapack
> > >>>>>> >>
> > >>>>>> >> The error message I get is:
> > >>>>>> >>
> > >>>>>> >> ============================================================
> > >>>>>> >> ===================
> > >>>>>> >>              Configuring PETSc to compile on your system
> > >>>>>> >> ============================================================
> > >>>>>> >> ===================
> > >>>>>> >> TESTING: checkInclude from config.headers(config/
> > >>>>>> >> BuildSystem/config/headers.py:86) ******************************
> > >>>>>> >> *************************************************
> > >>>>>> >>          UNABLE to CONFIGURE with GIVEN OPTIONS    (see
> > >>>>>> configure.log for
> > >>>>>> >> details):
> > >>>>>> >> ------------------------------------------------------------
> > >>>>>> >> -------------------
> > >>>>>> >> --with-thrust-dir=/usr/local/cuda-6.0/include/thrust did not work
> > >>>>>> >> ************************************************************
> > >>>>>> >> *******************
> > >>>>>> >>
> > >>>>>> >>
> > >>>>>> >> Did I miss something ?
> > >>>>>> >> My GPU is a GeForce Titan Black and I am using CUDA 6.0.
> > >>>>>> >>
> > >>>>>> >> Many thanks in advance for your help.
> > >>>>>> >> Christophe
> > >>>>>> >>
> > >>>>>> >>
> > >>>>>> >
> > >>>>>> > --
> > >>>>>> > Dominic Meiser
> > >>>>>> > Tech-X Corporation
> > >>>>>> > 5621 Arapahoe Avenue
> > >>>>>> > Boulder, CO 80303
> > >>>>>> > USA
> > >>>>>> > Telephone: 303-996-2036
> > >>>>>> > Fax: 303-448-7756
> > >>>>>> > www.txcorp.com
> > >>>>>> >
> > >>>>>> >
> > >>>>>> -------------- next part --------------
> > >>>>>> An HTML attachment was scrubbed...
> > >>>>>> URL: <
> > >>>>>> http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20140828/70c223ee/attachment-0001.html
> > >>>>>> >
> > >>>>>>
> > >>>>>> ------------------------------
> > >>>>>>
> > >>>>>> _______________________________________________
> > >>>>>> petsc-users mailing list
> > >>>>>> petsc-users at mcs.anl.gov
> > >>>>>> https://lists.mcs.anl.gov/mailman/listinfo/petsc-users
> > >>>>>>
> > >>>>>>
> > >>>>>> End of petsc-users Digest, Vol 68, Issue 63
> > >>>>>> *******************************************
> > >>>>>>
> > >>>>>> ----------------------------
> > >>>>>> Confidencialidad:
> > >>>>>> Este mensaje y sus ficheros adjuntos se dirige exclusivamente a su
> > >>>>>> destinatario y puede contener información privilegiada o confidencial. Si
> > >>>>>> no es vd. el destinatario indicado, queda notificado de que la utilización,
> > >>>>>> divulgación y/o copia sin autorización está prohibida en virtud de la
> > >>>>>> legislación vigente. Si ha recibido este mensaje por error, le rogamos que
> > >>>>>> nos lo comunique inmediatamente respondiendo al mensaje y proceda a su
> > >>>>>> destrucción.
> > >>>>>>
> > >>>>>> Disclaimer:
> > >>>>>> This message and its attached files is intended exclusively for its
> > >>>>>> recipients and may contain confidential information. If you received this
> > >>>>>> e-mail in error you are hereby notified that any dissemination, copy or
> > >>>>>> disclosure of this communication is strictly prohibited and may be
> > >>>>>> unlawful. In this case, please notify us by a reply and delete this email
> > >>>>>> and its contents immediately.
> > >>>>>> ----------------------------
> > >>>>>>
> > >>>>>>
> > >>>>>
> > >>>>
> > >>>>
> > >>>>   --
> > >>>> What most experimenters take for granted before they begin their
> > >>>> experiments is infinitely more interesting than any results to which their
> > >>>> experiments lead.
> > >>>> -- Norbert Wiener
> > >>>>
> > >>>
> > >>>
> > >>
> > >>
> > >>  --
> > >> What most experimenters take for granted before they begin their
> > >> experiments is infinitely more interesting than any results to which their
> > >> experiments lead.
> > >> -- Norbert Wiener
> > >>
> > >>
> > >>
> > >> --
> > >> Dominic Meiser
> > >> Tech-X Corporation
> > >> 5621 Arapahoe Avenue
> > >> Boulder, CO 80303
> > >> USA
> > >> Telephone: 303-996-2036
> > >> Fax: 303-448-7756www.txcorp.com
> > >>
> > >>
> > >
> > >
> > > --
> > > What most experimenters take for granted before they begin their
> > > experiments is infinitely more interesting than any results to which their
> > > experiments lead.
> > > -- Norbert Wiener
> > >
> > 
>

More information about the petsc-users mailing list