[petsc-users] Cannot eagerly initialize cuda, as doing so results in cuda error 35 (cudaErrorInsufficientDriver) : CUDA driver version is insufficient for CUDA runtime version
Matthew Knepley
knepley at gmail.com
Thu Jan 20 17:47:58 CST 2022
On Thu, Jan 20, 2022 at 6:44 PM Fande Kong <fdkong.jd at gmail.com> wrote:
> Thanks, Jed
>
> On Thu, Jan 20, 2022 at 4:34 PM Jed Brown <jed at jedbrown.org> wrote:
>
>> You can't create CUDA or Kokkos Vecs if you're running on a node without
>> a GPU.
>
>
> I am running the code on compute nodes that do have GPUs.
>
If you are actually running on GPUs, why would you need lazy
initialization? It would not break with GPUs present.
Matt
> With PETSc-3.16.1, I got good speedup by running GAMG on GPUs. That
> might be a bug of PETSc-main.
>
> Thanks,
>
> Fande
>
>
>
> KSPSetUp 13 1.0 6.4400e-01 1.0 2.02e+09 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 5 0 0 0 0 5 0 0 0 3140 64630 15 1.05e+02 5
> 3.49e+01 100
> KSPSolve 1 1.0 1.0109e+00 1.0 3.49e+10 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 87 0 0 0 0 87 0 0 0 34522 69556 4 4.35e-03 1
> 2.38e-03 100
> KSPGMRESOrthog 142 1.0 1.2674e-01 1.0 1.06e+10 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 27 0 0 0 0 27 0 0 0 83755 87801 0 0.00e+00 0
> 0.00e+00 100
> SNESSolve 1 1.0 4.4402e+01 1.0 4.00e+10 1.0 0.0e+00 0.0e+00
> 0.0e+00 21100 0 0 0 21100 0 0 0 901 51365 57 1.10e+03 52
> 8.78e+02 100
> SNESSetUp 1 1.0 3.9101e-05 1.0 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00e+00 0
> 0.00e+00 0
> SNESFunctionEval 2 1.0 1.7097e+01 1.0 1.60e+07 1.0 0.0e+00 0.0e+00
> 0.0e+00 8 0 0 0 0 8 0 0 0 0 1 0 0 0.00e+00 6
> 1.92e+02 0
> SNESJacobianEval 1 1.0 1.6213e+01 1.0 2.80e+07 1.0 0.0e+00 0.0e+00
> 0.0e+00 8 0 0 0 0 8 0 0 0 0 2 0 0 0.00e+00 1
> 3.20e+01 0
> SNESLineSearch 1 1.0 8.5582e+00 1.0 1.24e+08 1.0 0.0e+00 0.0e+00
> 0.0e+00 4 0 0 0 0 4 0 0 0 0 14 64153 1 3.20e+01 3
> 9.61e+01 94
> PCGAMGGraph_AGG 5 1.0 3.0509e+00 1.0 8.19e+07 1.0 0.0e+00 0.0e+00
> 0.0e+00 1 0 0 0 0 1 0 0 0 0 27 0 5 3.49e+01 9
> 7.43e+01 0
> PCGAMGCoarse_AGG 5 1.0 3.8711e+00 1.0 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00 2 0 0 0 0 2 0 0 0 0 0 0 0 0.00e+00 0
> 0.00e+00 0
> PCGAMGProl_AGG 5 1.0 7.0748e-01 1.0 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00e+00 0
> 0.00e+00 0
> PCGAMGPOpt_AGG 5 1.0 1.2904e+00 1.0 2.14e+09 1.0 0.0e+00 0.0e+00
> 0.0e+00 1 5 0 0 0 1 5 0 0 0 1661 29807 26 7.15e+02 20
> 2.90e+02 99
> GAMG: createProl 5 1.0 8.9489e+00 1.0 2.22e+09 1.0 0.0e+00 0.0e+00
> 0.0e+00 4 6 0 0 0 4 6 0 0 0 249 29666 31 7.50e+02 29
> 3.64e+02 96
> Graph 10 1.0 3.0478e+00 1.0 8.19e+07 1.0 0.0e+00 0.0e+00
> 0.0e+00 1 0 0 0 0 1 0 0 0 0 27 0 5 3.49e+01 9
> 7.43e+01 0
> MIS/Agg 5 1.0 4.1290e-01 1.0 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00e+00 0
> 0.00e+00 0
> SA: col data 5 1.0 1.9127e-02 1.0 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00e+00 0
> 0.00e+00 0
> SA: frmProl0 5 1.0 6.2662e-01 1.0 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 0 0 0 0.00e+00 0
> 0.00e+00 0
> SA: smooth 5 1.0 4.9595e-01 1.0 1.21e+08 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 244 2709 15 1.97e+02 15
> 2.55e+02 90
> GAMG: partLevel 5 1.0 4.7330e-01 1.0 6.98e+08 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 2 0 0 0 0 2 0 0 0 1475 4120 5 1.78e+02 10
> 2.55e+02 100
> PCGAMG Squ l00 1 1.0 2.6027e+00 1.0 0.00e+00 0.0 0.0e+00 0.0e+00
> 0.0e+00 1 0 0 0 0 1 0 0 0 0 0 0 0 0.00e+00 0
> 0.00e+00 0
> PCGAMG Gal l00 1 1.0 3.8406e-01 1.0 5.48e+08 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 1 0 0 0 0 1 0 0 0 1426 4270 1 1.48e+02 2
> 2.11e+02 100
> PCGAMG Opt l00 1 1.0 2.4932e-01 1.0 7.20e+07 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 289 2653 1 6.41e+01 1
> 1.13e+02 100
> PCGAMG Gal l01 1 1.0 6.6279e-02 1.0 1.09e+08 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 1645 3851 1 2.40e+01 2
> 3.64e+01 100
> PCGAMG Opt l01 1 1.0 2.9544e-02 1.0 7.15e+06 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 242 1671 1 4.84e+00 1
> 1.23e+01 100
> PCGAMG Gal l02 1 1.0 1.8874e-02 1.0 3.72e+07 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 1974 3636 1 5.04e+00 2
> 6.58e+00 100
> PCGAMG Opt l02 1 1.0 7.4353e-03 1.0 2.40e+06 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 323 1457 1 7.71e-01 1
> 2.30e+00 100
> PCGAMG Gal l03 1 1.0 2.8479e-03 1.0 4.10e+06 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 1440 2266 1 4.44e-01 2
> 5.51e-01 100
> PCGAMG Opt l03 1 1.0 8.2684e-04 1.0 2.80e+05 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 339 1667 1 6.72e-02 1
> 2.03e-01 100
> PCGAMG Gal l04 1 1.0 1.2238e-03 1.0 2.09e+05 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 170 244 1 2.05e-02 2
> 2.53e-02 100
> PCGAMG Opt l04 1 1.0 4.1008e-04 1.0 1.77e+04 1.0 0.0e+00 0.0e+00
> 0.0e+00 0 0 0 0 0 0 0 0 0 0 43 165 1 4.49e-03 1
> 1.19e-02 100
> PCSetUp 2 1.0 9.9632e+00 1.0 4.95e+09 1.0 0.0e+00 0.0e+00
> 0.0e+00 5 12 0 0 0 5 12 0 0 0 496 17826 55 1.03e+03 45
> 6.54e+02 98
> PCSetUpOnBlocks 44 1.0 9.9087e-04 1.0 2.88e+03 1.0
>
>
>
>
>
>
>> The point of lazy initialization is to make it possible to run a solve
>> that doesn't use a GPU in PETSC_ARCH that supports GPUs, regardless of
>> whether a GPU is actually present.
>>
>> Fande Kong <fdkong.jd at gmail.com> writes:
>>
>> > I spoke too soon. It seems that we have trouble creating cuda/kokkos
>> vecs
>> > now. Got Segmentation fault.
>> >
>> > Thanks,
>> >
>> > Fande
>> >
>> > Program received signal SIGSEGV, Segmentation fault.
>> > 0x00002aaab5558b11 in
>> >
>> Petsc::CUPMDevice<(Petsc::CUPMDeviceType)0>::CUPMDeviceInternal::initialize
>> > (this=0x1) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/sys/objects/device/impls/cupm/cupmdevice.cxx:54
>> > 54 PetscErrorCode CUPMDevice<T>::CUPMDeviceInternal::initialize()
>> noexcept
>> > Missing separate debuginfos, use: debuginfo-install
>> > bzip2-libs-1.0.6-13.el7.x86_64 elfutils-libelf-0.176-5.el7.x86_64
>> > elfutils-libs-0.176-5.el7.x86_64 glibc-2.17-325.el7_9.x86_64
>> > libX11-1.6.7-4.el7_9.x86_64 libXau-1.0.8-2.1.el7.x86_64
>> > libattr-2.4.46-13.el7.x86_64 libcap-2.22-11.el7.x86_64
>> > libibmad-5.4.0.MLNX20190423.1d917ae-0.1.49224.x86_64
>> > libibumad-43.1.1.MLNX20200211.078947f-0.1.49224.x86_64
>> > libibverbs-41mlnx1-OFED.4.9.0.0.7.49224.x86_64
>> > libmlx4-41mlnx1-OFED.4.7.3.0.3.49224.x86_64
>> > libmlx5-41mlnx1-OFED.4.9.0.1.2.49224.x86_64 libnl3-3.2.28-4.el7.x86_64
>> > librdmacm-41mlnx1-OFED.4.7.3.0.6.49224.x86_64
>> > librxe-41mlnx1-OFED.4.4.2.4.6.49224.x86_64 libxcb-1.13-1.el7.x86_64
>> > libxml2-2.9.1-6.el7_9.6.x86_64 numactl-libs-2.0.12-5.el7.x86_64
>> > systemd-libs-219-78.el7_9.3.x86_64 xz-libs-5.2.2-1.el7.x86_64
>> > zlib-1.2.7-19.el7_9.x86_64
>> > (gdb) bt
>> > #0 0x00002aaab5558b11 in
>> >
>> Petsc::CUPMDevice<(Petsc::CUPMDeviceType)0>::CUPMDeviceInternal::initialize
>> > (this=0x1) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/sys/objects/device/impls/cupm/cupmdevice.cxx:54
>> > #1 0x00002aaab5558db7 in
>> > Petsc::CUPMDevice<(Petsc::CUPMDeviceType)0>::getDevice
>> > (this=this at entry=0x2aaab7f37b70
>> > <CUDADevice>, device=0x115da00, id=-35, id at entry=-1) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/sys/objects/device/impls/cupm/cupmdevice.cxx:344
>> > #2 0x00002aaab55577de in PetscDeviceCreate (type=type at entry
>> =PETSC_DEVICE_CUDA,
>> > devid=devid at entry=-1, device=device at entry=0x2aaab7f37b48
>> > <defaultDevices+8>) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/sys/objects/device/interface/device.cxx:107
>> > #3 0x00002aaab5557b3a in PetscDeviceInitializeDefaultDevice_Internal
>> > (type=type at entry=PETSC_DEVICE_CUDA,
>> defaultDeviceId=defaultDeviceId at entry=-1)
>> > at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/sys/objects/device/interface/device.cxx:273
>> > #4 0x00002aaab5557bf6 in PetscDeviceInitialize
>> > (type=type at entry=PETSC_DEVICE_CUDA)
>> > at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/sys/objects/device/interface/device.cxx:234
>> > #5 0x00002aaab5661fcd in VecCreate_SeqCUDA (V=0x115d150) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/vec/vec/impls/seq/seqcuda/veccuda.c:244
>> > #6 0x00002aaab5649b40 in VecSetType (vec=vec at entry=0x115d150,
>> > method=method at entry=0x2aaab70b45b8 "seqcuda") at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/vec/vec/interface/vecreg.c:93
>> > #7 0x00002aaab579c33f in VecCreate_CUDA (v=0x115d150) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/vec/vec/impls/mpi/mpicuda/
>> > mpicuda.cu:214
>> > #8 0x00002aaab5649b40 in VecSetType (vec=vec at entry=0x115d150,
>> > method=method at entry=0x7fffffff9260 "cuda") at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/vec/vec/interface/vecreg.c:93
>> > #9 0x00002aaab5648bf1 in VecSetTypeFromOptions_Private (vec=0x115d150,
>> > PetscOptionsObject=0x7fffffff9210) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/vec/vec/interface/vector.c:1263
>> > #10 VecSetFromOptions (vec=0x115d150) at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/petsc/src/vec/vec/interface/vector.c:1297
>> > #11 0x00002aaab02ef227 in libMesh::PetscVector<double>::init
>> > (this=0x11cd1a0, n=441, n_local=441, fast=false,
>> ptype=libMesh::PARALLEL)
>> > at
>> >
>> /home/kongf/workhome/sawtooth/moosegpu/scripts/../libmesh/installed/include/libmesh/petsc_vector.h:693
>> >
>> > On Thu, Jan 20, 2022 at 1:09 PM Fande Kong <fdkong.jd at gmail.com> wrote:
>> >
>> >> Thanks, Jed,
>> >>
>> >> This worked!
>> >>
>> >> Fande
>> >>
>> >> On Wed, Jan 19, 2022 at 11:03 PM Jed Brown <jed at jedbrown.org> wrote:
>> >>
>> >>> Fande Kong <fdkong.jd at gmail.com> writes:
>> >>>
>> >>> > On Wed, Jan 19, 2022 at 11:39 AM Jacob Faibussowitsch <
>> >>> jacob.fai at gmail.com>
>> >>> > wrote:
>> >>> >
>> >>> >> Are you running on login nodes or compute nodes (I can’t seem to
>> tell
>> >>> from
>> >>> >> the configure.log)?
>> >>> >>
>> >>> >
>> >>> > I was compiling codes on login nodes, and running codes on compute
>> >>> nodes.
>> >>> > Login nodes do not have GPUs, but compute nodes do have GPUs.
>> >>> >
>> >>> > Just to be clear, the same thing (code, machine) with PETSc-3.16.1
>> >>> worked
>> >>> > perfectly. I have this trouble with PETSc-main.
>> >>>
>> >>> I assume you can
>> >>>
>> >>> export PETSC_OPTIONS='-device_enable lazy'
>> >>>
>> >>> and it'll work.
>> >>>
>> >>> I think this should be the default. The main complaint is that timing
>> the
>> >>> first GPU-using event isn't accurate if it includes initialization,
>> but I
>> >>> think this is mostly hypothetical because you can't trust any timing
>> that
>> >>> doesn't preload in some form and the first GPU-using event will almost
>> >>> always be something uninteresting so I think it will rarely lead to
>> >>> confusion. Meanwhile, eager initialization is viscerally disruptive
>> for
>> >>> lots of people.
>> >>>
>> >>
>>
>
--
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
https://www.cse.buffalo.edu/~knepley/ <http://www.cse.buffalo.edu/~knepley/>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20220120/fe894247/attachment-0001.html>
More information about the petsc-users
mailing list