[petsc-users] Cannot eagerly initialize cuda, as doing so results in cuda error 35 (cudaErrorInsufficientDriver) : CUDA driver version is insufficient for CUDA runtime version

Fande Kong fdkong.jd at gmail.com
Thu Jan 20 17:43:41 CST 2022


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.

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.
> >>>
> >>
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.mcs.anl.gov/pipermail/petsc-users/attachments/20220120/ca0a7e64/attachment.html>


More information about the petsc-users mailing list