February 2017 - petsc-dev - mailman3.cels.anl.gov

Re: [petsc-dev] no petsc on Edison
by Mark Adams 08 Feb '17

08 Feb '17

Thanks Hong, that sounds great. I am weary of silent optimizations like you suggest but 2x is big! and failure is very bad. So I would vote for your suggestion. ex56 is an elasticity problem. It would be nice, now that you have this experimental setup in place, to compare with hypre on a 3D scalar problem. Hypre might not spend much effort optimizing for block matrices. 3x better than hypre seems large to me. I have to suspect that hypre does not exploit blocked matrices as much as we do. Thanks again, On Wed, Feb 8, 2017 at 12:07 PM, Hong <hzhang(a)mcs.anl.gov> wrote: > I conducted tests on MatMatMult() and MatPtAP() using petsc/src/ksp/ksp/examples/tutorials/ex56.c > (gamg) on a 8-core machine (petsc machine). The output file is attached. > > Summary: > 1) non-scalable MatMatMult() for mpiaij format is 2x faster than scalable > version. The major difference between the two is dense-axpy vs. sparse-axpy. > > Currently, we set non-scalable as default, which leads to problem when > running large problems. > How about setting default as > - non-scalable for small to medium size matrices > - scalable for larger ones, e.g. > > + ierr = PetscOptionsEList("-matmatmult_via","Algorithmic > approach","MatMatMult",algTypes,nalg,algTypes[1],&alg,&flg); > > + if (!flg) { /* set default algorithm based on B->cmap->N */ > + PetscMPIInt size; > + ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); > + if ((PetscReal)(B->cmap->N)/size > 100000.0) alg = 0; /* scalable > algorithm */ > + } > > i.e., if user does NOT pick an algorithm, when ave cols per process > > 100k, use scalable implementation; otherwise, non-scalable version. > > 2) We do NOT have scalable implementation for MatPtAP() yet. > We have non-scalable PtAP and interface to Hypre's PtAP. Comparing the > two, > Petsc MatPtAP() is approx 3x faster than Hypre's. > > I'm writing a scalable MatPtAP() now. > > Hong > > On Thu, Feb 2, 2017 at 2:54 PM, Stefano Zampini <stefano.zampini(a)gmail.com > > wrote: > >> >> >> Il 02 Feb 2017 23:43, "Mark Adams" <mfadams(a)lbl.gov> ha scritto: >> >> >> >> On Thu, Feb 2, 2017 at 12:02 PM, Stefano Zampini < >> stefano.zampini(a)gmail.com> wrote: >> >>> Mark, >>> >>> I saw your configuration has hypre. If you could run with master, you >>> may try -matptap_via hypre. >>> >> >> This is worth trying. Does this even work with GAMG? >> >> >> Yes, it should work, except that the block sizes, if any, are not >> propagated to the resulting matrix. I can add it if you need it. >> >> >> >> Treb: try hypre anyway. It has its own RAP code. >> >> >> >> With that option, you will use hypre's RAP with MATAIJ >> >> >> It uses BoomerAMGBuildCoarseOperator directly with the AIJ matrices. >>> >>> Stefano >>> >>> On Feb 2, 2017, at 7:28 PM, Mark Adams <mfadams(a)lbl.gov> wrote: >>> >>> >>> >>> On Thu, Feb 2, 2017 at 11:13 AM, Hong <hzhang(a)mcs.anl.gov> wrote: >>> >>>> Mark: >>>> Try '-matmatmult_via scalable' first. If this works, should we set it >>>> as default? >>>> >>> >>> If it is robust I would say yes unless it is noticeably slower (say >>> >20%) small scale problems. >>> >>> >>> >> >> >

1 0

Re: [petsc-dev] using valgrind with new test harness
by Satish Balay 08 Feb '17

08 Feb '17

The old system was just makefile with runtargets using shell. So we had: - default 'VAR=foo' in petscvariables (i.e make variable) - and we could change the value on command line: 'make VAR=new' We avoided 'make -e' early-on [due to buggy make?] and continued to do so - to avoid picking up user env that could have conflicting stuff With the new system we have 'makefile -> gmakefile -> runex.sh' - and we can run tests at all 3 leves. I think its best if all 3 modes behave the same way (or pretty close). runex.sh will get default values from petsvariables via ./config/gmakegentest.py - so this case works fine. With the default usage of 'make' [i.e without -e] - things can become inconsistant between make and the script - when user env variables exist. For ex: $ cat makefile FOO=default_val_make all: @echo makevar: ${FOO} @echo scriptvar: `./script.sh` $ cat script.sh #!/bin/sh FOO=${FOO:-"default_val_script"} echo $FOO 1. Env variable FOO is not set [and we see the expected behavior. - we would have default_val_make and default_val_script be the same] $ ./script.sh default_val_script $ make makevar: default_val_make scriptvar: default_val_script $ 2. With 'FOO' specified to make (via command line) This is fine aswell. (no equivalent stuff for script) $ make FOO=cmd_var makevar: cmd_var scriptvar: cmd_var 3. Use env variable. (env var is ignored by make as expected - but script_var gets changed to make_var? I don't understand this - however it won't affect our usage as both should be same). But now script.sh is picking up this user env variable - which I don't think we would want to happen. $ export FOO=env_var $ make makevar: default_val_make scriptvar: default_val_make $ ./script.sh env_var There is also one more thing to worry about: the usage is: 'makefile -> gmakefile -> runex.sh' -> "bin/*.sh" and the scripts in bin can rely on env variables [PETSC_ARCH/PETSC_DIR] for defaults. So I fixed petscdiff for defaults but I'm avoiding petscmpiexec and using petsc_mpiexec_valgrind() in petsc_harness.sh instead. [petscdiff currently lets us override defaults with env variables - perhaps it will also have the same issue as above - so should be fixed - or we add petsc_diff() in petsc_harness.sh.] Satish On Wed, 8 Feb 2017, Scott Kruger wrote: > > > I have no strong objection, but a couple of comments. > > Barry said that the old test system could take env variables from the > command-line. Didn't the old system pull in petscvariables? If so, > how could it be overwritten. > > I think only documenting (and yes, I will need to update the > documentation when things have settled down) VALGRIND=1 but > allow the secret -e method would probably be OK. Users that > know about make -e should be smart enough (?) to know the danger. > > Scott > > > On 2/8/17 3:47 PM, Satish Balay wrote: > > Scott & Barry, > > > > I'd like to remove the 'env' variable change. We don't want this to be > > overwritten as it can be fragile [as users env can have incompatible > > stuff here] > > > > And petscmpiexec relies on env variables - so I'm not using it in the test > > harness. > > > > So I have a branch scott/test-harness-valgrind-v2 (with commits > > collapsed/fixed) > > > > So the following would be the way to use it. > > > > make -f gmakefile test VALGRIND=1 > > > > Do you have any objections to this? > > > > If we want them to be overwritted - then I guess we'll have to namespace > > them properly. > > > > Satish > > > > On Wed, 8 Feb 2017, Scott Kruger wrote: > > > > > > > > > > > Actually Jed's discussion of using env for DIFF had me consider another > > > solution than the one I had originally came up with. I have both mpiexec > > > and > > > diff using this new method to give what Barry > > > wants more directly: > > > https://bitbucket.org/petsc/petsc/pull-requests/631/ > > > > > > > > > Both of these now work at the makefile level: > > > > > > make -f gmakefile test VALGRIND=1 > > > > > > MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" make -e -f > > > gmakefile test > > > > > > Note that I am passing the -e flag to make which has > > > the environment variable override the value in > > > ${PETSC_ARCH}/lib/conf/petscvariables > > > This can be dangerous depending on how clean your environment is > > > because it'll override ALL variables. > > > > > > > > > And from the shell script level: > > > > > > runex1.sh -V > > > > > > MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" runex1.sh > > > > > > > > > > > > On 2/7/17 4:42 PM, Barry Smith wrote: > > > > > > > > > On Feb 7, 2017, at 5:34 PM, Satish Balay <balay(a)mcs.anl.gov> wrote: > > > > > > > > > > We cannot override stuff via env variables in the new test harness. > > > > > For > > > > > nightlybuilds > > > > > I added support to check if PETSC_ARCH has 'valgrind' string in it - > > > > > and > > > > > if so - use > > > > > valgrind. > > > > > > > > > > 26646c0bcc29652ef239540b066d00d1c8fc99c7 > > > > > > > > > > Perhaps we can force this to use valgrind with: > > > > > > > > > > ./config/gmakegentest.py --valgrind=1 > > > > > > > > I can't decide at the time I run the tests, but need to decide when I > > > > do > > > > the gmakegentest.py > > > > > > > > > > Ok - perhaps the attached patch. > > > > > > > > That is one heck of a patch, talk about code duplication resulting in > > > > endless bugs. > > > > > > > > > But then - sometimes gmakefile runs ./config/gmakegentest.py directly > > > > > [without valgrind] > > > > > > > > Fix it properly somehow; currently it is impossible and that is > > > > insanely > > > > horrible. > > > > > > > > > > > > > > Satish > > > > > > > > > > ------- > > > > > > > > > > > > > > > On Tue, 7 Feb 2017, Barry Smith wrote: > > > > > > > > > > > > > > > > > Both > > > > > > > > > > > > make -f gmakefile test > > > > > > MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec > > > > > > -valgrind " > > > > > > > > > > > > and > > > > > > > > > > > > MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind " make -f > > > > > > gmakefile test > > > > > > > > > > > > do not appear to use valgrind in the test harness (the old system > > > > > > did). > > > > > > > > > > > > How do I run all tests under valgrind automatically? > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > not ok dm_impls_plex_tests-ex18_1 > > > > > > # [0]PETSC ERROR: PetscTrFreeDefault() called from > > > > > > VecDestroy_MPI() line 18 in /sandbox/bsmith/petsc/src/vec\ > > > > > > /vec/impls/mpi/pdvec.c > > > > > > # [0]PETSC ERROR: Block [id=0(32)] at address 0xb54240 is > > > > > > # corrupted (probably write past end of array) > > > > > > # [0]PETSC ERROR: Block allocated in VecCreate_MPI_Private() > > > > > > # line > > > > > > # 497 in /sandbox/bsmith/petsc/src/vec/vec/i\ > > > > > > mpls/mpi/pbvec.c > > > > > > # [0]PETSC ERROR: --------------------- Error Message > > > > > > ------------------------------------------------------\ > > > > > > -------- > > > > > > # [0]PETSC ERROR: Memory corruption: > > > > > > # http://www.mcs.anl.gov/petsc/documentation/installation.html#valgrind > > > > > > # [0]PETSC ERROR: Corrupted memory > > > > > > # [0]PETSC ERROR: See > > > > > > # http://www.mcs.anl.gov/petsc/documentation/faq.html for > > > > > > # trouble > > > > > > # shooting. > > > > > > # [0]PETSC ERROR: Petsc Development GIT revision: > > > > > > # v3.7.5-3057-g2d11e0f GIT Date: 2017-02-07 15:39:04 -0600 > > > > > > # [0]PETSC ERROR: ../ex18 on a arch-single named es by bsmith > > > > > > # Tue > > > > > > # Feb 7 15:46:25 2017 > > > > > > # [0]PETSC ERROR: Configure options --download-ctetgen > > > > > > # --download-exodusii --download-hdf5 --download-mpich \ > > > > > > --download-netcdf PETSC_ARCH=arch-single --with-precision=single > > > > > > # [0]PETSC ERROR: #1 PetscTrFreeDefault() line 287 in > > > > > > # /sandbox/bsmith/petsc/src/sys/memory/mtr.c > > > > > > # [0]PETSC ERROR: #2 VecDestroy_MPI() line 18 in > > > > > > # /sandbox/bsmith/petsc/src/vec/vec/impls/mpi/pdvec.c > > > > > > # [0]PETSC ERROR: #3 VecDestroy() line 406 in > > > > > > # /sandbox/bsmith/petsc/src/vec/vec/interface/vector.c > > > > > > # [0]PETSC ERROR: #4 DMDestroy() line 681 in > > > > > > # /sandbox/bsmith/petsc/src/dm/interface/dm.c > > > > > > # [0]PETSC ERROR: #5 main() line 473 in > > > > > > # /sandbox/bsmith/petsc/src/dm/impls/plex/examples/tests/ex18.c > > > > > > # [0]PETSC ERROR: PETSc Option Table entries: > > > > > > # [0]PETSC ERROR: -dm_view ascii::ascii_info_detail > > > > > > # [0]PETSC ERROR: -interpolate > > > > > > > > > > > > > > > > > > > > > > > <val.patch> > > > > > > > > > > > >

1 0

Re: [petsc-dev] using valgrind with new test harness
by Scott Kruger 08 Feb '17

08 Feb '17

I have no strong objection, but a couple of comments. Barry said that the old test system could take env variables from the command-line. Didn't the old system pull in petscvariables? If so, how could it be overwritten. I think only documenting (and yes, I will need to update the documentation when things have settled down) VALGRIND=1 but allow the secret -e method would probably be OK. Users that know about make -e should be smart enough (?) to know the danger. Scott On 2/8/17 3:47 PM, Satish Balay wrote: > Scott & Barry, > > I'd like to remove the 'env' variable change. We don't want this to be > overwritten as it can be fragile [as users env can have incompatible > stuff here] > > And petscmpiexec relies on env variables - so I'm not using it in the test harness. > > So I have a branch scott/test-harness-valgrind-v2 (with commits collapsed/fixed) > > So the following would be the way to use it. > > make -f gmakefile test VALGRIND=1 > > Do you have any objections to this? > > If we want them to be overwritted - then I guess we'll have to namespace them properly. > > Satish > > On Wed, 8 Feb 2017, Scott Kruger wrote: > >> >> >> Actually Jed's discussion of using env for DIFF had me consider another >> solution than the one I had originally came up with. I have both mpiexec and >> diff using this new method to give what Barry >> wants more directly: >> https://bitbucket.org/petsc/petsc/pull-requests/631/ >> >> >> Both of these now work at the makefile level: >> >> make -f gmakefile test VALGRIND=1 >> >> MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" make -e -f >> gmakefile test >> >> Note that I am passing the -e flag to make which has >> the environment variable override the value in >> ${PETSC_ARCH}/lib/conf/petscvariables >> This can be dangerous depending on how clean your environment is >> because it'll override ALL variables. >> >> >> And from the shell script level: >> >> runex1.sh -V >> >> MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" runex1.sh >> >> >> >> On 2/7/17 4:42 PM, Barry Smith wrote: >>> >>>> On Feb 7, 2017, at 5:34 PM, Satish Balay <balay(a)mcs.anl.gov> wrote: >>>> >>>> We cannot override stuff via env variables in the new test harness. For >>>> nightlybuilds >>>> I added support to check if PETSC_ARCH has 'valgrind' string in it - and >>>> if so - use >>>> valgrind. >>>> >>>> 26646c0bcc29652ef239540b066d00d1c8fc99c7 >>>> >>>> Perhaps we can force this to use valgrind with: >>>> >>>> ./config/gmakegentest.py --valgrind=1 >>> >>> I can't decide at the time I run the tests, but need to decide when I do >>> the gmakegentest.py >>>> >>>> Ok - perhaps the attached patch. >>> >>> That is one heck of a patch, talk about code duplication resulting in >>> endless bugs. >>> >>>> But then - sometimes gmakefile runs ./config/gmakegentest.py directly >>>> [without valgrind] >>> >>> Fix it properly somehow; currently it is impossible and that is insanely >>> horrible. >>> >>>> >>>> Satish >>>> >>>> ------- >>>> >>>> >>>> On Tue, 7 Feb 2017, Barry Smith wrote: >>>> >>>>> >>>>> Both >>>>> >>>>> make -f gmakefile test MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec >>>>> -valgrind " >>>>> >>>>> and >>>>> >>>>> MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind " make -f >>>>> gmakefile test >>>>> >>>>> do not appear to use valgrind in the test harness (the old system did). >>>>> >>>>> How do I run all tests under valgrind automatically? >>>>> >>>>> >>>>> >>>>> >>>>> not ok dm_impls_plex_tests-ex18_1 >>>>> # [0]PETSC ERROR: PetscTrFreeDefault() called from >>>>> VecDestroy_MPI() line 18 in /sandbox/bsmith/petsc/src/vec\ >>>>> /vec/impls/mpi/pdvec.c >>>>> # [0]PETSC ERROR: Block [id=0(32)] at address 0xb54240 is >>>>> # corrupted (probably write past end of array) >>>>> # [0]PETSC ERROR: Block allocated in VecCreate_MPI_Private() line >>>>> # 497 in /sandbox/bsmith/petsc/src/vec/vec/i\ >>>>> mpls/mpi/pbvec.c >>>>> # [0]PETSC ERROR: --------------------- Error Message >>>>> ------------------------------------------------------\ >>>>> -------- >>>>> # [0]PETSC ERROR: Memory corruption: >>>>> # http://www.mcs.anl.gov/petsc/documentation/installation.html#valgrind >>>>> # [0]PETSC ERROR: Corrupted memory >>>>> # [0]PETSC ERROR: See >>>>> # http://www.mcs.anl.gov/petsc/documentation/faq.html for trouble >>>>> # shooting. >>>>> # [0]PETSC ERROR: Petsc Development GIT revision: >>>>> # v3.7.5-3057-g2d11e0f GIT Date: 2017-02-07 15:39:04 -0600 >>>>> # [0]PETSC ERROR: ../ex18 on a arch-single named es by bsmith Tue >>>>> # Feb 7 15:46:25 2017 >>>>> # [0]PETSC ERROR: Configure options --download-ctetgen >>>>> # --download-exodusii --download-hdf5 --download-mpich \ >>>>> --download-netcdf PETSC_ARCH=arch-single --with-precision=single >>>>> # [0]PETSC ERROR: #1 PetscTrFreeDefault() line 287 in >>>>> # /sandbox/bsmith/petsc/src/sys/memory/mtr.c >>>>> # [0]PETSC ERROR: #2 VecDestroy_MPI() line 18 in >>>>> # /sandbox/bsmith/petsc/src/vec/vec/impls/mpi/pdvec.c >>>>> # [0]PETSC ERROR: #3 VecDestroy() line 406 in >>>>> # /sandbox/bsmith/petsc/src/vec/vec/interface/vector.c >>>>> # [0]PETSC ERROR: #4 DMDestroy() line 681 in >>>>> # /sandbox/bsmith/petsc/src/dm/interface/dm.c >>>>> # [0]PETSC ERROR: #5 main() line 473 in >>>>> # /sandbox/bsmith/petsc/src/dm/impls/plex/examples/tests/ex18.c >>>>> # [0]PETSC ERROR: PETSc Option Table entries: >>>>> # [0]PETSC ERROR: -dm_view ascii::ascii_info_detail >>>>> # [0]PETSC ERROR: -interpolate >>>>> >>>>> >>>>> >>>> <val.patch> >>> >> >> -- Tech-X Corporation kruger(a)txcorp.com 5621 Arapahoe Ave, Suite A Phone: (720) 974-1841 Boulder, CO 80303 Fax: (303) 448-7756

1 0

Re: [petsc-dev] using valgrind with new test harness
by Satish Balay 08 Feb '17

08 Feb '17

Scott & Barry, I'd like to remove the 'env' variable change. We don't want this to be overwritten as it can be fragile [as users env can have incompatible stuff here] And petscmpiexec relies on env variables - so I'm not using it in the test harness. So I have a branch scott/test-harness-valgrind-v2 (with commits collapsed/fixed) So the following would be the way to use it. make -f gmakefile test VALGRIND=1 Do you have any objections to this? If we want them to be overwritted - then I guess we'll have to namespace them properly. Satish On Wed, 8 Feb 2017, Scott Kruger wrote: > > > Actually Jed's discussion of using env for DIFF had me consider another > solution than the one I had originally came up with. I have both mpiexec and > diff using this new method to give what Barry > wants more directly: > https://bitbucket.org/petsc/petsc/pull-requests/631/ > > > Both of these now work at the makefile level: > > make -f gmakefile test VALGRIND=1 > > MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" make -e -f > gmakefile test > > Note that I am passing the -e flag to make which has > the environment variable override the value in > ${PETSC_ARCH}/lib/conf/petscvariables > This can be dangerous depending on how clean your environment is > because it'll override ALL variables. > > > And from the shell script level: > > runex1.sh -V > > MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" runex1.sh > > > > On 2/7/17 4:42 PM, Barry Smith wrote: > > > > > On Feb 7, 2017, at 5:34 PM, Satish Balay <balay(a)mcs.anl.gov> wrote: > > > > > > We cannot override stuff via env variables in the new test harness. For > > > nightlybuilds > > > I added support to check if PETSC_ARCH has 'valgrind' string in it - and > > > if so - use > > > valgrind. > > > > > > 26646c0bcc29652ef239540b066d00d1c8fc99c7 > > > > > > Perhaps we can force this to use valgrind with: > > > > > > ./config/gmakegentest.py --valgrind=1 > > > > I can't decide at the time I run the tests, but need to decide when I do > > the gmakegentest.py > > > > > > Ok - perhaps the attached patch. > > > > That is one heck of a patch, talk about code duplication resulting in > > endless bugs. > > > > > But then - sometimes gmakefile runs ./config/gmakegentest.py directly > > > [without valgrind] > > > > Fix it properly somehow; currently it is impossible and that is insanely > > horrible. > > > > > > > > Satish > > > > > > ------- > > > > > > > > > On Tue, 7 Feb 2017, Barry Smith wrote: > > > > > > > > > > > Both > > > > > > > > make -f gmakefile test MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec > > > > -valgrind " > > > > > > > > and > > > > > > > > MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind " make -f > > > > gmakefile test > > > > > > > > do not appear to use valgrind in the test harness (the old system did). > > > > > > > > How do I run all tests under valgrind automatically? > > > > > > > > > > > > > > > > > > > > not ok dm_impls_plex_tests-ex18_1 > > > > # [0]PETSC ERROR: PetscTrFreeDefault() called from > > > > VecDestroy_MPI() line 18 in /sandbox/bsmith/petsc/src/vec\ > > > > /vec/impls/mpi/pdvec.c > > > > # [0]PETSC ERROR: Block [id=0(32)] at address 0xb54240 is > > > > # corrupted (probably write past end of array) > > > > # [0]PETSC ERROR: Block allocated in VecCreate_MPI_Private() line > > > > # 497 in /sandbox/bsmith/petsc/src/vec/vec/i\ > > > > mpls/mpi/pbvec.c > > > > # [0]PETSC ERROR: --------------------- Error Message > > > > ------------------------------------------------------\ > > > > -------- > > > > # [0]PETSC ERROR: Memory corruption: > > > > # http://www.mcs.anl.gov/petsc/documentation/installation.html#valgrind > > > > # [0]PETSC ERROR: Corrupted memory > > > > # [0]PETSC ERROR: See > > > > # http://www.mcs.anl.gov/petsc/documentation/faq.html for trouble > > > > # shooting. > > > > # [0]PETSC ERROR: Petsc Development GIT revision: > > > > # v3.7.5-3057-g2d11e0f GIT Date: 2017-02-07 15:39:04 -0600 > > > > # [0]PETSC ERROR: ../ex18 on a arch-single named es by bsmith Tue > > > > # Feb 7 15:46:25 2017 > > > > # [0]PETSC ERROR: Configure options --download-ctetgen > > > > # --download-exodusii --download-hdf5 --download-mpich \ > > > > --download-netcdf PETSC_ARCH=arch-single --with-precision=single > > > > # [0]PETSC ERROR: #1 PetscTrFreeDefault() line 287 in > > > > # /sandbox/bsmith/petsc/src/sys/memory/mtr.c > > > > # [0]PETSC ERROR: #2 VecDestroy_MPI() line 18 in > > > > # /sandbox/bsmith/petsc/src/vec/vec/impls/mpi/pdvec.c > > > > # [0]PETSC ERROR: #3 VecDestroy() line 406 in > > > > # /sandbox/bsmith/petsc/src/vec/vec/interface/vector.c > > > > # [0]PETSC ERROR: #4 DMDestroy() line 681 in > > > > # /sandbox/bsmith/petsc/src/dm/interface/dm.c > > > > # [0]PETSC ERROR: #5 main() line 473 in > > > > # /sandbox/bsmith/petsc/src/dm/impls/plex/examples/tests/ex18.c > > > > # [0]PETSC ERROR: PETSc Option Table entries: > > > > # [0]PETSC ERROR: -dm_view ascii::ascii_info_detail > > > > # [0]PETSC ERROR: -interpolate > > > > > > > > > > > > > > > <val.patch> > > > >

1 0

Re: [petsc-dev] no petsc on Edison
by Hong 08 Feb '17

08 Feb '17

I conducted tests on MatMatMult() and MatPtAP() using petsc/src/ksp/ksp/examples/tutorials/ex56.c (gamg) on a 8-core machine (petsc machine). The output file is attached. Summary: 1) non-scalable MatMatMult() for mpiaij format is 2x faster than scalable version. The major difference between the two is dense-axpy vs. sparse-axpy. Currently, we set non-scalable as default, which leads to problem when running large problems. How about setting default as - non-scalable for small to medium size matrices - scalable for larger ones, e.g. + ierr = PetscOptionsEList("-matmatmult_via","Algorithmic approach","MatMatMult",algTypes,nalg,algTypes[1],&alg,&flg); + if (!flg) { /* set default algorithm based on B->cmap->N */ + PetscMPIInt size; + ierr = MPI_Comm_size(comm,&size);CHKERRQ(ierr); + if ((PetscReal)(B->cmap->N)/size > 100000.0) alg = 0; /* scalable algorithm */ + } i.e., if user does NOT pick an algorithm, when ave cols per process > 100k, use scalable implementation; otherwise, non-scalable version. 2) We do NOT have scalable implementation for MatPtAP() yet. We have non-scalable PtAP and interface to Hypre's PtAP. Comparing the two, Petsc MatPtAP() is approx 3x faster than Hypre's. I'm writing a scalable MatPtAP() now. Hong On Thu, Feb 2, 2017 at 2:54 PM, Stefano Zampini <stefano.zampini(a)gmail.com> wrote: > > > Il 02 Feb 2017 23:43, "Mark Adams" <mfadams(a)lbl.gov> ha scritto: > > > > On Thu, Feb 2, 2017 at 12:02 PM, Stefano Zampini < > stefano.zampini(a)gmail.com> wrote: > >> Mark, >> >> I saw your configuration has hypre. If you could run with master, you may >> try -matptap_via hypre. >> > > This is worth trying. Does this even work with GAMG? > > > Yes, it should work, except that the block sizes, if any, are not > propagated to the resulting matrix. I can add it if you need it. > > > > Treb: try hypre anyway. It has its own RAP code. > > > > With that option, you will use hypre's RAP with MATAIJ > > > It uses BoomerAMGBuildCoarseOperator directly with the AIJ matrices. >> >> Stefano >> >> On Feb 2, 2017, at 7:28 PM, Mark Adams <mfadams(a)lbl.gov> wrote: >> >> >> >> On Thu, Feb 2, 2017 at 11:13 AM, Hong <hzhang(a)mcs.anl.gov> wrote: >> >>> Mark: >>> Try '-matmatmult_via scalable' first. If this works, should we set it as >>> default? >>> >> >> If it is robust I would say yes unless it is noticeably slower (say >20%) >> small scale problems. >> >> >> > >

1 0

Re: [petsc-dev] using valgrind with new test harness
by Scott Kruger 08 Feb '17

08 Feb '17

Actually Jed's discussion of using env for DIFF had me consider another solution than the one I had originally came up with. I have both mpiexec and diff using this new method to give what Barry wants more directly: https://bitbucket.org/petsc/petsc/pull-requests/631/ Both of these now work at the makefile level: make -f gmakefile test VALGRIND=1 MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" make -e -f gmakefile test Note that I am passing the -e flag to make which has the environment variable override the value in ${PETSC_ARCH}/lib/conf/petscvariables This can be dangerous depending on how clean your environment is because it'll override ALL variables. And from the shell script level: runex1.sh -V MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind" runex1.sh On 2/7/17 4:42 PM, Barry Smith wrote: > >> On Feb 7, 2017, at 5:34 PM, Satish Balay <balay(a)mcs.anl.gov> wrote: >> >> We cannot override stuff via env variables in the new test harness. For nightlybuilds >> I added support to check if PETSC_ARCH has 'valgrind' string in it - and if so - use >> valgrind. >> >> 26646c0bcc29652ef239540b066d00d1c8fc99c7 >> >> Perhaps we can force this to use valgrind with: >> >> ./config/gmakegentest.py --valgrind=1 > > I can't decide at the time I run the tests, but need to decide when I do the gmakegentest.py >> >> Ok - perhaps the attached patch. > > That is one heck of a patch, talk about code duplication resulting in endless bugs. > >> But then - sometimes gmakefile runs ./config/gmakegentest.py directly [without valgrind] > > Fix it properly somehow; currently it is impossible and that is insanely horrible. > >> >> Satish >> >> ------- >> >> >> On Tue, 7 Feb 2017, Barry Smith wrote: >> >>> >>> Both >>> >>> make -f gmakefile test MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind " >>> >>> and >>> >>> MPIEXEC="/sandbox/bsmith/petsc/bin/petscmpiexec -valgrind " make -f gmakefile test >>> >>> do not appear to use valgrind in the test harness (the old system did). >>> >>> How do I run all tests under valgrind automatically? >>> >>> >>> >>> >>> not ok dm_impls_plex_tests-ex18_1 >>> # [0]PETSC ERROR: PetscTrFreeDefault() called from VecDestroy_MPI() line 18 in /sandbox/bsmith/petsc/src/vec\ >>> /vec/impls/mpi/pdvec.c >>> # [0]PETSC ERROR: Block [id=0(32)] at address 0xb54240 is corrupted (probably write past end of array) >>> # [0]PETSC ERROR: Block allocated in VecCreate_MPI_Private() line 497 in /sandbox/bsmith/petsc/src/vec/vec/i\ >>> mpls/mpi/pbvec.c >>> # [0]PETSC ERROR: --------------------- Error Message ------------------------------------------------------\ >>> -------- >>> # [0]PETSC ERROR: Memory corruption: http://www.mcs.anl.gov/petsc/documentation/installation.html#valgrind >>> # [0]PETSC ERROR: Corrupted memory >>> # [0]PETSC ERROR: See http://www.mcs.anl.gov/petsc/documentation/faq.html for trouble shooting. >>> # [0]PETSC ERROR: Petsc Development GIT revision: v3.7.5-3057-g2d11e0f GIT Date: 2017-02-07 15:39:04 -0600 >>> # [0]PETSC ERROR: ../ex18 on a arch-single named es by bsmith Tue Feb 7 15:46:25 2017 >>> # [0]PETSC ERROR: Configure options --download-ctetgen --download-exodusii --download-hdf5 --download-mpich \ >>> --download-netcdf PETSC_ARCH=arch-single --with-precision=single >>> # [0]PETSC ERROR: #1 PetscTrFreeDefault() line 287 in /sandbox/bsmith/petsc/src/sys/memory/mtr.c >>> # [0]PETSC ERROR: #2 VecDestroy_MPI() line 18 in /sandbox/bsmith/petsc/src/vec/vec/impls/mpi/pdvec.c >>> # [0]PETSC ERROR: #3 VecDestroy() line 406 in /sandbox/bsmith/petsc/src/vec/vec/interface/vector.c >>> # [0]PETSC ERROR: #4 DMDestroy() line 681 in /sandbox/bsmith/petsc/src/dm/interface/dm.c >>> # [0]PETSC ERROR: #5 main() line 473 in /sandbox/bsmith/petsc/src/dm/impls/plex/examples/tests/ex18.c >>> # [0]PETSC ERROR: PETSc Option Table entries: >>> # [0]PETSC ERROR: -dm_view ascii::ascii_info_detail >>> # [0]PETSC ERROR: -interpolate >>> >>> >>> >> <val.patch> > -- Tech-X Corporation kruger(a)txcorp.com 5621 Arapahoe Ave, Suite A Phone: (720) 974-1841 Boulder, CO 80303 Fax: (303) 448-7756

1 0

Re: [petsc-dev] install petsc on windows 10 with mpi
by Satish Balay 08 Feb '17

08 Feb '17

Richard, The current issue is with MS compiler build - with MS-MPI. WSL doesn't help for this case. [assuming 'abaqus' requirement requires a native build with MS compilers] Wrt WSL/PETSc - yeah build goes through fine - but 'make alltests' give random hangs/errors in MPI calls for me.. Satish On Wed, 8 Feb 2017, Richard Mills wrote: > Hi Mohammad, > > Have you considered not using Cygwin and instead using the beta "Windows > Subsystem for Linux" that provides a layer for running ELF binaries and > using Ubuntu packages? I've not tried building PETSc with it yet, but I've > been impressed at how well it works when I have tried other things. Quick > Google search comes up with this page that briefly talks about it and how > to enable it: > > http://www.zdnet.com/article/ubuntu-and-bash-arrive-on-windows-10/ > > Best regards, > Richard > > On Sun, Feb 5, 2017 at 8:51 PM, Mohammad S Dodaran <mdodar1(a)lsu.edu> wrote: > > > Dear Petsc developers, > > > > I am trying to install Petsc on windows 10 but I have a problem. I > > installed MS-MPI v8 <http://go.microsoft.com/FWLink/p/?LinkID=389556> separately > > and in the configure options I use > > > > --with-mpi-include="/cygdrive/c/Program\ Files\ $x86$/Microsoft\ > > SDKs/MPI/Include" --with-mpi-lib="/cygdrive/c/Program\ Files\ > > $x86$/Microsoft\ SDKs/MPI/Lib/x64/msmpi.lib" > > > > > > But it gives error which says: > > > > *************************************************************** > > ERROR in COMMAND LINE ARGUMENT to ./configure > > ------------------------------------------------------------ > > ------------------- > > Invalid directory: [/cygdrive/c/Program\ Files\ $x86$/Microsoft\ > > SDKs/MPI/Include for key with-mpi-include > > > > How should I solve it? > > > > Mohammad > > >

1 0

Re: [petsc-dev] install petsc on windows 10 with mpi
by Richard Mills 08 Feb '17

08 Feb '17

Hi Mohammad, Have you considered not using Cygwin and instead using the beta "Windows Subsystem for Linux" that provides a layer for running ELF binaries and using Ubuntu packages? I've not tried building PETSc with it yet, but I've been impressed at how well it works when I have tried other things. Quick Google search comes up with this page that briefly talks about it and how to enable it: http://www.zdnet.com/article/ubuntu-and-bash-arrive-on-windows-10/ Best regards, Richard On Sun, Feb 5, 2017 at 8:51 PM, Mohammad S Dodaran <mdodar1(a)lsu.edu> wrote: > Dear Petsc developers, > > I am trying to install Petsc on windows 10 but I have a problem. I > installed MS-MPI v8 <http://go.microsoft.com/FWLink/p/?LinkID=389556> separately > and in the configure options I use > > --with-mpi-include="/cygdrive/c/Program\ Files\ $x86$/Microsoft\ > SDKs/MPI/Include" --with-mpi-lib="/cygdrive/c/Program\ Files\ > $x86$/Microsoft\ SDKs/MPI/Lib/x64/msmpi.lib" > > > But it gives error which says: > > *************************************************************** > ERROR in COMMAND LINE ARGUMENT to ./configure > ------------------------------------------------------------ > ------------------- > Invalid directory: [/cygdrive/c/Program\ Files\ $x86$/Microsoft\ > SDKs/MPI/Include for key with-mpi-include > > How should I solve it? > > Mohammad >

1 0

Re: [petsc-dev] install petsc on windows 10 with mpi
by Mohammad S Dodaran 07 Feb '17

07 Feb '17

I have already installed pets c on linux. The problem is that i dont have access to abaqus on linux. So i have to shift all these packages(damask petsc .....) to windows which clearly will cause endless pain Get Outlook for Android<https://aka.ms/ghei36> ________________________________ From: Satish Balay <balay(a)mcs.anl.gov> Sent: Tuesday, February 7, 2017 5:44:52 PM To: Mohammad S Dodaran Cc: petsc-dev Subject: Re: [petsc-dev] install petsc on windows 10 with mpi >>> C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but now appears to be compiling using a non-MPICH mpi.h"^M <<<<< I have no clue why you get this error - since I don't see the flag PETSC_HAVE_MPICH_NUMVERSION set anymore. Its best if you attempt a clean install [reboot, run cygwin setup.exe, start with a fresh tarball] BTW: do you have a requirement for windows - or is it just a convinence? Most of the externalpackages are untested on windows - and https://damask.mpie.de/Installation/PETSc doesn't claim windows build is test. You might want to explore linux build [perhaps via VM] Satish On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > would you please check this log file. > > > the previous one belong to the make that i did on cygwin terminal > > > this time i tried to make on the command prompt of windows parallel studio by doing cygwin.bat and then i did make this time again the mpi error appeared > > > i will try to patch the file that you send once again and do make > > ________________________________ > From: Satish Balay <balay(a)mcs.anl.gov> > Sent: Tuesday, February 7, 2017 5:10:05 PM > To: Mohammad S Dodaran > Cc: petsc-dev > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > Yeah - thats was I was suggesting. Before that - can you try the following and see if it works? > > make all-legacy > > For a clean install - I would: > > 1. reboot > 2. run cygwin setup.exe to completion [which should run rebaseall] > 3. start with a fresh traball and redo the configure and make > > Satish > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > by fresh tar ball you mean that its better to download the petsc and start everything from the scratch? > > > > > > > > ________________________________ > > From: Satish Balay <balay(a)mcs.anl.gov> > > Sent: Tuesday, February 7, 2017 5:00:18 PM > > To: Mohammad S Dodaran > > Cc: petsc-dev > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > I don't see the realvent messages - so I have no idea whats going wrong here. > > > > You could try manually running the compile command - but I'm not sure > > if that would help: > > > > /home/petsc-3.6.4/bin/win32fe/win32fe cl -c -MT -wd4996 -O2 -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 /home/petsc-3.6.4/src/sys/info/verboseinfo.c -o first/obj/src/sys/info/verboseinfo.o > > > > My suggestion is to start with a fresh tarball - to eliminate werid issues. > > > > Satish > > > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > > > is there anything else that i have to do? > > > > > > > > > ________________________________ > > > From: Mohammad S Dodaran > > > Sent: Tuesday, February 7, 2017 4:31:24 PM > > > To: petsc-dev > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > i did > > > > > > make clean > > > make V=1 MAKE_NP=1 > > > > > > > > > and i attached the make.log > > > > > > > > > > > > when I try to make the error that appears on the screen is like this: > > > > > > ----------------------------------------- > > > Using C/C++ compile: /home/petsc-3.6.4/bin/win32fe/win32fe cl -c -MT -wd4996 -O2 -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 > > > Using Fortran compile: /home/petsc-3.6.4/bin/win32fe/win32fe ifort -c -MT -O3 -fpp -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 > > > ----------------------------------------- > > > Using C/C++ linker: /home/petsc-3.6.4/bin/win32fe/win32fe cl > > > Using C/C++ flags: -MT -wd4996 -O2 > > > Using Fortran linker: /home/petsc-3.6.4/bin/win32fe/win32fe ifort > > > Using Fortran flags: -MT -O3 -fpp > > > ----------------------------------------- > > > Using libraries: -L/home/petsc-3.6.4/first/lib -lpetsc -lflapack -lfblas /cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpi.lib /cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpifec.lib Gdi32.lib User32.lib Advapi32.lib Kernel32.lib Ws2_32.lib > > > ------------------------------------------ > > > Using mpiexec: mpiexec > > > ========================================== > > > Building PETSc using GNU Make with 1 build threads > > > ========================================== > > > make[2]: Entering directory '/home/petsc-3.6.4' > > > /usr/bin/python ./config/gmakegen.py --petsc-arch=first > > > /home/petsc-3.6.4/bin/win32fe/win32fe cl -c -MT -wd4996 -O2 -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 /home/petsc-3.6.4/src/sys/info/verboseinfo.c -o first/obj/src/sys/info/verboseinfo.o > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/verboseinfo.o] Error 188 > > > make[2]: Leaving directory '/home/petsc-3.6.4' > > > make[1]: *** [/home/petsc-3.6.4/lib/petsc/conf/rules:107: gnumake] Error 2 > > > make[1]: Leaving directory '/home/petsc-3.6.4' > > > **************************ERROR************************************* > > > Error during compile, check first/lib/petsc/conf/make.log > > > Send it and first/lib/petsc/conf/configure.log to petsc-maint(a)mcs.anl.gov > > > ******************************************************************** > > > make: *** [makefile:30: all] Error 1 > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > > > > > > > > > ________________________________ > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > Sent: Tuesday, February 7, 2017 3:44:01 PM > > > To: Mohammad S Dodaran > > > Cc: petsc-dev > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > okay > > > > > > > > > > > > i patched the file you send > > > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > $ patch -Np1 < mpich-ver-test.patch > > > > (Stripping trailing CRs from patch; use --binary to disable.) > > > > patching file config/BuildSystem/config/packages/MPI.py > > > > > > wrt this patch - you would run configure again. But the error messages below are weird. > > > > > > Can you do: > > > > > > make clean > > > make V=1 MAKE_NP=1 > > > > > > and resend logs. > > > > > > Satish > > > > > > > > > > > > > > > the tried to make > > > > it failed > > > > > > > > Building PETSc using GNU Make with 4 build threads > > > > ========================================== > > > > make[2]: Entering directory '/home/petsc-3.6.4' > > > > Use "/usr/bin/make V=1" to see the verbose compile lines. > > > > CC first/obj/src/sys/info/verboseinfo.o > > > > CC first/obj/src/sys/info/ftn-custom/zverboseinfof.o > > > > CC first/obj/src/sys/info/ftn-auto/verboseinfof.o > > > > CC first/obj/src/sys/ftn-custom/zsys.o > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/ftn-custom/zverboseinfof.o] Error 188 > > > > make[2]: *** Waiting for unfinished jobs.... > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/verboseinfo.o] Error 188 > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/ftn-auto/verboseinfof.o] Err or 188 > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/ftn-custom/zsys.o] Error 188 > > > > make[2]: Leaving directory '/home/petsc-3.6.4' > > > > make[1]: *** [/home/petsc-3.6.4/lib/petsc/conf/rules:107: gnumake] Error 2 > > > > make[1]: Leaving directory '/home/petsc-3.6.4' > > > > **************************ERROR************************************* > > > > Error during compile, check first/lib/petsc/conf/make.log > > > > Send it and first/lib/petsc/conf/configure.log to petsc-maint(a)mcs.anl.gov > > > > ******************************************************************** > > > > make: *** [makefile:30: all] Error 1 > > > > > > > > i have attached the make.log > > > > > > > > > > > > > > > > ________________________________ > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > Sent: Tuesday, February 7, 2017 3:25:52 PM > > > > To: Mohammad S Dodaran > > > > Cc: petsc-dev > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > nope. > > > > > > > > satish > > > > > > > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > do you want me to remove the env and configure it once again and send you the log files? > > > > > > > > > > > > > > > ________________________________ > > > > > From: Mohammad S Dodaran > > > > > Sent: Tuesday, February 7, 2017 3:13:55 PM > > > > > To: petsc-dev > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > sorry for responding late. something came up and i couldn't start working on this earlier. > > > > > > > > > > env is still there, and i removed those three line > > > > > > > > > > then: tried to make > > > > > > > > > > i attached the configure and make log files > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > i tried to patch the file that you send but the patch command does not work in cygwin > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 ~ > > > > > $ cd /home/petsc-3.6.4 > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > > $ patch -Np1 < mpich-ver-test.patch > > > > > bash: patch: command not found > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > > $ > > > > > > > > > > > > > > > am i missing something? > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > Sent: Tuesday, February 7, 2017 12:09:38 AM > > > > > To: petsc-dev > > > > > Cc: Mohammad S Dodaran > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > Please try the attached patch - and see if it works [without removing IntelMPI from env] > > > > > > > > > > you can apply it with 'patch' command from cygwin > > > > > > > > > > cd /home/petsc-3.6.4 > > > > > patch -Np1 < mpich-ver-test.patch > > > > > > > > > > Satish > > > > > > > > > > On Mon, 6 Feb 2017, Satish Balay wrote: > > > > > > > > > > > Thinking about this - configure should not fail with MSMPI even if Intel-MPI is in the default path. > > > > > > > > > > > > A workarround is [after successful configure] - edit PETSC_ARCH/include/petscconf.h and remove the lines: > > > > > > > > > > > > #ifndef PETSC_HAVE_MPICH_NUMVERSION > > > > > > #define PETSC_HAVE_MPICH_NUMVERSION 30200300 > > > > > > #endif > > > > > > > > > > > > And then run 'make all' > > > > > > > > > > > > Will have to check on the fix.. > > > > > > > > > > > > Satish > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > I will remove intel mpi env tomorrow first thing in the morning and then i will send you the log files. > > > > > > > Mohammad > > > > > > > > > > > > > > Get Outlook for Android<https://aka.ms/ghei36> > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > Sent: Monday, February 6, 2017 11:34:00 PM > > > > > > > To: petsc-dev > > > > > > > Cc: Mohammad S Dodaran > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > BTW: Why not use petsc-3.7? > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Satish Balay wrote: > > > > > > > > > > > > > > > Please send the logs for the case that is supporsed to work - but failed. > > > > > > > > > > > > > > > > without the relavent logs - I have no idea what failed. > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > As i stated in my previous email. > > > > > > > > > " i tried to make with and without intel-mpi ENV . in both cases the error had been changed." And the error didnt have any thing to do with intel_mpi ENV. Do you still want me to remove the intel_ mpi env and send you the log file? > > > > > > > > > > > > > > > > > > Get Outlook for Android<https://aka.ms/ghei36> > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > Sent: Monday, February 6, 2017 11:15:19 PM > > > > > > > > > To: petsc-dev > > > > > > > > > Cc: Mohammad S Dodaran > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > never mind. > > > > > > > > > > > > > > > > > > >>>>>>>>>>>>>> > > > > > > > > > #line 2416 "c:\\program files (x86)\\intelswtools\\compilers_and_libraries_2017.1.143\\windows\\mpi\\intel64\\include\\mpi.h"^M > > > > > > > > > ^M > > > > > > > > > #line 2418 "c:\\program files (x86)\\intelswtools\\compilers_and_libraries_2017.1.143\\windows\\mpi\\intel64\\include\\mpi.h"^M > > > > > > > > > ^M > > > > > > > > > #line 4 "C:\\cygwin64\\tmp\\PE5639~1\\CONFIG~1.MPI\\conftest.c"^M > > > > > > > > > int mpich_ver = 30200300; > > > > > > > > > Defined "HAVE_MPICH_NUMVERSION" to "30200300" > > > > > > > > > Checking for functions [MPI_Alltoallw] in library ['/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpi.lib', '/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/ms\ > > > > > > > > > mpifec.lib'] [] > > > > > > > > > Pushing language C > > > > > > > > > > > > > > > > > > <<<<<<<<<< > > > > > > > > > > > > > > > > > > Its still picking up mpi.h from intel MPI > > > > > > > > > > > > > > > > > > >>>>>>>>>>>> > > > > > > > > > INCLUDE=C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\compiler\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\compiler\include\intel64;C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\INCLUDE;C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\ATLMFC\INCLUDE;C:\Program Files (x86)\Windows Kits\8.1\include\shared;C:\Program Files (x86)\Windows Kits\8.1\include\um;C:\Program Files (x86)\Windows Kits\8.1\include\winrt;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mpi\intel64\bin\..\..\intel64\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\ipp\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mkl\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\tbb\bin\..\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\daal\include;C:\Program Files (x86)\IntelSWTools\itac_2017\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mpi\intel64\bin\..\..\intel64\include; > > > > > > > > > <<<<<<<<<<< > > > > > > > > > > > > > > > > > > The env has "C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mpi\intel64\bin\..\..\intel64\include" from where mpi.h is getting picked up. > > > > > > > > > > > > > > > > > > If you can't remove it perhaps you can try using it instead of MSMPI > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Satish Balay wrote: > > > > > > > > > > > > > > > > > > > can you also send make.log? > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > i was trying to repair my parallel_studio_xe_2017. then i configure it once again and then i tried to make with and without intel-mpi ENV . in both cases the error had been changed. i have attached the configure.log > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > > > Sent: Monday, February 6, 2017 10:10:43 PM > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > Cc: petsc-dev > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > Them Intel-MPI was not removed. You can resend configure.log to confirm.. > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > I removed the INTEL-MPI from the environmental variables and I still have the same problem. When I configure it looks like it accepts the mpi_include and mpi_lib but when I want to build , it says the configure was with mpi but in build it appears to compile is using a non-mpich. > > > > > > > > > > > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > > > > Sent: Monday, February 6, 2017 6:13:04 PM > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > Cc: petsc-dev > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > 1. configure.log is a cygwin simlink - so native mail app is unable to > > > > > > > > > > > > attach correctly. The actual file is in > > > > > > > > > > > > PETSC_DIR/PETSC_ARCH/lib/petsc/conf > > > > > > > > > > > > > > > > > > > > > > > > 2. Suggest doing the build without fftw and hdf5 packages [once the > > > > > > > > > > > > basic build succeeds - you can look at adding these pacakges.] > > > > > > > > > > > > > > > > > > > > > > > > 3. remove INTEL-MPI from your env - before running configure. [you > > > > > > > > > > > > might be able to use intel-mpi instead of ms-mpi - but we don't have > > > > > > > > > > > > any experience with it] > > > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > > > now I'm trying to configure petsc 3.7 > > > > > > > > > > > > > i used the format you showd in your previous email for the directories of fftw and hdf5 as below > > > > > > > > > > > > > > > > > > > > > > > > > > --with-fftw-dir=/cygdrive/c/cygwin64/home/fftw --with-hdf5-dir=/cygdrive/c/cygwin64/home/hdf5 > > > > > > > > > > > > > > > > > > > > > > > > > > the result is the following > > > > > > > > > > > > > > > > > > > > > > > > > > TESTING: alternateConfigureLibrary from config.packages.glut(config/BuildSystem/config/package.py:742) TESTING: alternateConfigureLibrary from config.packages.giflib(config/BuildSystem/config/package.py:742) TESTING: alternateConfigureLibrary from config.packages.scientificpython(config/BuildSystem/config/package.py:742) TESTING: alternateConfigureLibrary from config.packages.fiat(config/BuildSystem/config/package.py:742) TESTING: configureLibrary from config.packages.fftw(config/BuildSystem/config/package.py:679) ******************************************************************************* > > > > > > > > > > > > > UNABLE to CONFIGURE with GIVEN OPTIONS (see configure.log for details): > > > > > > > > > > > > > ------------------------------------------------------------------------------- > > > > > > > > > > > > > --with-fftw-dir=/cygdrive/c/cygwin64/home/fftw did not work > > > > > > > > > > > > > ******************************************************************************* > > > > > > > > > > > > > > > > > > > > > > > > > > i also attached the configure.log > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > > > > > Sent: Monday, February 6, 2017 4:28:09 PM > > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > > Cc: Matthew Knepley; petsc-dev > > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > > > >>>> > > > > > > > > > > > > > > > > > > > > > > > > > > Configure Options: --configModules=PETSc.Configure --optionsModule=config.compilerOptions --with-cc="win32fe cl" --with-fc="win32fe ifort" --with-cxx="win32fe cl" --with-mpi-include="[/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include,/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64]" --with-mpi-lib="[/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpi.lib,/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpifec.lib]" --with-fftw-dir="[/cygdrive/c/cygwin64/home/fftw]" --with-hdf5-dir="[/cygdrive/c/cygwin64/home/hdf5]" --download-fblaslapack=1 --with-c2html=0 --with-debugging=0 --with-ssl=0 --with-x=0 PETSC_ARCH= PETSC_DIR=/home/petsc-3.6.4 > > > > > > > > > > > > > <<<< > > > > > > > > > > > > > > > > > > > > > > > > > > > 'PETSC_ARCH=' > > > > > > > > > > > > > > > > > > > > > > > > > > Can you rerun with a non-empty PETSC_ARCH string - or let configure choost default.. > > > > > > > > > > > > > > > > > > > > > > > > > > > --with-fftw-dir="[/cygdrive/c/cygwin64/home/fftw]" --with-hdf5-dir="[/cygdrive/c/cygwin64/home/hdf5]" > > > > > > > > > > > > > > > > > > > > > > > > > > I don't think I've built these packages with MS compilers > > > > > > > > > > > > > before. Somehow they are not picked up anyway. For one - the format > > > > > > > > > > > > > here should be the following [but I don't know if that will make > > > > > > > > > > > > > configure check for them]: > > > > > > > > > > > > > > > > > > > > > > > > > > --with-fftw-dir=/cygdrive/c/cygwin64/home/fftw --with-hdf5-dir=/cygdrive/c/cygwin64/home/hdf5 > > > > > > > > > > > > > > > > > > > > > > > > > > > PETSC_DIR=/home/petsc-3.6.4 > > > > > > > > > > > > > > > > > > > > > > > > > > Also - please use latest petsc-3.7 release. > > > > > > > > > > > > > > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > > > > > > > > > > > > > Hm - configure is picking up Intel-MPI from your 'INCLUDE' env > > > > > > > > > > > > > variable - but not 'make' - not sure whats hapenning - but I think its > > > > > > > > > > > > > best to start a clean build with petsc-3.7 [unless you need 3.6 for a > > > > > > > > > > > > > specific need] > > > > > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > hello > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > i configured petsc and i was trying to make then this error appeared > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Using mpiexec: mpiexec > > > > > > > > > > > > > > ========================================== > > > > > > > > > > > > > > Building PETSc using GNU Make with 4 build threads > > > > > > > > > > > > > > ========================================== > > > > > > > > > > > > > > make[2]: Entering directory '/home/petsc-3.6.4' > > > > > > > > > > > > > > /usr/bin/python ./config/gmakegen.py --petsc-arch=../petsc-3.6.4 > > > > > > > > > > > > > > Use "/usr/bin/make V=1" to see the verbose compile lines. > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/info/verboseinfo.o > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/info/ftn-auto/verboseinfof.o > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/info/ftn-custom/zverboseinfof.o > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/ftn-custom/zsys.o > > > > > > > > > > > > > > verboseinfo.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/info/verboseinfo.o] Error 2 > > > > > > > > > > > > > > make[2]: *** Waiting for unfinished jobs.... > > > > > > > > > > > > > > verboseinfof.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/info/ftn-auto/verboseinfof.o] Error 2 > > > > > > > > > > > > > > zverboseinfof.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/info/ftn-custom/zverboseinfof.o] Error 2 > > > > > > > > > > > > > > zsys.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/ftn-custom/zsys.o] Error 2 > > > > > > > > > > > > > > make[2]: Leaving directory '/home/petsc-3.6.4' > > > > > > > > > > > > > > make[1]: *** [/home/petsc-3.6.4/lib/petsc/conf/rules:107: gnumake] Error 2 > > > > > > > > > > > > > > make[1]: Leaving directory '/home/petsc-3.6.4' > > > > > > > > > > > > > > **************************ERROR************************************* > > > > > > > > > > > > > > Error during compile, check ../petsc-3.6.4/lib/petsc/conf/make.log > > > > > > > > > > > > > > Send it and ../petsc-3.6.4/lib/petsc/conf/configure.log to petsc-maint(a)mcs.anl.gov > > > > > > > > > > > > > > ******************************************************************** > > > > > > > > > > > > > > make: *** [makefile:30: all] Error 1 > > > > > > > > > > > > > > > > > > > > > > > > > > > > i have also attached the configure.log and make.log > > > > > > > > > > > > > > > > > > > > > > > > > > > > regards > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > > > From: Matthew Knepley <knepley(a)gmail.com> > > > > > > > > > > > > > > Sent: Monday, February 6, 2017 8:07:41 AM > > > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > > > Cc: petsc-dev > > > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Sun, Feb 5, 2017 at 11:26 PM, Mohammad S Dodaran <mdodar1(a)lsu.edu<mailto:[email protected]>> wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > Satish, > > > > > > > > > > > > > > > > > > > > > > > > > > > > Thanks for your quick response. as you see I can ls to the directory: > > > > > > > > > > > > > > > > > > > > > > > > > > > > ls /cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include/ > > > > > > > > > > > > > > > > > > > > > > > > > > > > Yes, but as Satish said, there are spaces in the path above. We do not allow spaces in the path. > > > > > > > > > > > > > > You cannot give a path to configure with spaces. > > > > > > > > > > > > > > > > > > > > > > > > > > > > You can remove the spaces using the cygpath command that Satish showed. Please use that command. > > > > > > > > > > > > > > Remove the spaces from this path and give the spaceless path to configure. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Thanks, > > > > > > > > > > > > > > > > > > > > > > > > > > > > Matt > > > > > > > > > > > > > > > > > > > > > > > > > > > > mpi.f90 mpi.h mpif.h mpio.h mspms.h pmidbg.h x64/ x86/ > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > but I do not know why still it has the same problem. and my configure file looks strange. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Best, > > > > > > > > > > > > > > > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov<mailto:[email protected]>> > > > > > > > > > > > > > > Sent: Sunday, February 5, 2017 10:58:35 PM > > > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > > > Cc: petsc-dev(a)mcs.anl.gov<mailto:[email protected]> > > > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > > > > > Try to figureout path without spaces. > > > > > > > > > > > > > > > > > > > > > > > > > > > > cygpath -u `cygpath -s -m PATH` > > > > > > > > > > > > > > > > > > > > > > > > > > > > Make sure you can do "ls" on this path. > > > > > > > > > > > > > > > > > > > > > > > > > > > > And use it with petsc configure. > > > > > > > > > > > > > > > > > > > > > > > > > > > > balay@ps4 ~ > > > > > > > > > > > > > > $ cygpath -u `cygpath -s -m /cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include` > > > > > > > > > > > > > > /cygdrive/c/PROGRA~2/MICROS~2/MPI/Include > > > > > > > > > > > > > > > > > > > > > > > > > > > > balay@ps4 ~ > > > > > > > > > > > > > > $ ls /cygdrive/c/PROGRA~2/MICROS~2/MPI/Include > > > > > > > > > > > > > > mpi.f90 mpi.h mpif.h mpio.h mspms.h pmidbg.h x64 x86 > > > > > > > > > > > > > > > > > > > > > > > > > > > > balay@ps4 ~ > > > > > > > > > > > > > > $ > > > > > > > > > > > > > > > > > > > > > > > > > > > > For ex: - I use: > > > > > > > > > > > > > > > > > > > > > > > > > > > > >>>>>>>> > > > > > > > > > > > > > > $ cat config/examples/arch-mswin-intel.py > > > > > > > > > > > > > > #!/usr/bin/python > > > > > > > > > > > > > > if __name__ == '__main__': > > > > > > > > > > > > > > import sys > > > > > > > > > > > > > > import os > > > > > > > > > > > > > > sys.path.insert(0, os.path.abspath('config')) > > > > > > > > > > > > > > import configure > > > > > > > > > > > > > > configure_options = [ > > > > > > > > > > > > > > '--download-fblaslapack=1', > > > > > > > > > > > > > > '--with-cc=win32fe icl', > > > > > > > > > > > > > > '--with-cxx=win32fe icl', > > > > > > > > > > > > > > '--with-fc=win32fe ifort', > > > > > > > > > > > > > > '--with-mpi-include=[/cygdrive/c/PROGRA~2/MICROS~2/MPI/Include/,/cygdrive/c/PROGRA~2/MICROS~2/MPI/Include/x64]', > > > > > > > > > > > > > > '--with-mpi-lib=[/cygdrive/c/PROGRA~2/MICROS~2/MPI/lib/x64/msmpifec.lib,/cygdrive/c/PROGRA~2/MICROS~2/MPI/lib/x64/msmpi.lib]', > > > > > > > > > > > > > > '--with-mpiexec=/cygdrive/c/PROGRA~1/MICROS~2/Bin/mpiexec', > > > > > > > > > > > > > > '--with-shared-libraries=0', > > > > > > > > > > > > > > 'DATAFILESPATH=c:/cygwin64/home/petsc/datafiles', > > > > > > > > > > > > > > ] > > > > > > > > > > > > > > configure.petsc_configure(configure_options) > > > > > > > > > > > > > > > > > > > > > > > > > > > > <<<<<<<<<<< > > > > > > > > > > > > > > > > > > > > > > > > > > > > if you still have issues - send us configure.log for the failure. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Sun, 5 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Dear Petsc developers, > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > I am trying to install Petsc on windows 10 but I have a problem. I installed MS-MPI v8<http://go.microsoft.com/FWLink/p/?LinkID=389556> separately and in the configure options I use > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > --with-mpi-include="/cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include" --with-mpi-lib="/cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Lib/x64/msmpi.lib" > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > But it gives error which says: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > *************************************************************** > > > > > > > > > > > > > > > ERROR in COMMAND LINE ARGUMENT to ./configure > > > > > > > > > > > > > > > ------------------------------------------------------------------------------- > > > > > > > > > > > > > > > Invalid directory: [/cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include for key with-mpi-include > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > How should I solve it? > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > > > > > > 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 > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >

1 0

Re: [petsc-dev] install petsc on windows 10 with mpi
by Satish Balay 07 Feb '17

07 Feb '17

>>> C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but now appears to be compiling using a non-MPICH mpi.h"^M <<<<< I have no clue why you get this error - since I don't see the flag PETSC_HAVE_MPICH_NUMVERSION set anymore. Its best if you attempt a clean install [reboot, run cygwin setup.exe, start with a fresh tarball] BTW: do you have a requirement for windows - or is it just a convinence? Most of the externalpackages are untested on windows - and https://damask.mpie.de/Installation/PETSc doesn't claim windows build is test. You might want to explore linux build [perhaps via VM] Satish On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > would you please check this log file. > > > the previous one belong to the make that i did on cygwin terminal > > > this time i tried to make on the command prompt of windows parallel studio by doing cygwin.bat and then i did make this time again the mpi error appeared > > > i will try to patch the file that you send once again and do make > > ________________________________ > From: Satish Balay <balay(a)mcs.anl.gov> > Sent: Tuesday, February 7, 2017 5:10:05 PM > To: Mohammad S Dodaran > Cc: petsc-dev > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > Yeah - thats was I was suggesting. Before that - can you try the following and see if it works? > > make all-legacy > > For a clean install - I would: > > 1. reboot > 2. run cygwin setup.exe to completion [which should run rebaseall] > 3. start with a fresh traball and redo the configure and make > > Satish > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > by fresh tar ball you mean that its better to download the petsc and start everything from the scratch? > > > > > > > > ________________________________ > > From: Satish Balay <balay(a)mcs.anl.gov> > > Sent: Tuesday, February 7, 2017 5:00:18 PM > > To: Mohammad S Dodaran > > Cc: petsc-dev > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > I don't see the realvent messages - so I have no idea whats going wrong here. > > > > You could try manually running the compile command - but I'm not sure > > if that would help: > > > > /home/petsc-3.6.4/bin/win32fe/win32fe cl -c -MT -wd4996 -O2 -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 /home/petsc-3.6.4/src/sys/info/verboseinfo.c -o first/obj/src/sys/info/verboseinfo.o > > > > My suggestion is to start with a fresh tarball - to eliminate werid issues. > > > > Satish > > > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > > > is there anything else that i have to do? > > > > > > > > > ________________________________ > > > From: Mohammad S Dodaran > > > Sent: Tuesday, February 7, 2017 4:31:24 PM > > > To: petsc-dev > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > i did > > > > > > make clean > > > make V=1 MAKE_NP=1 > > > > > > > > > and i attached the make.log > > > > > > > > > > > > when I try to make the error that appears on the screen is like this: > > > > > > ----------------------------------------- > > > Using C/C++ compile: /home/petsc-3.6.4/bin/win32fe/win32fe cl -c -MT -wd4996 -O2 -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 > > > Using Fortran compile: /home/petsc-3.6.4/bin/win32fe/win32fe ifort -c -MT -O3 -fpp -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 > > > ----------------------------------------- > > > Using C/C++ linker: /home/petsc-3.6.4/bin/win32fe/win32fe cl > > > Using C/C++ flags: -MT -wd4996 -O2 > > > Using Fortran linker: /home/petsc-3.6.4/bin/win32fe/win32fe ifort > > > Using Fortran flags: -MT -O3 -fpp > > > ----------------------------------------- > > > Using libraries: -L/home/petsc-3.6.4/first/lib -lpetsc -lflapack -lfblas /cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpi.lib /cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpifec.lib Gdi32.lib User32.lib Advapi32.lib Kernel32.lib Ws2_32.lib > > > ------------------------------------------ > > > Using mpiexec: mpiexec > > > ========================================== > > > Building PETSc using GNU Make with 1 build threads > > > ========================================== > > > make[2]: Entering directory '/home/petsc-3.6.4' > > > /usr/bin/python ./config/gmakegen.py --petsc-arch=first > > > /home/petsc-3.6.4/bin/win32fe/win32fe cl -c -MT -wd4996 -O2 -I/home/petsc-3.6.4/include -I/home/petsc-3.6.4/first/include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include -I/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64 /home/petsc-3.6.4/src/sys/info/verboseinfo.c -o first/obj/src/sys/info/verboseinfo.o > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/verboseinfo.o] Error 188 > > > make[2]: Leaving directory '/home/petsc-3.6.4' > > > make[1]: *** [/home/petsc-3.6.4/lib/petsc/conf/rules:107: gnumake] Error 2 > > > make[1]: Leaving directory '/home/petsc-3.6.4' > > > **************************ERROR************************************* > > > Error during compile, check first/lib/petsc/conf/make.log > > > Send it and first/lib/petsc/conf/configure.log to petsc-maint(a)mcs.anl.gov > > > ******************************************************************** > > > make: *** [makefile:30: all] Error 1 > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > > > > > > > > > ________________________________ > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > Sent: Tuesday, February 7, 2017 3:44:01 PM > > > To: Mohammad S Dodaran > > > Cc: petsc-dev > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > okay > > > > > > > > > > > > i patched the file you send > > > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > $ patch -Np1 < mpich-ver-test.patch > > > > (Stripping trailing CRs from patch; use --binary to disable.) > > > > patching file config/BuildSystem/config/packages/MPI.py > > > > > > wrt this patch - you would run configure again. But the error messages below are weird. > > > > > > Can you do: > > > > > > make clean > > > make V=1 MAKE_NP=1 > > > > > > and resend logs. > > > > > > Satish > > > > > > > > > > > > > > > the tried to make > > > > it failed > > > > > > > > Building PETSc using GNU Make with 4 build threads > > > > ========================================== > > > > make[2]: Entering directory '/home/petsc-3.6.4' > > > > Use "/usr/bin/make V=1" to see the verbose compile lines. > > > > CC first/obj/src/sys/info/verboseinfo.o > > > > CC first/obj/src/sys/info/ftn-custom/zverboseinfof.o > > > > CC first/obj/src/sys/info/ftn-auto/verboseinfof.o > > > > CC first/obj/src/sys/ftn-custom/zsys.o > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/ftn-custom/zverboseinfof.o] Error 188 > > > > make[2]: *** Waiting for unfinished jobs.... > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/verboseinfo.o] Error 188 > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/info/ftn-auto/verboseinfof.o] Err or 188 > > > > make[2]: *** [gmakefile:161: first/obj/src/sys/ftn-custom/zsys.o] Error 188 > > > > make[2]: Leaving directory '/home/petsc-3.6.4' > > > > make[1]: *** [/home/petsc-3.6.4/lib/petsc/conf/rules:107: gnumake] Error 2 > > > > make[1]: Leaving directory '/home/petsc-3.6.4' > > > > **************************ERROR************************************* > > > > Error during compile, check first/lib/petsc/conf/make.log > > > > Send it and first/lib/petsc/conf/configure.log to petsc-maint(a)mcs.anl.gov > > > > ******************************************************************** > > > > make: *** [makefile:30: all] Error 1 > > > > > > > > i have attached the make.log > > > > > > > > > > > > > > > > ________________________________ > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > Sent: Tuesday, February 7, 2017 3:25:52 PM > > > > To: Mohammad S Dodaran > > > > Cc: petsc-dev > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > nope. > > > > > > > > satish > > > > > > > > On Tue, 7 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > do you want me to remove the env and configure it once again and send you the log files? > > > > > > > > > > > > > > > ________________________________ > > > > > From: Mohammad S Dodaran > > > > > Sent: Tuesday, February 7, 2017 3:13:55 PM > > > > > To: petsc-dev > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > sorry for responding late. something came up and i couldn't start working on this earlier. > > > > > > > > > > env is still there, and i removed those three line > > > > > > > > > > then: tried to make > > > > > > > > > > i attached the configure and make log files > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > i tried to patch the file that you send but the patch command does not work in cygwin > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 ~ > > > > > $ cd /home/petsc-3.6.4 > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > > $ patch -Np1 < mpich-ver-test.patch > > > > > bash: patch: command not found > > > > > > > > > > MIE-HERA1+instadm@mie-hera1 /home/petsc-3.6.4 > > > > > $ > > > > > > > > > > > > > > > am i missing something? > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > Sent: Tuesday, February 7, 2017 12:09:38 AM > > > > > To: petsc-dev > > > > > Cc: Mohammad S Dodaran > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > Please try the attached patch - and see if it works [without removing IntelMPI from env] > > > > > > > > > > you can apply it with 'patch' command from cygwin > > > > > > > > > > cd /home/petsc-3.6.4 > > > > > patch -Np1 < mpich-ver-test.patch > > > > > > > > > > Satish > > > > > > > > > > On Mon, 6 Feb 2017, Satish Balay wrote: > > > > > > > > > > > Thinking about this - configure should not fail with MSMPI even if Intel-MPI is in the default path. > > > > > > > > > > > > A workarround is [after successful configure] - edit PETSC_ARCH/include/petscconf.h and remove the lines: > > > > > > > > > > > > #ifndef PETSC_HAVE_MPICH_NUMVERSION > > > > > > #define PETSC_HAVE_MPICH_NUMVERSION 30200300 > > > > > > #endif > > > > > > > > > > > > And then run 'make all' > > > > > > > > > > > > Will have to check on the fix.. > > > > > > > > > > > > Satish > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > I will remove intel mpi env tomorrow first thing in the morning and then i will send you the log files. > > > > > > > Mohammad > > > > > > > > > > > > > > Get Outlook for Android<https://aka.ms/ghei36> > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > Sent: Monday, February 6, 2017 11:34:00 PM > > > > > > > To: petsc-dev > > > > > > > Cc: Mohammad S Dodaran > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > BTW: Why not use petsc-3.7? > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Satish Balay wrote: > > > > > > > > > > > > > > > Please send the logs for the case that is supporsed to work - but failed. > > > > > > > > > > > > > > > > without the relavent logs - I have no idea what failed. > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > As i stated in my previous email. > > > > > > > > > " i tried to make with and without intel-mpi ENV . in both cases the error had been changed." And the error didnt have any thing to do with intel_mpi ENV. Do you still want me to remove the intel_ mpi env and send you the log file? > > > > > > > > > > > > > > > > > > Get Outlook for Android<https://aka.ms/ghei36> > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > Sent: Monday, February 6, 2017 11:15:19 PM > > > > > > > > > To: petsc-dev > > > > > > > > > Cc: Mohammad S Dodaran > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > never mind. > > > > > > > > > > > > > > > > > > >>>>>>>>>>>>>> > > > > > > > > > #line 2416 "c:\\program files (x86)\\intelswtools\\compilers_and_libraries_2017.1.143\\windows\\mpi\\intel64\\include\\mpi.h"^M > > > > > > > > > ^M > > > > > > > > > #line 2418 "c:\\program files (x86)\\intelswtools\\compilers_and_libraries_2017.1.143\\windows\\mpi\\intel64\\include\\mpi.h"^M > > > > > > > > > ^M > > > > > > > > > #line 4 "C:\\cygwin64\\tmp\\PE5639~1\\CONFIG~1.MPI\\conftest.c"^M > > > > > > > > > int mpich_ver = 30200300; > > > > > > > > > Defined "HAVE_MPICH_NUMVERSION" to "30200300" > > > > > > > > > Checking for functions [MPI_Alltoallw] in library ['/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpi.lib', '/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/ms\ > > > > > > > > > mpifec.lib'] [] > > > > > > > > > Pushing language C > > > > > > > > > > > > > > > > > > <<<<<<<<<< > > > > > > > > > > > > > > > > > > Its still picking up mpi.h from intel MPI > > > > > > > > > > > > > > > > > > >>>>>>>>>>>> > > > > > > > > > INCLUDE=C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\compiler\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\compiler\include\intel64;C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\INCLUDE;C:\Program Files (x86)\Microsoft Visual Studio 12.0\VC\ATLMFC\INCLUDE;C:\Program Files (x86)\Windows Kits\8.1\include\shared;C:\Program Files (x86)\Windows Kits\8.1\include\um;C:\Program Files (x86)\Windows Kits\8.1\include\winrt;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mpi\intel64\bin\..\..\intel64\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\ipp\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mkl\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\tbb\bin\..\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\daal\include;C:\Program Files (x86)\IntelSWTools\itac_2017\include;C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mpi\intel64\bin\..\..\intel64\include; > > > > > > > > > <<<<<<<<<<< > > > > > > > > > > > > > > > > > > The env has "C:\Program Files (x86)\IntelSWTools\compilers_and_libraries_2017.1.143\windows\mpi\intel64\bin\..\..\intel64\include" from where mpi.h is getting picked up. > > > > > > > > > > > > > > > > > > If you can't remove it perhaps you can try using it instead of MSMPI > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Satish Balay wrote: > > > > > > > > > > > > > > > > > > > can you also send make.log? > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > i was trying to repair my parallel_studio_xe_2017. then i configure it once again and then i tried to make with and without intel-mpi ENV . in both cases the error had been changed. i have attached the configure.log > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > > > Sent: Monday, February 6, 2017 10:10:43 PM > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > Cc: petsc-dev > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > Them Intel-MPI was not removed. You can resend configure.log to confirm.. > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > I removed the INTEL-MPI from the environmental variables and I still have the same problem. When I configure it looks like it accepts the mpi_include and mpi_lib but when I want to build , it says the configure was with mpi but in build it appears to compile is using a non-mpich. > > > > > > > > > > > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > > > > Sent: Monday, February 6, 2017 6:13:04 PM > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > Cc: petsc-dev > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > 1. configure.log is a cygwin simlink - so native mail app is unable to > > > > > > > > > > > > attach correctly. The actual file is in > > > > > > > > > > > > PETSC_DIR/PETSC_ARCH/lib/petsc/conf > > > > > > > > > > > > > > > > > > > > > > > > 2. Suggest doing the build without fftw and hdf5 packages [once the > > > > > > > > > > > > basic build succeeds - you can look at adding these pacakges.] > > > > > > > > > > > > > > > > > > > > > > > > 3. remove INTEL-MPI from your env - before running configure. [you > > > > > > > > > > > > might be able to use intel-mpi instead of ms-mpi - but we don't have > > > > > > > > > > > > any experience with it] > > > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > > > now I'm trying to configure petsc 3.7 > > > > > > > > > > > > > i used the format you showd in your previous email for the directories of fftw and hdf5 as below > > > > > > > > > > > > > > > > > > > > > > > > > > --with-fftw-dir=/cygdrive/c/cygwin64/home/fftw --with-hdf5-dir=/cygdrive/c/cygwin64/home/hdf5 > > > > > > > > > > > > > > > > > > > > > > > > > > the result is the following > > > > > > > > > > > > > > > > > > > > > > > > > > TESTING: alternateConfigureLibrary from config.packages.glut(config/BuildSystem/config/package.py:742) TESTING: alternateConfigureLibrary from config.packages.giflib(config/BuildSystem/config/package.py:742) TESTING: alternateConfigureLibrary from config.packages.scientificpython(config/BuildSystem/config/package.py:742) TESTING: alternateConfigureLibrary from config.packages.fiat(config/BuildSystem/config/package.py:742) TESTING: configureLibrary from config.packages.fftw(config/BuildSystem/config/package.py:679) ******************************************************************************* > > > > > > > > > > > > > UNABLE to CONFIGURE with GIVEN OPTIONS (see configure.log for details): > > > > > > > > > > > > > ------------------------------------------------------------------------------- > > > > > > > > > > > > > --with-fftw-dir=/cygdrive/c/cygwin64/home/fftw did not work > > > > > > > > > > > > > ******************************************************************************* > > > > > > > > > > > > > > > > > > > > > > > > > > i also attached the configure.log > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov> > > > > > > > > > > > > > Sent: Monday, February 6, 2017 4:28:09 PM > > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > > Cc: Matthew Knepley; petsc-dev > > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > > > >>>> > > > > > > > > > > > > > > > > > > > > > > > > > > Configure Options: --configModules=PETSc.Configure --optionsModule=config.compilerOptions --with-cc="win32fe cl" --with-fc="win32fe ifort" --with-cxx="win32fe cl" --with-mpi-include="[/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include,/cygdrive/c/PROGRA~2/MICROS~4/MPI/Include/x64]" --with-mpi-lib="[/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpi.lib,/cygdrive/c/PROGRA~2/MICROS~4/MPI/Lib/x64/msmpifec.lib]" --with-fftw-dir="[/cygdrive/c/cygwin64/home/fftw]" --with-hdf5-dir="[/cygdrive/c/cygwin64/home/hdf5]" --download-fblaslapack=1 --with-c2html=0 --with-debugging=0 --with-ssl=0 --with-x=0 PETSC_ARCH= PETSC_DIR=/home/petsc-3.6.4 > > > > > > > > > > > > > <<<< > > > > > > > > > > > > > > > > > > > > > > > > > > > 'PETSC_ARCH=' > > > > > > > > > > > > > > > > > > > > > > > > > > Can you rerun with a non-empty PETSC_ARCH string - or let configure choost default.. > > > > > > > > > > > > > > > > > > > > > > > > > > > --with-fftw-dir="[/cygdrive/c/cygwin64/home/fftw]" --with-hdf5-dir="[/cygdrive/c/cygwin64/home/hdf5]" > > > > > > > > > > > > > > > > > > > > > > > > > > I don't think I've built these packages with MS compilers > > > > > > > > > > > > > before. Somehow they are not picked up anyway. For one - the format > > > > > > > > > > > > > here should be the following [but I don't know if that will make > > > > > > > > > > > > > configure check for them]: > > > > > > > > > > > > > > > > > > > > > > > > > > --with-fftw-dir=/cygdrive/c/cygwin64/home/fftw --with-hdf5-dir=/cygdrive/c/cygwin64/home/hdf5 > > > > > > > > > > > > > > > > > > > > > > > > > > > PETSC_DIR=/home/petsc-3.6.4 > > > > > > > > > > > > > > > > > > > > > > > > > > Also - please use latest petsc-3.7 release. > > > > > > > > > > > > > > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > > > > > > > > > > > > > Hm - configure is picking up Intel-MPI from your 'INCLUDE' env > > > > > > > > > > > > > variable - but not 'make' - not sure whats hapenning - but I think its > > > > > > > > > > > > > best to start a clean build with petsc-3.7 [unless you need 3.6 for a > > > > > > > > > > > > > specific need] > > > > > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > > > On Mon, 6 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > hello > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > i configured petsc and i was trying to make then this error appeared > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Using mpiexec: mpiexec > > > > > > > > > > > > > > ========================================== > > > > > > > > > > > > > > Building PETSc using GNU Make with 4 build threads > > > > > > > > > > > > > > ========================================== > > > > > > > > > > > > > > make[2]: Entering directory '/home/petsc-3.6.4' > > > > > > > > > > > > > > /usr/bin/python ./config/gmakegen.py --petsc-arch=../petsc-3.6.4 > > > > > > > > > > > > > > Use "/usr/bin/make V=1" to see the verbose compile lines. > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/info/verboseinfo.o > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/info/ftn-auto/verboseinfof.o > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/info/ftn-custom/zverboseinfof.o > > > > > > > > > > > > > > CC ../petsc-3.6.4/obj/src/sys/ftn-custom/zsys.o > > > > > > > > > > > > > > verboseinfo.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/info/verboseinfo.o] Error 2 > > > > > > > > > > > > > > make[2]: *** Waiting for unfinished jobs.... > > > > > > > > > > > > > > verboseinfof.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/info/ftn-auto/verboseinfof.o] Error 2 > > > > > > > > > > > > > > zverboseinfof.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/info/ftn-custom/zverboseinfof.o] Error 2 > > > > > > > > > > > > > > zsys.c > > > > > > > > > > > > > > C:\cygwin64\home\PETSC-~1.4\include\petscsys.h(122) : fatal error C1189: #error : "PETSc was configured with MPICH but > > > > > > > > > > > > > > now appears to be compiling using a non-MPICH mpi.h" > > > > > > > > > > > > > > make[2]: *** [gmakefile:161: ../petsc-3.6.4/obj/src/sys/ftn-custom/zsys.o] Error 2 > > > > > > > > > > > > > > make[2]: Leaving directory '/home/petsc-3.6.4' > > > > > > > > > > > > > > make[1]: *** [/home/petsc-3.6.4/lib/petsc/conf/rules:107: gnumake] Error 2 > > > > > > > > > > > > > > make[1]: Leaving directory '/home/petsc-3.6.4' > > > > > > > > > > > > > > **************************ERROR************************************* > > > > > > > > > > > > > > Error during compile, check ../petsc-3.6.4/lib/petsc/conf/make.log > > > > > > > > > > > > > > Send it and ../petsc-3.6.4/lib/petsc/conf/configure.log to petsc-maint(a)mcs.anl.gov > > > > > > > > > > > > > > ******************************************************************** > > > > > > > > > > > > > > make: *** [makefile:30: all] Error 1 > > > > > > > > > > > > > > > > > > > > > > > > > > > > i have also attached the configure.log and make.log > > > > > > > > > > > > > > > > > > > > > > > > > > > > regards > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > > > From: Matthew Knepley <knepley(a)gmail.com> > > > > > > > > > > > > > > Sent: Monday, February 6, 2017 8:07:41 AM > > > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > > > Cc: petsc-dev > > > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Sun, Feb 5, 2017 at 11:26 PM, Mohammad S Dodaran <mdodar1(a)lsu.edu<mailto:[email protected]>> wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > Satish, > > > > > > > > > > > > > > > > > > > > > > > > > > > > Thanks for your quick response. as you see I can ls to the directory: > > > > > > > > > > > > > > > > > > > > > > > > > > > > ls /cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include/ > > > > > > > > > > > > > > > > > > > > > > > > > > > > Yes, but as Satish said, there are spaces in the path above. We do not allow spaces in the path. > > > > > > > > > > > > > > You cannot give a path to configure with spaces. > > > > > > > > > > > > > > > > > > > > > > > > > > > > You can remove the spaces using the cygpath command that Satish showed. Please use that command. > > > > > > > > > > > > > > Remove the spaces from this path and give the spaceless path to configure. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Thanks, > > > > > > > > > > > > > > > > > > > > > > > > > > > > Matt > > > > > > > > > > > > > > > > > > > > > > > > > > > > mpi.f90 mpi.h mpif.h mpio.h mspms.h pmidbg.h x64/ x86/ > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > but I do not know why still it has the same problem. and my configure file looks strange. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Best, > > > > > > > > > > > > > > > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > > > > > ________________________________ > > > > > > > > > > > > > > From: Satish Balay <balay(a)mcs.anl.gov<mailto:[email protected]>> > > > > > > > > > > > > > > Sent: Sunday, February 5, 2017 10:58:35 PM > > > > > > > > > > > > > > To: Mohammad S Dodaran > > > > > > > > > > > > > > Cc: petsc-dev(a)mcs.anl.gov<mailto:[email protected]> > > > > > > > > > > > > > > Subject: Re: [petsc-dev] install petsc on windows 10 with mpi > > > > > > > > > > > > > > > > > > > > > > > > > > > > Try to figureout path without spaces. > > > > > > > > > > > > > > > > > > > > > > > > > > > > cygpath -u `cygpath -s -m PATH` > > > > > > > > > > > > > > > > > > > > > > > > > > > > Make sure you can do "ls" on this path. > > > > > > > > > > > > > > > > > > > > > > > > > > > > And use it with petsc configure. > > > > > > > > > > > > > > > > > > > > > > > > > > > > balay@ps4 ~ > > > > > > > > > > > > > > $ cygpath -u `cygpath -s -m /cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include` > > > > > > > > > > > > > > /cygdrive/c/PROGRA~2/MICROS~2/MPI/Include > > > > > > > > > > > > > > > > > > > > > > > > > > > > balay@ps4 ~ > > > > > > > > > > > > > > $ ls /cygdrive/c/PROGRA~2/MICROS~2/MPI/Include > > > > > > > > > > > > > > mpi.f90 mpi.h mpif.h mpio.h mspms.h pmidbg.h x64 x86 > > > > > > > > > > > > > > > > > > > > > > > > > > > > balay@ps4 ~ > > > > > > > > > > > > > > $ > > > > > > > > > > > > > > > > > > > > > > > > > > > > For ex: - I use: > > > > > > > > > > > > > > > > > > > > > > > > > > > > >>>>>>>> > > > > > > > > > > > > > > $ cat config/examples/arch-mswin-intel.py > > > > > > > > > > > > > > #!/usr/bin/python > > > > > > > > > > > > > > if __name__ == '__main__': > > > > > > > > > > > > > > import sys > > > > > > > > > > > > > > import os > > > > > > > > > > > > > > sys.path.insert(0, os.path.abspath('config')) > > > > > > > > > > > > > > import configure > > > > > > > > > > > > > > configure_options = [ > > > > > > > > > > > > > > '--download-fblaslapack=1', > > > > > > > > > > > > > > '--with-cc=win32fe icl', > > > > > > > > > > > > > > '--with-cxx=win32fe icl', > > > > > > > > > > > > > > '--with-fc=win32fe ifort', > > > > > > > > > > > > > > '--with-mpi-include=[/cygdrive/c/PROGRA~2/MICROS~2/MPI/Include/,/cygdrive/c/PROGRA~2/MICROS~2/MPI/Include/x64]', > > > > > > > > > > > > > > '--with-mpi-lib=[/cygdrive/c/PROGRA~2/MICROS~2/MPI/lib/x64/msmpifec.lib,/cygdrive/c/PROGRA~2/MICROS~2/MPI/lib/x64/msmpi.lib]', > > > > > > > > > > > > > > '--with-mpiexec=/cygdrive/c/PROGRA~1/MICROS~2/Bin/mpiexec', > > > > > > > > > > > > > > '--with-shared-libraries=0', > > > > > > > > > > > > > > 'DATAFILESPATH=c:/cygwin64/home/petsc/datafiles', > > > > > > > > > > > > > > ] > > > > > > > > > > > > > > configure.petsc_configure(configure_options) > > > > > > > > > > > > > > > > > > > > > > > > > > > > <<<<<<<<<<< > > > > > > > > > > > > > > > > > > > > > > > > > > > > if you still have issues - send us configure.log for the failure. > > > > > > > > > > > > > > > > > > > > > > > > > > > > Satish > > > > > > > > > > > > > > > > > > > > > > > > > > > > On Sun, 5 Feb 2017, Mohammad S Dodaran wrote: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Dear Petsc developers, > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > I am trying to install Petsc on windows 10 but I have a problem. I installed MS-MPI v8<http://go.microsoft.com/FWLink/p/?LinkID=389556> separately and in the configure options I use > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > --with-mpi-include="/cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include" --with-mpi-lib="/cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Lib/x64/msmpi.lib" > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > But it gives error which says: > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > *************************************************************** > > > > > > > > > > > > > > > ERROR in COMMAND LINE ARGUMENT to ./configure > > > > > > > > > > > > > > > ------------------------------------------------------------------------------- > > > > > > > > > > > > > > > Invalid directory: [/cygdrive/c/Program\ Files\ $x86$/Microsoft\ SDKs/MPI/Include for key with-mpi-include > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > How should I solve it? > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > Mohammad > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > -- > > > > > > > > > > > > > > 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 > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > > >

1 0