March 2020 - petsc-dev - mailman3.cels.anl.gov

Re: [petsc-dev] Very Useful Description of MAT Types in Petsc Index
by Jacob Faibussowitsch 26 Mar '20

26 Mar '20

> Actually, the man pages for specific constructors actually yields more useful information (in nearly all cases) about the formats. > Take a look at these pages > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateS… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateS…> > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM…> > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM…> > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM…> > > Is that more informative? Yes this is pretty much what I had in mind! But I think it needs to moved away from the create calls and to the all upper-case impls sections. Currently the work-flow of the mat examples I have seen are as follows: MatCreate(); MatSetSizes(); MatSetType(); Which means new users will likely never look at the individual MatCreateXXX() routines and never see any of the useful information. They are far more likely to click on the links on the MatSetType calls to the various implementations, and be frustrated when there’s nothing useful there... Also as Junchao noted: > Even better, with examples showing a small matrix. I think this should be on every informative MAT page (wherever that ends up being). For example for the “base” MatCreateAIJ you have nice diagrams showing in broad strokes how things are internally structured. Say a user eventually implements MATMPIAIJMKL, given the above it isn’t guaranteed that they see the MatCreateAIJ which actually has the useful diagram, leading to headaches trying to figure out how to properly feed petsc the matrix. It’s certainly a bit of bloat, but every MAT impls should have a section with the diagram about its “base” implementation. Maybe we can be smart about it and “link” the text from another part of the src, but I don’t know much about how much you can do with the current documentation generation. Best regards, Jacob Faibussowitsch (Jacob Fai - booss - oh - vitch) Cell: (312) 694-3391 > On Mar 26, 2020, at 3:00 PM, Dave May <dave.mayhem23(a)gmail.com> wrote: > > > > On Thu, 26 Mar 2020 at 19:26, Jacob Faibussowitsch <jacob.fai(a)gmail.com <mailto:[email protected]>> wrote: > Hello, > > In keeping with PETSc design it would be nice to have *more* detail for all MAT implementations explaining in what the letters stand for, and use-cases it might be useful for. For example: > >> MATAIJSELL <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATAIJSELL.htm…> = "AIJSELL" - A matrix type to be used for sparse matrices. This matrix type is identical to MATSEQAIJSELL when constructed with a single process communicator, and MATMPIAIJSELL otherwise. As a result, for single process communicators, MatSeqAIJSetPreallocation <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSeqAIJSetPr…>() is supported, and similarly MatMPIAIJSetPreallocation <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatMPIAIJSetPr…>() is supported for communicators controlling multiple processes. It is recommended that you call both of the above preallocation routines for simplicity. > > > Agreed that is not particularly informative. > > I am no complete beginner but I am no computational matrix expert either. I have no idea what “SELL” is. Obviously googling "mat sell format” gives less than useful results. Other such types are MATAIJCRL, MATAIJMKL, MATAIJPERM, MATADJ. > > Actually, the man pages for specific constructors actually yields more useful information (in nearly all cases) about the formats. > Take a look at these pages > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateS… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateS…> > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM…> > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM…> > > https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… <https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM…> > > Is that more informative? > > Admittedly the SELL man page should include references to the papers which introduced the sliced ELLPACK stuff. > Possibly there are other matrix man pages which could also benefit from references. > Similarly the source code for these matrix types should ideally also use the function PetscCitationsRegister() to log / report these papers. > > > > Smallest change could be as simple as mirroring MATSBAIJ, which is clear and concise: > >> MATMPISBAIJ <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATMPISBAIJ.ht…> = "mpisbaij" - A matrix type to be used for distributed symmetric sparse block matrices, based on block compressed sparse row format. Only the upper triangular portion of the "diagonal" portion of the matrix is stored. For complex numbers by default this matrix is symmetric, NOT Hermitian symmetric. To make it Hermitian symmetric you can call MatSetOption <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSetOption.h…>(Mat <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/Mat.html#Mat>, MAT_HERMITIAN <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatOption.html…>); > > Clearest case would be: > >> MATMPISBAIJ = "mpisbaij" - MATrix MPI Symmetric Block AIJ. A matrix type to be used for distributed symmetric sparse block matrices, based on block compressed sparse row format. >> Only the upper triangular portion of the "diagonal" portion of the matrix is stored. For complex numbers by default this matrix is symmetric, NOT Hermitian symmetric. To make it Hermitian symmetric you can call MatSetOption(Mat, MAT_HERMITIAN); > > > Each format is useful in its own case, but people can’t use things if they don’t know what it is! > > Best regards, > > Jacob Faibussowitsch > (Jacob Fai - booss - oh - vitch) > Cell: (312) 694-3391 >

1 0

Re: [petsc-dev] Very Useful Description of MAT Types in Petsc Index
by Dave May 26 Mar '20

26 Mar '20

On Thu, 26 Mar 2020 at 19:26, Jacob Faibussowitsch <jacob.fai(a)gmail.com> wrote: > Hello, > > In keeping with PETSc design it would be nice to have *more* detail for > all MAT implementations explaining in what the letters stand for, and > use-cases it might be useful for. For example: > > MATAIJSELL > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATAIJSELL.htm…> > = "AIJSELL" - A matrix type to be used for sparse matrices. This matrix > type is identical to MATSEQAIJSELL when constructed with a single process > communicator, and MATMPIAIJSELL otherwise. As a result, for single process > communicators, MatSeqAIJSetPreallocation > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSeqAIJSetPr…>() > is supported, and similarly MatMPIAIJSetPreallocation > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatMPIAIJSetPr…>() > is supported for communicators controlling multiple processes. It is > recommended that you call both of the above preallocation routines for > simplicity. > > > Agreed that is not particularly informative. > I am no complete beginner but I am no computational matrix expert either. > I have no idea what “SELL” is. Obviously googling "mat sell format” gives > less than useful results. Other such types are MATAIJCRL, MATAIJMKL, > MATAIJPERM, MATADJ. > Actually, the man pages for specific constructors actually yields more useful information (in nearly all cases) about the formats. Take a look at these pages https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateS… https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… https://www.mcs.anl.gov/petsc/petsc-current/docs/manualpages/Mat/MatCreateM… Is that more informative? Admittedly the SELL man page should include references to the papers which introduced the sliced ELLPACK stuff. Possibly there are other matrix man pages which could also benefit from references. Similarly the source code for these matrix types should ideally also use the function PetscCitationsRegister() to log / report these papers. > > Smallest change could be as simple as mirroring MATSBAIJ, which is clear > and concise: > > MATMPISBAIJ > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATMPISBAIJ.ht…> > = "mpisbaij" - A matrix type to be used for distributed symmetric sparse > block matrices, based on block compressed sparse row format. Only the upper > triangular portion of the "diagonal" portion of the matrix is stored. For > complex numbers by default this matrix is symmetric, NOT Hermitian > symmetric. To make it Hermitian symmetric you can call MatSetOption > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSetOption.h…> > (Mat > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/Mat.html#Mat>, > MAT_HERMITIAN > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatOption.html…> > ); > > > Clearest case would be: > > MATMPISBAIJ = "mpisbaij" - *MATrix MPI Symmetric Block AIJ.* A matrix > type to be used for distributed symmetric sparse block matrices, based on > block compressed sparse row format. > > Only the upper triangular portion of the "diagonal" portion of the matrix > is stored. For complex numbers by default this matrix is symmetric, NOT > Hermitian symmetric. To make it Hermitian symmetric you > can call MatSetOption(Mat, MAT_HERMITIAN); > > > Each format is useful in its own case, but people can’t use things if they > don’t know what it is! > > Best regards, > > Jacob Faibussowitsch > (Jacob Fai - booss - oh - vitch) > Cell: (312) 694-3391 > >

1 0

Re: [petsc-dev] Very Useful Description of MAT Types in Petsc Index
by Junchao Zhang 26 Mar '20

26 Mar '20

Even better, with examples showing a small matrix. --Junchao Zhang On Thu, Mar 26, 2020 at 2:26 PM Jacob Faibussowitsch <jacob.fai(a)gmail.com> wrote: > Hello, > > In keeping with PETSc design it would be nice to have *more* detail for > all MAT implementations explaining in what the letters stand for, and > use-cases it might be useful for. For example: > > MATAIJSELL > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATAIJSELL.htm…> > = "AIJSELL" - A matrix type to be used for sparse matrices. This matrix > type is identical to MATSEQAIJSELL when constructed with a single process > communicator, and MATMPIAIJSELL otherwise. As a result, for single process > communicators, MatSeqAIJSetPreallocation > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSeqAIJSetPr…>() > is supported, and similarly MatMPIAIJSetPreallocation > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatMPIAIJSetPr…>() > is supported for communicators controlling multiple processes. It is > recommended that you call both of the above preallocation routines for > simplicity. > > > I am no complete beginner but I am no computational matrix expert either. > I have no idea what “SELL” is. Obviously googling "mat sell format” gives > less than useful results. Other such types are MATAIJCRL, MATAIJMKL, > MATAIJPERM, MATADJ. > > Smallest change could be as simple as mirroring MATSBAIJ, which is clear > and concise: > > MATMPISBAIJ > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATMPISBAIJ.ht…> > = "mpisbaij" - A matrix type to be used for distributed symmetric sparse > block matrices, based on block compressed sparse row format. Only the upper > triangular portion of the "diagonal" portion of the matrix is stored. For > complex numbers by default this matrix is symmetric, NOT Hermitian > symmetric. To make it Hermitian symmetric you can call MatSetOption > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSetOption.h…> > (Mat > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/Mat.html#Mat>, > MAT_HERMITIAN > <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatOption.html…> > ); > > > Clearest case would be: > > MATMPISBAIJ = "mpisbaij" - *MATrix MPI Symmetric Block AIJ.* A matrix > type to be used for distributed symmetric sparse block matrices, based on > block compressed sparse row format. > > Only the upper triangular portion of the "diagonal" portion of the matrix > is stored. For complex numbers by default this matrix is symmetric, NOT > Hermitian symmetric. To make it Hermitian symmetric you > can call MatSetOption(Mat, MAT_HERMITIAN); > > > Each format is useful in its own case, but people can’t use things if they > don’t know what it is! > > Best regards, > > Jacob Faibussowitsch > (Jacob Fai - booss - oh - vitch) > Cell: (312) 694-3391 > >

1 0

Very Useful Description of MAT Types in Petsc Index
by Jacob Faibussowitsch 26 Mar '20

26 Mar '20

Hello, In keeping with PETSc design it would be nice to have *more* detail for all MAT implementations explaining in what the letters stand for, and use-cases it might be useful for. For example: > MATAIJSELL <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATAIJSELL.htm…> = "AIJSELL" - A matrix type to be used for sparse matrices. This matrix type is identical to MATSEQAIJSELL when constructed with a single process communicator, and MATMPIAIJSELL otherwise. As a result, for single process communicators, MatSeqAIJSetPreallocation <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSeqAIJSetPr…>() is supported, and similarly MatMPIAIJSetPreallocation <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatMPIAIJSetPr…>() is supported for communicators controlling multiple processes. It is recommended that you call both of the above preallocation routines for simplicity. I am no complete beginner but I am no computational matrix expert either. I have no idea what “SELL” is. Obviously googling "mat sell format” gives less than useful results. Other such types are MATAIJCRL, MATAIJMKL, MATAIJPERM, MATADJ. Smallest change could be as simple as mirroring MATSBAIJ, which is clear and concise: > MATMPISBAIJ <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MATMPISBAIJ.ht…> = "mpisbaij" - A matrix type to be used for distributed symmetric sparse block matrices, based on block compressed sparse row format. Only the upper triangular portion of the "diagonal" portion of the matrix is stored. For complex numbers by default this matrix is symmetric, NOT Hermitian symmetric. To make it Hermitian symmetric you can call MatSetOption <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatSetOption.h…>(Mat <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/Mat.html#Mat>, MAT_HERMITIAN <https://www.mcs.anl.gov/petsc/petsc-dev/docs/manualpages/Mat/MatOption.html…>); Clearest case would be: > MATMPISBAIJ = "mpisbaij" - MATrix MPI Symmetric Block AIJ. A matrix type to be used for distributed symmetric sparse block matrices, based on block compressed sparse row format. > Only the upper triangular portion of the "diagonal" portion of the matrix is stored. For complex numbers by default this matrix is symmetric, NOT Hermitian symmetric. To make it Hermitian symmetric you can call MatSetOption(Mat, MAT_HERMITIAN); Each format is useful in its own case, but people can’t use things if they don’t know what it is! Best regards, Jacob Faibussowitsch (Jacob Fai - booss - oh - vitch) Cell: (312) 694-3391

1 0

Re: [petsc-dev] petsc release plan for march/2020
by Jed Brown 25 Mar '20

25 Mar '20

Matthew Knepley <knepley(a)gmail.com> writes: > On Wed, Mar 25, 2020 at 9:05 PM Jed Brown <jed(a)jedbrown.org> wrote: > >> Matthew Knepley <knepley(a)gmail.com> writes: >> >> > On Wed, Mar 25, 2020 at 8:23 PM Satish Balay via petsc-dev < >> > petsc-dev(a)mcs.anl.gov> wrote: >> > >> >> > > - feature freeze: March 27 say 5PM EST >> >> >> >> A reminder, feature freeze is in 2 days. I see the following MRs with >> >> milestone: v3.13-release >> >> >> >> !2598 Add Sphinx docs and port Developers Manual >> >> !2614 Knepley/feature dm remove hybrid >> >> !2626 More support for not needing to set PETSC_DIR yet build >> with >> >> PETSc, also... >> >> !2634 WIP: Balay/release 3.13 revert to dev >> >> !2554 Knepley/feature fe lagrange gll >> >> >> > >> > 2554 is not merging. I will change the target. >> >> If Toby's branch also isn't merging, is there a stop-gap we can have? Or >> is Toby's branch merging and the milestone is missing? >> > > This branch will not work because it requires fixes for DMForest which it > will not get in time. Toby said he was very close on his branch. Should its milestone be set? https://gitlab.com/petsc/petsc/-/merge_requests/2399

1 0

Re: [petsc-dev] petsc release plan for march/2020
by Matthew Knepley 25 Mar '20

25 Mar '20

On Wed, Mar 25, 2020 at 9:05 PM Jed Brown <jed(a)jedbrown.org> wrote: > Matthew Knepley <knepley(a)gmail.com> writes: > > > On Wed, Mar 25, 2020 at 8:23 PM Satish Balay via petsc-dev < > > petsc-dev(a)mcs.anl.gov> wrote: > > > >> > > - feature freeze: March 27 say 5PM EST > >> > >> A reminder, feature freeze is in 2 days. I see the following MRs with > >> milestone: v3.13-release > >> > >> !2598 Add Sphinx docs and port Developers Manual > >> !2614 Knepley/feature dm remove hybrid > >> !2626 More support for not needing to set PETSC_DIR yet build > with > >> PETSc, also... > >> !2634 WIP: Balay/release 3.13 revert to dev > >> !2554 Knepley/feature fe lagrange gll > >> > > > > 2554 is not merging. I will change the target. > > If Toby's branch also isn't merging, is there a stop-gap we can have? Or > is Toby's branch merging and the milestone is missing? > This branch will not work because it requires fixes for DMForest which it will not get in time. Toby said he was very close on his branch. Thanks, Matt -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener https://www.cse.buffalo.edu/~knepley/ <http://www.cse.buffalo.edu/~knepley/>

1 0

Re: [petsc-dev] petsc release plan for march/2020
by Jed Brown 25 Mar '20

25 Mar '20

Matthew Knepley <knepley(a)gmail.com> writes: > On Wed, Mar 25, 2020 at 8:23 PM Satish Balay via petsc-dev < > petsc-dev(a)mcs.anl.gov> wrote: > >> > > - feature freeze: March 27 say 5PM EST >> >> A reminder, feature freeze is in 2 days. I see the following MRs with >> milestone: v3.13-release >> >> !2598 Add Sphinx docs and port Developers Manual >> !2614 Knepley/feature dm remove hybrid >> !2626 More support for not needing to set PETSC_DIR yet build with >> PETSc, also... >> !2634 WIP: Balay/release 3.13 revert to dev >> !2554 Knepley/feature fe lagrange gll >> > > 2554 is not merging. I will change the target. If Toby's branch also isn't merging, is there a stop-gap we can have? Or is Toby's branch merging and the milestone is missing?

1 0

Re: [petsc-dev] petsc release plan for march/2020
by Matthew Knepley 25 Mar '20

25 Mar '20

On Wed, Mar 25, 2020 at 8:23 PM Satish Balay via petsc-dev < petsc-dev(a)mcs.anl.gov> wrote: > > > - feature freeze: March 27 say 5PM EST > > A reminder, feature freeze is in 2 days. I see the following MRs with > milestone: v3.13-release > > !2598 Add Sphinx docs and port Developers Manual > !2614 Knepley/feature dm remove hybrid > !2626 More support for not needing to set PETSC_DIR yet build with > PETSc, also... > !2634 WIP: Balay/release 3.13 revert to dev > !2554 Knepley/feature fe lagrange gll > 2554 is not merging. I will change the target. Thanks, Matt > !2367 WIP: Add sample file that demonstrates use of CMake and > pkg-config to build a PETSc application > > I'm hoping all v3.13-release MRs will complete testing and reviews > (approvals) before the freeze. > > Thanks, > Satish > > On Fri, 20 Mar 2020, Satish Balay via petsc-dev wrote: > > > An update: > > > > > https://gitlab.com/petsc/petsc/-/merge_requests?milestone_title=v3.13-relea… > > > > We have a week to go. And have about 47 MRs, with 12 marked as > 'v3.13-release' milestone. > > > > Likely the milestone is not correctly set for some MRs - if so - please > update the MR. > > > > And I see some MRs with significant code changes. Hopefully we can > process them early [and not have large changes close to freeze date]. > > > > Satish > > > > On Fri, 13 Mar 2020, Satish Balay via petsc-dev wrote: > > > > > All, > > > > > > We are to make a petsc release by the end of March. > > > > > > For this release [3.13], will work with the following dates: > > > > > > - feature freeze: March 27 say 5PM EST > > > - release: March 29 > > > > > > Merges after freeze should contain only fixes that would normally be > acceptable to maint workflow. > > > > > > I've changed the current milestone 'v3.13' to 'master' and created a > new one 'v3.13-release' > > > > > > If you are working on a MR with the goal of merging before release - > its best to use 'v3.13-release' tag with the MR. > > > > > > And it would be good to avoid merging large changes at the last > minute. And not have merge requests stuck in need of reviews, testing and > other necessary tasks. > > > > > > And I would think the testing/CI resources would get stressed in this > timeframe - so it would be good to use them judiciously if possible. > > > > > > - if there are failures in stage-2 or 3 - and its no longer necessary > to complete all the jobs - one can 'cancel' the pipeline. > > > - if a fix needs to be tested - one can first test with only the > failed jobs (if this is known) - before doing a full test pipeline. i.e: > > > - start pipeline > > > - immediately cancel the pipeline > > > - now toggle only the jobs that need to be run > > > - [on success of the selected jobs] if one wants to run the full > pipeleine - click 'retry' - and the remaining canceled jobs should now get > scheduled. > > > > > > thanks, > > > Satish > > > > > > > -- What most experimenters take for granted before they begin their experiments is infinitely more interesting than any results to which their experiments lead. -- Norbert Wiener https://www.cse.buffalo.edu/~knepley/ <http://www.cse.buffalo.edu/~knepley/>

1 0

Re: [petsc-dev] petsc release plan for march/2020
by Satish Balay 25 Mar '20

25 Mar '20

> > - feature freeze: March 27 say 5PM EST A reminder, feature freeze is in 2 days. I see the following MRs with milestone: v3.13-release !2598 Add Sphinx docs and port Developers Manual !2614 Knepley/feature dm remove hybrid !2626 More support for not needing to set PETSC_DIR yet build with PETSc, also... !2634 WIP: Balay/release 3.13 revert to dev !2554 Knepley/feature fe lagrange gll !2367 WIP: Add sample file that demonstrates use of CMake and pkg-config to build a PETSc application I'm hoping all v3.13-release MRs will complete testing and reviews (approvals) before the freeze. Thanks, Satish On Fri, 20 Mar 2020, Satish Balay via petsc-dev wrote: > An update: > > https://gitlab.com/petsc/petsc/-/merge_requests?milestone_title=v3.13-relea… > > We have a week to go. And have about 47 MRs, with 12 marked as 'v3.13-release' milestone. > > Likely the milestone is not correctly set for some MRs - if so - please update the MR. > > And I see some MRs with significant code changes. Hopefully we can process them early [and not have large changes close to freeze date]. > > Satish > > On Fri, 13 Mar 2020, Satish Balay via petsc-dev wrote: > > > All, > > > > We are to make a petsc release by the end of March. > > > > For this release [3.13], will work with the following dates: > > > > - feature freeze: March 27 say 5PM EST > > - release: March 29 > > > > Merges after freeze should contain only fixes that would normally be acceptable to maint workflow. > > > > I've changed the current milestone 'v3.13' to 'master' and created a new one 'v3.13-release' > > > > If you are working on a MR with the goal of merging before release - its best to use 'v3.13-release' tag with the MR. > > > > And it would be good to avoid merging large changes at the last minute. And not have merge requests stuck in need of reviews, testing and other necessary tasks. > > > > And I would think the testing/CI resources would get stressed in this timeframe - so it would be good to use them judiciously if possible. > > > > - if there are failures in stage-2 or 3 - and its no longer necessary to complete all the jobs - one can 'cancel' the pipeline. > > - if a fix needs to be tested - one can first test with only the failed jobs (if this is known) - before doing a full test pipeline. i.e: > > - start pipeline > > - immediately cancel the pipeline > > - now toggle only the jobs that need to be run > > - [on success of the selected jobs] if one wants to run the full pipeleine - click 'retry' - and the remaining canceled jobs should now get scheduled. > > > > thanks, > > Satish > > >

1 0

Re: [petsc-dev] Globsearch fails for me when running tests
by Jed Brown 25 Mar '20

25 Mar '20

The other strategy would be globsearch calls Python calls make to get the list on stdout, which it filters and returns on its stdout. An alternative would be to use $(guile ...) to do all the processing, but that will only work when Make is built with Guile support. Check make -f gmakefile print VAR=.FEATURES | grep guile to see if you have support. Mine has it, but Ubuntu's make package doesn't, for example. Scott Kruger <kruger(a)txcorp.com> writes: > Ugh -- this is ugly. > > Can't we just tell users to either use the '%' syntax or recompile their > linux kernel? > > Just kidding. I'll take a look. > > Scott > > > On 3/25/20 3:48 PM, Jed Brown wrote: >> Scott, you can't pass '$(alltesttargets)' on the command line like this. >> >> TESTTARGETS := $(shell $(PYTHON) -c"import sys,fnmatch,itertools; m=[fnmatch.filter(sys.argv[2].split(),p) for p in sys.argv[1].split()]; print(' '.join(list(itertools.chain.from_iterable(m))))" '$(globsearch)' '$(alltesttargets)') >> >> For this feature, probably put them in an argsfile >> >> $(file >$(TESTDIR)/globsearch.args,$(alltesttargets)) >> >> and make your Python read from that file. I don't know a way to pass it >> on stdin. >> >> Scott Kruger <kruger(a)txcorp.com> writes: >> >>> What platform? >>> >>> On 3/25/20 3:20 PM, Stefano Zampini wrote: >>>> This was working before.. >>>> >>>> [szampini@localhost petsc]$ make -f gmakefile.test test globsearch='dm*' >>>> make: execvp: /usr/bin/sh: Argument list too long >>>> Using MAKEFLAGS: -- globsearch=dm* >>>> # No tests run >>>> # No tests run >>>> # No tests run >>>> >>>> [szampini@localhost petsc]$ git branch >>>> * knepley/feature-dm-remove-hybrid >>>> >>>> [szampini@localhost petsc]$ make -v >>>> GNU Make 4.2.1 >>>> Built for x86_64-redhat-linux-gnu >>>> Copyright (C) 1988-2016 Free Software Foundation, Inc. >>>> License GPLv3+: GNU GPL version 3 or later >>>> <http://gnu.org/licenses/gpl.html> >>>> This is free software: you are free to change and redistribute it. >>>> There is NO WARRANTY, to the extent permitted by law. >>>> >>> >>> -- >>> Tech-X Corporation kruger(a)txcorp.com >>> 5621 Arapahoe Ave, Suite A Phone: (720) 974-1841 >>> Boulder, CO 80303 Fax: (303) 448-7756 > > -- > 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