January 2026 - petsc-dev - mailman3.cels.anl.gov

PETSc related MCP servers/agents
by Barry Smith 30 Jan '26

30 Jan '26

I have (with lots of Satish’s and Hong’s help) stood up three PETSc related MCP servers, see https://urldefense.us/v3/__https://gitlab.com/petsc/petsc_mcp_servers/-/blo… for details on their usage. I hope anyone using AI/LLM tools in their use of PETSc can try out these servers and see if they help (or hinder) the use of AI development systems. If you know of any available MCP servers that might help with PETSc please let us know. Any problems using these or suggestions on making them better let us know. Barry Note that MCP servers/agents serve multiple purposes - are used directly by developer systems like gemini-cli - can be used directly by other agents we wrote - can be used by modern RESTful interfaces for LLM based on the OpenAI response model MCP servers can be used in multiple ways - “directly” by an LLM, that is a LLM is given some text and a list of MCP tools and it decides exactly what tool to call with what arguments and - via the MCP Python API; in this case it is just like programming Python and calling functions (that are transparently) actually implemented somewhere else.

1 0

Re: [petsc-dev] Bug in make checkbadSource python code
by Barry Smith 25 Jan '26

25 Jan '26

It is not a “bad file” it is just some junk file left by a bad editor. Definitely the script should be ignoring such files. The script should probably be using git to get the list of files; it has no business checking files that are not in the repository. > On Jan 24, 2026, at 11:10 PM, Satish Balay <balay.anl(a)fastmail.org> wrote: > > In some sense - you had a bad file that was getting processed - and > the script gave you an error. - And from the stack trace - you could > determine and delete the bad file - so its a good thing? > > But I guess the functionality issue here is - this script should > ignore ".#*" (or ".*") files? [A bit hacky patch here]: > > Satish > --------- > balay@p1 /home/balay/petsc (main =) > $ touch include/.#petscsystypes.h > balay@p1 /home/balay/petsc (main =) > $ chmod a-rwx include/.#petscsystypes.h > balay@p1 /home/balay/petsc (main =) > $ ./lib/petsc/bin/maint/check_header_guard.py --action=check --kind=pragma_once -- ./src ./include > Traceback (most recent call last): > File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 508, in <module> > ret = command_line_main() > File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 495, in command_line_main > ret = main(args) > File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 436, in main > errors.extend(replace_in_file(dirpath / f, args, replacer_cls)) > ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 369, in replace_in_file > lines = path.read_text().splitlines() > ~~~~~~~~~~~~~~^^ > File "/usr/lib64/python3.14/pathlib/__init__.py", line 787, in read_text > with self.open(mode='r', encoding=encoding, errors=errors, newline=newline) as f: > ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > File "/usr/lib64/python3.14/pathlib/__init__.py", line 771, in open > return io.open(self, mode, buffering, encoding, errors, newline) > ~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ > PermissionError: [Errno 13] Permission denied: '/home/balay/petsc/include/.#petscsystypes.h' > balay@p1 /home/balay/petsc (main =) > $ cat fix.diff > diff --git a/lib/petsc/bin/maint/check_header_guard.py b/lib/petsc/bin/maint/check_header_guard.py > index f246fd62890..b56528e8bc3 100755 > --- a/lib/petsc/bin/maint/check_header_guard.py > +++ b/lib/petsc/bin/maint/check_header_guard.py > @@ -432,7 +432,7 @@ def main(args: argparse.Namespace) -> int: > dirs[:] = [d for d in dirs if d not in exclude_dirs] > dirpath = pathlib.Path(dirname) > for f in files: > - if f not in exclude_files: > + if f not in exclude_files and not f.startswith('.#'): > errors.extend(replace_in_file(dirpath / f, args, replacer_cls)) > > if errors: > balay@p1 /home/balay/petsc (main =) > $ patch -Np1 < fix.diff > patching file lib/petsc/bin/maint/check_header_guard.py > balay@p1 /home/balay/petsc (main *=) > $ ./lib/petsc/bin/maint/check_header_guard.py --action=check --kind=pragma_once -- ./src ./include > balay@p1 /home/balay/petsc (main *=) > $ echo $? > 0 > > > On Sat, 24 Jan 2026, Barry Smith wrote: > >> >> $ make checkbadSource >> Traceback (most recent call last): >> File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 508, in <module> >> ret = command_line_main() >> File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 495, in command_line_main >> ret = main(args) >> File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 436, in main >> errors.extend(replace_in_file(dirpath / f, args, replacer_cls)) >> File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 369, in replace_in_file >> lines = path.read_text().splitlines() >> File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1256, in read_text >> with self.open(mode='r', encoding=encoding, errors=errors) as f: >> File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1242, in open >> return io.open(self, mode, buffering, encoding, errors, newline, >> File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1110, in _opener >> return self._accessor.open(self, flags, mode) >> FileNotFoundError: [Errno 2] No such file or directory: '/Users/barrysmith/Src/petsc/include/.#petscsystypes.h' >> make[1]: [checkbadSource] Error 1 (ignored) >> ^Cmake: *** [checkbadSource] Interrupt: 2 >> >> ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions >> $ rm /Users/barrysmith/Src/petsc/include/.#petscsystypes.h >> ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions >> $ checkbadSource >> -bash: checkbadSource: command not found >> ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions >> $ make checkbadSource >> ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) >> >> >> >

1 0

Re: [petsc-dev] Bug in make checkbadSource python code
by Satish Balay 24 Jan '26

24 Jan '26

In some sense - you had a bad file that was getting processed - and the script gave you an error. - And from the stack trace - you could determine and delete the bad file - so its a good thing? But I guess the functionality issue here is - this script should ignore ".#*" (or ".*") files? [A bit hacky patch here]: Satish --------- balay@p1 /home/balay/petsc (main =) $ touch include/.#petscsystypes.h balay@p1 /home/balay/petsc (main =) $ chmod a-rwx include/.#petscsystypes.h balay@p1 /home/balay/petsc (main =) $ ./lib/petsc/bin/maint/check_header_guard.py --action=check --kind=pragma_once -- ./src ./include Traceback (most recent call last): File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 508, in <module> ret = command_line_main() File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 495, in command_line_main ret = main(args) File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 436, in main errors.extend(replace_in_file(dirpath / f, args, replacer_cls)) ~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/home/balay/petsc/./lib/petsc/bin/maint/check_header_guard.py", line 369, in replace_in_file lines = path.read_text().splitlines() ~~~~~~~~~~~~~~^^ File "/usr/lib64/python3.14/pathlib/__init__.py", line 787, in read_text with self.open(mode='r', encoding=encoding, errors=errors, newline=newline) as f: ~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ File "/usr/lib64/python3.14/pathlib/__init__.py", line 771, in open return io.open(self, mode, buffering, encoding, errors, newline) ~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^ PermissionError: [Errno 13] Permission denied: '/home/balay/petsc/include/.#petscsystypes.h' balay@p1 /home/balay/petsc (main =) $ cat fix.diff diff --git a/lib/petsc/bin/maint/check_header_guard.py b/lib/petsc/bin/maint/check_header_guard.py index f246fd62890..b56528e8bc3 100755 --- a/lib/petsc/bin/maint/check_header_guard.py +++ b/lib/petsc/bin/maint/check_header_guard.py @@ -432,7 +432,7 @@ def main(args: argparse.Namespace) -> int: dirs[:] = [d for d in dirs if d not in exclude_dirs] dirpath = pathlib.Path(dirname) for f in files: - if f not in exclude_files: + if f not in exclude_files and not f.startswith('.#'): errors.extend(replace_in_file(dirpath / f, args, replacer_cls)) if errors: balay@p1 /home/balay/petsc (main =) $ patch -Np1 < fix.diff patching file lib/petsc/bin/maint/check_header_guard.py balay@p1 /home/balay/petsc (main *=) $ ./lib/petsc/bin/maint/check_header_guard.py --action=check --kind=pragma_once -- ./src ./include balay@p1 /home/balay/petsc (main *=) $ echo $? 0 On Sat, 24 Jan 2026, Barry Smith wrote: > > $ make checkbadSource > Traceback (most recent call last): > File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 508, in <module> > ret = command_line_main() > File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 495, in command_line_main > ret = main(args) > File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 436, in main > errors.extend(replace_in_file(dirpath / f, args, replacer_cls)) > File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 369, in replace_in_file > lines = path.read_text().splitlines() > File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1256, in read_text > with self.open(mode='r', encoding=encoding, errors=errors) as f: > File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1242, in open > return io.open(self, mode, buffering, encoding, errors, newline, > File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1110, in _opener > return self._accessor.open(self, flags, mode) > FileNotFoundError: [Errno 2] No such file or directory: '/Users/barrysmith/Src/petsc/include/.#petscsystypes.h' > make[1]: [checkbadSource] Error 1 (ignored) > ^Cmake: *** [checkbadSource] Interrupt: 2 > > ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions > $ rm /Users/barrysmith/Src/petsc/include/.#petscsystypes.h > ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions > $ checkbadSource > -bash: checkbadSource: command not found > ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions > $ make checkbadSource > ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) > > >

1 0

Bug in make checkbadSource python code
by Barry Smith 24 Jan '26

24 Jan '26

$ make checkbadSource Traceback (most recent call last): File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 508, in <module> ret = command_line_main() File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 495, in command_line_main ret = main(args) File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 436, in main errors.extend(replace_in_file(dirpath / f, args, replacer_cls)) File "/Users/barrysmith/Src/petsc/lib/petsc/bin/maint/check_header_guard.py", line 369, in replace_in_file lines = path.read_text().splitlines() File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1256, in read_text with self.open(mode='r', encoding=encoding, errors=errors) as f: File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1242, in open return io.open(self, mode, buffering, encoding, errors, newline, File "/Library/Developer/CommandLineTools/Library/Frameworks/Python3.framework/Versions/3.9/lib/python3.9/pathlib.py", line 1110, in _opener return self._accessor.open(self, flags, mode) FileNotFoundError: [Errno 2] No such file or directory: '/Users/barrysmith/Src/petsc/include/.#petscsystypes.h' make[1]: [checkbadSource] Error 1 (ignored) ^Cmake: *** [checkbadSource] Interrupt: 2 ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions $ rm /Users/barrysmith/Src/petsc/include/.#petscsystypes.h ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions $ checkbadSource -bash: checkbadSource: command not found ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=) arch-ftn-generation-for-context-functions $ make checkbadSource ~/Src/petsc (barry/2026-01-13/ftn-generation-for-context-functions *=)

1 0

Fwd: Google Summer of Code 2026 & NumFOCUS
by Barry Smith 22 Jan '26

22 Jan '26

If anyone is thinking of Google summer of code projects this may be a way to apply. Barry > Begin forwarded message: > > From: Arliss Collins <arliss(a)numfocus.org> > Subject: Google Summer of Code 2026 & NumFOCUS > Date: January 22, 2026 at 9:50:23 AM EST > To: bsmith(a)petsc.dev > Reply-To: arliss(a)numfocus.org > Message-Id: <1769093422652.5f4eaa5d-f7e5-4791-94bc-035216e66f9b(a)bf08x.hubspotemail.net> > > > Hello, Projects! > > We hope you are well. Google has officially announced <https://urldefense.us/v3/__https://d4jbd104.na1.hubspotlinks.com/Ctc/OT*113… > the 22nd iteration of Google Summer of Code (GSoC), and we are pleased to share that NumFOCUS will be applying again as a mentoring organization for GSoC 2026. > > Key updates: > Organization applications: January 19 – February 3, 2026 > Project focus: GSoC continues to welcome a wide range of open source projects. For 2026, there is increased emphasis on AI/ML and Security, alongside areas such as science, data, cloud, development tools, medicine, and media. > Encouragement for individual applications: Google’s selection process is not well suited to umbrella organizations. We strongly encourage projects that can apply independently to do so. We are happy to provide guidance to first-time applicant projects. Please reach out early if interested. > Project sizes: Contributors may propose projects in three sizes - small (90 hours), medium (175 hours), or large (350 hours). However, because our available slots and administrative capacity are limited, we strongly encourage projects to prioritize medium and large proposals and to submit few, if any, ideas at the small level. If this guidance creates challenges for your project or potential contributors, please contact us so we can discuss options. > We are also delighted to share that Henry Kironde and Rohit Goswami will return as GSoC Administrators for 2026. > > This year we want to go bigger and better!. > If you would like to participate in GSoC 2026 under the NumFOCUS umbrella, please review the NumFOCUS GSoC GitHub <https://urldefense.us/v3/__https://d4jbd104.na1.hubspotlinks.com/Ctc/OT*113… > repository and contact the administrators as soon as possible to get your ideas added: > Henry Kironde <henrykironde(a)gmail.com <mailto:[email protected]>> > Rohit Goswami <r95g10(a)gmail.com <mailto:[email protected]>> > Thank you for your continued support. We look forward to another impactful year of collaboration and innovation. > Best regards, > The NumFOCUS GSoC Team > NumFOCUS, P.O. Box 90596, Austin, Texas 78709 > Unsubscribe <https://urldefense.us/v3/__https://hs-39929437.s.hubspotemail.net/hs/prefer… > Manage preferences <https://urldefense.us/v3/__https://hs-39929437.s.hubspotemail.net/hs/prefer… >

1 0

Re: [petsc-dev] PETSC_FORTRAN_CALLBACK_CLASS vs PETSC_FORTRAN_CALLBACK_SUBTYPE ?
by Barry Smith 20 Jan '26

20 Jan '26

https://urldefense.us/v3/__https://gitlab.com/petsc/petsc/-/merge_requests/… > On Jan 20, 2026, at 9:26 PM, Barry Smith <bsmith(a)petsc.dev> wrote: > > Thanks. I now understand why they must be handled differently > > >> On Jan 20, 2026, at 7:47 PM, Jed Brown <jed(a)jedbrown.org> wrote: >> >> Some callbacks are meant to persist beyond an XSetType() while others must be cleared when that is done. See the last line here. >> >> PetscErrorCode PetscObjectChangeTypeName(PetscObject obj, const char type_name[]) >> { >> PetscFunctionBegin; >> PetscValidHeader(obj, 1); >> PetscCall(PetscFree(obj->type_name)); >> PetscCall(PetscStrallocpy(type_name, &obj->type_name)); >> /* Clear all the old subtype callbacks so they can't accidentally be called (shouldn't happen anyway) */ >> PetscCall(PetscArrayzero(obj->fortrancallback[PETSC_FORTRAN_CALLBACK_SUBTYPE], obj->num_fortrancallback[PETSC_FORTRAN_CALLBACK_SUBTYPE])); >> PetscFunctionReturn(PETSC_SUCCESS); >> } >> >> >> The implementation would be more complicated if they went in the same list because you'd have to distinguish class-scope reserved slots from subtype-scope free slots. >> >> Barry Smith <bsmith(a)petsc.dev> writes: >> >>> Jed, >>> >>> How come you introduced this construct? >>> >>> typedef enum { >>> PETSC_FORTRAN_CALLBACK_CLASS, >>> PETSC_FORTRAN_CALLBACK_SUBTYPE, >>> PETSC_FORTRAN_CALLBACK_MAXTYPE >>> } PetscFortranCallbackType; >>> >>> That is why are there two types of callbacks, they seem to be managed the same way but go into two different lists in the object. I don't see why there cannot just be one list they all go in? >>> >>> Thanks >>> >>> Barry >

1 0

Re: [petsc-dev] PETSC_FORTRAN_CALLBACK_CLASS vs PETSC_FORTRAN_CALLBACK_SUBTYPE ?
by Barry Smith 20 Jan '26

20 Jan '26

Thanks. I now understand why they must be handled differently > On Jan 20, 2026, at 7:47 PM, Jed Brown <jed(a)jedbrown.org> wrote: > > Some callbacks are meant to persist beyond an XSetType() while others must be cleared when that is done. See the last line here. > > PetscErrorCode PetscObjectChangeTypeName(PetscObject obj, const char type_name[]) > { > PetscFunctionBegin; > PetscValidHeader(obj, 1); > PetscCall(PetscFree(obj->type_name)); > PetscCall(PetscStrallocpy(type_name, &obj->type_name)); > /* Clear all the old subtype callbacks so they can't accidentally be called (shouldn't happen anyway) */ > PetscCall(PetscArrayzero(obj->fortrancallback[PETSC_FORTRAN_CALLBACK_SUBTYPE], obj->num_fortrancallback[PETSC_FORTRAN_CALLBACK_SUBTYPE])); > PetscFunctionReturn(PETSC_SUCCESS); > } > > > The implementation would be more complicated if they went in the same list because you'd have to distinguish class-scope reserved slots from subtype-scope free slots. > > Barry Smith <bsmith(a)petsc.dev> writes: > >> Jed, >> >> How come you introduced this construct? >> >> typedef enum { >> PETSC_FORTRAN_CALLBACK_CLASS, >> PETSC_FORTRAN_CALLBACK_SUBTYPE, >> PETSC_FORTRAN_CALLBACK_MAXTYPE >> } PetscFortranCallbackType; >> >> That is why are there two types of callbacks, they seem to be managed the same way but go into two different lists in the object. I don't see why there cannot just be one list they all go in? >> >> Thanks >> >> Barry

1 0

Re: [petsc-dev] PETSC_FORTRAN_CALLBACK_CLASS vs PETSC_FORTRAN_CALLBACK_SUBTYPE ?
by Jed Brown 20 Jan '26

20 Jan '26

Some callbacks are meant to persist beyond an XSetType() while others must be cleared when that is done. See the last line here. PetscErrorCode PetscObjectChangeTypeName(PetscObject obj, const char type_name[]) { PetscFunctionBegin; PetscValidHeader(obj, 1); PetscCall(PetscFree(obj->type_name)); PetscCall(PetscStrallocpy(type_name, &obj->type_name)); /* Clear all the old subtype callbacks so they can't accidentally be called (shouldn't happen anyway) */ PetscCall(PetscArrayzero(obj->fortrancallback[PETSC_FORTRAN_CALLBACK_SUBTYPE], obj->num_fortrancallback[PETSC_FORTRAN_CALLBACK_SUBTYPE])); PetscFunctionReturn(PETSC_SUCCESS); } The implementation would be more complicated if they went in the same list because you'd have to distinguish class-scope reserved slots from subtype-scope free slots. Barry Smith <bsmith(a)petsc.dev> writes: > Jed, > > How come you introduced this construct? > > typedef enum { > PETSC_FORTRAN_CALLBACK_CLASS, > PETSC_FORTRAN_CALLBACK_SUBTYPE, > PETSC_FORTRAN_CALLBACK_MAXTYPE > } PetscFortranCallbackType; > > That is why are there two types of callbacks, they seem to be managed the same way but go into two different lists in the object. I don't see why there cannot just be one list they all go in? > > Thanks > > Barry

1 0

PETSC_FORTRAN_CALLBACK_CLASS vs PETSC_FORTRAN_CALLBACK_SUBTYPE ?
by Barry Smith 19 Jan '26

19 Jan '26

Jed, How come you introduced this construct? typedef enum { PETSC_FORTRAN_CALLBACK_CLASS, PETSC_FORTRAN_CALLBACK_SUBTYPE, PETSC_FORTRAN_CALLBACK_MAXTYPE } PetscFortranCallbackType; That is why are there two types of callbacks, they seem to be managed the same way but go into two different lists in the object. I don't see why there cannot just be one list they all go in? Thanks Barry

1 0