Libmesh update 20250304 #30042

roystgnr · 2025-03-06T20:33:54Z

Quadrature API changes. A few virtual QBase methods have had
changed signatures, and some methods have been deprecated in favor
of newer versions, requiring Elem inputs rather than mere
ElemType inputs, to enable support for elements like arbitrary
polygons and polyhedra where a single class can take on different
topologies at run time.
FE and FEInterface API changes. Some methods have been
deprecated in favor of newer versions, either in Elem or taking an
Elem rather than an ElemType argument, also for runtime-topology
compatibility.
Add APIs to control System name prefixing from C++, not just from
the command line. This will enable fixes for a MOOSE bug with
independent preconditioning of multiple nonlinear systems, affecting
the optimization module.
Added an integrate_slits option to JumpErrorEstimator-based
estimators
Added an OverlapCoupling ghosting functor, useful for integrating
on slit meshes or between manifolds of overlapping meshes.
Bug fix: DistributedMesh::add_point now operates correctly when
given a node id that was recently deleted from the mesh but has not
yet been cleaned from the internal node container.
Bug fix: Prism6::build_side_ptr() now sets subdomain_id(); this
fixes a failure in adaptive mesh refinement of Prism6 elements with
subdomain-restricted variables.
Bug fix: methods which elevated mesh element order no longer
renumber nodes and elements while doing so. This fixes potential
iterator invalidations in some use cases, and it's more efficient.
Bug fix: when copying nodes and elements from a source mesh into a
target DistributedMesh that is too small to use all its processors,
the target mesh was left in an invalid state. This fixes errors
triggered by certain mesh stitching operations.
Bug fix for range estimation bug in FParser optimizer
Minor FE code refactoring for more simplicity and readability
Minor optimization to some hash (and thus unordered_map) use cases
Cleaned up and commented code with high-precision floating-point
constants

@lindsayad was hoping to get one more libMesh PR squeezed into this, and we can wait for that before merging, but I want to get the API changes into CI right away and see what surprises may be in store.

libmesh

roystgnr · 2025-03-07T00:23:51Z

Yeah, those runtime --disable-deprecated failures are the sort of thing I was looking for.

moosebuild · 2025-03-07T01:58:55Z

Job Documentation, step Docs: sync website on 077bbab wanted to post the following:

View the site here

This comment will be updated on new commits.

moosebuild · 2025-03-07T02:18:52Z

Job Coverage, step Generate coverage on 077bbab wanted to post the following:

Framework coverage

	051868	#30042 077bba
	Total	Total	+/-	New
Rate	85.29%	85.29%	-0.00%	82.76%
Hits	109056	109041	-15	24
Misses	18809	18810	+1	5

Diff coverage report

Full coverage report

Modules coverage

Mutiple base shas were found when retrieving head reports:
92e4cfa
0518686

Full coverage reports

Reports

framework
chemical_reactions (FAILED)
combined
contact (FAILED)
electromagnetics (FAILED)
external_petsc_solver (FAILED)
fluid_properties (FAILED)
fsi (FAILED)
functional_expansion_tools (FAILED)
geochemistry (FAILED)
heat_transfer (FAILED)
level_set (FAILED)
misc (FAILED)
navier_stokes (FAILED)
optimization (FAILED)
peridynamics (FAILED)
phase_field (FAILED)
porous_flow (FAILED)
ray_tracing (FAILED)
rdg (FAILED)
reactor (FAILED)
richards
scalar_transport (FAILED)
solid_mechanics (FAILED)
solid_properties (FAILED)
stochastic_tools (FAILED)
subchannel (FAILED)
thermal_hydraulics (FAILED)
xfem (FAILED)

Warnings

framework new line coverage rate 82.76% is less than the suggested 90.0%

This comment will be updated on new commits.

- Quadrature API changes. A few virtual `QBase` methods have had changed signatures, and some methods have been deprecated in favor of newer versions, requiring `Elem` inputs rather than mere `ElemType` inputs, to enable support for elements like arbitrary polygons and polyhedra where a single class can take on different topologies at run time. - `FE` and `FEInterface` API changes. Some methods have been deprecated in favor of newer versions, either in `Elem` or taking an `Elem` rather than an `ElemType` argument, also for runtime-topology compatibility. - Add APIs to control `System` name prefixing from C++, not just from the command line. This will enable fixes for a MOOSE bug with independent preconditioning of multiple nonlinear systems, affecting the optimization module. - Added an `integrate_slits` option to `JumpErrorEstimator`-based estimators - Added an `OverlapCoupling` ghosting functor, useful for integrating on slit meshes or between manifolds of overlapping meshes. - Bug fix: `DistributedMesh::add_point` now operates correctly when given a node id that was recently deleted from the mesh but has not yet been cleaned from the internal node container. - Bug fix: `Prism6::build_side_ptr()` now sets `subdomain_id()`; this fixes a failure in adaptive mesh refinement of Prism6 elements with subdomain-restricted variables. - Bug fix: methods which elevated mesh element order no longer renumber nodes and elements while doing so. This fixes potential iterator invalidations in some use cases, and it's more efficient. - Bug fix: when copying nodes and elements from a source mesh into a target `DistributedMesh` that is too small to use all its processors, the target mesh was left in an invalid state. This fixes errors triggered by certain mesh stitching operations. - Bug fix for range estimation bug in FParser optimizer - Minor FE code refactoring for more simplicity and readability - Minor optimization to some hash (and thus `unordered_map`) use cases - Cleaned up and commented code with high-precision floating-point constants Refs #0

libmesh

loganharbour · 2025-03-13T20:27:02Z

Already?!

lindsayad · 2025-03-13T20:33:28Z

We needs it

libmesh

loganharbour · 2025-03-18T14:11:31Z

ETA on this?

roystgnr · 2025-03-18T15:16:10Z

Hopefully this week, but definitely not soon at this rate. I'm still working through the upgrades for --disable-deprecated, and I can't even imagine what could be triggering e.g. PC failed due to FACTOR_OUTMEMORY in the navier_stokes distributed test step.

libmesh

lindsayad · 2025-03-18T16:46:01Z

PC failed due to FACTOR_OUTMEMORY in the navier_stokes distributed test step.

sporadic failure

GiudGiud · 2025-03-18T17:21:21Z

sporadic failure

i ll try to patch this

libmesh

libMesh::FEInterface::inverse_map is now deprecated.

The former function is now deprecated as redundant

Refs idaholab#13921

This is also a little faster, skipping an unnecessary test. This is *broken* if users try to use a sufficiently higher quadrature order on faces than on interiors for some reason, but in that use case the existing code was broken too. Fixing this without a performance regression in typical use cases is going to require a new libMesh API.

We need n_dofs() with an actual Elem if we want to be able to handle polyhedra where the result depends on more than just the ElemType

I'm not looking forward to this breaking when someone tries to get_xyz() or something, but I can't justify adding even a little more expense in an inner loop based on a hypothetical future possibility.

libmesh

moosebuild · 2025-03-20T14:04:43Z

Job Precheck, step Versioner verify on 077bbab wanted to post the following:

Versioner templates

Found 16 templates, 0 failed

Versioner versions

Found 10 packages, 2 changed, 0 failed

package	status	hash	to hash	version	to version
libmesh	CHANGE	dabc922	c2aeeb2	2025.02.25 build 0	2025.03.06 build 0
moose-dev	CHANGE	fcaa73c	ca0a369	2025.03.03	2025.03.06

moosebuild · 2025-03-20T22:10:48Z

Job Modules parallel on 077bbab : invalidated by @roystgnr

Kicking "Modules parallel" - "test:finite_volume/ins/turbulence/lid-driven.lid_driven_turb_capped/result_physics: Solve Did NOT Converge!" from a "distributed threads" step

moosebuild · 2025-03-21T01:11:28Z

Job Coverage, step Verify coverage on 077bbab wanted to post the following:

The following coverage requirement(s) failed:

Failed to generate chemical_reactions coverage rate (required: 92.0%)
Failed to generate contact coverage rate (required: 87.0%)
Failed to generate electromagnetics coverage rate (required: 94.0%)
Failed to generate external_petsc_solver coverage rate (required: 84.0%)
Failed to generate fluid_properties coverage rate (required: 84.0%)
Failed to generate fsi coverage rate (required: 85.0%)
Failed to generate functional_expansion_tools coverage rate (required: 81.0%)
Failed to generate geochemistry coverage rate (required: 96.0%)
Failed to generate heat_transfer coverage rate (required: 87.0%)
Failed to generate level_set coverage rate (required: 85.0%)
Failed to generate misc coverage rate (required: 32.0%)
Failed to generate navier_stokes coverage rate (required: 77.0%)
Failed to generate optimization coverage rate (required: 86.0%)
Failed to generate peridynamics coverage rate (required: 77.0%)
Failed to generate phase_field coverage rate (required: 85.0%)
Failed to generate porous_flow coverage rate (required: 95.0%)
Failed to generate ray_tracing coverage rate (required: 93.0%)
Failed to generate rdg coverage rate (required: 63.0%)
Failed to generate reactor coverage rate (required: 90.0%)
Failed to generate scalar_transport coverage rate (required: 81.0%)
Failed to generate solid_mechanics coverage rate (required: 84.0%)
Failed to generate solid_properties coverage rate (required: 83.0%)
Failed to generate stochastic_tools coverage rate (required: 88.0%)
Failed to generate subchannel coverage rate (required: 87.0%)
Failed to generate thermal_hydraulics coverage rate (required: 88.0%)
Failed to generate xfem coverage rate (required: 80.0%)

roystgnr · 2025-03-21T02:37:02Z

The coverage errors here are the usual "Multiple base SHAs" issue

framework/src/geomsearch/FindContactPoint.C

loganharbour · 2025-03-24T13:57:27Z

Let's leave the final merge to @milljm

milljm · 2025-03-24T14:11:04Z

Minus next being busted for something @GiudGiud says he will fix (today: #28235), I am ready for this.

cticenhour · 2025-03-24T17:26:27Z

scripts/tests/versioner_hashes.yaml

@@ -863,3 +863,34 @@ fd5f7bc651f0d6a8e2d23e5d4c2c2dadca58aa0b: #29978
  wasp:
    full_version: 2025.02.25_0
    hash: 46ea1c0
+9c3c8d833ce9c5d8fbb748dc6b870a075f4b2827:


Noted that this is missing a ticket number comment while rebasing #29984. Will add in a commit in that PR for completeness.

This must be something missing in --build-template

I thought the specific ticket number was something we should be adding manually after copy-pasting the versioner.py --summary output into versioner_hashes.yaml.

You're right. I never scrolled that far in the wiki instructions 😆 I do think that this third step should also be somehow automated and included in --build-template

roystgnr added the PR: Do not merge label Mar 6, 2025

roystgnr requested review from cticenhour, lindsayad, milljm, loganharbour and GiudGiud as code owners March 6, 2025 20:33

moosebuild reviewed Mar 6, 2025

View reviewed changes

libmesh Show resolved Hide resolved

lindsayad mentioned this pull request Mar 7, 2025

Properly prefix multiple systems #29964

Merged

roystgnr added 3 commits March 13, 2025 11:56

Newsletter entries for libMesh submodule update

cf40e22

Update versioner templates

9c3c8d8

roystgnr force-pushed the libmesh_update_20250304 branch from 9d5194c to f18838d Compare March 13, 2025 17:57

moosebuild reviewed Mar 13, 2025

View reviewed changes

libmesh Show resolved Hide resolved

moosebuild reviewed Mar 13, 2025

View reviewed changes

libmesh Show resolved Hide resolved

moosebuild reviewed Mar 17, 2025

View reviewed changes

libmesh Show resolved Hide resolved

moosebuild reviewed Mar 18, 2025

View reviewed changes

libmesh Show resolved Hide resolved

miaoyinb mentioned this pull request Mar 18, 2025

Enable Batch Mesh Generation #29801

Merged

miaoyinb mentioned this pull request Mar 18, 2025

Distributed Mesh Mode Test Failures in #29801 #30129

Closed

moosebuild reviewed Mar 18, 2025

View reviewed changes

libmesh Show resolved Hide resolved

moosebuild reviewed Mar 19, 2025

View reviewed changes

libmesh Show resolved Hide resolved

Update versioner hashes

6849d93

roystgnr added 8 commits March 19, 2025 16:26

FEInterface -> FEMap::inverse_map

aed64a9

libMesh::FEInterface::inverse_map is now deprecated.

FEInterface -> Elem::on_reference_element()

4171be4

The former function is now deprecated as redundant

Use non-deprecated libMesh n_dofs_at_node()

637dd11

Refs idaholab#13921

Avoid deprecated n_shape_functions overload

bef5c00

We need n_dofs() with an actual Elem if we want to be able to handle polyhedra where the result depends on more than just the ElemType

Avoid deprecated n_shape_functions() call

43f6ed0

Init quadrature with the full element

0af6593

Tell libMesh the deprecated init is deliberate

7ae3d34

I'm not looking forward to this breaking when someone tries to get_xyz() or something, but I can't justify adding even a little more expense in an inner loop based on a hypothetical future possibility.

roystgnr force-pushed the libmesh_update_20250304 branch from 945292b to 3f4285c Compare March 19, 2025 21:27

moosebuild reviewed Mar 19, 2025

View reviewed changes

libmesh Show resolved Hide resolved

GiudGiud assigned roystgnr Mar 20, 2025

Update deprecated quadrature init()

077bbab

roystgnr force-pushed the libmesh_update_20250304 branch from 3f4285c to 077bbab Compare March 20, 2025 13:59

moosebuild reviewed Mar 20, 2025

View reviewed changes

libmesh Show resolved Hide resolved

loganharbour approved these changes Mar 24, 2025

View reviewed changes

framework/src/geomsearch/FindContactPoint.C Show resolved Hide resolved

milljm merged commit e410ade into idaholab:next Mar 24, 2025
98 of 99 checks passed

loganharbour mentioned this pull request Mar 24, 2025

Update seacas to 2025-02-27 release #29984

Merged

roystgnr deleted the libmesh_update_20250304 branch March 24, 2025 16:06

cticenhour reviewed Mar 24, 2025

View reviewed changes

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Libmesh update 20250304 #30042

Libmesh update 20250304 #30042

roystgnr commented Mar 6, 2025

roystgnr commented Mar 7, 2025

moosebuild commented Mar 7, 2025 •

edited

Loading

moosebuild commented Mar 7, 2025 •

edited

Loading

loganharbour commented Mar 13, 2025

lindsayad commented Mar 13, 2025

loganharbour commented Mar 18, 2025

roystgnr commented Mar 18, 2025

lindsayad commented Mar 18, 2025

GiudGiud commented Mar 18, 2025

moosebuild commented Mar 20, 2025

moosebuild commented Mar 20, 2025

moosebuild commented Mar 21, 2025

roystgnr commented Mar 21, 2025

loganharbour commented Mar 24, 2025

milljm commented Mar 24, 2025

cticenhour Mar 24, 2025

lindsayad Mar 24, 2025

cticenhour Mar 24, 2025

lindsayad Mar 24, 2025

Libmesh update 20250304 #30042

Libmesh update 20250304 #30042

Conversation

roystgnr commented Mar 6, 2025

roystgnr commented Mar 7, 2025

moosebuild commented Mar 7, 2025 • edited Loading

moosebuild commented Mar 7, 2025 • edited Loading

Framework coverage

Modules coverage

Full coverage reports

Warnings

loganharbour commented Mar 13, 2025

lindsayad commented Mar 13, 2025

loganharbour commented Mar 18, 2025

roystgnr commented Mar 18, 2025

lindsayad commented Mar 18, 2025

GiudGiud commented Mar 18, 2025

moosebuild commented Mar 20, 2025

Versioner templates

Versioner versions

moosebuild commented Mar 20, 2025

moosebuild commented Mar 21, 2025

roystgnr commented Mar 21, 2025

loganharbour commented Mar 24, 2025

milljm commented Mar 24, 2025

cticenhour Mar 24, 2025

Choose a reason for hiding this comment

lindsayad Mar 24, 2025

Choose a reason for hiding this comment

cticenhour Mar 24, 2025

Choose a reason for hiding this comment

lindsayad Mar 24, 2025

Choose a reason for hiding this comment

moosebuild commented Mar 7, 2025 •

edited

Loading

moosebuild commented Mar 7, 2025 •

edited

Loading