Update libmesh #15182

lindsayad · 2020-04-30T18:36:18Z

Summary of changes:

Make Elem::interior_parent always safe to call
Support shell preconditioning matrix in eigen system
Add dual shape functions
Add hash specialization for FEType
Added an API to retrieve eigenvalue
Fix an overzealous assertion in unpack_indexing
Update to MetaPhysicL 1.0.2
Make "NumericVector localize()" scalable if possible
Try appending to a VariableGroup in add_variables
Revert "Use push_parallel_packed_range in scatter_constraints"
Fixes for "make distcheck"
Don't undef or define MPI_REAL
Use push_parallel_packed_range in scatter_constraints
Use common m4 files from a submodule
Let user control sorting method used by BoundaryInfo::build_side_list()
Adding number of threads option to LibMeshInit.
QGaussLobatto: add fallback on regular Gauss quadrature
Add Tri::quality(SHAPE) metric.
Fixed valgrind issue with VecSetDM
Fix node unique_id in xda for libMesh 0.9.2
Use terser make_range() in place of IntRange where possible.
Add templated make_range() helper functions.
Don't renumber with parallel and non-contiguous numbering
AbaqusIO: Add support for "MASS" elements and "NODE" sidesets.
Respect empty variables in Nemesis_IO::write_nodal_data
More robust Partitioner::set_node_processor_ids
Introduce std::reference_wrapper for more examples
Use numerically stable Edge3::volume() algorithm
RB fix: fix to RBEIMConstruction
Multiple file support in meshtool
Small fix to RBConstruction
Introduce std::reference_wrapper in fem_system_ex1
Make Edge3::volume() calculation more robust.
Don't swap LIBS and CPPFLAGS
RBConstruction update to store an untransformed basis.
Smooth libmesh for PETSc-3.13
Fix flux error estimates when flux weights not in approximation space
Introduce std::reference_wrapper in fem_system_ex3
RBConstruction change: Small change to train_reduced_basis().
Update to post_process_elem_matrix_and_vector in RBConstruction
Bugfix in rb_construction_base.C
DofMap::add_constraint_row(): assert there is no diagonal entry
RBConstruction change to element post-processing
macrofice second-derivative quantities
Fixes for build_sphere(DistributedMesh)
Fix adjoint Dirichlet constraints with AMR
The 'extra_checks' parameter should match 'secure' for backwards comp…
FEMSystem assert should work with new casts
PetscSolverException: use specific error message if available
Add 'extra_checks' parameter to FEMap::inverse_map()
Add DirichletBoundary copy assignment operator.
FileSolutionHistory class
Testing unique_id uniqueness

Also, for conda, make sure we build libmesh without the gdb command.
Then I don't have to pass --without-gdb-backtrace every time.
Trying to spawn a gdb process takes forever. Nobody got time
for that.

Finally, in conda test and make sure we created a TIMPI library

Refs #14922

lindsayad · 2020-04-30T18:36:56Z

@milljm I think that's all I need to do with conda right?

moosebuild · 2020-04-30T18:39:46Z

libmesh

@@ -1 +1 @@
-Subproject commit 179e8c8c2dcbfcc01311bd6f5044a9534aa21584
+Subproject commit f8e214a055c88f1917903367693ba5b6ce4aeb70


Caution! This contains a submodule update

milljm · 2020-04-30T18:45:22Z

Correct!

lindsayad · 2020-04-30T20:36:56Z

This https://civet.inl.gov/job/514885/ looks like a real failure. It seems that the current TIMPI hash checked into libMesh is on an orphaned branch (@roystgnr ), so the unreachable error makes sense. I actually don't understand why this only shows up in the non-conda builds.

moosebuild · 2020-04-30T20:52:06Z

Job Documentation on 149255b wanted to post the following:

View the site here

This comment will be updated on new commits.

roystgnr · 2020-04-30T21:06:05Z

That TIMPI hash came in via libMesh/libmesh#2515 ...

Damn, this looks like my fault. I was futzing with TIMPI, was trying to push temporary tags to experiment with before submitting libMesh/libmesh#2534 , and must have unintentionally overwritten the previous tag and branch. Let me see what I can do.

roystgnr · 2020-04-30T21:10:19Z

Give it a try now?

moosebuild · 2020-04-30T21:19:10Z

Job Test timings on 5df2ade : invalidated by @lindsayad

lindsayad · 2020-04-30T21:31:57Z

These failures look like they're caused by libMesh/libmesh#2491. Perhaps they need a fix like that in #15012?

moosebuild · 2020-04-30T21:36:31Z

Job Code Coverage on 5df2ade wanted to post the following:

View code coverage site here

lindsayad · 2020-04-30T21:36:38Z

Wow the valgrind memory errors in https://civet.inl.gov/job/514877/ are unreadable. Is this a conda artifact?

lindsayad · 2020-04-30T21:38:47Z

Give it a try now?

This failed again. If I run git fetch; git fetch --tag in TIMPI I still don't see a corresponding branch or tag:

* acb414b 2020-04-30 Roy Stogner (tag: v1.1.1_bootstrapped) Add bootstrap output
* e9ef8f6 2020-04-30 Roy Stogner Don't ignore bootstrap output
* 8a09dd6 2020-04-30 Roy Stogner (tag: v1.1.1) Update NEWS
* b391e69 2020-04-30 Roy Stogner Bump up release number for autoconf refactoring
| * a27bd7e 2020-04-29 Roy Stogner (tag: v1.1.0_bootstrapped, up/branch_1.1.0) Add bootstrap output
| * a46b021 2020-04-29 Roy Stogner Don't ignore bootstrap output
|/  
*   64f6826 2020-04-29 roystgnr (tag: v1.1.0, up/master, up/HEAD) Merge pull request #30 from roystgnr/relative_submodule_path
|\  
| * 316ab2b 2020-04-29 Roy Stogner Use a relative path for submodules
|/  
*   7fb9089 2020-04-27 Alex Lindsay Merge pull request #29 from roystgnr/common_m4
|\  
| | * 9e59828 2020-04-24 Roy Stogner (up/branch_1.1.0_acsm) Add bootstrap output
| | * 85e2b62 2020-04-24 Roy Stogner Don't ignore bootstrap output
| |/  
| * 2e9e882 2020-04-24 Roy Stogner Remove obsolete macro
| * e683273 2020-04-24 Roy Stogner Use autoconf-submodule for ACSM+AX macros
|/  
| * 4054808 2020-04-21 Roy Stogner (HEAD) Add bootstrap output
| * 858381b 2020-04-21 Roy Stogner Don't ignore bootstrap output
|/  
* 157450a 2020-04-21 Roy Stogner Add missing header

roystgnr · 2020-04-30T22:09:00Z

You might need git fetch --tag --force? Just look at the releases page - at least on Github, libMesh/TIMPI@4054808 is again the hash associated with v1.1.0_bootstrapped.

roystgnr · 2020-04-30T22:09:27Z

These failures look like they're caused by libMesh/libmesh#2491. Perhaps they need a fix like that in #15012?

Let me see if I can replicate any of those.

lindsayad · 2020-04-30T23:14:57Z

You have to have a branch pointer in order for the commit to be reachable:

(moose-boost) lindad@pop-os:~/projects/moose/libmesh/contrib/timpi(master)$ git fsck --unreachable up | grep commit.*4054808
Checking object directories: 100% (256/256), done.
Checking objects: 100% (465/465), done.
unreachable commit 4054808355ba2511247b83ac0a8aa0ce2a404e8b

roystgnr · 2020-05-01T14:39:57Z

Okay; branch_1.1.0 is now force pushed to connect to the tag too.

moosebuild · 2020-05-01T15:37:23Z

Job Linux peacock testing on 5df2ade : invalidated by @lindsayad

should be able to fetch TIMPI now

moosebuild · 2020-05-01T15:37:32Z

Job Test timings on 5df2ade : invalidated by @lindsayad

should be able to fetch TIMPI now

moosebuild · 2020-05-01T16:09:26Z

Job Test timings on 149255b wanted to post the following:

View timings here

This comment will be updated on new commits.

This was working fine in new libMesh, and in older libMesh, but not in one of our intermediate file formats, which was breaking a CI test in idaholab/moose#15182 Refs idaholab/moose#15193

roystgnr · 2020-05-01T22:32:21Z

libMesh/libmesh#2542 is merged now and should fix that unique_id test. No idea what's going on with those valgrind failures though.

fdkong · 2020-05-01T23:09:55Z

libMesh/libmesh#2542 is merged now and should fix that unique_id test. No idea what's going on with those valgrind failures though.

#15202

moosebuild · 2020-05-01T23:51:05Z

libmesh

@@ -1 +1 @@
-Subproject commit 179e8c8c2dcbfcc01311bd6f5044a9534aa21584
+Subproject commit 61aef80a6e24f2ab1f3d7ec8668ca07eabd33fd4


Caution! This contains a submodule update

lindsayad · 2020-05-02T00:01:03Z

Yes, see now this will include the libmesh changes in all the conda targets 👍

moosebuild · 2020-05-14T23:42:25Z

Job Test timings on 149255b : invalidated by @lindsayad

moosebuild · 2020-05-14T23:44:09Z

All jobs on 149255b : invalidated by @lindsayad

Jason hates us

moosebuild · 2020-05-14T23:47:09Z

libmesh

@@ -1 +1 @@
-Subproject commit 179e8c8c2dcbfcc01311bd6f5044a9534aa21584
+Subproject commit 30d38fc9b5bc608655d033117a8d5cb59c57ae4c


Caution! This contains a submodule update

moosebuild · 2020-05-15T02:33:29Z

Job Code Coverage on 149255b wanted to post the following:

View code coverage site here

lindsayad · 2020-05-15T04:07:42Z

Oh yea beautiful pass

fdkong · 2020-05-15T15:48:33Z

@lindsayad Do you want me to merge it right now?

lindsayad · 2020-05-15T16:14:27Z

Yea

…

On Fri, May 15, 2020 at 8:48 AM Fande Kong ***@***.***> wrote: @lindsayad <https://github.com/lindsayad> Do you want me to merge it right now? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#15182 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/ACOGA4ELDHRTEW5XNI6PSSDRRVP6DANCNFSM4MWID7LA> .

fdkong · 2020-05-15T16:46:35Z

@milljm It is safe to get this in?

fdkong

Let us hold on for a moment

fdkong · 2020-05-15T16:55:36Z

I will let @milljm hit button on this one!

milljm · 2020-05-15T16:55:38Z

Lets make this change to the conda_build_config.cfg:

moose_petsc:
 - 3.12.5 build_2

The problem with this, will be that this version of PETSc isn't available yet. So we will need to wait until it is. Or merge into next, where it does exist.

This will fail PRs... but not in Next (as that is where Fande's PETSc update currently exists)

moosebuild · 2020-05-15T17:13:38Z

libmesh

@@ -1 +1 @@
-Subproject commit 179e8c8c2dcbfcc01311bd6f5044a9534aa21584
+Subproject commit 30d38fc9b5bc608655d033117a8d5cb59c57ae4c


Caution! This contains a submodule update

milljm · 2020-05-15T17:23:20Z

This should be merged anyway, as the error is due to PETSc 3.12.5 @ build_2 does not yet exist publicly (it exists in Next).

Or. We wait until Next, devel, master passes. Invalidate this, and then it will pass.

milljm

I am merging this now, because libMesh passed, and Fande's PR passed (updating PETSc which this PR now includes). We don't want to create another PR with both updates.

We discussed off-list

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

lindsayad mentioned this pull request Apr 30, 2020

Build libmesh without the gdb command #15084

Closed

moosebuild reviewed Apr 30, 2020

View reviewed changes

roystgnr mentioned this pull request May 1, 2020

Module unique_id setting isn't always unique #15193

Closed

roystgnr mentioned this pull request May 1, 2020

Fix node unique_id in xda for libMesh 0.9.2 libMesh/libmesh#2542

Merged

lindsayad mentioned this pull request May 1, 2020

Smooth libmesh for PETSc-3.13 libMesh/libmesh#2516

Merged

fdkong mentioned this pull request May 1, 2020

Fixed valgrind VecSetDM #15202

Closed

fdkong added the PR: Do not merge label May 1, 2020

moosebuild reviewed May 1, 2020

View reviewed changes

moosebuild reviewed May 14, 2020

View reviewed changes

moosebuild added the PR: Failed but allowed label May 15, 2020

lindsayad added the PR: Ready for review/merge label May 15, 2020

fdkong approved these changes May 15, 2020

View reviewed changes

fdkong previously requested changes May 15, 2020

View reviewed changes

fdkong requested a review from milljm May 15, 2020 16:50

Include Fande's PETSc update

cfa1081

This will fail PRs... but not in Next (as that is where Fande's PETSc update currently exists)

moosebuild removed the PR: Failed but allowed label May 15, 2020

moosebuild reviewed May 15, 2020

View reviewed changes

milljm approved these changes May 15, 2020

View reviewed changes

milljm merged commit 10eaddb into idaholab:next May 15, 2020

lindsayad mentioned this pull request May 28, 2020

Enforce Ordering of Nodal Variables During Nemesis Output libMesh/libmesh#2584

Merged

cticenhour added a commit to cticenhour/moose that referenced this pull request Aug 23, 2023

Remove temporary filters on CPPFLAGS and CXXFLAGS in libmesh script

9552fab

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

cticenhour added a commit to cticenhour/moose that referenced this pull request Sep 6, 2023

Remove temporary filters on CPPFLAGS and CXXFLAGS in libmesh script

d216233

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

cticenhour added a commit to cticenhour/moose that referenced this pull request Sep 19, 2023

Remove temporary filters on CPPFLAGS and CXXFLAGS in libmesh script

642d123

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

cticenhour added a commit to cticenhour/moose that referenced this pull request Sep 20, 2023

Remove temporary filters on CPPFLAGS and CXXFLAGS in libmesh script

d121538

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

cticenhour added a commit to cticenhour/moose that referenced this pull request Sep 21, 2023

Remove temporary filters on CPPFLAGS and CXXFLAGS in libmesh script

9ccb36f

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

oanaoana pushed a commit to oanaoana/moose that referenced this pull request Oct 19, 2023

Remove temporary filters on CPPFLAGS and CXXFLAGS in libmesh script

62012e0

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

oanaoana pushed a commit to oanaoana/moose that referenced this pull request Nov 7, 2023

Remove temporary filters on CPPFLAGS and CXXFLAGS in libmesh script

4512aeb

libMesh submodule was patched as of idaholab#15182 Refs idaholab#15120

		@@ -1 +1 @@
		Subproject commit 179e8c8c2dcbfcc01311bd6f5044a9534aa21584
		Subproject commit f8e214a055c88f1917903367693ba5b6ce4aeb70

		@@ -1 +1 @@
		Subproject commit 179e8c8c2dcbfcc01311bd6f5044a9534aa21584
		Subproject commit 61aef80a6e24f2ab1f3d7ec8668ca07eabd33fd4

		@@ -1 +1 @@
		Subproject commit 179e8c8c2dcbfcc01311bd6f5044a9534aa21584
		Subproject commit 30d38fc9b5bc608655d033117a8d5cb59c57ae4c

Update libmesh #15182

Update libmesh #15182

Conversation

lindsayad commented Apr 30, 2020 • edited

lindsayad commented Apr 30, 2020

moosebuild Apr 30, 2020

Choose a reason for hiding this comment

milljm commented Apr 30, 2020

lindsayad commented Apr 30, 2020

moosebuild commented Apr 30, 2020 • edited

roystgnr commented Apr 30, 2020

roystgnr commented Apr 30, 2020

moosebuild commented Apr 30, 2020

lindsayad commented Apr 30, 2020

moosebuild commented Apr 30, 2020

lindsayad commented Apr 30, 2020

lindsayad commented Apr 30, 2020

roystgnr commented Apr 30, 2020

roystgnr commented Apr 30, 2020

lindsayad commented Apr 30, 2020

roystgnr commented May 1, 2020

moosebuild commented May 1, 2020

moosebuild commented May 1, 2020

moosebuild commented May 1, 2020 • edited

roystgnr commented May 1, 2020

fdkong commented May 1, 2020

moosebuild May 1, 2020

Choose a reason for hiding this comment

lindsayad commented May 2, 2020

moosebuild commented May 14, 2020

moosebuild commented May 14, 2020

moosebuild May 14, 2020

Choose a reason for hiding this comment

moosebuild commented May 15, 2020

lindsayad commented May 15, 2020

fdkong commented May 15, 2020

lindsayad commented May 15, 2020 via email

fdkong commented May 15, 2020

fdkong left a comment

Choose a reason for hiding this comment

fdkong commented May 15, 2020

milljm commented May 15, 2020 • edited

moosebuild May 15, 2020

Choose a reason for hiding this comment

milljm commented May 15, 2020 • edited

milljm left a comment

Choose a reason for hiding this comment

lindsayad commented Apr 30, 2020 •

edited

moosebuild commented Apr 30, 2020 •

edited

moosebuild commented May 1, 2020 •

edited

milljm commented May 15, 2020 •

edited

milljm commented May 15, 2020 •

edited