Add MeshCAP implementation of Mesh2 to interface with Capstone #400

bobpaw · 2023-11-03T14:37:29Z

Add MeshCAP implementation of Mesh2 to interface with Capstone CRE meshes.

Add Mesh2 wrapper for Capstone CRE meshes
Adds capstone_cli for test meshes
Update scorec packages to latest and add capstone requirements

Now a Mesh2 object can be made from a capstone MeshDatabaseInterface

note that capstone's MeshSmartIterator is located at the current item, which needs to be returned before iterator_next is called. That way MeshCAP::iterate will work as intended.

Handles seem to be unique across entities. Thus, using them to convert M_MTopo to MeshEntity* and MeshEntity* to M_MTopo.

Since capstone does not provide attach data to mesh entities and we had to implement tags ourselves, before destroying an element tags must be removed manually.

this is because collapse does not work yet!

cwsmith

Looks good. Thank you.

My biggest concern is that there appears to be no tests added to ctest (git diff origin/develop | grep add_test). Before merging this we will have to add some that cover core functionality.

Some other minor edits/comments/suggestions are below.

CMakeLists.txt

cwsmith · 2023-11-06T01:29:59Z

apf/apfMesh.cc

 {{0,1}
 ,{1,2}
 ,{2,0}
 ,{0,3}
 ,{1,3}
 ,{2,3}};
+#else


IIUC, these changes are modifying the canonical ordering of entities. If that is correct, will enabling Capstone break support for Simmetrix meshes? If we haven't tested that combination, would you please try building with -DENABLE_SIMMETRIX=on and -DENABLE_CAPSTONE=on then running ctest?

Failures with Simmetrix (mpich3) (no change with nocxx11abi)

The following tests FAILED:
96 - parallel_meshgen (Failed)
97 - parallel_meshgen_surf (Failed)
98 - parallel_meshgen_vol (Failed)

Failures with Capstone

The following tests FAILED:
5 - bezierElevation (Failed)
6 - bezierMesh (Failed)
8 - bezierRefine (Failed)
9 - bezierSubdivision (Failed)
10 - bezierValidity (Failed)
75 - nedelec (Failed)
78 - h1_shape_serial (Failed)
79 - h1_shape_parallel (Failed)

Failures with Capstone and Simmetrix

The following tests FAILED:
5 - bezierElevation (Failed)
6 - bezierMesh (Failed)
8 - bezierRefine (Failed)
9 - bezierSubdivision (Failed)
10 - bezierValidity (Failed)
80 - nedelec (Failed)
83 - h1_shape_serial (Failed)
84 - h1_shape_parallel (Failed)
96 - parallel_meshgen (Failed)
97 - parallel_meshgen_surf (Failed)
98 - parallel_meshgen_vol (Failed)

My interpretation is that the combination of the two causes no extra breakage. However, the capstone failures merit investigation.

Thank you for testing. I have seen the parallel_meshgen_* on the rhel9 systems (#398); as you said, they are unrelated to this PR. Ideally, we can quickly resolve the capstone test failures but I won't be opposed to merging this and creating another PR to fix the errors if time is tight/limited.

I'm not sure why Morteza added this change (in commit 5def445). I need to make Capstone test meshes with tets because right now all new capstone tests pass when this section is reverted. Reverting just this section fixes bezierElevation, bezierSubdivision, nedelec, and h1_shape_*. However, it doesn't fix bezierMesh or bezierValidity which is beguiling. Is there capstone documentation that would back up the tet-reorder?

Ok, so I've checked out the capstone mesh topologies and these are the correct values. Not sure if this will create incompatibilities.

apf_cap/apfCAP.cc

cwsmith · 2023-11-06T01:32:53Z

apf_cap/apfCAP.cc

I didn't do an in-depth review the code added here and just marked some commented code that should be removed.

gmi_cap/gmi_cap.cc

gmi_cap/libgmi_cap.pc.in

apf_cap/libapf_cap.pc.in

cwsmith · 2023-11-06T02:36:42Z

test/1d.cc

Were there line ending changes in this file? All but a few lines are marked as edited, but they appear the same.

Yes, there were a few lines with CRLF.

pumi/GenTag.h

onkarsahni · 2023-11-06T15:44:39Z

@cwsmith We have some mesh-adaptation tests based on surface meshes (in .cre format). Where and how should we add those? I assume we do not want to put them in the "core" repo. Also, note we cannot do mesh-verify on them as of now (due to being only surface mesh for a 3D domain). We do have output mesh for each case, however, a striaght comparison to output mesh is not a good idea (it will not be exact same output mesh if we were to change compilers for example) and so question is what should be 'assert' equivalent for this, compare mesh stats between reference stats and stats on adapted/output mesh, with some tolerance value.

cwsmith · 2023-11-06T16:52:50Z

@onkarsahni Sounds good.

The input .cre files should be added to the pumi-meshes repo (in an appropriately named subdir). Please create a PR for adding these.

Any check for correctness on the output (adapted mesh stats, etc.) would be fine with me.

onkarsahni · 2023-11-06T23:05:56Z

@bobpaw I can think of three items: 1) add .cre meshes to repo Cameron indicated (need a PR for that), 2) add test code (e.g., analytic anisotropic sizefield) in ctest, and 3) for now add some assert statements which will look like: abs(numVerts-numVertsRef)<0.05numVertsRef and abs(numElms-numElmsRef)<0.05numElmsRef. numVertsRef and numElmsRef will be hard-coded and 0.05*numVertsRef implies different is within 5% (relative tolerance). numVerts and numElms are obtained for adapted mesh. One can also add similar assert statements with relative tolerance on mesh edge lengths, e.g., for minimum, average and maximum values.

Please proceed with these. Let me know if any questions, and keep us posted.

bobpaw · 2023-11-08T17:06:14Z

@onkarsahni Sounds good.

The input .cre files should be added to the pumi-meshes repo (in an appropriately named subdir). Please create a PR for adding these.

Any check for correctness on the output (adapted mesh stats, etc.) would be fine with me.

Hi @cwsmith, I've made a local branch but need push access to pumi-meshes. Or I can make a PR from a fork.

cwsmith · 2023-11-08T17:07:21Z

@bobpaw A PR from the fork is fine. Thank you.

Remove unused files apf_capConfig.h.in, moduletemplate, gmi_cap_config.h.in Update apf_cap/pkg_tribits.cmake

apf_cap/CMakeLists.txt

Fix unsigned integer errors in capStoneAttachSolution.cc Remove unused .pc.in files in apf_cap, gmi_cap Remove spurious whitespace in apfCAP.cc Signed-off-by: Aiden Woodruff <[email protected]>

Signed-off-by: Aiden Woodruff <[email protected]>

Test capstone_meshes cube, cylinder, and wing with anisoadapt against reference entity counts. Signed-off-by: Aiden Woodruff <[email protected]>

Updated https://github.com/SCOREC/core/wiki/General-Build-instructions Signed-off-by: Aiden Woodruff <[email protected]>

bobpaw · 2023-11-13T20:58:57Z

@cwsmith, discussed this PR with Prof. Sahni. Since all tests are passing for regular core and only a few tests are failing when ENABLE_CAPSTONE=True, I think this PR should be merged and a issues should be opened to add extra capstone test cases and sort out Bezier interaction. Moving on will give us better resources (i.e. ability to convert meshes from mds->cre) to come back and solve these issues. What are your thoughts?

cwsmith · 2023-11-15T19:33:06Z

@bobpaw That is OK with me. For the tests that pass, have they been run through valgrind to check for memory leaks or errors?

destroyNative calls gmi_destroy. MeshCAP::~MeshCAP calls destroyNative. This is the only resource MeshCAP manages. Signed-off-by: Aiden Woodruff <[email protected]>

cwsmith

Looks good. Thank you.

For the record, the issue for the capstone memory leaks is here: #404

Is the .gitmodule for pumi-meshes pointing at the latest version (that has the capstone meshes)?

Signed-off-by: Aiden Woodruff <[email protected]>

bobpaw · 2023-11-30T16:01:31Z

@cwsmith it is now!

cwsmith · 2023-11-30T16:14:41Z

@bobpaw Thank you. I approved the PR. If the 'merge pull request' button isn't grayed out, please go ahead and merge when you are ready. Otherwise, let me know if you are done and I'll click merge.

We meant to do this when merging #400 but it was accidentally reverted by commit f42cb13. Signed-off-by: Aiden Woodruff <[email protected]>

mortezah added 30 commits November 2, 2023 10:49

First attempt at cap-stone Mesh2 wrapper

29c4a49

Basic cmake setup for capstone wrapper

fc9304d

config file for capstone

17d290b

updated dependencies

68aea15

MeshCAP ctor + createMesh

d40df73

Now a Mesh2 object can be made from a capstone MeshDatabaseInterface

iterators + getPoint_

46461ab

first attempt at adjacencies

cd90e47

gmi wrappers for capstone

096f72b

fixed iterate in MeshCAP

ed74a33

note that capstone's MeshSmartIterator is located at the current item, which needs to be returned before iterator_next is called. That way MeshCAP::iterate will work as intended.

more gmi and apf wrappers implemented

1a54855

return function names for un-impl-ed gmi wrappers

ba02a7d

Determining periodicity using parametrization type

893c13f

reordered members in apfCAP.cc to match with .h

7daa993

MeshCAP::createVert_ was added

b21621b

rest of adjacencies added.

29e1518

base tag class + double tag

86c94f2

comment clean up

96c329e

int tag added

b582525

isOwned and isShared added

d0ff10e

using M_MTopo handles instead of a new class

9c31d65

Handles seem to be unique across entities. Thus, using them to convert M_MTopo to MeshEntity* and MeshEntity* to M_MTopo.

using M_GTopo handles instead of a new class

923a45c

reparam and first_derivative wrappers

662fa8e

setPoint_ and createEntity_ wrappers in apfCAP.cc

cf65617

Handling cases where input model is empty

b7760ec

buildElement->MeshCAP::createEntity when using cap

f779676

MeshCAP::destroy_ implemented

6cd320a

MeshCAP::{increment, isDone, deref} implemented

5c45b67

MeshCAP::setParam implemented

8e65151

tags must be destroyed manually

6bbe67e

Since capstone does not provide attach data to mesh entities and we had to implement tags ourselves, before destroying an element tags must be removed manually.

disable advance snapping for capstone meshes

db7829f

this is because collapse does not work yet!

bobpaw requested a review from cwsmith November 3, 2023 15:05

cwsmith self-assigned this Nov 6, 2023

cwsmith requested changes Nov 6, 2023

View reviewed changes

Remove comments and revert gentag, env-scorec

5a48535

Remove unused files apf_capConfig.h.in, moduletemplate, gmi_cap_config.h.in Update apf_cap/pkg_tribits.cmake

bobpaw commented Nov 8, 2023

View reviewed changes

apf_cap/CMakeLists.txt Show resolved Hide resolved

bobpaw added 4 commits November 9, 2023 12:22

Link apf_cap to capstone_clis

6801aba

Fix unsigned integer errors in capStoneAttachSolution.cc Remove unused .pc.in files in apf_cap, gmi_cap Remove spurious whitespace in apfCAP.cc Signed-off-by: Aiden Woodruff <[email protected]>

Convert find_package to bob_public_dep

453362b

Signed-off-by: Aiden Woodruff <[email protected]>

Add capstone tests

81bcc36

Test capstone_meshes cube, cylinder, and wing with anisoadapt against reference entity counts. Signed-off-by: Aiden Woodruff <[email protected]>

Remove capstone-config.sh and added info to wiki page

609e1ac

Updated https://github.com/SCOREC/core/wiki/General-Build-instructions Signed-off-by: Aiden Woodruff <[email protected]>

Defined MeshCAP::destroyNative

40ebce9

destroyNative calls gmi_destroy. MeshCAP::~MeshCAP calls destroyNative. This is the only resource MeshCAP manages. Signed-off-by: Aiden Woodruff <[email protected]>

bobpaw requested a review from cwsmith November 23, 2023 18:47

cwsmith reviewed Nov 30, 2023

View reviewed changes

Update pumi-meshes submodule

ba627b8

Signed-off-by: Aiden Woodruff <[email protected]>

cwsmith approved these changes Nov 30, 2023

View reviewed changes

cwsmith added the v2.2.9 label Nov 30, 2023

cwsmith mentioned this pull request Nov 30, 2023

PUMI 2.2.9 Release #405

Closed

bobpaw merged commit 25e1ae8 into develop Nov 30, 2023
4 checks passed

bobpaw added a commit that referenced this pull request May 8, 2024

fix: Update pumi-meshes submodule to latest upstream

b3c5b28

We meant to do this when merging #400 but it was accidentally reverted by commit f42cb13. Signed-off-by: Aiden Woodruff <[email protected]>

bobpaw mentioned this pull request May 8, 2024

fix: Update pumi-meshes submodule to latest upstream #428

Merged

bobpaw deleted the develop-capstone-rebase branch September 24, 2024 16:29

cwsmith added the capstone label Oct 23, 2024

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Add MeshCAP implementation of Mesh2 to interface with Capstone #400

Add MeshCAP implementation of Mesh2 to interface with Capstone #400

bobpaw commented Nov 3, 2023

cwsmith left a comment

cwsmith Nov 6, 2023

bobpaw Nov 9, 2023 •

edited

Loading

cwsmith Nov 9, 2023

bobpaw Nov 13, 2023

bobpaw Nov 23, 2023

cwsmith Nov 6, 2023

cwsmith Nov 6, 2023

bobpaw Nov 8, 2023

onkarsahni commented Nov 6, 2023

cwsmith commented Nov 6, 2023

onkarsahni commented Nov 6, 2023

bobpaw commented Nov 8, 2023

cwsmith commented Nov 8, 2023

bobpaw commented Nov 13, 2023

cwsmith commented Nov 15, 2023

cwsmith left a comment •

edited

Loading

bobpaw commented Nov 30, 2023

cwsmith commented Nov 30, 2023

Add MeshCAP implementation of Mesh2 to interface with Capstone #400

Add MeshCAP implementation of Mesh2 to interface with Capstone #400

Conversation

bobpaw commented Nov 3, 2023

Add MeshCAP implementation of Mesh2 to interface with Capstone CRE meshes.

cwsmith left a comment

Choose a reason for hiding this comment

cwsmith Nov 6, 2023

Choose a reason for hiding this comment

bobpaw Nov 9, 2023 • edited Loading

Choose a reason for hiding this comment

Failures with Simmetrix (mpich3) (no change with nocxx11abi)

Failures with Capstone

Failures with Capstone and Simmetrix

cwsmith Nov 9, 2023

Choose a reason for hiding this comment

bobpaw Nov 13, 2023

Choose a reason for hiding this comment

bobpaw Nov 23, 2023

Choose a reason for hiding this comment

cwsmith Nov 6, 2023

Choose a reason for hiding this comment

cwsmith Nov 6, 2023

Choose a reason for hiding this comment

bobpaw Nov 8, 2023

Choose a reason for hiding this comment

onkarsahni commented Nov 6, 2023

cwsmith commented Nov 6, 2023

onkarsahni commented Nov 6, 2023

bobpaw commented Nov 8, 2023

cwsmith commented Nov 8, 2023

bobpaw commented Nov 13, 2023

cwsmith commented Nov 15, 2023

cwsmith left a comment • edited Loading

Choose a reason for hiding this comment

bobpaw commented Nov 30, 2023

cwsmith commented Nov 30, 2023

bobpaw Nov 9, 2023 •

edited

Loading

cwsmith left a comment •

edited

Loading