BUG : foamToEnsight -faceZones sometimes fails in the develop branch

foamToEnsight -faceZones '(".*")' sometimes fails in the most recent commit (7582334f). The attached case worked in commit 57a1e6d2, but fails in 7582334f. This failure only occurs in parallel and only occurs when using the -faceZones flag.

The error I receive appears to be an infinite loop of...

[adpn029:214314] *** Process received signal ***
[adpn029:214314] Signal: Segmentation fault (11)
[adpn029:214314] Signal code:  (2102221680)
[adpn029:214314] Failing at address: (nil)
[adpn029:214314] [ 0] /lib64/libc.so.6[0x3d83c326a0]
[adpn029:214314] [ 1] /apps/OpenFOAM/.builds/OFPlus/dev/OpenFOAM-plus/platforms/linux64Gcc61DPInt32Opt/lib/libconversion.so(_ZNK4Foam11ensightMesh5writeERNS_14ensightGeoFileE+0xb4e)[0x7fc64ebc91fe]
[adpn029:214314] [ 2] foamToEnsight[0x4ecb4a]
[adpn029:214314] [ 3] /lib64/libc.so.6[0x3d83c326a0]
[adpn029:214317] [ 1] /apps/OpenFOAM/.builds/OFPlus/dev/OpenFOAM-plus/platforms/linux64Gcc61DPInt32Opt/lib/libconversion.so(_ZNK4Foam11ensightMesh5writeERNS_14ensightGeoFileE+0xb4e)[0x7fc374cae1fe]
[adpn029:214317] [ 2] foamToEnsight[0x4ecb4a]
[adpn029:214317] [ 3] /lib64/libc.so.6(__libc_start_main+0xfd)[0x3d83c1ed5d]
[adpn029:214317] [ 4] foamToEnsight[0x44d19d]
[adpn029:214317] *** End of error message ***

dam_break.tar.gz

Changed title: foamToEnsight -faceZones sometimes fails in develop → foamToEnsight -faceZones sometimes fails in the develop branch

By Justin Graupman on 2016-11-28T22:41:23 (imported from GitLab project)

Mentioned in issue #304 (closed)

By Justin Graupman on 2016-11-28T23:06:04 (imported from GitLab project)

Changed title: foamToEnsight -faceZones sometimes fails in the develop branch → BUG : foamToEnsight -faceZones sometimes fails in the develop branch

By Justin Graupman on 2016-11-28T23:06:49 (imported from GitLab project)

I'm trying to track down what is going wrong - I'll do more in the morning.

Mentioned in commit 1323f83e627e314958033decbcc8185f279d8c03

Thanks for spotting these problems. As you with sometimes rightly noted, they didn't show in my test cases (with few fields or different combinations of fields!) I traced the main problem to incorrect addressing in the faceZone / ensight conversion, so I trust that it is now fixed. For your test case, I used a modified Allrun, otherwise you can get weird things (the 0/ directory having patches for blockMesh, not snappy and the cell distribution being wrong).

Please let me know how your tests go.

Reassigned to @mark

@mark : Thanks for your prompt and knowledgeable support. I can confirm that this issue is resolved for my test case, I will let you know if I find any more problems. You may close this ticket.

By Justin Graupman on 2016-11-29T15:55:25 (imported from GitLab project)

Thanks for confirmation.

Status changed to closed

Mentioned in issue #334 (closed)

By Justin Graupman on 2016-12-09T17:49:15 (imported from GitLab project)