William R Tobin
William R Tobin
This issue might not occur when using the `apf::createNumbering(Mesh*,const char*,FieldShape*,int)` function instead, I'm looking into that.
Okay using the alternate `apf::createNumbering()` function avoids this issue, but using this workaround means that `apf::getField(num)` will return `NULL`. The bug in the simpler `apf::createNumbering()` function still needs to be...
I haven't checked but do we do any coverage analysis? In my thinking, that would be a good place to start for someone assigned to this issue.
I'm not gonna merge this in until we're sure it won't cause any issues with the accidental merge and the revert of PR #236 (#237).
Forgetting the test case .cc file would've broken everything anyway. At least Travis catches this when I PR on the correct branch....
In my particular case it is of course also possible not to copy the meshes at all and just get by with a single reference mesh with many fields for...
I never experienced a failure directly attributable to this issue, just an invalid read reported in valgrind, which probably makes this very low priority really. I'll see if I can...
should the default value for `delete_shr` be `false`? It is automatically set to true if passed-in `Sharing* == nullptr`
The most straightforward thing to do is expose the bool through the numbering synchronize interface and also have it express a default value of `true` I'll do that for now...
added in commit 4e65c1e297d75e8cfaf55db64dbd895255d3e8cc but this may not be the best way to solve this switching the default value in `synchronizeFieldData()` to `false` might be the better solution, it depends...