python3-openimageio fails with an ImportError: libjemalloc.so.2: cannot allocate memory in static TLS block

[Roger B]

Trying to use the python3-openimageio package inside python3 fails with the following problem:

```Shell
$ python3

Python 3.8.2 (default, Apr 27 2020, 15:53:34)
[GCC 9.3.0] on linux
Type "help", "copyright", "credits" or "license" for more information.
>>> import OpenImageIO
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
ImportError: /lib/x86_64-linux-gnu/libjemalloc.so.2: cannot allocate memory in static TLS block
```

Repro steps:
- Have Ubuntu 20.04 LTS installed
- sudo apt install python3-openimageio
- Attempt to import the OpenImageIO module

ProblemType: Bug
DistroRelease: Ubuntu 20.04
Package: python3-openimageio 2.1.12.0~dfsg0-1
ProcVersionSignature: User Name 5.4.0-1020.20-azure 5.4.44
Uname: Linux 5.4.0-1020-azure x86_64
ApportVersion: 2.20.11-0ubuntu27.3
Architecture: amd64
CasperMD5CheckResult: skip
CurrentDesktop: KDE
Date: Sat Jul 11 09:34:05 2020
SourcePackage: openimageio
UpgradeStatus: No upgrade log present (probably fresh install)

[Kai Kasurinen (kai-kasurinen)]

Fixed in groovy

openvdb (7.0.0-1) experimental; urgency=medium
  * d/control: Remove dependency to jemalloc. See #951704 for details

[Kai Kasurinen (kai-kasurinen)]

Thank you for taking the time to report this bug and helping to make Ubuntu better. However, I am closing it because the bug has been fixed in the latest development version of Ubuntu.

see also bug 1882998:
openvdb6.2 is non functionnal on Focal due to jemalloc

[Roger B (rboxman)]

Ugh, what good is the LTS then... yikes?

Should I just get the groovy package of python3-openimageio and have it work correctly?

[Roger B (rboxman)]

Also, it's not clear about the duplicated bug. That one is for OpenVDB, not OpenImageIO as here.

[Kai Kasurinen (kai-kasurinen)]

libOpenImageIO.so.2.1.12 is not directly linked to libjemalloc.so.2, but is linked to libopenvdb.so.6.2. And libopenvdb.so.6.2 is directly linked to libjemalloc.so.2.

[Richard Viney (richard-nz)]

This is fixed in focal by the debdiff recently posted in https://bugs.launchpad.net/ubuntu/+source/openvdb/+bug/1882998

I've filled in the SRU template there, done a debdiff, and added ubuntu-sponsors to that bug. Hopefully that's enough to get this fix into the focal LTS.

If it's possible to add a test for this to ensure no future regression I'll do that as well. Does anyone know if that's doable in this kind of situation?

Thanks.