Error when launching a Gaffer Python subprocess from Maya 2020.4

[Carlo Giesa]

Hi there!

This is a new issue which does not appear when using Maya 2019. When I test current pipeline tools from Maya 2020.4 where I launch a Python subprocess that runs Gaffer, I get following error:

RuntimeError: error when building the Gaffer template: Error: mayaCollision::initialize: Invalid TBB runtime interface version 10005 (must be >= 10006).

Any clue how I can fix this?

Thanks,
Carlo

[Carlo Giesa]

This happens with Gaffer 0.59.3.0 and 0.59.4.0. I guess that there is some incompatibility issue between the latest Maya version and Gaffer. Do you think it would be possible to update TBB in the next Gaffer release? In the meantime, I will see what I can do to start my subprocess from a clean environment without having Maya in the way.

Greets,

Carlo

--
You received this message because you are subscribed to a topic in the Google Groups "gaffer-dev" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/gaffer-dev/rJOKfYI3J6M/unsubscribe.
To unsubscribe from this group and all its topics, send an email to gaffer-dev+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gaffer-dev/fc7ba6fd-c1cf-416a-94eb-d7ac76cb2a53n%40googlegroups.com.

[John Haddon]

Hi Carlo,

We are overdue a TBB update, and we could/should probably include one in Gaffer 0.60. But if you're launching a separate subprocess, it really should be possible to avoid any conflicts anyway. The `$GAFFER_ROOT/bin/gaffer` wrapper puts our libraries first on the LD_LIBRARY_PATH, so should be loading our TBB in preference to anything Maya might have put on the path previously. So the question is - at least for me since I haven't head of it - what is `mayaCollision`, and why is it getting loaded? Is it maybe on the ARNOLD_PLUGIN_PATH, and Arnold is trying to load it as a plugin? If you don't know what it is either, it might be useful to set `LD_DEBUG=files` to see whenabouts it gets loaded and whereabouts it comes from...

Cheers...

John

[Carlo Giesa]

Hi John!

Thanks for your hints. I will give that a try tomorrow. I have no clue what mayaCollision is. There are (as far as I can tell) no custom plugins or Arnold procedurals (although I just installed Golaem for testing purposes). Anyway, tomorrow, I'll give it another try and see if I can isolate the issue.

I always had the impression (without knowing exactly why - so I might be completely wrong) that TBB is a bit of a special library that is not always as isolated as others might be.

Any clue when Gaffer 0.60.0.0 might be ready?

Greets,

Carlo

To view this discussion on the web visit https://groups.google.com/d/msgid/gaffer-dev/dd49700b-d1a4-4667-904c-20054d0704fcn%40googlegroups.com.

[John Haddon]

On Thu, Mar 4, 2021 at 6:46 PM 'Carlo Giesa' via gaffer-dev <gaffe...@googlegroups.com> wrote:

Any clue when Gaffer 0.60.0.0 might be ready?

We don't have any definite timescale, but I wouldn't anticipate it being for another month at least. The Qt upgrade in 0.59 has meant that we've only just got productions moved over to it. But as I said, because you're launching a separate process, we really should be able to deal with this particular problem without a TBB upgrade...

Cheers...

John

[Carlo Giesa]

Just a little heads-up. It came out, that my real issue was something else, completely unrelated to the error message that I posted here. Well, the error message is still here and ends up in my stderr of my subprocess. But it does actually not make fail my subprocess. Its return code is 0, as expected and my procedure does execute until its end.

So, we can chill and wait the release of Gaffer 0.60.0.0! ;)

Have a nice weekend!

Carlo

--
You received this message because you are subscribed to the Google Groups "gaffer-dev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gaffer-dev+...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gaffer-dev/CAB8pVgJSb%2B4Jo9iKgZdHZWVZxzhi6SKiCNsB3Cbv4naCqOMKWQ%40mail.gmail.com.