OTL problem: "Bad header in CPIO packet"

[Pavlo Penenko]

Hi,

I wonder if any of you guys encountered the following problem creating types defined ieCoreHoudini.otl. E.g.

======================================================================
ERROR: testCreateRmanInject (__main__.TestCortexRmanInject)
----------------------------------------------------------------------
Traceback (most recent call last):
  File "test/IECoreHoudini/CortexRmanInject.py", line 55, in testCreateRmanInject
    rmaninject = hou.node("/shop").createNode("cortexRmanInject")
  File "/mnt/software/houdini/hfs13.0.509/houdini/python2.6libs/hou.py", line 22595, in createNode
    return _hou.Node_createNode(*args, **kwargs)
MatchDefinitionError: Failed to match node type definition.
Error:       Bad header in CPIO packet ().

I seem to have fixed this by replacing -C with -c in this SConstruct line:

otlCommand = houdiniPluginEnv.Command( otlTarget, otlPath, "cp -r %s %s; $HOUDINI_BIN_PATH/hotl -C %s $TARGET; rm -rf %s" % ( otlPath, buildPath, buildPath, buildPath ) )

According to hotl usage: "The -C option causes Contents sections to be rebuilt from their expanded archives." Is it possible that the expanded files in git are corrupt? Suspecting some git storage problems, I've tried redownloading files in /otls/IECoreHoudini/ieCoreHoudini from both GitHub and Google Code and it didn't help.

We're using Cortex 8.4.2 with Houdini 13.

Thanks in advance,
Pavlo

[Andrew Kaufman]

I just tried running the tests with Cortex 8.5.2 and Houdini 13.0.376, and I got one failure, but the test you mentioned passed just fine. So I am able to create the Rman Shop, but I guess it isn't executing as expected...

We definitely need to -C option, as otherwise changes to the expanded contents won't be reflected in the build. The reason -c is working for you is because we also have the compressed contents sitting there, though we probably shouldn't as it gets re-written when you run a build.

I'm not really sure what this CPIO packet stuff is about... maybe that's a question SideFx could shed some light on?

Andrew


testCreateRmanInject (CortexRmanInject.TestCortexRmanInject) ... ok
testRenderRmanInject (CortexRmanInject.TestCortexRmanInject) ... FAIL

======================================================================
FAIL: testRenderRmanInject (CortexRmanInject.TestCortexRmanInject)

----------------------------------------------------------------------
Traceback (most recent call last):

  File "/disk1/cortexBuilds/cortex-8.5.2/test/IECoreHoudini/CortexRmanInject.py", line 97, in testRenderRmanInject
    self.failUnless( "['-radius', '1', '-theta', '360']" in procs[0] )
AssertionError: False is not true

----------------------------------------------------------------------
Ran 407 tests in 61.478s
FAILED (failures=1)

--
--
You received this message because you are subscribed to the "cortexdev" group.
To post to this group, send email to cort...@googlegroups.com
To unsubscribe from this group, send email to cortexdev-...@googlegroups.com
For more options, visit this group at http://groups.google.com/group/cortexdev?hl=en
---
You received this message because you are subscribed to the Google Groups "cortexdev" group.
To unsubscribe from this group and stop receiving emails from it, send an email to cortexdev+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Pavlo Penenko]

Thanks Andrew.

That was also my understanding: Contents.gz appears to be valid but Contents is causing the problem for me. When I replaced Contents with what's inside Contents.gz, the problem went away as well. Contents seems to be itself a CPIO archive, and the error is its header.

Thanks for running the tests! I guess the problem might be due to my environment. Another possibility would be that SCons fails to track some dependencies and doesn't rebuild or reinstall /plugins/houdini/ieCoreHoudini.otl for you (just a guess).

With my SCons hack (using the Contents.gz from git), I'm getting the exact same assertion error in the actual render test as you posted.