Re: [Boost-build] [boost-build] darwin linker hangs when building boost.python quickstart example

Subject: Re: [Boost-build] [boost-build] darwin linker hangs when building boost.python quickstart example
From: Steven Watanabe (watanabesj_at_[hidden])
Date: 2012-04-02 11:16:27

AMDG

On 04/02/2012 07:19 AM, Berg, Stuart wrote:
> <snip>
> I thought this issue was related to my inability to run boost-python extension modules, but if the command works correctly when run manually, then maybe it was a red herring and my real problem is a different one.
> 
> When I attempt to use an extension module, python acts as though I have a version mismatch (see fatal python error, below).  But as far as I can tell, the 'file' and 'otool' commands aren't showing anything unusual (unless there's a big difference between 2.7.1 and 2.7.2 -- scroll to the bottom of this output to see what otool says.)  I'm out of ideas here . . . Is there anything else I should look at to figure this out?
> 

The reason that a version mismatch can cause this
error, is that the two versions reside in different
shared libraries--you effectively have two separate
python interpreters in the same process.  It doesn't
matter how similar the versions are.  With that in
mind, that paths look suspicious to me.  Can you find
the output of bjam --debug-configuration <other options>
?

> <snip>
> capture-output bin/hello.test/clang-darwin-4.2.1/debug/hello
> /bin/sh: line 6: 19223 Abort trap: 6           "/Library/Frameworks/Python.framework/Versions/2.7/bin/python" "hello.py" > "bin/hello.test/clang-darwin-4.2.1/debug/hello.output" 2>&1
> ====== BEGIN OUTPUT ======
> Fatal Python error: Interpreter not initialized (version mismatch?)
> 
> EXIT STATUS: 134
> ====== END OUTPUT ======
> <snip>
> 

In Christ,
Steven Watanabe