Subject: Re: [boost] [Block Pointer] multithread help
From: Phil Bouchard (philippe_at_[hidden])
Date: 2011-05-13 13:10:43
On 5/13/2011 7:06 AM, Frank Mori Hess wrote:
> It looks like you might be affected by the "static initialization order
> fiasco". Your test creates a global block_ptr, whose constructor seems to
> eventually use some class static boost::mutex. It's not clear that the
> constructors for the static boost::mutex will necessarily be run before the
> constructor of your global block_ptr.
That's a good point... but technically it is declared in the header,
which is read first so it should be initialized first.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk