Re: [PATCH] Initialize m_hFile in FileMapping constructor.

On Fri, 2012-03-23 at 09:00 +0100, Stephan Bergmann wrote:
> That looks like a compiler error.  Have you tried with a plain 
> (non-OpenSUSE) GCC 4.6.2, too?  (What's also odd is the mention of 
> "lockbyte.css" -- is that a typo of yours, or is that a temporary file 
> that specific compiler generates internally?)

*our* Fedora gcc-4.6.2-1.fc16.x86_64 gives that exact same warning
too :-) Which was driving me wild yesterday with its
unhelpful"prephitmp.221" info. It doesn't appear with lower optimization
levels, and up until yesterday I had being building --with-dbgutil
--with-debug for the last few months, so the appearance of this warning
had slipped by me as did a handful of warnings from the additional 

export CFLAGS="-Wp,-D_FORTIFY_SOURCE=2 -fstack-protector
--param=ssp-buffer-size=4"
export CXXFLAGS="-Wp,-D_FORTIFY_SOURCE=2 -fstack-protector
--param=ssp-buffer-size=4"

that I stick in my env.

> ...but always initializing m_hFile is probably not a bad idea, anyway. 
> And if it helps your compiler, all the better.  ;)

It makes me very happy anyway as it was proving a bit of a pig to find
*what* line of code gcc was actually tripping over.

C.