Fix for spawning windows processes

[Fridrich Strba]

OK, when we are already doing defensive programming, maybe the attached
one can be considered, although not sure whether it will ever be useful
in real life situation, but who knows since that function is generic.

Cheers

Fridrich

On Fri, 2010-04-30 at 01:52 +0200, Marcus Brinkmann wrote: 
> On 04/29/2010 04:05 PM, Ralf Habacker wrote:
> > Fridrich Strba schrieb:
> >> Whichever test was testing the passing of environmental variables before
> >> should be ok for this one I guess. We are just passing the same
> >> environment that was passed when we used spanwnve, just in a format that
> >> is understood by CreateProcessA.
> >>   
> > Many thanks for your patch, which fixes also some bus-test service
> > starting problems.
> 
> I'm really sorry for having introduced this bug, and also am very thankful
> that not only it was found and reported but even fixed, in record time no less!
> 
> To make my amends, I reviewed the patch, looks great.  In the process, I also
> found an unchecked argv + 1 in my W32 CE code.  Shouldn't be a problem if argv
> is always sane, but I'd like to see the attached patch applied, just to make
> sure we catch any errors properly instead of just, uhm, crashing :)
> 
> Thanks,
> Marcus

-------------- next part --------------
A non-text attachment was scrubbed...
Name: 0001-Be-a-bit-more-defensive-in-computing-buffer-size.patch
Type: text/x-patch
Size: 940 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/dbus/attachments/20100430/c55110da/attachment.bin>