build failure on netbsd-5 with gtk2+-2.20.1nb

[Greg Troxel]

Ross Burton <ross at openedhand.com> writes:

> On Fri, 2010-06-25 at 20:27 -0400, Greg Troxel wrote:
>> I got a complaint about gvalue being used undefined, and sure enough the
>> code looks like that's possible at first glance.  The inference that one
>> or the other if branch will be taken because of g_return_val_if_fail
>> returning from the function is apparently too subtle for
>> 
>>   Using built-in specs.
>>   Target: i386--netbsdelf
>>   Configured with: /usr/src/tools/gcc/../../gnu/dist/gcc4/configure --enable-long-long --disable-multilib --enable-threads --disable-symvers --build=x86_64-unknown-netbsd4.99.72 --host=i386--netbsdelf --target=i386--netbsdelf --enable-__cxa_atexit
>>   Thread model: posix
>>   gcc version 4.1.3 20080704 prerelease (NetBSD nb2 20081120)
>> 
>> Here's a diff that points out the issues and makes it build for me.
>
> I'd actually prefer initialising value to NULL so that if you pass an
> invalid gconfvalue without warnings on you get a NULL pointer instead of
> a uninitialized gvalue.

Sure - what I tried to say is:

  I don't understand geoclue's code enough to know the right fix.  My
  diff merely points out where a compiler doesn't figure out that the
  value can't ever be uninitialized.  It would be cool if someone could
  fix this correctly.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 194 bytes
Desc: not available
URL: <http://lists.freedesktop.org/archives/geoclue/attachments/20100626/3414b79f/attachment.pgp>