Hi Marc,

Apparently, these tests were removed upstream. As we are going to
upgrade xserver-xorg package to 1.12.3, those fixes will be included
too. However, if you need the fix now, I attached the patch.

Thanks,
Laurentiu

On 07/16/2012 08:56 PM, Marc Ferland wrote:
> Hi all,
> 
> I'm having trouble building the xserver-xorg package. The build fails
> with:
> 
> | render2.c: In function '__glXDisp_Map1d':
> | render2.c:104:5: error: the comparison will always evaluate as 'true' for the address of 'u1' will never be NULL [-Werror=address]
> | render2.c:105:5: error: the comparison will always evaluate as 'true' for the address of 'u2' will never be NULL [-Werror=address]
> | render2.c: In function '__glXDisp_Map2d':
> | render2.c:147:5: error: the comparison will always evaluate as 'true' for the address of 'u1' will never be NULL [-Werror=address]
> | render2.c:148:5: error: the comparison will always evaluate as 'true' for the address of 'u2' will never be NULL [-Werror=address]
> | render2.c:149:5: error: the comparison will always evaluate as 'true' for the address of 'v1' will never be NULL [-Werror=address]
> | render2.c:150:5: error: the comparison will always evaluate as 'true' for the address of 'v2' will never be NULL [-Werror=address]
> | cc1: some warnings being treated as errors
> 
> The target machine is a x86_64. Looks like the compiler doesn't like how
> the __GLX_GET_DOUBLE macro tests the address of stack variables against
> NULL.
> 
> I would like to know if there is a patch upstream or if this is a known
> issue? What surprises me the most is that this machine looks a lot like
> the meta-sugarbay which builds correctly.
> 
> Regards,
> 
> Marc
> _______________________________________________
> yocto mailing list
> yocto@yoctoproject.org
> https://lists.yoctoproject.org/listinfo/yocto
>