Unresolved printf in libtype1.a and libdrm.a

Kristian Høgsberg krh at bitplanet.net
Thu Dec 2 10:08:16 PST 2004


Yes, the fix is in CVS now.

Sérgio M. Basto wrote:
> So correction of  "unresolved printf"  is already available on CVS ? 
> 
> 
> On Thu, 2004-12-02 at 03:38, Kristian Høgsberg wrote:
> 
>>Dave Airlie wrote:
>>
>>>>These modules should be using xf86MsgVerb() or ErrorF() at best --
>>>>certainly not printf().
>>>
>>>
>>>Perhaps, but they don't at the moment so I'd think the patch should be
>>>reverted until these are fixed up.... libdrm is used outside of X.org
>>>(or at least we have some crazy plans to do so..),
>>>
>>>or should it be including another X include at the top to get printf back?
>>
>>OK, I looked into this a bit more and understand the problem better now. 
>>  The #define's in xf86.h and os.h was actually there to work around the 
>>problem that the gcc __attribute((format(printf,a,b))) function 
>>attribute conflicts with
>>
>>	#define printf xf86printf
>>
>>so it was there to temporarily #undef the printf #define and #define it 
>>again at the end of the file.  The better fix for this is to use 
>>__attribute((format(__printf__,a,b))) instead, which I just committed.
>>
>>cheers,
>>Kristian
>>_______________________________________________
>>xorg mailing list
>>xorg at freedesktop.org
>>http://freedesktop.org/mailman/listinfo/xorg




More information about the xorg mailing list