[PATCH] xf86xv.h cannot be included without first including xorg-server.h.

[Thomas Hellstrom]

On 10/14/2014 05:37 AM, Michel Dänzer wrote:
> On 14.10.2014 12:25, Stefan Dirsch wrote:
>> On Tue, Oct 14, 2014 at 10:02:13AM +0900, Michel Dänzer wrote:
>>> On 13.10.2014 21:58, Stefan Dirsch wrote:
>>>> On Tue, Oct 07, 2014 at 11:56:55AM +0100, Daniel Stone wrote:
>>>>> Hi,
>>>>>
>>>>> On 7 October 2014 10:54, Michel Dänzer <michel at daenzer.net> wrote:
>>>>>
>>>>>      On 07.10.2014 18:46, Daniel Martin wrote:
>>>>>
>>>>>          On 7 October 2014 11:29, Stefan Dirsch <sndirsch at suse.de>
>>>>> wrote:
>>>>>
>>>>>              Without this the build fails on systems with the
>>>>> latest glibc,
>>>>>              throwing this error:
>>>>>              [...]
>>>>>
>>>>>              This is caused by HAVE_STRNDUP not being set (it is
>>>>> set from
>>>>>              xorg-server.h),
>>>>>              causing os.h to redefine it.
>>>>>
>>>>>
>>>>>          Have a look at this commit in xf86-video-intel:
>>>>>              
>>>>> https://urldefense.proofpoint.com/v1/url?u=http://cgit.freedesktop.org/xorg/driver/xf86-video-intel/commit/?&k=oIvRg1%2BdGAgOoM1BIlLLqw%3D%3D%0A&r=l5Ago9ekmVFZ3c4M6eauqrJWGwjf6fTb%2BP3CxbBFkVM%3D%0A&m=xI13rHs%2BLeuvv%2FqtZFARei8HUkkffhUHhhsWCS721DM%3D%0A&s=48be33ecca86cb741cedc5bcb504b7681cd4b992020571f5f47b932cd8018591
>>>>>          id=d209b03915e4d12ce445d72ac09bb4fe664ba664
>>>>>
>>>>>          With that HAVE_STRNDUP will be defined in config.h,
>>>>>          which looks like a better solution to me.
>>>>>
>>>>>
>>>>>      I disagree. xorg-server.h must be included before any other
>>>>> headers from
>>>>>      xorg/, so Stefan's patch is the proper solution.
>>>>>
>>>>>
>>>>> Yes, absolutely. xorg-server.h does things like define _XSERVER64,
>>>>> which
>>>>> changes the size of CARD32 and thus the entire server ABI. Not
>>>>> including
>>>>> xorg-server.h first is terminally broken; this is just one symptom.
>>>>>
>>>>> Stefan:
>>>>> Reviewed-by: Daniel Stone <daniels at collabora.com>
>>>>
>>>> Thanks! Anybody, who would like to push it?
>>>
>>> Thomas already pushed it. :)
>>
>> Hmm. Then I don't see it. Latest commit in git I can see is from Sep
>> 22nd.
>>
>>    vmware/saa: Fix saa_check_poly_fill_rect_noreadback v2
>
> Ah, I thought this was about the -vmmouse patch. Sorry for the noise. :(
>
>
Hmm,
I totally missed this patch was for the -vmware driver. Sorry for that.
I need to give it a spin in our ancient-server-compilation-tool and if
there's no problem I'll push it.

Thanks,
Thomas