[Mesa-dev] [PATCH] Android: define PIPE_OS_BSD

[Emil Velikov]

On 7 June 2017 at 16:22, Jonathan Gray <jsg at jsg.id.au> wrote:
> On Wed, Jun 07, 2017 at 11:18:53PM +0800, Chih-Wei Huang wrote:
>> 2017-06-07 22:47 GMT+08:00 Jonathan Gray <jsg at jsg.id.au>:
>> > On Wed, Jun 07, 2017 at 10:30:49PM +0800, Chih-Wei Huang wrote:
>> >> 2017-06-07 21:54 GMT+08:00 Jonathan Gray <jsg at jsg.id.au>:
>> >> > On Wed, Jun 07, 2017 at 02:45:15PM +0100, Emil Velikov wrote:
>> >> >> On 6 June 2017 at 07:30, Chih-Wei Huang <cwhuang at android-x86.org> wrote:
>> >> >> >>> These are are due to program_invocation_short_name/getprogname. How
>> >> >> >>> about updating the file alike xmlconfig.c?
>> >> >> >>
>> >> >> >> Oh... I didn't notice that.
>> >> >> >> Let me check.
>> >> >> >
>> >> >> > Well, I saw several BSD OSes define
>> >> >> > the same context:
>> >> >> >
>> >> >> > #elif defined(__NetBSD__) && defined(__NetBSD_Version__) &&
>> >> >> > (__NetBSD_Version__ >= 106000100)
>> >> >> > #    include <stdlib.h>
>> >> >> > #    define GET_PROGRAM_NAME() getprogname()
>> >> >> > #elif defined(__DragonFly__)
>> >> >> > #    include <stdlib.h>
>> >> >> > #    define GET_PROGRAM_NAME() getprogname()
>> >> >> > #elif defined(__APPLE__)
>> >> >> > #    include <stdlib.h>
>> >> >> > #    define GET_PROGRAM_NAME() getprogname()
>> >> >> > #elif ...
>> >> >> >
>> >> >> > Not sure why has no OpenBSD?
>> >> >> Your guess is as good as mine. Jonathan, any ideas if/who this is
>> >> >> supposed to work for you guys?
>> >> >
>> >> > I have no idea where that block comes from as the latest
>> >> > src/gallium/auxiliary/os/os_process.c
>> >>
>> >> Sorry. A little off-topic.
>> >> Yes, the original patch is about os_process.c.
>> >> But now we are talking about xmlconfig.c
>> >> (in src/mesa/drivers/dri/common/ )
>> >
>> > You miss the other relevant part of that file
>> >
>> > #if !defined(GET_PROGRAM_NAME)
>> > #    if defined(__OpenBSD__) || defined(NetBSD) || defined(__UCLIBC__) || defined(ANDROID)
>> > /* This is a hack. It's said to work on OpenBSD, NetBSD and GNU.
>> >  * Rogelio M.Serrano Jr. reported it's also working with UCLIBC. It's
>> >  * used as a last resort, if there is no documented facility available. */
>> > static const char *__getProgramName () {
>> >     extern const char *__progname;
>> >     char * arg = strrchr(__progname, '/');
>> >     if (arg)
>> >         return arg+1;
>> >     else
>> >         return __progname;
>> > }
>> > #        define GET_PROGRAM_NAME() __getProgramName()
>> > #    else
>> > #        define GET_PROGRAM_NAME() ""
>> > #        warning "Per application configuration won't work with your OS version."
>> > #    endif
>> > #endif
>> >
>> > __progname exists since BSD 4.4,  OpenBSD has had getprogname() since
>> > 5.4 in 2013.  This should all be simplified to look like or use the
>> > gallium code.
>>
>> No. I didn't miss the part.
>
> __progname was never mentioned in any mail I read.
>
>> That's exact what I asked.
>> OpenBSD has getprogname().
>> Why should it fall back to the hack?
>
> Clearly the code is older than 2013.
Right - so we were a bit to sparse :-\

Chih-Wei, Jonathan any interest in cleaning up the xmlconfig.c ifdef
spaghetti and adding a proper Android fix for gallium?

-Emil