xf86-video-glint: Changes to 'master'

[Matthieu Herrb]

 src/Makefile.am    |    2 +-
 src/glint_driver.c |   15 +++++++++------
 2 files changed, 10 insertions(+), 7 deletions(-)

New commits:
commit fc258c28f99101f7a08bac3f08a66b70ba273161
Author: Mark Kettenis <mark.kettenis at xs4all.nl>
Date:   Thu Jun 4 11:29:10 2009 -0600

    Fix DDC probe.
    
    The driver tried to do DDC stuff without having its registers mapped.
    
    Signed-off-by: Matthieu Herrb <matthieu.herrb at laas.fr>

commit 2c022e126c027d20b4b060e31cccb012265cec6a
Author: Mark Kettenis <mark.kettenis at xs4all.nl>
Date:   Thu Jun 4 11:27:07 2009 -0600

    Only map a 64k block of registers.
    
    The driver tries to map a block of registers that is too large.  The
    mmio registers consist of two blocks of 64k.  The first 64k provide a
    little-endian view, the second 64k provide a big-endian view.  However
    the driver always tries to map 128k.  And mapping 128k from the offset
    where the big-endian view starts will fail on OpenBSD.
    
    This changes things such that the driver maps just 64k.  There is no
    reason to map the big-endian view on little-endian machines.
    
    Signed-off-by: Matthieu Herrb <matthieu.herrb at laas.fr>

commit 1b36ca0c453191e4837cea25794731b166df19f7
Author: Mark Kettenis <mark.kettenis at xs4all.nl>
Date:   Thu Jun 4 11:24:15 2009 -0600

    Fix for big-endian machines.
    
    The problem is that the card provides a big-endian view on its
    registers, which the driver uses on big-endian machines.  However, the
    driver uses MMIO_OUT32() and friends to access these registers.  And
    on sparc64, these macros do byte swapping themselves.  The net result
    of course is that no byte swapping gets done at all, and things fail
    miserably.
    
    The fix is to make the macros not do byte swapping, which can be done
    by defining SPARC_MMIO_IS_BE.  There is a similar define for powerpc
    as well.  I assume these defines were accidentally dropped in the
    autoconfiscation of Xorg.
    
    Other people came up with the same conclusion see:
    
      http://www.mail-archive.com/opensolaris-discuss@opensolaris.org/msg20874.html
    
    and
    
      https://bugs.freedesktop.org/show_bug.cgi?id=10742
    
    Signed-off-by: Matthieu Herrb <matthieu.herrb at laas.fr>