Paul Berry <stereotype...@gmail.com> writes: > On 7 December 2013 17:11, Francisco Jerez <curroje...@riseup.net> wrote: > >> Paul Berry <stereotype...@gmail.com> writes: >[...] > > Wow, that explains why I never managed to pick up the convention by > osmosis. Thanks for explaining. > > BTW, after reading through format_pack.c, I found some exceptions to what > you said above. The following formats are always packed by format_pack.c > in reverse memory order, regardless of endianness: > > - MESA_FORMAT_RGB888 > - MESA_FORMAT_BGR888 > - MESA_FORMAT_SRGB8 > - MESA_FORMAT_XBGR16161616_UNORM > - MESA_FORMAT_XBGR16161616_SNORM > - MESA_FORMAT_XBGR16161616_FLOAT > - MESA_FORMAT_XBGR32323232_FLOAT > > Which means they get packed correctly on little-endian systems but > backwards on big-endian systems. This seems like it's a bug in > format_pack.c, so it's not relevant to your patch :)
I suspect that format_pack.c is correct and those formats were meant to be packed with a fixed memory ordering, it's just the naming that seems to be inconsistent.
pgpSjgvFN9CuG.pgp
Description: PGP signature
_______________________________________________ mesa-dev mailing list mesa-dev@lists.freedesktop.org http://lists.freedesktop.org/mailman/listinfo/mesa-dev
