[MPlayer-dev-eng] [PATCH] radeon_vid + fglrx support - try 2

[Vladimir Mosgalin]

On Tue, 6 Apr 2004, Alex Beregszaszi wrote:

AB>Hi,
AB>
AB>> OK, since pre4 will be soon, maybe this patch will be accepted?
AB>> Original patch was made long ago by Marcel Naziri, and it allowed to
AB>> use vidix output on radeon cards with ati binary driver (the best
AB>> driver for 8500 & up), which was messed otherwise.
AB>> 
AB>> It was said that this patch breaks compiling vidix without X11, hence
AB>> it is not acceptable to apply it, but it can be avoided by several
AB>> ifdefs.. Dunno why it wasn't done before.
AB>>  
AB>> Anyway, fixed version is here. All X11 calls are surrounded with
AB>> #ifdef HAVE_X11, defined in config.h, so the problem doesn't exist
AB>> anymore. Hope I (and all other ati users) won't need to apply it on
AB>> every mplayer compilation anymore.
AB>Imho it's safe to always shift without detection... The upper part may
AB>be reserved in several drivers.

Well, it is proved that only fglrx driver needs this for now. And isn't
the same code shared with rage128 cards? If variants with 4m or 8m
onboard memory exist (I don't remember), reducing address by 5m could
lead to problems - but I don't really know this code, maybe I'm
mistaken.

Since it already works great without shifting for all other cases, why
change?

-- 

Vladimir