[Ffmpeg-devel] gcc4 support & MMX fixups (from Debian) (original) (raw)
Michael Niedermayer michaelni
Thu Feb 2 03:14:56 CET 2006
- Previous message: [Ffmpeg-devel] gcc4 support & MMX fixups (from Debian)
- Next message: [Ffmpeg-devel] gcc4 support & MMX fixups (from Debian)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]
Hi
On Thu, Feb 02, 2006 at 01:47:09AM +0000, M?ns Rullg?rd wrote:
Michael Niedermayer writes:
> Hi > > On Wed, Feb 01, 2006 at 10:24:55PM +0100, Pawe?? Sikora wrote: >> Dnia Wednesday, 1 of February 2006 21:59, M?ns Rullg?rd napisa??: >> >> > > maybe we should typedef int int64t; on x86-64? arrays where space >> > > matters should be of the intXXt type or similar anyway >> > >> > Did you by any chance mean #define int int64t? The typedef you >> > suggest is illegal, and will break things. > > yes ... > >> >> I suggest we should use a sizet type for params like *stride >> to avoid (on x86-64) unneeded 32<->64 conversions. > > no, sizet is unsigned, stride must be signed! intfast32t?
if that is 64bits on x86_64 ...
[...]
-- Michael
- Previous message: [Ffmpeg-devel] gcc4 support & MMX fixups (from Debian)
- Next message: [Ffmpeg-devel] gcc4 support & MMX fixups (from Debian)
- Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]