[FFmpeg-cvslog] r22997 - trunk/libavcodec/rawdec.c

[Måns Rullgård]

Michael Niedermayer <michaelni at gmx.at> writes:

> On Sat, May 01, 2010 at 09:10:03AM +0200, Reimar D?ffinger wrote:
>> On Sat, May 01, 2010 at 12:09:42AM +0200, michael wrote:
>> > Author: michael
>> > Date: Sat May  1 00:09:42 2010
>> > New Revision: 22997
>> > 
>> > Log:
>> > avi bgr24 padding fix.
>> > Fixes issue1901
>> > 
>> > Modified:
>> >    trunk/libavcodec/rawdec.c
>> > 
>> > Modified: trunk/libavcodec/rawdec.c
>> > ==============================================================================
>> > --- trunk/libavcodec/rawdec.c	Fri Apr 30 23:43:29 2010	(r22996)
>> > +++ trunk/libavcodec/rawdec.c	Sat May  1 00:09:42 2010	(r22997)
>> > @@ -152,6 +152,8 @@ static int raw_decode(AVCodecContext *av
>> >          memcpy(frame->data[1], avctx->palctrl->palette, AVPALETTE_SIZE);
>> >          avctx->palctrl->palette_changed = 0;
>> >      }
>> > +    if(avctx->pix_fmt==PIX_FMT_BGR24 && ((frame->linesize[0]+3)&~3)*avctx->height <= buf_size)
>> > +        frame->linesize[0] = (frame->linesize[0]+3)&~3;
>> 
>> Nice hack :-), but shouldn't it use FFALIGN?
>
> iam all for factorizng code, but when the call is biger than the code
> itself iam not sure. As is anyone knowing C should understand the code
> with FFALIGN() the person must understand what FFALIGN does as well

Why do we have FFALIGN if we are not meant to use it?  In fact, you
approved the addition of it yourself:
http://article.gmane.org/gmane.comp.video.ffmpeg.devel/91101

-- 
M?ns Rullg?rd
mans at mansr.com