[FFmpeg-cvslog] r9307 - in trunk/libavcodec: dsputil.c mpegvideo.c

[Måns Rullgård]

Aurelien Jacobs <aurel at gnuage.org> writes:

> On Fri, 15 Jun 2007 14:52:35 +0200
> Michael Niedermayer <michaelni at gmx.at> wrote:
>
>> Hi
>> 
>> On Fri, Jun 15, 2007 at 01:22:09PM +0200, Aurelien Jacobs wrote:
>> > On Fri, 15 Jun 2007 12:56:10 +0200
>> > Reimar Doeffinger <Reimar.Doeffinger at stud.uni-karlsruhe.de> wrote:
>> > 
>> > > Hello,
>> > > On Fri, Jun 15, 2007 at 10:17:09AM +0200, Diego Biurrun wrote:
>> > > > On Fri, Jun 15, 2007 at 10:06:55AM +0200, Reimar Doeffinger wrote:
>> > > > > On Thu, Jun 14, 2007 at 10:44:41AM +0200, diego wrote:
>> > > > > > 
>> > > > > > Log:
>> > > > > > Simplify init preprocessor statements.
>> > > > > > patch by Albert Lee, trisk+xine acm.jhu edu
>> > > > > 
>> > > > > As discussed this has a sideeffect and IMO also removes a feature (using
>> > > > > accelerated functions from different "areas"), so I'd request this to be
>> > > > > reverted unless there is a really, really good reason to do it like
>> > > > > this (and even then the log message is not okay).
>> > > > 
>> > > > AFAICT all of these options except for mlib are mutually exclusive,
>> > > > right?  So the right solution IMO is to take just mlib out of the #if
>> > > > #elif construction.
>> > > 
>> > > I do not know, and what is the advantage that would justify trying to
>> > > find out? Just to get rid of 12 #endifs? If it bothers you as much,
>> > > IMO just switch to the "if (ENABLED_*)" construct used elsewhere, then
>> > > you can even put it all in one line.
>> > 
>> > Indeed this would looks much cleaner.
>> > First attached patch is needed to have ENABLE_* defined for MMX, etc...
>> > Then second patch cleanup this #ifdef mess.
>> 
>> looks ok
>
> applied

Neither myself nor Diego approved this configure change.  That's not
saying I disagree with it, just that Michael isn't playing by the same
rules he sets for others.

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