[MPlayer-users] does debug compile affect performance?

D Richard Felker III dalias at aerifal.cx
Tue Sep 9 17:25:46 CEST 2003


On Tue, Sep 09, 2003 at 07:00:29AM +0400, Vladimir Mosgalin wrote:
> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> On Mon, 8 Sep 2003, zimon at iki.fi wrote:
> 
> z>[Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> z>On Mon, Sep 08, 2003 at 11:46:46PM +0400, Vladimir Mosgalin wrote:
> z>> [Automatic answer: RTFM (read DOCS, FAQ), also read DOCS/bugreports.html]
> z>> Of course speed of mencoders (re-encoding file with lavc to /dev/null)
> z>> was the same.
> z>
> z>Try with time(1), how much cpu and other resources both test runs will
> z>take.
> 
> Lot of info, huh. I tried mencoder with internal bash command "time" and
> watched mplayer RSS with top. Trying /usr/bin/time and watching all this
> info in unnecessary - you'll just get the same results, check this
> yourself if you don't believe. It's not only practice, I already gave
> theoretical explanation before...
> 
> Programs compiled with -g/-g2/-g3 comparing to the ones compiled without
> it take no other resources except the disk space, period.

I maintain that this is not true. I've seen programs which, when
compiled with -g, have RSS of 900k, while the version without -g has
RSS of 500k or so. I have not investigated the cause, but it
definitely happens...

Rich



More information about the MPlayer-users mailing list