[MPlayer-dev-eng] [PATCH] printf -> mp_msg

[Reynaldo H. Verdejo Pinochet]

On Sun, Mar 19, 2006 at 11:35:36PM -0500, Rich Felker wrote:
> On Sun, Mar 19, 2006 at 06:54:27PM -0400, Reynaldo H. Verdejo Pinochet wrote:
> > On Sun, Mar 19, 2006 at 11:34:00PM +0100, Diego Biurrun wrote:
> > > On Sun, Mar 19, 2006 at 07:24:37PM +0100, Ötvös Attila wrote:
> > > > 
> > > > liba52: 
> > > > parse.c, ao_oss.c, ao_sun.c, audio_out.c
> > > 
> > > This is an imported library, don't patch it.
> > > 
> > > You have a few hunks like the following in the patch:
> > > 
> > > -    if(verbose > 4)
> > > -        printf("vd_ffmpeg::mc_get_buffer\n");
> > > +//    if(verbose > 4)
> > > +        mp_msg(MSGT_DECVIDEO, MSGL_DBG4, "vd_ffmpeg::mc_get_buffer\n");
> > > 
> > > Somebody correct me if I am wrong, but I believe this is incorrect,
> > > since now the call to mp_msg will not be skipped as it should.
> > > 
> > > Diego
> > 
> > I prefer the later, having if(verbose >|<|= X) clauses _and_ MSGL_X

a missing line here:
"just for preventing an mp_msg when a certain condition of verbosity
level is not meet"

> > is just error prone, duplicated work and useles code polution IMHO.
> 
> No, the few cases where this is done it's critical -- at least for the
> few I know. The performance penalty of calling mp_msg over and over in
> these instances was deemed unacceptable so it was put under
> conditional.
> 

from mp_msg.c:

void mp_msg(int mod, int lev, const char *format, ... ){
va_list va;
char tmp[MSGSIZE_MAX];
if (!mp_msg_test(mod, lev)) return; // do not display

I know this is somewhat slower than if() but, its really that slow? so to be 
forced to check for it outside the mp_msg call?
			
> Rich

 Best regards

    Reynaldo
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 189 bytes
Desc: Digital signature
URL: <http://lists.mplayerhq.hu/pipermail/mplayer-dev-eng/attachments/20060320/b9cc5dcb/attachment.pgp>