[Patch]ad-hwac3.c minor fix

When the crc for ac3 block failed, there is a very high chance it is wrong detected( ac3 has
shorter sync word length ), program should return error to give DTS decoder a decent chance to try
to decode correctly.

In one of my XBMC test Swedish Radio test sample, that is the case.

This patch will sort out the wrong detection of AC3, and relay the task to libDTS decoder(Not
passthrough code). If the passthrouh is still wanted, then this routine need to rewrite the
detection part to allow both AC3 and DTS has been tested before send back fail signal.

diff -u -r1.10 ad_hwac3.c
--- libmpcodecs/ad_hwac3.c	26 Sep 2003 19:21:08 -0000	1.10
+++ libmpcodecs/ad_hwac3.c	1 Nov 2004 00:01:07 -0000
 <at>  <at>  -92,7 +92,10  <at>  <at> 

   // TODO: is DTS also checksummed?
   if(isdts == 0 && crc16_block(sh_audio->a_in_buffer + 2, length - 2) != 0)
+  {
     mp_msg(MSGT_DECAUDIO, MSGL_STATUS, "a52: CRC check failed!  \n");
+    return -1;
+  }

   return length;
 }

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 

[Patch] vf_scale.c memleak

The local pointer need to be freed befroe the function exits, there is no other pointer store
those allocated memory pointers.

diff -u -r1.55 vf_scale.c
--- libmpcodecs/vf_scale.c	18 Sep 2004 00:08:16 -0000	1.55
+++ libmpcodecs/vf_scale.c	1 Nov 2004 00:23:28 -0000
 <at>  <at>  -198,6 +198,9  <at>  <at> 
 	    best,
 	    int_sws_flags | get_sws_cpuflags(), srcFilter, dstFilter, vf->priv->param);
     }
+    if (srcFilter) sws_freeFilter(srcFilter);
+    if (dstFilter) sws_freeFilter(dstFilter);
+    srcFilter=dstFilter=NULL;
     if(!vf->priv->ctx){
 	// error...
 	mp_msg(MSGT_VFILTER,MSGL_WARN,"Couldn't init SwScaler for this setup\n");

		
__________________________________
Do you Yahoo!?
Yahoo! Mail Address AutoComplete - You start. We finish.
http://promotions.yahoo.com/new_mail 

[Patch]demux_real.c memleak

A simple patch to free index data allocated

diff -u -r1.61 demux_real.c
--- libmpdemux/demux_real.c	10 Aug 2004 20:13:44 -0000	1.61
+++ libmpdemux/demux_real.c	1 Nov 2004 00:37:06 -0000
 <at>  <at>  -1516,11 +1516,15  <at>  <at> 

 void demux_close_real(demuxer_t *demuxer)
 {
+    int i;
     real_priv_t* priv = demuxer->priv;

-    if (priv)
+    if (priv){
+    	for(i=0; i<MAX_STREAMS; i++)
+	   if(priv->index_table[i])
+	      free(priv->index_table[i]);
 	free(priv);
-
+    }
     return;
 }

		
__________________________________
Do you Yahoo!?
Yahoo! Mail Address AutoComplete - You start. We finish.
http://promotions.yahoo.com/new_mail 

Re: [Patch] vf_scale.c memleak

Hi

On Monday 01 November 2004 01:26, Wei Jiang wrote:
> The local pointer need to be freed befroe the function exits, there is no
> other pointer store those allocated memory pointers.

static SwsFilter *src_filter= NULL;

>
> diff -u -r1.55 vf_scale.c
> --- libmpcodecs/vf_scale.c 18 Sep 2004 00:08:16 -0000 1.55
> +++ libmpcodecs/vf_scale.c 1 Nov 2004 00:23:28 -0000
>  <at>  <at>  -198,6 +198,9  <at>  <at> 
>       best,
>       int_sws_flags | get_sws_cpuflags(), srcFilter, dstFilter,
> vf->priv->param); }
> +    if (srcFilter) sws_freeFilter(srcFilter);
> +    if (dstFilter) sws_freeFilter(dstFilter);
> +    srcFilter=dstFilter=NULL;

rejected

[...]
--

-- 
Michael

"I do not agree with what you have to say, but I'll defend to the death your
right to say it." -- Voltaire

Re: [Patch]demux_real.c memleak

Wei Jiang wrote:

> A simple patch to free index data allocated
<snip inline patch>

If you're going to continue sending patches, please read
DOCS/tech/patches.txt; I'm thinking specifically of rule 8.

--

-- 
       The Wanderer

Warning: Simply because I argue an issue does not mean I agree with any
side of it.

A government exists to serve its citizens, not to control them.

Re: [PATCH] GNOME-VFS layer

hi,

On Tue, 26 Oct 2004 21:23:57 +0300
Jan Knutar <jknutar <at> nic.fi> wrote:

> Is there anyway to detect broken seeking, and only enable the
> reopen+dump with large warnings printed, if seeking is broken? That
> would be ideal.
> 

the demuxer set demuxer->seekable = 0 if seeking is disabled or
demuxer->seekable = 1 if enabled respectively maybe that helps ?

regards
reza jelveh

Re: [PATCH] ve_xvid4.c: have each option flag have its counterpart

Hi,
Le lundi 01 novembre 2004 à 00:13 +0100, Reimar Döffinger a écrit :
> > -.B greyscale
> > +.B greyscale | nogreyscale
> 
> How should those "no"-variants be indicated? It's cuurently
> inconsistent. I found 4 times this variant and 10 times something like
> (no)optionname

Good catch! I indeed followed Loren's style. Let me attach an updated
patch that features the correct style for all XviD's and x264's options
that have a (no)counterpart.

Regards,
Guillaume
PS: please wait for one last correction on the XviD section (about divx5
compatibility) before releasing!

_______________________________________________
MPlayer-dev-eng mailing list
MPlayer-dev-eng <at> mplayerhq.hu
http://mplayerhq.hu/mailman/listinfo/mplayer-dev-eng

[BUG]: compile error on Darwin/ppc 6.8

There are two failures:

1)

In file included from
/System/Library/Frameworks/CoreFoundation.framework/Headers/CFBase.h:18,
                 from
/System/Library/Frameworks/CoreFoundation.framework/Headers/CoreFoundation.h:5,
                 from vcd_read_darwin.h:2,
                 from
stream_vcd.c:23:
/System/Library/Frameworks/CoreServices.framework/Headers/../Frameworks/CarbonCore.framework/Headers/MacTypes.h:75:
error: redefinition of `struct
wide'
/System/Library/Frameworks/CoreServices.framework/Headers/../Frameworks/CarbonCore.framework/Headers/MacTypes.h:80:
error: redefinition of `struct UnsignedWide'
make[1]: *** [stream_vcd.o] Error 1
make: *** [libmpdemux/libmpdemux.a] Error 2

2) See TOOLS/subfont-c/Makefile
There's this line:

LDLIBS=-lm ../../osdep/timer-lx.o $(shell freetype-config --libs)

the correct line to work on darwin is this:

LDLIBS=-lm ../../osdep/timer-darwin.o $(shell freetype-config --libs)
-liconv

Re: Re: [PATCH] Separated LIVE.COM library and include directories

> > > > Will configuration continue to work properly if the user specifies
> > > > "--with-livelibdir" but not "--with-liveincdir"?
> > > >
> > >
> > > I have made a small modification to the patch that should work.
> >
> > Any comments about this patch ?
>
>Ross, could check this one?

Diego (et al),

Sorry for the delay in responding.

I checked the latest CVS version of MPlayer.  It works as expected.

	Ross Finlayson
	LIVE.COM
	<http://www.live.com/>

How to access svn://svn.videolan.org

Hi,
I am new to this mailing list and the mplayer. I am interested in
working with the x264 player. But I don't know how to access the
following file:
> svn co svn://svn.videolan.org/x264/trunk x264 --revision 26
Which browser can I used to access this site? I tried with IE but end
up with the error message 'The page you are looking for might have
been removed or had its name changed.' To be honest, I know nothing
about the 'svn' protocol.
HKA