flac dev - Sep 2004 - Re: [st.n@gmx.net: Bug#200435: xmms-flac: doesn't properly support long files]

severity 200435 normal
thanks

I received this bug report from a Debian user.  I can't think of any reason
offhand why the command line tool would work while the xmms plugin would
fail.

----- Forwarded message from Stephan Niemz <st.n@gmx.net> -----

Date: Tue, 8 Jul 2003 10:24:57 +0200
From: Stephan Niemz <st.n@gmx.net>
Resent-From: Stephan Niemz <st.n@gmx.net>
To: Debian Bug Tracking System <submit@bugs.debian.org>
Subject: Bug#200435: xmms-flac: doesn't properly support long files

Package: xmms-flac
Version: 1.1.0-4
Severity: important

    Xmms-flac has problems with files longer than 55 minutes and 54 seconds.
It shows the files (and length) correctly in the playlist, but as soon as I
start playing such a file, the length kind of switches to 55:54.  I can still
play these files completely, but the remaining time is obviously not correct
and I can't seek or jump to a position after this 55:54 barrier, which is
really bad (hence the severity important).  The files decode fine with the
command line tool.

What's happening here?

Thanks,

    - Stephan.


-- System Information:
Debian Release: testing/unstable
Architecture: i386
Kernel: Linux gauss.dali.inka.de 2.4.20 #4 Wed Jun 25 16:30:40 CEST 2003 i686
Locale: LANG=C, LC_CTYPE=C

Versions of packages xmms-flac depends on:
ii  libc6                      2.3.1-17      GNU C Library: Shared libraries an
ii  libflac4                   1.1.0-4       Free Lossless Audio Codec - runtim
ii  libglib1.2                 1.2.10-9      The GLib library of C routines
ii  libgtk1.2                  1.2.10-16     The GIMP Toolkit set of widgets fo
ii  libid3-3.8.3               3.8.3-2       Library for manipulating ID3v1 and
ii  libstdc++5                 1:3.3.1-0pre0 The GNU Standard C++ Library v3
ii  xlibs                      4.2.1-9       X Window System client libraries
ii  xmms                       1.2.7-3       Versatile X audio player that look
ii  zlib1g                     1:1.1.4-13    compression library - runtime



----- End forwarded message -----

-- 
 - mdz

--- Miroslav Lichvar <lichvarm@phoenix.inf.upol.cz>
wrote:
> On Tue, Jul 08, 2003 at 11:07:09AM -0400, Matt Zimmerman wrote:
> > severity 200435 normal
> > thanks
> > 
> > I received this bug report from a Debian user.  I can't think of
> any reason
> > offhand why the command line tool would work while the xmms plugin
> would
> > fail.
> ...
> 
> It's an overflow, this patch will fix it. 
> [...]
thanks, checked it in to CVS...

Josh


__________________________________
Do you Yahoo!?
SBC Yahoo! DSL - Now only $29.95 per month!
http://sbc.yahoo.com

On Tue, Jul 08, 2003 at 11:07:09AM -0400, Matt Zimmerman
wrote:
> severity 200435 normal
> thanks
> 
> I received this bug report from a Debian user.  I can't think of any
reason
> offhand why the command line tool would work while the xmms plugin would
> fail.
...

It's an overflow, this patch will fix it. 

--- plugin.c.orig	2003-05-20 21:57:04.000000000 +0200
+++ plugin.c	2003-07-08 22:03:37.000000000 +0200
@@ -537,7 +537,7 @@
 		file_info->bits_per_sample =
metadata->data.stream_info.bits_per_sample;
 		file_info->channels = metadata->data.stream_info.channels;
 		file_info->sample_rate = metadata->data.stream_info.sample_rate;
-		file_info->length_in_msec = file_info->total_samples * 10 /
(file_info->sample_rate / 100);
+		file_info->length_in_msec = (FLAC__uint64)file_info->total_samples * 10
/ (file_info->sample_rate / 100);
 	}
 	else if(metadata->type == FLAC__METADATA_TYPE_VORBIS_COMMENT) {
 		double gain, peak;

-- 
Miroslav Lichvar

On Tue, Jul 08, 2003 at 10:11:49PM +0200, Miroslav Lichvar wrote:
> On Tue, Jul 08, 2003 at 11:07:09AM -0400, Matt Zimmerman wrote:
> > severity 200435 normal
> > thanks
> > 
> > I received this bug report from a Debian user.  I can't think of
any reason
> > offhand why the command line tool would work while the xmms plugin
would
> > fail.
> ...
> 
> It's an overflow, this patch will fix it. 
> 
> --- plugin.c.orig	2003-05-20 21:57:04.000000000 +0200
> +++ plugin.c	2003-07-08 22:03:37.000000000 +0200
> @@ -537,7 +537,7 @@
>  		file_info->bits_per_sample =
metadata->data.stream_info.bits_per_sample;
>  		file_info->channels = metadata->data.stream_info.channels;
>  		file_info->sample_rate = metadata->data.stream_info.sample_rate;
> -		file_info->length_in_msec = file_info->total_samples * 10 /
(file_info->sample_rate / 100);
> +		file_info->length_in_msec = (FLAC__uint64)file_info->total_samples
* 10 / (file_info->sample_rate / 100);
>  	}
>  	else if(metadata->type == FLAC__METADATA_TYPE_VORBIS_COMMENT) {
>  		double gain, peak;
It seems like would be simpler to do something like:

(file_info->total_samples / file_info->sample_rate) * 1000;

-- 
 - mdz

Matt Zimmerman <mdz@debian.org> wrote:
> I received this bug report from a Debian user.
>     Xmms-flac has problems with files longer than 55 minutes and 54
seconds.
> It shows the files (and length) correctly in the playlist, but as soon as I
> start playing such a file, the length kind of switches to 55:54.  I can
still
> play these files completely, but the remaining time is obviously not
correct
> and I can't seek or jump to a position after this 55:54 barrier,
I cannot reproduce this (on OpenBSD/i386).  I ripped, encoded,
played, and seeked through "Made in Japan" (76:44).  Works fine.

-- 
Christian "naddy" Weisgerber                         
naddy@mips.inka.de

On Tue, Jul 08, 2003 at 08:59:13PM +0000, Christian Weisgerber
wrote:
> Matt Zimmerman <mdz@debian.org> wrote:
> 
> > I received this bug report from a Debian user.
> 
> >     Xmms-flac has problems with files longer than 55 minutes and 54
seconds.
> > It shows the files (and length) correctly in the playlist, but as soon
as I
> > start playing such a file, the length kind of switches to 55:54.  I
can still
> > play these files completely, but the remaining time is obviously not
correct
> > and I can't seek or jump to a position after this 55:54 barrier,
> 
> I cannot reproduce this (on OpenBSD/i386).  I ripped, encoded,
> played, and seeked through "Made in Japan" (76:44).  Works fine.
That's too short :).

2^32 / 10 / 44100 / 60 = 162.3 minutes.

-- 
Miroslav Lichvar