Bug in the generation of IEC958 subframes

Bug in the generation of IEC958 subframes

[Nicolas Boullis]

[-- Attachment #1: Type: text/plain, Size: 740 bytes --]

Hi,

While working on an ALSA driver for the audio part of em8300-based MPEG 
decoder boards (Sigma Designs' Hollywood+ boards and Creative Labs' DXR3 
boards), I think I discovered a bug in the code that generates IEC958 
subframes from linear samples (iec958 plugin): M and W preambles are 
swapped, and B preambles should only be used on even subframes. The 
attached patch fixes this issue.

Moreover, (using debian's 1.0.9-3 package, I haven't checked with 
1.0.10rc3), I get segfaults whenever I try to specify the preambles in 
my ~/.asoundrc file.

(https://bugtrack.alsa-project.org/alsa-bug/ times out, that's why I'm 
sending this report here.)


Cheers,

Nicolas

PS: please CC replies to me, as I don't read the alsa-devel ML.

[-- Attachment #2: alsa-lib.patch --]
[-- Type: text/plain, Size: 668 bytes --]

diff -ruN alsa-lib-1.0.10rc3.orig/src/pcm/pcm_iec958.c alsa-lib-1.0.10rc3/src/pcm/pcm_iec958.c
--- alsa-lib-1.0.10rc3.orig/src/pcm/pcm_iec958.c	2005-05-24 16:14:35.000000000 +0200
+++ alsa-lib-1.0.10rc3/src/pcm/pcm_iec958.c	2005-11-16 01:35:46.000000000 +0100
@@ -116,10 +116,10 @@
 		data |= 0x80000000;
 
 	/* Preamble */
-	if (! iec->counter)
-		data |= iec->preamble[PREAMBLE_Z];	/* Block start, 'Z' */
-	else if (! channel)
+	if (channel)
 		data |= iec->preamble[PREAMBLE_Y];	/* odd sub frame, 'Y' */
+	else if (! iec->counter)
+		data |= iec->preamble[PREAMBLE_Z];	/* Block start, 'Z' */
 	else
 		data |= iec->preamble[PREAMBLE_X];	/* even sub frame, 'X' */
 

Re: Bug in the generation of IEC958 subframes

[Takashi Iwai]

At Wed, 16 Nov 2005 01:45:01 +0100,
Nicolas Boullis wrote:
> 
> Hi,
> 
> While working on an ALSA driver for the audio part of em8300-based MPEG 
> decoder boards (Sigma Designs' Hollywood+ boards and Creative Labs' DXR3 
> boards), I think I discovered a bug in the code that generates IEC958 
> subframes from linear samples (iec958 plugin): M and W preambles are 
> swapped, and B preambles should only be used on even subframes. The 
> attached patch fixes this issue.

Thanks, applied to CVS now.

> Moreover, (using debian's 1.0.9-3 package, I haven't checked with 
> 1.0.10rc3), I get segfaults whenever I try to specify the preambles in 
> my ~/.asoundrc file.

Hmm, the segfault is more serious.  Could you get a backtrace via gdb?


Takashi


-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.  Get Certified Today
Register for a JBoss Training Course.  Free Certification Exam
for All Training Attendees Through End of 2005. For more info visit:
http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click

Re: Bug in the generation of IEC958 subframes

[Nicolas Boullis]

Hi,

Yesterday, I could manage to acces the bugtracking system, and file bugs 
0001570 and 0001571.

On Thu, Nov 17, 2005 at 01:09:13PM +0100, Takashi Iwai wrote:
> At Wed, 16 Nov 2005 01:45:01 +0100,
> Nicolas Boullis wrote:
> > 
> > Hi,
> > 
> > While working on an ALSA driver for the audio part of em8300-based MPEG 
> > decoder boards (Sigma Designs' Hollywood+ boards and Creative Labs' DXR3 
> > boards), I think I discovered a bug in the code that generates IEC958 
> > subframes from linear samples (iec958 plugin): M and W preambles are 
> > swapped, and B preambles should only be used on even subframes. The 
> > attached patch fixes this issue.
> 
> Thanks, applied to CVS now.

Thanks. I think you can close bug #0001570 then.


> > Moreover, (using debian's 1.0.9-3 package, I haven't checked with 
> > 1.0.10rc3), I get segfaults whenever I try to specify the preambles in 
> > my ~/.asoundrc file.
> 
> Hmm, the segfault is more serious.  Could you get a backtrace via gdb?

I submitted some test .asoundrc file + a backtrace I got with valgring 
in bug #0001571. Is it sufficient or should I also get a gdb backtrace?


Thanks,

Nicolas


-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.  Get Certified Today
Register for a JBoss Training Course.  Free Certification Exam
for All Training Attendees Through End of 2005. For more info visit:
http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click

Re: Bug in the generation of IEC958 subframes

[Takashi Iwai]

At Thu, 17 Nov 2005 20:47:37 +0100,
Nicolas Boullis wrote:
> 
> > > Moreover, (using debian's 1.0.9-3 package, I haven't checked with 
> > > 1.0.10rc3), I get segfaults whenever I try to specify the preambles in 
> > > my ~/.asoundrc file.
> > 
> > Hmm, the segfault is more serious.  Could you get a backtrace via gdb?
> 
> I submitted some test .asoundrc file + a backtrace I got with valgring 
> in bug #0001571. Is it sufficient or should I also get a gdb backtrace?

Yes, it's preferred.  We may be able to see more details.


thanks,

Takashi


-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.  Get Certified Today
Register for a JBoss Training Course.  Free Certification Exam
for All Training Attendees Through End of 2005. For more info visit:
http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click

Re: Bug in the generation of IEC958 subframes

[Nicolas Boullis]

On Fri, Nov 18, 2005 at 08:58:03PM +0100, Takashi Iwai wrote:
> 
> Yes, it's preferred.  We may be able to see more details.

OK, I just submitted the gdb backtrace to bug #0001571.
(To be honnest, the valgring backtrace looks much more informative than 
the gdb backtrace, as far as I can see...)


Cheers,

Nicolas


-------------------------------------------------------
This SF.Net email is sponsored by the JBoss Inc.  Get Certified Today
Register for a JBoss Training Course.  Free Certification Exam
for All Training Attendees Through End of 2005. For more info visit:
http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click