* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-01-27 13:45 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-01-27 13:45 UTC (permalink / raw)
To: alsa-devel
A NOTE has been added to this issue.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To:
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: new
Distribution:
Kernel Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 01-27-2006 14:45 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
----------------------------------------------------------------------
alanmann - 01-27-06 10:10
----------------------------------------------------------------------
The previous note has been added to, but the extra didn't come to
alsa-devel list.
----------------------------------------------------------------------
perex - 01-27-06 14:45
----------------------------------------------------------------------
Ok, I selected the second way. I implemented the initialization for
ice1712.c. Could you test it and change the cs8427 code if necessary when
rate changes?
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
01-27-06 14:30 perex File Added: out.txt
01-27-06 14:44 perex File Deleted: out.txt
01-27-06 14:44 perex File Added: out.txt
01-27-06 14:45 perex Note Added: 0007801
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-02-08 18:58 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-02-08 18:58 UTC (permalink / raw)
To: alsa-devel
The following issue has been set as RELATED TO issue 0001472.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To: perex
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: resolved
Distribution:
Kernel Version:
Resolution: fixed
Fixed in Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 02-02-2006 08:57 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
Relationships ID Summary
----------------------------------------------------------------------
related to 0001806 SPDIF and Wordclock I/O is broken on De...
related to 0001472 Constant distortion noise on digital mixer
======================================================================
----------------------------------------------------------------------
alanmann - 01-26-06 23:50
----------------------------------------------------------------------
I have traced the problems to the following, if I understand correctly:
1. In ice1712.c, function snd_ice1712_pro_internal_clock_put L1858:
if ((oval & ICE1712_SPDIF_MASTER) !=
(inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER)) {
tests if spdif has changed, and updates the chip if so.
When the driver is started, the test fails, so the chip is not sent spdif
set-up data, and operates wrong. Selecting spdif writes correct data and
all is OK .
2. When set to eg 44K, selecting eg 48K results in
snd_ice1712_pro_internal_clock_put calling snd_ice1712_set_pro_rate which
ends up in cs8427.c with function snd_cs8427_iec958_pcm. This sets spdif
status & rate data then performs a reset, writing to 0x04 and deleting the
spdif set-up. As the spdif has not been set changed, back in
snd_ice1712_pro_internal_clock_put the spdif test in 1. is failed again so
the chip has become un-set as in 1.
The easiest solution to both is to remove the test L1858 shown above from
ice1712.c so that the cs8427 is written to every time master clock is
changed. I have tried this and it does work. Then the cs8427 is written
or re-written each time, but is the test there for any other reason?
Otherwise I suppose extra spdif set-up is needed during driver start-up,
AND the reset routine in cs8427.c would need changing to perhaps read the
register state and re-establish the spdif setting.
The former solution sounds preferable to me if it doesn't break anything
else.
If this were agreed, I could prepare a patch to submit.
Please add comments.
----------------------------------------------------------------------
alanmann - 01-27-06 10:10
----------------------------------------------------------------------
The previous note has been added to, but the extra didn't come to
alsa-devel list.
----------------------------------------------------------------------
perex - 01-27-06 14:46
----------------------------------------------------------------------
Ok, I selected the second way. I implemented the initialization for
ice1712.c. Could you test it and change the cs8427 code if necessary when
rate changes?
My code is in out.txt file attached to this bug.
----------------------------------------------------------------------
nostar - 01-27-06 21:48
----------------------------------------------------------------------
I have a Delta 1010LT with broken SPDIF as well. I tried manually
overriding the conditional, and applying the patch. Neither of these mods
fix my problem. This test in question, "if ((oval & ICE1712_SPDIF_MASTER)
!= (inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER))" seems okay to me.
It's only purpose is to test if the old master clock source had changed
between SPDIF In and any of the Internal clock settings. By toggling
between 48k and 44.1k as in the example given, they are both Internal
clock settings and therefore toggling the SPDIF In signal on or off as the
master clock source is not necessary.
See my reply to "ice1712 -wordclock and SPDIF" on the alsa-devel list for
a description of the behavior I am seeing
----------------------------------------------------------------------
nostar - 01-27-06 22:15
----------------------------------------------------------------------
I will add my cards behavior here, since I have more to describe that I
didn't add to the list. First, the descripion I posted:
When trying to use SPDIF In as the clock source, I have to keep toggling
between any Int clock setting and SPDIF IN over and over, getting usually
static or barely audible audio (I've got audio coming from my digital
mixer into the SPDIF In during this procedure so I know when I get the
right sync), untill eventually i'll hit it dead on and it will be synced
perfectly (or very close, with an occasional pop). This behavior is the
same with and without your test removed from the driver code. I wonder if
you might try doing this too, toggling back n forth between any Int clock
setting and SPDIF In (without jack running of course so the changes will
actually take), using an external SPDIF signal coming in and spdif out
going to somewhere audible (directly to digital monitors, receiver, etc)
to make sure your problem is really fixed? Of course you can only do this
if you have the means available.
Also, In Envy24control mixer, I see (and hear) no audio on the level
meters for SPDIF In if I am set on any of the internal clock settings.
Once I set the master clock source to SPDIF In, then the signals appear on
the level meters, and the noise/static is heard, at which time I start the
toggle procedure described above untill I get an audio.
----------------------------------------------------------------------
alanmann - 02-01-06 23:28
----------------------------------------------------------------------
Good progress
1. re out.txt -it does the job, but when routine is called at
initialisation is_spdif_master(ice) does not return the correct value (is
always 0). It is OK though as later on in startup it is repeated with the
correct value. As an alternative, to avoid ambiguity at startup, consider
'bug1785-spdif-init-1.0.11rc2.diff, which has an extra parameter in the
new function, and a different name. Neither of these patches applies to
1.0.10 & earlier due to ice1712_t changes (I have an altered patch
available).
2. Neither of the above makes any difference to the end result without
changes to cs8427.c snd_cs8427_reset, since in L319 it sets the AES3 bit
(always) sourcing the clock from AES3 receiver, during internal clock freq
changes. The line may be unneccesary altogether, but
bug1785-cs8427-reset-AES3-changes.diff tests the bit at first and sets it
back at the end.
The combination of these 2 bug1785 patches resolves this problem AFAICT.
----------------------------------------------------------------------
perex - 02-02-06 08:57
----------------------------------------------------------------------
Thanks. Both patches were applied to CVS. Thank you for your work.
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
01-27-06 14:30 perex File Added: out.txt
01-27-06 14:44 perex File Deleted: out.txt
01-27-06 14:44 perex File Added: out.txt
01-27-06 14:45 perex Note Added: 0007801
01-27-06 14:46 perex Note Edited: 0007801
01-27-06 21:48 nostar Note Added: 0007806
01-27-06 22:00 nostar Issue Monitored: nostar
01-27-06 22:15 nostar Note Added: 0007807
02-01-06 23:28 alanmann Note Added: 0007864
02-01-06 23:29 alanmann File Added: bug1785-spdif-init-1.0.11rc2.diff
02-01-06 23:30 alanmann File Added: bug1785-cs8427-reset-AES3-changes.diff
02-02-06 08:57 perex Status new => resolved
02-02-06 08:57 perex Resolution open => fixed
02-02-06 08:57 perex Assigned To => perex
02-02-06 08:57 perex Note Added: 0007869
02-02-06 08:59 perex Relationship added related to 0001806
02-08-06 19:58 perex Relationship added related to 0001472
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-02-02 7:59 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-02-02 7:59 UTC (permalink / raw)
To: alsa-devel
The following issue has been set as RELATED TO issue 0001806.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To: perex
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: resolved
Distribution:
Kernel Version:
Resolution: fixed
Fixed in Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 02-02-2006 08:57 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
Relationships ID Summary
----------------------------------------------------------------------
related to 0001806 SPDIF and Wordclock I/O is broken on De...
======================================================================
----------------------------------------------------------------------
alanmann - 01-26-06 23:50
----------------------------------------------------------------------
I have traced the problems to the following, if I understand correctly:
1. In ice1712.c, function snd_ice1712_pro_internal_clock_put L1858:
if ((oval & ICE1712_SPDIF_MASTER) !=
(inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER)) {
tests if spdif has changed, and updates the chip if so.
When the driver is started, the test fails, so the chip is not sent spdif
set-up data, and operates wrong. Selecting spdif writes correct data and
all is OK .
2. When set to eg 44K, selecting eg 48K results in
snd_ice1712_pro_internal_clock_put calling snd_ice1712_set_pro_rate which
ends up in cs8427.c with function snd_cs8427_iec958_pcm. This sets spdif
status & rate data then performs a reset, writing to 0x04 and deleting the
spdif set-up. As the spdif has not been set changed, back in
snd_ice1712_pro_internal_clock_put the spdif test in 1. is failed again so
the chip has become un-set as in 1.
The easiest solution to both is to remove the test L1858 shown above from
ice1712.c so that the cs8427 is written to every time master clock is
changed. I have tried this and it does work. Then the cs8427 is written
or re-written each time, but is the test there for any other reason?
Otherwise I suppose extra spdif set-up is needed during driver start-up,
AND the reset routine in cs8427.c would need changing to perhaps read the
register state and re-establish the spdif setting.
The former solution sounds preferable to me if it doesn't break anything
else.
If this were agreed, I could prepare a patch to submit.
Please add comments.
----------------------------------------------------------------------
alanmann - 01-27-06 10:10
----------------------------------------------------------------------
The previous note has been added to, but the extra didn't come to
alsa-devel list.
----------------------------------------------------------------------
perex - 01-27-06 14:46
----------------------------------------------------------------------
Ok, I selected the second way. I implemented the initialization for
ice1712.c. Could you test it and change the cs8427 code if necessary when
rate changes?
My code is in out.txt file attached to this bug.
----------------------------------------------------------------------
nostar - 01-27-06 21:48
----------------------------------------------------------------------
I have a Delta 1010LT with broken SPDIF as well. I tried manually
overriding the conditional, and applying the patch. Neither of these mods
fix my problem. This test in question, "if ((oval & ICE1712_SPDIF_MASTER)
!= (inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER))" seems okay to me.
It's only purpose is to test if the old master clock source had changed
between SPDIF In and any of the Internal clock settings. By toggling
between 48k and 44.1k as in the example given, they are both Internal
clock settings and therefore toggling the SPDIF In signal on or off as the
master clock source is not necessary.
See my reply to "ice1712 -wordclock and SPDIF" on the alsa-devel list for
a description of the behavior I am seeing
----------------------------------------------------------------------
nostar - 01-27-06 22:15
----------------------------------------------------------------------
I will add my cards behavior here, since I have more to describe that I
didn't add to the list. First, the descripion I posted:
When trying to use SPDIF In as the clock source, I have to keep toggling
between any Int clock setting and SPDIF IN over and over, getting usually
static or barely audible audio (I've got audio coming from my digital
mixer into the SPDIF In during this procedure so I know when I get the
right sync), untill eventually i'll hit it dead on and it will be synced
perfectly (or very close, with an occasional pop). This behavior is the
same with and without your test removed from the driver code. I wonder if
you might try doing this too, toggling back n forth between any Int clock
setting and SPDIF In (without jack running of course so the changes will
actually take), using an external SPDIF signal coming in and spdif out
going to somewhere audible (directly to digital monitors, receiver, etc)
to make sure your problem is really fixed? Of course you can only do this
if you have the means available.
Also, In Envy24control mixer, I see (and hear) no audio on the level
meters for SPDIF In if I am set on any of the internal clock settings.
Once I set the master clock source to SPDIF In, then the signals appear on
the level meters, and the noise/static is heard, at which time I start the
toggle procedure described above untill I get an audio.
----------------------------------------------------------------------
alanmann - 02-01-06 23:28
----------------------------------------------------------------------
Good progress
1. re out.txt -it does the job, but when routine is called at
initialisation is_spdif_master(ice) does not return the correct value (is
always 0). It is OK though as later on in startup it is repeated with the
correct value. As an alternative, to avoid ambiguity at startup, consider
'bug1785-spdif-init-1.0.11rc2.diff, which has an extra parameter in the
new function, and a different name. Neither of these patches applies to
1.0.10 & earlier due to ice1712_t changes (I have an altered patch
available).
2. Neither of the above makes any difference to the end result without
changes to cs8427.c snd_cs8427_reset, since in L319 it sets the AES3 bit
(always) sourcing the clock from AES3 receiver, during internal clock freq
changes. The line may be unneccesary altogether, but
bug1785-cs8427-reset-AES3-changes.diff tests the bit at first and sets it
back at the end.
The combination of these 2 bug1785 patches resolves this problem AFAICT.
----------------------------------------------------------------------
perex - 02-02-06 08:57
----------------------------------------------------------------------
Thanks. Both patches were applied to CVS. Thank you for your work.
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
01-27-06 14:30 perex File Added: out.txt
01-27-06 14:44 perex File Deleted: out.txt
01-27-06 14:44 perex File Added: out.txt
01-27-06 14:45 perex Note Added: 0007801
01-27-06 14:46 perex Note Edited: 0007801
01-27-06 21:48 nostar Note Added: 0007806
01-27-06 22:00 nostar Issue Monitored: nostar
01-27-06 22:15 nostar Note Added: 0007807
02-01-06 23:28 alanmann Note Added: 0007864
02-01-06 23:29 alanmann File Added: bug1785-spdif-init-1.0.11rc2.diff
02-01-06 23:30 alanmann File Added: bug1785-cs8427-reset-AES3-changes.diff
02-02-06 08:57 perex Status new => resolved
02-02-06 08:57 perex Resolution open => fixed
02-02-06 08:57 perex Assigned To => perex
02-02-06 08:57 perex Note Added: 0007869
02-02-06 08:59 perex Relationship added related to 0001806
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-02-02 7:57 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-02-02 7:57 UTC (permalink / raw)
To: alsa-devel
The following issue has been RESOLVED.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To: perex
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: resolved
Distribution:
Kernel Version:
Resolution: fixed
Fixed in Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 02-02-2006 08:57 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
----------------------------------------------------------------------
alanmann - 02-01-06 23:28
----------------------------------------------------------------------
Good progress
1. re out.txt -it does the job, but when routine is called at
initialisation is_spdif_master(ice) does not return the correct value (is
always 0). It is OK though as later on in startup it is repeated with the
correct value. As an alternative, to avoid ambiguity at startup, consider
'bug1785-spdif-init-1.0.11rc2.diff, which has an extra parameter in the
new function, and a different name. Neither of these patches applies to
1.0.10 & earlier due to ice1712_t changes (I have an altered patch
available).
2. Neither of the above makes any difference to the end result without
changes to cs8427.c snd_cs8427_reset, since in L319 it sets the AES3 bit
(always) sourcing the clock from AES3 receiver, during internal clock freq
changes. The line may be unneccesary altogether, but
bug1785-cs8427-reset-AES3-changes.diff tests the bit at first and sets it
back at the end.
The combination of these 2 bug1785 patches resolves this problem AFAICT.
----------------------------------------------------------------------
perex - 02-02-06 08:57
----------------------------------------------------------------------
Thanks. Both patches were applied to CVS. Thank you for your work.
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
01-27-06 14:30 perex File Added: out.txt
01-27-06 14:44 perex File Deleted: out.txt
01-27-06 14:44 perex File Added: out.txt
01-27-06 14:45 perex Note Added: 0007801
01-27-06 14:46 perex Note Edited: 0007801
01-27-06 21:48 nostar Note Added: 0007806
01-27-06 22:00 nostar Issue Monitored: nostar
01-27-06 22:15 nostar Note Added: 0007807
02-01-06 23:28 alanmann Note Added: 0007864
02-01-06 23:29 alanmann File Added: bug1785-spdif-init-1.0.11rc2.diff
02-01-06 23:30 alanmann File Added: bug1785-cs8427-reset-AES3-changes.diff
02-02-06 08:57 perex Status new => resolved
02-02-06 08:57 perex Resolution open => fixed
02-02-06 08:57 perex Assigned To => perex
02-02-06 08:57 perex Note Added: 0007869
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-02-01 22:28 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-02-01 22:28 UTC (permalink / raw)
To: alsa-devel
A NOTE has been added to this issue.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To:
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: new
Distribution:
Kernel Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 02-01-2006 23:28 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
----------------------------------------------------------------------
nostar - 01-27-06 22:15
----------------------------------------------------------------------
I will add my cards behavior here, since I have more to describe that I
didn't add to the list. First, the descripion I posted:
When trying to use SPDIF In as the clock source, I have to keep toggling
between any Int clock setting and SPDIF IN over and over, getting usually
static or barely audible audio (I've got audio coming from my digital
mixer into the SPDIF In during this procedure so I know when I get the
right sync), untill eventually i'll hit it dead on and it will be synced
perfectly (or very close, with an occasional pop). This behavior is the
same with and without your test removed from the driver code. I wonder if
you might try doing this too, toggling back n forth between any Int clock
setting and SPDIF In (without jack running of course so the changes will
actually take), using an external SPDIF signal coming in and spdif out
going to somewhere audible (directly to digital monitors, receiver, etc)
to make sure your problem is really fixed? Of course you can only do this
if you have the means available.
Also, In Envy24control mixer, I see (and hear) no audio on the level
meters for SPDIF In if I am set on any of the internal clock settings.
Once I set the master clock source to SPDIF In, then the signals appear on
the level meters, and the noise/static is heard, at which time I start the
toggle procedure described above untill I get an audio.
----------------------------------------------------------------------
alanmann - 02-01-06 23:28
----------------------------------------------------------------------
Good progress
1. re out.txt -it does the job, but when routine is called at
initialisation is_spdif_master(ice) does not return the correct value (is
always 0). It is OK though as later on in startup it is repeated with the
correct value. As an alternative, to avoid ambiguity at startup, consider
'bug1785-spdif-init-1.0.11rc2.diff, which has an extra parameter in the
new function, and a different name. Neither of these patches applies to
1.0.10 & earlier due to ice1712_t changes (I have an altered patch
available).
2. Neither of the above makes any difference to the end result without
changes to cs8427.c snd_cs8427_reset, since in L319 it sets the AES3 bit
(always) sourcing the clock from AES3 receiver, during internal clock freq
changes. The line may be unneccesary altogether, but
bug1785-cs8427-reset-AES3-changes.diff tests the bit at first and sets it
back at the end.
The combination of these 2 bug1785 patches resolves this problem AFAICT.
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
01-27-06 14:30 perex File Added: out.txt
01-27-06 14:44 perex File Deleted: out.txt
01-27-06 14:44 perex File Added: out.txt
01-27-06 14:45 perex Note Added: 0007801
01-27-06 14:46 perex Note Edited: 0007801
01-27-06 21:48 nostar Note Added: 0007806
01-27-06 22:00 nostar Issue Monitored: nostar
01-27-06 22:15 nostar Note Added: 0007807
02-01-06 23:28 alanmann Note Added: 0007864
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-01-27 21:15 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-01-27 21:15 UTC (permalink / raw)
To: alsa-devel
A NOTE has been added to this issue.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To:
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: new
Distribution:
Kernel Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 01-27-2006 22:15 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
----------------------------------------------------------------------
nostar - 01-27-06 21:48
----------------------------------------------------------------------
I have a Delta 1010LT with broken SPDIF as well. I tried manually
overriding the conditional, and applying the patch. Neither of these mods
fix my problem. This test in question, "if ((oval & ICE1712_SPDIF_MASTER)
!= (inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER))" seems okay to me.
It's only purpose is to test if the old master clock source had changed
between SPDIF In and any of the Internal clock settings. By toggling
between 48k and 44.1k as in the example given, they are both Internal
clock settings and therefore toggling the SPDIF In signal on or off as the
master clock source is not necessary.
See my reply to "ice1712 -wordclock and SPDIF" on the alsa-devel list for
a description of the behavior I am seeing
----------------------------------------------------------------------
nostar - 01-27-06 22:15
----------------------------------------------------------------------
I will add my cards behavior here, since I have more to describe that I
didn't add to the list. First, the descripion I posted:
When trying to use SPDIF In as the clock source, I have to keep toggling
between any Int clock setting and SPDIF IN over and over, getting usually
static or barely audible audio (I've got audio coming from my digital
mixer into the SPDIF In during this procedure so I know when I get the
right sync), untill eventually i'll hit it dead on and it will be synced
perfectly (or very close, with an occasional pop). This behavior is the
same with and without your test removed from the driver code. I wonder if
you might try doing this too, toggling back n forth between any Int clock
setting and SPDIF In (without jack running of course so the changes will
actually take), using an external SPDIF signal coming in and spdif out
going to somewhere audible (directly to digital monitors, receiver, etc)
to make sure your problem is really fixed? Of course you can only do this
if you have the means available.
Also, In Envy24control mixer, I see (and hear) no audio on the level
meters for SPDIF In if I am set on any of the internal clock settings.
Once I set the master clock source to SPDIF In, then the signals appear on
the level meters, and the noise/static is heard, at which time I start the
toggle procedure described above untill I get an audio.
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
01-27-06 14:30 perex File Added: out.txt
01-27-06 14:44 perex File Deleted: out.txt
01-27-06 14:44 perex File Added: out.txt
01-27-06 14:45 perex Note Added: 0007801
01-27-06 14:46 perex Note Edited: 0007801
01-27-06 21:48 nostar Note Added: 0007806
01-27-06 22:00 nostar Issue Monitored: nostar
01-27-06 22:15 nostar Note Added: 0007807
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-01-27 20:48 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-01-27 20:48 UTC (permalink / raw)
To: alsa-devel
A NOTE has been added to this issue.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To:
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: new
Distribution:
Kernel Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 01-27-2006 21:48 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
----------------------------------------------------------------------
perex - 01-27-06 14:46
----------------------------------------------------------------------
Ok, I selected the second way. I implemented the initialization for
ice1712.c. Could you test it and change the cs8427 code if necessary when
rate changes?
My code is in out.txt file attached to this bug.
----------------------------------------------------------------------
nostar - 01-27-06 21:48
----------------------------------------------------------------------
I have a Delta 1010LT with broken SPDIF as well. I tried manually
overriding the conditional, and applying the patch. Neither of these mods
fix my problem. This test in question, "if ((oval & ICE1712_SPDIF_MASTER)
!= (inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER))" seems okay to me.
It's only purpose is to test if the old master clock source had changed
between SPDIF In and any of the Internal clock settings. By toggling
between 48k and 44.1k as in the example given, they are both Internal
clock settings and therefore toggling the SPDIF In signal on or off as the
master clock source is not necessary.
See my reply to "ice1712 -wordclock and SPDIF" on the alsa-devel list for
a description of the behavior I am seeing
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
01-27-06 14:30 perex File Added: out.txt
01-27-06 14:44 perex File Deleted: out.txt
01-27-06 14:44 perex File Added: out.txt
01-27-06 14:45 perex Note Added: 0007801
01-27-06 14:46 perex Note Edited: 0007801
01-27-06 21:48 nostar Note Added: 0007806
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-01-27 9:10 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-01-27 9:10 UTC (permalink / raw)
To: alsa-devel
A NOTE has been added to this issue.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To:
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: new
Distribution:
Kernel Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 01-27-2006 10:10 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
----------------------------------------------------------------------
alanmann - 01-26-06 23:50
----------------------------------------------------------------------
I have traced the problems to the following, if I understand correctly:
1. In ice1712.c, function snd_ice1712_pro_internal_clock_put L1858:
if ((oval & ICE1712_SPDIF_MASTER) !=
(inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER)) {
tests if spdif has changed, and updates the chip if so.
When the driver is started, the test fails, so the chip is not sent spdif
set-up data, and operates wrong. Selecting spdif writes correct data and
all is OK .
2. When set to eg 44K, selecting eg 48K results in
snd_ice1712_pro_internal_clock_put calling snd_ice1712_set_pro_rate which
ends up in cs8427.c with function snd_cs8427_iec958_pcm. This sets spdif
status & rate data then performs a reset, writing to 0x04 and deleting the
spdif set-up. As the spdif has not been set changed, back in
snd_ice1712_pro_internal_clock_put the spdif test in 1. is failed again so
the chip has become un-set as in 1.
The easiest solution to both is to remove the test L1858 shown above from
ice1712.c so that the cs8427 is written to every time master clock is
changed. I have tried this and it does work. Then the cs8427 is written
or re-written each time, but is the test there for any other reason?
Otherwise I suppose extra spdif set-up is needed during driver start-up,
AND the reset routine in cs8427.c would need changing to perhaps read the
register state and re-establish the spdif setting.
The former solution sounds preferable to me if it doesn't break anything
else.
If this were agreed, I could prepare a patch to submit.
Please add comments.
----------------------------------------------------------------------
alanmann - 01-27-06 10:10
----------------------------------------------------------------------
The previous note has been added to, but the extra didn't come to
alsa-devel list.
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
01-26-06 23:50 alanmann Note Edited: 0007795
01-27-06 10:10 alanmann Note Added: 0007797
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-01-26 22:41 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-01-26 22:41 UTC (permalink / raw)
To: alsa-devel
A NOTE has been added to this issue.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To:
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: new
Distribution:
Kernel Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 01-26-2006 23:41 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
----------------------------------------------------------------------
alanmann - 01-26-06 23:41
----------------------------------------------------------------------
I have traced the problems to the following, if I understnd correctly:
1. In ice1712.c, function snd_ice1712_pro_internal_clock_put L1858:
if ((oval & ICE1712_SPDIF_MASTER) !=
(inb(ICEMT(ice, RATE)) & ICE1712_SPDIF_MASTER)) {
tests if spdif has changed, and updates the chip if so.
When the driver is started, the test fails, so the chip is not sent spdif
set-up data, and operates wrong. Selecting spdif writes correct data and
all is OK .
2. When set to eg 44K, selecting eg 48K results in
snd_ice1712_pro_internal_clock_put calling snd_ice1712_set_pro_rate which
ends up in cs8427.c with function snd_cs8427_iec958_pcm. This sets spdif
status & rate data then performs a reset, writing to 0x04 and deleting the
spdif set-up. As the spdif has not been set changed, back in
snd_ice1712_pro_internal_clock_put the spdif test in 1. is failed again so
the chip has become un-set as in 1.
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
01-26-06 23:41 alanmann Note Added: 0007795
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
* [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly
@ 2006-01-26 22:27 bugtrack
0 siblings, 0 replies; 10+ messages in thread
From: bugtrack @ 2006-01-26 22:27 UTC (permalink / raw)
To: alsa-devel
The following issue has been SUBMITTED.
======================================================================
<https://bugtrack.alsa-project.org/alsa-bug/view.php?id=1785>
======================================================================
Reported By: alanmann
Assigned To:
======================================================================
Project: ALSA - driver
Issue ID: 1785
Category: PCI - ice1712
Reproducibility: always
Severity: minor
Priority: normal
Status: new
Distribution:
Kernel Version:
======================================================================
Date Submitted: 01-26-2006 23:27 CET
Last Modified: 01-26-2006 23:27 CET
======================================================================
Summary: ice1712 driver -cs8427 SPDIF chip not always set up
correctly
Description:
There are regular occasions when SPDIF IN operates spuriously, on Terratec
DMX6fire (& presumably others, like delta1010LT?).
At start-up or when changing between non-spdif master clock settings,
Digital Input stream has incorrect signal present, real enough to be
recorded with audacity. The cause is errors in writing to the cs8427
spdif chip. Will add notes with my findings.
======================================================================
Issue History
Date Modified Username Field Change
======================================================================
01-26-06 23:27 alanmann New Issue
======================================================================
-------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc. Do you grep through log files
for problems? Stop! Download the new AJAX search engine that makes
searching your log files as easy as surfing the web. DOWNLOAD SPLUNK!
http://sel.as-us.falkag.net/sel?cmd=lnk&kid=103432&bid=230486&dat=121642
^ permalink raw reply [flat|nested] 10+ messages in thread
end of thread, other threads:[~2006-02-08 18:58 UTC | newest]
Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2006-01-27 13:45 [ALSA - driver 0001785]: ice1712 driver -cs8427 SPDIF chip not always set up correctly bugtrack
-- strict thread matches above, loose matches on Subject: below --
2006-02-08 18:58 bugtrack
2006-02-02 7:59 bugtrack
2006-02-02 7:57 bugtrack
2006-02-01 22:28 bugtrack
2006-01-27 21:15 bugtrack
2006-01-27 20:48 bugtrack
2006-01-27 9:10 bugtrack
2006-01-26 22:41 bugtrack
2006-01-26 22:27 bugtrack
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.