AC97 bus interface for ad-hoc drivers
PATCH 1/1 AC97 bus interface for ad-hoc drivers
This patch adds support for ad-hoc AC97 device drivers (e.g WM97xx and UCB touchscreen drivers) and was originally posted to the list by Nicolas Pitre as an RFC.
Changes from RFC version :-
o Now matches codec name within codec group.
o Added ac97_dev_release() to stop kernel complaining about no release
method for device.
Liam
include/sound/ac97_codec.h +6 -0
@@ -26,6 +26,7 @@*/ #include <linux/bitops.h> +#include <linux/device.h> #include "pcm.h" #include "control.h" #include "info.h"
@@ -520,6 +521,7 @@ struct _snd_ac97 {/* jack-sharing info */ unsigned char indep_surround;unsigned char channel_mode; +struct device dev; };/* conditions */
@@ -599,4 +601,8 @@ struct ac97_enum {unsigned short mask; const char **texts; }; + +/* ad hoc AC97 device driver access */ +extern struct bus_type ac97_bus_type; +#endif /* __SOUND_AC97_CODEC_H */
sound/pci/ac97/Makefile +1 -1
@@ -12,7 +12,7 @@ endif snd-ak4531-codec-objs := ak4531_codec.o # Toplevel Module Dependency -obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o +obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o ac97_bus.o obj-$(CONFIG_SND_ENS1370) += snd-ak4531-codec.o obj-m := $(sort $(obj-m))
sound/pci/ac97/ac97_bus.c new file +83 -0
83 lines not on this page · raw message
sound/pci/ac97/ac97_codec.c +15 -0
@@ -992,6 +992,8 @@ static int snd_ac97_free(ac97_t *ac97) {if (ac97) {snd_ac97_proc_done (ac97); +if (ac97->dev.bus) +device_unregister (&ac97->dev);if (ac97->bus) { ac97->bus->codec[ac97->num] =NULL ;if (ac97->bus->shared_type) {
@@ -1007,6 +1009,11 @@ static int snd_ac97_free(ac97_t *ac97)return 0 ; } +/* Stops no dev release warning */ +static void ac97_dev_release (struct device * dev) +{ +} +static int snd_ac97_dev_free (snd_device_t *device) {ac97_t *ac97 = device->device_data;
@@ -2117,6 +2124,14 @@ int snd_ac97_mixer(ac97_bus_t *bus, ac97snd_ac97_free (ac97);return err; } + + ac97->dev.bus = &ac97_bus_type; + ac97->dev.parent = ac97->bus->card->dev; + ac97->dev.platform_data = ac97; + ac97->dev.release = ac97_dev_release; +strncpy (ac97->dev.bus_id,snd_ac97_get_short_name (ac97),BUS_ID_SIZE ); +if ((err =device_register (&ac97->dev)) <0 ) +return err; *rac97 = ac97;if (bus->shared_type) {
At Tue, 19 Jul 2005 13:20:24 +0100,
Quoted 10 lines · Liam Girdwood
This patch adds support for ad-hoc AC97 device drivers (e.g WM97xx and UCB touchscreen drivers) and was originally posted to the list by Nicolas Pitre as an RFC.
Changes from RFC version :-
o Now matches codec name within codec group.
o Addedac97_dev_release()to stop kernel complaining about no release
method for device.
Thanks for the patch. The idea is fine, and the code looks almost OK except for the below.
Could you describe the full changelog (or paste Nicolas' RFC) ?
sound/pci/ac97/Makefile endif · line 15
Quoted 9 lines
sound/pci/ac97/Makefile +1
-1@@ -12,7 +12,7 @@ endif snd-ak4531-codec-objs := ak4531_codec.o # Toplevel Module Dependency -obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o +obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o ac97_bus.o
Better to keep snd- prefix for ac97_bus module (although it can be used independently from ALSA)...? It's an open question.
Quoted 4 lines
sound/pci/ac97/ac97_bus.c new file +0
-0
(snip)
sound/pci/ac97/ac97_bus.c line 21
Quoted 8 lines
sound/pci/ac97/ac97_bus.c new file +8
-0+#include <linux/module.h> +#include <linux/init.h> +#include <linux/device.h> +#include <linux/string.h> + +#include <sound/driver.h> +#include <sound/core.h> +#include <sound/ac97_codec.h>
Don't include unnecessary header files.
Quoted 3 lines
sound/pci/ac97/ac97_codec.c +0
-0
(snip)
sound/pci/ac97/ac97_codec.c int snd_ac97_mixer(ac97_bus_t *bus, ac97 · line 2134
Quoted 12 lines
sound/pci/ac97/ac97_codec.c +8
-0@@ -2117,6 +2124,14 @@ int snd_ac97_mixer(ac97_bus_t *bus, ac97snd_ac97_free (ac97);return err; } + + ac97->dev.bus = &ac97_bus_type; + ac97->dev.parent = ac97->bus->card->dev; + ac97->dev.platform_data = ac97; + ac97->dev.release = ac97_dev_release; +strncpy (ac97->dev.bus_id,snd_ac97_get_short_name (ac97),BUS_ID_SIZE ); +if ((err =device_register (&ac97->dev)) <0 ) +return err;
Need to call snd_ac97_free() in the error path (but reset dev.bus to NULL to avoid unregister).
Takashi
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
Quoted 25 lines · Takashi Iwai
At Tue, 19 Jul 2005 13:20:24 +0100,
Quoted 10 lines · Liam Girdwood
This patch adds support for ad-hoc AC97 device drivers (e.g WM97xx and UCB touchscreen drivers) and was originally posted to the list by Nicolas Pitre as an RFC.
Changes from RFC version :-
o Now matches codec name within codec group.
o Addedac97_dev_release()to stop kernel complaining about no release
method for device.Thanks for the patch. The idea is fine, and the code looks almost OK except for the below.
Could you describe the full changelog (or paste Nicolas' RFC) ?
Quoted 6 lines
sound/pci/ac97/Makefile +0
-0@@ -12,7 +12,7 @@ endif snd-ak4531-codec-objs := ak4531_codec.o
Quoted 6 lines
sound/pci/ac97/Makefile endif · line 15
Quoted 3 lines
sound/pci/ac97/Makefile +1
-1# Toplevel Module Dependency -obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o +obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o ac97_bus.oBetter to keep snd- prefix for
ac97_busmodule (although it can be used independently from ALSA)...? It's an open question.
The idea is to be able to use ALSA as modules and the extra function driver linked in, or vice versa. Therefore ac97_bus.o should probably be selected with its own config symbol allowing for things like:
config UCB1400_TS
tristate "UCB1400 touchscreen interface"
select AC97_BUS
And similarly:
config SND_AC97_CODEC
... select AC97_BUS
This way there is no hard linkage depedency between ALSA and any ad-hoc codec function driver.
Nicolas
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
At Wed, 27 Jul 2005 09:05:33 -0400 (EDT),
Quoted 44 lines · Nicolas Pitre
Quoted 25 lines · Takashi Iwai
At Tue, 19 Jul 2005 13:20:24 +0100,
Quoted 10 lines · Liam Girdwood
This patch adds support for ad-hoc AC97 device drivers (e.g WM97xx and UCB touchscreen drivers) and was originally posted to the list by Nicolas Pitre as an RFC.
Changes from RFC version :-
o Now matches codec name within codec group.
o Addedac97_dev_release()to stop kernel complaining about no release
method for device.Thanks for the patch. The idea is fine, and the code looks almost OK except for the below.
Could you describe the full changelog (or paste Nicolas' RFC) ?
Quoted 6 lines
sound/pci/ac97/Makefile +0
-0@@ -12,7 +12,7 @@ endif snd-ak4531-codec-objs := ak4531_codec.oQuoted 6 lines
sound/pci/ac97/Makefile endif · line 15
Quoted 3 lines
sound/pci/ac97/Makefile +1
-1# Toplevel Module Dependency -obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o +obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o ac97_bus.oBetter to keep snd- prefix for
ac97_busmodule (although it can be used independently from ALSA)...? It's an open question.The idea is to be able to use ALSA as modules and the extra function driver linked in, or vice versa. Therefore ac97_bus.o should probably be selected with its own config symbol allowing for things like:
config
UCB1400_TS
tristate "UCB1400 touchscreen interface"
selectAC97_BUS
Quoted 8 lines
And similarly:
config
SND_AC97_CODEC... select AC97_BUSThis way there is no hard linkage depedency between ALSA and any ad-hoc codec function driver.
Yes, I understand the purpose. My concern was whether we keep a consistent module name rule for the modules in sound directory, or simply use as it is. I have no special preference here, and would like to hear from others.
Takashi
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
Quoted 52 lines · Takashi Iwai
At Wed, 27 Jul 2005 09:05:33 -0400 (EDT),
Quoted 50 lines · Nicolas Pitre
Quoted 25 lines · Takashi Iwai
At Tue, 19 Jul 2005 13:20:24 +0100,
Quoted 10 lines · Liam Girdwood
This patch adds support for ad-hoc AC97 device drivers (e.g WM97xx and UCB touchscreen drivers) and was originally posted to the list by Nicolas Pitre as an RFC.
Changes from RFC version :-
o Now matches codec name within codec group.
o Addedac97_dev_release()to stop kernel complaining about no release
method for device.Thanks for the patch. The idea is fine, and the code looks almost OK except for the below.
Could you describe the full changelog (or paste Nicolas' RFC) ?
Quoted 6 lines
sound/pci/ac97/Makefile +0
-0@@ -12,7 +12,7 @@ endif snd-ak4531-codec-objs := ak4531_codec.oQuoted 6 lines
sound/pci/ac97/Makefile endif · line 15
Quoted 3 lines
sound/pci/ac97/Makefile +1
-1# Toplevel Module Dependency -obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o +obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o ac97_bus.oBetter to keep snd- prefix for
ac97_busmodule (although it can be used independently from ALSA)...? It's an open question.The idea is to be able to use ALSA as modules and the extra function driver linked in, or vice versa. Therefore ac97_bus.o should probably be selected with its own config symbol allowing for things like:
config
UCB1400_TS
tristate "UCB1400 touchscreen interface"
selectAC97_BUSAnd similarly:
config
SND_AC97_CODEC... select AC97_BUS
Quoted 8 lines
Quoted 2 lines
This way there is no hard linkage depedency between ALSA and any ad-hoc codec function driver.
Yes, I understand the purpose. My concern was whether we keep a consistent module name rule for the modules in sound directory, or simply use as it is. I have no special preference here, and would like to hear from others.
We should probably stick with the snd- naming convention because the ad-hoc bus module does require the sound drivers to be loaded before any ad-hoc driver will work. This does not mean the sound modules or bus module needs to be loaded before the ad-hoc driver, it just means the ad-hoc driver has a functional dependency on the others.
Liam
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click
Quoted 14 lines · Nicolas Pitre
Quoted 3 lines · Takashi Iwai
Better to keep snd- prefix for
ac97_busmodule (although it can be used independently from ALSA)...? It's an open question.The idea is to be able to use ALSA as modules and the extra function driver linked in, or vice versa. Therefore ac97_bus.o should probably be selected with its own config symbol allowing for things like:
config
UCB1400_TS
tristate "UCB1400 touchscreen interface"
selectAC97_BUS
Quoted 8 lines
And similarly:
config
SND_AC97_CODEC... select AC97_BUSThis way there is no hard linkage depedency between ALSA and any ad-hoc codec function driver.
I've just unsuccessfully tried to create this dependency in the Kconfig system. It looks like you cannot do:-
config SND_SOME_DRIVER
tristate "some driver"
select SND_AC97_CODEC
config SND_AC97_CODEC
tristate
select SND_PCM
select SND_AC97_BUS
config SND_AC97_BUS
tristate
In this scenario SND_SOME_DRIVER selects SND_AC97_CODEC selects SND_AC97_BUS and I could never get a SND_AC97_BUS=m or otherwise in my .config. I was only ever successful by adding the select SND_AC97_BUS line to SND_SOME_DRIVER.
I now don't think that the "select SND_PCM" is having any effect in the AC97_CODEC config, although it still builds fine when remove it.
With this in mind (I'm not sure if this is intentional in the build system or not), it might be better to stick with the original Makefile and rename the module with a snd- prefix.
Liam
-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
At Thu, 28 Jul 2005 21:26:39 +0100,
Quoted 37 lines · Liam Girdwood
Quoted 14 lines · Nicolas Pitre
Quoted 3 lines · Takashi Iwai
Better to keep snd- prefix for
ac97_busmodule (although it can be used independently from ALSA)...? It's an open question.The idea is to be able to use ALSA as modules and the extra function driver linked in, or vice versa. Therefore ac97_bus.o should probably be selected with its own config symbol allowing for things like:
config
UCB1400_TS
tristate "UCB1400 touchscreen interface"
selectAC97_BUSQuoted 8 lines
And similarly:
config
SND_AC97_CODEC... select AC97_BUSThis way there is no hard linkage depedency between ALSA and any ad-hoc codec function driver.
I've just unsuccessfully tried to create this dependency in the
Kconfigsystem. It looks like you cannot do:-config
SND_SOME_DRIVER
tristate "some driver"
selectSND_AC97_CODECconfig
SND_AC97_CODEC
tristate
selectSND_PCM
selectSND_AC97_BUS
Quoted 7 lines
config
SND_AC97_BUS
tristateIn this scenario
SND_SOME_DRIVERselectsSND_AC97_CODECselectsSND_AC97_BUSand I could never get a SND_AC97_BUS=m or otherwise in my .config. I was only ever successful by adding the selectSND_AC97_BUSline toSND_SOME_DRIVER.
Is it a PCI driver? Since SND_AC97_CODEC and SND_AC97_BUS is in pci/Kconfig, it might the case that it's be selected only when PCI!=n.
Quoted 2 lines
I now don't think that the "select
SND_PCM" is having any effect in theAC97_CODECconfig, although it still builds fine when remove it.
Hmm, we didn't have problems with other similar cases...
Takashi
-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
Quoted 4 lines · Takashi Iwai
At Thu, 28 Jul 2005 21:26:39 +0100,
Quoted 1 line · Liam Girdwood
Quoted 10 lines
Quoted 4 lines
In this scenario
SND_SOME_DRIVERselectsSND_AC97_CODECselectsSND_AC97_BUSand I could never get a SND_AC97_BUS=m or otherwise in my .config. I was only ever successful by adding the selectSND_AC97_BUSline toSND_SOME_DRIVER.Is it a PCI driver? Since
SND_AC97_CODECandSND_AC97_BUSis in pci/Kconfig, it might the case that it's be selected only when PCI!=n.
It's an Arm (PXA27x) based driver, hence PCI=n.
You are correct. I had unsuccessfully tried the config SND_AC97_BUS out with the PCI=n clause. It only works when _both_ SND_AC97_CODEC and SND_AC97_BUS are out with PCI=n.
If nobody objects, I can resubmit with the SND_AC97_CODEC out with the PCI=n in sound/pci/Kconfig. Will this cause any problems for other cards ?
Quoted 4 lines
Quoted 2 lines
I now don't think that the "select
SND_PCM" is having any effect in theAC97_CODECconfig, although it still builds fine when remove it.Hmm, we didn't have problems with other similar cases...
It was being included by the other Arm drivers.
Liam
-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
At Fri, 29 Jul 2005 12:20:40 +0100,
Quoted 19 lines · Liam Girdwood
Quoted 4 lines · Takashi Iwai
At Thu, 28 Jul 2005 21:26:39 +0100,
Quoted 1 line · Liam Girdwood
Quoted 10 lines
Quoted 4 lines
In this scenario
SND_SOME_DRIVERselectsSND_AC97_CODECselectsSND_AC97_BUSand I could never get a SND_AC97_BUS=m or otherwise in my .config. I was only ever successful by adding the selectSND_AC97_BUSline toSND_SOME_DRIVER.Is it a PCI driver? Since
SND_AC97_CODECandSND_AC97_BUSis in pci/Kconfig, it might the case that it's be selected only when PCI!=n.It's an Arm (PXA27x) based driver, hence PCI=n.
Quoted 7 lines
You are correct. I had unsuccessfully tried the config
SND_AC97_BUSout with the PCI=n clause. It only works when _both_SND_AC97_CODECandSND_AC97_BUSare out with PCI=n.If nobody objects, I can resubmit with the
SND_AC97_CODECout with the PCI=n insound/pci/Kconfig. Will this cause any problems for other cards ?
Please post it. That change should be harmless.
Takashi
-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
I've made the review changes and as requested I've pasted the RFC by Nicolas below:-
"I would like to know what people think of the following patch. It allows for a codec on an AC97 bus to be shared with other drivers which are completely unrelated to audio. It registers a new bus type, and whenever a codec instance is created then a device for it is also registered with the driver model using that bus type. This allows, for example, to use the extra features of the UCB1400 like the touchscreen interface and the additional GPIOs and ADCs available on that chip for battery monitoring. I have a working UCB1400 touchscreen driver here that simply registers with the driver model happily working alongside with audio features using this."
Changes over RFC:-
o Now matches codec name within codec group.
o Added ac97_dev_release() to stop kernel complaining about no release
method for device.
o Added "config SND_AC97_BUS" to sound/pci/Kconfig and moved "configSND_AC97_CODEC" out with the PCI=n statement.
o module is now called snd-ac97-bus
Liam
sound/pci/Kconfig +7 -3
@@ -1,11 +1,15 @@ # ALSA PCI drivers -menu "PCI devices" - depends on SND!=n && PCI - config SND_AC97_CODEC tristate select SND_PCM + select SND_AC97_BUS + +config SND_AC97_BUS + tristate + +menu "PCI devices" + depends on SND!=n && PCI config SND_ALI5451 tristate "ALi M5451 PCI Audio Controller"
sound/pci/ac97/ac97_bus.c new file +79 -0
79 lines not on this page · raw message
sound/pci/ac97/ac97_codec.c +18 -0
@@ -992,6 +992,8 @@ static int snd_ac97_free(ac97_t *ac97) {if (ac97) {snd_ac97_proc_done (ac97); +if (ac97->dev.bus) +device_unregister (&ac97->dev);if (ac97->bus) { ac97->bus->codec[ac97->num] =NULL ;if (ac97->bus->shared_type) {
@@ -1007,6 +1009,11 @@ static int snd_ac97_free(ac97_t *ac97)return 0 ; } +/* stop no dev release warning */ +static void ac97_dev_release (struct device * dev) +{ +} +static int snd_ac97_dev_free (snd_device_t *device) {ac97_t *ac97 = device->device_data;
@@ -2117,6 +2136,17 @@ int snd_ac97_mixer(ac97_bus_t *bus, ac97snd_ac97_free (ac97);return err; } + + ac97->dev.bus = &ac97_bus_type; + ac97->dev.parent = ac97->bus->card->dev; + ac97->dev.platform_data = ac97; + ac97->dev.release = ac97_dev_release; +strncpy (ac97->dev.bus_id,snd_ac97_get_short_name (ac97),BUS_ID_SIZE ); +if ((err =device_register (&ac97->dev)) <0 ) { + ac97->dev.bus =NULL ; +snd_ac97_free (ac97); +return err; + } *rac97 = ac97;if (bus->shared_type) {
include/sound/ac97_codec.h +6 -0
@@ -26,6 +26,7 @@*/ #include <linux/bitops.h> +#include <linux/device.h> #include "pcm.h" #include "control.h" #include "info.h"
@@ -520,6 +524,7 @@ struct _snd_ac97 {/* jack-sharing info */ unsigned char indep_surround;unsigned char channel_mode; +struct device dev; };/* conditions */
@@ -599,4 +604,8 @@ struct ac97_enum {unsigned short mask; const char **texts; }; + +/* ad hoc AC97 device driver access */ +extern struct bus_type ac97_bus_type; +#endif /* __SOUND_AC97_CODEC_H */
sound/pci/ac97/Makefile +2 -0
@@ -10,9 +10,11 @@ snd-ac97-codec-objs += ac97_proc.o endif snd-ak4531-codec-objs := ak4531_codec.o +snd-ac97-bus-objs := ac97_bus.o # Toplevel Module Dependency obj-$(CONFIG_SND_AC97_CODEC) += snd-ac97-codec.o obj-$(CONFIG_SND_ENS1370) += snd-ak4531-codec.o +obj-$(CONFIG_SND_AC97_BUS) += snd-ac97-bus.o obj-m := $(sort $(obj-m))
At Fri, 29 Jul 2005 13:51:56 +0100,
Quoted 24 lines · Liam Girdwood
I've made the review changes and as requested I've pasted the RFC by Nicolas below:-
"I would like to know what people think of the following patch. It allows for a codec on an AC97 bus to be shared with other drivers which are completely unrelated to audio. It registers a new bus type, and whenever a codec instance is created then a device for it is also registered with the driver model using that bus type. This allows, for example, to use the extra features of the UCB1400 like the touchscreen interface and the additional GPIOs and ADCs available on that chip for battery monitoring. I have a working UCB1400 touchscreen driver here that simply registers with the driver model happily working alongside with audio features using this."
Changes over RFC:-
o Now matches codec name within codec group.
o Addedac97_dev_release()to stop kernel complaining about no release
method for device.
o Added "configSND_AC97_BUS" tosound/pci/Kconfigand moved "configSND_AC97_CODEC" out with the PCI=n statement.
o module is now called snd-ac97-bus
Quoted 2 lines
Signed-off-by Liam Girdwood <[email protected]>Signed-off-by Nicolas Pitre <[email protected]>
Thanks. I applied it to CVS with some modifications.
- the registration and unregistration to ac97_bus is done in dev_register and dev_unregister callbacks. This assures that card->dev is set.
- 2.2/2.4 workaronds in alsa-driver tree.
Please check whether it's OK (later, after cvs tree is sync'ed)
Takashi
-------------------------------------------------------
SF.Net email is Sponsored by the Better Software Conference & EXPO September 19-22, 2005 * San Francisco, CA * Development Lifecycle Practices Agile & Plan-Driven Development * Managing Projects & Teams * Testing & QA Security * Process Improvement & Measurement * http://www.sqe.com/bsce5sf
Quoted 10 lines · Takashi Iwai
Thanks. I applied it to CVS with some modifications.
- the registration and unregistration to
ac97_busis done indev_registeranddev_unregistercallbacks. This assures thatcard->devis set.- 2.2/2.4 workaronds in alsa-driver tree.
Please check whether it's OK (later, after cvs tree is sync'ed)
Works as expected.
Thanks
Liam
-------------------------------------------------------
SF.Net email is sponsored by: Discover Easy Linux Migration Strategies from IBM. Find simple to follow Roadmaps, straightforward articles, informative Webcasts and more! Get everything you need to get up to speed, fast. http://ads.osdn.com/?ad_id=7477&alloc_id=16492&op=click