[alsa-devel] [RFC] virtual master control for HD-audio
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
Takashi
This patch adds the virtual master control to sound core.
Takashi
diff -r 1207d46e60b3 core/Kconfig --- a/core/Kconfig Mon Dec 03 15:15:59 2007 +0100 +++ b/core/Kconfig Mon Dec 03 15:46:17 2007 +0100 @@ -15,6 +15,9 @@ config SND_RAWMIDI config SND_RAWMIDI tristate depends on SND + +config SND_VMASTER + bool
config SND_SEQUENCER tristate "Sequencer support" diff -r 1207d46e60b3 core/Makefile --- a/core/Makefile Mon Dec 03 15:15:59 2007 +0100 +++ b/core/Makefile Mon Dec 03 15:46:17 2007 +0100 @@ -6,6 +6,7 @@ snd-y := sound.o init.o memory.o inf snd-y := sound.o init.o memory.o info.o control.o misc.o device.o snd-$(CONFIG_ISA_DMA_API) += isadma.o snd-$(CONFIG_SND_OSSEMUL) += sound_oss.o info_oss.o +snd-$(CONFIG_SND_VMASTER) += vmaster.o
snd-pcm-objs := pcm.o pcm_native.o pcm_lib.o pcm_timer.o pcm_misc.o \ pcm_memory.o diff -r 1207d46e60b3 core/vmaster.c --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/core/vmaster.c Mon Dec 03 15:46:17 2007 +0100 @@ -0,0 +1,379 @@ +/* + * Virtual master and slave controls + * + * Copyright (c) by Takashi Iwai tiwai@suse.de + * + * This program is free software; you can redistribute it and/or modify + * it under the terms of the GNU General Public License as published by + * the Free Software Foundation; either version 2 of the License, or + * (at your option) any later version. + * + * This program is distributed in the hope that it will be useful, + * but WITHOUT ANY WARRANTY; without even the implied warranty of + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + * GNU General Public License for more details. + * + * You should have received a copy of the GNU General Public License + * along with this program; if not, write to the Free Software + * Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + * + */ + +#include <sound/driver.h> +#include <linux/slab.h> +#include <sound/core.h> +#include <sound/control.h> + +/* + * a subset of information returned via ctl info callback + */ +struct link_ctl_info { + int type; /* value type */ + int count; /* item count */ + int min_val, max_val; /* min, max values */ +}; + +/* + * link master - this contains a list of slave controls that are + * identical types, i.e. info returns the same value type and value + * ranges, but may have different number of counts. + * + * The master control is so far only mono volume/switch for simplicity. + * The same value will be applied to all slaves. + */ +struct link_master { + struct list_head slaves; + struct link_ctl_info info; + int val; /* the master value */ +}; + +/* + * link slave - this contains a slave control element + * + * It fakes the control callbacsk with additional attenuation by the + * master control. A slave may have either one or two channels. + */ + +struct link_slave { + struct list_head list; + struct link_master *master; + struct link_ctl_info info; + int vals[2]; /* current values */ + struct snd_kcontrol slave; /* the copy of original control entry */ +}; + +/* get the slave ctl info and save the initial values */ +static int slave_init(struct link_slave *slave) +{ + struct snd_ctl_elem_info *uinfo; + struct snd_ctl_elem_value *uctl; + int err, ch; + + if (slave->info.count) + return 0; /* already initialized */ + + uinfo = kmalloc(sizeof(*uinfo), GFP_KERNEL); + if (!uinfo) + return -ENOMEM; + uinfo->id = slave->slave.id; + err = slave->slave.info(&slave->slave, uinfo); + if (err < 0) { + kfree(uinfo); + return err; + } + slave->info.type = uinfo->type; + slave->info.count = uinfo->count; + if (slave->info.count > 2 || + (slave->info.type != SNDRV_CTL_ELEM_TYPE_INTEGER && + slave->info.type != SNDRV_CTL_ELEM_TYPE_BOOLEAN)) { + snd_printk(KERN_ERR "invalid slave element\n"); + kfree(uinfo); + return -EINVAL; + } + slave->info.min_val = uinfo->value.integer.min; + slave->info.max_val = uinfo->value.integer.max; + kfree(uinfo); + + uctl = kmalloc(sizeof(*uctl), GFP_KERNEL); + if (!uctl) + return -ENOMEM; + uctl->id = slave->slave.id; + err = slave->slave.get(&slave->slave, uctl); + for (ch = 0; ch < slave->info.count; ch++) + slave->vals[ch] = uctl->value.integer.value[ch]; + kfree(uctl); + return 0; +} + +/* initialize master volume */ +static int master_init(struct link_master *master) +{ + struct link_slave *slave; + + if (master->info.count) + return 0; /* already initialized */ + + list_for_each_entry(slave, &master->slaves, list) { + int err = slave_init(slave); + if (err < 0) + return err; + master->info = slave->info; + master->info.count = 1; /* always mono */ + /* set full volume as default (= no attenuation) */ + master->val = master->info.max_val; + return 0; + } + return -ENOENT; +} + +static int slave_get_val(struct link_slave *slave, + struct snd_ctl_elem_value *ucontrol) +{ + int err, ch; + + err = slave_init(slave); + if (err < 0) + return err; + for (ch = 0; ch < slave->info.count; ch++) + ucontrol->value.integer.value[ch] = slave->vals[ch]; + return 0; +} + +static int slave_put_val(struct link_slave *slave, + struct snd_ctl_elem_value *ucontrol) +{ + int err, ch, vol; + + err = master_init(slave->master); + if (err < 0) + return err; + + switch (slave->info.type) { + case SNDRV_CTL_ELEM_TYPE_BOOLEAN: + for (ch = 0; ch < slave->info.count; ch++) + ucontrol->value.integer.value[ch] &= + !!slave->master->val; + break; + case SNDRV_CTL_ELEM_TYPE_INTEGER: + for (ch = 0; ch < slave->info.count; ch++) { + /* max master volume is supposed to be 0 dB */ + vol = ucontrol->value.integer.value[ch]; + vol += slave->master->val - slave->master->info.max_val; + if (vol < slave->info.min_val) + vol = slave->info.min_val; + else if (vol > slave->info.max_val) + vol = slave->info.max_val; + ucontrol->value.integer.value[ch] = vol; + } + break; + } + return slave->slave.put(&slave->slave, ucontrol); +} + +/* + * ctl callbacks for slaves + */ +static int slave_info(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_info *uinfo) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + return slave->slave.info(&slave->slave, uinfo); +} + +static int slave_get(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + return slave_get_val(slave, ucontrol); +} + +static int slave_put(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + int err, ch, changed = 0; + + err = slave_init(slave); + if (err < 0) + return err; + for (ch = 0; ch < slave->info.count; ch++) { + if (slave->vals[ch] != ucontrol->value.integer.value[ch]) { + changed = 1; + slave->vals[ch] = ucontrol->value.integer.value[ch]; + } + } + if (!changed) + return 0; + return slave_put_val(slave, ucontrol); +} + +static int slave_tlv_cmd(struct snd_kcontrol *kcontrol, + int op_flag, unsigned int size, + unsigned int __user *tlv) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + /* FIXME: this assumes that the max volume is 0 dB */ + return slave->slave.tlv.c(&slave->slave, op_flag, size, tlv); +} + +static void slave_free(struct snd_kcontrol *kcontrol) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + if (slave->slave.private_free) + slave->slave.private_free(&slave->slave); + if (slave->master) + list_del(&slave->list); + kfree(slave); +} + +/* + * Add a slave control to the group with the given master control + * + * All slaves must be the same type (returning the same information + * via info callback). The fucntion doesn't check it, so it's your + * responsibility. + * + * Also, some additional limitations: + * - at most two channels + * - logarithmic volume control (dB level), no linear volume + * - master can only attenuate the volume, no gain + */ +int snd_ctl_add_slave(struct snd_kcontrol *master, struct snd_kcontrol *slave) +{ + struct link_master *master_link = snd_kcontrol_chip(master); + struct link_slave *srec; + + srec = kzalloc(sizeof(*srec) + + slave->count * sizeof(*slave->vd), GFP_KERNEL); + if (!srec) + return -ENOMEM; + srec->slave = *slave; + memcpy(srec->slave.vd, slave->vd, slave->count * sizeof(*slave->vd)); + srec->master = master_link; + + /* override callbacks */ + slave->info = slave_info; + slave->get = slave_get; + slave->put = slave_put; + if (slave->vd[0].access & SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK) + slave->tlv.c = slave_tlv_cmd; + slave->private_data = srec; + slave->private_free = slave_free; + + list_add_tail(&srec->list, &master_link->slaves); + return 0; +} + +EXPORT_SYMBOL(snd_ctl_add_slave); + +/* + * ctl callbacks for master controls + */ +static int master_info(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_info *uinfo) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + int ret; + + ret = master_init(master); + if (ret < 0) + return ret; + uinfo->type = master->info.type; + uinfo->count = master->info.count; + uinfo->value.integer.min = master->info.min_val; + uinfo->value.integer.max = master->info.max_val; + return 0; +} + +static int master_get(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + int err = master_init(master); + if (err < 0) + return err; + ucontrol->value.integer.value[0] = master->val; + return 0; +} + +static int master_put(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + struct link_slave *slave; + struct snd_ctl_elem_value *uval; + int err, old_val; + + err = master_init(master); + if (err < 0) + return err; + old_val = master->val; + if (ucontrol->value.integer.value[0] == old_val) + return 0; + + uval = kmalloc(sizeof(*uval), GFP_KERNEL); + if (!uval) + return -ENOMEM; + list_for_each_entry(slave, &master->slaves, list) { + master->val = old_val; + uval->id = slave->slave.id; + slave_get_val(slave, uval); + master->val = ucontrol->value.integer.value[0]; + slave_put_val(slave, uval); + } + kfree(uval); + return 1; +} + +static void master_free(struct snd_kcontrol *kcontrol) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + struct link_slave *slave; + + list_for_each_entry(slave, &master->slaves, list) + slave->master = NULL; + kfree(master); +} + + +/* + * Create a virtual master control with the given name + */ +struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, + const unsigned int *tlv) +{ + struct link_master *master; + struct snd_kcontrol *kctl; + struct snd_kcontrol_new knew; + + memset(&knew, 0, sizeof(knew)); + knew.iface = SNDRV_CTL_ELEM_IFACE_MIXER; + knew.name = name; + knew.info = master_info; + + master = kzalloc(sizeof(*master), GFP_KERNEL); + if (!master) + return NULL; + INIT_LIST_HEAD(&master->slaves); + + kctl = snd_ctl_new1(&knew, master); + if (!kctl) { + kfree(master); + return NULL; + } + /* override some callbacks */ + kctl->info = master_info; + kctl->get = master_get; + kctl->put = master_put; + kctl->private_free = master_free; + + /* additional (constant) TLV read */ + if (tlv) { + /* FIXME: this assumes that the max volume is 0 dB */ + kctl->vd[0].access |= SNDRV_CTL_ELEM_ACCESS_TLV_READ; + kctl->tlv.p = tlv; + } + return kctl; +} + +EXPORT_SYMBOL(snd_ctl_make_virtual_master); diff -r 1207d46e60b3 include/control.h --- a/include/control.h Mon Dec 03 15:15:59 2007 +0100 +++ b/include/control.h Mon Dec 03 15:46:17 2007 +0100 @@ -169,4 +169,12 @@ int snd_ctl_boolean_stereo_info(struct s int snd_ctl_boolean_stereo_info(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_info *uinfo);
+/* + * virtual master control + */ +struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, + const unsigned int *tlv); +int snd_ctl_add_slave(struct snd_kcontrol *master, struct snd_kcontrol *slave); + + #endif /* __SOUND_CONTROL_H */
On Thu, Dec 20, 2007 at 04:45:17PM +0100, Takashi Iwai wrote:
This patch adds the virtual master control to sound core.
Takashi
diff -r 1207d46e60b3 core/Kconfig --- a/core/Kconfig Mon Dec 03 15:15:59 2007 +0100 +++ b/core/Kconfig Mon Dec 03 15:46:17 2007 +0100 @@ -15,6 +15,9 @@ config SND_RAWMIDI config SND_RAWMIDI tristate depends on SND
+config SND_VMASTER
- bool
config SND_SEQUENCER tristate "Sequencer support"
When you committed this to hg, you put vmaster.c into hda instead of core - could you explain why? I'd like to use it with ice1724.
At Wed, 30 Jan 2008 10:19:10 -0500, Daniel Jacobowitz wrote:
On Thu, Dec 20, 2007 at 04:45:17PM +0100, Takashi Iwai wrote:
This patch adds the virtual master control to sound core.
Takashi
diff -r 1207d46e60b3 core/Kconfig --- a/core/Kconfig Mon Dec 03 15:15:59 2007 +0100 +++ b/core/Kconfig Mon Dec 03 15:46:17 2007 +0100 @@ -15,6 +15,9 @@ config SND_RAWMIDI config SND_RAWMIDI tristate depends on SND
+config SND_VMASTER
- bool
config SND_SEQUENCER tristate "Sequencer support"
When you committed this to hg, you put vmaster.c into hda instead of core - could you explain why? I'd like to use it with ice1724.
Because HD-audio is the only user, so far. I don't want to expand the core stuff (and kconfig) unless really needed.
Takashi
On Wed, Jan 30, 2008 at 04:27:07PM +0100, Takashi Iwai wrote:
Because HD-audio is the only user, so far. I don't want to expand the core stuff (and kconfig) unless really needed.
I said I'd want this for ice1724 / vt1616, and I finally got time to do it. What do you think of this patch? I'm happily using it.
From: Daniel Jacobowitz dan@codesourcery.com
Move VMASTER from HDA out to core, and enable it for VT1616.
Signed-off-by: Daniel Jacobowitz dan@codesourcery.com
diff --git a/include/sound/control.h b/include/sound/control.h index e79baa6..3dc1291 100644 --- a/include/sound/control.h +++ b/include/sound/control.h @@ -169,4 +169,11 @@ int snd_ctl_boolean_mono_info(struct snd_kcontrol *kcontrol, int snd_ctl_boolean_stereo_info(struct snd_kcontrol *kcontrol, struct snd_ctl_elem_info *uinfo);
+/* + * virtual master control + */ +struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, + const unsigned int *tlv); +int snd_ctl_add_slave(struct snd_kcontrol *master, struct snd_kcontrol *slave); + #endif /* __SOUND_CONTROL_H */ diff --git a/sound/core/Kconfig b/sound/core/Kconfig index 829ca38..b0528c4 100644 --- a/sound/core/Kconfig +++ b/sound/core/Kconfig @@ -16,6 +16,9 @@ config SND_RAWMIDI tristate depends on SND
+config SND_VMASTER + bool + config SND_SEQUENCER tristate "Sequencer support" depends on SND diff --git a/sound/core/Makefile b/sound/core/Makefile index 267039a..da8e685 100644 --- a/sound/core/Makefile +++ b/sound/core/Makefile @@ -6,6 +6,7 @@ snd-y := sound.o init.o memory.o info.o control.o misc.o device.o snd-$(CONFIG_ISA_DMA_API) += isadma.o snd-$(CONFIG_SND_OSSEMUL) += sound_oss.o info_oss.o +snd-$(CONFIG_SND_VMASTER) += vmaster.o
snd-pcm-objs := pcm.o pcm_native.o pcm_lib.o pcm_timer.o pcm_misc.o \ pcm_memory.o diff --git a/sound/core/vmaster.c b/sound/core/vmaster.c new file mode 100644 index 0000000..7cfd8b8 --- /dev/null +++ b/sound/core/vmaster.c @@ -0,0 +1,368 @@ +/* + * Virtual master and slave controls + * + * Copyright (c) 2008 by Takashi Iwai tiwai@suse.de + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of the GNU General Public License as + * published by the Free Software Foundation, version 2. + * + */ + +#include <linux/slab.h> +#include <sound/core.h> +#include <sound/control.h> + +/* + * a subset of information returned via ctl info callback + */ +struct link_ctl_info { + int type; /* value type */ + int count; /* item count */ + int min_val, max_val; /* min, max values */ +}; + +/* + * link master - this contains a list of slave controls that are + * identical types, i.e. info returns the same value type and value + * ranges, but may have different number of counts. + * + * The master control is so far only mono volume/switch for simplicity. + * The same value will be applied to all slaves. + */ +struct link_master { + struct list_head slaves; + struct link_ctl_info info; + int val; /* the master value */ +}; + +/* + * link slave - this contains a slave control element + * + * It fakes the control callbacsk with additional attenuation by the + * master control. A slave may have either one or two channels. + */ + +struct link_slave { + struct list_head list; + struct link_master *master; + struct link_ctl_info info; + int vals[2]; /* current values */ + struct snd_kcontrol slave; /* the copy of original control entry */ +}; + +/* get the slave ctl info and save the initial values */ +static int slave_init(struct link_slave *slave) +{ + struct snd_ctl_elem_info *uinfo; + struct snd_ctl_elem_value *uctl; + int err, ch; + + if (slave->info.count) + return 0; /* already initialized */ + + uinfo = kmalloc(sizeof(*uinfo), GFP_KERNEL); + if (!uinfo) + return -ENOMEM; + uinfo->id = slave->slave.id; + err = slave->slave.info(&slave->slave, uinfo); + if (err < 0) { + kfree(uinfo); + return err; + } + slave->info.type = uinfo->type; + slave->info.count = uinfo->count; + if (slave->info.count > 2 || + (slave->info.type != SNDRV_CTL_ELEM_TYPE_INTEGER && + slave->info.type != SNDRV_CTL_ELEM_TYPE_BOOLEAN)) { + snd_printk(KERN_ERR "invalid slave element\n"); + kfree(uinfo); + return -EINVAL; + } + slave->info.min_val = uinfo->value.integer.min; + slave->info.max_val = uinfo->value.integer.max; + kfree(uinfo); + + uctl = kmalloc(sizeof(*uctl), GFP_KERNEL); + if (!uctl) + return -ENOMEM; + uctl->id = slave->slave.id; + err = slave->slave.get(&slave->slave, uctl); + for (ch = 0; ch < slave->info.count; ch++) + slave->vals[ch] = uctl->value.integer.value[ch]; + kfree(uctl); + return 0; +} + +/* initialize master volume */ +static int master_init(struct link_master *master) +{ + struct link_slave *slave; + + if (master->info.count) + return 0; /* already initialized */ + + list_for_each_entry(slave, &master->slaves, list) { + int err = slave_init(slave); + if (err < 0) + return err; + master->info = slave->info; + master->info.count = 1; /* always mono */ + /* set full volume as default (= no attenuation) */ + master->val = master->info.max_val; + return 0; + } + return -ENOENT; +} + +static int slave_get_val(struct link_slave *slave, + struct snd_ctl_elem_value *ucontrol) +{ + int err, ch; + + err = slave_init(slave); + if (err < 0) + return err; + for (ch = 0; ch < slave->info.count; ch++) + ucontrol->value.integer.value[ch] = slave->vals[ch]; + return 0; +} + +static int slave_put_val(struct link_slave *slave, + struct snd_ctl_elem_value *ucontrol) +{ + int err, ch, vol; + + err = master_init(slave->master); + if (err < 0) + return err; + + switch (slave->info.type) { + case SNDRV_CTL_ELEM_TYPE_BOOLEAN: + for (ch = 0; ch < slave->info.count; ch++) + ucontrol->value.integer.value[ch] &= + !!slave->master->val; + break; + case SNDRV_CTL_ELEM_TYPE_INTEGER: + for (ch = 0; ch < slave->info.count; ch++) { + /* max master volume is supposed to be 0 dB */ + vol = ucontrol->value.integer.value[ch]; + vol += slave->master->val - slave->master->info.max_val; + if (vol < slave->info.min_val) + vol = slave->info.min_val; + else if (vol > slave->info.max_val) + vol = slave->info.max_val; + ucontrol->value.integer.value[ch] = vol; + } + break; + } + return slave->slave.put(&slave->slave, ucontrol); +} + +/* + * ctl callbacks for slaves + */ +static int slave_info(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_info *uinfo) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + return slave->slave.info(&slave->slave, uinfo); +} + +static int slave_get(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + return slave_get_val(slave, ucontrol); +} + +static int slave_put(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + int err, ch, changed = 0; + + err = slave_init(slave); + if (err < 0) + return err; + for (ch = 0; ch < slave->info.count; ch++) { + if (slave->vals[ch] != ucontrol->value.integer.value[ch]) { + changed = 1; + slave->vals[ch] = ucontrol->value.integer.value[ch]; + } + } + if (!changed) + return 0; + return slave_put_val(slave, ucontrol); +} + +static int slave_tlv_cmd(struct snd_kcontrol *kcontrol, + int op_flag, unsigned int size, + unsigned int __user *tlv) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + /* FIXME: this assumes that the max volume is 0 dB */ + return slave->slave.tlv.c(&slave->slave, op_flag, size, tlv); +} + +static void slave_free(struct snd_kcontrol *kcontrol) +{ + struct link_slave *slave = snd_kcontrol_chip(kcontrol); + if (slave->slave.private_free) + slave->slave.private_free(&slave->slave); + if (slave->master) + list_del(&slave->list); + kfree(slave); +} + +/* + * Add a slave control to the group with the given master control + * + * All slaves must be the same type (returning the same information + * via info callback). The fucntion doesn't check it, so it's your + * responsibility. + * + * Also, some additional limitations: + * - at most two channels + * - logarithmic volume control (dB level), no linear volume + * - master can only attenuate the volume, no gain + */ +int snd_ctl_add_slave(struct snd_kcontrol *master, struct snd_kcontrol *slave) +{ + struct link_master *master_link = snd_kcontrol_chip(master); + struct link_slave *srec; + + srec = kzalloc(sizeof(*srec) + + slave->count * sizeof(*slave->vd), GFP_KERNEL); + if (!srec) + return -ENOMEM; + srec->slave = *slave; + memcpy(srec->slave.vd, slave->vd, slave->count * sizeof(*slave->vd)); + srec->master = master_link; + + /* override callbacks */ + slave->info = slave_info; + slave->get = slave_get; + slave->put = slave_put; + if (slave->vd[0].access & SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK) + slave->tlv.c = slave_tlv_cmd; + slave->private_data = srec; + slave->private_free = slave_free; + + list_add_tail(&srec->list, &master_link->slaves); + return 0; +} + +EXPORT_SYMBOL(snd_ctl_add_slave); + +/* + * ctl callbacks for master controls + */ +static int master_info(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_info *uinfo) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + int ret; + + ret = master_init(master); + if (ret < 0) + return ret; + uinfo->type = master->info.type; + uinfo->count = master->info.count; + uinfo->value.integer.min = master->info.min_val; + uinfo->value.integer.max = master->info.max_val; + return 0; +} + +static int master_get(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + int err = master_init(master); + if (err < 0) + return err; + ucontrol->value.integer.value[0] = master->val; + return 0; +} + +static int master_put(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + struct link_slave *slave; + struct snd_ctl_elem_value *uval; + int err, old_val; + + err = master_init(master); + if (err < 0) + return err; + old_val = master->val; + if (ucontrol->value.integer.value[0] == old_val) + return 0; + + uval = kmalloc(sizeof(*uval), GFP_KERNEL); + if (!uval) + return -ENOMEM; + list_for_each_entry(slave, &master->slaves, list) { + master->val = old_val; + uval->id = slave->slave.id; + slave_get_val(slave, uval); + master->val = ucontrol->value.integer.value[0]; + slave_put_val(slave, uval); + } + kfree(uval); + return 1; +} + +static void master_free(struct snd_kcontrol *kcontrol) +{ + struct link_master *master = snd_kcontrol_chip(kcontrol); + struct link_slave *slave; + + list_for_each_entry(slave, &master->slaves, list) + slave->master = NULL; + kfree(master); +} + + +/* + * Create a virtual master control with the given name + */ +struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, + const unsigned int *tlv) +{ + struct link_master *master; + struct snd_kcontrol *kctl; + struct snd_kcontrol_new knew; + + memset(&knew, 0, sizeof(knew)); + knew.iface = SNDRV_CTL_ELEM_IFACE_MIXER; + knew.name = name; + knew.info = master_info; + + master = kzalloc(sizeof(*master), GFP_KERNEL); + if (!master) + return NULL; + INIT_LIST_HEAD(&master->slaves); + + kctl = snd_ctl_new1(&knew, master); + if (!kctl) { + kfree(master); + return NULL; + } + /* override some callbacks */ + kctl->info = master_info; + kctl->get = master_get; + kctl->put = master_put; + kctl->private_free = master_free; + + /* additional (constant) TLV read */ + if (tlv) { + /* FIXME: this assumes that the max volume is 0 dB */ + kctl->vd[0].access |= SNDRV_CTL_ELEM_ACCESS_TLV_READ; + kctl->tlv.p = tlv; + } + return kctl; +} + +EXPORT_SYMBOL(snd_ctl_make_virtual_master); diff --git a/sound/drivers/Kconfig b/sound/drivers/Kconfig index 75d4fe0..154e0ae 100644 --- a/sound/drivers/Kconfig +++ b/sound/drivers/Kconfig @@ -27,6 +27,7 @@ config SND_AC97_CODEC tristate select SND_PCM select AC97_BUS + select SND_VMASTER
config SND_DUMMY tristate "Dummy (/dev/null) soundcard" diff --git a/sound/pci/Kconfig b/sound/pci/Kconfig index 812085d..48296d9 100644 --- a/sound/pci/Kconfig +++ b/sound/pci/Kconfig @@ -517,6 +517,7 @@ config SND_HDA_INTEL tristate "Intel HD Audio" depends on SND select SND_PCM + select SND_VMASTER help Say Y here to include support for Intel "High Definition Audio" (Azalia) motherboard devices. diff --git a/sound/pci/ac97/ac97_patch.c b/sound/pci/ac97/ac97_patch.c index 50c637e..2019602 100644 --- a/sound/pci/ac97/ac97_patch.c +++ b/sound/pci/ac97/ac97_patch.c @@ -3312,8 +3312,66 @@ AC97_SINGLE("Downmix LFE and Center to Front", 0x5a, 12, 1, 0), AC97_SINGLE("Downmix Surround to Front", 0x5a, 11, 1, 0), };
+static const char *slave_vols_vt1616[] = { + "Front Playback Volume", + "Surround Playback Volume", + "Center Playback Volume", + "LFE Playback Volume", + NULL +}; + +static const char *slave_sws_vt1616[] = { + "Front Playback Switch", + "Surround Playback Switch", + "Center Playback Switch", + "LFE Playback Switch", + NULL +}; + +/* find a mixer control element with the given name */ +static struct snd_kcontrol *snd_ac97_find_mixer_ctl(struct snd_ac97 *ac97, + const char *name) +{ + struct snd_ctl_elem_id id; + memset(&id, 0, sizeof(id)); + id.iface = SNDRV_CTL_ELEM_IFACE_MIXER; + strcpy(id.name, name); + return snd_ctl_find_id(ac97->bus->card, &id); +} + +/* create a virtual master control and add slaves */ +int snd_ac97_add_vmaster(struct snd_ac97 *ac97, char *name, + const unsigned int *tlv, const char **slaves) +{ + struct snd_kcontrol *kctl; + const char **s; + int err; + + kctl = snd_ctl_make_virtual_master(name, tlv); + if (!kctl) + return -ENOMEM; + err = snd_ctl_add(ac97->bus->card, kctl); + if (err < 0) + return err; + + for (s = slaves; *s; s++) { + struct snd_kcontrol *sctl; + + sctl = snd_ac97_find_mixer_ctl(ac97, *s); + if (!sctl) { + snd_printdd("Cannot find slave %s, skipped\n", *s); + continue; + } + err = snd_ctl_add_slave(kctl, sctl); + if (err < 0) + return err; + } + return 0; +} + static int patch_vt1616_specific(struct snd_ac97 * ac97) { + struct snd_kcontrol *kctl; int err;
if (snd_ac97_try_bit(ac97, 0x5a, 9)) @@ -3321,6 +3379,24 @@ static int patch_vt1616_specific(struct snd_ac97 * ac97) return err; if ((err = patch_build_controls(ac97, &snd_ac97_controls_vt1616[1], ARRAY_SIZE(snd_ac97_controls_vt1616) - 1)) < 0) return err; + + /* There is already a misnamed master switch. Rename it. */ + kctl = snd_ac97_find_mixer_ctl(ac97, "Master Playback Volume"); + if (!kctl) + return -EINVAL; + + snd_ac97_rename_vol_ctl(ac97, "Master Playback", "Front Playback"); + + err = snd_ac97_add_vmaster(ac97, "Master Playback Volume", + kctl->tlv.p, slave_vols_vt1616); + if (err < 0) + return err; + + err = snd_ac97_add_vmaster(ac97, "Master Playback Switch", + NULL, slave_sws_vt1616); + if (err < 0) + return err; + return 0; }
diff --git a/sound/pci/hda/Makefile b/sound/pci/hda/Makefile index 9e0d8a1..ab0c726 100644 --- a/sound/pci/hda/Makefile +++ b/sound/pci/hda/Makefile @@ -2,7 +2,7 @@ snd-hda-intel-y := hda_intel.o # since snd-hda-intel is the only driver using hda-codec, # merge it into a single module although it was originally # designed to be individual modules -snd-hda-intel-y += hda_codec.o vmaster.o +snd-hda-intel-y += hda_codec.o snd-hda-intel-$(CONFIG_PROC_FS) += hda_proc.o snd-hda-intel-$(CONFIG_SND_HDA_HWDEP) += hda_hwdep.o snd-hda-intel-$(CONFIG_SND_HDA_GENERIC) += hda_generic.o diff --git a/sound/pci/hda/hda_local.h b/sound/pci/hda/hda_local.h index ad0014a..5a859d8 100644 --- a/sound/pci/hda/hda_local.h +++ b/sound/pci/hda/hda_local.h @@ -407,11 +407,4 @@ int snd_hda_check_amp_list_power(struct hda_codec *codec, hda_nid_t nid); #endif /* CONFIG_SND_HDA_POWER_SAVE */
-/* - * virtual master control - */ -struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, - const unsigned int *tlv); -int snd_ctl_add_slave(struct snd_kcontrol *master, struct snd_kcontrol *slave); - #endif /* __SOUND_HDA_LOCAL_H */ diff --git a/sound/pci/hda/vmaster.c b/sound/pci/hda/vmaster.c deleted file mode 100644 index 2da49d2..0000000 --- a/sound/pci/hda/vmaster.c +++ /dev/null @@ -1,364 +0,0 @@ -/* - * Virtual master and slave controls - * - * Copyright (c) 2008 by Takashi Iwai tiwai@suse.de - * - * This program is free software; you can redistribute it and/or - * modify it under the terms of the GNU General Public License as - * published by the Free Software Foundation, version 2. - * - */ - -#include <linux/slab.h> -#include <sound/core.h> -#include <sound/control.h> - -/* - * a subset of information returned via ctl info callback - */ -struct link_ctl_info { - int type; /* value type */ - int count; /* item count */ - int min_val, max_val; /* min, max values */ -}; - -/* - * link master - this contains a list of slave controls that are - * identical types, i.e. info returns the same value type and value - * ranges, but may have different number of counts. - * - * The master control is so far only mono volume/switch for simplicity. - * The same value will be applied to all slaves. - */ -struct link_master { - struct list_head slaves; - struct link_ctl_info info; - int val; /* the master value */ -}; - -/* - * link slave - this contains a slave control element - * - * It fakes the control callbacsk with additional attenuation by the - * master control. A slave may have either one or two channels. - */ - -struct link_slave { - struct list_head list; - struct link_master *master; - struct link_ctl_info info; - int vals[2]; /* current values */ - struct snd_kcontrol slave; /* the copy of original control entry */ -}; - -/* get the slave ctl info and save the initial values */ -static int slave_init(struct link_slave *slave) -{ - struct snd_ctl_elem_info *uinfo; - struct snd_ctl_elem_value *uctl; - int err, ch; - - if (slave->info.count) - return 0; /* already initialized */ - - uinfo = kmalloc(sizeof(*uinfo), GFP_KERNEL); - if (!uinfo) - return -ENOMEM; - uinfo->id = slave->slave.id; - err = slave->slave.info(&slave->slave, uinfo); - if (err < 0) { - kfree(uinfo); - return err; - } - slave->info.type = uinfo->type; - slave->info.count = uinfo->count; - if (slave->info.count > 2 || - (slave->info.type != SNDRV_CTL_ELEM_TYPE_INTEGER && - slave->info.type != SNDRV_CTL_ELEM_TYPE_BOOLEAN)) { - snd_printk(KERN_ERR "invalid slave element\n"); - kfree(uinfo); - return -EINVAL; - } - slave->info.min_val = uinfo->value.integer.min; - slave->info.max_val = uinfo->value.integer.max; - kfree(uinfo); - - uctl = kmalloc(sizeof(*uctl), GFP_KERNEL); - if (!uctl) - return -ENOMEM; - uctl->id = slave->slave.id; - err = slave->slave.get(&slave->slave, uctl); - for (ch = 0; ch < slave->info.count; ch++) - slave->vals[ch] = uctl->value.integer.value[ch]; - kfree(uctl); - return 0; -} - -/* initialize master volume */ -static int master_init(struct link_master *master) -{ - struct link_slave *slave; - - if (master->info.count) - return 0; /* already initialized */ - - list_for_each_entry(slave, &master->slaves, list) { - int err = slave_init(slave); - if (err < 0) - return err; - master->info = slave->info; - master->info.count = 1; /* always mono */ - /* set full volume as default (= no attenuation) */ - master->val = master->info.max_val; - return 0; - } - return -ENOENT; -} - -static int slave_get_val(struct link_slave *slave, - struct snd_ctl_elem_value *ucontrol) -{ - int err, ch; - - err = slave_init(slave); - if (err < 0) - return err; - for (ch = 0; ch < slave->info.count; ch++) - ucontrol->value.integer.value[ch] = slave->vals[ch]; - return 0; -} - -static int slave_put_val(struct link_slave *slave, - struct snd_ctl_elem_value *ucontrol) -{ - int err, ch, vol; - - err = master_init(slave->master); - if (err < 0) - return err; - - switch (slave->info.type) { - case SNDRV_CTL_ELEM_TYPE_BOOLEAN: - for (ch = 0; ch < slave->info.count; ch++) - ucontrol->value.integer.value[ch] &= - !!slave->master->val; - break; - case SNDRV_CTL_ELEM_TYPE_INTEGER: - for (ch = 0; ch < slave->info.count; ch++) { - /* max master volume is supposed to be 0 dB */ - vol = ucontrol->value.integer.value[ch]; - vol += slave->master->val - slave->master->info.max_val; - if (vol < slave->info.min_val) - vol = slave->info.min_val; - else if (vol > slave->info.max_val) - vol = slave->info.max_val; - ucontrol->value.integer.value[ch] = vol; - } - break; - } - return slave->slave.put(&slave->slave, ucontrol); -} - -/* - * ctl callbacks for slaves - */ -static int slave_info(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_info *uinfo) -{ - struct link_slave *slave = snd_kcontrol_chip(kcontrol); - return slave->slave.info(&slave->slave, uinfo); -} - -static int slave_get(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct link_slave *slave = snd_kcontrol_chip(kcontrol); - return slave_get_val(slave, ucontrol); -} - -static int slave_put(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct link_slave *slave = snd_kcontrol_chip(kcontrol); - int err, ch, changed = 0; - - err = slave_init(slave); - if (err < 0) - return err; - for (ch = 0; ch < slave->info.count; ch++) { - if (slave->vals[ch] != ucontrol->value.integer.value[ch]) { - changed = 1; - slave->vals[ch] = ucontrol->value.integer.value[ch]; - } - } - if (!changed) - return 0; - return slave_put_val(slave, ucontrol); -} - -static int slave_tlv_cmd(struct snd_kcontrol *kcontrol, - int op_flag, unsigned int size, - unsigned int __user *tlv) -{ - struct link_slave *slave = snd_kcontrol_chip(kcontrol); - /* FIXME: this assumes that the max volume is 0 dB */ - return slave->slave.tlv.c(&slave->slave, op_flag, size, tlv); -} - -static void slave_free(struct snd_kcontrol *kcontrol) -{ - struct link_slave *slave = snd_kcontrol_chip(kcontrol); - if (slave->slave.private_free) - slave->slave.private_free(&slave->slave); - if (slave->master) - list_del(&slave->list); - kfree(slave); -} - -/* - * Add a slave control to the group with the given master control - * - * All slaves must be the same type (returning the same information - * via info callback). The fucntion doesn't check it, so it's your - * responsibility. - * - * Also, some additional limitations: - * - at most two channels - * - logarithmic volume control (dB level), no linear volume - * - master can only attenuate the volume, no gain - */ -int snd_ctl_add_slave(struct snd_kcontrol *master, struct snd_kcontrol *slave) -{ - struct link_master *master_link = snd_kcontrol_chip(master); - struct link_slave *srec; - - srec = kzalloc(sizeof(*srec) + - slave->count * sizeof(*slave->vd), GFP_KERNEL); - if (!srec) - return -ENOMEM; - srec->slave = *slave; - memcpy(srec->slave.vd, slave->vd, slave->count * sizeof(*slave->vd)); - srec->master = master_link; - - /* override callbacks */ - slave->info = slave_info; - slave->get = slave_get; - slave->put = slave_put; - if (slave->vd[0].access & SNDRV_CTL_ELEM_ACCESS_TLV_CALLBACK) - slave->tlv.c = slave_tlv_cmd; - slave->private_data = srec; - slave->private_free = slave_free; - - list_add_tail(&srec->list, &master_link->slaves); - return 0; -} - -/* - * ctl callbacks for master controls - */ -static int master_info(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_info *uinfo) -{ - struct link_master *master = snd_kcontrol_chip(kcontrol); - int ret; - - ret = master_init(master); - if (ret < 0) - return ret; - uinfo->type = master->info.type; - uinfo->count = master->info.count; - uinfo->value.integer.min = master->info.min_val; - uinfo->value.integer.max = master->info.max_val; - return 0; -} - -static int master_get(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct link_master *master = snd_kcontrol_chip(kcontrol); - int err = master_init(master); - if (err < 0) - return err; - ucontrol->value.integer.value[0] = master->val; - return 0; -} - -static int master_put(struct snd_kcontrol *kcontrol, - struct snd_ctl_elem_value *ucontrol) -{ - struct link_master *master = snd_kcontrol_chip(kcontrol); - struct link_slave *slave; - struct snd_ctl_elem_value *uval; - int err, old_val; - - err = master_init(master); - if (err < 0) - return err; - old_val = master->val; - if (ucontrol->value.integer.value[0] == old_val) - return 0; - - uval = kmalloc(sizeof(*uval), GFP_KERNEL); - if (!uval) - return -ENOMEM; - list_for_each_entry(slave, &master->slaves, list) { - master->val = old_val; - uval->id = slave->slave.id; - slave_get_val(slave, uval); - master->val = ucontrol->value.integer.value[0]; - slave_put_val(slave, uval); - } - kfree(uval); - return 1; -} - -static void master_free(struct snd_kcontrol *kcontrol) -{ - struct link_master *master = snd_kcontrol_chip(kcontrol); - struct link_slave *slave; - - list_for_each_entry(slave, &master->slaves, list) - slave->master = NULL; - kfree(master); -} - - -/* - * Create a virtual master control with the given name - */ -struct snd_kcontrol *snd_ctl_make_virtual_master(char *name, - const unsigned int *tlv) -{ - struct link_master *master; - struct snd_kcontrol *kctl; - struct snd_kcontrol_new knew; - - memset(&knew, 0, sizeof(knew)); - knew.iface = SNDRV_CTL_ELEM_IFACE_MIXER; - knew.name = name; - knew.info = master_info; - - master = kzalloc(sizeof(*master), GFP_KERNEL); - if (!master) - return NULL; - INIT_LIST_HEAD(&master->slaves); - - kctl = snd_ctl_new1(&knew, master); - if (!kctl) { - kfree(master); - return NULL; - } - /* override some callbacks */ - kctl->info = master_info; - kctl->get = master_get; - kctl->put = master_put; - kctl->private_free = master_free; - - /* additional (constant) TLV read */ - if (tlv) { - /* FIXME: this assumes that the max volume is 0 dB */ - kctl->vd[0].access |= SNDRV_CTL_ELEM_ACCESS_TLV_READ; - kctl->tlv.p = tlv; - } - return kctl; -}
At Fri, 21 Mar 2008 15:08:55 -0400, Daniel Jacobowitz wrote:
On Wed, Jan 30, 2008 at 04:27:07PM +0100, Takashi Iwai wrote:
Because HD-audio is the only user, so far. I don't want to expand the core stuff (and kconfig) unless really needed.
I said I'd want this for ice1724 / vt1616, and I finally got time to do it. What do you think of this patch? I'm happily using it.
Well, don't forget to check HG version before you submit a patch :)
Takashi
On Sat, Mar 22, 2008 at 10:09:23AM +0100, Takashi Iwai wrote:
At Fri, 21 Mar 2008 15:08:55 -0400, Daniel Jacobowitz wrote:
On Wed, Jan 30, 2008 at 04:27:07PM +0100, Takashi Iwai wrote:
Because HD-audio is the only user, so far. I don't want to expand the core stuff (and kconfig) unless really needed.
I said I'd want this for ice1724 / vt1616, and I finally got time to do it. What do you think of this patch? I'm happily using it.
Well, don't forget to check HG version before you submit a patch :)
Whoops! This version works with the hg tree. The new snd_ac97_add_vmaster is mostly similar to the hda version so it might make sense to put that in vmaster.c instead.
From: Daniel Jacobowitz dan@codesourcery.com
Enable VMASTER for VT1616 / VT1617A.
Signed-off-by: Daniel Jacobowitz dan@codesourcery.com
diff -r 0d5f43585ca7 drivers/Kconfig --- a/drivers/Kconfig Sat Mar 22 10:26:05 2008 +0100 +++ b/drivers/Kconfig Sat Mar 22 11:02:16 2008 -0400 @@ -44,6 +44,7 @@ config SND_AC97_CODEC tristate select SND_PCM select AC97_BUS + select SND_VMASTER
config SND_DUMMY tristate "Dummy (/dev/null) soundcard" diff -r 0d5f43585ca7 pci/ac97/ac97_patch.c --- a/pci/ac97/ac97_patch.c Sat Mar 22 10:26:05 2008 +0100 +++ b/pci/ac97/ac97_patch.c Sat Mar 22 11:02:16 2008 -0400 @@ -3332,8 +3332,66 @@ AC97_SINGLE("Downmix Surround to Front", AC97_SINGLE("Downmix Surround to Front", 0x5a, 11, 1, 0), };
+static const char *slave_vols_vt1616[] = { + "Front Playback Volume", + "Surround Playback Volume", + "Center Playback Volume", + "LFE Playback Volume", + NULL +}; + +static const char *slave_sws_vt1616[] = { + "Front Playback Switch", + "Surround Playback Switch", + "Center Playback Switch", + "LFE Playback Switch", + NULL +}; + +/* find a mixer control element with the given name */ +static struct snd_kcontrol *snd_ac97_find_mixer_ctl(struct snd_ac97 *ac97, + const char *name) +{ + struct snd_ctl_elem_id id; + memset(&id, 0, sizeof(id)); + id.iface = SNDRV_CTL_ELEM_IFACE_MIXER; + strcpy(id.name, name); + return snd_ctl_find_id(ac97->bus->card, &id); +} + +/* create a virtual master control and add slaves */ +int snd_ac97_add_vmaster(struct snd_ac97 *ac97, char *name, + const unsigned int *tlv, const char **slaves) +{ + struct snd_kcontrol *kctl; + const char **s; + int err; + + kctl = snd_ctl_make_virtual_master(name, tlv); + if (!kctl) + return -ENOMEM; + err = snd_ctl_add(ac97->bus->card, kctl); + if (err < 0) + return err; + + for (s = slaves; *s; s++) { + struct snd_kcontrol *sctl; + + sctl = snd_ac97_find_mixer_ctl(ac97, *s); + if (!sctl) { + snd_printdd("Cannot find slave %s, skipped\n", *s); + continue; + } + err = snd_ctl_add_slave(kctl, sctl); + if (err < 0) + return err; + } + return 0; +} + static int patch_vt1616_specific(struct snd_ac97 * ac97) { + struct snd_kcontrol *kctl; int err;
if (snd_ac97_try_bit(ac97, 0x5a, 9)) @@ -3341,6 +3399,24 @@ static int patch_vt1616_specific(struct return err; if ((err = patch_build_controls(ac97, &snd_ac97_controls_vt1616[1], ARRAY_SIZE(snd_ac97_controls_vt1616) - 1)) < 0) return err; + + /* There is already a misnamed master switch. Rename it. */ + kctl = snd_ac97_find_mixer_ctl(ac97, "Master Playback Volume"); + if (!kctl) + return -EINVAL; + + snd_ac97_rename_vol_ctl(ac97, "Master Playback", "Front Playback"); + + err = snd_ac97_add_vmaster(ac97, "Master Playback Volume", + kctl->tlv.p, slave_vols_vt1616); + if (err < 0) + return err; + + err = snd_ac97_add_vmaster(ac97, "Master Playback Switch", + NULL, slave_sws_vt1616); + if (err < 0) + return err; + return 0; }
At Sat, 22 Mar 2008 11:30:46 -0400, Daniel Jacobowitz wrote:
On Sat, Mar 22, 2008 at 10:09:23AM +0100, Takashi Iwai wrote:
At Fri, 21 Mar 2008 15:08:55 -0400, Daniel Jacobowitz wrote:
On Wed, Jan 30, 2008 at 04:27:07PM +0100, Takashi Iwai wrote:
Because HD-audio is the only user, so far. I don't want to expand the core stuff (and kconfig) unless really needed.
I said I'd want this for ice1724 / vt1616, and I finally got time to do it. What do you think of this patch? I'm happily using it.
Well, don't forget to check HG version before you submit a patch :)
Whoops! This version works with the hg tree. The new snd_ac97_add_vmaster is mostly similar to the hda version so it might make sense to put that in vmaster.c instead.
From: Daniel Jacobowitz dan@codesourcery.com
Enable VMASTER for VT1616 / VT1617A.
Signed-off-by: Daniel Jacobowitz dan@codesourcery.com
Thanks. This kind of change is basically good to apply, I think, but a couple of things related are in my mind:
- other AC97 codecs need similar hacks, so it could be done cleaner (in a more generic way) than in each patch_*()? - headphone and line-out have the similar issue - it's handled as a bind-control right now, though - what about if the device actually uses only two channels although codec supports more than two? Then the additional master control makes little sense. But, we have no way to know it from the codec itself. Another quirk table? That's to be avoided...
Since this is a non-urgent fix and I'll be on vacation for three weeks from tomorrow, I'd like to keep this pending unless other developers want to put this in. Is it OK?
Takashi
diff -r 0d5f43585ca7 drivers/Kconfig --- a/drivers/Kconfig Sat Mar 22 10:26:05 2008 +0100 +++ b/drivers/Kconfig Sat Mar 22 11:02:16 2008 -0400 @@ -44,6 +44,7 @@ config SND_AC97_CODEC tristate select SND_PCM select AC97_BUS
- select SND_VMASTER
config SND_DUMMY tristate "Dummy (/dev/null) soundcard" diff -r 0d5f43585ca7 pci/ac97/ac97_patch.c --- a/pci/ac97/ac97_patch.c Sat Mar 22 10:26:05 2008 +0100 +++ b/pci/ac97/ac97_patch.c Sat Mar 22 11:02:16 2008 -0400 @@ -3332,8 +3332,66 @@ AC97_SINGLE("Downmix Surround to Front", AC97_SINGLE("Downmix Surround to Front", 0x5a, 11, 1, 0), };
+static const char *slave_vols_vt1616[] = {
- "Front Playback Volume",
- "Surround Playback Volume",
- "Center Playback Volume",
- "LFE Playback Volume",
- NULL
+};
+static const char *slave_sws_vt1616[] = {
- "Front Playback Switch",
- "Surround Playback Switch",
- "Center Playback Switch",
- "LFE Playback Switch",
- NULL
+};
+/* find a mixer control element with the given name */ +static struct snd_kcontrol *snd_ac97_find_mixer_ctl(struct snd_ac97 *ac97,
const char *name)+{
- struct snd_ctl_elem_id id;
- memset(&id, 0, sizeof(id));
- id.iface = SNDRV_CTL_ELEM_IFACE_MIXER;
- strcpy(id.name, name);
- return snd_ctl_find_id(ac97->bus->card, &id);
+}
+/* create a virtual master control and add slaves */ +int snd_ac97_add_vmaster(struct snd_ac97 *ac97, char *name,
const unsigned int *tlv, const char **slaves)+{
- struct snd_kcontrol *kctl;
- const char **s;
- int err;
- kctl = snd_ctl_make_virtual_master(name, tlv);
- if (!kctl)
return -ENOMEM;- err = snd_ctl_add(ac97->bus->card, kctl);
- if (err < 0)
return err;- for (s = slaves; *s; s++) {
struct snd_kcontrol *sctl;sctl = snd_ac97_find_mixer_ctl(ac97, *s);if (!sctl) {snd_printdd("Cannot find slave %s, skipped\n", *s);continue;}err = snd_ctl_add_slave(kctl, sctl);if (err < 0)return err;- }
- return 0;
+}
static int patch_vt1616_specific(struct snd_ac97 * ac97) {
struct snd_kcontrol *kctl; int err;
if (snd_ac97_try_bit(ac97, 0x5a, 9))
@@ -3341,6 +3399,24 @@ static int patch_vt1616_specific(struct return err; if ((err = patch_build_controls(ac97, &snd_ac97_controls_vt1616[1], ARRAY_SIZE(snd_ac97_controls_vt1616) - 1)) < 0) return err;
- /* There is already a misnamed master switch. Rename it. */
- kctl = snd_ac97_find_mixer_ctl(ac97, "Master Playback Volume");
- if (!kctl)
return -EINVAL;- snd_ac97_rename_vol_ctl(ac97, "Master Playback", "Front Playback");
- err = snd_ac97_add_vmaster(ac97, "Master Playback Volume",
kctl->tlv.p, slave_vols_vt1616);- if (err < 0)
return err;- err = snd_ac97_add_vmaster(ac97, "Master Playback Switch",
NULL, slave_sws_vt1616);- if (err < 0)
return err;- return 0;
}
-- Daniel Jacobowitz CodeSourcery
On Sat, Mar 22, 2008 at 05:31:21PM +0100, Takashi Iwai wrote:
Thanks. This kind of change is basically good to apply, I think, but a couple of things related are in my mind:
- other AC97 codecs need similar hacks, so it could be done cleaner (in a more generic way) than in each patch_*()?
- headphone and line-out have the similar issue - it's handled as a bind-control right now, though
- what about if the device actually uses only two channels although codec supports more than two? Then the additional master control makes little sense. But, we have no way to know it from the codec itself. Another quirk table? That's to be avoided...
Since this is a non-urgent fix and I'll be on vacation for three weeks from tomorrow, I'd like to keep this pending unless other developers want to put this in. Is it OK?
Fine with me. I don't think I can make broader changes than this kind, because I don't know much about AC97; maybe there can be a flag on recognized codecs saying which channels the front volume really controls?
I noticed that the PCM mixer also only affected my front speakers.
On Sat, Mar 22, 2008 at 03:15:51PM -0400, Daniel Jacobowitz wrote:
On Sat, Mar 22, 2008 at 05:31:21PM +0100, Takashi Iwai wrote:
Thanks. This kind of change is basically good to apply, I think, but a couple of things related are in my mind:
- other AC97 codecs need similar hacks, so it could be done cleaner (in a more generic way) than in each patch_*()?
- headphone and line-out have the similar issue - it's handled as a bind-control right now, though
- what about if the device actually uses only two channels although codec supports more than two? Then the additional master control makes little sense. But, we have no way to know it from the codec itself. Another quirk table? That's to be avoided...
Since this is a non-urgent fix and I'll be on vacation for three weeks from tomorrow, I'd like to keep this pending unless other developers want to put this in. Is it OK?
Fine with me. I don't think I can make broader changes than this kind, because I don't know much about AC97; maybe there can be a flag on recognized codecs saying which channels the front volume really controls?
I noticed that the PCM mixer also only affected my front speakers.
Hi Takashi,
Have you had another chance to look at this?
At Fri, 2 May 2008 14:01:54 -0400, Daniel Jacobowitz wrote:
On Sat, Mar 22, 2008 at 03:15:51PM -0400, Daniel Jacobowitz wrote:
On Sat, Mar 22, 2008 at 05:31:21PM +0100, Takashi Iwai wrote:
Thanks. This kind of change is basically good to apply, I think, but a couple of things related are in my mind:
- other AC97 codecs need similar hacks, so it could be done cleaner (in a more generic way) than in each patch_*()?
- headphone and line-out have the similar issue - it's handled as a bind-control right now, though
- what about if the device actually uses only two channels although codec supports more than two? Then the additional master control makes little sense. But, we have no way to know it from the codec itself. Another quirk table? That's to be avoided...
Since this is a non-urgent fix and I'll be on vacation for three weeks from tomorrow, I'd like to keep this pending unless other developers want to put this in. Is it OK?
Fine with me. I don't think I can make broader changes than this kind, because I don't know much about AC97; maybe there can be a flag on recognized codecs saying which channels the front volume really controls?
I noticed that the PCM mixer also only affected my front speakers.
Hi Takashi,
Have you had another chance to look at this?
Not quite, unfortunately. But, in the end, I think it's relatively harmless to take your patch as is, since it's pretty local to VT161x codecs.
I'll put it on my local tree and make testing, then merge to the public tree soon later.
thanks,
Takashi
On Tue, May 06, 2008 at 05:17:31PM +0200, Takashi Iwai wrote:
Not quite, unfortunately. But, in the end, I think it's relatively harmless to take your patch as is, since it's pretty local to VT161x codecs.
I'll put it on my local tree and make testing, then merge to the public tree soon later.
Great, thanks a lot! I agree it could be improved - but I'll leave that for another future user :-)
This patch adds the virtual master controls to some codecs. The virtual master controls are added only when no real "Master" control is present.
Takashi
diff -r 861c951ac951 pci/Kconfig --- a/pci/Kconfig Tue Sep 18 15:48:16 2007 +0200 +++ b/pci/Kconfig Tue Sep 18 15:48:29 2007 +0200 @@ -512,6 +512,7 @@ config SND_HDA_CODEC_REALTEK config SND_HDA_CODEC_REALTEK bool "Build Realtek HD-audio codec support" depends on SND_HDA_INTEL + select SND_VMASTER default y help Say Y here to include Realtek HD-audio codec support in @@ -520,6 +521,7 @@ config SND_HDA_CODEC_ANALOG config SND_HDA_CODEC_ANALOG bool "Build Analog Device HD-audio codec support" depends on SND_HDA_INTEL + select SND_VMASTER default y help Say Y here to include Analog Device HD-audio codec support in @@ -528,6 +530,7 @@ config SND_HDA_CODEC_SIGMATEL config SND_HDA_CODEC_SIGMATEL bool "Build IDT/Sigmatel HD-audio codec support" depends on SND_HDA_INTEL + select SND_VMASTER default y help Say Y here to include IDT (Sigmatel) HD-audio codec support in diff -r 861c951ac951 pci/hda/hda_codec.c --- a/pci/hda/hda_codec.c Tue Sep 18 15:48:16 2007 +0200 +++ b/pci/hda/hda_codec.c Tue Sep 18 15:48:29 2007 +0200 @@ -1016,6 +1016,68 @@ int snd_hda_mixer_amp_tlv(struct snd_kco return -EFAULT; return 0; } + +#ifdef CONFIG_SND_VMASTER +/* + * set (static) TLV for virtual master volume; recalculated as max 0dB + */ +void snd_hda_set_vmaster_tlv(struct hda_codec *codec, hda_nid_t nid, int dir, + unsigned int *tlv) +{ + u32 caps; + int nums, step; + + caps = query_amp_caps(codec, nid, dir); + nums = (caps & AC_AMPCAP_NUM_STEPS) >> AC_AMPCAP_NUM_STEPS_SHIFT; + step = (caps & AC_AMPCAP_STEP_SIZE) >> AC_AMPCAP_STEP_SIZE_SHIFT; + step = (step + 1) * 25; + tlv[0] = SNDRV_CTL_TLVT_DB_SCALE; + tlv[1] = 2 * sizeof(unsigned int); + tlv[2] = -nums * step; + tlv[3] = step; +} + +/* find a mixer control element with the given name */ +struct snd_kcontrol *snd_hda_find_mixer_ctl(struct hda_codec *codec, + const char *name) +{ + struct snd_ctl_elem_id id; + memset(&id, 0, sizeof(id)); + id.iface = SNDRV_CTL_ELEM_IFACE_MIXER; + strcpy(id.name, name); + return snd_ctl_find_id(codec->bus->card, &id); +} + +/* create a virtual master control and add slaves */ +int snd_hda_add_vmaster(struct hda_codec *codec, char *name, + unsigned int *tlv, const char **slaves) +{ + struct snd_kcontrol *kctl; + const char **s; + int err; + + kctl = snd_ctl_make_virtual_master(name, tlv); + if (!kctl) + return -ENOMEM; + err = snd_ctl_add(codec->bus->card, kctl); + if (err < 0) + return err; + + for (s = slaves; *s; s++) { + struct snd_kcontrol *sctl; + + sctl = snd_hda_find_mixer_ctl(codec, *s); + if (!sctl) { + snd_printdd("Cannot find slave %s, skipped\n", *s); + continue; + } + err = snd_ctl_add_slave(kctl, sctl); + if (err < 0) + return err; + } + return 0; +} +#endif /* CONFIG_SND_VMASTER */
/* switch */ int snd_hda_mixer_amp_switch_info(struct snd_kcontrol *kcontrol, diff -r 861c951ac951 pci/hda/hda_local.h --- a/pci/hda/hda_local.h Tue Sep 18 15:48:16 2007 +0200 +++ b/pci/hda/hda_local.h Tue Sep 18 15:48:29 2007 +0200 @@ -90,6 +90,13 @@ void snd_hda_codec_resume_amp(struct hda void snd_hda_codec_resume_amp(struct hda_codec *codec); #endif
+void snd_hda_set_vmaster_tlv(struct hda_codec *codec, hda_nid_t nid, int dir, + unsigned int *tlv); +struct snd_kcontrol *snd_hda_find_mixer_ctl(struct hda_codec *codec, + const char *name); +int snd_hda_add_vmaster(struct hda_codec *codec, char *name, + unsigned int *tlv, const char **slaves); + /* amp value bits */ #define HDA_AMP_MUTE 0x80 #define HDA_AMP_UNMUTE 0x00 diff -r 861c951ac951 pci/hda/patch_analog.c --- a/pci/hda/patch_analog.c Tue Sep 18 15:48:16 2007 +0200 +++ b/pci/hda/patch_analog.c Tue Sep 18 15:48:29 2007 +0200 @@ -79,6 +79,11 @@ struct ad198x_spec { #ifdef CONFIG_SND_HDA_POWER_SAVE struct hda_loopback_check loopback; #endif + /* for virtual master */ + hda_nid_t vmaster_nid; + u32 vmaster_tlv[4]; + const char **slave_vols; + const char **slave_sws; };
/* @@ -125,6 +130,28 @@ static int ad198x_init(struct hda_codec snd_hda_sequence_write(codec, spec->init_verbs[i]); return 0; } + +static const char *ad_slave_vols[] = { + "Front Playback Volume", + "Surround Playback Volume", + "Center Playback Volume", + "LFE Playback Volume", + "Side Playback Volume", + "Headphone Playback Volume", + "Mono Playback Volume", + NULL +}; + +static const char *ad_slave_sws[] = { + "Front Playback Switch", + "Surround Playback Switch", + "Center Playback Switch", + "LFE Playback Switch", + "Side Playback Switch", + "Headphone Playback Switch", + "Mono Playback Switch", + NULL +};
static int ad198x_build_controls(struct hda_codec *codec) { @@ -147,6 +174,27 @@ static int ad198x_build_controls(struct if (err < 0) return err; } + + /* if we have no master control, let's create it */ + if (!snd_hda_find_mixer_ctl(codec, "Master Playback Volume")) { + snd_hda_set_vmaster_tlv(codec, spec->vmaster_nid, + HDA_OUTPUT, spec->vmaster_tlv); + err = snd_hda_add_vmaster(codec, "Master Playback Volume", + spec->vmaster_tlv, + (spec->slave_vols ? + spec->slave_vols : ad_slave_vols)); + if (err < 0) + return err; + } + if (!snd_hda_find_mixer_ctl(codec, "Master Playback Switch")) { + err = snd_hda_add_vmaster(codec, "Master Playback Switch", + NULL, + (spec->slave_sws ? + spec->slave_sws : ad_slave_sws)); + if (err < 0) + return err; + } + return 0; }
@@ -897,6 +945,7 @@ static int patch_ad1986a(struct hda_code #ifdef CONFIG_SND_HDA_POWER_SAVE spec->loopback.amplist = ad1986a_loopbacks; #endif + spec->vmaster_nid = 0x1b;
codec->patch_ops = ad198x_patch_ops;
@@ -1129,6 +1178,7 @@ static int patch_ad1983(struct hda_codec #ifdef CONFIG_SND_HDA_POWER_SAVE spec->loopback.amplist = ad1983_loopbacks; #endif + spec->vmaster_nid = 0x05;
codec->patch_ops = ad198x_patch_ops;
@@ -1525,6 +1575,7 @@ static int patch_ad1981(struct hda_codec #ifdef CONFIG_SND_HDA_POWER_SAVE spec->loopback.amplist = ad1981_loopbacks; #endif + spec->vmaster_nid = 0x05;
codec->patch_ops = ad198x_patch_ops;
@@ -2834,6 +2885,7 @@ static int patch_ad1988(struct hda_codec #ifdef CONFIG_SND_HDA_POWER_SAVE spec->loopback.amplist = ad1988_loopbacks; #endif + spec->vmaster_nid = 0x04;
return 0; } @@ -3000,6 +3052,19 @@ static struct hda_amp_list ad1884_loopba }; #endif
+static const char *ad1884_slave_vols[] = { + "PCM Playback Volume", + "Mic Playback Volume", + "Mono Playback Volume", + "Front Mic Playback Volume", + "Mic Playback Volume", + "CD Playback Volume", + "Internal Mic Playback Volume", + "Docking Mic Playback Volume" + "Beep Playback Volume", + NULL +}; + static int patch_ad1884(struct hda_codec *codec) { struct ad198x_spec *spec; @@ -3027,6 +3092,9 @@ static int patch_ad1884(struct hda_codec #ifdef CONFIG_SND_HDA_POWER_SAVE spec->loopback.amplist = ad1884_loopbacks; #endif + spec->vmaster_nid = 0x04; + /* we need to cover all playback volumes */ + spec->slave_vols = ad1884_slave_vols;
codec->patch_ops = ad198x_patch_ops;
@@ -3459,6 +3527,7 @@ static int patch_ad1882(struct hda_codec #ifdef CONFIG_SND_HDA_POWER_SAVE spec->loopback.amplist = ad1882_loopbacks; #endif + spec->vmaster_nid = 0x04;
codec->patch_ops = ad198x_patch_ops;
diff -r 861c951ac951 pci/hda/patch_realtek.c --- a/pci/hda/patch_realtek.c Tue Sep 18 15:48:16 2007 +0200 +++ b/pci/hda/patch_realtek.c Tue Sep 18 15:48:29 2007 +0200 @@ -247,6 +247,9 @@ struct alc_spec { unsigned int sense_updated: 1; unsigned int jack_present: 1;
+ /* for virtual master */ + hda_nid_t vmaster_nid; + u32 vmaster_tlv[4]; #ifdef CONFIG_SND_HDA_POWER_SAVE struct hda_loopback_check loopback; #endif @@ -1098,8 +1101,8 @@ static struct snd_kcontrol_new alc880_f1 static struct snd_kcontrol_new alc880_f1734_mixer[] = { HDA_CODEC_VOLUME("Headphone Playback Volume", 0x0c, 0x0, HDA_OUTPUT), HDA_BIND_MUTE("Headphone Playback Switch", 0x0c, 2, HDA_INPUT), - HDA_CODEC_VOLUME("Internal Speaker Playback Volume", 0x0d, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("Internal Speaker Playback Switch", 0x0d, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Speaker Playback Volume", 0x0d, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Speaker Playback Switch", 0x0d, 2, HDA_INPUT), HDA_CODEC_VOLUME("CD Playback Volume", 0x0b, 0x04, HDA_INPUT), HDA_CODEC_MUTE("CD Playback Switch", 0x0b, 0x04, HDA_INPUT), HDA_CODEC_VOLUME("Mic Playback Volume", 0x0b, 0x0, HDA_INPUT), @@ -1197,10 +1200,10 @@ static struct snd_kcontrol_new alc880_tc
/* Uniwill */ static struct snd_kcontrol_new alc880_uniwill_mixer[] = { - HDA_CODEC_VOLUME("HPhone Playback Volume", 0x0c, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("HPhone Playback Switch", 0x0c, 2, HDA_INPUT), - HDA_CODEC_VOLUME("iSpeaker Playback Volume", 0x0d, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("iSpeaker Playback Switch", 0x0d, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Headphone Playback Volume", 0x0c, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Headphone Playback Switch", 0x0c, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Speaker Playback Volume", 0x0d, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Speaker Playback Switch", 0x0d, 2, HDA_INPUT), HDA_CODEC_VOLUME_MONO("Center Playback Volume", 0x0e, 1, 0x0, HDA_OUTPUT), HDA_CODEC_VOLUME_MONO("LFE Playback Volume", 0x0e, 2, 0x0, HDA_OUTPUT), HDA_BIND_MUTE_MONO("Center Playback Switch", 0x0e, 1, 2, HDA_INPUT), @@ -1240,13 +1243,47 @@ static struct snd_kcontrol_new alc880_fu };
static struct snd_kcontrol_new alc880_uniwill_p53_mixer[] = { - HDA_CODEC_VOLUME("HPhone Playback Volume", 0x0c, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("HPhone Playback Switch", 0x0c, 2, HDA_INPUT), - HDA_CODEC_VOLUME("iSpeaker Playback Volume", 0x0d, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("iSpeaker Playback Switch", 0x0d, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Headphone Playback Volume", 0x0c, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Headphone Playback Switch", 0x0c, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Speaker Playback Volume", 0x0d, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Speaker Playback Switch", 0x0d, 2, HDA_INPUT), HDA_CODEC_VOLUME("Mic Playback Volume", 0x0b, 0x0, HDA_INPUT), HDA_CODEC_MUTE("Mic Playback Switch", 0x0b, 0x0, HDA_INPUT), { } /* end */ +}; + +/* + * virtual master controls + */ + +/* + * slave controls for virtual master + */ +static const char *alc_slave_vols[] = { + "Front Playback Volume", + "Surround Playback Volume", + "Center Playback Volume", + "LFE Playback Volume", + "Side Playback Volume", + "Headphone Playback Volume", + "Speaker Playback Volume", + "Mono Playback Volume", + "iSpeaker Playback Volume", + "Line-Out Playback Volume", + NULL, +}; + +static const char *alc_slave_sws[] = { + "Front Playback Switch", + "Surround Playback Switch", + "Center Playback Switch", + "LFE Playback Switch", + "Side Playback Switch", + "Headphone Playback Switch", + "Speaker Playback Switch", + "Mono Playback Switch", + "iSpeaker Playback Switch", + NULL, };
/* @@ -1275,6 +1312,23 @@ static int alc_build_controls(struct hda if (err < 0) return err; } + + /* if we have no master control, let's create it */ + if (!snd_hda_find_mixer_ctl(codec, "Master Playback Volume")) { + snd_hda_set_vmaster_tlv(codec, spec->vmaster_nid, + HDA_OUTPUT, spec->vmaster_tlv); + err = snd_hda_add_vmaster(codec, "Master Playback Volume", + spec->vmaster_tlv, alc_slave_vols); + if (err < 0) + return err; + } + if (!snd_hda_find_mixer_ctl(codec, "Master Playback Switch")) { + err = snd_hda_add_vmaster(codec, "Master Playback Switch", + NULL, alc_slave_sws); + if (err < 0) + return err; + } + return 0; }
@@ -1823,8 +1877,8 @@ static struct hda_channel_mode alc880_lg
static struct snd_kcontrol_new alc880_lg_mixer[] = { /* FIXME: it's not really "master" but front channels */ - HDA_CODEC_VOLUME("Master Playback Volume", 0x0f, 0x0, HDA_OUTPUT), - HDA_BIND_MUTE("Master Playback Switch", 0x0f, 2, HDA_INPUT), + HDA_CODEC_VOLUME("Front Playback Volume", 0x0f, 0x0, HDA_OUTPUT), + HDA_BIND_MUTE("Front Playback Switch", 0x0f, 2, HDA_INPUT), HDA_CODEC_VOLUME("Surround Playback Volume", 0x0c, 0x0, HDA_OUTPUT), HDA_BIND_MUTE("Surround Playback Switch", 0x0c, 2, HDA_INPUT), HDA_CODEC_VOLUME_MONO("Center Playback Volume", 0x0d, 1, 0x0, HDA_OUTPUT), @@ -3390,6 +3444,8 @@ static int patch_alc880(struct hda_codec spec->num_mixers++; } } + + spec->vmaster_nid = 0x0c;
codec->patch_ops = alc_patch_ops; if (board_config == ALC880_AUTO) @@ -4762,6 +4818,8 @@ static int patch_alc260(struct hda_codec spec->stream_digital_playback = &alc260_pcm_digital_playback; spec->stream_digital_capture = &alc260_pcm_digital_capture;
+ spec->vmaster_nid = 0x08; + codec->patch_ops = alc_patch_ops; if (board_config == ALC260_AUTO) spec->init_hook = alc260_auto_init; @@ -4962,15 +5020,15 @@ static struct snd_kcontrol_new alc882_ba };
static struct snd_kcontrol_new alc885_mbp3_mixer[] = { - HDA_CODEC_VOLUME("Master Volume", 0x0c, 0x00, HDA_OUTPUT), - HDA_BIND_MUTE ("Master Switch", 0x0c, 0x02, HDA_INPUT), - HDA_CODEC_MUTE ("Speaker Switch", 0x14, 0x00, HDA_OUTPUT), - HDA_CODEC_VOLUME("Line Out Volume", 0x0d,0x00, HDA_OUTPUT), - HDA_CODEC_VOLUME("Line In Playback Volume", 0x0b, 0x02, HDA_INPUT), - HDA_CODEC_MUTE ("Line In Playback Switch", 0x0b, 0x02, HDA_INPUT), + HDA_CODEC_VOLUME("Front Playback Volume", 0x0c, 0x00, HDA_OUTPUT), + HDA_BIND_MUTE ("Front Playback Switch", 0x0c, 0x02, HDA_INPUT), + HDA_CODEC_MUTE ("Speaker Playback Switch", 0x14, 0x00, HDA_OUTPUT), + HDA_CODEC_VOLUME("Line-Out Playback Volume", 0x0d, 0x00, HDA_OUTPUT), + HDA_CODEC_VOLUME("Line Playback Volume", 0x0b, 0x02, HDA_INPUT), + HDA_CODEC_MUTE ("Line Playback Switch", 0x0b, 0x02, HDA_INPUT), HDA_CODEC_VOLUME("Mic Playback Volume", 0x0b, 0x00, HDA_INPUT), HDA_CODEC_MUTE ("Mic Playback Switch", 0x0b, 0x00, HDA_INPUT), - HDA_CODEC_VOLUME("Line In Boost", 0x1a, 0x00, HDA_INPUT), + HDA_CODEC_VOLUME("Line Boost", 0x1a, 0x00, HDA_INPUT), HDA_CODEC_VOLUME("Mic Boost", 0x18, 0x00, HDA_INPUT), { } /* end */ }; @@ -5971,6 +6029,8 @@ static int patch_alc882(struct hda_codec spec->num_mixers++; } } + + spec->vmaster_nid = 0x0c;
codec->patch_ops = alc_patch_ops; if (board_config == ALC882_AUTO) @@ -7446,6 +7506,8 @@ static int patch_alc883(struct hda_codec spec->adc_nids = alc883_adc_nids; spec->num_adc_nids = ARRAY_SIZE(alc883_adc_nids); } + + spec->vmaster_nid = 0x0c;
codec->patch_ops = alc_patch_ops; if (board_config == ALC883_AUTO) @@ -8566,6 +8628,8 @@ static int patch_alc262(struct hda_codec } }
+ spec->vmaster_nid = 0x0c; + codec->patch_ops = alc_patch_ops; if (board_config == ALC262_AUTO) spec->init_hook = alc262_auto_init; @@ -9240,6 +9304,9 @@ static int patch_alc268(struct hda_codec } } } + + spec->vmaster_nid = 0x02; + codec->patch_ops = alc_patch_ops; if (board_config == ALC268_AUTO) spec->init_hook = alc268_auto_init; @@ -10394,6 +10461,8 @@ static int patch_alc861(struct hda_codec spec->stream_digital_playback = &alc861_pcm_digital_playback; spec->stream_digital_capture = &alc861_pcm_digital_capture;
+ spec->vmaster_nid = 0x03; + codec->patch_ops = alc_patch_ops; if (board_config == ALC861_AUTO) spec->init_hook = alc861_auto_init; @@ -11370,6 +11439,8 @@ static int patch_alc861vd(struct hda_cod
spec->mixers[spec->num_mixers] = alc861vd_capture_mixer; spec->num_mixers++; + + spec->vmaster_nid = 0x02;
codec->patch_ops = alc_patch_ops;
@@ -12222,6 +12293,8 @@ static int patch_alc662(struct hda_codec spec->num_adc_nids = ARRAY_SIZE(alc662_adc_nids); }
+ spec->vmaster_nid = 0x02; + codec->patch_ops = alc_patch_ops; if (board_config == ALC662_AUTO) spec->init_hook = alc662_auto_init; diff -r 861c951ac951 pci/hda/patch_sigmatel.c --- a/pci/hda/patch_sigmatel.c Tue Sep 18 15:48:16 2007 +0200 +++ b/pci/hda/patch_sigmatel.c Tue Sep 18 15:48:29 2007 +0200 @@ -156,6 +156,9 @@ struct sigmatel_spec { struct snd_kcontrol_new *kctl_alloc; struct hda_input_mux private_dimux; struct hda_input_mux private_imux; + + /* virtual master */ + unsigned int vmaster_tlv[4]; };
static hda_nid_t stac9200_adc_nids[1] = { @@ -521,6 +524,34 @@ static struct snd_kcontrol_new stac927x_ { } /* end */ };
+static const char *slave_vols[] = { + "Front Playback Volume", + "Surround Playback Volume", + "Center Playback Volume", + "LFE Playback Volume", + "Side Playback Volume", + "Headphone Playback Volume", + "Headphone Playback Volume", + "Speaker Playback Volume", + "External Speaker Playback Volume", + "Speaker2 Playback Volume", + NULL +}; + +static const char *slave_sws[] = { + "Front Playback Switch", + "Surround Playback Switch", + "Center Playback Switch", + "LFE Playback Switch", + "Side Playback Switch", + "Headphone Playback Switch", + "Headphone Playback Switch", + "Speaker Playback Switch", + "External Speaker Playback Switch", + "Speaker2 Playback Switch", + NULL +}; + static int stac92xx_build_controls(struct hda_codec *codec) { struct sigmatel_spec *spec = codec->spec; @@ -547,6 +578,23 @@ static int stac92xx_build_controls(struc if (err < 0) return err; } + + /* if we have no master control, let's create it */ + if (!snd_hda_find_mixer_ctl(codec, "Master Playback Volume")) { + snd_hda_set_vmaster_tlv(codec, spec->multiout.dac_nids[0], + HDA_OUTPUT, spec->vmaster_tlv); + err = snd_hda_add_vmaster(codec, "Master Playback Volume", + spec->vmaster_tlv, slave_vols); + if (err < 0) + return err; + } + if (!snd_hda_find_mixer_ctl(codec, "Master Playback Switch")) { + err = snd_hda_add_vmaster(codec, "Master Playback Switch", + NULL, slave_sws); + if (err < 0) + return err; + } + return 0; }
On Thu, Dec 20, 2007 at 04:46:35PM +0100, Takashi Iwai wrote:
static struct snd_kcontrol_new alc880_lg_mixer[] = { /* FIXME: it's not really "master" but front channels */
- HDA_CODEC_VOLUME("Master Playback Volume", 0x0f, 0x0, HDA_OUTPUT),
- HDA_BIND_MUTE("Master Playback Switch", 0x0f, 2, HDA_INPUT),
- HDA_CODEC_VOLUME("Front Playback Volume", 0x0f, 0x0, HDA_OUTPUT),
- HDA_BIND_MUTE("Front Playback Switch", 0x0f, 2, HDA_INPUT),
You've just fixed that FIXME :-)
At Sat, 12 Jan 2008 15:21:22 -0500, Daniel Jacobowitz wrote:
On Thu, Dec 20, 2007 at 04:46:35PM +0100, Takashi Iwai wrote:
static struct snd_kcontrol_new alc880_lg_mixer[] = { /* FIXME: it's not really "master" but front channels */
- HDA_CODEC_VOLUME("Master Playback Volume", 0x0f, 0x0, HDA_OUTPUT),
- HDA_BIND_MUTE("Master Playback Switch", 0x0f, 2, HDA_INPUT),
- HDA_CODEC_VOLUME("Front Playback Volume", 0x0f, 0x0, HDA_OUTPUT),
- HDA_BIND_MUTE("Front Playback Switch", 0x0f, 2, HDA_INPUT),
You've just fixed that FIXME :-)
Yeah, fixed FIXME now ;)
Takashi
This is the patch to alsa-driver tree. Simply adds a stub to build vmaster object.
Rebuild from scratch (e.g. ./hgcompile) after vmaster patches because you'll need to re-create configure script.
Takashi
diff -r 06162eef8936 acore/vmaster.c --- /dev/null Thu Jan 01 00:00:00 1970 +0000 +++ b/acore/vmaster.c Mon Oct 08 18:26:21 2007 +0200 @@ -0,0 +1,2 @@ +#define __NO_VERSION__ +#include "../alsa-kernel/core/vmaster.c"
[Resent as I don't see the message back; if you already received it, please disregard this]
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
Takashi
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
Jaroslav
----- Jaroslav Kysela perex@perex.cz Linux Kernel Sound Maintainer ALSA Project
Jaroslav Kysela wrote:
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
Aren't the patches supposed to be useful also for OSS-only apps, where userspace solutions can't work without LD_PRELOAD hacks? Currently, without the patch and the "aoss" wrapper, OSS apps sound too loud. I have not tested the patch, but I am too lazy to change each .desktop file for each OSS-only application provided by the distribution.
On Dec 22, 2007 2:39 AM, Alexander E. Patrakov patrakov@gmail.com wrote:
Aren't the patches supposed to be useful also for OSS-only apps, where userspace solutions can't work without LD_PRELOAD hacks? Currently, without the patch and the "aoss" wrapper, OSS apps sound too loud. I have not tested the patch, but I am too lazy to change each .desktop file for each OSS-only application provided by the distribution.
Please file a bug report with your distro. It should be considered a bug for a modern distro to ship OSS only apps withut using the aoss wrapper. It causes all sorts of problems, like blocking the soundcard for ALSA apps.
Lee
At Fri, 21 Dec 2007 20:37:09 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
A user-space implementation of virtual mixer elements would be far more complicated than the simplistic kernel-space patch. I've considered it many times, even tried to implement it, but got that conclusion. You'll see obviously the following difficulties:
1. Many user-space virtual elements
Each slave control element needs a virtual element (eventually a user-space one) because we need to keep both raw and virtual values to handle saturation. That is, the same number of additional controls would be added. Significant for 7.1 outputs.
2. Easy incosistency between virtual and raw elements
Even if the mixer abstraction hides the virtual elements, both virtual and raw elements are exposed on control API. This can cause the value-inconsistency between them quite easily, because many apps access directly via control API (even some mixer apps), and they likely change only raw values. The similar situation is for kernel OSS emulation.
3. Complicated configuration
The requirement of virtual master controls is very much dependent on the hardware. In the case of HD-audio, it depends on codec chip types, and even on the preset model chosen via PCI SSID or a module option. Implementing such a complex conditional in alsa-lib configuration is a clear overhead.
4. More resource requirement
Clearly a user-space solution requires more layers in alsa-lib, which is invoked by each process, and even more memory footprint than the kernel solution, not only the additional complexity.
... and think again what is the benifit of the user-space solution in this case, in comparison with the demerits above.
If you could implement the same feature on user-space with less amount of codes, I'd happily take it. But, I'm sure it's very hard.
Takashi
At Sat, 22 Dec 2007 15:55:19 +0100, I wrote:
At Fri, 21 Dec 2007 20:37:09 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
A user-space implementation of virtual mixer elements would be far more complicated than the simplistic kernel-space patch. I've considered it many times, even tried to implement it, but got that conclusion. You'll see obviously the following difficulties:
- Many user-space virtual elements
Each slave control element needs a virtual element (eventually a user-space one) because we need to keep both raw and virtual values to handle saturation. That is, the same number of additional controls would be added. Significant for 7.1 outputs.
- Easy incosistency between virtual and raw elements
Even if the mixer abstraction hides the virtual elements, both virtual and raw elements are exposed on control API. This can cause the value-inconsistency between them quite easily, because many apps access directly via control API (even some mixer apps), and they likely change only raw values. The similar situation is for kernel OSS emulation.
- Complicated configuration
The requirement of virtual master controls is very much dependent on the hardware. In the case of HD-audio, it depends on codec chip types, and even on the preset model chosen via PCI SSID or a module option. Implementing such a complex conditional in alsa-lib configuration is a clear overhead.
- More resource requirement
Clearly a user-space solution requires more layers in alsa-lib, which is invoked by each process, and even more memory footprint than the kernel solution, not only the additional complexity.
... and think again what is the benifit of the user-space solution in this case, in comparison with the demerits above.
If you could implement the same feature on user-space with less amount of codes, I'd happily take it. But, I'm sure it's very hard.
The master control is really important feature. Unless you have really a better patch, I'd like to commit the vmaster patch soon, so that more people can test.
thanks,
Takashi
On Sat, 22 Dec 2007, Takashi Iwai wrote:
At Fri, 21 Dec 2007 20:37:09 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
A user-space implementation of virtual mixer elements would be far more complicated than the simplistic kernel-space patch. I've considered it many times, even tried to implement it, but got that conclusion. You'll see obviously the following difficulties:
- Many user-space virtual elements
Each slave control element needs a virtual element (eventually a user-space one) because we need to keep both raw and virtual values to handle saturation. That is, the same number of additional controls would be added. Significant for 7.1 outputs.
Note that these "extra" values can be handled together using only one user-space control element without "Volume/Switch" in name.
- Easy incosistency between virtual and raw elements
Even if the mixer abstraction hides the virtual elements, both virtual and raw elements are exposed on control API. This can cause the value-inconsistency between them quite easily, because many apps access directly via control API (even some mixer apps), and they likely change only raw values. The similar situation is for kernel OSS emulation.
I have basically two ideas to handle this.
1) create a plugin system directly in the snd_ctl_* interface
- basically code might be very same code as you proposed in kernel
2) create virtual things only in smixer
- it's something I strongly prefer - if mixers uses CTL interface, it's not our bussiness because we haven't claimed to create virtual layers in CTL
- Complicated configuration
The requirement of virtual master controls is very much dependent on the hardware. In the case of HD-audio, it depends on codec chip types, and even on the preset model chosen via PCI SSID or a module option. Implementing such a complex conditional in alsa-lib configuration is a clear overhead.
I don't agree much here. Of course, we might need some hints from the driver (using information about components in sndrv_ctl_card_info) or we might analyze available controls using their names if Master control is present or we can eventually create a configuration tool saving hints to alsa-lib's configuration files which controls should be added for given hardware (this configuration tool might create a list of useable PCM devices handled in the driver, too). Just idea.
- More resource requirement
Clearly a user-space solution requires more layers in alsa-lib, which is invoked by each process, and even more memory footprint than the kernel solution, not only the additional complexity.
... and think again what is the benifit of the user-space solution in this case, in comparison with the demerits above.
The benefit is flexibility for future extensions.
If you could implement the same feature on user-space with less amount of codes, I'd happily take it. But, I'm sure it's very hard.
The question is also how to behave in future and if we should allow to break our basic rule in the current ALSA design - everything that can be moved to user space will be moved to user space.
Jaroslav
----- Jaroslav Kysela perex@perex.cz Linux Kernel Sound Maintainer ALSA Project, Red Hat, Inc.
At Thu, 10 Jan 2008 11:29:39 +0100 (CET), Jaroslav Kysela wrote:
On Sat, 22 Dec 2007, Takashi Iwai wrote:
At Fri, 21 Dec 2007 20:37:09 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
A user-space implementation of virtual mixer elements would be far more complicated than the simplistic kernel-space patch. I've considered it many times, even tried to implement it, but got that conclusion. You'll see obviously the following difficulties:
- Many user-space virtual elements
Each slave control element needs a virtual element (eventually a user-space one) because we need to keep both raw and virtual values to handle saturation. That is, the same number of additional controls would be added. Significant for 7.1 outputs.
Note that these "extra" values can be handled together using only one user-space control element without "Volume/Switch" in name.
This method is very likely fragile. Think what happens if the number of elements changes, or if the element numid changes. Then the packed values will be screwed up.
- Easy incosistency between virtual and raw elements
Even if the mixer abstraction hides the virtual elements, both virtual and raw elements are exposed on control API. This can cause the value-inconsistency between them quite easily, because many apps access directly via control API (even some mixer apps), and they likely change only raw values. The similar situation is for kernel OSS emulation.
I have basically two ideas to handle this.
create a plugin system directly in the snd_ctl_* interface
- basically code might be very same code as you proposed in kernel
No. The big difference is that the raw values are still exposed via control API (otherwise how do you read/write raw values?). That is, any user-space program may break consistency. OTOH, the kernel solution doesn't expose raw values.
create virtual things only in smixer
- it's something I strongly prefer
- if mixers uses CTL interface, it's not our bussiness because we haven't claimed to create virtual layers in CTL
Hm, I have to say it's a bad attitude. It's our business to keep the whole system consistent. We shouldn't build such a weak and fragile system any more.
And what about the kernel OSS mixer? There are still many apps to handle it.
- Complicated configuration
The requirement of virtual master controls is very much dependent on the hardware. In the case of HD-audio, it depends on codec chip types, and even on the preset model chosen via PCI SSID or a module option. Implementing such a complex conditional in alsa-lib configuration is a clear overhead.
I don't agree much here. Of course, we might need some hints from the driver (using information about components in sndrv_ctl_card_info) or we might analyze available controls using their names if Master control is present or we can eventually create a configuration tool saving hints to alsa-lib's configuration files which controls should be added for given hardware (this configuration tool might create a list of useable PCM devices handled in the driver, too). Just idea.
But it *is* an overhead.
- More resource requirement
Clearly a user-space solution requires more layers in alsa-lib, which is invoked by each process, and even more memory footprint than the kernel solution, not only the additional complexity.
... and think again what is the benifit of the user-space solution in this case, in comparison with the demerits above.
The benefit is flexibility for future extensions.
The demerit is complexity even longing for future :)
And the worst thing is that it is not implemented yet *at all*. How many years have we told that this kind of thing must be implemented in user-space?
Seriously, we need a solution RIGHT NOW.
If you could implement the same feature on user-space with less amount of codes, I'd happily take it. But, I'm sure it's very hard.
The question is also how to behave in future and if we should allow to break our basic rule in the current ALSA design - everything that can be moved to user space will be moved to user space.
Hey, it's just the question of balance. Insisting on a rule makes you blind. The rule exists not for following rules. The rule exists for creating a better system.
Note taht this isn't for solving every mixer problem. It doesn't solve the world hunger. But, it solves nicely the really annoying problem we face right now, without messy, fragile and overly complex implementation.
Takashi
On Thu, 10 Jan 2008, Takashi Iwai wrote:
At Thu, 10 Jan 2008 11:29:39 +0100 (CET), Jaroslav Kysela wrote:
On Sat, 22 Dec 2007, Takashi Iwai wrote:
At Fri, 21 Dec 2007 20:37:09 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
A user-space implementation of virtual mixer elements would be far more complicated than the simplistic kernel-space patch. I've considered it many times, even tried to implement it, but got that conclusion. You'll see obviously the following difficulties:
- Many user-space virtual elements
Each slave control element needs a virtual element (eventually a user-space one) because we need to keep both raw and virtual values to handle saturation. That is, the same number of additional controls would be added. Significant for 7.1 outputs.
Note that these "extra" values can be handled together using only one user-space control element without "Volume/Switch" in name.
This method is very likely fragile. Think what happens if the number of elements changes, or if the element numid changes. Then the packed values will be screwed up.
- Easy incosistency between virtual and raw elements
Even if the mixer abstraction hides the virtual elements, both virtual and raw elements are exposed on control API. This can cause the value-inconsistency between them quite easily, because many apps access directly via control API (even some mixer apps), and they likely change only raw values. The similar situation is for kernel OSS emulation.
I have basically two ideas to handle this.
create a plugin system directly in the snd_ctl_* interface
- basically code might be very same code as you proposed in kernel
No. The big difference is that the raw values are still exposed via control API (otherwise how do you read/write raw values?). That is, any user-space program may break consistency. OTOH, the kernel solution doesn't expose raw values.
Alsa-lib can filter raw writes from applications, but it's really not something I prefer, of course.
create virtual things only in smixer
- it's something I strongly prefer
- if mixers uses CTL interface, it's not our bussiness because we haven't claimed to create virtual layers in CTL
Hm, I have to say it's a bad attitude. It's our business to keep the whole system consistent. We shouldn't build such a weak and fragile system any more.
But the definition of consistency is relative in this case. If we have universal control API (which is not a mixer API, but just API to pass runtime configuration parameters to driver) and second one API (simple mixer - fully abstract) then everything is consistent.
And what about the kernel OSS mixer? There are still many apps to handle it.
Yes, OSS mixer support still requires a rework. My idea is to create a helper userspace application which will start automatically by udev when /dev/mixer devices are created passing values through smixer API in user space.
For OSS PCM emulation, similar thing can be done using snd-aloop module.
- Complicated configuration
The requirement of virtual master controls is very much dependent on the hardware. In the case of HD-audio, it depends on codec chip types, and even on the preset model chosen via PCI SSID or a module option. Implementing such a complex conditional in alsa-lib configuration is a clear overhead.
I don't agree much here. Of course, we might need some hints from the driver (using information about components in sndrv_ctl_card_info) or we might analyze available controls using their names if Master control is present or we can eventually create a configuration tool saving hints to alsa-lib's configuration files which controls should be added for given hardware (this configuration tool might create a list of useable PCM devices handled in the driver, too). Just idea.
But it *is* an overhead.
Maybe for this simple case. I would be really happy to see a system which allows me to split for example six channel hardware to three stereo soundcards with full abstract and independent mixers.
If we agree that it's the way to go, it does not make much sense to implement these things again in the kernel space, because virtual master implemented there becomes unuseable.
Note taht this isn't for solving every mixer problem. It doesn't solve the world hunger. But, it solves nicely the really annoying problem we face right now, without messy, fragile and overly complex implementation.
My point is that if we allow this now, we might have trouble to remove it in future when it becomes obsolete. It is much more difficult for the kernel code. We definitely need to settle a clear way about future mixer directions now and decide where the abstraction should be.
I'm sure I can hack smixer - simple_none.c - to add virtual master support in few days. If I fail for a reason or code is really ugly, I give you my ack for your implementation, ok?
Jaroslav
----- Jaroslav Kysela perex@perex.cz Linux Kernel Sound Maintainer ALSA Project, Red Hat, Inc.
At Thu, 10 Jan 2008 15:09:46 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 10 Jan 2008, Takashi Iwai wrote:
At Thu, 10 Jan 2008 11:29:39 +0100 (CET), Jaroslav Kysela wrote:
On Sat, 22 Dec 2007, Takashi Iwai wrote:
At Fri, 21 Dec 2007 20:37:09 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 20 Dec 2007, Takashi Iwai wrote:
Hi,
another thing I'd like to push into the next kernel is the virtual master volume control. As I posted in earlier posts, it adds virtual controls that have several slave controls with the same types, e.g. Front, Surround, Center, LFE, etc. Then these are adjusted simultaneously via the master control.
It'd be appreciated if some one can test the patches with HD-audio h/w that has no master control yet (e.g. some STAC codecs).
Note that this won't add the master volumes if there are no real volume controls. Some codecs have really no volume control, and this won't help for such devices.
Two (and one for driver) patches will follow after this.
NAK from my side. I am convinced that this code can be implemented in the user space even without any daemon just in the mixer abstract layer using standard control elements and using eventually new user controls to store data for virtual mixer controls.
A user-space implementation of virtual mixer elements would be far more complicated than the simplistic kernel-space patch. I've considered it many times, even tried to implement it, but got that conclusion. You'll see obviously the following difficulties:
- Many user-space virtual elements
Each slave control element needs a virtual element (eventually a user-space one) because we need to keep both raw and virtual values to handle saturation. That is, the same number of additional controls would be added. Significant for 7.1 outputs.
Note that these "extra" values can be handled together using only one user-space control element without "Volume/Switch" in name.
This method is very likely fragile. Think what happens if the number of elements changes, or if the element numid changes. Then the packed values will be screwed up.
- Easy incosistency between virtual and raw elements
Even if the mixer abstraction hides the virtual elements, both virtual and raw elements are exposed on control API. This can cause the value-inconsistency between them quite easily, because many apps access directly via control API (even some mixer apps), and they likely change only raw values. The similar situation is for kernel OSS emulation.
I have basically two ideas to handle this.
create a plugin system directly in the snd_ctl_* interface
- basically code might be very same code as you proposed in kernel
No. The big difference is that the raw values are still exposed via control API (otherwise how do you read/write raw values?). That is, any user-space program may break consistency. OTOH, the kernel solution doesn't expose raw values.
Alsa-lib can filter raw writes from applications, but it's really not something I prefer, of course.
Right. It'd be really bad to do that.
create virtual things only in smixer
- it's something I strongly prefer
- if mixers uses CTL interface, it's not our bussiness because we haven't claimed to create virtual layers in CTL
Hm, I have to say it's a bad attitude. It's our business to keep the whole system consistent. We shouldn't build such a weak and fragile system any more.
But the definition of consistency is relative in this case. If we have universal control API (which is not a mixer API, but just API to pass runtime configuration parameters to driver) and second one API (simple mixer - fully abstract) then everything is consistent.
What I meant is that this would create bogus values easily, that is, screw up the volumes. The kernel OSS can change the raw value while mixer changes the value at the same time.
And what about the kernel OSS mixer? There are still many apps to handle it.
Yes, OSS mixer support still requires a rework. My idea is to create a helper userspace application which will start automatically by udev when /dev/mixer devices are created passing values through smixer API in user space.
For OSS PCM emulation, similar thing can be done using snd-aloop module.
Yeah, that's good. But this needs a lot of work and more resources. And don't forget that many people hate daemons.
- Complicated configuration
The requirement of virtual master controls is very much dependent on the hardware. In the case of HD-audio, it depends on codec chip types, and even on the preset model chosen via PCI SSID or a module option. Implementing such a complex conditional in alsa-lib configuration is a clear overhead.
I don't agree much here. Of course, we might need some hints from the driver (using information about components in sndrv_ctl_card_info) or we might analyze available controls using their names if Master control is present or we can eventually create a configuration tool saving hints to alsa-lib's configuration files which controls should be added for given hardware (this configuration tool might create a list of useable PCM devices handled in the driver, too). Just idea.
But it *is* an overhead.
Maybe for this simple case.
Exactly, that's the point. This case is DAMN SIMPLE.
The virtual master patch simply fills the hole in the driver. It should have been there, but I didn't put it. That's my fault.
It has nothing to do with the virtual cards or complicated layering. It just adds a control for binding several registers. We don't need too much work for this simple purpose.
I would be really happy to see a system which allows me to split for example six channel hardware to three stereo soundcards with full abstract and independent mixers.
I'm afraid this would result in a similar situation like PCM plugins. Without a proper setup utility, the whole thing could be too complicated to use in practice...
If we agree that it's the way to go, it does not make much sense to implement these things again in the kernel space, because virtual master implemented there becomes unuseable.
The virtual master in the kernel doesn't conflict with the user-space additions. It's simply a master volume in the end, nothing different from the real master volume on hardware. So, it's fine to have both.
Note taht this isn't for solving every mixer problem. It doesn't solve the world hunger. But, it solves nicely the really annoying problem we face right now, without messy, fragile and overly complex implementation.
My point is that if we allow this now, we might have trouble to remove it in future when it becomes obsolete. It is much more difficult for the kernel code. We definitely need to settle a clear way about future mixer directions now and decide where the abstraction should be.
It doesn't have to be obsoleted. The user-space solution requires more resources for the same purpose. Then why do we need to switch? And above all, both can coexist peacefully.
I'm sure I can hack smixer - simple_none.c - to add virtual master support in few days. If I fail for a reason or code is really ugly, I give you my ack for your implementation, ok?
Well, as I wrote, smixer alone won't solve all the corner cases. Imlementing smixer would be relatively easy. But it can't be solid like the kernel driver, and we have still problems with control API and kernel mixer OSS.
Takashi
On Thu, 10 Jan 2008, Takashi Iwai wrote:
I'm sure I can hack smixer - simple_none.c - to add virtual master support in few days. If I fail for a reason or code is really ugly, I give you my ack for your implementation, ok?
Well, as I wrote, smixer alone won't solve all the corner cases. Imlementing smixer would be relatively easy. But it can't be solid like the kernel driver, and we have still problems with control API and kernel mixer OSS.
Ok, I switch to rather neutral mode in this issue. Although I don't think that it's the right solution. You can commit this code to repo. Time shows us if it was good or not. Back to coding.
Jaroslav
----- Jaroslav Kysela perex@perex.cz Linux Kernel Sound Maintainer ALSA Project, Red Hat, Inc.
At Thu, 10 Jan 2008 16:43:24 +0100 (CET), Jaroslav Kysela wrote:
On Thu, 10 Jan 2008, Takashi Iwai wrote:
I'm sure I can hack smixer - simple_none.c - to add virtual master support in few days. If I fail for a reason or code is really ugly, I give you my ack for your implementation, ok?
Well, as I wrote, smixer alone won't solve all the corner cases. Imlementing smixer would be relatively easy. But it can't be solid like the kernel driver, and we have still problems with control API and kernel mixer OSS.
Ok, I switch to rather neutral mode in this issue. Although I don't think that it's the right solution. You can commit this code to repo. Time shows us if it was good or not. Back to coding.
Glad to hear that.
The user-space solution would be of course nice, but in this particular case, it results in far more codes and may give even more problems. Once after we get a better infrastructure for mixer abstraction, let's clean up the whole stuff.
Takashi
participants (5)
-
Alexander E. Patrakov -
Daniel Jacobowitz -
Jaroslav Kysela -
Lee Revell -
Takashi Iwai