| Message ID | 1375180329-4860-3-git-send-email-richard.genoud@gmail.com (mailing list archive) |
|---|---|
| State | New, archived |
| Headers | show |
On 07/30/2013 04:32 AM, Richard Genoud wrote: > From: Nicolas Ferre <nicolas.ferre@atmel.com> > > Description of the Asoc machine driver for an at91sam9x5 based board > with a wm8731 audio DAC. Wm8731 is clocked by a crystal and used as a > master on the SSC/I2S interface. Its connections are a headphone jack > and an Line input jack. > > [Richard: this is based on an old patch from Nicolas that I forward > ported and reworked to use only device tree] > diff --git a/Documentation/devicetree/bindings/sound/atmel-sam9x5-wm8731-audio.txt b/Documentation/devicetree/bindings/sound/atmel-sam9x5-wm8731-audio.txt > +* Atmel at91sam9x5ek wm8731 audio complex > + > +Required properties: > + - compatible: "atmel,sam9x5-wm8731-audio" > + - atmel,model: The user-visible name of this sound complex. > + - atmel,ssc-controller: The phandle of the SSC controller > + - atmel,audio-codec: The phandle of the WM8731 audio codec > + - atmel,audio-routing: A list of the connections between audio components. > + Each entry is a pair of strings, the first being the connection's sink, > + the second being the connection's source. > + > +Available audio endpoints for the audio-routing table: > + > +Board connectors: > + * Headphone Jack > + * Line In Jack > + > +wm8731 pins: > +cf Documentation/devicetree/bindings/sound/wm8731.txt In the Tegra bindings, I deliberately put the list of CODEC pins into the audio complex binding rather than the CODEC binding. That's because I'm not sure that in the long-term we want to use strings to identify the CODEC pins, rather than using integers. By putting the list orf CODEC pins in the audio complex binding rather than the CODEC binding, I didn't lumber the CODEC binding with a list of strings that it had to support forever. One reason that strings are problematic is because they can't be mixed with integers/phandles in the same property, so if we ever end up with more generic audio bindings where the routing table is expressed more like: audio-routes = <&component_a $a_pin &component_b $b_pin>; ... in order to allow completely arbitrary and fully name-spaced routing specification[1], then $a_pin and $b_pin need to be integers not strings. [1] and perhaps we can get rid of properties like audio-codec/ssc-controller, and automatically deduce which components are needed simply by finding all phandles in the audio-routes property. Perhaps the solution here is to allow mixing phandles/integers/strings in one property, but that's potentially quite a large change to the DTB format; we'd need to introduce type fields into the property data, and other data format changes.
On Tue, Jul 30, 2013 at 11:45:12AM -0600, Stephen Warren wrote: > On 07/30/2013 04:32 AM, Richard Genoud wrote: > > +wm8731 pins: > > +cf Documentation/devicetree/bindings/sound/wm8731.txt > In the Tegra bindings, I deliberately put the list of CODEC pins into > the audio complex binding rather than the CODEC binding. That's because You really shouldn't do this, it's not accomplishing much. > I'm not sure that in the long-term we want to use strings to identify > the CODEC pins, rather than using integers. By putting the list orf > CODEC pins in the audio complex binding rather than the CODEC binding, I > didn't lumber the CODEC binding with a list of strings that it had to > support forever. How would you create the numbers - you can't use the pin numbers since BGA type packages have alphanumeric ball identifiers? Even with numerical pin numbering ou'd need to specify some defines for this not to be totally awful at which point you may as well have the strings documented since you'd end up writing a table in the binding document that's basically a mapping of pin names to numbers, > One reason that strings are problematic is because they can't be mixed > with integers/phandles in the same property, so if we ever end up with > more generic audio bindings where the routing table is expressed more like: > audio-routes = <&component_a $a_pin &component_b $b_pin>; > ... in order to allow completely arbitrary and fully name-spaced routing > specification[1], then $a_pin and $b_pin need to be integers not strings. The above is going to be legibility challenged without defines at which point the strings end up appearing in the binding document anyway. Besides, you're stuck with the names you picked anyway so you may as well put them in the CODEC driver binding document so that anyone else using the same CODEC can reuse that bit of work. > [1] and perhaps we can get rid of properties like > audio-codec/ssc-controller, and automatically deduce which components > are needed simply by finding all phandles in the audio-routes property. > Perhaps the solution here is to allow mixing phandles/integers/strings > in one property, but that's potentially quite a large change to the DTB > format; we'd need to introduce type fields into the property data, and > other data format changes. I really do think this is a DT failure which ought to be addressed there; people are working around this with things like parallel arrays with names and phandles which aren't awesome once the arrays get to be any size. One option that seems sensible is to introduce syntactic sugar which will do the parallel arrays trick automatically - the actual DT that gets output could still be two arrays with indexes that match up (so no impact on parsers) but the DT author would be able to write an array of (phandle, string) tuples or something instead of having to line up two arrays.
On 07/30/2013 02:45 PM, Mark Brown wrote: > On Tue, Jul 30, 2013 at 11:45:12AM -0600, Stephen Warren wrote: >> On 07/30/2013 04:32 AM, Richard Genoud wrote: > >>> +wm8731 pins: +cf >>> Documentation/devicetree/bindings/sound/wm8731.txt > >> In the Tegra bindings, I deliberately put the list of CODEC pins >> into the audio complex binding rather than the CODEC binding. >> That's because > > You really shouldn't do this, it's not accomplishing much. What it accomplishes is not polluting the CODEC's binding with a set of strings that must always be supported since DT is an ABI. The strings are isolated into the specific audio complex binding, which might possibly become deprecated and replaced with something better and more generic. >> I'm not sure that in the long-term we want to use strings to >> identify the CODEC pins, rather than using integers. By putting >> the list orf CODEC pins in the audio complex binding rather than >> the CODEC binding, I didn't lumber the CODEC binding with a list >> of strings that it had to support forever. > > How would you create the numbers - you can't use the pin numbers > since BGA type packages have alphanumeric ball identifiers? The numbering would be arbitrary, except perhaps for packages that had simple numeric pin IDs in which case those could be used. There would be a file in <dt-bindings/audio/...> that defined it, which DTs would include to create the properties, and CODEC/... drivers would include and use whenever it needed to identify the pins, e.g. in some kind of .of_xlate() function. > Even with numerical pin numbering ou'd need to specify some defines > for this not to be totally awful at which point you may as well > have the strings documented since you'd end up writing a table in > the binding document that's basically a mapping of pin names to > numbers, > >> One reason that strings are problematic is because they can't be >> mixed with integers/phandles in the same property, so if we ever >> end up with more generic audio bindings where the routing table >> is expressed more like: > >> audio-routes = <&component_a $a_pin &component_b $b_pin>; > >> ... in order to allow completely arbitrary and fully name-spaced >> routing specification[1], then $a_pin and $b_pin need to be >> integers not strings. > > The above is going to be legibility challenged without defines at > which point the strings end up appearing in the binding document > anyway. Yes, you'd certainly have to use defines not raw integers. The strings wouldn't appear in the binding doc so much as the numbers.
On Tue, Jul 30, 2013 at 03:29:17PM -0600, Stephen Warren wrote: > On 07/30/2013 02:45 PM, Mark Brown wrote: > > You really shouldn't do this, it's not accomplishing much. > What it accomplishes is not polluting the CODEC's binding with a set > of strings that must always be supported since DT is an ABI. The > strings are isolated into the specific audio complex binding, which > might possibly become deprecated and replaced with something better > and more generic. You seem to be assuming that the strings are a bad thing. I'm not sure that this is the case modulo the tooling issues... Besides the reality is that we're using strings right now and they're being baked into bindings right now, I can't see anything being saved by listing them in the board binding rather than in the CODEC binding since both are ABIs especially once you get to the second board binding. The ship has sailed on that one. All the CODEC binding does is say "there are pins called this", it's up to some other binding to define how they're used and the CODEC binding can always add extra information to that list, including a mapping to integers. > > The above is going to be legibility challenged without defines at > > which point the strings end up appearing in the binding document > > anyway. > Yes, you'd certainly have to use defines not raw integers. The strings > wouldn't appear in the binding doc so much as the numbers. The strings would be there in the text explaining what the numbers mean - whatever happens the pin names are going to be physically present in the document, it's just a formatting difference. The CODEC document has to list all the pins, we're not going to avoid that by defining a mapping to integers since adding the integers will in some way end up defining the names even if just as part of the name of a constant. We could start adding the integers right now - something like: The X CODEC has the following numbered input and output pins: 1. HPOUTL 2. HPOUTR ... (see datasheet for details), a header file is provided with constants X_PIN_foo for convenience. would work with either approach to identifying the pins or if we're being less verbose and just reference the header file for the definitions that becomes something like: The X CODEC has input and output pins listed with numerical identifiers in the form the X_PIN_foo defined in the X.h header file where foo is the name of the pin. (that's not good boilerplate wording but you get my drift) which still ends up defining a set of known strings for pins. In fact now that I think about this why don't we just go ahead and do all this, starting by putting the numbers into the bindings for the CODECs since that's the simplest thing and doesn't involve writing code? I even have several boards on my desk that run DT ASoC...
On 07/30/2013 04:58 PM, Mark Brown wrote: > On Tue, Jul 30, 2013 at 03:29:17PM -0600, Stephen Warren wrote: >> On 07/30/2013 02:45 PM, Mark Brown wrote: > >>> You really shouldn't do this, it's not accomplishing much. > >> What it accomplishes is not polluting the CODEC's binding with a >> set of strings that must always be supported since DT is an ABI. >> The strings are isolated into the specific audio complex binding, >> which might possibly become deprecated and replaced with >> something better and more generic. > > You seem to be assuming that the strings are a bad thing. I'm not > sure that this is the case modulo the tooling issues... I do tend to think that using strings is pretty evil... > We could start adding the integers right now - something like: > > The X CODEC has the following numbered input and output pins: > > 1. HPOUTL 2. HPOUTR ... > > (see datasheet for details), a header file is provided with > constants X_PIN_foo for convenience. > > would work with either approach to identifying the pins or if > we're being less verbose and just reference the header file for > the definitions that becomes something like: > > The X CODEC has input and output pins listed with numerical > identifiers in the form the X_PIN_foo defined in the X.h header > file where foo is the name of the pin. > > (that's not good boilerplate wording but you get my drift) which > still ends up defining a set of known strings for pins. > > In fact now that I think about this why don't we just go ahead and > do all this, starting by putting the numbers into the bindings for > the CODECs since that's the simplest thing and doesn't involve > writing code? I even have several boards on my desk that run DT > ASoC... I'm certainly fine with that compromise.
On Wed, Jul 31, 2013 at 04:16:28PM -0600, Stephen Warren wrote: > On 07/30/2013 04:58 PM, Mark Brown wrote: > > You seem to be assuming that the strings are a bad thing. I'm not > > sure that this is the case modulo the tooling issues... > I do tend to think that using strings is pretty evil... Well, they do avoid the legibility issues that tend to plague device tree - I don't really mind if they're in the property names (as with regulator supply properties) or in the data but having something that is directly legible is a substantial win. Constants deal with this in the source but they make runtime debugging more painful which isn't ideal.
On Tue, Jul 30, 2013 at 12:32:03PM +0200, Richard Genoud wrote: > From: Nicolas Ferre <nicolas.ferre@atmel.com> > > Description of the Asoc machine driver for an at91sam9x5 based board > with a wm8731 audio DAC. Wm8731 is clocked by a crystal and used as a > master on the SSC/I2S interface. Its connections are a headphone jack > and an Line input jack. Applied, thanks. We should really be able to merge this with the sam9g20ek driver that's here already but we can do that later.
2013/8/6 Mark Brown <broonie@kernel.org>: > On Tue, Jul 30, 2013 at 12:32:03PM +0200, Richard Genoud wrote: >> From: Nicolas Ferre <nicolas.ferre@atmel.com> >> >> Description of the Asoc machine driver for an at91sam9x5 based board >> with a wm8731 audio DAC. Wm8731 is clocked by a crystal and used as a >> master on the SSC/I2S interface. Its connections are a headphone jack >> and an Line input jack. > > Applied, thanks. We should really be able to merge this with the > sam9g20ek driver that's here already but we can do that later. I think it will be easier after the switch to the common clock framework. I'll look into that. Thanks !
diff --git a/Documentation/devicetree/bindings/sound/atmel-sam9x5-wm8731-audio.txt b/Documentation/devicetree/bindings/sound/atmel-sam9x5-wm8731-audio.txt new file mode 100644 index 0000000..0720857 --- /dev/null +++ b/Documentation/devicetree/bindings/sound/atmel-sam9x5-wm8731-audio.txt @@ -0,0 +1,35 @@ +* Atmel at91sam9x5ek wm8731 audio complex + +Required properties: + - compatible: "atmel,sam9x5-wm8731-audio" + - atmel,model: The user-visible name of this sound complex. + - atmel,ssc-controller: The phandle of the SSC controller + - atmel,audio-codec: The phandle of the WM8731 audio codec + - atmel,audio-routing: A list of the connections between audio components. + Each entry is a pair of strings, the first being the connection's sink, + the second being the connection's source. + +Available audio endpoints for the audio-routing table: + +Board connectors: + * Headphone Jack + * Line In Jack + +wm8731 pins: +cf Documentation/devicetree/bindings/sound/wm8731.txt + +Example: +sound { + compatible = "atmel,sam9x5-wm8731-audio"; + + atmel,model = "wm8731 @ AT91SAM9X5EK"; + + atmel,audio-routing = + "Headphone Jack", "RHPOUT", + "Headphone Jack", "LHPOUT", + "LLINEIN", "Line In Jack", + "RLINEIN", "Line In Jack"; + + atmel,ssc-controller = <&ssc0>; + atmel,audio-codec = <&wm8731>; +}; diff --git a/Documentation/devicetree/bindings/sound/wm8731.txt b/Documentation/devicetree/bindings/sound/wm8731.txt index 15f7004..236690e 100644 --- a/Documentation/devicetree/bindings/sound/wm8731.txt +++ b/Documentation/devicetree/bindings/sound/wm8731.txt @@ -16,3 +16,12 @@ codec: wm8731@1a { compatible = "wlf,wm8731"; reg = <0x1a>; }; + +Available audio endpoints for an audio-routing table: + * LOUT: Left Channel Line Output + * ROUT: Right Channel Line Output + * LHPOUT: Left Channel Headphone Output + * RHPOUT: Right Channel Headphone Output + * LLINEIN: Left Channel Line Input + * RLINEIN: Right Channel Line Input + * MICIN: Microphone Input diff --git a/sound/soc/atmel/Kconfig b/sound/soc/atmel/Kconfig index 986323b..e48d38a 100644 --- a/sound/soc/atmel/Kconfig +++ b/sound/soc/atmel/Kconfig @@ -43,6 +43,16 @@ config SND_ATMEL_SOC_WM8904 Say Y if you want to add support for Atmel ASoC driver for boards using WM8904 codec. +config SND_AT91_SOC_SAM9X5_WM8731 + tristate "SoC Audio support for WM8731-based at91sam9x5 board" + depends on ATMEL_SSC && SND_ATMEL_SOC && SOC_AT91SAM9X5 + select SND_ATMEL_SOC_SSC + select SND_ATMEL_SOC_DMA + select SND_SOC_WM8731 + help + Say Y if you want to add support for audio SoC on an + at91sam9x5 based board that is using WM8731 codec. + config SND_AT91_SOC_AFEB9260 tristate "SoC Audio support for AFEB9260 board" depends on ARCH_AT91 && ATMEL_SSC && ARCH_AT91 && MACH_AFEB9260 && SND_ATMEL_SOC diff --git a/sound/soc/atmel/Makefile b/sound/soc/atmel/Makefile index 922d4da..5baabc8 100644 --- a/sound/soc/atmel/Makefile +++ b/sound/soc/atmel/Makefile @@ -12,7 +12,9 @@ obj-$(CONFIG_SND_ATMEL_SOC_SSC) += snd-soc-atmel_ssc_dai.o # AT91 Machine Support snd-soc-sam9g20-wm8731-objs := sam9g20_wm8731.o snd-atmel-soc-wm8904-objs := atmel_wm8904.o +snd-soc-sam9x5-wm8731-objs := sam9x5_wm8731.o obj-$(CONFIG_SND_AT91_SOC_SAM9G20_WM8731) += snd-soc-sam9g20-wm8731.o obj-$(CONFIG_SND_ATMEL_SOC_WM8904) += snd-atmel-soc-wm8904.o +obj-$(CONFIG_SND_AT91_SOC_SAM9X5_WM8731) += snd-soc-sam9x5-wm8731.o obj-$(CONFIG_SND_AT91_SOC_AFEB9260) += snd-soc-afeb9260.o diff --git a/sound/soc/atmel/sam9x5_wm8731.c b/sound/soc/atmel/sam9x5_wm8731.c new file mode 100644 index 0000000..992ae38 --- /dev/null +++ b/sound/soc/atmel/sam9x5_wm8731.c @@ -0,0 +1,208 @@ +/* + * sam9x5_wm8731 -- SoC audio for AT91SAM9X5-based boards + * that are using WM8731 as codec. + * + * Copyright (C) 2011 Atmel, + * Nicolas Ferre <nicolas.ferre@atmel.com> + * + * Copyright (C) 2013 Paratronic, + * Richard Genoud <richard.genoud@gmail.com> + * + * Based on sam9g20_wm8731.c by: + * Sedji Gaouaou <sedji.gaouaou@atmel.com> + * + * 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. + * + */ +#include <linux/of.h> +#include <linux/export.h> +#include <linux/module.h> +#include <linux/mod_devicetable.h> +#include <linux/platform_device.h> +#include <linux/device.h> + +#include <sound/soc.h> +#include <sound/soc-dai.h> +#include <sound/soc-dapm.h> + +#include "../codecs/wm8731.h" +#include "atmel_ssc_dai.h" + + +#define MCLK_RATE 12288000 + +#define DRV_NAME "sam9x5-snd-wm8731" + +struct sam9x5_drvdata { + int ssc_id; +}; + +/* + * Logic for a wm8731 as connected on a at91sam9x5ek based board. + */ +static int sam9x5_wm8731_init(struct snd_soc_pcm_runtime *rtd) +{ + struct snd_soc_dai *codec_dai = rtd->codec_dai; + struct device *dev = rtd->dev; + int ret; + + dev_dbg(dev, "ASoC: %s called\n", __func__); + + /* set the codec system clock for DAC and ADC */ + ret = snd_soc_dai_set_sysclk(codec_dai, WM8731_SYSCLK_XTAL, + MCLK_RATE, SND_SOC_CLOCK_IN); + if (ret < 0) { + dev_err(dev, "ASoC: Failed to set WM8731 SYSCLK: %d\n", ret); + return ret; + } + + return 0; +} + +/* + * Audio paths on at91sam9x5ek board: + * + * |A| ------------> | | ---R----> Headphone Jack + * |T| <----\ | WM | ---L--/ + * |9| ---> CLK <--> | 8731 | <--R----- Line In Jack + * |1| <------------ | | <--L--/ + */ +static const struct snd_soc_dapm_widget sam9x5_dapm_widgets[] = { + SND_SOC_DAPM_HP("Headphone Jack", NULL), + SND_SOC_DAPM_LINE("Line In Jack", NULL), +}; + +static int sam9x5_wm8731_driver_probe(struct platform_device *pdev) +{ + struct device_node *np = pdev->dev.of_node; + struct device_node *codec_np, *cpu_np; + struct snd_soc_card *card; + struct snd_soc_dai_link *dai; + struct sam9x5_drvdata *priv; + int ret; + + if (!np) { + dev_err(&pdev->dev, "No device node supplied\n"); + return -EINVAL; + } + + card = devm_kzalloc(&pdev->dev, sizeof(*card), GFP_KERNEL); + priv = devm_kzalloc(&pdev->dev, sizeof(*priv), GFP_KERNEL); + dai = devm_kzalloc(&pdev->dev, sizeof(*dai), GFP_KERNEL); + if (!dai || !card || !priv) { + ret = -ENOMEM; + goto out; + } + + card->dev = &pdev->dev; + card->owner = THIS_MODULE; + card->dai_link = dai; + card->num_links = 1; + card->dapm_widgets = sam9x5_dapm_widgets; + card->num_dapm_widgets = ARRAY_SIZE(sam9x5_dapm_widgets); + dai->name = "WM8731"; + dai->stream_name = "WM8731 PCM"; + dai->codec_dai_name = "wm8731-hifi"; + dai->init = sam9x5_wm8731_init; + dai->dai_fmt = SND_SOC_DAIFMT_I2S | SND_SOC_DAIFMT_NB_NF + | SND_SOC_DAIFMT_CBM_CFM; + + ret = snd_soc_of_parse_card_name(card, "atmel,model"); + if (ret) { + dev_err(&pdev->dev, "atmel,model node missing\n"); + goto out; + } + + ret = snd_soc_of_parse_audio_routing(card, "atmel,audio-routing"); + if (ret) { + dev_err(&pdev->dev, "atmel,audio-routing node missing\n"); + goto out; + } + + codec_np = of_parse_phandle(np, "atmel,audio-codec", 0); + if (!codec_np) { + dev_err(&pdev->dev, "atmel,audio-codec node missing\n"); + ret = -EINVAL; + goto out; + } + + dai->codec_of_node = codec_np; + + cpu_np = of_parse_phandle(np, "atmel,ssc-controller", 0); + if (!cpu_np) { + dev_err(&pdev->dev, "atmel,ssc-controller node missing\n"); + ret = -EINVAL; + goto out; + } + dai->cpu_of_node = cpu_np; + dai->platform_of_node = cpu_np; + + priv->ssc_id = of_alias_get_id(cpu_np, "ssc"); + + ret = atmel_ssc_set_audio(priv->ssc_id); + if (ret != 0) { + dev_err(&pdev->dev, + "ASoC: Failed to set SSC %d for audio: %d\n", + ret, priv->ssc_id); + goto out; + } + + of_node_put(codec_np); + of_node_put(cpu_np); + + platform_set_drvdata(pdev, card); + + ret = snd_soc_register_card(card); + if (ret) { + dev_err(&pdev->dev, + "ASoC: Platform device allocation failed\n"); + goto out_put_audio; + } + + dev_dbg(&pdev->dev, "ASoC: %s ok\n", __func__); + + return ret; + +out_put_audio: + atmel_ssc_put_audio(priv->ssc_id); +out: + return ret; +} + +static int sam9x5_wm8731_driver_remove(struct platform_device *pdev) +{ + struct snd_soc_card *card = platform_get_drvdata(pdev); + struct sam9x5_drvdata *priv = card->drvdata; + + snd_soc_unregister_card(card); + atmel_ssc_put_audio(priv->ssc_id); + + return 0; +} + +static const struct of_device_id sam9x5_wm8731_of_match[] = { + { .compatible = "atmel,sam9x5-wm8731-audio", }, + {}, +}; +MODULE_DEVICE_TABLE(of, sam9x5_wm8731_of_match); + +static struct platform_driver sam9x5_wm8731_driver = { + .driver = { + .name = DRV_NAME, + .owner = THIS_MODULE, + .of_match_table = of_match_ptr(sam9x5_wm8731_of_match), + }, + .probe = sam9x5_wm8731_driver_probe, + .remove = sam9x5_wm8731_driver_remove, +}; +module_platform_driver(sam9x5_wm8731_driver); + +/* Module information */ +MODULE_AUTHOR("Nicolas Ferre <nicolas.ferre@atmel.com>"); +MODULE_AUTHOR("Richard Genoud <richard.genoud@gmail.com>"); +MODULE_DESCRIPTION("ALSA SoC machine driver for AT91SAM9x5 - WM8731"); +MODULE_LICENSE("GPL"); +MODULE_ALIAS("platform:" DRV_NAME);