March 2014 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH 0/19] ASoC: rsnd: new module path method and DT support
by Kuninori Morimoto 25 Mar '14

25 Mar '14

Hi Mark These are Renesas R-Car sound driver patches. #1 - #7 are clean up patches. #8 - #18 adds new module path method #19 adds DT support Now, Mark's ALSA branch doesn't include SH-ARM platform branch. Thus, these patches are keeping compatible for it at this point. Compatible code will be removed after merge. Kuninori Morimoto (19): 1) ASoC: rsnd: move priv member settings to upper side 2) ASoC: rsnd: move rsnd_mod_call() macro 3) ASoC: rsnd: remove verbose function parameter 4) ASoC: rsnd: remove verbose debug message from scu/ssi 5) ASoC: rsnd: unify rdai naming 6) ASoC: rsnd: add .set_sysclk on snd_soc_dai_ops 7) ASoC: rsnd: tidyup RSND_SSI_xxx flags 8) ASoC: rsnd: run rsnd_path_init() when probe() timing 9) ASoC: rsnd: use mod array instead of list on rdai 10) ASoC: rsnd: get ssi/scu from rsnd_dai_stream 11) ASoC: rsnd: add struct rsnd_dai_platform_info 12) ASoC: rsnd: use function pointer for each probe 13) ASoC: rsnd: Get correct SCU ID 14) ASoC: rsnd: add rsnd_scu_enable_ssi_irq() 15) ASoC: rsnd: call rsnd_scu_ssi_mode_init() from SSI 16) ASoC: rsnd: add probe/remove callback on rsnd_mod_ops 17) ASoC: rsnd: use mod probe method on SCU 18) ASoC: rsnd: use mod probe method on SSI 19) ASoC: rsnd: add DeviceTree support .../devicetree/bindings/sound/renesas,rsnd.txt | 101 ++++++ include/sound/rcar_snd.h | 21 +- sound/soc/sh/rcar/adg.c | 2 +- sound/soc/sh/rcar/core.c | 380 ++++++++++++++++---- sound/soc/sh/rcar/gen.c | 78 +--- sound/soc/sh/rcar/rsnd.h | 86 +++-- sound/soc/sh/rcar/scu.c | 188 ++++++---- sound/soc/sh/rcar/ssi.c | 191 +++++++--- 8 files changed, 759 insertions(+), 288 deletions(-)

3 67

[alsa-devel] [PATCH v2 00/12] Cleanup card DAPM and control setup
by Lars-Peter Clausen 25 Mar '14

25 Mar '14

This patch series converts all remaining machine drivers that still register their controls and DAPM elements with the CODEC to registering them with the card. The reason for doing this is to be able to remove the hack in the ASoC core that un-sets the name_prefix of a CODEC before calling the init callback for a DAI link. This comes from a time where it was not possible to register controls or DAPM elements at a card level. And removing this hack makes going forward with the componentisation easier and it also looks like the DAI multi-CODEC patch could benefit from this as well. The series depends on topic/pxa and topic/dapm - Lars Lars-Peter Clausen (12): ASoC: ams-delta: Fix compile error ASoC: ams-delta: Convert to table based DAPM and control setup ASoC: mfld_machine: Convert to table based DAPM and control setup ASoC: omap-abe-twl6040: Register machine level DMIC DAPM routes with the card ASoC: omap: rx51: Convert to table based control and DAPM setup ASoC: omap3pandora: Convert to table based DAPM setup ASoC: davinci-evm: Register machine level DAPM elements with the card ASoC: snappercl15: Convert to table based DAPM setup ASoC: s6105-ipcam: Convert to table based DAPM setup ASoC: pxa: magician: Convert to table based DAPM and control setup ASoC: pxa: tosa: Convert to table based DAPM and control setup ASoC: Remove name_prefix un-set during DAI link init hack sound/soc/cirrus/snappercl15.c | 18 +++-------- sound/soc/davinci/davinci-evm.c | 22 +++++-------- sound/soc/intel/mfld_machine.c | 49 ++++++++++++----------------- sound/soc/omap/ams-delta.c | 66 +++++++++++++-------------------------- sound/soc/omap/omap-abe-twl6040.c | 3 +- sound/soc/omap/omap3pandora.c | 31 ++++++------------ sound/soc/omap/rx51.c | 47 +++++----------------------- sound/soc/pxa/magician.c | 34 +++++++------------- sound/soc/pxa/tosa.c | 35 ++++++++------------- sound/soc/s6000/s6105-ipcam.c | 28 ++++++----------- sound/soc/soc-core.c | 7 +---- 11 files changed, 108 insertions(+), 232 deletions(-) -- 1.8.0

3 25

[alsa-devel] [PATCH v3 0/3] ASoC: davinci: perparation for eDMA dmaengine PCM
by Peter Ujfalusi 25 Mar '14

25 Mar '14

Hi, Changes since v2: - Constraint only placed for the period size [1] - Patch for davinci-pcm.h to add empty function in case of !DAVINCI platforms When the edma arch and dmaengine patches has been taken only one patch is needed to switch AM335x/AM437x to use the edma-pcm. [1]. I was not able to reporoduce the issue I have seen with mplayer anymore. As it was thought constraint for the period size is enough and there's no need to place the same constraint on buffer size. Regards, Peter --- Peter Ujfalusi (3): ASoC: davinci-mcasp: Constraint on the period and buffer size based on FIFO usage ASoC: davinci-mcasp: Assign the dma_data earlier in dai_probe callback ASoC: davinci-pcm: Add empty functions for !CONFIG_SND_DAVINCI_SOC builds sound/soc/davinci/davinci-mcasp.c | 33 +++++++++++++++++++++++++++++---- sound/soc/davinci/davinci-pcm.h | 10 ++++++++++ 2 files changed, 39 insertions(+), 4 deletions(-) -- 1.9.0

3 13

[alsa-devel] [PATCH] ALSA: hda - Inform the unexpectedly ignored pins by auto-parser
by Takashi Iwai 25 Mar '14

25 Mar '14

The auto-parser may ignore some pins that could be valid when they don't match with the assumption or if there are way too many pins assigned to the same output type. So far, such a pin has been silently ignored, but it's better to leave a message, which would help for debugging and understanding the problem. Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/pci/hda/hda_auto_parser.c | 30 +++++++++++++++++++++++++----- 1 file changed, 25 insertions(+), 5 deletions(-) diff --git a/sound/pci/hda/hda_auto_parser.c b/sound/pci/hda/hda_auto_parser.c index 8de7cfa460b9..90d2fda6c8f9 100644 --- a/sound/pci/hda/hda_auto_parser.c +++ b/sound/pci/hda/hda_auto_parser.c @@ -227,10 +227,18 @@ int snd_hda_parse_pin_defcfg(struct hda_codec *codec, continue; if (!assoc_line_out) assoc_line_out = assoc; - else if (assoc_line_out != assoc) + else if (assoc_line_out != assoc) { + codec_info(codec, + "ignore pin 0x%x with mismatching assoc# 0x%x vs 0x%x\n", + nid, assoc, assoc_line_out); continue; - if (cfg->line_outs >= ARRAY_SIZE(cfg->line_out_pins)) + } + if (cfg->line_outs >= ARRAY_SIZE(cfg->line_out_pins)) { + codec_info(codec, + "ignore pin 0x%x, too many assigned pins\n", + nid); continue; + } line_out[cfg->line_outs].pin = nid; line_out[cfg->line_outs].seq = seq; cfg->line_outs++; @@ -238,8 +246,12 @@ int snd_hda_parse_pin_defcfg(struct hda_codec *codec, case AC_JACK_SPEAKER: seq = get_defcfg_sequence(def_conf); assoc = get_defcfg_association(def_conf); - if (cfg->speaker_outs >= ARRAY_SIZE(cfg->speaker_pins)) + if (cfg->speaker_outs >= ARRAY_SIZE(cfg->speaker_pins)) { + codec_info(codec, + "ignore pin 0x%x, too many assigned pins\n", + nid); continue; + } speaker_out[cfg->speaker_outs].pin = nid; speaker_out[cfg->speaker_outs].seq = (assoc << 4) | seq; cfg->speaker_outs++; @@ -247,8 +259,12 @@ int snd_hda_parse_pin_defcfg(struct hda_codec *codec, case AC_JACK_HP_OUT: seq = get_defcfg_sequence(def_conf); assoc = get_defcfg_association(def_conf); - if (cfg->hp_outs >= ARRAY_SIZE(cfg->hp_pins)) + if (cfg->hp_outs >= ARRAY_SIZE(cfg->hp_pins)) { + codec_info(codec, + "ignore pin 0x%x, too many assigned pins\n", + nid); continue; + } hp_out[cfg->hp_outs].pin = nid; hp_out[cfg->hp_outs].seq = (assoc << 4) | seq; cfg->hp_outs++; @@ -267,8 +283,12 @@ int snd_hda_parse_pin_defcfg(struct hda_codec *codec, break; case AC_JACK_SPDIF_OUT: case AC_JACK_DIG_OTHER_OUT: - if (cfg->dig_outs >= ARRAY_SIZE(cfg->dig_out_pins)) + if (cfg->dig_outs >= ARRAY_SIZE(cfg->dig_out_pins)) { + codec_info(codec, + "ignore pin 0x%x, too many assigned pins\n", + nid); continue; + } cfg->dig_out_pins[cfg->dig_outs] = nid; cfg->dig_out_type[cfg->dig_outs] = (loc == AC_JACK_LOC_HDMI) ? -- 1.9.1

1 0

[alsa-devel] [RFC PATCH] ASoC: wm8904: add CCF support
by Bo Shen 25 Mar '14

25 Mar '14

Enable WM8904 to support common clock framework. And also supports different clk configuration. Signed-off-by: Bo Shen <voice.shen(a)atmel.com> --- Documentation/devicetree/bindings/sound/wm8904.txt | 57 ++++++++++++++++++++++ sound/soc/codecs/wm8904.c | 39 +++++++++++++++ 2 files changed, 96 insertions(+) create mode 100644 Documentation/devicetree/bindings/sound/wm8904.txt diff --git a/Documentation/devicetree/bindings/sound/wm8904.txt b/Documentation/devicetree/bindings/sound/wm8904.txt new file mode 100644 index 0000000..039374d --- /dev/null +++ b/Documentation/devicetree/bindings/sound/wm8904.txt @@ -0,0 +1,57 @@ +WM8904 audio CODEC + +This device supports I2C only. + +Required properties: + - compatible: "wlf,wm8904" + - reg: the I2C address of the device. + + if work with CCF, the following properties as required: + - wlf,sysclk-from-mclk: set the sys clock is driven from mclk, + - wlf,mclk-use-xtal: if the mclk is generated by crystal. + if without this property, the mclk is generated from SOC. + - clocks: which clock generated clock to mclk. + - clock-names: the clock names use for retrieve. + - wlf,mclk-freq: mclk's frequency + + +Pins on the device (for linking into audio routes): + + * IN1L + * IN1R + * IN2L + * IN2R + * IN3L + * IN3R + * HPOUTL + * HPOUTR + * LINEOUTL + * LINEOUTR + * MICBIAS + +Examples: + +1. General usage: +codec: wm8904@1a { + compatible = "wlf,wm8904"; + reg = <0x1a>; +}; + +2. Working with CCF supports and using crystall provide mclk: +codec: wm8904@1a { + compatible = "wlf,wm8904"; + reg = <0x1a>; + wlf,sysclk-from-mclk; + wlf,mclk-use-xtal; + wlf,mclk-freq = <12000000>; +}; + +3. Working with CCF supports and using SoC provide mclk: +codec: wm8904@1a { + compatible = "wlf,wm8904"; + reg = <0x1a>; + wlf,sysclk-from-mclk; + clocks = <&pck0>; + clock-names = "mclk"; + wlf,mclk-freq = <32768>; +}; diff --git a/sound/soc/codecs/wm8904.c b/sound/soc/codecs/wm8904.c index 49c35c3..99e3e90 100644 --- a/sound/soc/codecs/wm8904.c +++ b/sound/soc/codecs/wm8904.c @@ -11,6 +11,7 @@ * published by the Free Software Foundation. */ +#include <linux/clk.h> #include <linux/module.h> #include <linux/moduleparam.h> #include <linux/init.h> @@ -49,6 +50,10 @@ static const char *wm8904_supply_names[WM8904_NUM_SUPPLIES] = { /* codec private data */ struct wm8904_priv { struct regmap *regmap; + struct clk *mclk; + bool sysclk_from_mclk; + bool mclk_use_xtal; + u32 mclk_freq; enum wm8904_type devtype; @@ -1828,6 +1833,9 @@ static int wm8904_set_bias_level(struct snd_soc_codec *codec, switch (level) { case SND_SOC_BIAS_ON: + if (IS_ENABLED(CONFIG_COMMON_CLK)) + if (!wm8904->mclk_use_xtal) + clk_prepare_enable(wm8904->mclk); break; case SND_SOC_BIAS_PREPARE: @@ -1894,6 +1902,9 @@ static int wm8904_set_bias_level(struct snd_soc_codec *codec, regulator_bulk_disable(ARRAY_SIZE(wm8904->supplies), wm8904->supplies); + if (IS_ENABLED(CONFIG_COMMON_CLK)) + if (!wm8904->mclk_use_xtal) + clk_disable_unprepare(wm8904->mclk); break; } codec->dapm.bias_level = level; @@ -2139,6 +2150,34 @@ static int wm8904_i2c_probe(struct i2c_client *i2c, return ret; } + if (IS_ENABLED(CONFIG_COMMON_CLK)) { + struct device_node *np = i2c->dev.of_node; + wm8904->sysclk_from_mclk = + of_property_read_bool(np, "wlf,sysclk-from-mclk"); + if (wm8904->sysclk_from_mclk) { + wm8904->mclk_use_xtal = + of_property_read_bool(np, "wlf,mclk-use-xtal"); + + ret = of_property_read_u32(np, "wlf,mclk-freq", + &wm8904->mclk_freq); + if (ret) { + dev_err(&i2c->dev, "Failed to read mclk freq"); + goto err_enable; + } + + if (!wm8904->mclk_use_xtal) { + wm8904->mclk = devm_clk_get(&i2c->dev, "mclk"); + if (IS_ERR(wm8904->mclk)) { + dev_err(&i2c->dev, "Failed to get MCLK\n"); + ret = PTR_ERR(wm8904->mclk); + goto err_enable; + } + + clk_set_rate(wm8904->mclk, wm8904->mclk_freq); + } + } + } + ret = regmap_read(wm8904->regmap, WM8904_SW_RESET_AND_ID, &val); if (ret < 0) { dev_err(&i2c->dev, "Failed to read ID register: %d\n", ret); -- 1.8.5.2

4 8

[alsa-devel] [PATCH] ASoC: rcar: bugfix: it cares about the non-src case
by Kuninori Morimoto 24 Mar '14

24 Mar '14

From: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> src might not exist. kernel will be hung-up without this patch in such case. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- >> Mark This is for mark/topic/rcar branch sound/soc/sh/rcar/src.c | 18 +++++++++++------- 1 file changed, 11 insertions(+), 7 deletions(-) diff --git a/sound/soc/sh/rcar/src.c b/sound/soc/sh/rcar/src.c index eee75eb..6232b7d 100644 --- a/sound/soc/sh/rcar/src.c +++ b/sound/soc/sh/rcar/src.c @@ -182,16 +182,20 @@ unsigned int rsnd_src_get_ssi_rate(struct rsnd_priv *priv, struct rsnd_dai_stream *io, struct snd_pcm_runtime *runtime) { + struct rsnd_mod *src_mod = rsnd_io_to_mod_src(io); struct rsnd_src *src; - unsigned int rate; + unsigned int rate = 0; - src = rsnd_mod_to_src(rsnd_io_to_mod_src(io)); + if (src_mod) { + src = rsnd_mod_to_src(src_mod); + + /* + * return convert rate if SRC is used, + * otherwise, return runtime->rate as usual + */ + rate = rsnd_src_convert_rate(src); + } - /* - * return convert rate if SRC is used, - * otherwise, return runtime->rate as usual - */ - rate = rsnd_src_convert_rate(src); if (!rate) rate = runtime->rate; -- 1.7.9.5

2 1

[alsa-devel] [PATCH] ARM: bockw: fixup SND_SOC_DAIFMT_CBx_CFx flags
by Kuninori Morimoto 24 Mar '14

24 Mar '14

From: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> SND_SOC_DAIFMT_CBx_CFx means "codec" side master/slave mode. Then, rcar will be master mode if it was SND_SOC_DAIFMT_CBS_CFS. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- >> Mark, Simon I'm so sorry. current Mark's fix/rcar branch has SND_SOC_DAIFMT_CBx_CFx fixup patches, but it didn't care about bockw board. This patch is for it. This Subject is using "ARM: bockw:" prefix, but, should it be "ASoC: rcar" ? arch/arm/mach-shmobile/board-bockw.c | 16 +++++----------- 1 file changed, 5 insertions(+), 11 deletions(-) diff --git a/arch/arm/mach-shmobile/board-bockw.c b/arch/arm/mach-shmobile/board-bockw.c index b5867b7..f444be2 100644 --- a/arch/arm/mach-shmobile/board-bockw.c +++ b/arch/arm/mach-shmobile/board-bockw.c @@ -467,14 +467,12 @@ static struct asoc_simple_card_info rsnd_card_info[] = { .card = "SSI56-AK4643", .codec = "ak4642-codec.0-0012", .platform = "rcar_sound", - .daifmt = SND_SOC_DAIFMT_LEFT_J, + .daifmt = SND_SOC_DAIFMT_LEFT_J | SND_SOC_DAIFMT_CBM_CFM, .cpu_dai = { .name = "rsnd-dai.0", - .fmt = SND_SOC_DAIFMT_CBS_CFS, }, .codec_dai = { .name = "ak4642-hifi", - .fmt = SND_SOC_DAIFMT_CBM_CFM, .sysclk = 11289600, }, }, @@ -484,10 +482,9 @@ static struct asoc_simple_card_info rsnd_card_info[] = { .card = "SSI3-AK4554(playback)", .codec = "ak4554-adc-dac.0", .platform = "rcar_sound", + .daifmt = SND_SOC_DAIFMT_CBS_CFS | SND_SOC_DAIFMT_RIGHT_J, .cpu_dai = { .name = "rsnd-dai.1", - .fmt = SND_SOC_DAIFMT_CBM_CFM | - SND_SOC_DAIFMT_RIGHT_J, }, .codec_dai = { .name = "ak4554-hifi", @@ -499,10 +496,9 @@ static struct asoc_simple_card_info rsnd_card_info[] = { .card = "SSI4-AK4554(capture)", .codec = "ak4554-adc-dac.0", .platform = "rcar_sound", + .daifmt = SND_SOC_DAIFMT_CBS_CFS | SND_SOC_DAIFMT_LEFT_J, .cpu_dai = { .name = "rsnd-dai.2", - .fmt = SND_SOC_DAIFMT_CBM_CFM | - SND_SOC_DAIFMT_LEFT_J, }, .codec_dai = { .name = "ak4554-hifi", @@ -514,10 +510,9 @@ static struct asoc_simple_card_info rsnd_card_info[] = { .card = "SSI7-AK4554(playback)", .codec = "ak4554-adc-dac.1", .platform = "rcar_sound", + .daifmt = SND_SOC_DAIFMT_CBS_CFS | SND_SOC_DAIFMT_RIGHT_J, .cpu_dai = { .name = "rsnd-dai.3", - .fmt = SND_SOC_DAIFMT_CBM_CFM | - SND_SOC_DAIFMT_RIGHT_J, }, .codec_dai = { .name = "ak4554-hifi", @@ -529,10 +524,9 @@ static struct asoc_simple_card_info rsnd_card_info[] = { .card = "SSI8-AK4554(capture)", .codec = "ak4554-adc-dac.1", .platform = "rcar_sound", + .daifmt = SND_SOC_DAIFMT_CBS_CFS | SND_SOC_DAIFMT_LEFT_J, .cpu_dai = { .name = "rsnd-dai.4", - .fmt = SND_SOC_DAIFMT_CBM_CFM | - SND_SOC_DAIFMT_LEFT_J, }, .codec_dai = { .name = "ak4554-hifi", -- 1.7.9.5

2 2

[alsa-devel] [PATCH RFC 0/2] Move dai-link level properties away from dai subnodes
by Jyri Sarha 24 Mar '14

24 Mar '14

This patch is implemented on top of late patches from Jean-Francois Moine [1]. The patches implement the main part of the simple-card changes discussed in alsa-devel mailing list. [1] http://mailman.alsa-project.org/pipermail/alsa-devel/2014-March/074592.html Best regards, Jyri Jyri Sarha (2): ASoC: core: Update snd_soc_of_parse_daifmt() interface ASoC: simple-card: Move dai-link level properties away from dai subnodes .../devicetree/bindings/sound/simple-card.txt | 91 ++++---- include/sound/soc.h | 4 +- sound/soc/generic/simple-card.c | 237 ++++++++++++-------- sound/soc/soc-core.c | 8 +- 4 files changed, 201 insertions(+), 139 deletions(-) -- 1.7.9.5

3 6

[alsa-devel] [PATCH RFC v2 0/2] Move dai-link level properties away from dai subnodes
by Jyri Sarha 24 Mar '14

24 Mar '14

This patch is implemented on top of late patches from Jean-Francois Moine [1]. These patches implement the main part of the simple-card changes discussed in alsa-devel mailing list [2]. Since RFC version of the patches: - ASoC: core: Update snd_soc_of_parse_daifmt() interface - Make minimal changes to simple-card for it to work with the updated snd_soc_of_parse_daifmt(). (Kuninori Morimoto's comment [3]). - ASoC: simple-card: Move dai-link level properties away from dai subnodes - Fix typo in DT binding document (Jean-Francois Moine's comment [4]). - Remove #define DEBUG - Remove accidental line wrap at simple-audio-card,name property descripition in the DT binding document [1] http://mailman.alsa-project.org/pipermail/alsa-devel/2014-March/074592.html [2] http://mailman.alsa-project.org/pipermail/alsa-devel/2014-March/074388.html [3] http://mailman.alsa-project.org/pipermail/alsa-devel/2014-March/074699.html [4] http://mailman.alsa-project.org/pipermail/alsa-devel/2014-March/074697.html Best regards, Jyri Jyri Sarha (2): ASoC: core: Update snd_soc_of_parse_daifmt() interface ASoC: simple-card: Move dai-link level properties away from dai subnodes .../devicetree/bindings/sound/simple-card.txt | 88 ++++---- include/sound/soc.h | 4 +- sound/soc/generic/simple-card.c | 236 ++++++++++++-------- sound/soc/soc-core.c | 8 +- 4 files changed, 198 insertions(+), 138 deletions(-) -- 1.7.9.5

1 2

[alsa-devel] [hsw dp audio] Muting S/PDIF turns off DP
by Daniel Martin 24 Mar '14

24 Mar '14

Hi, I've a bug here: when playing audio via the speakers of a monitor attached to a DisplayPort and I mute the corresponding S/PDIF (i.e. with alsamixer) then the monitor switches off into standby, unmuting the S/PDIF turns the monitor back on. When muting the S/PDIF while no audio is playing everything is fine. There's no audio daemon running and my ~/.asoundrc contains only: pcm.!default { type hw card 0 device 3 } I've already filed a fdo bug report (drm/intel) for this as I thought it might be related to it: Bug 75729 - [hsw dp audio] Muting S/PDIF turns off DP https://bugs.freedesktop.org/show_bug.cgi?id=75729 Attached to this bug is a `alsa-info.log` too. As stated in the bug report it's not a showstopper for me, but obviously something seems to go wrong and maybe someone has an idea. Thanks, Daniel Martin

2 1