November 2011 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] First test of the hda-jack branch
by David Henningsson 21 Nov '11

21 Nov '11

Hi Takashi, After having some troubles with the kernel scripts (on the Ubuntu side) I've managed to try out your test/hda-jack branch. I'm attaching a patch I had to apply to make the modules load properly. After having applied the attached patch, I get stuck at an OOPS, which I don't have time to troubleshoot further today. I'm attaching the OOPS log as well (alsa-info for the system I'm testing on is the same as I sent you a few days ago). -- David Henningsson, Canonical Ltd. http://launchpad.net/~diwic

4 10

[alsa-devel] [PATCH] ASoC: Mark wm8753 register cache as dirty when suspending
by Mark Brown 21 Nov '11

21 Nov '11

Ensures that it is resynced properly on resume. Signed-off-by: Mark Brown <broonie(a)opensource.wolfsonmicro.com> --- sound/soc/codecs/wm8753.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/sound/soc/codecs/wm8753.c b/sound/soc/codecs/wm8753.c index 3a629d0..3f3541b 100644 --- a/sound/soc/codecs/wm8753.c +++ b/sound/soc/codecs/wm8753.c @@ -1395,6 +1395,7 @@ static void wm8753_work(struct work_struct *work) static int wm8753_suspend(struct snd_soc_codec *codec, pm_message_t state) { wm8753_set_bias_level(codec, SND_SOC_BIAS_OFF); + codec->cache_sync = 1; return 0; } -- 1.7.7.3

1 0

[alsa-devel] [PATCH] ASoC: Use table based init for wm8731_snd_controls
by Mark Brown 21 Nov '11

21 Nov '11

Signed-off-by: Mark Brown <broonie(a)opensource.wolfsonmicro.com> --- sound/soc/codecs/wm8731.c | 5 ++--- 1 files changed, 2 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/wm8731.c b/sound/soc/codecs/wm8731.c index 4030a9c..b7abe33 100644 --- a/sound/soc/codecs/wm8731.c +++ b/sound/soc/codecs/wm8731.c @@ -559,9 +559,6 @@ static int wm8731_probe(struct snd_soc_codec *codec) /* Disable bypass path by default */ snd_soc_update_bits(codec, WM8731_APANA, 0x8, 0); - snd_soc_add_controls(codec, wm8731_snd_controls, - ARRAY_SIZE(wm8731_snd_controls)); - /* Regulators will have been enabled by bias management */ regulator_bulk_disable(ARRAY_SIZE(wm8731->supplies), wm8731->supplies); @@ -598,6 +595,8 @@ static struct snd_soc_codec_driver soc_codec_dev_wm8731 = { .num_dapm_widgets = ARRAY_SIZE(wm8731_dapm_widgets), .dapm_routes = wm8731_intercon, .num_dapm_routes = ARRAY_SIZE(wm8731_intercon), + .controls = wm8731_snd_controls, + .num_controls = ARRAY_SIZE(wm8731_snd_controls), }; static const struct of_device_id wm8731_of_match[] = { -- 1.7.7.3

1 0

[alsa-devel] [PATCH] ASoC: Ensure WM8731 register cache is synced when resuming from disabled
by Mark Brown 21 Nov '11

21 Nov '11

Signed-off-by: Mark Brown <broonie(a)opensource.wolfsonmicro.com> --- sound/soc/codecs/wm8731.c | 1 + 1 files changed, 1 insertions(+), 0 deletions(-) diff --git a/sound/soc/codecs/wm8731.c b/sound/soc/codecs/wm8731.c index 7e5ec03..a7c9ae1 100644 --- a/sound/soc/codecs/wm8731.c +++ b/sound/soc/codecs/wm8731.c @@ -453,6 +453,7 @@ static int wm8731_set_bias_level(struct snd_soc_codec *codec, snd_soc_write(codec, WM8731_PWR, 0xffff); regulator_bulk_disable(ARRAY_SIZE(wm8731->supplies), wm8731->supplies); + codec->cache_sync = 1; break; } codec->dapm.bias_level = level; -- 1.7.7.3

1 0

[alsa-devel] [patch] ALSA: cs5535 - Fix an endianness conversion
by Dan Carpenter 21 Nov '11

21 Nov '11

desc->size is supposed to be a le16 type. On a big endian system the current code will set ->size to zero. We fixed a similar bug on the next line but missed this one. Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> --- This is a static checker bug. I haven't tested it. diff --git a/sound/pci/cs5535audio/cs5535audio_pcm.c b/sound/pci/cs5535audio/cs5535audio_pcm.c index e083122..dbf94b1 100644 --- a/sound/pci/cs5535audio/cs5535audio_pcm.c +++ b/sound/pci/cs5535audio/cs5535audio_pcm.c @@ -148,7 +148,7 @@ static int cs5535audio_build_dma_packets(struct cs5535audio *cs5535au, struct cs5535audio_dma_desc *desc = &((struct cs5535audio_dma_desc *) dma->desc_buf.area)[i]; desc->addr = cpu_to_le32(addr); - desc->size = cpu_to_le32(period_bytes); + desc->size = cpu_to_le16(period_bytes); desc->ctlreserved = cpu_to_le16(PRD_EOP); desc_addr += sizeof(struct cs5535audio_dma_desc); addr += period_bytes;

2 1

[alsa-devel] [GIT PULL] fixes for wrong DB_RANGE_HEADER sizes
by Clemens Ladisch 21 Nov '11

21 Nov '11

The following changes since commit 1ea6b8f48918282bdca0b32a34095504ee65bab5: Linux 3.2-rc1 (2011-11-07 16:16:02 -0800) are available in the git repository at: git://git.alsa-project.org/alsa-kprivate.git DB_RANGE-size-fixes Clemens Ladisch (7): ASoC: adau1373: fix DB_RANGE size ASoC: rt5631: fix DB_RANGE size ASoC: sgtl5000: fix DB_RANGE size ASoC: wm8962: fix DB_RANGE size ASoC: wm8993: fix DB_RANGE size ASoC: wm9090: fix DB_RANGE size ASoC: wm_hubs: fix DB_RANGE size sound/soc/codecs/adau1373.c | 2 +- sound/soc/codecs/rt5631.c | 2 +- sound/soc/codecs/sgtl5000.c | 2 +- sound/soc/codecs/wm8962.c | 4 ++-- sound/soc/codecs/wm8993.c | 2 +- sound/soc/codecs/wm9090.c | 6 +++--- sound/soc/codecs/wm_hubs.c | 2 +- 7 files changed, 10 insertions(+), 10 deletions(-)

2 3

Re: [alsa-devel] [PATCH 1/3] sound/soc/codecs: add LAPIS Semiconductor ML26124
by Mark Brown 21 Nov '11

21 Nov '11

On Mon, Nov 21, 2011 at 01:08:50PM +0900, Tomoya MORINAGA wrote: > ML26124-01HB/ML26124-02GD is 16bit monaural audio CODEC which has high > resistance to voltage noise. On chip regulator realizes power supply > rejection So, in general the big overall issue with this code is that it's really not following the general coding styles and standards for a modern ASoC driver. I've pointed out some issues below but there's rather a lot of them. In general you should always try to ensure that the code you're submitting looks like other code for the relevant subsystem. The big thing that's sticking out at me is the register write sequence you've got in the middle of the driver - it looks awfully like it's hard coding a particular use case. > +struct ml26124_priv { > + enum snd_soc_control_type control_type; > + spinlock_t lock; > + struct snd_pcm_substream *substream; > + unsigned int rate; > + unsigned int ch; > + struct mutex i2c_mutex; What are these locks for? Especially the spinlock looks highly suspicous. > +static const char *ml26124_lrmcon[] = {"Use L", "Use R", "Use (L+R)", "Use (L+R)/2"}; The "Use" isn't terribly idomatic, remove it. > +static const struct soc_enum ml26124_enum[] = { > +/* #define SOC_ENUM_SINGLE(xreg, xshift, xmax, xtexts) */ > + SOC_ENUM_SINGLE(ML26124_MIXER_VOL_CTL, 0, 4, ml26124_lrmcon), > + SOC_ENUM_SINGLE(ML26124_MIXER_VOL_CTL, 4, 15, ml26124_dvfconcon), Don't use a big table of enums, indexing into it is illegible and error prone. > +static const struct snd_kcontrol_new ml26124_snd_controls[] = { > + SOC_SINGLE_TLV("Record Digital Volume", ML26124_RECORD_DIG_VOL, 0, 0xff, 1, rec_play_digi_vol), Capture, not Record. > + SOC_SINGLE_TLV("EQ Band1 Gain Setting", ML26124_EQ_GAIN_BRAND1, 0, 0xff, 1, eq_band_gain), Volume not Gain Setting. As a rule anything with TLV information should be a Volume. > +static const struct snd_kcontrol_new ml26124_dsp_controls[] = { > + SOC_SINGLE("Play Limitter ON/OFF", ML26124_FILTER_EN, 0, 1, 0), Switch. > + SOC_SINGLE("Ditital Volume MUTE", ML26124_FILTER_EN, 4, 1, 0), Switch. > + SOC_SINGLE("Set ALC position", ML26124_FILTER_EN, 5, 1, 0), What does this actually do? From the name it *really* doesn't look like a mixer input. > + SND_SOC_DAPM_SPK("Speaker", NULL), > + SND_SOC_DAPM_LINE("LINEOUT", NULL), These should be _OUTPUT() pins - SPK and LINE widgets should only be used by machine drivers. > +#define CODEC_DEV_ADDR (0x1A) > +static struct i2c_board_info ioh_hwmon_info[] = { > + {I2C_BOARD_INFO("ioh_i2c-0", CODEC_DEV_ADDR + 1)}, > + {I2C_BOARD_INFO("ioh_i2c-1", CODEC_DEV_ADDR + 2)}, > + {I2C_BOARD_INFO("ioh_i2c-2", CODEC_DEV_ADDR + 3)}, > + {I2C_BOARD_INFO("ioh_i2c-3", CODEC_DEV_ADDR + 4)}, > + {I2C_BOARD_INFO("ioh_i2c-4", CODEC_DEV_ADDR + 5)}, > + {I2C_BOARD_INFO("ioh_i2c-5", CODEC_DEV_ADDR + 0)}, > +}; Absolutely no. Your driver should be a standard I2C driver, device registration should be done by the board not the chip driver. > +static const struct snd_soc_dapm_route intercon[] = { > +}; I can't see how any of the DAPM widgets you specify will ever work with no interconnections. > + > +static int snd_card_codec_reg_read(struct ml26124_priv *priv, > + unsigned char reg, unsigned char *val) > +{ > + unsigned char data; > + struct i2c_client *i2c; Use the standard register access code, don't open code things in your driver unless there's a good reason to. Current best practice for most I2C or SPI devices is to use regmap, see any recently added driver for examples. > +static int snd_card_codec_set(int number, struct ml26124_priv *priv) > +{ Namespacing - this isn't a generic ALSA or ASoC function. > + unsigned char data; > + unsigned int rate = priv->rate; > + unsigned int channels = priv->ch; > + > + snd_card_codec_reg_read(priv, 0x30, &data); /* Read MICVIAS Voltage */ > + > + snd_card_codec_reg_write(priv, 0x10, 0x01); /* soft reset assert */ > + snd_card_codec_reg_write(priv, 0x10, 0x00); /* soft reset negate */ > + > + snd_card_codec_reg_write(priv, 0x0c, 0x00); /* Stop clock */ Use snd_soc_update_bits() and the other standard register access functions. > + switch (rate) { > + case 16000: > + snd_card_codec_reg_write(priv, 0x00, 0x03); > + snd_card_codec_reg_write(priv, 0x02, 0x0c); > + snd_card_codec_reg_write(priv, 0x04, 0x00); > + snd_card_codec_reg_write(priv, 0x06, 0x20); > + snd_card_codec_reg_write(priv, 0x08, 0x00); > + snd_card_codec_reg_write(priv, 0x0a, 0x04); > + break; This sort of magic number based code without even registers being named isn't really of the standard required for maintainability or review. > + default: > + pr_err("%s:this rate is no support for ml26124\n", __func__); > + break; > + } If the driver has detected an error it should return an error; printing the rate that was requested as part of the error message would also be useful. > + snd_card_codec_reg_write(priv, 0x64, 0x00); /* master/slave set slave */ This looks like it should be in set_dai_fmt(). > + > + snd_card_codec_reg_write(priv, 0x20, 0x02); /* VMID on. normal mode */ > + msleep(50); This and much of the code looks like it should be in set_bias_level(). > + snd_card_codec_reg_write(priv, 0x26, 0x1f); /* Speaker Amplified Poer Management */ > + snd_card_codec_reg_write(priv, 0x28, 0x02); /* LOUT Control Regsister */ > + > + snd_card_codec_reg_write(priv, 0x54, 0x02); /* Speaker Amplifier output Control */ > + snd_card_codec_reg_write(priv, 0x5a, 0x00); /* Mic Interface Control Register */ This looks awfully like it's supposed to be DAPM. > +static int ml26124_hw_params(struct snd_pcm_substream *substream, > + struct snd_pcm_hw_params *hw_params, > + struct snd_soc_dai *dai) > +{ > + struct snd_soc_codec *codec = dai->codec; > + struct ml26124_priv *priv = snd_soc_codec_get_drvdata(codec); > + > + if (snd_card_codec_set(substream->number, priv)) > + return -1; Return real error codes if you're going to detect errors. > + > + return snd_pcm_lib_malloc_pages(substream, > + params_buffer_bytes(hw_params)); Why is a CODEC driver allocing memory to buffer audio? > +static int ml26124_pcm_prepare(struct snd_pcm_substream *substream, > + struct snd_soc_dai *dai) > +{ > + return 0; > +} Remove empty functions. > +static int ml26124_mute(struct snd_soc_dai *dai, int mute) > +{ > + struct snd_soc_codec *codec = dai->codec; > + struct ml26124_priv *priv = snd_soc_codec_get_drvdata(codec); > + > + if (mute) > + snd_card_codec_reg_write(priv, ML26124_DVOL_CTL, > + DVOL_CTL_DVMUTE_ON); > + else > + snd_card_codec_reg_write(priv, ML26124_DVOL_CTL, > + DVOL_CTL_DVMUTE_OFF); I suspect you mean to use snd_soc_update_bits() here. > +static int ml26124_set_bias_level(struct snd_soc_codec *codec, > + enum snd_soc_bias_level level) > +{ > + struct ml26124_priv *priv = snd_soc_codec_get_drvdata(codec); > + > + switch (level) { > + case SND_SOC_BIAS_ON: > + case SND_SOC_BIAS_PREPARE: > + case SND_SOC_BIAS_STANDBY: > + snd_card_codec_reg_write(priv, ML26124_REF_PM, REF_PM_VMID_ON); > + msleep(50); > + snd_card_codec_reg_write(priv, ML26124_REF_PM, > + REF_PM_VMID_ON | REF_PM_MICBEN); This looks much more like what I'd expect in terms of coding style - meaingful defines and use of set_bias_level(). Also note that during audio startup you'll transition STANDBY->PREPARE->ON and the reverse on shutdown, I really don't think you want to be doing this for anything except STANDBY. I also can't see how this will play nicely with the above register write sequence - you probably want to just remove the register write sequence. > +static int ml26124_probe(struct snd_soc_codec *codec) > +{ > + snd_soc_add_controls(codec, ml26124_snd_controls, > + ARRAY_SIZE(ml26124_snd_controls)); > + > + snd_soc_dapm_new_controls(codec, ml26124_dapm_widgets, > + ARRAY_SIZE(ml26124_dapm_widgets)); Initialise these from your codec driver structure. > +#if defined(CONFIG_I2C) || defined(CONFIG_I2C_MODULE) > +static __devinit int ml26124_i2c_probe(struct i2c_client *i2c, > + const struct i2c_device_id *id) Since there's no visible SPI support just unconditionally use I2C. > +static struct i2c_driver ml26124_i2c_driver = { > + .driver = { > + .name = "ml26124-codec", Remove the -codec, it's redundant.

1 0

[alsa-devel] [PATCH v6 0/3] ARM: mxs: add recording support for saif
by Dong Aisheng 21 Nov '11

21 Nov '11

Changes since v5: Only one change is that: * remove the unneccesary parenthesis for [PATCH 1/3] suggested by Marek Vasut. Changes since v4: * use new added master_id and master_mode in platform_data. The new changes depend on another patch: 0001-ASoC-mxs-saif-remove-function-in-platform_data.patch Changes since v3: * remove unneeded locking according to Sascha. The patches are based on imx-features branch since commit: ca4e419c2. Changes since v2: * separate clkmux code into another patch according to Uwe * add lock according to Wolfram * Other minus fixes suggested by Uwe and Wolfram. * remove Wolfram's fix saif clock setting patch which is in v2 series since Wolfram will reform the clock code and send it out himself. For test purpose, user still needs that patch. People can get it from v2 series. The patch is: [PATCH v2 3/3] arm: mxs: disable clock-gates when setting saif-clocks The patches are based on imx-features branch since commit: f4f01e31835f. Changes since v1: The main changes are move mach-specific code(clkmux in DIGCTL) from saif driver to mach-specific layer based on Wolfram's suggestion. Note that the last patch is a RFC patch and sent out for testing since without that patch the saif may not work. Dong Aisheng (3): ARM: mxs: add saif clkmux functions ARM: mx28evk: add platform data for saif ARM: mx28evk: set a initial clock rate for saif arch/arm/mach-mxs/clock-mx28.c | 38 +++++++++++++++++++++++ arch/arm/mach-mxs/devices-mx28.h | 3 +- arch/arm/mach-mxs/devices/platform-mxs-saif.c | 5 ++- arch/arm/mach-mxs/include/mach/common.h | 1 + arch/arm/mach-mxs/include/mach/devices-common.h | 4 ++- arch/arm/mach-mxs/include/mach/digctl.h | 21 ++++++++++++ arch/arm/mach-mxs/mach-mx28evk.c | 18 +++++++++- 7 files changed, 84 insertions(+), 6 deletions(-) create mode 100644 arch/arm/mach-mxs/include/mach/digctl.h

3 5

[alsa-devel] No sound on Dell StudioXPS 1340
by auto30512278＠hushmail.com 21 Nov '11

21 Nov '11

Hi, I've this problem from months ago (I think it started with an ubuntu upgrade), I reinstalled Ubuntu at least 2 times since and I couldn't get the sound working again.Now I'm on a fresh installation of xubuntu, and still have the same problem. Recently I asked for help at #alsa and after a couple of minutes of troubleshooting with a nice guy without luck, he recommend me to post the problem here with the link below. Short story: No sound from internal speakers, neither from headphones jack (both works fine on win). Applications seems to reproduce the audio without throwing errors but I never heard a sound. We tried a lot of thing, I can recall:- Disable pulseaudio (echo "autospawn = false" > ~/.pulse/client.conf && pactl exit)- I ran speaker-test -c 2 -d plughw0,0, no luck- I see the meters moving on pavucontrol (while running speaker-test -d pulse -c 2) http://www.alsa-project.org/db/?f=d9612384067fb86a13668607496e52ebcb90415e Anyone? It's really annoying :(Thanks

1 0

[alsa-devel] [PATCH] ASoC: cs4271: Fix wrong mask parameter in some snd_soc_update_bits calls
by Axel Lin 20 Nov '11

20 Nov '11

Signed-off-by: Axel Lin <axel.lin(a)gmail.com> --- sound/soc/codecs/cs4271.c | 8 +++++--- 1 files changed, 5 insertions(+), 3 deletions(-) diff --git a/sound/soc/codecs/cs4271.c b/sound/soc/codecs/cs4271.c index 23d1bd5..69fde15 100644 --- a/sound/soc/codecs/cs4271.c +++ b/sound/soc/codecs/cs4271.c @@ -434,7 +434,8 @@ static int cs4271_soc_suspend(struct snd_soc_codec *codec, pm_message_t mesg) { int ret; /* Set power-down bit */ - ret = snd_soc_update_bits(codec, CS4271_MODE2, 0, CS4271_MODE2_PDN); + ret = snd_soc_update_bits(codec, CS4271_MODE2, CS4271_MODE2_PDN, + CS4271_MODE2_PDN); if (ret < 0) return ret; return 0; @@ -501,8 +502,9 @@ static int cs4271_probe(struct snd_soc_codec *codec) return ret; } - ret = snd_soc_update_bits(codec, CS4271_MODE2, 0, - CS4271_MODE2_PDN | CS4271_MODE2_CPEN); + ret = snd_soc_update_bits(codec, CS4271_MODE2, + CS4271_MODE2_PDN | CS4271_MODE2_CPEN, + CS4271_MODE2_PDN | CS4271_MODE2_CPEN); if (ret < 0) return ret; ret = snd_soc_update_bits(codec, CS4271_MODE2, CS4271_MODE2_PDN, 0); -- 1.7.5.4

3 2