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

Re: [alsa-devel] [PATCH] ASoC: samsung: drop unknown Kconfig symbol
by Mark Brown 24 Nov '11

24 Nov '11

On Thu, Nov 24, 2011 at 08:59:08PM +0100, Paul Bolle wrote: > On Thu, 2011-11-24 at 19:19 +0000, Mark Brown wrote: > > The cost of extra platforms is really > > very low and people actually send me patches for the OpenMoko devices > > from time to time (much more frequently than for most other boards) so > > at least someone cares. > But - repeating my message - none of those patches made the code in the > mainline tree buildable. Which is a problem because... (and note that you mean "selectable in Kconfig" here). > I don't have any stake in the GTA01 code. So I don't really mind > deferring this discussion to (say) the end of the v3.3 merge period. > But, surely, there must be a point where one can conclude that > unbuildable code is just that? No. Removing this code has negative value. There is no meaningful cost in having it in mainline; your patch has been the biggest piece of work so far other than updates done by people actively working on the machine. On the other hand the OpenMoko guys from time to time send useful patches and report interesting bugs (they sent a patch only last week in fact) and are definitely a user for the code (which is more than can be said for most of the legacy boards in mainline). Please take a step back, think about what value would be brought by making this change.

1 0

[alsa-devel] PCM usage / seg fault in pcm_params.c
by Borg Onion 24 Nov '11

24 Nov '11

Hello, Program terminated with signal 11, Segmentation fault. #0 0x00534502 in snd_pcm_hw_refine (pcm=0x94b36f0, params=0x94b36f0) at pcm_params.c:2291 2291 res = pcm->ops->hw_refine(pcm->op_arg, params); (gdb) bt #0 0x00534502 in snd_pcm_hw_refine (pcm=0x94b36f0, params=0x94b36f0) at pcm_params.c:2291 #1 0x0052a1e2 in snd_pcm_hw_params_any (pcm=0x94b36f0, params=0x94b36f0) at pcm.c:3178 #2 0x080497a7 in print_pcm_hw_params (pcm=0x94b36f0) at list.c:109 #3 0x080499ef in print_pcm_devices (handle=0x94af638) at list.c:147 #4 0x08049362 in main (argc=1, argv=0xbfcf66b4) at list.c:227 (gdb) print pcm->ops $1 = (const snd_pcm_ops_t *) 0x0 Source code at http://pastebin.com/rqBCDza7 Sound card is opened, PCM device is opened, hw params were just allocated, and now I'm calling snd_pcm_hw_params_any(). Trying to populate the empty params struct so as to interrogate it and find out the card's capabilities. I'm sure I'm mis-using alsa-lib, but also shouldn't this have an assert() on pcm->ops ? What's the right call sequence to get capabilities of this PCM channel (sample rates, bit depths, etc)? Thanks for any help! --BO

2 2

Re: [alsa-devel] [PATCH] ASoC: samsung: drop unknown Kconfig symbol
by Mark Brown 24 Nov '11

24 Nov '11

On Thu, Nov 24, 2011 at 08:09:14PM +0100, Paul Bolle wrote: > - Kconfig"), both dated May 14, 2007. So I'm touching code here that has > been unbuildable (in the mainline tree) for over four years. Isn't that > in itself sufficient to drop this code (and the incomplete kbuild stuff > supporting it)? No, not really. None of the OpenMoko code (including GTA02) has ever been usable in mainline as there's a reasonable patch stack for it out of tree and the more of their code is in mainline the more likely it is that that situation will improve. The cost of extra platforms is really very low and people actually send me patches for the OpenMoko devices from time to time (much more frequently than for most other boards) so at least someone cares. > > I'd suggest discussing this with the OpenMoko guys if you want to do it. > Who would that be? Google tells me http://www.openmoko.org is a good start - you might want to ask on the openmoko-kernel list.

1 0

[alsa-devel] [PATCH] ASoC: Staticise non-exported symbols in sta32x
by Mark Brown 24 Nov '11

24 Nov '11

Signed-off-by: Mark Brown <broonie(a)opensource.wolfsonmicro.com> --- sound/soc/codecs/sta32x.c | 4 ++-- 1 files changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/sta32x.c b/sound/soc/codecs/sta32x.c index e2b1cde..edcbeef 100644 --- a/sound/soc/codecs/sta32x.c +++ b/sound/soc/codecs/sta32x.c @@ -265,7 +265,7 @@ static int sta32x_coefficient_put(struct snd_kcontrol *kcontrol, return 0; } -int sta32x_sync_coef_shadow(struct snd_soc_codec *codec) +static int sta32x_sync_coef_shadow(struct snd_soc_codec *codec) { struct sta32x_priv *sta32x = snd_soc_codec_get_drvdata(codec); unsigned int cfud; @@ -290,7 +290,7 @@ int sta32x_sync_coef_shadow(struct snd_soc_codec *codec) return 0; } -int sta32x_cache_sync(struct snd_soc_codec *codec) +static int sta32x_cache_sync(struct snd_soc_codec *codec) { unsigned int mute; int rc; -- 1.7.7.3

2 1

[alsa-devel] [PATCH] ASoC: Use devm_kzalloc() in wm5100
by Mark Brown 24 Nov '11

24 Nov '11

Signed-off-by: Mark Brown <broonie(a)opensource.wolfsonmicro.com> --- sound/soc/codecs/wm5100.c | 8 ++++---- 1 files changed, 4 insertions(+), 4 deletions(-) diff --git a/sound/soc/codecs/wm5100.c b/sound/soc/codecs/wm5100.c index 46aca14..a9d5ac0 100644 --- a/sound/soc/codecs/wm5100.c +++ b/sound/soc/codecs/wm5100.c @@ -2743,7 +2743,8 @@ static __devinit int wm5100_i2c_probe(struct i2c_client *i2c, unsigned int reg; int ret, i; - wm5100 = kzalloc(sizeof(struct wm5100_priv), GFP_KERNEL); + wm5100 = devm_kzalloc(&i2c->dev, sizeof(struct wm5100_priv), + GFP_KERNEL); if (wm5100 == NULL) return -ENOMEM; @@ -2752,7 +2753,7 @@ static __devinit int wm5100_i2c_probe(struct i2c_client *i2c, ret = PTR_ERR(wm5100->regmap); dev_err(&i2c->dev, "Failed to allocate register map: %d\n", ret); - goto err_alloc; + goto err; } for (i = 0; i < ARRAY_SIZE(wm5100->fll); i++) @@ -2910,8 +2911,7 @@ err_core: wm5100->core_supplies); err_regmap: regmap_exit(wm5100->regmap); -err_alloc: - kfree(wm5100); +err: return ret; } -- 1.7.7.3

1 0

Re: [alsa-devel] [PATCH] ASoC: samsung: drop unknown Kconfig symbol
by Mark Brown 24 Nov '11

24 Nov '11

On Thu, Nov 24, 2011 at 06:17:17PM +0100, Paul Bolle wrote: > There's no Kconfig symbol 'MACH_NEO1973_GTA01'. So drop all Kconfig > references to that symbol and all code that depends on > CONFIG_MACH_NEO1973_GTA01. This also allows for some related cleanups > (eg, drop some checks for CONFIG_MACH_NEO1973_GTA02, as they have become > useless). Update a few strings too, to make clear that the Openmoko Neo > Freerunner (GTA02) is supported while the Openmoko Neo 1973 (GTA01) > isn't. There's people somewhat actively working on the OpenMoko devices out of tree (and I personally have a GTA01) and I'd rather not do anything that causes any more hassle for them. I'd suggest discussing this with the OpenMoko guys if you want to do it.

1 0

[alsa-devel] WM8974 on AT91SAM9263EK
by Matthias Wieloch 24 Nov '11

24 Nov '11

Hello, I hope you can help me find the missing part of my problem. I would like to the WM8974 sound codec instead of the AC97 codec of the eval board. After starting the kernel (2.6.38) no error messages are shown and the sound card got detected by ALSA. But as soon as I run either aplay or arecord I run into a timeout due to pcm_write/pcm_read error. Additionally the dmesg buffer gets messed up. After analysing with an oscilloscope I'm sure that I2C is working and that the codecs responses at least with ACK. But I couldn't see any changes on the SSC lines. So far I disabled all peripherals and GPIOs that use the same pins as SSC1. Looking at the register values the SSC peripheral seems not to be initialised although the application flow is similar to the AT91SAM9g20EK I based my code on. What I find strange is that the WM8974 codec does not provide a .set_sysclk function although most other codecs do. Do I have to call other functions to set up the SSC or rather where should these be set? My current source file located in <kernel>/sound/soc/atmel can be found at http://pastebin.com/mrFFxxdF Regards Matthias

2 1

[alsa-devel] What's the deal with elements?
by Borg Onion 24 Nov '11

24 Nov '11

Hi, Still writing some basic ALSA software, trying to get to know the API. I wrote some software to list cards and their properties, with output like this: $ ./list Card #0: id=ICH6, driver=ICH4, name=Intel ICH6, longname=Intel ICH6 with AD1981B at irq 17, mixername=Analog Devices AD1981B, components=AC97a:41445374 Element used = 0, count = 28 Power state = D0 CTL type = Kernel level CTL Card #1: id=Device, driver=USB-Audio, name=C-Media USB Audio Device, longname=C-Media USB Audio Device at usb-0000:00:1d.0-1, full speed, mixername=USB Mixer, components=USB0d8c:0008 Element used = 0, count = 7 Power state = D0 CTL type = Kernel level CTL No more cards! Whenever I try to access the elements by their index though, I get errors. What's the meaning of "used" vs. "count"? What are these elements used for? I would assume to discover what devices are on a card, but I could be wrong. How does one do device discovery? Thanks, --BO

2 3

[alsa-devel] [patch] ALSA: hda - cut and paste typo in cs420x_models[]
by Dan Carpenter 24 Nov '11

24 Nov '11

The CS420X_IMAC27 was copied from the line before but CS420X_APPLE was clearly intented. Signed-off-by: Dan Carpenter <dan.carpenter(a)oracle.com> diff --git a/sound/pci/hda/patch_cirrus.c b/sound/pci/hda/patch_cirrus.c index 7bd2a52..70a7abd 100644 --- a/sound/pci/hda/patch_cirrus.c +++ b/sound/pci/hda/patch_cirrus.c @@ -1278,7 +1278,7 @@ static const char * const cs420x_models[CS420X_MODELS] = { [CS420X_MBP53] = "mbp53", [CS420X_MBP55] = "mbp55", [CS420X_IMAC27] = "imac27", - [CS420X_IMAC27] = "apple", + [CS420X_APPLE] = "apple", [CS420X_AUTO] = "auto", };

2 1

[alsa-devel] [PATCH] ALSA: Convert mips directory to module_platform_driver
by Axel Lin 24 Nov '11

24 Nov '11

Factor out some boilerplate code. Signed-off-by: Axel Lin <axel.lin(a)gmail.com> --- This is a resend of "ASoC: Convert mips directory to module_platform_driver". Fix the subject line and Cc Takashi. Axel sound/mips/hal2.c | 13 +------------ sound/mips/sgio2audio.c | 13 +------------ 2 files changed, 2 insertions(+), 24 deletions(-) diff --git a/sound/mips/hal2.c b/sound/mips/hal2.c index 2e6c858..5f88d1f 100644 --- a/sound/mips/hal2.c +++ b/sound/mips/hal2.c @@ -935,15 +935,4 @@ static struct platform_driver hal2_driver = { } }; -static int __init alsa_card_hal2_init(void) -{ - return platform_driver_register(&hal2_driver); -} - -static void __exit alsa_card_hal2_exit(void) -{ - platform_driver_unregister(&hal2_driver); -} - -module_init(alsa_card_hal2_init); -module_exit(alsa_card_hal2_exit); +module_platform_driver(hal2_driver); diff --git a/sound/mips/sgio2audio.c b/sound/mips/sgio2audio.c index 69425d4..ceaa593 100644 --- a/sound/mips/sgio2audio.c +++ b/sound/mips/sgio2audio.c @@ -976,15 +976,4 @@ static struct platform_driver sgio2audio_driver = { } }; -static int __init alsa_card_sgio2audio_init(void) -{ - return platform_driver_register(&sgio2audio_driver); -} - -static void __exit alsa_card_sgio2audio_exit(void) -{ - platform_driver_unregister(&sgio2audio_driver); -} - -module_init(alsa_card_sgio2audio_init) -module_exit(alsa_card_sgio2audio_exit) +module_platform_driver(sgio2audio_driver); -- 1.7.5.4

2 1