October 2012 - Alsa-devel - mailman.alsa-project.org

Re: [alsa-devel] [PATCH 5/5] snd-ice1712: Fix resume on ice1724
by Ondrej Zary 31 Oct '12

31 Oct '12

On Wednesday 17 October 2012 08:57:39 Takashi Iwai wrote: > At Sun, 14 Oct 2012 21:09:23 +0200, > > Ondrej Zary wrote: > > set_pro_rate() is called from hw_params() but not from prepare(), > > breaking running PCM on suspend/resume. Call it from prepare() if PCM was > > suspended to fix the problem. > > > > Signed-off-by: Ondrej Zary <linux(a)rainbow-software.org> > > This should be fixed in the resume callback. Could you check the > patch below instead? I've now verified that this patch fixes a real problem. When "multi track rate locking" is disabled and some sound is played at non-default (non-48kHz) frequency during suspend, the frequency is wrong (48kHz) after resume. This patch fixes that. > thanks, > > Takashi > > --- > diff --git a/sound/pci/ice1712/ice1724.c b/sound/pci/ice1712/ice1724.c > index 161c8f6..5cc3fa5 100644 > --- a/sound/pci/ice1712/ice1724.c > +++ b/sound/pci/ice1712/ice1724.c > @@ -2867,7 +2867,12 @@ static int snd_vt1724_resume(struct device *dev) > ice->set_spdif_clock(ice, 0); > } else { > /* internal on-card clock */ > - snd_vt1724_set_pro_rate(ice, ice->pro_rate_default, 1); > + int rate; > + if (ice->cur_rate) > + rate = ice->cur_rate; > + else > + rate = ice->pro_rate_default; > + snd_vt1724_set_pro_rate(ice, rate, 1); > } > > update_spdif_bits(ice, ice->pm_saved_spdif_ctrl); > -- > To unsubscribe from this list: send the line "unsubscribe linux-kernel" in > the body of a message to majordomo(a)vger.kernel.org > More majordomo info at http://vger.kernel.org/majordomo-info.html > Please read the FAQ at http://www.tux.org/lkml/ -- Ondrej Zary

2 1

[alsa-devel] possibly memory leak in alsa-lib
by Ivan Sorokin 30 Oct '12

30 Oct '12

The following program shows a few leaks in valgrind (ubuntu 12.04, alsa-lib 1.0.25, eglibc 2.15): #include <alsa/asoundlib.h> int main() { snd_pcm_t* pcm = 0; snd_pcm_open(&pcm, "default", SND_PCM_STREAM_PLAYBACK, 0); snd_pcm_close(pcm); snd_config_update_free_global(); } Do I close pcm correctly? In file MEMORY-LEAK it is written that program should call snd_config_update_free_global() to free global configuration cache and I do so. I'm not sure, but probably some leaks are caused by glibc.

2 2

[alsa-devel] Store codec capabilites in alsa
by Bharath Ramachandramurthy 30 Oct '12

30 Oct '12

Hello Is there a mechanism for the codec driver to register the codec capabilities to the ALSA framework ? The BE CPU dai driver can then read the capabilities of the codec and talk to the DSP. Thanks Bharath

4 7

[alsa-devel] [PATCH 1/5] ARM: at91: atmel-ssc: add platform device id table
by Bo Shen 30 Oct '12

30 Oct '12

Using platform device id to check whether the Atmel SOC support dma for data transfer If match "at91rm9200_ssc" which does not support DMA transfer If match "at91sam9g45_ssc" which support DMA transfer Signed-off-by: Bo Shen <voice.shen(a)atmel.com> --- arch/arm/mach-at91/at91rm9200.c | 6 +++--- arch/arm/mach-at91/at91rm9200_devices.c | 6 +++--- arch/arm/mach-at91/at91sam9260.c | 2 +- arch/arm/mach-at91/at91sam9260_devices.c | 2 +- arch/arm/mach-at91/at91sam9261.c | 6 +++--- arch/arm/mach-at91/at91sam9261_devices.c | 6 +++--- arch/arm/mach-at91/at91sam9263.c | 4 ++-- arch/arm/mach-at91/at91sam9263_devices.c | 4 ++-- arch/arm/mach-at91/at91sam9g45.c | 4 ++-- arch/arm/mach-at91/at91sam9g45_devices.c | 4 ++-- arch/arm/mach-at91/at91sam9rl.c | 4 ++-- arch/arm/mach-at91/at91sam9rl_devices.c | 4 ++-- drivers/misc/atmel-ssc.c | 23 +++++++++++++++++++++++ include/linux/atmel-ssc.h | 5 +++++ 14 files changed, 54 insertions(+), 26 deletions(-) diff --git a/arch/arm/mach-at91/at91rm9200.c b/arch/arm/mach-at91/at91rm9200.c index b4f0565..85d53c5 100644 --- a/arch/arm/mach-at91/at91rm9200.c +++ b/arch/arm/mach-at91/at91rm9200.c @@ -184,9 +184,9 @@ static struct clk_lookup periph_clocks_lookups[] = { CLKDEV_CON_DEV_ID("t0_clk", "atmel_tcb.1", &tc3_clk), CLKDEV_CON_DEV_ID("t1_clk", "atmel_tcb.1", &tc4_clk), CLKDEV_CON_DEV_ID("t2_clk", "atmel_tcb.1", &tc5_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.0", &ssc0_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.1", &ssc1_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.2", &ssc2_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.0", &ssc0_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.1", &ssc1_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.2", &ssc2_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91rm9200", &twi_clk), /* fake hclk clock */ CLKDEV_CON_DEV_ID("hclk", "at91_ohci", &ohci_clk), diff --git a/arch/arm/mach-at91/at91rm9200_devices.c b/arch/arm/mach-at91/at91rm9200_devices.c index a563189..59ceea1 100644 --- a/arch/arm/mach-at91/at91rm9200_devices.c +++ b/arch/arm/mach-at91/at91rm9200_devices.c @@ -752,7 +752,7 @@ static struct resource ssc0_resources[] = { }; static struct platform_device at91rm9200_ssc0_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 0, .dev = { .dma_mask = &ssc0_dmamask, @@ -794,7 +794,7 @@ static struct resource ssc1_resources[] = { }; static struct platform_device at91rm9200_ssc1_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 1, .dev = { .dma_mask = &ssc1_dmamask, @@ -836,7 +836,7 @@ static struct resource ssc2_resources[] = { }; static struct platform_device at91rm9200_ssc2_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 2, .dev = { .dma_mask = &ssc2_dmamask, diff --git a/arch/arm/mach-at91/at91sam9260.c b/arch/arm/mach-at91/at91sam9260.c index ad29f93..2c8aab0 100644 --- a/arch/arm/mach-at91/at91sam9260.c +++ b/arch/arm/mach-at91/at91sam9260.c @@ -210,7 +210,7 @@ static struct clk_lookup periph_clocks_lookups[] = { CLKDEV_CON_DEV_ID("t0_clk", "atmel_tcb.1", &tc3_clk), CLKDEV_CON_DEV_ID("t1_clk", "atmel_tcb.1", &tc4_clk), CLKDEV_CON_DEV_ID("t2_clk", "atmel_tcb.1", &tc5_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.0", &ssc_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.0", &ssc_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9260", &twi_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9g20", &twi_clk), /* more usart lookup table for DT entries */ diff --git a/arch/arm/mach-at91/at91sam9260_devices.c b/arch/arm/mach-at91/at91sam9260_devices.c index 805ef95..9cfdc3f 100644 --- a/arch/arm/mach-at91/at91sam9260_devices.c +++ b/arch/arm/mach-at91/at91sam9260_devices.c @@ -742,7 +742,7 @@ static struct resource ssc_resources[] = { }; static struct platform_device at91sam9260_ssc_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 0, .dev = { .dma_mask = &ssc_dmamask, diff --git a/arch/arm/mach-at91/at91sam9261.c b/arch/arm/mach-at91/at91sam9261.c index 8d999eb..4e8c56e 100644 --- a/arch/arm/mach-at91/at91sam9261.c +++ b/arch/arm/mach-at91/at91sam9261.c @@ -174,9 +174,9 @@ static struct clk_lookup periph_clocks_lookups[] = { CLKDEV_CON_DEV_ID("t0_clk", "atmel_tcb.0", &tc0_clk), CLKDEV_CON_DEV_ID("t1_clk", "atmel_tcb.0", &tc1_clk), CLKDEV_CON_DEV_ID("t2_clk", "atmel_tcb.0", &tc2_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.0", &ssc0_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.1", &ssc1_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.2", &ssc2_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.0", &ssc0_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.1", &ssc1_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.2", &ssc2_clk), CLKDEV_CON_DEV_ID("hclk", "at91_ohci", &hck0), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9261", &twi_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9g10", &twi_clk), diff --git a/arch/arm/mach-at91/at91sam9261_devices.c b/arch/arm/mach-at91/at91sam9261_devices.c index 9752f17..299637f 100644 --- a/arch/arm/mach-at91/at91sam9261_devices.c +++ b/arch/arm/mach-at91/at91sam9261_devices.c @@ -706,7 +706,7 @@ static struct resource ssc0_resources[] = { }; static struct platform_device at91sam9261_ssc0_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 0, .dev = { .dma_mask = &ssc0_dmamask, @@ -748,7 +748,7 @@ static struct resource ssc1_resources[] = { }; static struct platform_device at91sam9261_ssc1_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 1, .dev = { .dma_mask = &ssc1_dmamask, @@ -790,7 +790,7 @@ static struct resource ssc2_resources[] = { }; static struct platform_device at91sam9261_ssc2_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 2, .dev = { .dma_mask = &ssc2_dmamask, diff --git a/arch/arm/mach-at91/at91sam9263.c b/arch/arm/mach-at91/at91sam9263.c index 6a01d03..95a5471 100644 --- a/arch/arm/mach-at91/at91sam9263.c +++ b/arch/arm/mach-at91/at91sam9263.c @@ -186,8 +186,8 @@ static struct clk *periph_clocks[] __initdata = { static struct clk_lookup periph_clocks_lookups[] = { /* One additional fake clock for macb_hclk */ CLKDEV_CON_ID("hclk", &macb_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.0", &ssc0_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.1", &ssc1_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.0", &ssc0_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.1", &ssc1_clk), CLKDEV_CON_DEV_ID("mci_clk", "atmel_mci.0", &mmc0_clk), CLKDEV_CON_DEV_ID("mci_clk", "atmel_mci.1", &mmc1_clk), CLKDEV_CON_DEV_ID("spi_clk", "atmel_spi.0", &spi0_clk), diff --git a/arch/arm/mach-at91/at91sam9263_devices.c b/arch/arm/mach-at91/at91sam9263_devices.c index 8dde220..df89a00 100644 --- a/arch/arm/mach-at91/at91sam9263_devices.c +++ b/arch/arm/mach-at91/at91sam9263_devices.c @@ -1199,7 +1199,7 @@ static struct resource ssc0_resources[] = { }; static struct platform_device at91sam9263_ssc0_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 0, .dev = { .dma_mask = &ssc0_dmamask, @@ -1241,7 +1241,7 @@ static struct resource ssc1_resources[] = { }; static struct platform_device at91sam9263_ssc1_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 1, .dev = { .dma_mask = &ssc1_dmamask, diff --git a/arch/arm/mach-at91/at91sam9g45.c b/arch/arm/mach-at91/at91sam9g45.c index 84af1b5..f4f96a6 100644 --- a/arch/arm/mach-at91/at91sam9g45.c +++ b/arch/arm/mach-at91/at91sam9g45.c @@ -239,8 +239,8 @@ static struct clk_lookup periph_clocks_lookups[] = { CLKDEV_CON_DEV_ID("t0_clk", "atmel_tcb.1", &tcb0_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9g10.0", &twi0_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9g10.1", &twi1_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.0", &ssc0_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.1", &ssc1_clk), + CLKDEV_CON_DEV_ID("pclk", "at91sam9g45_ssc.0", &ssc0_clk), + CLKDEV_CON_DEV_ID("pclk", "at91sam9g45_ssc.1", &ssc1_clk), CLKDEV_CON_DEV_ID(NULL, "atmel-trng", &trng_clk), CLKDEV_CON_DEV_ID(NULL, "atmel_sha", &aestdessha_clk), CLKDEV_CON_DEV_ID(NULL, "atmel_tdes", &aestdessha_clk), diff --git a/arch/arm/mach-at91/at91sam9g45_devices.c b/arch/arm/mach-at91/at91sam9g45_devices.c index b159607..27e3bf6 100644 --- a/arch/arm/mach-at91/at91sam9g45_devices.c +++ b/arch/arm/mach-at91/at91sam9g45_devices.c @@ -1459,7 +1459,7 @@ static struct resource ssc0_resources[] = { }; static struct platform_device at91sam9g45_ssc0_device = { - .name = "ssc", + .name = "at91sam9g45_ssc", .id = 0, .dev = { .dma_mask = &ssc0_dmamask, @@ -1501,7 +1501,7 @@ static struct resource ssc1_resources[] = { }; static struct platform_device at91sam9g45_ssc1_device = { - .name = "ssc", + .name = "at91sam9g45_ssc", .id = 1, .dev = { .dma_mask = &ssc1_dmamask, diff --git a/arch/arm/mach-at91/at91sam9rl.c b/arch/arm/mach-at91/at91sam9rl.c index 72e9084..4110b54 100644 --- a/arch/arm/mach-at91/at91sam9rl.c +++ b/arch/arm/mach-at91/at91sam9rl.c @@ -184,8 +184,8 @@ static struct clk_lookup periph_clocks_lookups[] = { CLKDEV_CON_DEV_ID("t0_clk", "atmel_tcb.0", &tc0_clk), CLKDEV_CON_DEV_ID("t1_clk", "atmel_tcb.0", &tc1_clk), CLKDEV_CON_DEV_ID("t2_clk", "atmel_tcb.0", &tc2_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.0", &ssc0_clk), - CLKDEV_CON_DEV_ID("pclk", "ssc.1", &ssc1_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.0", &ssc0_clk), + CLKDEV_CON_DEV_ID("pclk", "at91rm9200_ssc.1", &ssc1_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9g20.0", &twi0_clk), CLKDEV_CON_DEV_ID(NULL, "i2c-at91sam9g20.1", &twi1_clk), CLKDEV_CON_ID("pioA", &pioA_clk), diff --git a/arch/arm/mach-at91/at91sam9rl_devices.c b/arch/arm/mach-at91/at91sam9rl_devices.c index d6ca054..01220c7 100644 --- a/arch/arm/mach-at91/at91sam9rl_devices.c +++ b/arch/arm/mach-at91/at91sam9rl_devices.c @@ -832,7 +832,7 @@ static struct resource ssc0_resources[] = { }; static struct platform_device at91sam9rl_ssc0_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 0, .dev = { .dma_mask = &ssc0_dmamask, @@ -874,7 +874,7 @@ static struct resource ssc1_resources[] = { }; static struct platform_device at91sam9rl_ssc1_device = { - .name = "ssc", + .name = "at91rm9200_ssc", .id = 1, .dev = { .dma_mask = &ssc1_dmamask, diff --git a/drivers/misc/atmel-ssc.c b/drivers/misc/atmel-ssc.c index ac00f83..f40abd8 100644 --- a/drivers/misc/atmel-ssc.c +++ b/drivers/misc/atmel-ssc.c @@ -68,6 +68,26 @@ void ssc_free(struct ssc_device *ssc) } EXPORT_SYMBOL(ssc_free); +static struct atmel_ssc_platform_data at91rm9200_config = { + .use_dma = 0, +}; + +static struct atmel_ssc_platform_data at91sam9g45_config = { + .use_dma = 1, +}; + +static const struct platform_device_id atmel_ssc_devtypes[] = { + { + .name = "at91rm9200_ssc", + .driver_data = (unsigned long) &at91rm9200_config, + }, { + .name = "at91sam9g45_ssc", + .driver_data = (unsigned long) &at91sam9g45_config, + }, { + /* sentinel */ + } +}; + static int ssc_probe(struct platform_device *pdev) { struct resource *regs; @@ -80,6 +100,8 @@ static int ssc_probe(struct platform_device *pdev) } ssc->pdev = pdev; + ssc->pdata = (struct atmel_ssc_platform_data *) + platform_get_device_id(pdev)->driver_data; regs = platform_get_resource(pdev, IORESOURCE_MEM, 0); if (!regs) { @@ -139,6 +161,7 @@ static struct platform_driver ssc_driver = { .name = "ssc", .owner = THIS_MODULE, }, + .id_table = atmel_ssc_devtypes, .probe = ssc_probe, .remove = __devexit_p(ssc_remove), }; diff --git a/include/linux/atmel-ssc.h b/include/linux/atmel-ssc.h index 4eb3175..1ca0e32 100644 --- a/include/linux/atmel-ssc.h +++ b/include/linux/atmel-ssc.h @@ -5,10 +5,15 @@ #include <linux/list.h> #include <linux/io.h> +struct atmel_ssc_platform_data { + int use_dma; +}; + struct ssc_device { struct list_head list; void __iomem *regs; struct platform_device *pdev; + struct atmel_ssc_platform_data *pdata; struct clk *clk; int user; int irq; -- 1.7.9.5

3 11

[alsa-devel] [PATCH 1/2] sound: soc: tlv320aic32x4: Fix problem with first capture.
by Javier Martin 30 Oct '12

30 Oct '12

In its previous status, the first capture didn't work properly; nothing was actually recorded from the microphone. This behaviour was observed using a Visstrim M10 board. In order to solve this BUG a workaround has been added that, during the initialization process of the codec, powers on and off the ADC. The issue seems related to a HW BUG or some behavior that is not documented in the datasheet. Signed-off-by: Javier Martin <javier.martin(a)vista-silicon.com> --- sound/soc/codecs/tlv320aic32x4.c | 10 ++++++++++ sound/soc/codecs/tlv320aic32x4.h | 3 +++ 2 files changed, 13 insertions(+) diff --git a/sound/soc/codecs/tlv320aic32x4.c b/sound/soc/codecs/tlv320aic32x4.c index b0a73d3..aad92f9 100644 --- a/sound/soc/codecs/tlv320aic32x4.c +++ b/sound/soc/codecs/tlv320aic32x4.c @@ -675,6 +675,16 @@ static int aic32x4_probe(struct snd_soc_codec *codec) ARRAY_SIZE(aic32x4_snd_controls)); aic32x4_add_widgets(codec); + /* + * Workaround: for an unknown reason, the ADC needs to be powered up + * and down for the first capture to work properly. It seems related to + * a HW BUG or some kind of behavior not documented in the datasheet. + */ + tmp_reg = snd_soc_read(codec, AIC32X4_ADCSETUP); + snd_soc_write(codec, AIC32X4_ADCSETUP, tmp_reg | + AIC32X4_LADC_EN | AIC32X4_RADC_EN); + snd_soc_write(codec, AIC32X4_ADCSETUP, tmp_reg); + return 0; } diff --git a/sound/soc/codecs/tlv320aic32x4.h b/sound/soc/codecs/tlv320aic32x4.h index aae2b24..3577422 100644 --- a/sound/soc/codecs/tlv320aic32x4.h +++ b/sound/soc/codecs/tlv320aic32x4.h @@ -94,6 +94,9 @@ #define AIC32X4_WORD_LEN_24BITS 0x02 #define AIC32X4_WORD_LEN_32BITS 0x03 +#define AIC32X4_LADC_EN (1 << 7) +#define AIC32X4_RADC_EN (1 << 6) + #define AIC32X4_I2S_MODE 0x00 #define AIC32X4_DSP_MODE 0x01 #define AIC32X4_RIGHT_JUSTIFIED_MODE 0x02 -- 1.7.9.5

2 2

[alsa-devel] emu 1212m devel, firmware
by oeaiyu 30 Oct '12

30 Oct '12

Aloha! the thing is about emu 1212m - yep i know it's not working, but i can help to debug some and try to do something with it, i'm programming sometimes but not C, but i can debug and send you bugtrack and i can log win7 x64 if you really need it, i'm trying to not using it to spend some time to discover linux bugs - it's funny thing =)) alotofw3rkz hea i got 4 sound cards in comp i'm musician - you can listen some of my music @ symbiants.com i can send you some if you wish to. and for writing music 7 years ago or so i've bought that emu card +creative x-fi + ati r6xx hdmi + hda ati sb ad1988b not sure about r6xx. but all others are working, but emu it shows me e-mu 1010b PCI MAEM8960 but it is e-mu 1212m pci actually - the chip is similar but it's other it is like 1818m, of course the chip itself is audigy2 but the connectors are others it is very simple - i think just need to reroute some in / out also i know exactly that in latest drivers there was a firmware update (2.0.0.18 or something latest for win7) so the firmware is in the bundle that can be extracted from installer. well about in out some more - i got some front / surround /center / lfe / side sliders and more - i can send print screen if needed, but in hardware level i got adat/spdif in/out left \ right in jacks left \ right out jacks midi in \ out and special for micro-dock (this is why it is ":m") input - ethernet like that is not active also there are special pins on the card that splits card for two pieces and those ananlog in+out are with midi together and all spdif adat on the other part of card that interconnected via harddrive like ide-cable. well right now the main bug for me is that when i'm enabling this emu card, the emu-firmware is loading and it gets about 25% of CPU well this is a 1 full core, so i think that it would take the whole 100% if those cores wasn't separated or something. well as i suppose this is the firmware problem that is not suited for this hardware. -- RA OEAI

1 0

[alsa-devel] usb audio race at disconnect time
by Matthieu CASTET 30 Oct '12

30 Oct '12

Hi, while doing some monkey tests on a product we found races in usb audio code when the device in unplugged from usb (on linus master tree). This can be reproduced with usb_audio_show_race.diff and CONFIG_DEBUG_SLAB. With this patch, start a stream : # arecord -D hw:0 -r 44100 -c 2 -f S16_LE > /dev/null you will see the kernel log : "in snd_usb_hw_params sleeping" Unplug the device before "in snd_usb_hw_params sleeping exit", and you will see an oops in snd_pcm_hw_params Instead of using CONFIG_DEBUG_SLAB, usb_audio_show_use_after_free.diff can show use after free by setting usb_device pointer to NULL in snd_usb_audio_disconnect. [1] In order to protect from all the races, before using any usb_device we need to check if it is not freed. What's the best way to do that ? A trival fix would be to take a mutex in all snd_pcm_ops callback that access usb : { mutex_lock(&chip->shutdown_mutex); if (!chip->dev) { mutex_unlock(&chip->shutdown_mutex); return -ENODEV; } [...] mutex_unlock(&chip->shutdown_mutex); } But that will serialize all snd_pcm_ops callbacks. Another solution could be to use refcounting or rwlock patern : - snd_pcm_ops callbacks call rdlock_trylock/rdlock_unlock - usb_driver disconnect callback take rwlock_lock and never release it This will make "usb_driver disconnect" wait that all "snd_pcm_ops callback" are finished and abort all future call to "snd_pcm_ops callback". I believe similar problems exist in other drivers, (we saw another one in hidraw [2]), so it could be nice to find a generic pattern to help driver to manage correctly disconnection. Are there any recommended pattern ? Do you have any comment ? Thanks, Matthieu [1] [ 1366.315307] usb 1-2.1.3: new full-speed USB device number 4 using ehci-omap [ 1366.556579] input: USB Audio as /devices/platform/usbhs_omap/ehci-omap.0/usb1 /1-2/1-2.1/1-2.1.3/1-2.1.3:1.3/input/input1 [ 1366.604553] hid-generic 0003:06F8:C000.0001: input,hidraw0: USB HID v1.00 Dev ice [USB Audio] on usb-ehci-omap.0-2.1.3/input3 # # # arecord -D hw:0 -r 44100 -c 2 -f S16_LE > /dev/null Recording WAVE '[ 1368.937774] in snd_usb_hw_params sleeping stdin' : Signed 16 bit Little Endian, Rate 44100 Hz, Stereo Warning: rate is not accurate (requested = 44100Hz, got = 48000Hz) please, try the plug plugin [ 1372.385375] hid-generic 0003:06F8:C000.0001: can't reset device, ehci-omap.0- 2.1.3/input3, status -71 [ 1372.439514] usb 1-2.1.3: USB disconnect, device number 4 [ 1372.449005] hid-generic 0003:06F8:C000.0001: can't reset device, ehci-omap.0- 2.1.3/input3, status -71 [ 1378.947235] in snd_usb_hw_params sleeping exit [ 1378.952301] Unable to handle kernel NULL pointer dereference at virtual addre ss 0000029c [ 1378.961029] pgd = c31e8000 [ 1378.964019] [0000029c] *pgd=8e885831, *pte=00000000, *ppte=00000000 [ 1378.970825] Internal error: Oops: 17 [#1] SMP ARM [ 1378.975799] Modules linked in: [ 1378.979064] CPU: 0 Not tainted (3.6.0-03889-ge8dc7a6-dirty #5) [ 1378.985595] PC is at usb_ifnum_to_if+0xc/0xd4 [ 1378.990234] LR is at snd_usb_hw_params+0x1bc/0x764 [ 1378.995300] pc : [<c0275338>] lr : [<c030816c>] psr: 00000013 [ 1378.995300] sp : cf3fddf0 ip : cf3fde08 fp : cf3fde04 [ 1379.007415] r10: 0000bb80 r9 : 00000001 r8 : 00000000 [ 1379.012908] r7 : cf972f40 r6 : cf3e1f14 r5 : ce8ad000 r4 : cf3e1eb0 [ 1379.019805] r3 : 00000004 r2 : 00000005 r1 : 00000002 r0 : 00000000 [ 1379.026702] Flags: nzcv IRQs on FIQs on Mode SVC_32 ISA ARM Segment user [ 1379.034240] Control: 10c5387d Table: 831e8019 DAC: 00000015 [ 1379.040313] Process arecord (pid: 636, stack limit = 0xcf3fc2f8) [ 1379.046661] Stack: (0xcf3fddf0 to 0xcf3fe000) [ 1379.051269] dde0: cf3e1eb0 ce8ad000 cf3fd e5c cf3fde08 [ 1379.059906] de00: c030816c c0275338 cf3fde3c cf3fde18 c02f5d68 c02f59e4 c05b4 c60 ce8ad000 [ 1379.068542] de20: cf3e1800 00000001 00000004 00000001 cf3fde5c cf3e1800 ce8ad 800 c05b8310 [ 1379.077178] de40: ce8ad000 c0014728 cf3fc000 00000000 cf3fde7c cf3fde60 c02f0 b8c c0307fbc [ 1379.085815] de60: cf3e1800 be91b890 be91b890 ce8ad000 cf3fdec4 cf3fde80 c02f1 778 c02f0ad4 [ 1379.094451] de80: cf3fc000 60000013 cf97b540 00000002 c0126484 00000000 cf3fd ed4 cf3e1800 [ 1379.103088] dea0: be91b890 cf3e5078 c0045877 c0014728 cf3fc000 00000000 cf3fd efc cf3fdec8 [ 1379.111755] dec0: c02f2150 c02f1228 cf3fdf3c cf3fded8 c0126634 c0068ef0 00000 001 00000000 [ 1379.120391] dee0: cf87f280 be91b890 cf3e5078 c0045877 cf3fdf0c cf3fdf00 c02f2 5dc c02f2124 [ 1379.129028] df00: cf3fdf7c cf3fdf10 c00fd6b4 c02f25ac c03d7328 cf8b49c8 00000 026 00000000 [ 1379.137664] df20: cd673738 00000002 cf3fc000 00000000 cf3fdf6c cf3fdf40 c00ec 138 c0126420 [ 1379.146301] df40: 00000000 00000000 00000006 cf8b49c0 be919140 00000000 cf87f 280 be91b890 [ 1379.154907] df60: c25c4111 00000004 c0014728 cf3fc000 cf3fdfa4 cf3fdf80 c00fd c6c c00fd638 [ 1379.163543] df80: c00183f0 00000000 00028288 b6fc68a0 00028238 00000036 00000 000 cf3fdfa8 [ 1379.172180] dfa0: c00145a0 c00fdc38 00028288 b6fc68a0 00000004 c25c4111 be91b 890 00020001 [ 1379.180816] dfc0: 00028288 b6fc68a0 00028238 00000036 be91b890 0000bb80 0000a c44 be91bb74 [ 1379.189453] dfe0: 00028288 be91b760 b6fa41c8 b6e05aec 20000010 00000004 007ff f00 00ffdf00 [ 1379.198089] Backtrace: [ 1379.200683] [<c027532c>] (usb_ifnum_to_if+0x0/0xd4) from [<c030816c>] (snd_us b_hw_params+0x1bc/0x764) [ 1379.210418] r5:ce8ad000 r4:cf3e1eb0 [ 1379.214263] [<c0307fb0>] (snd_usb_hw_params+0x0/0x764) from [<c02f0b8c>] (snd _pcm_hw_params+0xc4/0x368) [ 1379.224182] [<c02f0ac8>] (snd_pcm_hw_params+0x0/0x368) from [<c02f1778>] (snd _pcm_common_ioctl1+0x55c/0xefc) [ 1379.234558] r7:ce8ad000 r6:be91b890 r5:be91b890 r4:cf3e1800 [ 1379.240600] [<c02f121c>] (snd_pcm_common_ioctl1+0x0/0xefc) from [<c02f2150>] (snd_pcm_capture_ioctl1+0x38/0x488) [ 1379.251342] [<c02f2118>] (snd_pcm_capture_ioctl1+0x0/0x488) from [<c02f25dc>] (snd_pcm_capture_ioctl+0x3c/0x40) [ 1379.261962] r7:c0045877 r6:cf3e5078 r5:be91b890 r4:cf87f280 [ 1379.268005] [<c02f25a0>] (snd_pcm_capture_ioctl+0x0/0x40) from [<c00fd6b4>] ( do_vfs_ioctl+0x88/0x600) [ 1379.277770] [<c00fd62c>] (do_vfs_ioctl+0x0/0x600) from [<c00fdc6c>] (sys_ioct l+0x40/0x68) [ 1379.286376] r9:cf3fc000 r8:c0014728 r7:00000004 r6:c25c4111 r5:be91b890 r4:cf87f280 [ 1379.294738] [<c00fdc2c>] (sys_ioctl+0x0/0x68) from [<c00145a0>] (ret_fast_sys call+0x0/0x30) [ 1379.303558] r7:00000036 r6:00028238 r5:b6fc68a0 r4:00028288 [ 1379.309600] Code: e89da818 e1a0c00d e92dd830 e24cb004 (e590329c) [ 1379.316223] ---[ end trace 48c30133f24e0b57 ]--- [ 1379.321105] Kernel panic - not syncing: Fatal exception [2] [ 349.573974] ------------[ cut here ]------------ [ 349.578796] WARNING: at /opt/mcastet/CIA/fidji-beta28/raptor/kernel/omap/lib/list_debug.c:48 list_del+0x30/0x8c() [ 349.589477] list_del corruption. prev->next should be d95efc44, but was 6b6b6b6b [ 349.597167] Modules linked in: blackberry cdc_acm sierra option usb_wwan hso atmel_mxt_ts cp210x pl2303 usbserial 88w8688_wlan tun omap2_mcspi omap_hsmmc [ 349.611572] [<c013e608>] (unwind_backtrace+0x0/0xf0) from [<c016f7b4>] (warn_slowpath_common+0x4c/0x64) [ 349.621368] [<c016f7b4>] (warn_slowpath_common+0x4c/0x64) from [<c016f84c>] (warn_slowpath_fmt+0x2c/0x3c) [ 349.631347] [<c016f84c>] (warn_slowpath_fmt+0x2c/0x3c) from [<c0298200>] (list_del+0x30/0x8c) [ 349.640228] [<c0298200>] (list_del+0x30/0x8c) from [<c0185d50>] (remove_wait_queue+0x2c/0x6c) [ 349.649108] [<c0185d50>] (remove_wait_queue+0x2c/0x6c) from [<c01f4e64>] (free_poll_entry+0x14/0x20) [ 349.658630] [<c01f4e64>] (free_poll_entry+0x14/0x20) from [<c01f4e88>] (poll_freewait+0x18/0x6c) [ 349.667785] [<c01f4e88>] (poll_freewait+0x18/0x6c) from [<c01f5448>] (do_sys_poll+0x314/0x3a4) [ 349.676757] [<c01f5448>] (do_sys_poll+0x314/0x3a4) from [<c01f56fc>] (sys_poll+0x5c/0xbc) [ 349.685272] [<c01f56fc>] (sys_poll+0x5c/0xbc) from [<c0138f40>] (ret_fast_syscall+0x0/0x30) [ 349.693969] ---[ end trace 1b75b31a2719ed1e ]--- [ 349.698791] ------------[ cut here ]------------ [ 349.703582] WARNING: at /opt/mcastet/CIA/fidji-beta28/raptor/kernel/omap/lib/list_debug.c:51 list_del+0x58/0x8c() [ 349.714263] list_del corruption. next->prev should be d95efc44, but was 6b6b6b6b [ 349.721954] Modules linked in: blackberry cdc_acm sierra option usb_wwan hso atmel_mxt_ts cp210x pl2303 usbserial 88w8688_wlan tun omap2_mcspi omap_hsmmc [ 349.736328] [<c013e608>] (unwind_backtrace+0x0/0xf0) from [<c016f7b4>] (warn_slowpath_common+0x4c/0x64) [ 349.746124] [<c016f7b4>] (warn_slowpath_common+0x4c/0x64) from [<c016f84c>] (warn_slowpath_fmt+0x2c/0x3c) [ 349.756103] [<c016f84c>] (warn_slowpath_fmt+0x2c/0x3c) from [<c0298228>] (list_del+0x58/0x8c) [ 349.764984] [<c0298228>] (list_del+0x58/0x8c) from [<c0185d50>] (remove_wait_queue+0x2c/0x6c) [ 349.773864] [<c0185d50>] (remove_wait_queue+0x2c/0x6c) from [<c01f4e64>] (free_poll_entry+0x14/0x20) [ 349.783386] [<c01f4e64>] (free_poll_entry+0x14/0x20) from [<c01f4e88>] (poll_freewait+0x18/0x6c) [ 349.792541] [<c01f4e88>] (poll_freewait+0x18/0x6c) from [<c01f5448>] (do_sys_poll+0x314/0x3a4) [ 349.801513] [<c01f5448>] (do_sys_poll+0x314/0x3a4) from [<c01f56fc>] (sys_poll+0x5c/0xbc) [ 349.810028] [<c01f56fc>] (sys_poll+0x5c/0xbc) from [<c0138f40>] (ret_fast_syscall+0x0/0x30) [ 349.818725] ---[ end trace 1b75b31a2719ed1f ]--- [ 349.830444] hub 1-1:1.0: unable to enumerate USB device on port 2 [ 350.412475] Slab corruption: size-64 start=d7189ac0, len=64 [ 350.420166] 000: 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b cc 9a 18 d7 [ 350.426544] 010: cc 9a 18 d7 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b 6b [ 350.435699] Prev obj: start=d7189a80, len=64 [ 350.441284] 000: 02 26 00 00 c5 91 18 d7 84 93 18 d7 c4 9d 18 d7 [ 350.447601] 010: 50 91 ad d3 50 9f 18 d7 50 95 ad d3 00 00 00 00 [ 350.454650] Next obj: start=d7189b00, len=64 [ 350.460083] 000: 5d 25 00 00 05 78 e5 d8 c4 9e 18 d7 44 13 6a d9 [ 350.466613] 010: 91 13 6a d9 00 00 00 00 00 00 00 00 00 00 00 00

2 14

[alsa-devel] [PATCH] ASoC: fsi: fsi_set_master_clk() was called from fsi_hw_xxx() only
by Kuninori Morimoto 29 Oct '12

29 Oct '12

Current FSI driver is using fsi_set_master_clk() if it needs system clock. But this function was called from fsi_hw_shutdown()/fsi_dai_trigger()/fsi_resume() without a sense of unity. Because of this, sound playback after suspend failed sometimes. To keep consistency, fsi_master_clk() was called from fsi_hw_start/stop() only now. Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- sound/soc/sh/fsi.c | 23 ++++++++--------------- 1 file changed, 8 insertions(+), 15 deletions(-) diff --git a/sound/soc/sh/fsi.c b/sound/soc/sh/fsi.c index a33d483..e97501b 100644 --- a/sound/soc/sh/fsi.c +++ b/sound/soc/sh/fsi.c @@ -1364,12 +1364,17 @@ static int fsi_hw_startup(struct fsi_priv *fsi, /* fifo init */ fsi_fifo_init(fsi, io, dev); + /* start master clock */ + if (fsi_is_clk_master(fsi)) + fsi_set_master_clk(dev, fsi, fsi->rate, 1); + return 0; } static void fsi_hw_shutdown(struct fsi_priv *fsi, struct device *dev) { + /* stop master clock */ if (fsi_is_clk_master(fsi)) fsi_set_master_clk(dev, fsi, fsi->rate, 0); } @@ -1492,19 +1497,11 @@ static int fsi_dai_hw_params(struct snd_pcm_substream *substream, struct snd_soc_dai *dai) { struct fsi_priv *fsi = fsi_get_priv(substream); - long rate = params_rate(params); - int ret; - - if (!fsi_is_clk_master(fsi)) - return 0; - - ret = fsi_set_master_clk(dai->dev, fsi, rate, 1); - if (ret < 0) - return ret; - fsi->rate = rate; + if (fsi_is_clk_master(fsi)) + fsi->rate = params_rate(params); - return ret; + return 0; } static const struct snd_soc_dai_ops fsi_dai_ops = { @@ -1806,10 +1803,6 @@ static void __fsi_resume(struct fsi_priv *fsi, return; fsi_hw_startup(fsi, io, dev); - - if (fsi_is_clk_master(fsi) && fsi->rate) - fsi_set_master_clk(dev, fsi, fsi->rate, 1); - fsi_stream_start(fsi, io); } -- 1.7.9.5

2 1

[alsa-devel] snd_pcm_sw_params_set_start_threshold() functionality
by irfanks 29 Oct '12

29 Oct '12

Hello, snd_pcm_sw_params_set_start_threshold() API description explains, "PCM is automatically started when playback frames available to PCM are >= threshold" I want to know, whether above pre-buffering(threshold) mechanism is only during the start of playback ? OR Also if my playback device goes to XRUN state, Again next time when i prepare my playback device using snd_pcm_prepare(). Again the prebuffering (threshold) will occur since i have set snd_pcm_sw_params_set_start_threshold() to N periods in software params. Best Regards, Thanks in advance, Irfan...

1 1

[alsa-devel] [PATCH 4/4] ASoC: ak5702: Makefile and Kconfig for ak5702
by Paolo Doz 29 Oct '12

29 Oct '12

ASoC: ak5702: Makefile and Kconfig for ak5702 Signed-off-by: Paolo Doz <paolo.doz(a)gmail.com> --- sound/soc/codecs/Kconfig | 4 ++++ sound/soc/codecs/Makefile | 2 ++ 2 files changed, 6 insertions(+) diff --git a/sound/soc/codecs/Kconfig b/sound/soc/codecs/Kconfig index 6f4f284..c029778 100644 --- a/sound/soc/codecs/Kconfig +++ b/sound/soc/codecs/Kconfig @@ -26,6 +26,7 @@ config SND_SOC_ALL_CODECS select SND_SOC_AK4641 if I2C select SND_SOC_AK4642 if I2C select SND_SOC_AK4671 if I2C + select SND_SOC_AK5702 if I2C select SND_SOC_ALC5623 if I2C select SND_SOC_ALC5632 if I2C select SND_SOC_CQ0093VC if MFD_DAVINCI_VOICECODEC @@ -194,6 +195,9 @@ config SND_SOC_AK4642 config SND_SOC_AK4671 tristate +config SND_SOC_AK5702 + tristate + config SND_SOC_ALC5623 tristate config SND_SOC_ALC5632 diff --git a/sound/soc/codecs/Makefile b/sound/soc/codecs/Makefile index c55d2d7..cc201a0 100644 --- a/sound/soc/codecs/Makefile +++ b/sound/soc/codecs/Makefile @@ -14,6 +14,7 @@ snd-soc-ak4535-objs := ak4535.o snd-soc-ak4641-objs := ak4641.o snd-soc-ak4642-objs := ak4642.o snd-soc-ak4671-objs := ak4671.o +snd-soc-ak5702-objs := ak5702.o snd-soc-arizona-objs := arizona.o snd-soc-cq93vc-objs := cq93vc.o snd-soc-cs42l51-objs := cs42l51.o @@ -134,6 +135,7 @@ obj-$(CONFIG_SND_SOC_AK4535) += snd-soc-ak4535.o obj-$(CONFIG_SND_SOC_AK4641) += snd-soc-ak4641.o obj-$(CONFIG_SND_SOC_AK4642) += snd-soc-ak4642.o obj-$(CONFIG_SND_SOC_AK4671) += snd-soc-ak4671.o +obj-$(CONFIG_SND_SOC_AK5702) += snd-soc-ak5702.o obj-$(CONFIG_SND_SOC_ALC5623) += snd-soc-alc5623.o obj-$(CONFIG_SND_SOC_ALC5632) += snd-soc-alc5632.o obj-$(CONFIG_SND_SOC_ARIZONA) += snd-soc-arizona.o -- 1.7.9.5

1 0