November 2016 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] Modifying PCM buffer before DMA
by Ricard Wanderlof 30 Nov '16

30 Nov '16

Working with a SoC which has an I2S interface which reuquires 16-bit-swapped 32 bit samples (i.e. the first and 16 halfwords are swapped, but within the halfwords the bytes themselves are little endian), I'm trying to figure out a way to adjust the data in the PCM driver so that ALSA sees standard S32_LE samples. My approach is to use a tight loop to halfword-swap the data. This works fine for capture, where the swap routine can be run triggered by the DMA interrupt as each period arrives from the I2S interface, before calling snd_pcm_period_elapsed(). The problem I have is for playback. I want to have SNDRV_PCM_INFO_MMAP set in snd_pcm_hardware.info in order to be use plugins such as dmix, which in practice means that the PCM driver doesn't really know exactly how much data the application has written before the DMA starts writing it out to the I2S interface. As far as I can tell, the application layer asks the driver for the hw_ptr but there is no way the driver can actually know how much data the application has in fact written. Contrary to the capture case, once the DMA interrupt for a previous period has been triggered, the data for the current period will already be on its way out to the I2S interface, courtesy of the DMA. Disabling MMAP seems to be one way to get the PCM driver to take greater control of the transfer (for instance via snd_pcm_ops->copy() ), but I want to avoid that as mentioned above. My question is, is there another way? /Ricard -- Ricard Wolf Wanderlöf ricardw(at)axis.com Axis Communications AB, Lund, Sweden www.axis.com Phone +46 46 272 2016 Fax +46 46 13 61 30

1 0

Re: [alsa-devel] [PATCH v2] ALSA SoC MAX98927 driver - Revision
by Mark Brown 30 Nov '16

30 Nov '16

On Fri, Nov 25, 2016 at 02:02:47PM +0900, Ryan Lee wrote: > Signed-off-by: Ryan Lee <RyanS.Lee(a)maximintegrated.com> > --- > .../devicetree/bindings/sound/max98927.txt | 33 +- > sound/soc/codecs/max98927.c | 839 +++++------ > sound/soc/codecs/max98927.h | 1463 ++++---------------- This is modifying existing code, you need to at least attempt to supply a description of the changes you are trying to make and most likely send a series of individual, split out changes. Please, review the patch submission process in SubmittingPatches before sending anything further.

1 0

[alsa-devel] [PATCH] ASoC: intel: Replace kthread with work
by Takashi Iwai 30 Nov '16

30 Nov '16

The usage pattern of kthread worker in Intel SST drivers can be replaced gracefully with the normal workqueue, which is more light- weight and easier to manage in general. Let's do it. While in the replacement, move the schedule_work() call inside the spinlock for excluding the race, too. Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- Vinod, I tested this only lightly, so it'd be helpful if the patch is checked on the affected devices before merging. sound/soc/intel/baytrail/sst-baytrail-ipc.c | 3 +- sound/soc/intel/common/sst-ipc.c | 58 ++++++++++------------------- sound/soc/intel/common/sst-ipc.h | 4 +- sound/soc/intel/haswell/sst-haswell-ipc.c | 3 +- sound/soc/intel/skylake/skl-sst-cldma.c | 1 - sound/soc/intel/skylake/skl-sst-ipc.c | 2 +- sound/soc/intel/skylake/skl-sst-ipc.h | 1 - 7 files changed, 24 insertions(+), 48 deletions(-) diff --git a/sound/soc/intel/baytrail/sst-baytrail-ipc.c b/sound/soc/intel/baytrail/sst-baytrail-ipc.c index 7ab14ce65a73..260447da32b8 100644 --- a/sound/soc/intel/baytrail/sst-baytrail-ipc.c +++ b/sound/soc/intel/baytrail/sst-baytrail-ipc.c @@ -23,7 +23,6 @@ #include <linux/slab.h> #include <linux/delay.h> #include <linux/platform_device.h> -#include <linux/kthread.h> #include <linux/firmware.h> #include <linux/io.h> #include <asm/div64.h> @@ -338,7 +337,7 @@ static irqreturn_t sst_byt_irq_thread(int irq, void *context) spin_unlock_irqrestore(&sst->spinlock, flags); /* continue to send any remaining messages... */ - kthread_queue_work(&ipc->kworker, &ipc->kwork); + schedule_work(&ipc->kwork); return IRQ_HANDLED; } diff --git a/sound/soc/intel/common/sst-ipc.c b/sound/soc/intel/common/sst-ipc.c index 6c672ac79cce..d20356255d0c 100644 --- a/sound/soc/intel/common/sst-ipc.c +++ b/sound/soc/intel/common/sst-ipc.c @@ -26,7 +26,6 @@ #include <linux/sched.h> #include <linux/delay.h> #include <linux/platform_device.h> -#include <linux/kthread.h> #include <sound/asound.h> #include "sst-dsp.h" @@ -109,10 +108,9 @@ static int ipc_tx_message(struct sst_generic_ipc *ipc, u64 header, ipc->ops.tx_data_copy(msg, tx_data, tx_bytes); list_add_tail(&msg->list, &ipc->tx_list); + schedule_work(&ipc->kwork); spin_unlock_irqrestore(&ipc->dsp->spinlock, flags); - kthread_queue_work(&ipc->kworker, &ipc->kwork); - if (wait) return tx_wait_done(ipc, msg, rx_data); else @@ -156,35 +154,32 @@ static int msg_empty_list_init(struct sst_generic_ipc *ipc) return -ENOMEM; } -static void ipc_tx_msgs(struct kthread_work *work) +static void ipc_tx_msgs(struct work_struct *work) { struct sst_generic_ipc *ipc = container_of(work, struct sst_generic_ipc, kwork); struct ipc_message *msg; - unsigned long flags; - spin_lock_irqsave(&ipc->dsp->spinlock, flags); + spin_lock_irq(&ipc->dsp->spinlock); - if (list_empty(&ipc->tx_list) || ipc->pending) { - spin_unlock_irqrestore(&ipc->dsp->spinlock, flags); - return; - } - - /* if the DSP is busy, we will TX messages after IRQ. - * also postpone if we are in the middle of procesing completion irq*/ - if (ipc->ops.is_dsp_busy && ipc->ops.is_dsp_busy(ipc->dsp)) { - dev_dbg(ipc->dev, "ipc_tx_msgs dsp busy\n"); - spin_unlock_irqrestore(&ipc->dsp->spinlock, flags); - return; - } + while (!list_empty(&ipc->tx_list) && !ipc->pending) { + /* if the DSP is busy, we will TX messages after IRQ. + * also postpone if we are in the middle of processing + * completion irq + */ + if (ipc->ops.is_dsp_busy && ipc->ops.is_dsp_busy(ipc->dsp)) { + dev_dbg(ipc->dev, "ipc_tx_msgs dsp busy\n"); + break; + } - msg = list_first_entry(&ipc->tx_list, struct ipc_message, list); - list_move(&msg->list, &ipc->rx_list); + msg = list_first_entry(&ipc->tx_list, struct ipc_message, list); + list_move(&msg->list, &ipc->rx_list); - if (ipc->ops.tx_msg != NULL) - ipc->ops.tx_msg(ipc, msg); + if (ipc->ops.tx_msg != NULL) + ipc->ops.tx_msg(ipc, msg); + } - spin_unlock_irqrestore(&ipc->dsp->spinlock, flags); + spin_unlock_irq(&ipc->dsp->spinlock); } int sst_ipc_tx_message_wait(struct sst_generic_ipc *ipc, u64 header, @@ -280,19 +275,7 @@ int sst_ipc_init(struct sst_generic_ipc *ipc) if (ret < 0) return -ENOMEM; - /* start the IPC message thread */ - kthread_init_worker(&ipc->kworker); - ipc->tx_thread = kthread_run(kthread_worker_fn, - &ipc->kworker, "%s", - dev_name(ipc->dev)); - if (IS_ERR(ipc->tx_thread)) { - dev_err(ipc->dev, "error: failed to create message TX task\n"); - ret = PTR_ERR(ipc->tx_thread); - kfree(ipc->msg); - return ret; - } - - kthread_init_work(&ipc->kwork, ipc_tx_msgs); + INIT_WORK(&ipc->kwork, ipc_tx_msgs); return 0; } EXPORT_SYMBOL_GPL(sst_ipc_init); @@ -301,8 +284,7 @@ void sst_ipc_fini(struct sst_generic_ipc *ipc) { int i; - if (ipc->tx_thread) - kthread_stop(ipc->tx_thread); + cancel_work_sync(&ipc->kwork); if (ipc->msg) { for (i = 0; i < IPC_EMPTY_LIST_SIZE; i++) { diff --git a/sound/soc/intel/common/sst-ipc.h b/sound/soc/intel/common/sst-ipc.h index ceb7e468a3fa..ea79cff5c6cc 100644 --- a/sound/soc/intel/common/sst-ipc.h +++ b/sound/soc/intel/common/sst-ipc.h @@ -23,7 +23,6 @@ #include <linux/list.h> #include <linux/workqueue.h> #include <linux/sched.h> -#include <linux/kthread.h> #define IPC_MAX_MAILBOX_BYTES 256 @@ -65,8 +64,7 @@ struct sst_generic_ipc { struct list_head empty_list; wait_queue_head_t wait_txq; struct task_struct *tx_thread; - struct kthread_worker kworker; - struct kthread_work kwork; + struct work_struct kwork; bool pending; struct ipc_message *msg; int tx_data_max_size; diff --git a/sound/soc/intel/haswell/sst-haswell-ipc.c b/sound/soc/intel/haswell/sst-haswell-ipc.c index e432a31fd9f2..a3459d1682a6 100644 --- a/sound/soc/intel/haswell/sst-haswell-ipc.c +++ b/sound/soc/intel/haswell/sst-haswell-ipc.c @@ -26,7 +26,6 @@ #include <linux/delay.h> #include <linux/sched.h> #include <linux/platform_device.h> -#include <linux/kthread.h> #include <linux/firmware.h> #include <linux/dma-mapping.h> #include <linux/debugfs.h> @@ -818,7 +817,7 @@ static irqreturn_t hsw_irq_thread(int irq, void *context) spin_unlock_irqrestore(&sst->spinlock, flags); /* continue to send any remaining messages... */ - kthread_queue_work(&ipc->kworker, &ipc->kwork); + schedule_work(&ipc->kwork); return IRQ_HANDLED; } diff --git a/sound/soc/intel/skylake/skl-sst-cldma.c b/sound/soc/intel/skylake/skl-sst-cldma.c index efa2532114ba..c9f6d87381db 100644 --- a/sound/soc/intel/skylake/skl-sst-cldma.c +++ b/sound/soc/intel/skylake/skl-sst-cldma.c @@ -17,7 +17,6 @@ #include <linux/device.h> #include <linux/mm.h> -#include <linux/kthread.h> #include <linux/delay.h> #include "../common/sst-dsp.h" #include "../common/sst-dsp-priv.h" diff --git a/sound/soc/intel/skylake/skl-sst-ipc.c b/sound/soc/intel/skylake/skl-sst-ipc.c index 797cf4053235..499c0b15e5d0 100644 --- a/sound/soc/intel/skylake/skl-sst-ipc.c +++ b/sound/soc/intel/skylake/skl-sst-ipc.c @@ -464,7 +464,7 @@ irqreturn_t skl_dsp_irq_thread_handler(int irq, void *context) skl_ipc_int_enable(dsp); /* continue to send any remaining messages... */ - kthread_queue_work(&ipc->kworker, &ipc->kwork); + schedule_work(&ipc->kwork); return IRQ_HANDLED; } diff --git a/sound/soc/intel/skylake/skl-sst-ipc.h b/sound/soc/intel/skylake/skl-sst-ipc.h index 0334ed4af031..5c37442a6db4 100644 --- a/sound/soc/intel/skylake/skl-sst-ipc.h +++ b/sound/soc/intel/skylake/skl-sst-ipc.h @@ -16,7 +16,6 @@ #ifndef __SKL_IPC_H #define __SKL_IPC_H -#include <linux/kthread.h> #include <linux/irqreturn.h> #include "../common/sst-ipc.h" -- 2.10.2

3 3

[alsa-devel] [PATCH] ASoC: nau8825: lock longer to avoid playback pop upon resume
by John Hsu 30 Nov '16

30 Nov '16

There is an issue about pop noise in NAU88L25 as follows. Issue 54078: Chell_headphone pop back from S3 (1)Play directly to hw, bypassing CRAS: sox -b 16 -n -t alsa hw:0,0 synth sine 200 sine 200 (2)Close lid or powerd_dbus_suspend, then press a key to resume. (3)no audio after resume (4)Audio will be back after close then reopen the pcm device. After verification, we find one defect is that semaphone lock is not long enough and expired. In this situation, the playback comes back early but pauses a while to wait for the crosstalk detection done. But the detection spends too much time and lock time is up. Therefore, the playback and jack detection sequence interfere to each other. That breaks sequence and makes noise. The driver extends the lock time for the issue. Signed-off-by: John Hsu <KCHSU0(a)nuvoton.com> --- sound/soc/codecs/nau8825.c | 12 ++++++------ 1 file changed, 6 insertions(+), 6 deletions(-) diff --git a/sound/soc/codecs/nau8825.c b/sound/soc/codecs/nau8825.c index 3f9137f..2b89569 100644 --- a/sound/soc/codecs/nau8825.c +++ b/sound/soc/codecs/nau8825.c @@ -1233,7 +1233,7 @@ static int nau8825_hw_params(struct snd_pcm_substream *substream, struct nau8825 *nau8825 = snd_soc_codec_get_drvdata(codec); unsigned int val_len = 0, osr; - nau8825_sema_acquire(nau8825, 2 * HZ); + nau8825_sema_acquire(nau8825, 3 * HZ); /* CLK_DAC or CLK_ADC = OSR * FS * DAC or ADC clock frequency is defined as Over Sampling Rate (OSR) @@ -1293,7 +1293,7 @@ static int nau8825_set_dai_fmt(struct snd_soc_dai *codec_dai, unsigned int fmt) struct nau8825 *nau8825 = snd_soc_codec_get_drvdata(codec); unsigned int ctrl1_val = 0, ctrl2_val = 0; - nau8825_sema_acquire(nau8825, 2 * HZ); + nau8825_sema_acquire(nau8825, 3 * HZ); switch (fmt & SND_SOC_DAIFMT_MASTER_MASK) { case SND_SOC_DAIFMT_CBM_CFM: @@ -2121,7 +2121,7 @@ static int nau8825_configure_sysclk(struct nau8825 *nau8825, int clk_id, * fered by cross talk process, the driver make the playback * preparation halted until cross talk process finish. */ - nau8825_sema_acquire(nau8825, 2 * HZ); + nau8825_sema_acquire(nau8825, 3 * HZ); nau8825_configure_mclk_as_sysclk(regmap); /* MCLK not changed by clock tree */ regmap_update_bits(regmap, NAU8825_REG_CLK_DIVIDER, @@ -2169,7 +2169,7 @@ static int nau8825_configure_sysclk(struct nau8825 *nau8825, int clk_id, * fered by cross talk process, the driver make the playback * preparation halted until cross talk process finish. */ - nau8825_sema_acquire(nau8825, 2 * HZ); + nau8825_sema_acquire(nau8825, 3 * HZ); /* Higher FLL reference input frequency can only set lower * gain error, such as 0000 for input reference from MCLK * 12.288Mhz. @@ -2191,7 +2191,7 @@ static int nau8825_configure_sysclk(struct nau8825 *nau8825, int clk_id, * fered by cross talk process, the driver make the playback * preparation halted until cross talk process finish. */ - nau8825_sema_acquire(nau8825, 2 * HZ); + nau8825_sema_acquire(nau8825, 3 * HZ); /* If FLL reference input is from low frequency source, * higher error gain can apply such as 0xf which has * the most sensitive gain error correction threshold, @@ -2217,7 +2217,7 @@ static int nau8825_configure_sysclk(struct nau8825 *nau8825, int clk_id, * fered by cross talk process, the driver make the playback * preparation halted until cross talk process finish. */ - nau8825_sema_acquire(nau8825, 2 * HZ); + nau8825_sema_acquire(nau8825, 3 * HZ); /* If FLL reference input is from low frequency source, * higher error gain can apply such as 0xf which has * the most sensitive gain error correction threshold, -- 2.6.4

2 1

[alsa-devel] [PATCH v2] ASoC: intel: Fix crash at suspend/resume without card registration
by Takashi Iwai 30 Nov '16

30 Nov '16

When ASoC Intel SST Medfield driver is probed but without codec / card assigned, it causes an Oops and freezes the kernel at suspend/resume, PM: Suspending system (freeze) Suspending console(s) (use no_console_suspend to debug) BUG: unable to handle kernel NULL pointer dereference at 0000000000000018 IP: [<ffffffffc09d9409>] sst_soc_prepare+0x19/0xa0 [snd_soc_sst_mfld_platform] Oops: 0000 [#1] PREEMPT SMP CPU: 0 PID: 1552 Comm: systemd-sleep Tainted: G W 4.9.0-rc6-1.g5f5c2ad-default #1 Call Trace: [<ffffffffb45318f9>] dpm_prepare+0x209/0x460 [<ffffffffb4531b61>] dpm_suspend_start+0x11/0x60 [<ffffffffb40d3cc2>] suspend_devices_and_enter+0xb2/0x710 [<ffffffffb40d462e>] pm_suspend+0x30e/0x390 [<ffffffffb40d2eba>] state_store+0x8a/0x90 [<ffffffffb43c670f>] kobj_attr_store+0xf/0x20 [<ffffffffb42b0d97>] sysfs_kf_write+0x37/0x40 [<ffffffffb42b02bc>] kernfs_fop_write+0x11c/0x1b0 [<ffffffffb422be68>] __vfs_write+0x28/0x140 [<ffffffffb43728a8>] ? apparmor_file_permission+0x18/0x20 [<ffffffffb433b2ab>] ? security_file_permission+0x3b/0xc0 [<ffffffffb422d095>] vfs_write+0xb5/0x1a0 [<ffffffffb422e3d6>] SyS_write+0x46/0xa0 [<ffffffffb4719fbb>] entry_SYSCALL_64_fastpath+0x1e/0xad Add proper NULL checks in the PM code of mdfld driver. Cc: <stable(a)vger.kernel.org> Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- v1->v2: drop the bogus return code from sst_soc_complete(). sound/soc/intel/atom/sst-mfld-platform-pcm.c | 6 ++++++ 1 file changed, 6 insertions(+) diff --git a/sound/soc/intel/atom/sst-mfld-platform-pcm.c b/sound/soc/intel/atom/sst-mfld-platform-pcm.c index 25c6d87c818e..76bcb947b31c 100644 --- a/sound/soc/intel/atom/sst-mfld-platform-pcm.c +++ b/sound/soc/intel/atom/sst-mfld-platform-pcm.c @@ -771,6 +771,9 @@ static int sst_soc_prepare(struct device *dev) struct sst_data *drv = dev_get_drvdata(dev); struct snd_soc_pcm_runtime *rtd; + if (!drv->soc_card) + return 0; + /* suspend all pcms first */ snd_soc_suspend(drv->soc_card->dev); snd_soc_poweroff(drv->soc_card->dev); @@ -793,6 +796,9 @@ static void sst_soc_complete(struct device *dev) struct sst_data *drv = dev_get_drvdata(dev); struct snd_soc_pcm_runtime *rtd; + if (!drv->soc_card) + return; + /* restart SSPs */ list_for_each_entry(rtd, &drv->soc_card->rtd_list, list) { struct snd_soc_dai *dai = rtd->cpu_dai; -- 2.10.2

3 2

[alsa-devel] [PATCH] ASoC: cht_bsw_rt5672: Use HID translation unit
by Takashi Iwai 30 Nov '16

30 Nov '16

Instead of hard-coded "i2c-10EC5670:00", use the translation helper to avoid the mismatch between i2c-codec and ACPI strings just like what we've done for bytcr_rt5640. This gives more robust binding on funky devices like Dell Wyse 3040. Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/soc/intel/boards/cht_bsw_rt5672.c | 30 ++++++++++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-) diff --git a/sound/soc/intel/boards/cht_bsw_rt5672.c b/sound/soc/intel/boards/cht_bsw_rt5672.c index df9d254baa18..fe607d8701a0 100644 --- a/sound/soc/intel/boards/cht_bsw_rt5672.c +++ b/sound/soc/intel/boards/cht_bsw_rt5672.c @@ -25,12 +25,14 @@ #include <sound/jack.h> #include "../../codecs/rt5670.h" #include "../atom/sst-atom-controls.h" +#include "../common/sst-acpi.h" /* The platform clock #3 outputs 19.2Mhz clock to codec as I2S MCLK */ #define CHT_PLAT_CLK_3_HZ 19200000 #define CHT_CODEC_DAI "rt5670-aif1" static struct snd_soc_jack cht_bsw_headset; +static char cht_bsw_codec_name[16]; /* Headset jack detection DAPM pins */ static struct snd_soc_jack_pin cht_bsw_headset_pins[] = { @@ -295,7 +297,7 @@ static int cht_suspend_pre(struct snd_soc_card *card) struct snd_soc_codec *codec; list_for_each_entry(codec, &card->codec_dev_list, card_list) { - if (!strcmp(codec->component.name, "i2c-10EC5670:00")) { + if (!strcmp(codec->component.name, cht_bsw_codec_name)) { dev_dbg(codec->dev, "disabling jack detect before going to suspend.\n"); rt5670_jack_suspend(codec); break; @@ -309,7 +311,7 @@ static int cht_resume_post(struct snd_soc_card *card) struct snd_soc_codec *codec; list_for_each_entry(codec, &card->codec_dev_list, card_list) { - if (!strcmp(codec->component.name, "i2c-10EC5670:00")) { + if (!strcmp(codec->component.name, cht_bsw_codec_name)) { dev_dbg(codec->dev, "enabling jack detect for resume.\n"); rt5670_jack_resume(codec); break; @@ -335,9 +337,33 @@ static struct snd_soc_card snd_soc_card_cht = { .resume_post = cht_resume_post, }; +#define RT5672_I2C_DEFAULT "i2c-10EC5670:00" + static int snd_cht_mc_probe(struct platform_device *pdev) { int ret_val = 0; + struct sst_acpi_mach *mach = pdev->dev.platform_data; + const char *i2c_name; + int i; + + strcpy(cht_bsw_codec_name, RT5672_I2C_DEFAULT); + + /* fixup codec name based on HID */ + if (mach) { + i2c_name = sst_acpi_find_name_from_hid(mach->id); + if (i2c_name) { + snprintf(cht_bsw_codec_name, sizeof(cht_bsw_codec_name), + "i2c-%s", i2c_name); + for (i = 0; i < ARRAY_SIZE(cht_dailink); i++) { + if (!strcmp(cht_dailink[i].codec_name, + RT5672_I2C_DEFAULT)) { + cht_dailink[i].codec_name = + cht_bsw_codec_name; + break; + } + } + } + } /* register the soc card */ snd_soc_card_cht.dev = &pdev->dev; -- 2.10.2

4 4

[alsa-devel] [PATCH 00/10] ASoC: sunxi: Add support for audio codec in A23/H3 SoCs
by Chen-Yu Tsai 30 Nov '16

30 Nov '16

Hi everyone, This series adds support for the audio codec found in Allwinner A23 and H3 SoCs. The design and data paths are similar to the audio codec found in earlier SoCs such as the A31. The analog audio paths are symmetrical with left/right channels and down-mix selectors for mono differential output. What deviates from previous SoCs is that the analog path controls have been moved to a separate control bus, accessed through a message box like register interface in the PRCM block. This necessitates writing a separate component driver for it, which is then tied into the sound card as an ASoC auxiliary device. Patches 1 and 2 add the binding and driver for the analog path control block. This is the more complete version. Previously I provided a not fully tested version to Mylene Josserand from Free Electrons to use with the A33. Their version was then trimmed down and switched to SOC_DAPM_SINGLE controls. Mine implements all the commonly used paths, and uses the new stereo SOC_DAPM_DOUBLE controls. I also have a separate patch for "Phone In" in case anyone wants to test them. It is not included as my hardware does not use that input. As for the A33, the analog controls are exactly the same as the A23, so this driver can be reused, but the PCM and DAI bits are completely different. Patch 3 adds the analog path controls block to the sun6i-prcm driver as a sub-device, for the A23. The H3 currently does not use the PRCM driver. Patch 4 adds PCM and card support for the A23 codec to the sun4i-codec driver. Patch 5 adds a device node for the analog path controls block to the A23 dtsi. Patch 6 adds a device node for the audio codec, and the phandle for the analog path controls block to the A23 dtsi. Patch 7 enables the audio codec for the A23 Q8 tablets. On these tablets the headphone output is driven in DC coupled, or "direct drive", mode. Patch 8 adds PCM and card support for the H3 codec to the sun4i-codec driver. Patch 9 adds device nodes for the audio codec and analog path controls block to the H3 dtsi. Patch 10 enables the audio codec on the Orange Pi PC. The audio output jack on the board is tied to the line out pins on the SoC. Please take a look and let me know what you think. In addition, the sun4i-codec driver is getting pretty large. Maybe we want to split the different parts into different files? Regards ChenYu Chen-Yu Tsai (10): ASoC: sunxi: Add bindings for A23/A33/H3 codec's analog path controls ASoC: sunxi: Add support for A23/A33/H3 codec's analog path controls mfd: sun6i-prcm: Add codec analog controls sub-device for Allwinner A23 ASoC: sun4i-codec: Add support for A23 codec ARM: dts: sun8i: Add codec analog path controls node in PRCM for A23/A33 ARM: dts: sun8i-a23: Add device node for internal audio codec ARM: dts: sun8i-a23: q8-tablet: Enable internal audio codec ASoC: sun4i-codec: Add support for H3 codec ARM: dts: sun8i-h3: Add device nodes for audio codec and its analog controls ARM: dts: sun8i-h3: orange-pi-pc: Enable audio codec .../devicetree/bindings/sound/sun4i-codec.txt | 14 +- .../bindings/sound/sun8i-codec-analog.txt | 16 + arch/arm/boot/dts/sun8i-a23-a33.dtsi | 4 + arch/arm/boot/dts/sun8i-a23-q8-tablet.dts | 23 + arch/arm/boot/dts/sun8i-a23.dtsi | 16 + arch/arm/boot/dts/sun8i-h3-orangepi-pc.dts | 8 + arch/arm/boot/dts/sun8i-h3.dtsi | 19 + drivers/mfd/sun6i-prcm.c | 17 + sound/soc/sunxi/Kconfig | 8 + sound/soc/sunxi/Makefile | 1 + sound/soc/sunxi/sun4i-codec.c | 179 ++++++ sound/soc/sunxi/sun8i-codec-analog.c | 665 +++++++++++++++++++++ 12 files changed, 968 insertions(+), 2 deletions(-) create mode 100644 Documentation/devicetree/bindings/sound/sun8i-codec-analog.txt create mode 100644 sound/soc/sunxi/sun8i-codec-analog.c -- 2.10.2

5 21

[alsa-devel] [PATCH resend] ASoC: Intel: Add missing 10EC5672 ACPI ID matching for Cherry Trail
by Takashi Iwai 30 Nov '16

30 Nov '16

Add the missing ACPI ID 10EC5672 for Cherry Trail, which bounds with cht-bsw-rt5672 driver. This combination was found on Dell Wyse 3040. Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- Resent, as I forgot Cc to alsa-devel ML, sorry! sound/soc/intel/atom/sst/sst_acpi.c | 2 ++ 1 file changed, 2 insertions(+) diff --git a/sound/soc/intel/atom/sst/sst_acpi.c b/sound/soc/intel/atom/sst/sst_acpi.c index 0a88537ca58a..f4d92bbc5373 100644 --- a/sound/soc/intel/atom/sst/sst_acpi.c +++ b/sound/soc/intel/atom/sst/sst_acpi.c @@ -452,6 +452,8 @@ static struct sst_acpi_mach sst_acpi_bytcr[] = { static struct sst_acpi_mach sst_acpi_chv[] = { {"10EC5670", "cht-bsw-rt5672", "intel/fw_sst_22a8.bin", "cht-bsw", NULL, &chv_platform_data }, + {"10EC5672", "cht-bsw-rt5672", "intel/fw_sst_22a8.bin", "cht-bsw", NULL, + &chv_platform_data }, {"10EC5645", "cht-bsw-rt5645", "intel/fw_sst_22a8.bin", "cht-bsw", NULL, &chv_platform_data }, {"10EC5650", "cht-bsw-rt5645", "intel/fw_sst_22a8.bin", "cht-bsw", NULL, -- 2.10.2

3 2

[alsa-devel] [PATCH] ASoC: rt5670: Add missing 10EC5072 ACPI ID
by Takashi Iwai 30 Nov '16

30 Nov '16

rt5670 driver supports also RT5672 codec, but its ACPI ID is missing. This was found on Dell Wyse 3040 box. Signed-off-by: Takashi Iwai <tiwai(a)suse.de> --- sound/soc/codecs/rt5670.c | 1 + 1 file changed, 1 insertion(+) diff --git a/sound/soc/codecs/rt5670.c b/sound/soc/codecs/rt5670.c index 12d3bc14a558..98c7efbc6ee0 100644 --- a/sound/soc/codecs/rt5670.c +++ b/sound/soc/codecs/rt5670.c @@ -2813,6 +2813,7 @@ MODULE_DEVICE_TABLE(i2c, rt5670_i2c_id); #ifdef CONFIG_ACPI static const struct acpi_device_id rt5670_acpi_match[] = { { "10EC5670", 0}, + { "10EC5672", 0}, { }, }; MODULE_DEVICE_TABLE(acpi, rt5670_acpi_match); -- 2.10.2

2 1

[alsa-devel] [PATCH 1/1] ALSA: SOC: DMA: increment buffer pointer atomically
by Jiada Wang 30 Nov '16

30 Nov '16

From: Andreas Pape <apape(a)de.adit-jv.com> Setting pointer and afterwards check for wrap around leads to the possibility of returning the inconsistent pointer position. This patch increments buffer pointer atomically to avoid this issue. Signed-off-by: Andreas Pape <apape(a)de.adit-jv.com> Signed-off-by: Jiada Wang <jiada_wang(a)mentor.com> --- sound/core/pcm_dmaengine.c | 8 +++++--- 1 file changed, 5 insertions(+), 3 deletions(-) diff --git a/sound/core/pcm_dmaengine.c b/sound/core/pcm_dmaengine.c index 8eb58c7..6f6da11 100644 --- a/sound/core/pcm_dmaengine.c +++ b/sound/core/pcm_dmaengine.c @@ -139,12 +139,14 @@ EXPORT_SYMBOL_GPL(snd_dmaengine_pcm_set_config_from_dai_data); static void dmaengine_pcm_dma_complete(void *arg) { + unsigned int new_pos; struct snd_pcm_substream *substream = arg; struct dmaengine_pcm_runtime_data *prtd = substream_to_prtd(substream); - prtd->pos += snd_pcm_lib_period_bytes(substream); - if (prtd->pos >= snd_pcm_lib_buffer_bytes(substream)) - prtd->pos = 0; + new_pos = prtd->pos + snd_pcm_lib_period_bytes(substream); + if (new_pos >= snd_pcm_lib_buffer_bytes(substream)) + new_pos = 0; + prtd->pos = new_pos; snd_pcm_period_elapsed(substream); } -- 2.9.3

2 3