January 2017 - Alsa-devel - mailman.alsa-project.org

[alsa-devel] [PATCH V2] ASoC: rt5651: use msleep for large delays
by Nicholas Mc Guire 11 Jan '17

11 Jan '17

ulseep_range() uses hrtimers and provides no advantage over msleep() for larger delays. Fix up the 70/80ms delays here to use msleep() and reduce the load on the hrtimer subsystem. Link: http://lkml.org/lkml/2017/1/11/377 Signed-off-by: Nicholas Mc Guire <hofrat(a)osadl.org> --- V2: Mark Brown <broonie(a)kernel.org> suggested to take the lower time stamp for such conversions as this is the behavior developers expect rather than the upper that would be closer to the current (somewhat odd) usleep_range() behavior. As these are non-atomic regions and a delay of 70/80 ms implies that there is almost certainty multiple context switches occurring in the sleep time and thus relatively high jitter must be assumed with respect to actual wakeup time implying that there is little point in using high-resolution timers here. Patch was compile tested with: x86_64_defconfig + SND_SOC=m SND_SOC_INTEL_BYTCR_RT5651_MACH=m (implies CONFIG_SND_SOC_RT5651) Patch is aginast 4.10-rc3 (localversion-next is next-20170111) sound/soc/codecs/rt5651.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-) diff --git a/sound/soc/codecs/rt5651.c b/sound/soc/codecs/rt5651.c index f5d3415..fb592b0 100644 --- a/sound/soc/codecs/rt5651.c +++ b/sound/soc/codecs/rt5651.c @@ -802,7 +802,7 @@ static int rt5651_hp_event(struct snd_soc_dapm_widget *w, case SND_SOC_DAPM_PRE_PMD: rt5651->hp_mute = 1; - usleep_range(70000, 75000); + msleep(70); break; default: @@ -822,7 +822,7 @@ static int rt5651_hp_post_event(struct snd_soc_dapm_widget *w, switch (event) { case SND_SOC_DAPM_POST_PMU: if (!rt5651->hp_mute) - usleep_range(80000, 85000); + msleep(80); break; -- 2.1.4

1 0

[alsa-devel] [GIT PULL] ASoC fixes for v4.10
by Mark Brown 11 Jan '17

11 Jan '17

The following changes since commit a121103c922847ba5010819a3f250f1f7fc84ab8: Linux 4.10-rc3 (2017-01-08 14:18:17 -0800) are available in the git repository at: git://git.kernel.org/pub/scm/linux/kernel/git/broonie/sound.git tags/asoc-fix-v4.10-rc3 for you to fetch changes up to 42e0ebdef53d91891939a475c6ef23970829ff99: Merge remote-tracking branches 'asoc/fix/nau8825', 'asoc/fix/rt5645', 'asoc/fix/tlv320aic3x' and 'asoc/fix/topology' into asoc-linus (2017-01-10 10:47:54 +0000) ---------------------------------------------------------------- ASoC: Fixes for v4.10 As well as the usual smattering of driver specific fixes collected since the merge window this has one particularly important fix to the core for handling of aux_devs which was broken during the merge window by some of the componentization refactoring. ---------------------------------------------------------------- Bard Liao (1): ASoC: rt5645: set sel_i2s_pre_div1 to 2 Caleb Crome (1): ASoC: fsl_ssi: set fifo watermark to more reliable value Colin Ian King (2): ASoC: rsnd: don't double free kctrl ASoC: topology: kfree kcontrol->private_value before freeing kcontrol G Kranthi (1): ASoC: Intel: Skylake: Fix to fail safely if module not available in path Jeeja KP (1): ASoC: Intel: Skylake: Release FW ctx in cleanup John Hsu (2): ASoC: nau8825: correct the function name of register ASoC: nau8825: fix invalid configuration in Pre-Scalar of FLL Jose Abreu (1): ASoC: dwc: Fix PIO mode initialization Mark Brown (5): Merge remote-tracking branch 'asoc/fix/component' into asoc-linus Merge remote-tracking branch 'asoc/fix/intel' into asoc-linus Merge remote-tracking branch 'asoc/fix/rcar' into asoc-linus Merge remote-tracking branches 'asoc/fix/arizona', 'asoc/fix/dpcm', 'asoc/fix/dwc', 'asoc/fix/fsl-ssi' and 'asoc/fix/hdmi-codec' into asoc-linus Merge remote-tracking branches 'asoc/fix/nau8825', 'asoc/fix/rt5645', 'asoc/fix/tlv320aic3x' and 'asoc/fix/topology' into asoc-linus Patrick Lai (1): ASoC: dpcm: Avoid putting stream state to STOP when FE stream is paused Peter Ujfalusi (1): ASoC: tlv320aic3x: Mark the RESET register as volatile Pierre-Louis Bossart (2): ASoC: Intel: bytcr_rt5640: fallback mechanism if MCLK is not enabled ASoC: Intel: bytcr-rt5640: fix settings in internal clock mode Richard Fitzgerald (1): ASoC: wm_adsp: Don't overrun firmware file buffer when reading region data Sylwester Nawrocki (1): ASoC: Fix binding and probing of auxiliary components Takashi Sakamoto (1): ASoC: hdmi-codec: use unsigned type to structure members with bit-field include/sound/hdmi-codec.h | 8 ++-- include/sound/soc.h | 3 ++ sound/soc/codecs/nau8825.c | 9 +++-- sound/soc/codecs/nau8825.h | 7 ++-- sound/soc/codecs/rt5645.c | 3 ++ sound/soc/codecs/tlv320aic3x.c | 13 ++++++ sound/soc/codecs/wm_adsp.c | 25 +++++++++++- sound/soc/dwc/designware_i2s.c | 25 ++++++------ sound/soc/fsl/fsl_ssi.c | 74 +++++++++++++++++++++++++---------- sound/soc/intel/boards/bytcr_rt5640.c | 18 +++++++-- sound/soc/intel/skylake/skl-pcm.c | 3 ++ sound/soc/intel/skylake/skl-sst.c | 3 ++ sound/soc/sh/rcar/core.c | 4 +- sound/soc/soc-core.c | 10 ++--- sound/soc/soc-pcm.c | 4 +- sound/soc/soc-topology.c | 3 +- 16 files changed, 150 insertions(+), 62 deletions(-)

2 1

[alsa-devel] [PATCH] ALSA: hda/ca0132 - fix possible NULL pointer use
by Arnd Bergmann 11 Jan '17

11 Jan '17

gcc-7 caught what it considers a NULL pointer dereference: sound/pci/hda/patch_ca0132.c: In function 'dspio_scp.constprop': sound/pci/hda/patch_ca0132.c:1487:4: error: argument 1 null where non-null expected [-Werror=nonnull] This is plausible from looking at the function, as we compare 'reply' to NULL earlier in it. I have not tried to analyze if there are constraints that make it impossible to hit the bug, but adding another NULL check in the end kills the warning and makes the function more robust. Signed-off-by: Arnd Bergmann <arnd(a)arndb.de> --- sound/pci/hda/patch_ca0132.c | 3 +++ 1 file changed, 3 insertions(+) diff --git a/sound/pci/hda/patch_ca0132.c b/sound/pci/hda/patch_ca0132.c index bd7c29f8ab1e..07a9deb17477 100644 --- a/sound/pci/hda/patch_ca0132.c +++ b/sound/pci/hda/patch_ca0132.c @@ -1482,6 +1482,9 @@ static int dspio_scp(struct hda_codec *codec, } else if (ret_size != reply_data_size) { codec_dbg(codec, "RetLen and HdrLen .NE.\n"); return -EINVAL; + } else if (!reply) { + codec_dbg(codec, "NULL reply\n"); + return -EINVAL; } else { *reply_len = ret_size*sizeof(unsigned int); memcpy(reply, scp_reply.data, *reply_len); -- 2.9.0

2 1

[alsa-devel] [PATCH v3 0/2] ASoC: Use DMI name in sound card long name
by mengdong.lin＠linux.intel.com 11 Jan '17

11 Jan '17

From: Mengdong Lin <mengdong.lin(a)linux.intel.com> This series is based on Liam's idea to use DMI name as the sound card short name in the audio workshop. Now we add DMI info to the card long name: - Not change the card name (short name), keep it simple. So the card ID won't change as well. - Use DMI name to make the card long name. Machine drivers can use a new API to get this feature. In user space, when the sound server ask UCM to open a sound card, UCM will try to find the best machine-specific configuration file based on the card long name, and only if not avaiable, fallback to the default configuration file that matches the card name (short name). Please see patch: ucm: Automatically load the best config file based on the card long name History: v1 from RFC: Give up using a table to map DMI info to a simple string, since we cannot maintain such a table for each machine. Just do verbatim copy of DMI info and let user space decide how to use the DMI info in card long name to find the best UCM config file. If device driver adds a flavor string, just append it after the DMI info fields to the card long name. v2: Exclude card name and only keep DMI name in the long name, to save the space since the long name cannot exceed 80 characters. Trim special characters like SPACE in the long name, because UCM configuration files & directories may use the same long name and Autoconf cannot handle special characters. v3: Also include DMI product version in the long name, since vendors like Lenovo may only put self-explanatory name in this field. Liam Girdwood (1): ASoC: core: Add API to use DMI name in sound card long name Mengdong Lin (1): ASoC: Intel: Use DMI name for sound card long name in Broadwell machine driver include/sound/soc.h | 4 ++ sound/soc/intel/boards/broadwell.c | 2 + sound/soc/soc-core.c | 122 +++++++++++++++++++++++++++++++++++++ 3 files changed, 128 insertions(+) -- 2.5.0

4 8

Re: [alsa-devel] [PATCH 2/2] ASoC: cx2072x Add driver for CX2072X CODEC
by Mark Brown 11 Jan '17

11 Jan '17

On Sat, Dec 17, 2016 at 03:52:33PM +0800, simon.ho.cnxt(a)gmail.com wrote: > +/*FIXME: need to move the default settings to device tree*/ No, we'd expect things to be configured from userspace via a binary control. > +static unsigned char cx2072x_eq_coeff_array[MAX_EQ_BAND][MAC_EQ_COEFF] = { > + {0x77, 0x26, 0x13, 0xb3, 0x76, 0x26, 0x0a, 0x3d, 0xd4, 0xe2, 0x04}, > + {0x97, 0x3e, 0xb3, 0x86, 0xc2, 0x3b, 0x4d, 0x79, 0xa7, 0xc5, 0x03}, > + {0x0f, 0x39, 0x76, 0xa3, 0x1b, 0x2b, 0x89, 0x5c, 0xd7, 0xdb, 0x03}, > + {0x21, 0x46, 0xfe, 0xa6, 0xec, 0x24, 0x01, 0x59, 0xf4, 0xd4, 0x03}, > + {0xe9, 0x78, 0x9c, 0xb0, 0x8a, 0x56, 0x64, 0x4f, 0x8d, 0xb0, 0x02}, > + {0x60, 0x6e, 0x57, 0xee, 0xec, 0x18, 0xa8, 0x11, 0xb5, 0xf8, 0x02}, > + {0x5a, 0x14, 0x68, 0xe9, 0x1d, 0x06, 0xb9, 0x5f, 0x68, 0xdc, 0x03}, > +}; > + > +static unsigned char cx2072x_drc_array[MAX_DRC_REGS] = { > + 0x65, 0x55, 0x3C, 0x01, 0x05, 0x39, 0x76, 0x1A, 0x00 > +}; Just use the chip defaults, this avoids any confusion or debate about the values in the kernel. > +#define get_cx2072x_priv(_codec_) \ > + ((struct cx2072x_priv *) snd_soc_codec_get_drvdata(_codec_)) There is no need to cast void pointers, this will at most hide actual errors, and there's no real need for the macro at all really. > + { 36864000, 7 },/* Don't use div 7*/ Does something enforce that? Also coding style, missing space. > +static const struct reg_default cx2072x_reg_defaults[] = { > + { 0x0414, 0x00000003 }, /*2072X_AFG_POWER_STATE */ We have defines for the register names which are helpfully commented here, why not just actually use the defines? > + if (reg == CX2072X_UM_INTERRUPT_CRTL_E) { > + /* Update the MSB byte only */ > + reg += 3; > + size = 1; > + value >>= 24; > + } Use a switch here in case you find any more magic registers. > + ret = i2c_master_send(client, buf, size + 2); > + if (ret == size + 2) { > + ret = 0; > + } else if (ret < 0) { > + dev_err(dev, > + "I2C write address failed\n"); Please print the error code, it makes life easier for people looking at logs with errors. > +static unsigned int get_div_from_mclk(unsigned int mclk) > +{ > + unsigned int div = 8; > + int i = 0; > + > + for (i = 0; i < ARRAY_SIZE(MCLK_PRE_DIV); i++) { > + if (mclk <= MCLK_PRE_DIV[i].mclk) { > + div = MCLK_PRE_DIV[i].div; > + break; > + } > + } > + return div; > +} Why is this function in the middle of all the register I/O functions? > +static int cx2072x_config_headset_det(struct cx2072x_priv *cx2072x) > +{ > + const int interrupt_gpio_pin = 1; > + > + dev_dbg(cx2072x->dev, > + "Configure interrupt pin: %d\n", interrupt_gpio_pin); > + /*No-sticky input type*/ Coding style. There's *lots* of really obvious and bad coding style problems throughout the driver, this is really not good - as well as making things harder to read poor coding style is normally a good indicator that there are other problems. > + /* support both nokia and apple headset set. Monitor time = 275 ms*/ > + regmap_write(cx2072x->regmap, CX2072X_DIGITAL_TEST15, 0x73); > + > + /* Disable TIP detection*/ > + regmap_write(cx2072x->regmap, CX2072X_ANALOG_TEST12, 0x300); > + > + /* Switch MusicD3Live pin to GPIO */ > + regmap_write(cx2072x->regmap, CX2072X_DIGITAL_TEST1, 0); These look awfully like they should be configured by the system integrator via DT. > + regmap_write(cx2072x->regmap, CX2072X_ANALOG_TEST8, > + (unsigned char)int_div & 0xffff); This cast is worrying, why is it needed? > + > + /* clock inversion */ > + switch (fmt & SND_SOC_DAIFMT_INV_MASK) { > + case SND_SOC_DAIFMT_NB_NF: > + is_frame_inv = is_i2s ? 1 : 0; > + is_bclk_inv = is_i2s ? 1 : 0; Please don't abuse the ternery operator like this, write normal logic statements with straightforward assignments. Readers will thank you. > + regmap_write(cx2072x->regmap, CX2072X_I2SPCM_CONTROL1, reg1.ulVal); Please don't use Hungarian notation. > + /*Enables bclk and EAPD pin*/ > + if (cx2072x->rev_id == CX2072X_REV_A2) > + regmap_update_bits(cx2072x->regmap, CX2072X_DIGITAL_BIOS_TEST2, > + 0x84, 0xFF); > + else > + regmap_write(cx2072x->regmap, CX2072X_DIGITAL_BIOS_TEST2, > + regDBT2.ulVal); Again, switch statement. > +static int portg_power_ev(struct snd_soc_dapm_widget *w, > + struct snd_kcontrol *kcontrol, int event) > +{ > + struct snd_soc_codec *codec = snd_soc_dapm_to_codec(w->dapm); > + > + switch (event) { > + case SND_SOC_DAPM_POST_PMU: > + cx2072x_update_dsp(codec); > + break; > + default: > + break; > + } > + return 0; > +} > +static int cx2072x_plbk_dsp_info(struct snd_kcontrol *kcontrol, Blank lines between functions. > + mutex_lock(&cx2072x->eq_coeff_lock); > + memcpy(cache, param, CX2072X_PLBK_EQ_COEF_LEN); > + cx2072x->plbk_dsp_changed = true; Shouldn't this involve a memcmp() to check if the new value is different? > +static const struct snd_kcontrol_new cx2072x_snd_controls[] = { > + > + SOC_DOUBLE_R_TLV("PortD Boost", CX2072X_PORTD_GAIN_LEFT, > + CX2072X_PORTD_GAIN_RIGHT, 0, 3, 0, boost_tlv), All volume controls should end in Volume so userspace knows how to handle them. Also please use normal kernel indentation for the continuation lines. > + SOC_DOUBLE_R("DAC1 Mute", CX2072X_DAC1_AMP_GAIN_LEFT, > + CX2072X_DAC1_AMP_GAIN_RIGHT, 7, 1, 0), All boolean controls should end in Switch so userspace knows how to handle them. > + CX2072X_PLBK_EQ_COEF("DACL EQ 0", 0, 0), > + CX2072X_PLBK_EQ_COEF("DACL EQ 1", 0, 1), > + CX2072X_PLBK_EQ_COEF("DACL EQ 2", 0, 2), > + CX2072X_PLBK_EQ_COEF("DACL EQ 3", 0, 3), > + CX2072X_PLBK_EQ_COEF("DACL EQ 4", 0, 4), > + CX2072X_PLBK_EQ_COEF("DACL EQ 5", 0, 5), > + CX2072X_PLBK_EQ_COEF("DACL EQ 6", 0, 6), > + CX2072X_PLBK_EQ_COEF("DACR EQ 0", 1, 0), > + CX2072X_PLBK_EQ_COEF("DACR EQ 1", 1, 1), > + CX2072X_PLBK_EQ_COEF("DACR EQ 2", 1, 2), > + CX2072X_PLBK_EQ_COEF("DACR EQ 3", 1, 3), > + CX2072X_PLBK_EQ_COEF("DACR EQ 4", 1, 4), > + CX2072X_PLBK_EQ_COEF("DACR EQ 5", 1, 5), > + CX2072X_PLBK_EQ_COEF("DACR EQ 6", 1, 6), This looks like it should be two binary controls, one per EQ. > +int cx2072x_hs_jack_report(struct snd_soc_codec *codec) > +{ > +} > +EXPORT_SYMBOL_GPL(cx2072x_hs_jack_report); We have a standard jack detection interface in the kernel, the driver should be using it rather than writing its own. > +static int cx2072x_set_tdm_slot(struct snd_soc_dai *dai, unsigned int tx_mask, > + unsigned int rx_mask, int slots, int slot_width) > +{ > + struct snd_soc_codec *codec = dai->codec; > + struct cx2072x_priv *cx2072x = snd_soc_codec_get_drvdata(codec); > + > + if (slots == 0) > + goto out; > + > + > + switch (rx_mask) { > + case 1 ... 0xf: > + default: > + return -EINVAL; > + } So all possible values are invalid? That doesn't seemw right... > + switch (clk_id) { > + case CX2072X_MCLK_EXTERNAL_PLL: > + if (cx2072x->mclk && clk_set_rate(cx2072x->mclk, freq)) > + return -EINVAL; Please write sensible error handling, log what went wrong if you get an error and pass back error codes you get from other APIs. This will make it a lot easier for users to figure out what's happened. > +#define CX2072X_DAPM_SUPPLY_S(wname, wsubseq, wreg, wshift, wmask, won_val, \ > + woff_val, wevent, wflags) \ > + {.id = snd_soc_dapm_supply, .name = wname, .kcontrol_news = NULL, \ > + .num_kcontrols = 0, .reg = wreg, .shift = wshift, .mask = wmask, \ > + .on_val = won_val, .off_val = woff_val, \ > + .subseq = wsubseq, .event = wevent, .event_flags = wflags} There appears to be nothing device specific in here, why is this not being added as a generic supply type? > +static void cx2072x_sw_reset(struct cx2072x_priv *cx2072x) > +{ > + > + regmap_write(cx2072x->regmap, CX2072X_AFG_FUNCTION_RESET, 0x01); Coding style, random blank line hre. > + /* reduce the jack monitor time*/ > + regmap_update_bits(cx2072x->regmap, CX2072X_DIGITAL_TEST15, > + 0x00, 0x06); Coding style, please intent contiuation lines... > + cx2072x_config_headset_det(cx2072x); > + /* configre PortC as input device */ > + regmap_update_bits(cx2072x->regmap, CX2072X_PORTC_PIN_CTRL, > + 0x20, 0x20); ...this isn't even consistent with the rest of the same function :( > + switch (level) { > + case SND_SOC_BIAS_ON: > + /* Enable Headset Mic Bias */ > + if (cx2072x->is_biason == 0) > + cx2072x->is_biason = 1; > + break; This looks worrying but since nothing else uses that driver data I'm guessing it doesn't actually do anything. > + case SND_SOC_BIAS_PREPARE: > + if (old_level == SND_SOC_BIAS_STANDBY) { > + if (cx2072x->mclk) { > + dev_dbg(cx2072x->dev, "Turn on MCLK\n"); > + ret = clk_prepare_enable(cx2072x->mclk); > + if (ret) > + return ret; > + } > + } > + break; > + case SND_SOC_BIAS_STANDBY: > + > + if (old_level == SND_SOC_BIAS_OFF) { > + if (cx2072x->mclk) { > + dev_dbg(cx2072x->dev, "Turn on MCLK\n"); > + ret = clk_prepare_enable(cx2072x->mclk); > + if (ret) > + return ret; > + } So we enable MCLK both when coming out of _OFF and when coming out of _STANDBY? Why (and note that we're missing some disables here... > + regmap_read(cx2072x->regmap, CX2072X_VENDOR_ID, &ven_id); > + regmap_read(cx2072x->regmap, CX2072X_REVISION_ID, &cx2072x->rev_id); > + dev_dbg(codec->dev, "codec version: %08x,%08x\n", > + ven_id, cx2072x->rev_id); Log this at info level. > +#ifdef CONFIG_PM > +static int cx2072x_runtime_suspend(struct device *dev) > +{ > + struct cx2072x_priv *cx2072x = dev_get_drvdata(dev); > + > + cx2072x_set_bias_level(cx2072x->codec, SND_SOC_BIAS_OFF); Let the framework take care of managing bias level for you. > +static bool cx2072x_readable_register(struct device *dev, unsigned int reg) > +{ > + switch (reg) { Why is this nowhere near the rest of the regmap code? > +static int cx2072x_i2c_probe(struct i2c_client *i2c, > + const struct i2c_device_id *id) > +{ > + int ret = -1; > + struct cx2072x_priv *cx2072x; > + > + cx2072x = (struct cx2072x_priv *)devm_kzalloc( > + &i2c->dev, sizeof(struct cx2072x_priv), GFP_KERNEL); No need to cast away from void. > + if (cx2072x == NULL) { > + dev_err(&i2c->dev, "Out of memory!\n"); kzalloc() will already be very noisy if this happens. > +static const struct i2c_device_id cx2072x_i2c_id[] = { > + { "cx20721", 0 }, > + { "cx20723", 0 }, > + { "14F10720", 0 }, > + {} > +}; > +MODULE_DEVICE_TABLE(i2c, cx2072x_i2c_id); One of those looks like an ACPI ID not an I2C ID. > + > +static const struct of_device_id cx2072x_of_match[] = { > + { .compatible = "cnxt,cx20721", }, > + { .compatible = "cnxt,cx20723", }, > + { .compatible = "cnxt,cx7601", }, > + {} Why is cx7601 not an I2C ID? > +#ifdef CONFIG_ACPI > + .acpi_match_table = ACPI_PTR(cx2072x_acpi_match), > +#endif No need for the defines here.

3 4

[alsa-devel] [PATCH v4 0/3] support DP MST audio
by libin.yang＠intel.com 11 Jan '17

11 Jan '17

From: Libin Yang <libin.yang(a)intel.com> This patchset starts to support DP MST audio. This patchset is based on drm-tip on git://anongit.freedesktop.org/drm-tip change log: v4: refine some comments in 0001 and 0002 patches. refine the 0003 patch for the new ALSA-Configuration document Libin Yang (3): ALSA: hda - add DP mst verb support ALSA: hda - add DP MST audio support ALSA: Documentation about HDA DP MST pin init and connection Documentation/sound/hd-audio/dp-mst.rst | 17 +++ sound/pci/hda/hda_codec.c | 76 +++++++++- sound/pci/hda/hda_codec.h | 3 + sound/pci/hda/patch_hdmi.c | 245 +++++++++++++++++++++++++------- 4 files changed, 285 insertions(+), 56 deletions(-) -- 2.7.4

4 15

[alsa-devel] [PATCH v6] ASoC: omap-mcbsp: Add PM QoS support for McBSP to prevent glitches
by Matt Ranostay 11 Jan '17

11 Jan '17

We can get audio errors if hitting deeper idle states on omaps: [alsa.c:230] error: Fatal problem with alsa output, error -5. [audio.c:614] error: Error in writing audio (Input/output error?)! This seems to happen with off mode idle enabled as power for the whole SoC may get cut off between filling the McBSP fifo using DMA. While active DMA blocks deeper idle states in hardware, McBSP activity does not seem to do so. Basing the QoS latency calculation on the FIFO size, threshold, sample rate, and channels. Based on the original patch by Tony Lindgren Link: https://patchwork.kernel.org/patch/9305867/ Cc: Tony Lindgren <tony(a)atomide.com> Cc: Peter Ujfalusi <peter.ujfalusi(a)ti.com> Signed-off-by: Matt Ranostay <matt(a)ranostay.consulting> --- hanges from v1: * add calculations for latency per number of FIFO locations Changes from v2: * add missing mcbsp.h header change Changes from v3: * base the latency calculations on threshold, buffer size, sample rate, and channels Changes from v4: * using Peter Ujfalusi's suggestions for restoring a higher latency on audio stream completion, or if not applicable remove the QoS request Changes from v5: * clean up latency checking logic * move logic to .prepare and .shutdown to avoid functions that can sleep sound/soc/omap/mcbsp.c | 4 ++++ sound/soc/omap/mcbsp.h | 3 +++ sound/soc/omap/omap-mcbsp.c | 44 +++++++++++++++++++++++++++++++++++++++++++- 3 files changed, 50 insertions(+), 1 deletion(-) diff --git a/sound/soc/omap/mcbsp.c b/sound/soc/omap/mcbsp.c index 06fec5699cc8..e1c3d21dc5ed 100644 --- a/sound/soc/omap/mcbsp.c +++ b/sound/soc/omap/mcbsp.c @@ -25,6 +25,7 @@ #include <linux/io.h> #include <linux/slab.h> #include <linux/pm_runtime.h> +#include <linux/pm_qos.h> #include <linux/platform_data/asoc-ti-mcbsp.h> @@ -1098,6 +1099,9 @@ int omap_mcbsp_init(struct platform_device *pdev) void omap_mcbsp_cleanup(struct omap_mcbsp *mcbsp) { + if (pm_qos_request_active(&mcbsp->pm_qos_req)) + pm_qos_remove_request(&mcbsp->pm_qos_req); + if (mcbsp->pdata->buffer_size) sysfs_remove_group(&mcbsp->dev->kobj, &additional_attr_group); diff --git a/sound/soc/omap/mcbsp.h b/sound/soc/omap/mcbsp.h index 61e93b1c185d..46ae1269a698 100644 --- a/sound/soc/omap/mcbsp.h +++ b/sound/soc/omap/mcbsp.h @@ -323,8 +323,11 @@ struct omap_mcbsp { unsigned int fmt; unsigned int in_freq; + unsigned int latency[2]; int clk_div; int wlen; + + struct pm_qos_request pm_qos_req; }; void omap_mcbsp_config(struct omap_mcbsp *mcbsp, diff --git a/sound/soc/omap/omap-mcbsp.c b/sound/soc/omap/omap-mcbsp.c index d018e966e533..3814a234ef61 100644 --- a/sound/soc/omap/omap-mcbsp.c +++ b/sound/soc/omap/omap-mcbsp.c @@ -157,6 +157,16 @@ static void omap_mcbsp_dai_shutdown(struct snd_pcm_substream *substream, struct snd_soc_dai *cpu_dai) { struct omap_mcbsp *mcbsp = snd_soc_dai_get_drvdata(cpu_dai); + int tx = (substream->stream == SNDRV_PCM_STREAM_PLAYBACK); + int stream1 = tx ? SNDRV_PCM_STREAM_PLAYBACK : SNDRV_PCM_STREAM_CAPTURE; + int stream2 = tx ? SNDRV_PCM_STREAM_CAPTURE : SNDRV_PCM_STREAM_PLAYBACK; + + mcbsp->latency[stream1] = 0; + if (mcbsp->latency[stream2]) + pm_qos_update_request(&mcbsp->pm_qos_req, + mcbsp->latency[stream2]); + else + pm_qos_remove_request(&mcbsp->pm_qos_req); if (!cpu_dai->active) { omap_mcbsp_free(mcbsp); @@ -164,6 +174,28 @@ static void omap_mcbsp_dai_shutdown(struct snd_pcm_substream *substream, } } +static int omap_mcbsp_dai_prepare(struct snd_pcm_substream *substream, + struct snd_soc_dai *cpu_dai) +{ + struct omap_mcbsp *mcbsp = snd_soc_dai_get_drvdata(cpu_dai); + struct pm_qos_request *pm_qos_req = &mcbsp->pm_qos_req; + int tx = (substream->stream == SNDRV_PCM_STREAM_PLAYBACK); + int stream1 = tx ? SNDRV_PCM_STREAM_PLAYBACK : SNDRV_PCM_STREAM_CAPTURE; + int stream2 = tx ? SNDRV_PCM_STREAM_CAPTURE : SNDRV_PCM_STREAM_PLAYBACK; + int latency = mcbsp->latency[stream2]; + + /* Prevent omap hardware from hitting off between FIFO fills */ + if (!latency || mcbsp->latency[stream1] < latency) + latency = mcbsp->latency[stream1]; + + if (pm_qos_request_active(pm_qos_req)) + pm_qos_update_request(pm_qos_req, latency); + else if (latency) + pm_qos_add_request(pm_qos_req, PM_QOS_CPU_DMA_LATENCY, latency); + + return 0; +} + static int omap_mcbsp_dai_trigger(struct snd_pcm_substream *substream, int cmd, struct snd_soc_dai *cpu_dai) { @@ -226,6 +258,7 @@ static int omap_mcbsp_dai_hw_params(struct snd_pcm_substream *substream, int wlen, channels, wpf; int pkt_size = 0; unsigned int format, div, framesize, master; + unsigned int buffer_size = mcbsp->pdata->buffer_size; dma_data = snd_soc_dai_get_dma_data(cpu_dai, substream); channels = params_channels(params); @@ -240,7 +273,9 @@ static int omap_mcbsp_dai_hw_params(struct snd_pcm_substream *substream, default: return -EINVAL; } - if (mcbsp->pdata->buffer_size) { + if (buffer_size) { + int latency; + if (mcbsp->dma_op_mode == MCBSP_DMA_MODE_THRESHOLD) { int period_words, max_thrsh; int divider = 0; @@ -271,6 +306,12 @@ static int omap_mcbsp_dai_hw_params(struct snd_pcm_substream *substream, /* Use packet mode for non mono streams */ pkt_size = channels; } + + latency = ((((buffer_size - pkt_size) / channels) * 1000) + / (params->rate_num / params->rate_den)); + + mcbsp->latency[substream->stream] = latency; + omap_mcbsp_set_threshold(substream, pkt_size); } @@ -554,6 +595,7 @@ static int omap_mcbsp_dai_set_dai_sysclk(struct snd_soc_dai *cpu_dai, static const struct snd_soc_dai_ops mcbsp_dai_ops = { .startup = omap_mcbsp_dai_startup, .shutdown = omap_mcbsp_dai_shutdown, + .prepare = omap_mcbsp_dai_prepare, .trigger = omap_mcbsp_dai_trigger, .delay = omap_mcbsp_dai_delay, .hw_params = omap_mcbsp_dai_hw_params, -- 2.10.2

2 2

[alsa-devel] [PATCH 0/6] ASoC: platform cleanup
by Kuninori Morimoto 11 Jan '17

11 Jan '17

Hi Mark, Lars-Peter These removes un-used platform feature, and move last platform feature to component. If these are OK, all platform feature goes to component. Last platform piece is platform_list on soc-core.c Kuninori Morimoto (6): ASoC: remove .delay from snd_soc_platform_driver ASoC: remove .bespoke_trigger from snd_soc_platform_driver ASoC: remove snd_soc_platform_trigger() ASoC: add for_each_component{_safe} ASoC: add Component level struct snd_compr_ops ASoC: add Component level struct snd_pcm_ops include/sound/soc.h | 23 +++-- sound/soc/soc-compress.c | 228 ++++++++++++++++++++++++++++--------------- sound/soc/soc-core.c | 14 +-- sound/soc/soc-pcm.c | 244 +++++++++++++++++++++++++++++++++-------------- 4 files changed, 339 insertions(+), 170 deletions(-) -- 1.9.1

3 18

[alsa-devel] [PATCH v3] ASoC: ak4642: Replace mdelay function to msleep
by Kuninori Morimoto 11 Jan '17

11 Jan '17

From: Harunobu Kurokawa <harunobu.kurokawa.dn(a)renesas.com> Replace mdelay to msleep to avoid busy loop on ak4642_lout_event(). Otherwise, sometimes playback doesn't work correctly when pulseaudio was used. Signed-off-by: Harunobu Kurokawa <harunobu.kurokawa.dn(a)renesas.com> Signed-off-by: Kuninori Morimoto <kuninori.morimoto.gx(a)renesas.com> --- v2 -> v3 - based on latest kernel sound/soc/codecs/ak4642.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/ak4642.c b/sound/soc/codecs/ak4642.c index 2609f95..23ab964 100644 --- a/sound/soc/codecs/ak4642.c +++ b/sound/soc/codecs/ak4642.c @@ -189,7 +189,7 @@ static int ak4642_lout_event(struct snd_soc_dapm_widget *w, case SND_SOC_DAPM_POST_PMU: case SND_SOC_DAPM_POST_PMD: /* Power save mode OFF */ - mdelay(300); + msleep(300); snd_soc_update_bits(codec, SG_SL2, LOPS, 0); break; } -- 1.9.1

1 0

[alsa-devel] [PATCH] ASoC: mpc5200_psc_ac97: Remove unused DAI ID defines
by Lars-Peter Clausen 10 Jan '17

10 Jan '17

The DAI ID defines are back from the time when DAIs were referenced by a numerical ID. These days a string is used instead and the defines are unused. The last user of these defines was removed in commit f0fba2ad1b6b ("ASoC: multi-component - ASoC Multi-Component Support"). So remove the defines as well. This also means the mpc5200_psc_ac97.h file no longer has any content and can be removed. Signed-off-by: Lars-Peter Clausen <lars(a)metafoo.de> --- sound/soc/fsl/efika-audio-fabric.c | 1 - sound/soc/fsl/mpc5200_psc_ac97.c | 1 - sound/soc/fsl/mpc5200_psc_ac97.h | 13 ------------- 3 files changed, 15 deletions(-) delete mode 100644 sound/soc/fsl/mpc5200_psc_ac97.h diff --git a/sound/soc/fsl/efika-audio-fabric.c b/sound/soc/fsl/efika-audio-fabric.c index f200d1c..667f421 100644 --- a/sound/soc/fsl/efika-audio-fabric.c +++ b/sound/soc/fsl/efika-audio-fabric.c @@ -26,7 +26,6 @@ #include <sound/soc.h> #include "mpc5200_dma.h" -#include "mpc5200_psc_ac97.h" #define DRV_NAME "efika-audio-fabric" diff --git a/sound/soc/fsl/mpc5200_psc_ac97.c b/sound/soc/fsl/mpc5200_psc_ac97.c index 243700c..07ee355 100644 --- a/sound/soc/fsl/mpc5200_psc_ac97.c +++ b/sound/soc/fsl/mpc5200_psc_ac97.c @@ -25,7 +25,6 @@ #include <asm/mpc52xx_psc.h> #include "mpc5200_dma.h" -#include "mpc5200_psc_ac97.h" #define DRV_NAME "mpc5200-psc-ac97" diff --git a/sound/soc/fsl/mpc5200_psc_ac97.h b/sound/soc/fsl/mpc5200_psc_ac97.h deleted file mode 100644 index e881e78..0000000 --- a/sound/soc/fsl/mpc5200_psc_ac97.h +++ /dev/null @@ -1,13 +0,0 @@ -/* - * Freescale MPC5200 PSC in AC97 mode - * ALSA SoC Digital Audio Interface (DAI) driver - * - */ - -#ifndef __SOUND_SOC_FSL_MPC52xx_PSC_AC97_H__ -#define __SOUND_SOC_FSL_MPC52xx_PSC_AC97_H__ - -#define MPC5200_AC97_NORMAL 0 -#define MPC5200_AC97_SPDIF 1 - -#endif /* __SOUND_SOC_FSL_MPC52xx_PSC_AC97_H__ */ -- 2.1.4

1 0