On 30/07/2019 13:38, Charles Keepax wrote: > On Tue, Jul 30, 2019 at 01:09:36PM +0100, Thomas Preston wrote: >> Add an I2C based codec driver for ST TDA7802 amplifier. The amplifier >> supports 4 audio channels but can support up to 16 with multiple >> devices. >> >> Signed-off-by: Thomas Preston <thomas.preston@xxxxxxxxxxxxxxx> >> Cc: Patrick Glaser <pglaser@xxxxxxxxx> >> Cc: Rob Duncan <rduncan@xxxxxxxxx> >> Cc: Nate Case <ncase@xxxxxxxxx> >> --- >> Changes since v1: >> - Use ALSA kcontrol interface to expose device controls to userland >> - Gain >> - Channel diagnostic mode >> - Impedance efficiency optimiser. I decided against setting this >> as a DT property since it seems like something that can be >> changed on the fly. >> - Add regmap default values >> - Channel unmute by default is added in a downstream patch. >> - I'm not sure if I should keep this since they're all zero, >> although there are other drivers will all-zero reg_defaults. >> - I believe the "//" style is used for SPDX headers in normal C source files. >> https://lwn.net/Articles/739183/ >> - Drop the "enable" sysfs device attribute. >> - Don't set TDM format using magic numbers. >> - Set sample rate using hw_params. >> - Remove unecessary defines. >> - Use DAPM to handle AMP_ON. >> - Cosmetic fixups >> >> sound/soc/codecs/Kconfig | 6 + >> sound/soc/codecs/Makefile | 2 + >> sound/soc/codecs/tda7802.c | 509 +++++++++++++++++++++++++++++++++++++ >> 3 files changed, 517 insertions(+) >> create mode 100644 sound/soc/codecs/tda7802.c >> >> +++ b/sound/soc/codecs/tda7802.c >> @@ -0,0 +1,509 @@ >> +// SPDX-License-Identifier: GPL-2.0 >> +/* >> + * tda7802.c -- codec driver for ST TDA7802 >> + * >> + * Copyright (C) 2016-2019 Tesla Motors, Inc. >> + */ > > Better to make the whole comment // see something like > sound/soc/codecs/cs47l35.c for an example. > I will update to "//" style. Is this a new standard? There aren't many comments like that in 4.14 (my target kernel) - I can see a lot more in 5.3. My intention was: 1. Apply the SPDX rules to SPDX bit. Documentation/process/license-rules.rst 2. Use multi-line comments for the rest. Documentation/process/coding-style.rst >> +static int tda7802_set_bias_level(struct snd_soc_component *component, >> + enum snd_soc_bias_level level) >> +{ >> + const struct tda7802_priv *tda7802 = >> + snd_soc_component_get_drvdata(component); >> + struct snd_soc_dapm_context *dapm_context = >> + snd_soc_component_get_dapm(component); >> + const enum snd_soc_bias_level oldlevel = >> + snd_soc_dapm_get_bias_level(dapm_context); >> + int err = 0; >> + >> + dev_dbg(component->dev, "%s level %d\n", __func__, level); >> + >> + switch (level) { >> + case SND_SOC_BIAS_ON: >> + break; >> + case SND_SOC_BIAS_PREPARE: >> + break; >> + case SND_SOC_BIAS_STANDBY: >> + err = regulator_enable(tda7802->enable_reg); >> + if (err < 0) { >> + dev_err(component->dev, "Could not enable.\n"); >> + return err; >> + } >> + dev_dbg(component->dev, "Regulator enabled\n"); >> + msleep(ENABLE_DELAY_MS); >> + >> + if (oldlevel == SND_SOC_BIAS_OFF) { >> + dev_dbg(component->dev, "Syncing regcache\n"); >> + err = regcache_sync(component->regmap); >> + if (err < 0) >> + dev_err(component->dev, >> + "Could not sync regcache, %d\n", err); > > If your doing a regcache_sync I would probably have expected to > see calls to regcache_cache_only. > > If the device needs syncing that implies the hardware registers > have lost state, so there is little point in writing to them > if they are unavailable/about to loose their state. > Ah, from the comments I thought I only needed to call regcache_mark_dirty... >> + } >> + break; >> + case SND_SOC_BIAS_OFF: >> + regcache_mark_dirty(component->regmap); >> + err = regulator_disable(tda7802->enable_reg); >> + if (err < 0) >> + dev_err(component->dev, "Could not disable.\n"); >> + break; >> + } >> + >> + return err; >> +} So I think the correct order is: device_off: regcache_cache_only power-off (enable) regcache_mark_dirty device_on: power-on (enable) regcache_sync I will double-check the register state is actually lost too. Fiddling with the cache might be completely unnecessary. Many thanks
