On 04/03/2021 1:49 pm, Takashi Iwai wrote:

On Wed, 03 Mar 2021 19:29:59 +0100, Vitaly Rodionov wrote:

...

+static int cs8409_cs42l42_volume_get(struct snd_kcontrol *kcontrol, + struct snd_ctl_elem_value *ucontrol) +{ + struct hda_codec *codec = snd_kcontrol_chip(kcontrol); + hda_nid_t nid = get_amp_nid(kcontrol); + int chs = get_amp_channels(kcontrol); + long *valp = ucontrol->value.integer.value; + char vol = 0; + + codec_dbg(codec, "%s() nid: %d\n", __func__, nid); + snd_hda_power_up(codec); + switch (nid) { + case CS8409_CS42L42_HP_PIN_NID: + mutex_lock(&cs8409_i2c_mux); + if (chs & 1) { + vol = -(cs8409_i2c_read(codec, CS42L42_I2C_ADDR, + CS8409_CS42L42_REG_HS_VOLUME_CHA, 1)); Better to cache the values instead of i2c read at each time? Then the unnecessary power up/down sequence can be avoided, too. Yes, agree. Will be fixed in next version of patch.

thanks,

Takashi

Thank you, Vitaly