Assuming the calibration state gets lost after system suspend. Resets cali_done when suspending. The rt1015_priv is alloced by kzalloc. No need to initialize cali_done to 0 in component probe callback. Signed-off-by: Tzung-Bi Shih <tzungbi@xxxxxxxxxx> --- The idea inspired from previous discussion: (https://mailman.alsa-project.org/pipermail/alsa-devel/2020-December/178229.html) The device could lose the calibration state after suspend/resume. It is better to reset the flag and get chance to calibrate again. The patch applies after the series: https://mailman.alsa-project.org/pipermail/alsa-devel/2020-December/178608.html sound/soc/codecs/rt1015.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/sound/soc/codecs/rt1015.c b/sound/soc/codecs/rt1015.c index 4a9e2eaebe30..2dcb7b0fba60 100644 --- a/sound/soc/codecs/rt1015.c +++ b/sound/soc/codecs/rt1015.c @@ -1055,7 +1055,6 @@ static int rt1015_probe(struct snd_soc_component *component) rt1015->component = component; rt1015->bclk_ratio = 0; - rt1015->cali_done = 0; INIT_DELAYED_WORK(&rt1015->flush_work, rt1015_flush_work); @@ -1103,6 +1102,7 @@ static int rt1015_suspend(struct snd_soc_component *component) regcache_cache_only(rt1015->regmap, true); regcache_mark_dirty(rt1015->regmap); + rt1015->cali_done = 0; return 0; } -- 2.29.2.729.g45daf8777d-goog