| Message ID | 20160828173446.32728-2-contact@paulk.fr |
|---|---|
| State | Deferred |
| Headers | show |
Hi, On Sun, Aug 28, 2016 at 07:34:46PM +0200, Paul Kocialkowski wrote: > When the charger is missing (disconnected), it is safe to assume that > the charger chip is no charging. > > This is especially relevant when a status GPIO is present and the > charger is getting disconnected. bq24735_charger_is_charging will be > triggered due to the interrupt then, it will attempt to read whether it > is charging through i2c, which will fail as the charger is disconnected. > > This also fixes that specific issue. Thanks, I queued this one into my for-next branch. -- Sebastian
diff --git a/drivers/power/bq24735-charger.c b/drivers/power/bq24735-charger.c index 361a047..0b5b247 100644 --- a/drivers/power/bq24735-charger.c +++ b/drivers/power/bq24735-charger.c @@ -201,8 +201,12 @@ static bool bq24735_charger_is_present(struct bq24735 *charger) static int bq24735_charger_is_charging(struct bq24735 *charger) { - int ret = bq24735_read_word(charger->client, BQ24735_CHG_OPT); + int ret; + + if (!bq24735_charger_is_present(charger)) + return 0; + ret = bq24735_read_word(charger->client, BQ24735_CHG_OPT); if (ret < 0) return ret;
When the charger is missing (disconnected), it is safe to assume that the charger chip is no charging. This is especially relevant when a status GPIO is present and the charger is getting disconnected. bq24735_charger_is_charging will be triggered due to the interrupt then, it will attempt to read whether it is charging through i2c, which will fail as the charger is disconnected. This also fixes that specific issue. Signed-off-by: Paul Kocialkowski <contact@paulk.fr> --- drivers/power/bq24735-charger.c | 6 +++++- 1 file changed, 5 insertions(+), 1 deletion(-)