From: Jon Hunter Date: Thu, 21 Apr 2016 16:11:57 +0000 (+0100) Subject: regulator: core: Don't terminate supply resolution early X-Git-Url: http://git.cdn.openwrt.org/?a=commitdiff_plain;h=7ddede6a58a0bd26efcfd2a5055611195411f514;p=openwrt%2Fstaging%2Fblogic.git regulator: core: Don't terminate supply resolution early The function regulator_register_resolve_supply() is called from the context of class_for_each_dev() (during the regulator registration) to resolve any supplies added. regulator_register_resolve_supply() will return an error if a regulator's supply cannot be resolved and this will terminate the loop in class_for_each_dev(). This means that we will not attempt to resolve any other supplies after one has failed. Hence, this may delay the resolution of other regulator supplies until the failing one itself can be resolved. Rather than terminating the loop early, don't return an error code and keep attempting to resolve any other supplies for regulators that have been registered. Signed-off-by: Jon Hunter Signed-off-by: Mark Brown --- diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c index fd0e4e37f4e1..9922922ce6bd 100644 --- a/drivers/regulator/core.c +++ b/drivers/regulator/core.c @@ -3842,7 +3842,12 @@ static void rdev_init_debugfs(struct regulator_dev *rdev) static int regulator_register_resolve_supply(struct device *dev, void *data) { - return regulator_resolve_supply(dev_to_rdev(dev)); + struct regulator_dev *rdev = dev_to_rdev(dev); + + if (regulator_resolve_supply(rdev)) + rdev_dbg(rdev, "unable to resolve supply\n"); + + return 0; } /**