On error, callers of regulator_bulk_get() pass the error up and don't
use the pointers in consumers[]. The function is documented to release
all regulators if any request fails.

Note: if an i-th regulator_get() failed only the i-th pointer was
cleared. This is another suggestion that the clearing was unnecessary.

Signed-off-by: Michał Mirosław <mirq-linux@rere.qmqm.pl>
---
 drivers/regulator/core.c | 7 +------
 1 file changed, 1 insertion(+), 6 deletions(-)

diff --git a/drivers/regulator/core.c b/drivers/regulator/core.c
index 3d7147fabbed..a1573a7ff2b2 100644
--- a/drivers/regulator/core.c
+++ b/drivers/regulator/core.c
@@ -4872,11 +4872,7 @@ static int _notifier_call_chain(struct regulator_dev *rdev,
 int _regulator_bulk_get(struct device *dev, int num_consumers,
 			struct regulator_bulk_data *consumers, enum regulator_get_type get_type)
 {
-	int i;
-	int ret;
-
-	for (i = 0; i < num_consumers; i++)
-		consumers[i].consumer = NULL;
+	int ret, i;
 
 	for (i = 0; i < num_consumers; i++) {
 		consumers[i].consumer = _regulator_get(dev,
@@ -4885,7 +4881,6 @@ int _regulator_bulk_get(struct device *dev, int num_consumers,
 			ret = dev_err_probe(dev, PTR_ERR(consumers[i].consumer),
 					    "Failed to get supply '%s'",
 					    consumers[i].supply);
-			consumers[i].consumer = NULL;
 			goto err;
 		}
 
-- 
2.39.2

On Mon, Apr 29, 2024 at 04:45:28PM +0200, Michał Mirosław wrote:
> On error, callers of regulator_bulk_get() pass the error up and don't
> use the pointers in consumers[]. The function is documented to release
> all regulators if any request fails.

This doesn't seem good from a robustness point of view and should be
nowhere near a fast path.

On Tue, Apr 30, 2024 at 01:01:28AM +0900, Mark Brown wrote:
> On Mon, Apr 29, 2024 at 04:45:28PM +0200, Michał Mirosław wrote:
> > On error, callers of regulator_bulk_get() pass the error up and don't
> > use the pointers in consumers[]. The function is documented to release
> > all regulators if any request fails.
> 
> This doesn't seem good from a robustness point of view and should be
> nowhere near a fast path.

I take it you'd prefer to have the function fixed to clear the other
pointers? I'll do that in the v3 then.

Best Regards
Michał Mirosław