Re: [PATCH 3/3] regulator: twl: Regulator mode should not depend on regulator enabled state

From: Ivaylo Dimitrov
Date: Sun Apr 24 2016 - 13:01:45 EST


Hi,

On 24.04.2016 19:11, Pavel Machek wrote:
On Sat 2016-03-26 10:28:15, Ivaylo Dimitrov wrote:
When machine constraints are applied, regulator framework first sets
initial mode (if any) and then enables the regulator if needed. The current
code in twl4030reg_set_mode always checks if the regulator is enabled
before applying the mode. That results in -EACCES error returned for
"always-on" regulators which have "initial-mode" set in the board DTS. Fix
that by removing the unneeded check.

Should we keep the check for the regulators that are _not_ always-on?


I don't think so, as there is no other way to set mode to a regulator but through regulator framework and board DTS (for twl4030 that is) - and this is done only once during boot, see https://lkml.org/lkml/2016/4/17/78 as well. So, no matter if the regulator is always-on or not, that check is redundant IMO.

Ivo




Signed-off-by: Ivaylo Dimitrov <ivo.g.dimitrov.75@xxxxxxxxx>
---
drivers/regulator/twl-regulator.c | 8 --------
1 file changed, 8 deletions(-)

diff --git a/drivers/regulator/twl-regulator.c b/drivers/regulator/twl-regulator.c
index be8d05e..d8f6ad6 100644
--- a/drivers/regulator/twl-regulator.c
+++ b/drivers/regulator/twl-regulator.c
@@ -371,7 +371,6 @@ static int twl4030reg_set_mode(struct regulator_dev *rdev, unsigned mode)
{
struct twlreg_info *info = rdev_get_drvdata(rdev);
unsigned message;
- int status;

/* We can only set the mode through state machine commands... */
switch (mode) {
@@ -385,13 +384,6 @@ static int twl4030reg_set_mode(struct regulator_dev *rdev, unsigned mode)
return -EINVAL;
}

- /* Ensure the resource is associated with some group */
- status = twlreg_grp(rdev);
- if (status < 0)
- return status;
- if (!(status & (P3_GRP_4030 | P2_GRP_4030 | P1_GRP_4030)))
- return -EACCES;
-
return twl4030_send_pb_msg(message);
}