Store a pointer to the pdata so that if it needs to be used after
probe it can be. In the case of booting from DeviceTree, the
dev.platform_data may be null, so we need to store this pointer
explicitly.
Based on work by Joseph Lo.
Cc: Joseph Lo <[email protected]>
Signed-off-by: Rhyland Klein <[email protected]>
---
v2:
- Renamed patch from "regulator: palmas: fix pdata ptr not be updated
after it has been allocated" to be appropriate to the current change.
- Instead of updating dev.platform_data, store the pdata pointer in the
pmic struct which we already have access to via drvdata.
drivers/regulator/palmas-regulator.c | 2 +-
include/linux/mfd/palmas.h | 2 ++
2 files changed, 3 insertions(+), 1 deletion(-)
diff --git a/drivers/regulator/palmas-regulator.c b/drivers/regulator/palmas-regulator.c
index d0c8785..750ce55 100644
--- a/drivers/regulator/palmas-regulator.c
+++ b/drivers/regulator/palmas-regulator.c
@@ -761,7 +761,6 @@ static void palmas_dt_to_pdata(struct device *dev,
pdata->ldo6_vibrator = of_property_read_bool(node, "ti,ldo6-vibrator");
}
-
static int palmas_regulators_probe(struct platform_device *pdev)
{
struct palmas *palmas = dev_get_drvdata(pdev->dev.parent);
@@ -789,6 +788,7 @@ static int palmas_regulators_probe(struct platform_device *pdev)
pmic->dev = &pdev->dev;
pmic->palmas = palmas;
+ pmic->pdata = pdata;
palmas->pmic = pmic;
platform_set_drvdata(pdev, pmic);
diff --git a/include/linux/mfd/palmas.h b/include/linux/mfd/palmas.h
index 1a8dd7a..1703f18 100644
--- a/include/linux/mfd/palmas.h
+++ b/include/linux/mfd/palmas.h
@@ -352,6 +352,8 @@ struct palmas_pmic {
struct regulator_dev *rdev[PALMAS_NUM_REGS];
struct mutex mutex;
+ struct palmas_pmic_platform_data *pdata;
+
int smps123;
int smps457;
--
1.7.9.5
On Thu, Jul 25, 2013 at 01:04:59PM -0400, Rhyland Klein wrote:
> Store a pointer to the pdata so that if it needs to be used after
> probe it can be. In the case of booting from DeviceTree, the
> dev.platform_data may be null, so we need to store this pointer
> explicitly.
This doesn't seem to adjust any places where pdata is used to refer to
the new pointer so does it actually have any effect?
On 7/25/2013 2:02 PM, Mark Brown wrote:
> * PGP Signed by an unknown key
>
> On Thu, Jul 25, 2013 at 01:04:59PM -0400, Rhyland Klein wrote:
>> Store a pointer to the pdata so that if it needs to be used after
>> probe it can be. In the case of booting from DeviceTree, the
>> dev.platform_data may be null, so we need to store this pointer
>> explicitly.
>
> This doesn't seem to adjust any places where pdata is used to refer to
> the new pointer so does it actually have any effect?
I was thinking the same thing after I sent this. Right now, no this has
no effect. The expectation is that when ldo8-tracking support is posted,
it needs to have access to the pdata. I suppose it could make sense then
to either make this part of that series, or to flatten this directly
into it.
I think it can be fine to merge on its own, but it might make sense then
to include the ldo8 tracking support. Let me see how ready that is to be
posted, and if it is close, this can join it in a patchset.
-rhyland
>
> * Unknown Key
> * 0x7EA229BD
>
--
nvpublic