Thread (9 messages) 9 messages, 2 authors, 2014-09-23

Re: [PATCH v2 2/3] power: gpio-charger: add device tree support

From: Doug Anderson <dianders@chromium.org>
Date: 2014-09-22 16:48:38
Also in: linux-pm, lkml

Heiko

On Sun, Sep 21, 2014 at 1:05 PM, Heiko Stuebner [off-list ref] wrote:
quoted hunk ↗ jump to hunk
From: Heiko Stuebner <redacted>

Add the ability to parse gpio-charger data from a devicetree node.

Signed-off-by: Heiko Stuebner <redacted>
---
 drivers/power/gpio-charger.c | 66 ++++++++++++++++++++++++++++++++++++++++++--
 1 file changed, 63 insertions(+), 3 deletions(-)
diff --git a/drivers/power/gpio-charger.c b/drivers/power/gpio-charger.c
index a0024b2..5fe6879 100644
--- a/drivers/power/gpio-charger.c
+++ b/drivers/power/gpio-charger.c
@@ -22,6 +22,8 @@
 #include <linux/platform_device.h>
 #include <linux/power_supply.h>
 #include <linux/slab.h>
+#include <linux/of.h>
+#include <linux/of_gpio.h>

 #include <linux/power/gpio-charger.h>
@@ -69,17 +71,68 @@ static enum power_supply_property gpio_charger_properties[] = {
        POWER_SUPPLY_PROP_ONLINE,
 };

+static
+struct gpio_charger_platform_data *gpio_charger_parse_dt(struct device *dev)
+{
+       struct device_node *np = dev->of_node;
+       struct gpio_charger_platform_data *pdata;
+       const char *chargetype;
+       enum of_gpio_flags flags;
+       int ret;
+
+       if (!np)
+               return ERR_PTR(-ENOENT);
+
+       pdata = devm_kzalloc(dev, sizeof(struct gpio_charger_platform_data),
nit: sizeof(*pdata) is smaller / cleaner.
+                            GFP_KERNEL);
+       if (!pdata)
+               return ERR_PTR(-ENOMEM);
+
+       pdata->name = np->name;
+
+       pdata->gpio = of_get_gpio_flags(np, 0, &flags);
It would be nice to check the gpio for errors here, including handling
the possibility of EPROBE_DEFER.

+       pdata->gpio_active_low = flags & OF_GPIO_ACTIVE_LOW;
+
+       pdata->type = POWER_SUPPLY_TYPE_UNKNOWN;
+       ret = of_property_read_string(np, "charger-type", &chargetype);
+       if (ret >= 0) {
+               if (!strncmp("unknown", chargetype, 7))
+                       pdata->type = POWER_SUPPLY_TYPE_UNKNOWN;
+               else if (!strncmp("battery", chargetype, 7))
+                       pdata->type = POWER_SUPPLY_TYPE_BATTERY;
+               else if (!strncmp("ups", chargetype, 3))
+                       pdata->type = POWER_SUPPLY_TYPE_UPS;
+               else if (!strncmp("mains", chargetype, 5))
+                       pdata->type = POWER_SUPPLY_TYPE_MAINS;
+               else if (!strncmp("usb-sdp", chargetype, 7))
+                       pdata->type = POWER_SUPPLY_TYPE_USB;
+               else if (!strncmp("usb-dcp", chargetype, 7))
+                       pdata->type = POWER_SUPPLY_TYPE_USB_DCP;
+               else if (!strncmp("usb-cdp", chargetype, 7))
+                       pdata->type = POWER_SUPPLY_TYPE_USB_CDP;
+               else if (!strncmp("usb-aca", chargetype, 7))
+                       pdata->type = POWER_SUPPLY_TYPE_USB_ACA;
+               else
+                       dev_warn(dev, "unknown charger type %s\n", chargetype);
+       }
+
+       return pdata;
+}
+
 static int gpio_charger_probe(struct platform_device *pdev)
 {
-       const struct gpio_charger_platform_data *pdata = pdev->dev.platform_data;
+       struct gpio_charger_platform_data *pdata = pdev->dev.platform_data;
Do you really need to remove const?  If I remember correctly the const
here is not saying that the variable "pdata" won't change but that the
data pointed to by pdata won't change.  I think that's still the case
(from the perspective of this function).

        struct gpio_charger *gpio_charger;
        struct power_supply *charger;
        int ret;
        int irq;

        if (!pdata) {
-               dev_err(&pdev->dev, "No platform data\n");
-               return -EINVAL;
+               pdata = gpio_charger_parse_dt(&pdev->dev);
+               if (IS_ERR(pdata)) {
+                       dev_err(&pdev->dev, "No platform data\n");
+                       return -EINVAL;
Probably should pass back the error code since it might be
EPROBE_DEFER.  ...and <sigh> I guess you need the special case of
"don't print if it's EPROBE_DEFER".  I almost wish we had a helper
function for that.  ;)

quoted hunk ↗ jump to hunk
+               }
        }

        if (!gpio_is_valid(pdata->gpio)) {
@@ -189,6 +242,12 @@ static int gpio_charger_resume(struct device *dev)
 static SIMPLE_DEV_PM_OPS(gpio_charger_pm_ops,
                gpio_charger_suspend, gpio_charger_resume);

+static const struct of_device_id gpio_charger_match[] = {
+       { .compatible = "gpio-charger" },
+       { }
+};
+MODULE_DEVICE_TABLE(of, gpio_charger_match);
+
 static struct platform_driver gpio_charger_driver = {
        .probe = gpio_charger_probe,
        .remove = gpio_charger_remove,
@@ -196,6 +255,7 @@ static struct platform_driver gpio_charger_driver = {
                .name = "gpio-charger",
                .owner = THIS_MODULE,
                .pm = &gpio_charger_pm_ops,
+               .of_match_table = of_match_ptr(gpio_charger_match),
Given that you don't have any #ifdefs with "CONFIG_OF", I think
gpio_charger_match will always exist.  It seems like you should remove
the of_match_ptr or add some #ifdefs.  I can't quite keep up with what
the currently suggested best practice is here, though.

-Doug
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help