
Hi Hans,
On 18 August 2015 at 03:29, Hans de Goede hdegoede@redhat.com wrote:
Hi,
On 18-08-15 00:14, Simon Glass wrote:
Hi Hans,
On 6 August 2015 at 12:13, Hans de Goede hdegoede@redhat.com wrote:
sun6i and later have a couple of io-blocks which are shared between the main CPU core and the "R" cpu which is small embedded cpu which can be active while the main system is suspended.
These gpio banks sit at a different mmio address then the normal banks, and have a separate devicetree node and compatible, this adds support for these banks to the sunxi-gpio code when built with device-model support.
Signed-off-by: Hans de Goede hdegoede@redhat.com
drivers/gpio/sunxi_gpio.c | 22 ++++++++++++++++++---- 1 file changed, 18 insertions(+), 4 deletions(-)
diff --git a/drivers/gpio/sunxi_gpio.c b/drivers/gpio/sunxi_gpio.c index afa165a..57b78e5 100644 --- a/drivers/gpio/sunxi_gpio.c +++ b/drivers/gpio/sunxi_gpio.c @@ -266,16 +266,28 @@ static int gpio_sunxi_bind(struct udevice *parent) { struct sunxi_gpio_platdata *plat = parent->platdata; struct sunxi_gpio_reg *ctlr;
int bank;
int ret;
int bank, no_banks, ret, start; /* If this is a child device, there is nothing to do here */ if (plat) return 0;
if (fdt_node_check_compatible(gd->fdt_blob, parent->of_offset,
"allwinner,sun6i-a31-r-pinctrl") == 0) {
If I understand what you are doing here, the correct way is to use a .data field in sunxi_gpio_ids[] below.
Yes that would work too, but that indirection makes it harder to follow what the code is doing without giving us much benefit IMHO.
In any case you should not be checking the compatible string of a parent.
AFAICT this is the right way to check a compatible string in a bind function, the parent variable name here is confusing as this actually points to the udevice passed into the bind function, not its parent. I guess it is named parent rather then device to distinguish it from the per bank gpio udevices the bind function is instantiating.
Sounds fine.
So I'm not really checking the parents compatible but its own, it just looks that way because of the bind function parameter name being parent.
See for example, drivers/i2c/tegra_i2c.c:
static const struct udevice_id tegra_i2c_ids[] = { { .compatible = "nvidia,tegra114-i2c", .data = TYPE_114 }, { .compatible = "nvidia,tegra20-i2c", .data = TYPE_STD }, { .compatible = "nvidia,tegra20-i2c-dvc", .data = TYPE_DVC }, { } };
...
i2c_bus->type = dev_get_driver_data(dev);
which gets the 'data' value. So it's pretty easy to use and is the standard way to make a driver support multiple hardware types.
Regards,
Hans
start = 'L' - 'A';
no_banks = 2; /* L & M */
} else if (fdt_node_check_compatible(gd->fdt_blob,
parent->of_offset,
"allwinner,sun8i-a23-r-pinctrl") == 0) {
start = 'L' - 'A';
no_banks = 1; /* L only */
} else {
start = 0;
no_banks = SUNXI_GPIO_BANKS;
}
ctlr = (struct sunxi_gpio_reg *)fdtdec_get_addr(gd->fdt_blob, parent->of_offset,
"reg");
for (bank = 0; bank < SUNXI_GPIO_BANKS; bank++) {
for (bank = 0; bank < no_banks; bank++) { struct sunxi_gpio_platdata *plat; struct udevice *dev;
@@ -283,7 +295,7 @@ static int gpio_sunxi_bind(struct udevice *parent) if (!plat) return -ENOMEM; plat->regs = &ctlr->gpio_bank[bank];
plat->bank_name = gpio_bank_name(bank);
plat->bank_name = gpio_bank_name(start + bank); plat->gpio_count = SUNXI_GPIOS_PER_BANK; ret = device_bind(parent, parent->driver,
@@ -306,6 +318,8 @@ static const struct udevice_id sunxi_gpio_ids[] = { { .compatible = "allwinner,sun8i-a23-pinctrl" }, { .compatible = "allwinner,sun8i-a33-pinctrl" }, { .compatible = "allwinner,sun9i-a80-pinctrl" },
{ .compatible = "allwinner,sun6i-a31-r-pinctrl" },
};{ .compatible = "allwinner,sun8i-a23-r-pinctrl" }, { }
-- 2.4.3
U-Boot mailing list U-Boot@lists.denx.de http://lists.denx.de/mailman/listinfo/u-boot
Regards, Simon