Hi Ben,

On 3 August 2011 14:42, Ben Dooks [off-list ref] wrote:

On Wed, Aug 03, 2011 at 12:08:27AM +0100, Thomas Abraham wrote:

quoted

For device tree based probe, the dependecy on pdev->id to attach a
corresponding default port info to the driver's private data is
removed. The fifosize parameter is obtained from the device tree
node and the next available instance of port info is updated
with the fifosize value and attached to the driver's private data.
The default platform data is selected based on the compatible property.

CC: Ben Dooks <ben-linux@fluff.org>
Signed-off-by: Thomas Abraham <redacted>
---
 .../devicetree/bindings/serial/samsung_uart.txt    |   16 +++++++
 drivers/tty/serial/s5pv210.c                       |   43 +++++++++++++++++++-
 drivers/tty/serial/samsung.c                       |    5 ++-
 3 files changed, 62 insertions(+), 2 deletions(-)
 create mode 100644 Documentation/devicetree/bindings/serial/samsung_uart.txt

[...]

quoted

 /* device management */
 static int s5p_serial_probe(struct platform_device *pdev)
 {
-     return s3c24xx_serial_probe(pdev, s5p_uart_inf[pdev->id]);
+     static unsigned int probe_index;
+     unsigned int port = pdev->id;
+     const struct of_device_id *match;
+     struct s3c2410_uartcfg *cfg;
+
+     if (pdev->dev.of_node) {
+             if (of_property_read_u32(pdev->dev.of_node,
+                             "samsung,uart-fifosize",
+                             &s5p_uart_inf[probe_index]->fifosize))
+                     return -EINVAL;

I'd rather see the fifo size either being a property of the soc itself
or being inferred by the compatible field.

When using the compatible field to infer the fifosize of the
controller, the code looks as listed below.

struct s3c24xx_uart_dt_compat_data {
        unsigned int fifosize;
        struct s3c2410_uartcfg *uartcfg;
};

static struct s3c2410_uartcfg s5pv310_uart_defcfg = {
        .ucon   = 0x3c5,
        .ufcon  = 0x111,
        .flags  = NO_NEED_CHECK_CLKSRC,
        .has_fracval = 1,
};

static struct s3c24xx_uart_dt_compat_data s5pv210_compat_fs256 = {
        .fifosize = 256;
        .uartcfg = &s5pv310_uart_defcfg;
};

static struct s3c24xx_uart_dt_compat_data s5pv210_compat_fs64 = {
        .fifosize = 64;
        .uartcfg = &s5pv310_uart_defcfg;
};

static struct s3c24xx_uart_dt_compat_data s5pv210_compat_fs16 = {
        .fifosize = 16;
        .uartcfg = &s5pv310_uart_defcfg;
};

static const struct of_device_id s5pv210_uart_dt_match[] = {
        { .compatible = "samsung,s5pv310-uart-fs256", .data =
&s5pv210_compat_fs256 },
        { .compatible = "samsung,s5pv310-uart-fs64", .data =
s5pv210_compat_fs64 },
        { .compatible = "samsung,s5pv310-uart-fs16", .data =
s5pv210_compat_fs16 },
        {},
};
MODULE_DEVICE_TABLE(of, s5pv210_uart_match);


This requires a new structure definition and additional data in the
driver. So I still prefer to use a property in the uart device node to
define the fifosize of the controller.

What would you be your preference? Or is there a better way to obtain
the fifosize?

Thanks,
Thomas.

quoted

      .driver         = {
              .name   = "s5pv210-uart",
              .owner  = THIS_MODULE,
+             .of_match_table = s5pv210_uart_dt_match,

I think maybe doing something like

               .of_match_table = of_match_ptr(5pv210_uart_dt_match),

so we can avoid having the #else and #define 5pv210_uart_dt_match NULL
in a number of places.

quoted

      },
 };

--
Ben Dooks, ben@fluff.org, http://www.fluff.org/ben/

Large Hadron Colada: A large Pina Colada that makes the universe disappear.