Re: [PATCH v2] nand: Add NAND driver for Mikrotik RB4xx series boards

From: Frans Klaver
Date: Fri May 08 2015 - 06:14:45 EST


On Thu, May 7, 2015 at 2:31 PM, Bert Vermeulen <bert@xxxxxxxx> wrote:

> +static int rb4xx_nand_probe(struct platform_device *pdev)
> +{
> + struct device *spi_dev;
> + struct rb4xx_nand_info *info;
> + struct rb4xx_cpld_platform_data *pdata;
> + int ret;
> +
> + /*
> + * NAND chip is behind the CPLD on SPI bus 0 CS 1. Allow this
> + * to silently fail the second time around.
> + */
> + spi_register_driver(&rb4xx_cpld_driver);
> +
> + /*
> + * Getting a handle on the SPI device above will fail the first time,
> + * but should succeed whenever that driver gets probed.
> + */
> + spi_dev = bus_find_device_by_name(&spi_bus_type, NULL, "spi0.1");
> + if (!spi_dev)
> + return -EPROBE_DEFER;
> +
> + pdata = dev_get_platdata(spi_dev);
> + if (!pdata)
> + return -ENODATA;
> +
> + info = devm_kzalloc(&pdev->dev, sizeof(*info), GFP_KERNEL);
> + if (!info)
> + return -ENOMEM;
> +
> + info->gpio_ale = devm_gpiod_get(&pdev->dev, "NAND ALE", GPIOD_OUT_LOW);
> + if (IS_ERR(info->gpio_ale)) {
> + dev_err(&pdev->dev, "unable to request gpio NAND ALE: %ld\n",
> + PTR_ERR(info->gpio_ale));
> + return -ENOENT;
> + }
> +
> + info->gpio_cle = devm_gpiod_get(&pdev->dev, "NAND CLE", GPIOD_OUT_LOW);
> + if (IS_ERR(info->gpio_cle)) {
> + dev_err(&pdev->dev, "unable to request gpio NAND CLE: %ld\n",
> + PTR_ERR(info->gpio_cle));
> + return -ENOENT;
> + }
> +
> + info->gpio_nce = devm_gpiod_get(&pdev->dev, "NAND NCE", GPIOD_OUT_LOW);
> + if (IS_ERR(info->gpio_nce)) {
> + dev_err(&pdev->dev, "unable to request gpio NAND NCE: %ld\n",
> + PTR_ERR(info->gpio_nce));
> + return -ENOENT;
> + }
> +
> + /* Platform GPIO, has no device to manage it. */
> + info->gpio_rdy = gpiod_get(NULL, "NAND RDY", GPIOD_IN);
> + if (IS_ERR(info->gpio_rdy)) {
> + dev_err(&pdev->dev, "unable to request gpio RDY: %ld\n",
> + PTR_ERR(info->gpio_rdy));
> + return -ENOENT;
> + }
> +
> + info->spi_dev = container_of(spi_dev, struct spi_device, dev);
> +
> + info->chip.priv = info;
> + info->mtd.priv = &info->chip;
> + info->mtd.owner = THIS_MODULE;

If you should respin, could you add

info->mtd.dev.parent = &pdev->dev;

This will have a symlink to the parent device show up in sysfs. Seems
some people feel rather strongly about that.

Thanks,
Frans
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/