Re: [PATCH v1 2/2] mfd: syscon: Switch to use devm_ioremap_resource()
From: Arnd Bergmann
Date: Fri Nov 15 2019 - 04:13:17 EST
On Fri, Nov 15, 2019 at 9:49 AM Andy Shevchenko
<andriy.shevchenko@xxxxxxxxxxxxxxx> wrote:
>
> Instead of checking resource pointer for being NULL and
> report some not very standard error codes in this case,
> switch to devm_ioremap_resource() API.
>
> Signed-off-by: Andy Shevchenko <andriy.shevchenko@xxxxxxxxxxxxxxx>
IIRC there are some slightly odd uses of syscon that rely on on us not calling
devm_request_mem_region here, which is implied by devm_ioremap_resource()
but not devm_ioremap().
A patch to add a comment about this might be helpful though.
Arnd
> ---
> drivers/mfd/syscon.c | 9 +++------
> 1 file changed, 3 insertions(+), 6 deletions(-)
>
> diff --git a/drivers/mfd/syscon.c b/drivers/mfd/syscon.c
> index 13626bb2d432..fad961b2e4a5 100644
> --- a/drivers/mfd/syscon.c
> +++ b/drivers/mfd/syscon.c
> @@ -238,12 +238,9 @@ static int syscon_probe(struct platform_device *pdev)
> return -ENOMEM;
>
> res = platform_get_resource(pdev, IORESOURCE_MEM, 0);
> - if (!res)
> - return -ENOENT;
> -
> - base = devm_ioremap(dev, res->start, resource_size(res));
> - if (!base)
> - return -ENOMEM;
> + base = devm_ioremap_resource(dev, res);
> + if (IS_ERR(base))
> + return PTR_ERR(base);
>
> syscon_config.max_register = resource_size(res) - 4;
> if (pdata)
> --
> 2.24.0
>