Re: [PATCH 7/7] usb: musb: convert to devm_platform_ioremap_resource_byname

[Chunfeng Yun <chunfeng.yun@mediatek.com>]

On Thu, 2020-11-12 at 15:08 +0100, Geert Uytterhoeven wrote:
> Hi Chunfeng,
> 
> On Wed, Jul 22, 2020 at 9:54 AM Chunfeng Yun <chunfeng.yun@mediatek.com> wrote:
> > Use devm_platform_ioremap_resource_byname() to simplify code
> >
> > Cc: Bin Liu <b-liu@ti.com>
> > Signed-off-by: Chunfeng Yun <chunfeng.yun@mediatek.com>
> 
> Thanks for your patch, which is now commit 2d30e408a2a6b344 ("usb: musb:
> convert to devm_platform_ioremap_resource_byname") in v5.9.
> 
> > --- a/drivers/usb/musb/musb_dsps.c
> > +++ b/drivers/usb/musb/musb_dsps.c
> > @@ -429,12 +429,10 @@ static int dsps_musb_init(struct musb *musb)
> >         struct platform_device *parent = to_platform_device(dev->parent);
> >         const struct dsps_musb_wrapper *wrp = glue->wrp;
> >         void __iomem *reg_base;
> > -       struct resource *r;
> >         u32 rev, val;
> >         int ret;
> >
> > -       r = platform_get_resource_byname(parent, IORESOURCE_MEM, "control");
> > -       reg_base = devm_ioremap_resource(dev, r);
> > +       reg_base = devm_platform_ioremap_resource_byname(parent, "control");
> >         if (IS_ERR(reg_base))
> >                 return PTR_ERR(reg_base);
> >         musb->ctrl_base = reg_base;
> 
> On Beaglebone Black, where each interface has 2 children:
> 
>     musb-dsps 47401c00.usb: can't request region for resource [mem
> 0x47401800-0x474019ff]
>     musb-hdrc musb-hdrc.1: musb_init_controller failed with status -16
>     musb-hdrc: probe of musb-hdrc.1 failed with error -16
>     musb-dsps 47401400.usb: can't request region for resource [mem
> 0x47401000-0x474011ff]
>     musb-hdrc musb-hdrc.0: musb_init_controller failed with status -16
>     musb-hdrc: probe of musb-hdrc.0 failed with error -16
> 
> Before, devm_ioremap_resource() was called on "dev" ("musb-hdrc.0" or
> "musb-hdrc.1"), after it is called on "&pdev->dev" ("47401400.usb" or
> "47401c00.usb"), leading to a duplicate region request, which fails.
> 
My bad, I didn't notice that, thanks a lot.

And sorry, I missed your message

> I have sent a revert:
> https://lore.kernel.org/linux-usb/20201112135900.3822599-1-geert+renesas@glider.be/
> 
> Gr{oetje,eeting}s,
> 
>                         Geert
>