| Message ID | 20171212072421.8867-1-wens@csie.org |
|---|---|
| State | Superseded |
| Delegated to: | Marek Vasut |
| Headers | show |
| Series | [U-Boot] musb: sunxi: Use base address from device tree | expand |
On Tue, Dec 12, 2017 at 03:24:21PM +0800, Chen-Yu Tsai wrote: > Now that the musb sunxi glue driver is completely device model / device > tree driven, we should use the base address from the device tree, > instead of hard-coding it in the source code. > > Fixes: 3a61b080acee ("musb: sunxi: switch to the device model") > Signed-off-by: Chen-Yu Tsai <wens@csie.org> Acked-by: Maxime Ripard <maxime.ripard@free-electrons.com> We should probably do the same for the phy. Maxime
On 12/12/2017 08:24 AM, Chen-Yu Tsai wrote: > Now that the musb sunxi glue driver is completely device model / device > tree driven, we should use the base address from the device tree, > instead of hard-coding it in the source code. > > Fixes: 3a61b080acee ("musb: sunxi: switch to the device model") > Signed-off-by: Chen-Yu Tsai <wens@csie.org> > --- > drivers/usb/musb-new/sunxi.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) > > diff --git a/drivers/usb/musb-new/sunxi.c b/drivers/usb/musb-new/sunxi.c > index 7ee44ea91900..ae6a54118cd1 100644 > --- a/drivers/usb/musb-new/sunxi.c > +++ b/drivers/usb/musb-new/sunxi.c > @@ -318,7 +318,7 @@ static int musb_usb_probe(struct udevice *dev) > > #ifdef CONFIG_USB_MUSB_HOST > host->host = musb_init_controller(&musb_plat, NULL, > - (void *)SUNXI_USB0_BASE); > + dev_read_addr_ptr(dev)); What happens if dev_read_addr_ptr() returns some non-standard value? > if (!host->host) > return -EIO; > > @@ -326,7 +326,7 @@ static int musb_usb_probe(struct udevice *dev) > if (!ret) > printf("Allwinner mUSB OTG (Host)\n"); > #else > - ret = musb_register(&musb_plat, NULL, (void *)SUNXI_USB0_BASE); > + ret = musb_register(&musb_plat, NULL, dev_read_addr_ptr(dev)); > if (!ret) > printf("Allwinner mUSB OTG (Peripheral)\n"); > #endif >
On 12 December 2017 at 00:24, Chen-Yu Tsai <wens@csie.org> wrote: > Now that the musb sunxi glue driver is completely device model / device > tree driven, we should use the base address from the device tree, > instead of hard-coding it in the source code. > > Fixes: 3a61b080acee ("musb: sunxi: switch to the device model") > Signed-off-by: Chen-Yu Tsai <wens@csie.org> > --- > drivers/usb/musb-new/sunxi.c | 4 ++-- > 1 file changed, 2 insertions(+), 2 deletions(-) Reviewed-by: Simon Glass <sjg@chromium.org>
On Tue, Dec 12, 2017 at 11:56 PM, Marek Vasut <marex@denx.de> wrote: > On 12/12/2017 08:24 AM, Chen-Yu Tsai wrote: >> Now that the musb sunxi glue driver is completely device model / device >> tree driven, we should use the base address from the device tree, >> instead of hard-coding it in the source code. >> >> Fixes: 3a61b080acee ("musb: sunxi: switch to the device model") >> Signed-off-by: Chen-Yu Tsai <wens@csie.org> >> --- >> drivers/usb/musb-new/sunxi.c | 4 ++-- >> 1 file changed, 2 insertions(+), 2 deletions(-) >> >> diff --git a/drivers/usb/musb-new/sunxi.c b/drivers/usb/musb-new/sunxi.c >> index 7ee44ea91900..ae6a54118cd1 100644 >> --- a/drivers/usb/musb-new/sunxi.c >> +++ b/drivers/usb/musb-new/sunxi.c >> @@ -318,7 +318,7 @@ static int musb_usb_probe(struct udevice *dev) >> >> #ifdef CONFIG_USB_MUSB_HOST >> host->host = musb_init_controller(&musb_plat, NULL, >> - (void *)SUNXI_USB0_BASE); >> + dev_read_addr_ptr(dev)); > > What happens if dev_read_addr_ptr() returns some non-standard value? You mean like an error value? That would likely be a signal that the device tree is broken, like a missing property or invalid value, or a corrupt device tree. In all cases I think blowing up should be an expected result for a device tree driven driver? ChenYu > >> if (!host->host) >> return -EIO; >> >> @@ -326,7 +326,7 @@ static int musb_usb_probe(struct udevice *dev) >> if (!ret) >> printf("Allwinner mUSB OTG (Host)\n"); >> #else >> - ret = musb_register(&musb_plat, NULL, (void *)SUNXI_USB0_BASE); >> + ret = musb_register(&musb_plat, NULL, dev_read_addr_ptr(dev)); >> if (!ret) >> printf("Allwinner mUSB OTG (Peripheral)\n"); >> #endif >> > > > -- > Best regards, > Marek Vasut
On 12/19/2017 05:04 PM, Chen-Yu Tsai wrote: > On Tue, Dec 12, 2017 at 11:56 PM, Marek Vasut <marex@denx.de> wrote: >> On 12/12/2017 08:24 AM, Chen-Yu Tsai wrote: >>> Now that the musb sunxi glue driver is completely device model / device >>> tree driven, we should use the base address from the device tree, >>> instead of hard-coding it in the source code. >>> >>> Fixes: 3a61b080acee ("musb: sunxi: switch to the device model") >>> Signed-off-by: Chen-Yu Tsai <wens@csie.org> >>> --- >>> drivers/usb/musb-new/sunxi.c | 4 ++-- >>> 1 file changed, 2 insertions(+), 2 deletions(-) >>> >>> diff --git a/drivers/usb/musb-new/sunxi.c b/drivers/usb/musb-new/sunxi.c >>> index 7ee44ea91900..ae6a54118cd1 100644 >>> --- a/drivers/usb/musb-new/sunxi.c >>> +++ b/drivers/usb/musb-new/sunxi.c >>> @@ -318,7 +318,7 @@ static int musb_usb_probe(struct udevice *dev) >>> >>> #ifdef CONFIG_USB_MUSB_HOST >>> host->host = musb_init_controller(&musb_plat, NULL, >>> - (void *)SUNXI_USB0_BASE); >>> + dev_read_addr_ptr(dev)); >> >> What happens if dev_read_addr_ptr() returns some non-standard value? > > You mean like an error value? That would likely be a signal that > the device tree is broken, like a missing property or invalid value, > or a corrupt device tree. In all cases I think blowing up should be > an expected result for a device tree driven driver? What about handling the errors instead ? :)
diff --git a/drivers/usb/musb-new/sunxi.c b/drivers/usb/musb-new/sunxi.c index 7ee44ea91900..ae6a54118cd1 100644 --- a/drivers/usb/musb-new/sunxi.c +++ b/drivers/usb/musb-new/sunxi.c @@ -318,7 +318,7 @@ static int musb_usb_probe(struct udevice *dev) #ifdef CONFIG_USB_MUSB_HOST host->host = musb_init_controller(&musb_plat, NULL, - (void *)SUNXI_USB0_BASE); + dev_read_addr_ptr(dev)); if (!host->host) return -EIO; @@ -326,7 +326,7 @@ static int musb_usb_probe(struct udevice *dev) if (!ret) printf("Allwinner mUSB OTG (Host)\n"); #else - ret = musb_register(&musb_plat, NULL, (void *)SUNXI_USB0_BASE); + ret = musb_register(&musb_plat, NULL, dev_read_addr_ptr(dev)); if (!ret) printf("Allwinner mUSB OTG (Peripheral)\n"); #endif
Now that the musb sunxi glue driver is completely device model / device tree driven, we should use the base address from the device tree, instead of hard-coding it in the source code. Fixes: 3a61b080acee ("musb: sunxi: switch to the device model") Signed-off-by: Chen-Yu Tsai <wens@csie.org> --- drivers/usb/musb-new/sunxi.c | 4 ++-- 1 file changed, 2 insertions(+), 2 deletions(-)