[U-Boot] [PATCH v2] musb: sunxi: Use base address from device tree
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 --- Changes since v1:
- Check return value dev_read_addr_ptr()
IMHO, having NULL represent an error for dev_read_addr_ptr() doesn't work so well as in the kernel, because U-boot doesn't acually map addresses. So NULL (or 0x0) is in fact a valid address. Something like ~0x0 might work better, but that's a whole other changeset.
--- drivers/usb/musb-new/sunxi.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/usb/musb-new/sunxi.c b/drivers/usb/musb-new/sunxi.c index 7ee44ea91900..aedc24b93711 100644 --- a/drivers/usb/musb-new/sunxi.c +++ b/drivers/usb/musb-new/sunxi.c @@ -312,13 +312,16 @@ static int musb_usb_probe(struct udevice *dev) { struct musb_host_data *host = dev_get_priv(dev); struct usb_bus_priv *priv = dev_get_uclass_priv(dev); + void *base = dev_read_addr_ptr(dev); int ret;
+ if (!base) + return -EINVAL; + priv->desc_before_addr = true;
#ifdef CONFIG_USB_MUSB_HOST - host->host = musb_init_controller(&musb_plat, NULL, - (void *)SUNXI_USB0_BASE); + host->host = musb_init_controller(&musb_plat, NULL, base); if (!host->host) return -EIO;
@@ -326,7 +329,7 @@ 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, base); if (!ret) printf("Allwinner mUSB OTG (Peripheral)\n"); #endif
On Sat, Dec 30, 2017 at 08:44:07PM +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
Thanks! Maxime
On Thu, Jan 4, 2018 at 6:55 PM, Maxime Ripard maxime.ripard@free-electrons.com wrote:
On Sat, Dec 30, 2017 at 08:44:07PM +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
Applied to u-boot-sunxi/master
On 01/10/2018 07:10 AM, Jagan Teki wrote:
On Thu, Jan 4, 2018 at 6:55 PM, Maxime Ripard maxime.ripard@free-electrons.com wrote:
On Sat, Dec 30, 2017 at 08:44:07PM +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
Applied to u-boot-sunxi/master
This is a USB patch, this goes through USB tree, so drop it.
2017-12-30 20:44 GMT+08:00 Chen-Yu Tsai wens@csie.org:
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
Changes since v1:
- Check return value dev_read_addr_ptr()
IMHO, having NULL represent an error for dev_read_addr_ptr() doesn't work so well as in the kernel, because U-boot doesn't acually map addresses. So NULL (or 0x0) is in fact a valid address. Something like ~0x0 might work better, but that's a whole other changeset.
drivers/usb/musb-new/sunxi.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/usb/musb-new/sunxi.c b/drivers/usb/musb-new/sunxi.c index 7ee44ea91900..aedc24b93711 100644 --- a/drivers/usb/musb-new/sunxi.c +++ b/drivers/usb/musb-new/sunxi.c @@ -312,13 +312,16 @@ static int musb_usb_probe(struct udevice *dev) { struct musb_host_data *host = dev_get_priv(dev); struct usb_bus_priv *priv = dev_get_uclass_priv(dev);
void *base = dev_read_addr_ptr(dev); int ret;
Chenyu,
I am not familiar with MUSB and not sure whether this driver support both host and peripheral mode. I suppose it support both mode in different build config according to macro CONFIG_USB_MUSB_HOST. H3 user manual says register address space is different for the two modes, 0x1c19000 for device mode and 0x1c1a000 for host mode. If my understanding is correct, we need different address in dts for different mode. Or I misunderstand anything? Thank you for explaining it!
Jun
On Wed, Jan 10, 2018 at 10:01 PM, Jun Nie jun.nie@linaro.org wrote:
2017-12-30 20:44 GMT+08:00 Chen-Yu Tsai wens@csie.org:
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
Changes since v1:
- Check return value dev_read_addr_ptr()
IMHO, having NULL represent an error for dev_read_addr_ptr() doesn't work so well as in the kernel, because U-boot doesn't acually map addresses. So NULL (or 0x0) is in fact a valid address. Something like ~0x0 might work better, but that's a whole other changeset.
drivers/usb/musb-new/sunxi.c | 9 ++++++--- 1 file changed, 6 insertions(+), 3 deletions(-)
diff --git a/drivers/usb/musb-new/sunxi.c b/drivers/usb/musb-new/sunxi.c index 7ee44ea91900..aedc24b93711 100644 --- a/drivers/usb/musb-new/sunxi.c +++ b/drivers/usb/musb-new/sunxi.c @@ -312,13 +312,16 @@ static int musb_usb_probe(struct udevice *dev) { struct musb_host_data *host = dev_get_priv(dev); struct usb_bus_priv *priv = dev_get_uclass_priv(dev);
void *base = dev_read_addr_ptr(dev); int ret;Chenyu,
I am not familiar with MUSB and not sure whether this driver support both host and peripheral mode. I suppose it support both mode in different build config according to macro CONFIG_USB_MUSB_HOST. H3 user manual says register address space is different for the two modes, 0x1c19000 for device mode and 0x1c1a000 for host mode. If my understanding is correct, we need different address in dts for different mode. Or I misunderstand anything? Thank you for explaining it!
For host mode, the SoC provides a proper EHCI/OHCI host pair. Those are found at 0x1c1a000. However the MUSB controller can do host mode. It's just not as good as a proper host pair.
ChenYu
participants (5)
-
Chen-Yu Tsai -
Jagan Teki -
Jun Nie -
Marek Vasut -
Maxime Ripard