diff mbox

[U-Boot] usb: host: ehci: samsung: Move hcor initialization after usb phy setup

Message ID 1449761545-4824-1-git-send-email-l.majewski@samsung.com
State Accepted
Delegated to: Marek Vasut
Headers show

Commit Message

Łukasz Majewski Dec. 10, 2015, 3:32 p.m. UTC 
With the old order of initialization the hcor pointer has been setup to
the same address as Exynos EHCI base address (0x12110000 instead of
0x12110010).
Such behaviour was caused by reading value of 0 instead of 0x10 from EHCI
HCCPBASE register without doing proper clock initialization before.

To fix this problem hcor initialization has been moved after USB PHY setup.
Now ehci_readl(&ctx->hcd->cr_capbase) returns correct value.

Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
---
 drivers/usb/host/ehci-exynos.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

Comments

Marek Vasut Dec. 10, 2015, 4:20 p.m. UTC | #1 
On Thursday, December 10, 2015 at 04:32:25 PM, Lukasz Majewski wrote:
> With the old order of initialization the hcor pointer has been setup to
> the same address as Exynos EHCI base address (0x12110000 instead of
> 0x12110010).
> Such behaviour was caused by reading value of 0 instead of 0x10 from EHCI
> HCCPBASE register without doing proper clock initialization before.
> 
> To fix this problem hcor initialization has been moved after USB PHY setup.
> Now ehci_readl(&ctx->hcd->cr_capbase) returns correct value.
> 
> Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
> ---
>  drivers/usb/host/ehci-exynos.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/usb/host/ehci-exynos.c
> b/drivers/usb/host/ehci-exynos.c index 18e9251..bede04b 100644
> --- a/drivers/usb/host/ehci-exynos.c
> +++ b/drivers/usb/host/ehci-exynos.c
> @@ -219,14 +219,14 @@ static int ehci_usb_probe(struct udevice *dev)
> 
>  	ctx->hcd = (struct ehci_hccr *)plat->hcd_base;
>  	ctx->usb = (struct exynos_usb_phy *)plat->phy_base;
> -	hcor = (struct ehci_hcor *)((uint32_t)ctx->hcd +
> -			HC_LENGTH(ehci_readl(&ctx->hcd->cr_capbase)));
> 
>  	/* setup the Vbus gpio here */
>  	if (dm_gpio_is_valid(&plat->vbus_gpio))
>  		dm_gpio_set_value(&plat->vbus_gpio, 1);
> 
>  	setup_usb_phy(ctx->usb);
> +	hcor = (struct ehci_hcor *)((uint32_t)ctx->hcd +
> +			HC_LENGTH(ehci_readl(&ctx->hcd->cr_capbase)));

Acked-by: Marek Vasut <marex@denx.de>

Best regards,
Marek Vasut
Simon Glass Dec. 10, 2015, 5:39 p.m. UTC | #2 
On 10 December 2015 at 08:32, Lukasz Majewski <l.majewski@samsung.com> wrote:
> With the old order of initialization the hcor pointer has been setup to
> the same address as Exynos EHCI base address (0x12110000 instead of
> 0x12110010).
> Such behaviour was caused by reading value of 0 instead of 0x10 from EHCI
> HCCPBASE register without doing proper clock initialization before.
>
> To fix this problem hcor initialization has been moved after USB PHY setup.
> Now ehci_readl(&ctx->hcd->cr_capbase) returns correct value.
>
> Signed-off-by: Lukasz Majewski <l.majewski@samsung.com>
> ---
>  drivers/usb/host/ehci-exynos.c | 4 ++--
>  1 file changed, 2 insertions(+), 2 deletions(-)

Reviewed-by: Simon Glass <sjg@chromium.org>
diff mbox

Patch

diff --git a/drivers/usb/host/ehci-exynos.c b/drivers/usb/host/ehci-exynos.c
index 18e9251..bede04b 100644
--- a/drivers/usb/host/ehci-exynos.c
+++ b/drivers/usb/host/ehci-exynos.c
@@ -219,14 +219,14 @@  static int ehci_usb_probe(struct udevice *dev)
 
 	ctx->hcd = (struct ehci_hccr *)plat->hcd_base;
 	ctx->usb = (struct exynos_usb_phy *)plat->phy_base;
-	hcor = (struct ehci_hcor *)((uint32_t)ctx->hcd +
-			HC_LENGTH(ehci_readl(&ctx->hcd->cr_capbase)));
 
 	/* setup the Vbus gpio here */
 	if (dm_gpio_is_valid(&plat->vbus_gpio))
 		dm_gpio_set_value(&plat->vbus_gpio, 1);
 
 	setup_usb_phy(ctx->usb);
+	hcor = (struct ehci_hcor *)((uint32_t)ctx->hcd +
+			HC_LENGTH(ehci_readl(&ctx->hcd->cr_capbase)));
 
 	return ehci_register(dev, ctx->hcd, hcor, NULL, 0, USB_INIT_HOST);
 }