2018-10-03 05:19:04

by Kai-Heng Feng

Subject: [PATCH] usb: core: Check portchange C_PORT_RESET after resetting

Based on USB2.0 Spec Section 11.24.2.7.2.5:

"This bit is set when the port transitions from the Resetting state (or,
if present, the Speed_eval state) to the Enabled state."

Also Section 11.24.2.13:

"Setting the reset feature PORT_RESET causes the hub to signal reset on
that port. When the reset signaling is complete, the hub sets the
C_PORT_RESET status change and immediately enables the port."

So let's also check C_PORT_RESET for reset completion.

Signed-off-by: Kai-Heng Feng <[email protected]>
---
drivers/usb/core/hub.c | 3 ++-
1 file changed, 2 insertions(+), 1 deletion(-)

diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
index 7801bb30bdba..d96058372280 100644
--- a/drivers/usb/core/hub.c
+++ b/drivers/usb/core/hub.c
@@ -2721,7 +2721,8 @@ static int hub_port_wait_reset(struct usb_hub *hub, int port1,
* so also wait for the connection to be re-established.
*/
if (!(portstatus & USB_PORT_STAT_RESET) &&
- (portstatus & USB_PORT_STAT_CONNECTION))
+ (portstatus & USB_PORT_STAT_CONNECTION) &&
+ (portchange & USB_PORT_STAT_C_RESET))
break;

/* switch to the long delay after two short delay failures */
--
2.17.1

2018-10-03 13:57:29

by Alan Stern

Subject: Re: [PATCH] usb: core: Check portchange C_PORT_RESET after resetting

On Wed, 3 Oct 2018, Kai-Heng Feng wrote:

> Based on USB2.0 Spec Section 11.24.2.7.2.5:
>
> "This bit is set when the port transitions from the Resetting state (or,
> if present, the Speed_eval state) to the Enabled state."
>
> Also Section 11.24.2.13:
>
> "Setting the reset feature PORT_RESET causes the hub to signal reset on
> that port. When the reset signaling is complete, the hub sets the
> C_PORT_RESET status change and immediately enables the port."
>
> So let's also check C_PORT_RESET for reset completion.
>
> Signed-off-by: Kai-Heng Feng <[email protected]>
> ---
> drivers/usb/core/hub.c | 3 ++-
> 1 file changed, 2 insertions(+), 1 deletion(-)
>
> diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
> index 7801bb30bdba..d96058372280 100644
> --- a/drivers/usb/core/hub.c
> +++ b/drivers/usb/core/hub.c
> @@ -2721,7 +2721,8 @@ static int hub_port_wait_reset(struct usb_hub *hub, int port1,
> * so also wait for the connection to be re-established.
> */
> if (!(portstatus & USB_PORT_STAT_RESET) &&
> - (portstatus & USB_PORT_STAT_CONNECTION))
> + (portstatus & USB_PORT_STAT_CONNECTION) &&
> + (portchange & USB_PORT_STAT_C_RESET))
> break;
>
> /* switch to the long delay after two short delay failures */

What is the purpose of this patch? Does it fix a problem?

Alan Stern