[PATCH v4 0/2] USB: propose a generic fix for PORT_SUSPEND set feature timeout

[PATCH v4 0/2] USB: propose a generic fix for PORT_SUSPEND set feature timeout

[chris.chiu]

From: Chris Chiu <chris.chiu@canonical.com>

For the Realtek Hub which fails to resume the port which has wakeup
enable descendants, trying to propose a more generic and better fix
to have the runtime suspend/resume work instead of a reset-resume quirk.

Chris Chiu (2):
  USB: Verify the port status when timeout happens during port suspend
  Revert "USB: Add reset-resume quirk for WD19's Realtek Hub"

 drivers/usb/core/hub.c    | 21 +++++++++++++++++++++
 drivers/usb/core/quirks.c |  1 -
 2 files changed, 21 insertions(+), 1 deletion(-)

-- 
2.20.1

[PATCH v4 1/2] USB: Verify the port status when timeout happens during port suspend

[chris.chiu]

From: Chris Chiu <chris.chiu@canonical.com>

On the Realtek high-speed Hub(0bda:5487), the port which has wakeup
enabled_descendants will sometimes timeout when setting PORT_SUSPEND
feature. After checking the PORT_SUSPEND bit in wPortStatus, it is
already set which means the port has been suspended. We should treat
it suspended to make sure it will be resumed correctly.

Signed-off-by: Chris Chiu <chris.chiu@canonical.com>
---

Changelog:
  v4:
    - add comment for the timeout handling
    - intialize the portstatus and portchange
    - revise the commit message to reflect the change
  v3:
    - create a new goto target for the timeout case instead of
      reset_resume
    - Revise the commit title/message because reset_resume is not
      required.
  v2:
    - create a new variable to keep the result of hub_port_status
      when suspend timeout.

 drivers/usb/core/hub.c | 21 +++++++++++++++++++++
 1 file changed, 21 insertions(+)

diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
index b2bc4b7c4289..e739f7b5991a 100644
--- a/drivers/usb/core/hub.c
+++ b/drivers/usb/core/hub.c
@@ -3385,6 +3385,26 @@ int usb_port_suspend(struct usb_device *udev, pm_message_t msg)
 		status = 0;
 	}
 	if (status) {
+		/* Check if the port has been suspended for the timeout case
+		 * to prevent the suspended port from incorrect handling.
+		 */
+		if (status == -ETIMEDOUT) {
+			int ret;
+			u16 portstatus, portchange;
+
+			portstatus = portchange = 0;
+			ret = hub_port_status(hub, port1, &portstatus,
+					&portchange);
+
+			dev_dbg(&port_dev->dev,
+				"suspend timeout, status %04x\n", portstatus);
+
+			if (ret == 0 && port_is_suspended(hub, portstatus)) {
+				status = 0;
+				goto suspend_done;
+			}
+		}
+
 		dev_dbg(&port_dev->dev, "can't suspend, status %d\n", status);
 
 		/* Try to enable USB3 LTM again */
@@ -3401,6 +3421,7 @@ int usb_port_suspend(struct usb_device *udev, pm_message_t msg)
 		if (!PMSG_IS_AUTO(msg))
 			status = 0;
 	} else {
+ suspend_done:
 		dev_dbg(&udev->dev, "usb %ssuspend, wakeup %d\n",
 				(PMSG_IS_AUTO(msg) ? "auto-" : ""),
 				udev->do_remote_wakeup);
-- 
2.20.1

[PATCH v4 2/2] Revert "USB: Add reset-resume quirk for WD19's Realtek Hub"

[chris.chiu]

From: Chris Chiu <chris.chiu@canonical.com>

This reverts commit ca91fd8c7643 ("USB: Add reset-resume quirk for
WD19's Realtek Hub"). The previous patch in the series now handles
the problematic hubs by checking the port status and handling it
accordingly when PORT_SUSPEND timeout occurs. We don't need to use
reset-resume all the time.

Signed-off-by: Chris Chiu <chris.chiu@canonical.com>
---

Changelog:
  v4:
    - correct the format of commit reference in the commit message
    - rephrase the commit message properly

 drivers/usb/core/quirks.c | 1 -
 1 file changed, 1 deletion(-)

diff --git a/drivers/usb/core/quirks.c b/drivers/usb/core/quirks.c
index 21e7522655ac..6114cf83bb44 100644
--- a/drivers/usb/core/quirks.c
+++ b/drivers/usb/core/quirks.c
@@ -406,7 +406,6 @@ static const struct usb_device_id usb_quirk_list[] = {
 
 	/* Realtek hub in Dell WD19 (Type-C) */
 	{ USB_DEVICE(0x0bda, 0x0487), .driver_info = USB_QUIRK_NO_LPM },
-	{ USB_DEVICE(0x0bda, 0x5487), .driver_info = USB_QUIRK_RESET_RESUME },
 
 	/* Generic RTL8153 based ethernet adapters */
 	{ USB_DEVICE(0x0bda, 0x8153), .driver_info = USB_QUIRK_NO_LPM },
-- 
2.20.1

Re: [PATCH v4 0/2] USB: propose a generic fix for PORT_SUSPEND set feature timeout

[Alan Stern]

On Fri, May 14, 2021 at 12:54:03PM +0800, chris.chiu@canonical.com wrote:
> From: Chris Chiu <chris.chiu@canonical.com>
> 
> For the Realtek Hub which fails to resume the port which has wakeup
> enable descendants, trying to propose a more generic and better fix
> to have the runtime suspend/resume work instead of a reset-resume quirk.
> 
> Chris Chiu (2):
>   USB: Verify the port status when timeout happens during port suspend
>   Revert "USB: Add reset-resume quirk for WD19's Realtek Hub"
> 
>  drivers/usb/core/hub.c    | 21 +++++++++++++++++++++
>  drivers/usb/core/quirks.c |  1 -
>  2 files changed, 21 insertions(+), 1 deletion(-)

For both patches:

Acked-by: Alan Stern <stern@rowland.harvard.edu>