Re: [PATCH 2/2] usb: core: Attempt power cycle when port is in eSS.Disabled state

[Kai Heng Feng <kai.heng.feng@canonical.com>]

> On Nov 4, 2019, at 10:38 PM, Greg KH <gregkh@linuxfoundation.org> wrote:
> 
> On Tue, Oct 08, 2019 at 02:28:40AM +0800, Kai-Heng Feng wrote:
>> On Dell TB16, Realtek USB ethernet (r8152) connects to an SMSC hub which
>> then connects to ASMedia xHCI's root hub:
>> 
>> /:  Bus 04.Port 1: Dev 1, Class=root_hub, Driver=xhci_hcd/2p, 5000M
>>    |__ Port 1: Dev 2, If 0, Class=Hub, Driver=hub/7p, 5000M
>>            |__ Port 2: Dev 3, If 0, Class=Vendor Specific Class, Driver=r8152, 5000M
>> 
>> Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
>> Bus 004 Device 002: ID 0424:5537 Standard Microsystems Corp. USB5537B
>> Bus 004 Device 003: ID 0bda:8153 Realtek Semiconductor Corp. RTL8153 Gigabit Ethernet Adapter
>> 
>> The SMSC hub may disconnect after system resume from suspend. When this
>> happens, the reset resume attempt fails, and the last resort to disable
>> the port and see something comes up later, also fails.
>> 
>> When the issue occurs, the link state stays in eSS.Disabled state
>> despite the warm reset attempts. The USB spec mentioned this can be
>> caused by invalid VBus, and after some expiremets, it does show that the
>> SMSC hub can be brought back after a power cycle.
>> 
>> So let's power cycle the port at the end of reset resume attempt, if
>> it's in eSS.Disabled state.
>> 
>> Signed-off-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
>> ---
>> drivers/usb/core/hub.c | 21 +++++++++++++++++++--
>> 1 file changed, 19 insertions(+), 2 deletions(-)
>> 
>> diff --git a/drivers/usb/core/hub.c b/drivers/usb/core/hub.c
>> index 6655a6a1651b..5f50aca7cf67 100644
>> --- a/drivers/usb/core/hub.c
>> +++ b/drivers/usb/core/hub.c
>> @@ -2739,20 +2739,33 @@ static bool hub_port_warm_reset_required(struct usb_hub *hub, int port1,
>> 		|| link_state == USB_SS_PORT_LS_COMP_MOD;
>> }
>> 
>> +static bool hub_port_power_cycle_required(struct usb_hub *hub, int port1,
>> +		u16 portstatus)
>> +{
>> +	u16 link_state;
>> +
>> +	if (!hub_is_superspeed(hub->hdev))
>> +		return false;
>> +
>> +	link_state = portstatus & USB_PORT_STAT_LINK_STATE;
>> +	return link_state == USB_SS_PORT_LS_SS_DISABLED;
>> +}
>> +
>> static void hub_port_power_cycle(struct usb_hub *hub, int port1)
>> {
>> +	struct usb_port *port_dev = hub->ports[port1  - 1];
>> 	int ret;
>> 
>> 	ret = usb_hub_set_port_power(hub, port1, false);
>> 	if (ret) {
>> -		dev_info(&udev->dev, "failed to disable port power\n");
>> +		dev_info(&port_dev->dev, "failed to disable port power\n");
>> 		return;
>> 	}
>> 
>> 	msleep(2 * hub_power_on_good_delay(hub));
>> 	ret = usb_hub_set_port_power(hub, port1, true);
>> 	if (ret) {
>> -		dev_info(&udev->dev, "failed to enable port power\n");
>> +		dev_info(&port_dev->dev, "failed to enable port power\n");
>> 		return;
>> 	}
>> 
>> @@ -3600,6 +3613,10 @@ int usb_port_resume(struct usb_device *udev, pm_message_t msg)
>> 	if (status < 0) {
>> 		dev_dbg(&udev->dev, "can't resume, status %d\n", status);
>> 		hub_port_logical_disconnect(hub, port1);
>> +		if (hub_port_power_cycle_required(hub, port1, portstatus)) {
>> +			dev_info(&udev->dev, "device in disabled state, attempt power cycle\n");
> 
> Why dev_info()?  Shouldn't we only care if this fails?

I’ll lower the level to dev_dbg().

> 
>> +			hub_port_power_cycle(hub, port1);
> 
> Weren't we only going to do this for the broken types of devices?  And
> not for everything?

From what I can understand from the spec, if the device is in eSS.Disabled state, there’s no way out.
So "power cycling as a last resort” is indeed targets everything.

Kai-Heng

> thanks,
> 
> greg k-h