Re: [RFT PATCH 0/2] xhci: clean up endpoint rings after disconnect detect

[Mathias Nyman <mathias.nyman@linux.intel.com>]

>> Only Reset Endpoint and Set TR Dequeue Pointer commands are issued;
>> ep0 is not restarted while the disconnect is pending.
> 
> That is what the log shows here.  Reset Endpoint on the disconnected ep0
> completed with success, Set TR Dequeue Pointer succeeded, the control
> URB was given back with -71, and later URBs were refused until hub_wq
> re-enumerated the device.  The other endpoints were given back without a
> reset.  I saw no controller hang or command timeout, so this Intel xHC
> is fine with Reset Endpoint on a disconnected ep0 (it ran twice).
> 
>> If these patches work, they might be squashed into one before going
>> upstream.
> 
> Squashing is fine with me.  1/2 alone leaves the stale TRBs on the ep0
> ring, as the cover letter says, so I would not apply it separately.
> 
> Tested-by: Yiyu Zhou <yiyu@yiyuzhou.io>
> 
Thanks for testing, I'll squash them together and send them forward

-Mathias