Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
Date:   Thu, 5 Aug 2021 11:53:06 +0200
From:   Greg KH <gregkh@linuxfoundation.org>
To:     Mathias Nyman <mathias.nyman@linux.intel.com>
Cc:     linux-usb@vger.kernel.org, linux-kernel@vger.kernel.org
Subject: Re: USB xhci crash under load on 5.14-rc3
Message-ID: <YQu1AgtuCWaW32U9@kroah.com>
References: <YQpJLk42b+0g7WFB@kroah.com>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <YQpJLk42b+0g7WFB@kroah.com>
Precedence: bulk

On Wed, Aug 04, 2021 at 10:00:46AM +0200, Greg KH wrote:
> Hi,
> 
> I was doing some filesystem backups from one USB device to another one
> this weekend and kept running into the problem of the xhci controller
> shutting down after an hour or so of high volume traffic.
> 
> I finally captured the problem in the kernel log as this would also take
> out my keyboard, making it hard to recover from :)
> 
> The log is below for when the problem happens, and then the devices are
> disconnected from the bus (ignore the filesystem errors, those are
> expected when i/o is in flight and we disconnect a device.
> 
> Any hint as to what the IO_PAGE_FAULT error messages are?
> 
> I'll go back to 5.13.y now and see if I can reproduce it there or not,
> as my backups are not yet done...

5.13.7 seemed to work, but locked up overnight when running this same
workload, so it is a bit worse.  Let me split the USB devices across PCI
devices and see if that helps...