Bug#782442: USB3 external HDD not recognized [regression]
- To: Mathias Nyman <mathias.nyman@linux.intel.com>, Alan Stern <stern@rowland.harvard.edu>, "Wang, Yu" <yu.y.wang@intel.com>, Mathias Nyman <mathias.nyman@intel.com>
- Cc: USB list <linux-usb@vger.kernel.org>, 782442@bugs.debian.org, baolu.lu@linux.intel.com, "Zhuang, Jin Can" <jin.can.zhuang@intel.com>
- Subject: Bug#782442: USB3 external HDD not recognized [regression]
- From: Ralf Jung <post@ralfj.de>
- Date: Tue, 02 Jun 2015 14:59:51 +0200
- Message-id: <[🔎] 556DA8C7.100@ralfj.de>
- Reply-to: Ralf Jung <post@ralfj.de>, 782442@bugs.debian.org
- In-reply-to: <[🔎] 556DA0A6.2040009@linux.intel.com>
- References: <Pine.LNX.4.44L0.1505101104400.23927-100000@netrider.rowland.org> <5550BD3B.10809@linux.intel.com> <5550C039.6060202@ralfj.de> <5551E458.1090007@linux.intel.com> <55526125.2030104@ralfj.de> <55685D28.2030200@ralfj.de> <[🔎] 556DA0A6.2040009@linux.intel.com>
Hi,
>>> Indeed doing "echo 'on' | sudo tee /sys/bus/pci/devices/*/power/control"
>>> fixes the problem: When I subsequently plug in the disk, it is
>>> recognized correctly. I confirmed this with an older kernel (the 3.16
>>> Debian kernel, to be more precise). That's good to know, now I don't
>>> need to reboot to a different kernel anymore to do fast backups ;-)
>>
>> Is there any news on this? Any further information I can provide to help
>> fix the issue?
>>
>
> I got some untested and not yet reviewed usb3 resume race fixes from Zhuang Jin Can
> If you can try them out and see if they help it would be great.
> They are sitting in a topic branch called "xhci_usb3_pm_fixes" in my tree:
> (5 extra patches on top of current usb-linus)
>
> git://git.kernel.org/pub/scm/linux/kernel/git/mnyman/xhci.git xhci_usb3_pm_fixes
This is my production and (main and only) work machine. If you tell me
testing that branch is safe and won't mess up my file systems or
hardware, I can give it a try.
Kind regards,
Ralf
>
> If not, then the we need to change xhci resume code so that we resume the roothubs as default
> except if we are runtime suspended and woken up just to be (system) suspended.
>
> -Mathias
>
>
>
>
Reply to: