On Sat, 2020-06-27 at 12:07 -0700, Samuel Sieb wrote:
On 6/27/20 3:43 AM, Patrick O'Callaghan wrote:
> On Fri, 2020-06-26 at 16:00 -0700, Samuel Sieb wrote:
> > I messed around
> > enough that I confused that USB controller and that entire bus was dead.
> > Fortunately, I managed to reset the controller and got it back by
> > telling the driver to remove it and then add it back.
>
> How did you do that? Did it involve physically pulling and reinserting
> the drive?
Somehow I ended up in /sys/bus/pci/drivers/ehci-pci, but in your case it
might need to be /sys/bus/pci/drivers/xhci-pci since you're dealing with
a USB 3 device. In there are links to the USB pci devices that are
managed by this driver. From the logs you included, it would be
"0000:00:14.0" in your case. Doing "echo 0000:00:14.0 > unbind"
will
disconnect the entire usb bus on that device and "echo 0000:00:14.0 >
bind" will re-initialize it.
Interesting. I'll look into that.
> (Note the message from mdadm. The RAID array has /dev/sdd and
/dev/sde,
> but, again, RAID has already been stopped before running udisksctl.)
I really have my doubts about that. Something definitely still has the
drives open at this point. How did you stop it?
echo 1 > /sys/block/sdd/device/delete
echo 1 > /sys/block/sde/device/delete
poc