[PATCH] PCI: ensure the PCI device is locked over ->reset_notify calls

[Christoph Hellwig]

On Tue, May 30, 2017 at 05:28:44PM -0500, Bjorn Helgaas wrote:
> [+cc Alex]
> 
> On Tue, May 23, 2017 at 07:42:02AM +0200, Christoph Hellwig wrote:
> > Without this ->notify_reset instance may race with ->remove calls,
> 
> Do you mean the .reset_notify() method in struct pci_error_handlers?
> I don't see a "notify_reset" symbol.

Yes.

> Can you elaborate on exactly how this race happens?  I'm trying to
> figure out whether this is also a problem or potential problem with
> other reset paths like pci_try_reset_function(), pci_reset_bus(),
> pci_try_reset_bus(), pci_reset_slot(), and pci_try_reset_slot().
> 
> What does the race look like when it happens?  Oops, panic, etc?
> 
> Can this also be triggered via the sysfs "reset" file?

Here is the original bug report:

http://lists.infradead.org/pipermail/linux-nvme/2017-May/010345.html

the issue is triggered through sysfs.