[PATCH] nvme: drop bogus nid quirk for multipath devices which passed id test

Christoph Hellwig hch at lst.de
Wed Oct 29 02:59:07 PDT 2025


On Wed, Oct 29, 2025 at 10:45:57AM +0100, Daniel Wagner wrote:
> On Wed, Oct 29, 2025 at 08:44:03AM +0100, Christoph Hellwig wrote:
> > But more imporantly dropping the quirk for multipath devices is
> > fundamentally the wrong thing to do, as we really need proper IDs for
> > multipathing. So just add another entry for the Dell device using
> > the dell subvendor/subdevice id that does not have the quirk.
> 
> The problem it is not possible to distinguish between the two devices
> based on subvendor/subdevice ids. I've tried to find some way to keep
> those two devices apart, serial numbers but I don't think there is
> something usable. Thus the idea to rely our own sanity checks and
> enable the feature if these pass. 

The problem is that there is no such thing as a sanity check.  Otherwise
life would be easy.  But I really can't believe that Dell is incompetent
enough to not leave a mark of their OEM Firmware version anywhere.



More information about the Linux-nvme mailing list