On Wed, Jun 13, 2018 at 10:56:32AM -0600, Keith Busch wrote: > On Wed, Jun 13, 2018 at 05:37:35PM +0100, Ben Hutchings wrote: > > On Wed, 2018-06-13 at 10:30 -0600, Keith Busch wrote: > > > Thank you very much the catching that. This particular patch is indeed > > > not correct without the following: > > > > > > 161b8be2bd6ab ("nvme-pci: initialize queue memory before interrupts") > > > > Oh, now I see. Do you think that commit is suitable for stable > > branches? > > Yes, we can safely include that commit 4.4-stable. The scenario it fixes > isn't very common in 4.4, but it could happen with quirky hardware or > legacy interrupts. It does not apply to 4.4.y right now, but it does apply to 4.9.y, so I've added it there. If it needs to get to 4.4.y, can someone do a backport? thanks, greg k-h