Backporting of 3659fb5ac29a5e6102bebe494ac789fd47fb78f4 to 5.4?

[Keith Busch]

On Tue, Sep 30, 2025 at 10:34:35AM +0200, Rolf Eike Beer wrote:
> Hi all,
> 
> I just noticed that 3659fb5ac29a5e6102bebe494ac789fd47fb78f4 ("nvme: fix 
> multipath crash caused by flush request when blktrace is enabled") has not been 
> backported into 5.4 stable kernels, leaving CVE-2022-50388 unfixed. I assume 
> this has happened because it did not trivially apply.
> 
> Is this assumption correct? In this case please see the patch I'll send as 
> reply. If not, what is the reason?

It doesn't apply, though it's a trivial conflict resolution.