[bug report] WARNING: possible circular locking dependency detected with blktests nvme-tcp nvme/013

[Sagi Grimberg]

>> Also, I might have a sleeping function called in atomic context report
>> with a similar enough call trace to take a closer look at the direct send.
> 
> That looks to be a backport run with device mapper multipath, and a
> repeat of the dm-rq BLK_MQ_F_BLOCKING discussion on linux-block.
> 
> I currently don't see a reason why the lockdep annotation wouldn't be
> enough for this.  I'll send a real patch if that's the way we want to go.

Hey Chris,

Are you sending a patch for this? You can add my review tag when you do.