On 4/14/20 3:39 PM, Pavel Begunkov wrote: > [4/4] is dirty, but fixes the issue. And there is still "SQ vs CQ" > problem, solving which can effectively revert it, so I suggest to > postpone the last patch for a while. I'll rebase if it'd be necessary. > Thanks, I've applied 1-3 for now. Looks good to me, and also tests out fine. -- Jens Axboe