Shaohua Li writes:
>> I like the general approach. I do wonder whether we should hold back a
>> single I/O at all times, or whether we should do something similar to
>> what Mike Snitzer did in the dm-mpath driver, where he keeps track of
>> the end position of the last I/O (without holding the
On Mon, Apr 20, 2015 at 01:34:24PM -0400, Jeff Moyer wrote:
> Shaohua Li writes:
>
> > Previous post of the patch is lost, so I repost. This is helpful, for
> > example, using scsi-mq for a sata drive.
> >
> > plug is still helpful for workload with IO merge, but it can be harmful
> > otherwise e
Shaohua Li writes:
> Previous post of the patch is lost, so I repost. This is helpful, for
> example, using scsi-mq for a sata drive.
>
> plug is still helpful for workload with IO merge, but it can be harmful
> otherwise especially with multiple hardware queues, as there is (supposed) no
> lock
3 matches
Mail list logo
