On Mon, Dec 19, 2011 at 2:08 PM, Tejun Heo <tj@xxxxxxxxxx> wrote: > Hello, > > On Mon, Dec 19, 2011 at 02:01:47PM -0800, Dan Williams wrote: >> > Dan, as I replied before, I'm not a big fan of this approach. >> >> Sorry I must have missed it, I can't seem to find a reply in the archives? > > Yeah, I can't find it either. I definitely remember writing it. > Hmmm... weird. Either I'm finally losing my mind or it didn't get out > for some reason. Sorry. :) As long as it's you losing your mind and not me :-). >> > For >> > now, I think it would be best to add private wrapper in libsas to >> > support deferring unchained work items while draining. >> >> Ok, a form of this was nak'd by James before [1], but I can try again >> with pushing this chained submission checking down into scsi. > > The issues I see with the proposed change is, > > * There doesn't seem to be high demand for it. > > * It isn't implemented the right way - it introduces unnecessary and > hidden ordering between chained work items being drained and newly > queued unchained ones. We can try to do it properly without > affecting new unchained work items but I'm not sure the added > complexity is justified given the first issue. Fair enough. > I don't think adding a wrapper which defers queueing while draining is > going on would be too complex, right? No, I think it should be pretty straightforward to teach libsas not to throw anything else at the queue while a drain is pending. Thanks, Dan -- To unsubscribe from this list: send the line "unsubscribe linux-scsi" in the body of a message to majordomo@xxxxxxxxxxxxxxx More majordomo info at http://vger.kernel.org/majordomo-info.html
