On Mon, 2006-12-11 at 17:22 +0100, Andre Brinkmann wrote: > > -----Ursprüngliche Nachricht----- > > Von: Jens Axboe [mailto:[EMAIL PROTECTED] > > Gesendet: Montag, 11. Dezember 2006 17:12 > > An: Ross S. W. Walker > > Cc: [EMAIL PROTECTED]; [email protected]; > > [EMAIL PROTECTED] > > Betreff: Re: how to explain this? > > > > On Mon, Dec 11 2006, Ross S. W. Walker wrote: > > > > > > So the only way to assure order with bio is use of BIO_RW_BARRIER? > > > > (don't top post) > > > > The answer is yes. We probably could split the BIO_RW_BARRIER to only > > provide io scheduler ordering like the REQ_SOFTBARRIER does for > > requests. So far there hasn't been a use for it. > > > > Dumb question: It is possible that a request to block A outruns a request to > block B. But is it also possible that a read request to block A outruns a > write request to block A? > > Isn't there any kind of checking inside the elevator algorithms?
as Jens pointed in previous email, this is upper layer responsibility. > > André > > > -- > > Jens Axboe > > > - To unsubscribe from this list: send the line "unsubscribe linux-btrace" in the body of a message to [EMAIL PROTECTED] More majordomo info at http://vger.kernel.org/majordomo-info.html
