Re: Disk scheduling policy (Re: NEW_BUFQ_STRATEGY)

On Mon, Dec 01, 2003 at 04:14:54PM -0800, Jason Thorpe wrote:
>
> Really, it's not clear that the elevator sort buys you much anyway, 
> when you're talking to raw disks, because disks don't really expose 
> their real geometry anymore.

Well, I have yet to encounter one that randomly shuffles blocks
such that sorting in order of increasing block number would be
_harmful_.  Indeed, since we can see more information about future
requests than the disk can, it's not clear to me why the elevator
sort is not highly beneficial.

> I would also like to see a disk sorting algorithm that could coalesce 
> adjacent writes or reads into single requests (perhaps enqueueing an 
> uber-buf that pointed to a list of sub-bufs that were treated as s/g 
> elements, or something).  As part of this, I'd really like to add a 

I thought I suggested this to you about a year ago and you were rather
strongly opposed, due to the VM system tricks involved in the obvious
way one would do this given the rest of our current implementation?  I
am _definitely_ in favor of request coalescing in disksort.

-- 
 Thor Lancelot Simon	                                      [email protected]
   But as he knew no bad language, he had called him all the names of common
 objects that he could think of, and had screamed: "You lamp!  You towel!  You
 plate!" and so on.              --Sigmund Freud