lkml.org 
[lkml]   [2000]   [Feb]   [11]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: elevator-starvation-4 (2.2.14 && 2.3.42)
Bruce Thompson wrote:
>
> I tend to disagree. If we assume that cat /dev/zero > file fills
> consecutive blocks with zeroes then I would argue that the relative
> priority between the driver space handling block requests and user
> space generating requests is slightly screwy. In a case where
> requests are being generated faster than the driver can deal with
> them then no matter how we order the request queue, we've got a
> problem.
>
> I would argue that as long as the relative priority between user
> space and driver space is such that a swamped driver can handle at
> least N + 1 requests in the time that user space can generate N
> requests then we have a stable situation and the driver can
> eventually catch up and service all requests. Otherwise, I claim that
> I can come up with a pathological case for any aging scheme that will
> result in requests being starved. [...]

Surely you realize that the case where userland is generates requests
faster than the driver can fulfill is a common case...

Jeff



--
Jeff Garzik | "Vegetarian" is the Indian word
Building 1024 | for 'lousy hunter.'
MandrakeSoft, Inc. |
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu
Please read the FAQ at http://www.tux.org/lkml/

\
 
 \ /
  Last update: 2005-03-22 13:56    [from the cache]
©2003-2011 Jasper Spaans