lkml.org 
[lkml]   [1998]   [Apr]   [18]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
From
SubjectRe: Reproducable 2.1.95 checksum error
Date
On 17 Apr 1998 21:29:41 +0200, Benny Amorsen <amorsen@sscnet.com> wrote:

>>>>>> "MR" == Meelis Roos <mroos@tartu.cyber.ee> writes:
>
>>>>>> "L" == <linker@nightshade.ml.org> writes:
>
>L>> Isn't it against some RFC to log bad checksums? It certantly does
>L>> no good for logs.. I was under the impression that this was a
>L>> debugging feature that would go away with 2.2? Though it should
>L>> perhaps stay as a /proc/sysctl option..
>
>MR> Yes, AFAIK this is a debugging feature that will probably go away.
>MR> But You can't say that it's no good - it's one of the things that
>MR> helps You to see that something is wrong somehere. (yes, I know
>MR> the ifconfig output, but how often do You check it?)
>
>I was specifically talking about the ifconfig output, or rather its
>equivalent /proc/net/dev.
>
>Does the 2.0 kernel not increment the error count when it receives a
>frame with a bad checksum? -- If a standard says that bad checksums
>must not be counted, the standard is broken and should not be
>followed.
>
>The actual logging to syslog of wrong checksums is bad, of course, and
>should go for 2.2. I haven't seen anyone saying otherwise.
>
The current checksum logging seems to be exposing actual problems. Some
are local code problems and some are remote site problems. And some are
occaisional local link problems. It seems to me that the results are at
least worth an option to turn logging on.

john alvord

-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.rutgers.edu

\
 
 \ /
  Last update: 2005-03-22 13:42    [W:0.465 / U:0.104 seconds]
©2003-2020 Jasper Spaans|hosted at Digital Ocean and TransIP|Read the blog|Advertise on this site