lkml.org 
[lkml]   [1998]   [Mar]   [25]   [last100]   RSS Feed
Views: [wrap][no wrap]   [headers]  [forward] 
 
Messages in this thread
/
Date
From
SubjectRe: "Greased TCPenguin"
Andi Kleen writes:
> Richard Gooch <rgooch@atnf.CSIRO.AU> writes:
>
> > Andrea Arcangeli writes:
> > > On Tue, 24 Mar 1998, Richard Gooch wrote:
> > >
> > > >Something still not right:
> > > >
> > > >TCPv4 bad checksum from 130.155.198.47:03ff to 130.155.191.26:03fd, len=32/32/52
> > > >TCPv4 bad checksum from 130.155.198.47:03ff to 130.155.191.26:03fd, len=32/32/52
> > > >TCPv4 bad checksum from 130.155.198.47:0448 to 130.155.191.26:1770, len=32/32/52
> > > >TCPv4 bad checksum from 130.155.198.47:0201 to 130.155.191.26:03ff, len=20/20/4
> > >
> > > I read messages such us this a lot of times but I never related it to a
> > > possible TCP bug, I always think that the ppp/slip is a poor connection
> > > and can miss something (we use checksum for that cases no?).
> >
> > My modem link has compression and error checking, so in theory I have
> ^^^^^^^^^
> (but your serial cable between computer and modem hasn't), also note
> that theory and practice often differ. Another possibility is that
> the packets got corrupted further down the link.

The modem cable is properly shielded. The serial link runs at 115 k.
Is this really likely?

The router box at home is running 2.1.86. Is this perhaps causing
corruption?

> > a perfect link. I also didn't get these messages with earlier versions
> > of the kernel (it seems 2.1.86 is the first time I got it, according
> > to my kernel logs).
>
> I assume that is because 2.1.91-pre stresses your PPP link much more
> than earlier 2.1 kernels. Note that 2.0 doesn't look these messages,
> because they're for debugging only (and actually forbidden by RFC1122
> - which specifies that these packets should be "silently discarded").

BTW: I'm using CSLIP, not PPP. I don't get any more performance from
my serial link than what I did with earlier 2.1.x kernels (the one
before the TCP breakage).

> To state it again - some checksum errors are perfectly ok. One should
> only worry when they happen extensively.

I got 13 in the space of 50 minutes. All I was doing was reading email
(Emacs ViewMail over lbx) and transferred (rcp) two small files (one
around 9 kBytes and the other under 1 kByte).

Regards,

Richard....
-
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 13:41    [from the cache]
©2003-2011 Jasper Spaans