Re: TCPv4 bad checksum

Tim P. Gerla (timg@means.net)
Thu, 19 Jun 1997 13:52:19 -0500 (CDT)

On Wed, 18 Jun 1997, Chris Molnar wrote:

> Jun 18 16:08:41 newn kernel: eth0 changes netmask 00FFFFFF -> F8FFFFFF
> Jun 18 16:09:26 newn kernel: ppp0 UP fl=00000051 pa=11103DD1/FFFFFFFF brd=00000000 dst=1E213CCF
> Jun 18 16:09:26 newn kernel: pppd forgot to specify route netmask.
> Jun 18 16:15:50 newn kernel: TCPv4 bad checksum from 0e5a45c6:0014 to 11103dd1:044e, len=532/532/552
> Jun 18 16:16:35 newn last message repeated 16 times
> Jun 18 16:16:37 newn last message repeated 6 times
> Jun 18 16:18:36 newn kernel: TCPv4 bad checksum from 0e5a45c6:0014 to 11103dd1:046a, len=532/532/552
> Jun 18 16:19:07 newn last message repeated 17 times
> Jun 18 16:19:09 newn last message repeated 8 times
>
> Does anyone have any idea what this means.... (the bad checksum lines)....
> I am on a pentium machine, using ne2000 card, and has 32meg ram, 8 Gig hd.
> Also kernel version is 2.1.42.
>
I get these messages a lot too, but I'm connected to a normal PPP dialup
link. They don't seem to cause any problems though.

> If these errors are documented anyplace public can someone point me in the
> right direction? Also, GCC error 11 is something else I see a lot.
>
That *could* be a hardware problem. I was having many of those, and I had
to replace my ram. Now it never happens. There's an entire web site on
this, but I forgot the url.. I'm sure someone will remember. :)

-Tim

--
Tim P. Gerla | timg@means.net | http://timg.home.ml.org/ | Tybstar@IRC


Use NT if you want to hack out a way to restart your telnetd
when it crashes. Use Linux if you don't want your telnetd to
crash. http://www.linux.com/