Messages in this thread | | | Subject | Re: Massive TCPv4 bad checksum | From | Andi Kleen <> | Date | 18 Oct 1998 12:45:41 +0200 |
| |
In article <m0zUl2h-000396C@ocean.lucon.org>, hjl@lucon.org (H.J. Lu) writes: >> >> On Sat, 17 Oct 1998, H.J. Lu wrote: >> >> > I have massive TCPv4 bad checksum on my CSLIP link from >> > my CSLIP server to my machine ever since I upgraded to >> > linux 2.1.1xx. Any suggestions? >> >> 2.0 just never spat out a message when it got a bad checksum, so this is >> probably why you never saw it before. If you're noticing that there's an >> actual packet data problem that didn't exist before with 2.0, then you >> might want to investigate. The message will be probably be removed by >> 2.2. >>
> I can tell my SLIP speed is slower than 2.0 and there are quite > some corrupted packets come over SLIP link.
Some broken modem server have buggy VJ compression functions that can't cope with the new RFC1323/SACK TCP options Linux 2.1 sends. Try turning them off with
echo 0 >/proc/sys/net/ipv4/tcp_sack echo 0 >/proc/sys/net/ipv4/tcp_timestamps echo 0 >/proc/sys/net/ipv4/tcp_window_scaling
Will only take effect for new TCP connections afterward.
BTW, you can check for checksum errors even on 2.0 with a newer nettools and netstat -s or by directly looking at /proc/net/snmp.
-Andi
- 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/
| |