Re: Bug in VJ compression

Marijn Meijles (marijn@stack.nl)
Sun, 8 Mar 1998 18:51:15 +0100

You wrote:
> Actually we do support the timestamp and windowscale options in 2.1.x
> But back to the original point, is this with 2.1.x or 2.0.x VJ
> compression code? As far as I can see when there are options present
> the code sends the packet out uncompressed.
>
> Or could you be a little more detailed what is exactly happening in
> this case where you see the problem?

I use 2.0.33. I have a tcpdump from my local FreeBSD box and a remote one.
The names don't match, but I didn't make a local dump when the remote one
was made and I'm not root on the remote box.
Local:
16:56:03.328555 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: S 150336389:150336389(0) win 16384 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF) [tos 0x10]
16:56:03.518877 toad.stack.nl.telnet > hoop.student.tue.nl.blackjack: S 3420661761:3420661761(0) ack 150336390 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
The following packet is lost, although a packet goes over the line.
16:56:03.519128 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: . ack 1 win 17280 <nop,nop,timestamp 1581 1359161,nop,nop,eol,[|tcp]> (DF) [tos 0x10]
16:56:03.520822 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: P 1:25(24) ack 1 win 17280 <nop,nop,timestamp 1581 1359161,nop,nop,eol,[|tcp]> (DF) [tos 0x10]
16:56:06.420179 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: P 1:25(24) ack 1 win 17280 <nop,nop,timestamp 1586 1359161,nop,nop,[|tcp]> (DF) [tos 0x10]
16:56:10.796600 toad.stack.nl.telnet > hoop.student.tue.nl.blackjack: S 3420661761:3420661761(0) ack 150336390 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
16:56:10.796813 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: . ack 1 win 17280 <nop,nop,timestamp 1595 1359172,nop,nop,eol,[|tcp]> (DF) [tos 0x10]
16:56:12.420174 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: P 1:25(24) ack 1 win 17280 <nop,nop,timestamp 1598 1359172,nop,nop,[|tcp]> (DF) [tos 0x10]
16:56:24.420193 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: P 1:25(24) ack 1 win 17280 <nop,nop,timestamp 1622 1359172,nop,nop,[|tcp]> (DF) [tos 0x10]
16:56:35.628880 toad.stack.nl.telnet > hoop.student.tue.nl.blackjack: S 3420661761:3420661761(0) ack 150336390 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
16:56:35.629094 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: . ack 1 win 17280 <nop,nop,timestamp 1645 1359220,nop,nop,eol,[|tcp]> (DF) [tos 0x10]
16:56:48.420203 hoop.student.tue.nl.blackjack > toad.stack.nl.telnet: P 1:25(24) ack 1 win 17280 <nop,nop,timestamp 1670 1359220,nop,nop,[|tcp]> (DF) [tos 0x10]
16:57:19.485353 toad.stack.nl.telnet > hoop.student.tue.nl.blackjack: R 1:1(0) ack 1 win 17280 (DF)

The remote one, also FreeBSD although the latest Solaris seems to support
them also:
14:08:51.730958 unox.student.tue.nl.61001 > svgewis01.win.tue.nl.telnet: S 2980040854:2980040854(0) win 16384 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF) [tos 0x10]
14:08:51.731169 svgewis01.win.tue.nl.telnet > unox.student.tue.nl.61001: S 1309645250:1309645250(0) ack 2980040855 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
14:08:54.680098 svgewis01.win.tue.nl.telnet > unox.student.tue.nl.61001: S 1309645250:1309645250(0) ack 2980040855 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
14:09:00.680080 svgewis01.win.tue.nl.telnet > unox.student.tue.nl.61001: S 1309645250:1309645250(0) ack 2980040855 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
14:09:12.680106 svgewis01.win.tue.nl.telnet > unox.student.tue.nl.61001: S 1309645250:1309645250(0) ack 2980040855 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
14:09:36.680106 svgewis01.win.tue.nl.telnet > unox.student.tue.nl.61001: S 1309645250:1309645250(0) ack 2980040855 win 17280 <mss 1460,nop,wscale 0,nop,nop,timestamp[|tcp]> (DF)
14:10:06.680095 svgewis01.win.tue.nl.telnet > unox.student.tue.nl.61001: R 1309645251:1309645251(0) ack 2980040855 win 17280 (DF)

-- 
Marijn

---
Getting up is merely a stage of going to bed




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