Re: [ECOS] network test problem on ipaq

[Jonathan Larmour <jlarmour at redhat dot com>]

Weilong Li wrote:
> 
> The ip adddress of ipaq is 192.168.1.20. After redboot
> boots on ipaq, I issued tcp_source and tcp_sink
> command
> to target on my redhat host.
> 
> # ./ecos/packages/net/tcpip/current/tests/tcp_sink
> 192.168.1.20
> Start TCP test - SINK mode to 192.168.1.20
> 
> At another terminal:
> # ./ecos/packages/net/tcpip/current/tests/tcp_source
> 192.168.1.20
> Start TCP test - SOURCE mode to 192.168.1.20
> 
> And I pinged 192.168.1.20, and I got response.
> 
> Then I issued gdb command on tcp_echo. This time it
> can load, while immediately after I type "continue",
> and tcp_echo is running, tcp_sink and tcp_source
> will report "no route to 192.168.1.20", and
> I cannot ping ipaq from linux host.
> Here is the screen.
> #arm-elf-gdb
> ./ipaq/install/tests/net/tcpip/current/tests/tcp_echo
> GNU gdb 5.0
> Copyright 2000 Free Software Foundation, Inc.
> GDB is free software, covered by the GNU General
> Public License, and you are
> welcome to change it and/or distribute copies of it
> under certain conditions.
> Type "show copying" to see the conditions.
> There is absolutely no warranty for GDB.  Type "show
> warranty" for details.
> This GDB was configured as "--host=i586-pc-linux-gnu
> --target=arm-elf"...
> (gdb) set remotebaud 38400
> (gdb) target remote /dev/ttyS0
> Remote debugging using /dev/ttyS0
> 0x50046f30 in ?? ()
> (gdb) load
> Loading section .rom_vectors, size 0x40 lma 0x20000
> Loading section .text, size 0x2ec2c lma 0x20040
> Loading section .rodata, size 0x1d18 lma 0x4ec6c
> Loading section .data, size 0xc3c lma 0x50984
> Start address 0x20040 , load size 202176
> Transfer rate: 26956 bits/sec, 305 bytes/write.
> (gdb) continue
> Continuing.
> Start TCP test - ECHO mode
> No load = 723469
> Set background load = 50% starting 10 threads
> Set no background load
> High Load[100] = 686113 => 6%
> Set background load = 50% starting 10 threads
> Set no background load
> High Load[200] = 655463 => 10%
> Set background load = 50% starting 10 threads
> Set no background load
> High Load[400] = 593289 => 18%
> Set background load = 50% starting 10 threads
> Set no background load
> High Load[800] = 473787 => 35%
> Set background load = 50% starting 10 threads
> Set no background load
> High Load[1600] = 234626 => 68%
> Set background load = 50% starting 10 threads
> Set no background load
> Load[1200] = 354165 => 52%
> Set background load = 50% starting 10 threads
> Set no background load
> Final load[1153] = 368145 => 50%
> 
> Here are two screens from tcp_sink and tcp_source.
> #./ecos/packages/net/tcpip/current/tests/tcp_source
> 192.168.1.20
> Start TCP test - SOURCE mode to 192.168.1.20
> 
> Can't connect to target: No route to host
> 
> #./ecos/packages/net/tcpip/current/tests/tcp_sink
> 192.168.1.20
> Start TCP test - SOURCE mode to 192.168.1.20
> 
> Can't connect to target: No route to host
> 
> The problem I'm suspecious of is that: after tcp_echo
> is running through serial port, the ipaq seems stuck
> there and not responding to other connections (even
> from ethernet card interface.)
> But I don't know what is behind.

Firstly, are you making sure that tcp_echo is running _before_ starting
tcp_sink and tcp_source? Secondly when you talk about the IP address of the
ipaq, there are two to consider: RedBoot's and eCos's. Their addresses
should not be the same. The latter one is the relevant one for the net
test.

Jifl
-- 
Red Hat, Rustat House, Clifton Road, Cambridge, UK. Tel: +44 (1223) 271062
Maybe this world is another planet's Hell -Aldous Huxley || Opinions==mine
Come to the Red Hat TechWorld open source conference in Brussels!
Keynotes, techie talks and exhibitions    http://www.redhat-techworld.com/