Sunha Choi
06/24/2008 8:39 AM
post9559
|
Hi.
The platform which I’d like to use io-pkt-v4-hc on is Lite5200b.
My goal is to make the platform working as an Access Point connecting wireless clients to the server on the other side.
laptop A(Server) <-> ethernet <-> MPC5200B <-> wireless <-> laptop B(Client)
I built a binary from the latest version of source code(ver. 409) and didn’t see any problem up to the point.
io-pkt and related network driver(mpc5200) have been successfully loaded using the script as attached below.
When I perform UDP test by running “iperf” on both laptop A and B as a server and client respectively to get an idea
of network bandwidth, the problem comes up. At the beginning, the UDP test seems alright, but it couldn’t last any
longer than 1000sec. When the test is interrupted, the error message shows up and io-pkt-v4-hc is unloaded automatically
. After this happens, io-pkt-v4-hc can be loaded again manually by command.
Process 8204 (io-pkt-v4-hc) terminated SIGSEGV code=1 fltno=11 ip=480d467c ref=00000000
In either case of performing TCP test using the same tool “iperf” or replacing the Ethernet connection between laptop
A and MPC5200B with wireless connection, the error that I described above doesn’t occur.
The following is the script that I use for running io-pkt-v4-hc
io-pkt-v4-hc random -dath -dmpc5200 mac=001122334455 -ptcpip
waitfor /dev/io-net/en0 10
ifconfig en0 192.168.1.222 netmask 255.255.255.0 up // ethernet
ifconfig ath0 192.168.0.111 netmask 255.255.255.0 up // artheros ar5413 wireless lan
ifconfig ath0 media OFDM54 mode 11g mediaopt hostap chan 1
ifconfig ath0 ssid MGT5200
sysctl -w net.inet.ip.forwarding=1
Thanks & Regards
-------------------------------------------------------------------
# ifconfig
lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192
inet 127.0.0.1 netmask 0xff000000
ath0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
ssid MGT5200 apbridge
powersave off
bssid 00:03:7f:07:84:bc chan 1
address: 00:03:7f:07:84:bc
media: IEEE802.11 OFDM54 mode 11g hostap
status: active
inet 192.168.0.111 netmask 0xffffff00 broadcast 192.168.0.255
en0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
address: 00:11:22:33:44:44
inet 192.168.1.222 netmask 0xffffff00 broadcast 192.168.1.255
#
#
# pidin
pid tid name prio STATE Blocked
1 1 procnto 0f READY
1 2 procnto 255r RECEIVE 1
1 3 procnto 255r RECEIVE 1
1 4 procnto 10r RECEIVE 1
1 5 procnto 10r RUNNING
1 6 procnto 10r RECEIVE 1
1 7 procnto 10r RECEIVE 1
4098 1 c/boot/devc-serpsc 10r RECEIVE 1
4099 1 proc/boot/slogger 21r RECEIVE 1
4100 1 proc/boot/pipe 10r SIGWAITINFO
4100 2 proc/boot/pipe 10r RECEIVE 1
4100 3 proc/boot/pipe 10r RECEIVE 1
4101 1 c/boot/pci-mgt5200 21r RECEIVE 1
4102 1 proc/boot/devc-pty 10r RECEIVE 1
4103 1 proc/boot/qconn 10r SIGWAITINFO
4103 2 proc/boot/qconn 10r CONDVAR 48060168
4103 3 proc/boot/qconn 10r RECEIVE 1
4103 4 proc/boot/qconn 10r RECEIVE 3
4104 1 /boot/devf-mgt5200 10r SIGWAITINFO
4104 2 /boot/devf-mgt5200 10r RECEIVE 1
4106 1 proc/boot/io-usb 10r SIGWAITINFO
4106 2 proc/boot/io-usb 21r RECEIVE 4
4106 ...
View Full Message
|
|
|
Sean Boudreau(deleted)
06/24/2008 9:22 AM
post9563
|
On Tue, Jun 24, 2008 at 08:39:16AM -0400, Sunha Choi wrote:
> Hi.
>
> The platform which I???d like to use io-pkt-v4-hc on is Lite5200b.
> My goal is to make the platform working as an Access Point connecting wireless clients to the server on the other side
.
>
> laptop A(Server) <-> ethernet <-> MPC5200B <-> wireless <-> laptop B(Client)
>
> I built a binary from the latest version of source code(ver. 409) and didn???t see any problem up to the point.
> io-pkt and related network driver(mpc5200) have been successfully loaded using the script as attached below.
>
> When I perform UDP test by running ???iperf??? on both laptop A and B as a server and client respectively to get an
idea of network bandwidth, the problem comes up. At the beginning, the UDP test seems alright, but it couldn???t last
any longer than 1000sec. When the test is interrupted, the error message shows up and io-pkt-v4-hc is unloaded
automatically. After this happens, io-pkt-v4-hc can be loaded again manually by command.
>
> Process 8204 (io-pkt-v4-hc) terminated SIGSEGV code=1 fltno=11 ip=480d467c ref=00000000
So it's the wired ethernet that causes the issue? Can
you build debug variants, run with just the wired driver
and send me the resulting io-pkt-v4-hc, devn-mpc5200 and
io-pkt-v4-hc.core. Also send the libc.so.2 you're using.
# make CCOPTS=-O0 DEBUG=-g
< make sure 'dumper' is running on the target>
# dumper -d /dev/shmem
< run test >
< should be a /dev/shmem/io-pkt-v4-hc.core >
-seanb
|
|
|
Robert Craig
06/24/2008 10:03 AM
post9575
|
Another thing to check is to do a "pidin -p io-pkt-v4-hc mem" before the
crash and have a look at where things are loaded in memory. The
"ip=480d467c" can point to what part of the stack is crashing. This
isn't always useful since the driver can do "something" that causes
stack code to crash, but it's a good first step.
Robert.
-----Original Message-----
From: Sunha Choi [mailto:shchoi@lge.com]
Sent: Tuesday, June 24, 2008 8:39 AM
To: drivers-networking
Subject: io-pkt-v4-hc hang...
Hi.
The platform which I'd like to use io-pkt-v4-hc on is Lite5200b.
My goal is to make the platform working as an Access Point connecting
wireless clients to the server on the other side.
laptop A(Server) <-> ethernet <-> MPC5200B <-> wireless <-> laptop
B(Client)
I built a binary from the latest version of source code(ver. 409) and
didn't see any problem up to the point.
io-pkt and related network driver(mpc5200) have been successfully loaded
using the script as attached below.
When I perform UDP test by running "iperf" on both laptop A and B as a
server and client respectively to get an idea of network bandwidth, the
problem comes up. At the beginning, the UDP test seems alright, but it
couldn't last any longer than 1000sec. When the test is interrupted, the
error message shows up and io-pkt-v4-hc is unloaded automatically. After
this happens, io-pkt-v4-hc can be loaded again manually by command.
Process 8204 (io-pkt-v4-hc) terminated SIGSEGV code=1 fltno=11
ip=480d467c ref=00000000
In either case of performing TCP test using the same tool "iperf" or
replacing the Ethernet connection between laptop A and MPC5200B with
wireless connection, the error that I described above doesn't occur.
The following is the script that I use for running io-pkt-v4-hc
io-pkt-v4-hc random -dath -dmpc5200 mac=001122334455 -ptcpip
waitfor /dev/io-net/en0 10
ifconfig en0 192.168.1.222 netmask 255.255.255.0 up // ethernet
ifconfig ath0 192.168.0.111 netmask 255.255.255.0 up // artheros
ar5413 wireless lan
ifconfig ath0 media OFDM54 mode 11g mediaopt hostap chan 1
ifconfig ath0 ssid MGT5200
sysctl -w net.inet.ip.forwarding=1
Thanks & Regards
-------------------------------------------------------------------
# ifconfig
lo0: flags=8049<UP,LOOPBACK,RUNNING,MULTICAST> mtu 33192
inet 127.0.0.1 netmask 0xff000000
ath0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
ssid MGT5200 apbridge
powersave off
bssid 00:03:7f:07:84:bc chan 1
address: 00:03:7f:07:84:bc
media: IEEE802.11 OFDM54 mode 11g hostap
status: active
inet 192.168.0.111 netmask 0xffffff00 broadcast 192.168.0.255
en0: flags=8843<UP,BROADCAST,RUNNING,SIMPLEX,MULTICAST> mtu 1500
address: 00:11:22:33:44:44
inet 192.168.1.222 netmask 0xffffff00 broadcast 192.168.1.255
#
#
# pidin
pid tid name prio STATE Blocked
1 1 procnto 0f READY
1 2 procnto 255r RECEIVE 1
1 3 procnto 255r RECEIVE 1
1 4 procnto 10r RECEIVE 1
1 5 procnto 10r RUNNING
1 6 procnto 10r RECEIVE 1
1 7 procnto 10r RECEIVE 1
4098 1 c/boot/devc-serpsc 10r RECEIVE 1
4099 1 proc/boot/slogger 21r RECEIVE 1
4100 1 proc/boot/pipe 10r SIGWAITINFO
4100 2 proc/boot/pipe 10r RECEIVE 1
4100 3 proc/boot/pipe 10r RECEIVE 1
4101 1 c/boot/pci-mgt5200 21r RECEIVE 1
4102 1 proc/boot/devc-pty 10r RECEIVE 1
4103 1...
View Full Message
|
|
|
Sunha Choi
|
Re: RE: io-pkt-v4-hc hang...
|
Sunha Choi
07/02/2008 6:43 AM
post9897
|
Re: RE: io-pkt-v4-hc hang...
Hi.
Thank you for reply.
Sean, i built version 425 using 'CCOPTS=-O0 DEBUG=-g CPULIST=ppc'.
i send requested files(io-pkt-v4-hc, io-pkt-v4-hc.core, devn-mpc5200 and libc.so.2).
Robert, i done "pidin -p io-pkt-v4-hc mem" before the crash.
The result is following
----------------------------------------------------------------------------------------------
# pidin -p io-pkt-v4-hc mem
pid tid name prio STATE code data stack
8204 1 sbin/io-pkt-v4-hc 21r SIGWAITINFO 1604K 788K 8192(516K)*
8204 2 sbin/io-pkt-v4-hc 21r RECEIVE 1604K 788K 8192(132K)
8204 3 sbin/io-pkt-v4-hc 21r RECEIVE 1604K 788K 4096(132K)
libc.so.2 @fe300000 436K 32K
devnp-ath.so @fe378000 260K 80K
devnp-shim.so @fe3cd000 44K 8192
devn-mpc5200.so @fe3da000 36K 8192
-bestcomm5200.so.1 @fe3e5000 28K 12K
lsm-pf-v4.so @fe3ef000 256K 16K
sbin/io-pkt-v4-hc @4803e000 ( 0) 8192
sbin/io-pkt-v4-hc @481d7000 ( 0) 52K
sbin/io-pkt-v4-hc @fe375000 ( 0) 12K
#
< Iperf Test >
#
Process 8204 (io-pkt-v4-hc) terminated SIGSEGV code=1 fltno=11 ip=4812fc3c ref=0
0000000
#
----------------------------------------------------------------------------------------------
I used MPC5200B reference board from freescale with 3Com Wireless LAN card and ethernet. Reference board acts as a WAP.
laptop A(Server) <-> ethernet <-> MPC5200B Ref. Board <-> wireless <-> laptop B(Client)
Iperf Test // Iperf UDP Bandwidth Test
laptop A(Server)
> iperf -u -s
laptop B(Client)
> iperf -u -c Server IP -b 10m -i 1 -t 3600
0 - 1sec 1.14MByte 9.57Mbit/s
1 - 2sec 1.19MByte 10Mbit/s
...
...
crash
|
|
|
Sunha Choi
|
Re: RE: io-pkt-v4-hc hang...
|
Sunha Choi
07/02/2008 7:13 AM
post9898
|
Re: RE: io-pkt-v4-hc hang...
i'm sorry. i sent wrong file.
This attachment file is correct.
|
|
|
Sean Boudreau(deleted)
|
Re: RE: io-pkt-v4-hc hang...
|
Sean Boudreau(deleted)
07/02/2008 10:57 AM
post9938
|
Re: RE: io-pkt-v4-hc hang...
On Wed, Jul 02, 2008 at 06:44:45AM -0400, Sunha Choi wrote:
> Hi.
> Thank you for reply.
>
> Sean, i built version 425 using 'CCOPTS=-O0 DEBUG=-g CPULIST=ppc'.
> i send requested files(io-pkt-v4-hc, io-pkt-v4-hc.core, devn-mpc5200 and libc.so.2).
>
> Robert, i done "pidin -p io-pkt-v4-hc mem" before the crash.
> The result is following
> ----------------------------------------------------------------------------------------------
> # pidin -p io-pkt-v4-hc mem
> pid tid name prio STATE code data stack
> 8204 1 sbin/io-pkt-v4-hc 21r SIGWAITINFO 1604K 788K 8192(516K)*
> 8204 2 sbin/io-pkt-v4-hc 21r RECEIVE 1604K 788K 8192(132K)
> 8204 3 sbin/io-pkt-v4-hc 21r RECEIVE 1604K 788K 4096(132K)
> libc.so.2 @fe300000 436K 32K
> devnp-ath.so @fe378000 260K 80K
> devnp-shim.so @fe3cd000 44K 8192
> devn-mpc5200.so @fe3da000 36K 8192
> -bestcomm5200.so.1 @fe3e5000 28K 12K
> lsm-pf-v4.so @fe3ef000 256K 16K
> sbin/io-pkt-v4-hc @4803e000 ( 0) 8192
> sbin/io-pkt-v4-hc @481d7000 ( 0) 52K
> sbin/io-pkt-v4-hc @fe375000 ( 0) 12K
> #
> < Iperf Test >
> #
> Process 8204 (io-pkt-v4-hc) terminated SIGSEGV code=1 fltno=11 ip=4812fc3c ref=0
> 0000000
> #
It looks to have died while cleaning up the mpc5200 driver in response
to a SIGTERM. Do you say io-pkt after the iperf test? Can you also
send your devnp-shim.so and retry without loading the atheros and
lsm-pf dlls.
Thanks,
-seanb
|
|
|
|