Net.ether905 problem

Im having trouble configuring my system with a 3com 905B ethernet card on
qnx425 patch D (tried E as well).

nettrap query returns

Net &

Net.ether905 -pd800 -i255 -l1 &
netmap -f
<<

If I use “nettrap start” I get a screen dump and the system is locked solid.

I tried setting the IRQ manually to 10 in my bios and then started the
driver like…
<<
Net &
Net.ether905 -i10 -l1 -v &
netmap -f
<<

The verbose data from net ether905 reports…
irq 10
port d800 (same as nettrap query)
mac address looks good, no errors

Now Im off to start my TCP stack
<<
Socklet hostname & (hostname defined in /etc/hosts as 129.168.0.11)
dhcp.client hostname &

The dhcp client reaches my dhcp server and populates the /etc/resolv.conf
file with the correct nameservers.

next I try
ping localhost
or
ping qnx.com
or
ping hostname

all iterations block until I hit Ctrl^C

It seems to me that if the dhcp.client can reach my dhcp server that
Net.ether905 is working, is this correct ?
I also tried configuring the socket manually but still get the same effect.

Thanks in advance
John

OK, I reset the BIOS to defaults and now “nettrap query” matches show_pci in
regards to irq and IO.

now when I try to ping I get “No route to host”

my resolv.conf file looks like…
lookup file bind
nameserver xxx.xxx.xxx.xxx
nameserver xxx.xxx.xxx.xxx

and my host looks like…
127.1 localhost
192.168.0.11 my_hostname

any ideas ?

Thanks again
John

“John Love” <john@tolltex.com> wrote in message
news:a3s55s$7q4$1@inn.qnx.com

Im having trouble configuring my system with a 3com 905B ethernet card on
qnx425 patch D (tried E as well).

nettrap query returns

Net &
Net.ether905 -pd800 -i255 -l1 &
netmap -f


If I use “nettrap start” I get a screen dump and the system is locked
solid.

I tried setting the IRQ manually to 10 in my bios and then started the
driver like…

Net &
Net.ether905 -i10 -l1 -v &
netmap -f


The verbose data from net ether905 reports…
irq 10
port d800 (same as nettrap query)
mac address looks good, no errors

Now Im off to start my TCP stack

Socklet hostname & (hostname defined in /etc/hosts as 129.168.0.11)
dhcp.client hostname &

The dhcp client reaches my dhcp server and populates the /etc/resolv.conf
file with the correct nameservers.

next I try
ping localhost
or
ping qnx.com
or
ping hostname

all iterations block until I hit Ctrl^C

It seems to me that if the dhcp.client can reach my dhcp server that
Net.ether905 is working, is this correct ?
I also tried configuring the socket manually but still get the same
effect.

Thanks in advance
John

If you are trying to ping an IP address on the same subnet, it should
work, but if the destination is elsewhere or you need to access a DNS
server on another net, you need to define a gateway
by:
/usr/ucb/route add default xxx.xxx.xxx.xxx

You may also want to add “lookup file bind” to the end of resolv.conf.

Richard

John Love wrote:

OK, I reset the BIOS to defaults and now “nettrap query” matches show_pci in
regards to irq and IO.

now when I try to ping I get “No route to host”

my resolv.conf file looks like…
lookup file bind
nameserver xxx.xxx.xxx.xxx
nameserver xxx.xxx.xxx.xxx

and my host looks like…
127.1 localhost
192.168.0.11 my_hostname

any ideas ?

Thanks again
John

“John Love” <> john@tolltex.com> > wrote in message
news:a3s55s$7q4$> 1@inn.qnx.com> …

Im having trouble configuring my system with a 3com 905B ethernet card on
qnx425 patch D (tried E as well).

nettrap query returns

Net &
Net.ether905 -pd800 -i255 -l1 &
netmap -f


If I use “nettrap start” I get a screen dump and the system is locked

solid.

I tried setting the IRQ manually to 10 in my bios and then started the
driver like…

Net &
Net.ether905 -i10 -l1 -v &
netmap -f


The verbose data from net ether905 reports…
irq 10
port d800 (same as nettrap query)
mac address looks good, no errors

Now Im off to start my TCP stack

Socklet hostname & (hostname defined in /etc/hosts as 129.168.0.11)
dhcp.client hostname &

The dhcp client reaches my dhcp server and populates the /etc/resolv.conf
file with the correct nameservers.

next I try
ping localhost
or
ping qnx.com
or
ping hostname

all iterations block until I hit Ctrl^C

It seems to me that if the dhcp.client can reach my dhcp server that
Net.ether905 is working, is this correct ?
I also tried configuring the socket manually but still get the same

effect.

Thanks in advance
John


\

Thanks for the response Richard.

I was using both of those suggestions. I downloaded the latest tcprt update
from quics and that fixed most of my problems. I still can’t ping “some”
domains, qnx.com comes to mind. I can ping locally and some domains like
tolltex.com (our website hosting server). Whats really strange, is I can
ftp, telnet and web browse qnx.com, I just cant ping them.

For now my system is working well enough for me to get my work done, but I’m
still interested in the problem and what if any effect it has on my system.
I’m not real clear on how ping resolves its network address, or why it can’t
reach qnx.com, but I’m open to suggestions.

UPDATE: I just tried pinging qnx.com from work on a windows box and can’t
get through there either. The name is resolved to a net address, then the
requests just time out.

Thanks
John Love
Tolltex.com


“Richard R. Kramer” <rrkramer@kramer-smilko.com> wrote in message
news:3C61CA62.9060002@kramer-smilko.com

If you are trying to ping an IP address on the same subnet, it should
work, but if the destination is elsewhere or you need to access a DNS
server on another net, you need to define a gateway
by:
/usr/ucb/route add default xxx.xxx.xxx.xxx

You may also want to add “lookup file bind” to the end of resolv.conf.

Richard

John Love wrote:

OK, I reset the BIOS to defaults and now “nettrap query” matches
show_pci in
regards to irq and IO.

now when I try to ping I get “No route to host”

my resolv.conf file looks like…
lookup file bind
nameserver xxx.xxx.xxx.xxx
nameserver xxx.xxx.xxx.xxx

and my host looks like…
127.1 localhost
192.168.0.11 my_hostname

any ideas ?

Thanks again
John

“John Love” <> john@tolltex.com> > wrote in message
news:a3s55s$7q4$> 1@inn.qnx.com> …

Im having trouble configuring my system with a 3com 905B ethernet card
on
qnx425 patch D (tried E as well).

nettrap query returns

Net &
Net.ether905 -pd800 -i255 -l1 &
netmap -f


If I use “nettrap start” I get a screen dump and the system is locked

solid.

I tried setting the IRQ manually to 10 in my bios and then started the
driver like…

Net &
Net.ether905 -i10 -l1 -v &
netmap -f


The verbose data from net ether905 reports…
irq 10
port d800 (same as nettrap query)
mac address looks good, no errors

Now Im off to start my TCP stack

Socklet hostname & (hostname defined in /etc/hosts as 129.168.0.11)
dhcp.client hostname &

The dhcp client reaches my dhcp server and populates the
/etc/resolv.conf
file with the correct nameservers.

next I try
ping localhost
or
ping qnx.com
or
ping hostname

all iterations block until I hit Ctrl^C

It seems to me that if the dhcp.client can reach my dhcp server that
Net.ether905 is working, is this correct ?
I also tried configuring the socket manually but still get the same

effect.

Thanks in advance
John



\

Right - many/most routers and firewalls can be configured to ignore ping
or traceroute packets - a security thing.

Richard

John wrote:

Thanks for the response Richard.

I was using both of those suggestions. I downloaded the latest tcprt update
from quics and that fixed most of my problems. I still can’t ping “some”
domains, qnx.com comes to mind. I can ping locally and some domains like
tolltex.com (our website hosting server). Whats really strange, is I can
ftp, telnet and web browse qnx.com, I just cant ping them.

For now my system is working well enough for me to get my work done, but I’m
still interested in the problem and what if any effect it has on my system.
I’m not real clear on how ping resolves its network address, or why it can’t
reach qnx.com, but I’m open to suggestions.

UPDATE: I just tried pinging qnx.com from work on a windows box and can’t
get through there either. The name is resolved to a net address, then the
requests just time out.

Thanks
John Love
Tolltex.com


“Richard R. Kramer” <> rrkramer@kramer-smilko.com> > wrote in message
news:> 3C61CA62.9060002@kramer-smilko.com> …

If you are trying to ping an IP address on the same subnet, it should
work, but if the destination is elsewhere or you need to access a DNS
server on another net, you need to define a gateway
by:
/usr/ucb/route add default xxx.xxx.xxx.xxx

You may also want to add “lookup file bind” to the end of resolv.conf.

Richard

John Love wrote:


OK, I reset the BIOS to defaults and now “nettrap query” matches

show_pci in

regards to irq and IO.

now when I try to ping I get “No route to host”

my resolv.conf file looks like…
lookup file bind
nameserver xxx.xxx.xxx.xxx
nameserver xxx.xxx.xxx.xxx

and my host looks like…
127.1 localhost
192.168.0.11 my_hostname

any ideas ?

Thanks again
John

“John Love” <> john@tolltex.com> > wrote in message
news:a3s55s$7q4$> 1@inn.qnx.com> …


Im having trouble configuring my system with a 3com 905B ethernet card

on

qnx425 patch D (tried E as well).

nettrap query returns

Net &
Net.ether905 -pd800 -i255 -l1 &
netmap -f


If I use “nettrap start” I get a screen dump and the system is locked


solid.


I tried setting the IRQ manually to 10 in my bios and then started the
driver like…

Net &
Net.ether905 -i10 -l1 -v &
netmap -f


The verbose data from net ether905 reports…
irq 10
port d800 (same as nettrap query)
mac address looks good, no errors

Now Im off to start my TCP stack

Socklet hostname & (hostname defined in /etc/hosts as 129.168.0.11)
dhcp.client hostname &

The dhcp client reaches my dhcp server and populates the

/etc/resolv.conf

file with the correct nameservers.

next I try
ping localhost
or
ping qnx.com
or
ping hostname

all iterations block until I hit Ctrl^C

It seems to me that if the dhcp.client can reach my dhcp server that
Net.ether905 is working, is this correct ?
I also tried configuring the socket manually but still get the same


effect.


Thanks in advance
John




\