Qnet protocol messages

QNX Newsgroups (Archives) qnx.rtos
1

Hi,

Here is output from sloginfo utility:
…
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
…

As sloginfo reports there is an issue with data transfer.
How can I solve this problem?

Regards,
Jacek

2

Does anybody know where the source of these messages come from?

Standard & Tiny TCP stack works fine with my network card.
I can start telnet session, ping the target and work with it by
using IDE (through qconn).

Qnet protocol works too, but there are two issues:

  1. error message:
    Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
    problem: had to memove data 2 bytes

  2. I can list workstation filesystem from my target
    #ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

After this any command typed on the target
(i.e. #pidin) hangs the board.

Very interesting is that with slogger running
on the target ls command works in both ways.

I’m using QNX 6.3.0, my target is PXA255.

Please help,
Jacek

Hi,

Here is output from sloginfo utility:
…
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
…

As sloginfo reports there is an issue with data transfer.
How can I solve this problem?

Regards,
Jacek

3

Is there any chance for solve this problem?

I’m waiting for your feedback and I’m totally
for your disposition.

Jacek

U¿ytkownik “Jacek Rudnicki” <jacek.rudnicki@quantum.com.pl> napisa³ w
wiadomo¶ci news:dadbqs$fpf$1@inn.qnx.com…

Does anybody know where the source of these messages come from?

Standard & Tiny TCP stack works fine with my network card.
I can start telnet session, ping the target and work with it by
using IDE (through qconn).

Qnet protocol works too, but there are two issues:

  1. error message:
    Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
    problem: had to memove data 2 bytes

  2. I can list workstation filesystem from my target
    #ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

After this any command typed on the target
(i.e. #pidin) hangs the board.

Very interesting is that with slogger running
on the target ls command works in both ways.

I’m using QNX 6.3.0, my target is PXA255.

Please help,
Jacek

Hi,

Here is output from sloginfo utility:
…
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
…

As sloginfo reports there is an issue with data transfer.
How can I solve this problem?

Regards,
Jacek

\

4

I did two more tests:

  1. Target running 6.3 + qnet protocol from 6.2.1 release
    connected with workstation running 6.2.1.

The ls command works perfect in both directions:
#ls /net/target
#ls /net/workstation

and the sloginfo doesn’t report any errors:
#sloginfo
Time Sev Major Minor Args

  1. Target running 6.3 + npm-qnet-compat.so protocol
    connected with workstation running 6.2.1

Both machines are visible under /net directory only
on the target. When I type #ls /net/workstation
then system prints something like this:
“… No route to host…”

Here is sloginfo outputs:
Jul 06 10:58:16 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

Jul 06 10:58:17 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

Jul 06 10:58:39 7 15 0 npm-qnet(L4): lr_resolve(): starting for nd 4
toshiba.www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 1 looking for nd 4 toshiba.www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 2 looking for nd 4 toshiba.www.quantum.com.pl

Jul 06 10:58:40 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet(): 2
retries exceeded for nd 4 toshiba.www.quantum.com.pl found 0 interfaces

Jul 06 10:58:46 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

Jul 06 10:58:48 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

It seems that npm-qnet.so and npm-qnet-compat.so v6.3 is broken.
Can anybody comment this?

Jacek

U¿ytkownik “Jacek Rudnicki” <jacek.rudnicki@quantum.com.pl> napisa³ w
wiadomo¶ci news:dah4da$d5s$1@inn.qnx.com…

Is there any chance for solve this problem?

I’m waiting for your feedback and I’m totally
for your disposition.

Jacek

U¿ytkownik “Jacek Rudnicki” <> jacek.rudnicki@quantum.com.pl> > napisa³ w
wiadomo¶ci news:dadbqs$fpf$> 1@inn.qnx.com> …
Does anybody know where the source of these messages come from?

Standard & Tiny TCP stack works fine with my network card.
I can start telnet session, ping the target and work with it by
using IDE (through qconn).

Qnet protocol works too, but there are two issues:

  1. error message:
    Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():
    align
    problem: had to memove data 2 bytes

  2. I can list workstation filesystem from my target
    #ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

After this any command typed on the target
(i.e. #pidin) hangs the board.

Very interesting is that with slogger running
on the target ls command works in both ways.

I’m using QNX 6.3.0, my target is PXA255.

Please help,
Jacek

Hi,

Here is output from sloginfo utility:
…
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
…

As sloginfo reports there is an issue with data transfer.
How can I solve this problem?

Regards,
Jacek



\

5

Jacek Rudnicki <jacek.rudnicki@quantum.com.pl> wrote:

[snip]

  1. Target running 6.3 + npm-qnet-compat.so protocol
    connected with workstation running 6.2.1

Are you sure the messages below is comming out of a
npm-qnet-compat.so? These messages only exsit in
npm-qnet-l4_lite.so module.

What is your exactly goal? You want your 6.3 target
talking to a 6.2.1 host? In this case, you HAVE TO
use npm-qnet-compat.so, cause the npm-qnet.so (which
is linked to npm-qnet-l4_lite.so) is not compatible
with 6.2.1 QNET.

-Xiaodan Tang
ES, QSS

Both machines are visible under /net directory only
on the target. When I type #ls /net/workstation
then system prints something like this:
“… No route to host…”

Here is sloginfo outputs:
Jul 06 10:58:16 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

Jul 06 10:58:17 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

Jul 06 10:58:39 7 15 0 npm-qnet(L4): lr_resolve(): starting for nd 4
toshiba.> www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 1 looking for nd 4 toshiba.> www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 2 looking for nd 4 toshiba.> www.quantum.com.pl

Jul 06 10:58:40 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet(): 2
retries exceeded for nd 4 toshiba.> www.quantum.com.pl > found 0 interfaces

Jul 06 10:58:46 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

Jul 06 10:58:48 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem: had to
memove data 2 bytes

It seems that npm-qnet.so and npm-qnet-compat.so v6.3 is broken.
Can anybody comment this?

Jacek

U?ytkownik “Jacek Rudnicki” <> jacek.rudnicki@quantum.com.pl> > napisa? w
wiadomo?ci news:dah4da$d5s$> 1@inn.qnx.com> …
Is there any chance for solve this problem?

I’m waiting for your feedback and I’m totally
for your disposition.

Jacek

U?ytkownik “Jacek Rudnicki” <> jacek.rudnicki@quantum.com.pl> > napisa? w
wiadomo?ci news:dadbqs$fpf$> 1@inn.qnx.com> …
Does anybody know where the source of these messages come from?

Standard & Tiny TCP stack works fine with my network card.
I can start telnet session, ping the target and work with it by
using IDE (through qconn).

Qnet protocol works too, but there are two issues:

  1. error message:
    Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():
    align
    problem: had to memove data 2 bytes

  2. I can list workstation filesystem from my target
    #ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

After this any command typed on the target
(i.e. #pidin) hangs the board.

Very interesting is that with slogger running
on the target ls command works in both ways.

I’m using QNX 6.3.0, my target is PXA255.

Please help,
Jacek

Hi,

Here is output from sloginfo utility:
…
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2 bytes
…

As sloginfo reports there is an issue with data transfer.
How can I solve this problem?

Regards,
Jacek






\

6

Are you sure the messages below is comming out of a
npm-qnet-compat.so? These messages only exsit in
npm-qnet-l4_lite.so module.

Yes, these messages were captured when npm-qnet-compat.so
protocol was mounted by io-net.

What is your exactly goal?

My goal is to get npm-qnet properly working with my target and
network card.

First question where the source of this error:
"Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():
align problem: had to memove data 2 bytes "
come from?

What could be a reason?
I don’t know what to do next while TCP/IP (standard & tiny)
stack works perfect.

I need Your (qnet expert) and probably someone else help.

When we solve problem with qnet then qnet-compat issue
will be solved too.

Thank you,
Jacek


You want your 6.3 target
talking to a 6.2.1 host? In this case, you HAVE TO
use npm-qnet-compat.so, cause the npm-qnet.so (which
is linked to npm-qnet-l4_lite.so) is not compatible
with 6.2.1 QNET.

-Xiaodan Tang
ES, QSS

Both machines are visible under /net directory only
on the target. When I type #ls /net/workstation
then system prints something like this:
“… No route to host…”

Here is sloginfo outputs:
Jul 06 10:58:16 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:
had to
memove data 2 bytes

Jul 06 10:58:17 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:
had to
memove data 2 bytes

Jul 06 10:58:39 7 15 0 npm-qnet(L4): lr_resolve(): starting for nd 4
toshiba.> www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 1 looking for nd 4 toshiba.> www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 2 looking for nd 4 toshiba.> www.quantum.com.pl

Jul 06 10:58:40 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet(): 2
retries exceeded for nd 4 toshiba.> www.quantum.com.pl > found 0 interfaces

Jul 06 10:58:46 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:
had to
memove data 2 bytes

Jul 06 10:58:48 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:
had to
memove data 2 bytes

It seems that npm-qnet.so and npm-qnet-compat.so v6.3 is broken.
Can anybody comment this?

Jacek

U?ytkownik “Jacek Rudnicki” <> jacek.rudnicki@quantum.com.pl> > napisa? w
wiadomo?ci news:dah4da$d5s$> 1@inn.qnx.com> …
Is there any chance for solve this problem?

I’m waiting for your feedback and I’m totally
for your disposition.

Jacek

U?ytkownik “Jacek Rudnicki” <> jacek.rudnicki@quantum.com.pl> > napisa? w
wiadomo?ci news:dadbqs$fpf$> 1@inn.qnx.com> …
Does anybody know where the source of these messages come from?

Standard & Tiny TCP stack works fine with my network card.
I can start telnet session, ping the target and work with it by
using IDE (through qconn).

Qnet protocol works too, but there are two issues:

  1. error message:
    Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():
    align
    problem: had to memove data 2 bytes

  2. I can list workstation filesystem from my target
    #ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

After this any command typed on the target
(i.e. #pidin) hangs the board.

Very interesting is that with slogger running
on the target ls command works in both ways.

I’m using QNX 6.3.0, my target is PXA255.

Please help,
Jacek

Hi,

Here is output from sloginfo utility:
…
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():
align
problem: had to memove data 2 bytes
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():
align
problem: had to memove data 2 bytes
…

As sloginfo reports there is an issue with data transfer.
How can I solve this problem?

Regards,
Jacek






\

7

jacko wrote:

Are you sure the messages below is comming out of a
npm-qnet-compat.so? These messages only exsit in
npm-qnet-l4_lite.so module.


Yes, these messages were captured when npm-qnet-compat.so
protocol was mounted by io-net.

As Xiaodan stated, these message are only in npm-qnet-compat.so - please post
the method you are using to mount qnet.

Try

mount -Tio-net /usr/qnx630/target/qnx6/armle/lib/dll/npm-qnet-compat.so

What is your exactly goal?


My goal is to get npm-qnet properly working with my target and
network card.

First question where the source of this error:
"Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():
align problem: had to memove data 2 bytes "
come from?

What could be a reason?
I don’t know what to do next while TCP/IP (standard & tiny)
stack works perfect.

I need Your (qnet expert) and probably someone else help.

When we solve problem with qnet then qnet-compat issue
will be solved too.

Thank you,
Jacek



You want your 6.3 target
talking to a 6.2.1 host? In this case, you HAVE TO
use npm-qnet-compat.so, cause the npm-qnet.so (which
is linked to npm-qnet-l4_lite.so) is not compatible
with 6.2.1 QNET.

-Xiaodan Tang
ES, QSS


Both machines are visible under /net directory only
on the target. When I type #ls /net/workstation
then system prints something like this:
“… No route to host…”

Here is sloginfo outputs:
Jul 06 10:58:16 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:

had to

memove data 2 bytes

Jul 06 10:58:17 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:

had to

memove data 2 bytes

Jul 06 10:58:39 7 15 0 npm-qnet(L4): lr_resolve(): starting for nd 4
toshiba.> www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 1 looking for nd 4 toshiba.> www.quantum.com.pl

Jul 06 10:58:39 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet():
timeout: re-tx 2 looking for nd 4 toshiba.> www.quantum.com.pl

Jul 06 10:58:40 7 15 0 npm-qnet(L4): l4_resolve_periodic_en_ionet(): 2
retries exceeded for nd 4 toshiba.> www.quantum.com.pl > found 0 interfaces

Jul 06 10:58:46 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:

had to

memove data 2 bytes

Jul 06 10:58:48 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align problem:

had to

memove data 2 bytes

It seems that npm-qnet.so and npm-qnet-compat.so v6.3 is broken.
Can anybody comment this?

Jacek

U?ytkownik “Jacek Rudnicki” <> jacek.rudnicki@quantum.com.pl> > napisa? w
wiadomo?ci news:dah4da$d5s$> 1@inn.qnx.com> …

Is there any chance for solve this problem?

I’m waiting for your feedback and I’m totally
for your disposition.

Jacek

U?ytkownik “Jacek Rudnicki” <> jacek.rudnicki@quantum.com.pl> > napisa? w
wiadomo?ci news:dadbqs$fpf$> 1@inn.qnx.com> …

Does anybody know where the source of these messages come from?

Standard & Tiny TCP stack works fine with my network card.
I can start telnet session, ping the target and work with it by
using IDE (through qconn).

Qnet protocol works too, but there are two issues:

  1. error message:
    Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():

align

problem: had to memove data 2 bytes

  1. I can list workstation filesystem from my target
    #ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

After this any command typed on the target
(i.e. #pidin) hangs the board.

Very interesting is that with slogger running
on the target ls command works in both ways.

I’m using QNX 6.3.0, my target is PXA255.

Please help,
Jacek


Hi,

Here is output from sloginfo utility:
…
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():

align

problem: had to memove data 2 bytes
Feb 01 00:00:28 7 15 0 npm-qnet(L4): en_ionet_rx_up():

align

problem: had to memove data 2 bytes
…

As sloginfo reports there is an issue with data transfer.
How can I solve this problem?

Regards,
Jacek







\

–
cburgess@qnx.com

8

As Xiaodan stated, these message are only in npm-qnet-compat.so - please
post
the method you are using to mount qnet.

Try

mount -Tio-net /usr/qnx630/target/qnx6/armle/lib/dll/npm-qnet-compat.so

Hi Colin,

Please leave npm-qnet-compat.so at the moment
and go back to the second post of this thread
(npm-qnet.so issue).

Here is description of my problem:
…
I can list workstation filesystem from my target
#ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

after this any command typed on the target
(i.e. pidin) breaks a terminal session.

Very interesting is that with slogger running
on the target ls command works in both ways
without any problem.
…

What does the slogger realy do during qnet activity?


Regards,
Jacek

9

I have no idea - probably just changes the timing of things…

Jacek Rudnicki wrote:

As Xiaodan stated, these message are only in npm-qnet-compat.so - please

post

the method you are using to mount qnet.

Try

mount -Tio-net /usr/qnx630/target/qnx6/armle/lib/dll/npm-qnet-compat.so


Hi Colin,

Please leave npm-qnet-compat.so at the moment
and go back to the second post of this thread
(npm-qnet.so issue).

Here is description of my problem:
…
I can list workstation filesystem from my target
#ls /net/workstation

If I try to do the same from workstation
#ls /net/target

then system prints:
ls: readdir of ‘net/target’ failed (Host is down)

after this any command typed on the target
(i.e. pidin) breaks a terminal session.

Very interesting is that with slogger running
on the target ls command works in both ways
without any problem.
…

What does the slogger realy do during qnet activity?


Regards,
Jacek

\

–
cburgess@qnx.com

10

I have no idea - probably just changes the timing of things…

Ok, so what the next step should be?
How can I solve this problem?

Network driver modification …?

Jacek

11

I am have this messages in logs too.
i am use qnx6.3 and npm-qnet-l4_lite.so and ARM(le) cpu

io-net started like:

io-net -pqnet -dmy_driver

npm-qnet.so links on npm-qnet-l4_lite.so

if i am not run slogger on target then targed hangs then i try acces
from host-pc to target over qnet.
in slogger qnet spam this messages:
Jul 25 15:09:36 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2
bytes
Jul 25 15:09:36 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2
bytes
Jul 25 15:09:36 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2
bytes
Jul 25 15:09:36 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2
bytes
Jul 25 15:09:36 7 15 0 npm-qnet(L4): en_ionet_rx_up(): align
problem: had to memove data 2
bytes

TCP\IP works fine!