* [Bluez-users] How is pairing completed in rfcomm?
@ 2004-06-21 19:51 David Feldman
2004-06-21 21:08 ` Marcel Holtmann
0 siblings, 1 reply; 4+ messages in thread
From: David Feldman @ 2004-06-21 19:51 UTC (permalink / raw)
To: bluez-users; +Cc: wb0gaz
I'm working on getting my RH9 laptop to communicate with T68 (not i),
and I made
another step forward, but I can't complete the "pairing" process (
which is where the
phone picks up the identity of a PC and makes that PC "OK" for future
data operation.)
My steps:
insert card, do cardctl insert.
hciconfig hci0 up
rfcomm /dev/rfcomm0 MAC-ADDRESS 1 (1 is for DUN on this phone)
Now at this step I gave the phone the instruction to add to paired
devices, with
the option "phone accepts".
Then, I did this:
cat </dev/rfcomm0
at which point the phone displays the PC's name, then asks for a PIN.
I entered the PIN (a 4-digit code) that I had previously configured
Now the phone says "pairing..." and waits forever.
If I interrupt the cat </dev/rfcomm0 with Control-C,
then the following is seen:
1. Any output (echo, cat, etc.) to /dev/rfcomm0 produces the error
message:
"bash: /dev/rfcomm0: Operation already in progress"
2. The rfcomm command produces the message:
rfcomm0: MAC-ADDRESS channel 1 disconnecting
Marcel had advised me to update to the latest kernel - I am not in a
position to
do this immediately, and it I hope the pairing process would be able
to proceed
in any event, if I am using the correct procedure. I am using Kernel 2.
4.20-31.9,
which was current as of 04/2004; as the phone is nearly 2 years old,
I'm hoping
I can validate it's operation at least through the pairing process.
So my core question - how do I complete the pairing process, when
initiated
by the PC and accepted by the phone?
Dave
wb0gaz@hotmail.com
-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital self defense, top technical experts, no vendor pitches,
unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [Bluez-users] How is pairing completed in rfcomm?
2004-06-21 19:51 [Bluez-users] How is pairing completed in rfcomm? David Feldman
@ 2004-06-21 21:08 ` Marcel Holtmann
2004-06-22 15:20 ` David Feldman
2004-06-22 15:30 ` Solved pairing problem - big success! David Feldman
0 siblings, 2 replies; 4+ messages in thread
From: Marcel Holtmann @ 2004-06-21 21:08 UTC (permalink / raw)
To: David Feldman; +Cc: BlueZ Mailing List
Hi David,
> I'm working on getting my RH9 laptop to communicate with T68 (not i),
> and I made
> another step forward, but I can't complete the "pairing" process (
> which is where the
> phone picks up the identity of a PC and makes that PC "OK" for future
> data operation.)
>
> My steps:
>
> insert card, do cardctl insert.
> hciconfig hci0 up
> rfcomm /dev/rfcomm0 MAC-ADDRESS 1 (1 is for DUN on this phone)
it is "rfcomm bind 0 <bdaddr> [channel]".
> Now at this step I gave the phone the instruction to add to paired
> devices, with
> the option "phone accepts".
>
> Then, I did this:
>
> cat </dev/rfcomm0
>
> at which point the phone displays the PC's name, then asks for a PIN.
>
> I entered the PIN (a 4-digit code) that I had previously configured
>
> Now the phone says "pairing..." and waits forever.
>
> If I interrupt the cat </dev/rfcomm0 with Control-C,
>
> then the following is seen:
>
> 1. Any output (echo, cat, etc.) to /dev/rfcomm0 produces the error
> message:
> "bash: /dev/rfcomm0: Operation already in progress"
>
> 2. The rfcomm command produces the message:
> rfcomm0: MAC-ADDRESS channel 1 disconnecting
Until you upgraded to a newer kernel I can't guarantee that this will
work. Your RFCOMM code seems to be buggy for Bluetooth 1.0b phones.
> Marcel had advised me to update to the latest kernel - I am not in a
> position to
> do this immediately, and it I hope the pairing process would be able
> to proceed
> in any event, if I am using the correct procedure. I am using Kernel 2.
> 4.20-31.9,
> which was current as of 04/2004; as the phone is nearly 2 years old,
> I'm hoping
> I can validate it's operation at least through the pairing process.
The phone will work. I don't see any problem here because the Bluetooth
stack of the T68 and the T39m are identical and the T39m is working for
me.
> So my core question - how do I complete the pairing process, when
> initiated
> by the PC and accepted by the phone?
The phone uses security level 2 and this means the security routines
kick in when you request access to the RFCOMM channel. If you want them
earlier then the use of security mode 3 is needed. Use "hciconfig hci0
auth" to activate it on your device.
Regards
Marcel
-------------------------------------------------------
This SF.Net email sponsored by Black Hat Briefings & Training.
Attend Black Hat Briefings & Training, Las Vegas July 24-29 -
digital self defense, top technical experts, no vendor pitches,
unmatched networking opportunities. Visit www.blackhat.com
_______________________________________________
Bluez-users mailing list
Bluez-users@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-users
^ permalink raw reply [flat|nested] 4+ messages in thread
* Re: [Bluez-users] How is pairing completed in rfcomm?
2004-06-21 21:08 ` Marcel Holtmann
@ 2004-06-22 15:20 ` David Feldman
2004-06-22 15:30 ` Solved pairing problem - big success! David Feldman
1 sibling, 0 replies; 4+ messages in thread
From: David Feldman @ 2004-06-22 15:20 UTC (permalink / raw)
To: Marcel Holtmann; +Cc: BlueZ Mailing List, wb0gaz
Hello Marcel/List:
I have made more progress!
It is not working, but I think I am now close. I still cannot
complete pairing.
I did not realize that it was necessary to run "hcid" daemon... I
could not
find a man page... Then when I was beginning to use hcid it gave error
that it could not execute pin helper, so I fixed the path to
executable file
that generates my selected PIN code. When I manually execute the pin_
helper
file, it outputs 4 digits to the console, which is my test PIN. It is
executable.
Anyway, now I run hcid before my testing procedure.
At this point, here is my situation:
1. Prepare phone: Set bluetooth ON (not discoverable, by now I know
the
MAC address!), then choose Linked Devices->Add Device->Phone Accepts.
Phone says "Ready to Pair"
2. Prepare PC: bluetooth card inserted, then do "cardctl insert", and
verify
messages on /var/log/messages (hcid reports HCI dev 0 registered, hci
dev 0 up, then Starting security manager 0.)
3. Execute script:
hciconfig hci0 up
hciconfig hci0 auth 3 (per Marcel suggestion)
rfcomm bind 0 MAC-ADDRESS 1 (note I changed parameter #2 to 0,
not /dev/rfcomm0)
---> reports "rfcomm0: MAC-ADDRESS channel 1 clean"
cat </dev/rfcomm0
4. At this point the computer device name (from hcid.conf) appears on
the phone,
and the phone says "Add to paired devices?" I select Yes. I enter my
PIN digits
(as configured in pin helper script file) and phone says "Pairing
failed, Retry?"
5. The cat command reports: "Warning: null TTY for (d8:00) in tty_
fasync" and
"-bash: /dev/rfcomm0: Permission denied"
6. At this point I went back to /var/log/messages and I see these
messages in
the file:
kernel: application bug: hcid(PID) has SIGCHLD set to SIG_IGN but
calls wait()
kernel: (see the NOTES section of 'man 2 wait'). Workaround activated.
kernel: Warning: null TTY for (d8:00) in tty_fasync.
As an alternative, I tried using l2ping instead of cat </dev/rfcomm0
to get the
security routine to start. l2ping reported the error message "Can't
connect: Permission denied"
and /var/log/messages included two of the same kernel messages (
application bug
about wait) but not the message about null TTY.
My question: What should I do to start the security sequence?
Should I update hcid (I am now using version 2.2 as reported by hcid -?)
I hope I can make more progress without installing a new kernel,
because I searched
for RPM for later kernel in 2.4 series (I am now using 2.4.20-31.9)
but there is no
later version RPM, and my machine does not set up with compiler, so I
cannot try
to compile new kernel.
Dave
wb0gaz@hotmail.com
###
>> So my core question - how do I complete the pairing process, when
>> initiated
>> by the PC and accepted by the phone?
>
>The phone uses security level 2 and this means the security routines
>kick in when you request access to the RFCOMM channel. If you want them
>earlier then the use of security mode 3 is needed. Use "hciconfig hci0
>auth" to activate it on your device.
>
>Regards
>
>Marcel
^ permalink raw reply [flat|nested] 4+ messages in thread
* Solved pairing problem - big success!
2004-06-21 21:08 ` Marcel Holtmann
2004-06-22 15:20 ` David Feldman
@ 2004-06-22 15:30 ` David Feldman
1 sibling, 0 replies; 4+ messages in thread
From: David Feldman @ 2004-06-22 15:30 UTC (permalink / raw)
To: Marcel Holtmann; +Cc: BlueZ Mailing List, wb0gaz
Hello Marcel/List:
I sent my last message too soon - sorry for the noise on the list.
I did more reading, and discovered there was an error in my pin
helper file.
It appears the message generated must be PIN:xxxx, not just xxxx, so I
changed that.
This made pairing successful!
Then, I configured minicom to communicate with /dev/rfcomm0 at 19200
bps, and the phone bluetooth went on-line. I was able to manually
dial to
my home linux server phone number, and log in!
It is a very big success!!!
Thank you VERY MUCH for all of your fine help and advice, I now have
beginning bluetooth connection on my old Omnibook 800.
Soon, I will get ppp running, then I can access internet directly...
Dave
wb0gaz@hotmail.com
^ permalink raw reply [flat|nested] 4+ messages in thread
end of thread, other threads:[~2004-06-22 15:30 UTC | newest]
Thread overview: 4+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2004-06-21 19:51 [Bluez-users] How is pairing completed in rfcomm? David Feldman
2004-06-21 21:08 ` Marcel Holtmann
2004-06-22 15:20 ` David Feldman
2004-06-22 15:30 ` Solved pairing problem - big success! David Feldman
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox