Hi.
I'm prototyping an application that should eventually become a
Voip/Bluethooth gateway.
Among the various problems I'm facing (i.e. occasional bizzare errors on
SCO connections, or inability to open more than one SCO link through the
same BT dongle), the currently most serious one is the following.
After a while my program is running (repeatedly dialing a phone number,
opening and closing the audio link), the following error shows up
kernel: __tx_submit: hci0 tx submit failed urb c4c6b014 type 3 err -19
followed, a few seconds later, by an "Unable to handle kernel NULL
pointer..." message, which does not get logged.
Of course, I'm not asking what the cause is. I just need some suggestions
as to how to further investigate the problem and provide a more useful
report.
My system is Debian with kernel package 2.6.10-1-686.
Thanks,
Maxim
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Hi Maxim,
> I'm prototyping an application that should eventually become a
> Voip/Bluethooth gateway.
>
> Among the various problems I'm facing (i.e. occasional bizzare errors on
> SCO connections, or inability to open more than one SCO link through the
> same BT dongle), the currently most serious one is the following.
>
> After a while my program is running (repeatedly dialing a phone number,
> opening and closing the audio link), the following error shows up
>
> kernel: __tx_submit: hci0 tx submit failed urb c4c6b014 type 3 err -19
>
> followed, a few seconds later, by an "Unable to handle kernel NULL
> pointer..." message, which does not get logged.
>
> Of course, I'm not asking what the cause is. I just need some suggestions
> as to how to further investigate the problem and provide a more useful
> report.
>
> My system is Debian with kernel package 2.6.10-1-686.
build your own 2.6.11-rc5 kernel and use netconsole to get the full
oops.
Regards
Marcel
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Hi Marco,
> > The reason I'm using multiple dongles is that I need to control more than
> > one phone, and apparently I'm unable to open concurrent SCO links to
> > different phones through the same dongle.
>
> Well, in my opinion this is possible... I'm able to connect with two mobile
> phones to one laptop with one dongle at the same time. maybe you have a
> dongle that doesn't support multiple connections?
the SCO connections are different than the ACL links. The SCO links are
attached to a previous established ACL link and you can have three SCO
links attached to one ACL, but only three SCO max are allowed inside a
piconet. However BlueZ only supports one SCO connection per ACL link.
This is a stupid limitation, but so far nobody removed it.
Regards
Marcel
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Hi Max,
> > this is weird. The first one is related to sysfs and actually these
> > should all be fixed by now. The others are also not quite clear.
> >
> > How many dongles are you using at the same time? I saw hci2 and hci1 in
> > the logs. What does "hciconfig -a" say about them?
>
> Yes, I'm using three dongles. See:
>
> http://cassandra.comm2000.it/~maxim/pub/bluez-oops/hciconfig.txt
>
> The reason I'm using multiple dongles is that I need to control more than
> one phone, and apparently I'm unable to open concurrent SCO links to
> different phones through the same dongle.
>
> Theoretically, this should be possible, shouldn't it? Nevertheless, if I
> connect through different dongles (by bind'ing each socket to the
> appropriate dongle address), everything works more or less as expected
> (except for occasional oops, and some infrequent errors during SCO
> connect's). If I try to use a single dongle, either not binding the
> sockets, or binding them to BDADDR_ANY, I always get "Too many links" on
> the second SCO connection I try to open. Am I possibly doing something
> wrong?
>
> Besides, warnings very often get logged mentioning SCO handle numbers
> which look bogus to me:
>
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 65023
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 63231
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 65535
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 65280
>
> See here for more messages like the above:
>
> http://cassandra.comm2000.it/~maxim/pub/bluez-oops/log-0.txt
I think the firmware of the dongles is too old. What does lsusb shows
you? Maybe it is possible to update them.
Regards
Marcel
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Max M. wrote:
> On Tue, 1 Mar 2005, Marcel Holtmann wrote:
>
>>this is weird. The first one is related to sysfs and actually these
>>should all be fixed by now. The others are also not quite clear.
>>
>>How many dongles are you using at the same time? I saw hci2 and hci1 in
>>the logs. What does "hciconfig -a" say about them?
>
>
> Yes, I'm using three dongles. See:
>
> http://cassandra.comm2000.it/~maxim/pub/bluez-oops/hciconfig.txt
>
> The reason I'm using multiple dongles is that I need to control more than
> one phone, and apparently I'm unable to open concurrent SCO links to
> different phones through the same dongle.
Well, in my opinion this is possible... I'm able to connect with two mobile
phones to one laptop with one dongle at the same time. maybe you have a
dongle that doesn't support multiple connections?
if you try to etablish more than two connections at the same time, you have
to pay attention to the master/slave roles. But you said you can't even
connect to two mobile phones at a time, so you problem should be somewere else.
regards
Marco
> Theoretically, this should be possible, shouldn't it? Nevertheless, if I
> connect through different dongles (by bind'ing each socket to the
> appropriate dongle address), everything works more or less as expected
> (except for occasional oops, and some infrequent errors during SCO
> connect's). If I try to use a single dongle, either not binding the
> sockets, or binding them to BDADDR_ANY, I always get "Too many links" on
> the second SCO connection I try to open. Am I possibly doing something
> wrong?
>
> Besides, warnings very often get logged mentioning SCO handle numbers
> which look bogus to me:
>
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 65023
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 63231
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 65535
> hci_scodata_packet: hci0 SCO packet for unknown connection handle 65280
>
> See here for more messages like the above:
>
> http://cassandra.comm2000.it/~maxim/pub/bluez-oops/log-0.txt
>
> Max
>
>
>
>
>
> -------------------------------------------------------
> SF email is sponsored by - The IT Product Guide
> Read honest & candid reviews on hundreds of IT Products from real users.
> Discover which products truly live up to the hype. Start reading now.
> http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
> _______________________________________________
> Bluez-users mailing list
> [email protected]
> https://lists.sourceforge.net/lists/listinfo/bluez-users
>
>
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
On Tue, 1 Mar 2005, Marcel Holtmann wrote:
> this is weird. The first one is related to sysfs and actually these
> should all be fixed by now. The others are also not quite clear.
>
> How many dongles are you using at the same time? I saw hci2 and hci1 in
> the logs. What does "hciconfig -a" say about them?
Yes, I'm using three dongles. See:
http://cassandra.comm2000.it/~maxim/pub/bluez-oops/hciconfig.txt
The reason I'm using multiple dongles is that I need to control more than
one phone, and apparently I'm unable to open concurrent SCO links to
different phones through the same dongle.
Theoretically, this should be possible, shouldn't it? Nevertheless, if I
connect through different dongles (by bind'ing each socket to the
appropriate dongle address), everything works more or less as expected
(except for occasional oops, and some infrequent errors during SCO
connect's). If I try to use a single dongle, either not binding the
sockets, or binding them to BDADDR_ANY, I always get "Too many links" on
the second SCO connection I try to open. Am I possibly doing something
wrong?
Besides, warnings very often get logged mentioning SCO handle numbers
which look bogus to me:
hci_scodata_packet: hci0 SCO packet for unknown connection handle 65023
hci_scodata_packet: hci0 SCO packet for unknown connection handle 63231
hci_scodata_packet: hci0 SCO packet for unknown connection handle 65535
hci_scodata_packet: hci0 SCO packet for unknown connection handle 65280
See here for more messages like the above:
http://cassandra.comm2000.it/~maxim/pub/bluez-oops/log-0.txt
Max
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Hi Max,
> > build your own 2.6.11-rc5 kernel and use netconsole to get the full
> > oops.
>
> Thanks. I've been running my program under 2.6.11-rc5 now. I got three
> 'oops' in a few hours of continuos running. They look different each time,
> yet they only occurred right after some errors like the one mentioned in
> the subject line, and those errors only occurred right before an oops. I
> put the logs here: http://cassandra.comm2000.it/~maxim/pub/bluez-oops/.
this is weird. The first one is related to sysfs and actually these
should all be fixed by now. The others are also not quite clear.
How many dongles are you using at the same time? I saw hci2 and hci1 in
the logs. What does "hciconfig -a" say about them?
Regards
Marcel
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
On Mon, 28 Feb 2005, Marcel Holtmann wrote:
> >
> > My system is Debian with kernel package 2.6.10-1-686.
>
> build your own 2.6.11-rc5 kernel and use netconsole to get the full
> oops.
Thanks. I've been running my program under 2.6.11-rc5 now. I got three
'oops' in a few hours of continuos running. They look different each time,
yet they only occurred right after some errors like the one mentioned in
the subject line, and those errors only occurred right before an oops. I
put the logs here: http://cassandra.comm2000.it/~maxim/pub/bluez-oops/.
Max
-------------------------------------------------------
SF email is sponsored by - The IT Product Guide
Read honest & candid reviews on hundreds of IT Products from real users.
Discover which products truly live up to the hype. Start reading now.
http://ads.osdn.com/?ad_id=6595&alloc_id=14396&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users