Sorry if this is mildly off-topic...
Every time I remove a bluetooth USB key from my desktop machine, I get
a hard lock. I've tried this with a Belkin key and D-Link key (not the
one with the Broadcom chipset). I don't even see a kernel panic in
text mode. However, my laptop (running exactly the same kernel
version) works fine on remove. So far, this only seems to happen with
bluetooth USB keys. I've tried a number of other USB devices on my
desktop, and they all remove cleanly (even after several such cycles).
The desktop (where the problems are occuring) has an A-Bit motherboard
with all Via components, and has both a USB 1.1 and 2.0 controller
(I'm pretty sure I'm plugging the key into one of the 1.1 ports).
The laptop is a Toshiba Libretto L5 with ALI components (USB 1.1).
I'm running kernel 2.6.3 with no patches applied on both systems.
This is most likely a module unloading bug triggered by the hotplug
event, but I figured I'd ask here first, as it is bluetooth related.
If anyone else has a similar hardware combination, would you mind
attempting to remove the key a couple of times? (occasionally the
desktop will only freeze after the second unload)
Does anyone have any strategies for extracting some debugging
information? I've not played with SysReq or tracebacks, so I'm not
sure if they'd help in this case.
Thanks,
Russell
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
On Thursday 11 March 2004 10:32, Matthias Thomae wrote:
Hi,
you were right.
I've installed the 2.6.4 and I have the same problem too, this happens only
with 2.6.4 and not with 2.6.3.
Someone wrote to the kernel mailing list too:
http://www.ussg.iu.edu/hypermail/linux/kernel/0403.1/1519.html
Bye!
--
Simone Gotti
<[email protected]>
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Hi all,
I also get a crash when unplugging my bt usb adapter (Acer BT-700,
CSR-based) on a VIA Chipset machine with 2.6.4-rc3 kernel.
CONFIG_BT_HCIUSB_SCO is *not* set, CONFIG_BT_SCO is.
Regards.
Matthias
Russell Neches wrote:
> Now, by SCO support, do we mean the module (CONFIG_BT_SCO) or
> CONFIG_BT_HCIUSB_SCO? I'm actually getting the crash without the sco
> module loaded, and I have my kernel built with CONFIG_BT_HCIUSB_SCO not
> set. I don't think SCO shows up anywhere else in the kernel. Sorry if
> I'm beginning to sound pedantic.
>
> Maybe I should figure out how to use SysReq after all...
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
> yes, i enabled force remove because the module shows up as in use. i
> think it's unsafe, but how much worse can it be than crashing the kernel?
Indeed.
> i am compiling the modules without SCO support right now to see if that
> helps.
Now, by SCO support, do we mean the module (CONFIG_BT_SCO) or
CONFIG_BT_HCIUSB_SCO? I'm actually getting the crash without the sco
module loaded, and I have my kernel built with CONFIG_BT_HCIUSB_SCO not
set. I don't think SCO shows up anywhere else in the kernel. Sorry if
I'm beginning to sound pedantic.
Maybe I should figure out how to use SysReq after all...
Russell
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
On Thu, Mar 11, 2004 at 01:11:31AM +0100, Simone Gotti wrote:
> On Thursday 11 March 2004 00:41, Russell Neches wrote:
> > > Hi!, I think it's a SCO support problem .
> >
> > Those bastards! Death to SCO!
> >
> > Oh. SCO. Synchronous Connection Oriented.
> >
> > > I think that Marcel Holtmann spoked about this a lot of times on the
> > > mailinglists. You have to disable the SCO support in the hci_usb
> > > module to avoid this problem. I've read that probably it's a usb uchi
> > > isochronous problem.
> >
> > Hmm... Do you have a link?
>
> I've found this, but there are many others.
> http://www.ussg.iu.edu/hypermail/linux/kernel/0311.0/0135.html
Are you sure it's related to SCO? Then shouldn't my laptop exhibit the
same behavior? It's interesting to note that he is also using an Abit
motherboard, though his board uses the Intel 875P chipset, not the Via
chipset (like mine). Maybe it's related to USB 2.0?
Russell
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Russell
>>safely remove a usb adapter is to stop the bluetooth daemons, force
>>remove hci_usb (there always seems to be one reference pinning it down
>>that I can't find), then unplug the adapter.
>
>
> Do you have to do a force remove, or will it unload cleanly?
>
> What do you mean by a "reference"? Does it claim to be in use when you
> you rmmod it?
yes, i enabled force remove because the module shows up as in use. i
think it's unsafe, but how much worse can it be than crashing the kernel?
i am compiling the modules without SCO support right now to see if that
helps.
brad
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
On Thursday 11 March 2004 00:41, Russell Neches wrote:
> > Hi!, I think it's a SCO support problem .
>
> Those bastards! Death to SCO!
>
> Oh. SCO. Synchronous Connection Oriented.
>
> > I think that Marcel Holtmann spoked about this a lot of times on the
> > mailinglists. You have to disable the SCO support in the hci_usb
> > module to avoid this problem. I've read that probably it's a usb uchi
> > isochronous problem.
>
> Hmm... Do you have a link?
I've found this, but there are many others.
http://www.ussg.iu.edu/hypermail/linux/kernel/0311.0/0135.html
Bye!
--
Simone Gotti
<[email protected]>
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
> Hi!, I think it's a SCO support problem .
Those bastards! Death to SCO!
Oh. SCO. Synchronous Connection Oriented.
> I think that Marcel Holtmann spoked about this a lot of times on the
> mailinglists. You have to disable the SCO support in the hci_usb
> module to avoid this problem. I've read that probably it's a usb uchi
> isochronous problem.
Hmm... Do you have a link?
Russell
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
On Wednesday 10 March 2004 23:35, Russell Neches wrote:
> > >Every time I remove a bluetooth USB key from my desktop machine, I
> > >get a hard lock. I'm running kernel 2.6.3 with no patches applied on
> > >both systems.
> >
> > I also see something similar on 2.6.3-mh4. I think the only way I can
> > safely remove a usb adapter is to stop the bluetooth daemons, force
> > remove hci_usb (there always seems to be one reference pinning it down
> > that I can't find), then unplug the adapter.
>
> Do you have to do a force remove, or will it unload cleanly? I have
> forced module unloading turned off in my kernel, which may be the cause
> of my problem...
>
> What do you mean by a "reference"? Does it claim to be in use when you
> you rmmod it?
>
> > I'm using a toshiba m200. The internal adapter is also USB so I think
> > I have to do that before (ACPI-)suspending the laptop. I haven't tried
> > the external since I got internal working. Lately either suspending or
> > resuming crashes my machine so that may not be helping.
>
> Yeah, I get crashes on resume too... Sigh. I never thought I would, but
> I actually miss APM. The lack of APM support on my laptop has been a
> never-ending headache.
>
> Russell
Hi!, I think it's a SCO support problem .
I think that Marcel Holtmann spoked about this a lot of times on the
mailinglists. You have to disable the SCO support in the hci_usb module to
avoid this problem. I've read that probably it's a usb uchi isochronous
problem.
Bye!
--
Simone Gotti
<[email protected]>
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
> >Every time I remove a bluetooth USB key from my desktop machine, I
> >get a hard lock. I'm running kernel 2.6.3 with no patches applied on
> >both systems.
>
> I also see something similar on 2.6.3-mh4. I think the only way I can
> safely remove a usb adapter is to stop the bluetooth daemons, force
> remove hci_usb (there always seems to be one reference pinning it down
> that I can't find), then unplug the adapter.
Do you have to do a force remove, or will it unload cleanly? I have
forced module unloading turned off in my kernel, which may be the cause
of my problem...
What do you mean by a "reference"? Does it claim to be in use when you
you rmmod it?
> I'm using a toshiba m200. The internal adapter is also USB so I think
> I have to do that before (ACPI-)suspending the laptop. I haven't tried
> the external since I got internal working. Lately either suspending or
> resuming crashes my machine so that may not be helping.
Yeah, I get crashes on resume too... Sigh. I never thought I would, but
I actually miss APM. The lack of APM support on my laptop has been a
never-ending headache.
Russell
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users
Russell
> Every time I remove a bluetooth USB key from my desktop machine, I get
> a hard lock.
> I'm running kernel 2.6.3 with no patches applied on both systems.
I also see something similar on 2.6.3-mh4. I think the only way I can
safely remove a usb adapter is to stop the bluetooth daemons, force
remove hci_usb (there always seems to be one reference pinning it down
that I can't find), then unplug the adapter.
I'm using a toshiba m200. The internal adapter is also USB so I think I
have to do that before (ACPI-)suspending the laptop. I haven't tried the
external since I got internal working. Lately either suspending or
resuming crashes my machine so that may not be helping.
Brad
-------------------------------------------------------
This SF.Net email is sponsored by: IBM Linux Tutorials
Free Linux tutorial presented by Daniel Robbins, President and CEO of
GenToo technologies. Learn everything from fundamentals to system
administration.http://ads.osdn.com/?ad_id=1470&alloc_id=3638&op=click
_______________________________________________
Bluez-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/bluez-users