[Bluez-devel] Patch for kernel oops while disconnecting devices (hci_usb.c)

[Bluez-devel] Patch for kernel oops while disconnecting devices (hci_usb.c)

[Alexander Holler]

Hi,

I've a small patch for hci_usb.c (from 2.6-test11 and maybe below) which 
removes the kernel oops when disconnecting a csr-usb-devices (in my case 
with a button on my laptop).

----------------
--- hci_usb.c.orig      2003-12-15 18:03:24.000000000 +0100
+++ hci_usb.c   2003-12-17 00:10:29.000000000 +0100
@@ -781,7 +781,7 @@
        struct hci_dev *hdev;
        int i, a, e, size, ifn, isoc_ifnum, isoc_alts;

-       BT_DBG("udev %p ifnum %d", udev, ifnum);
+/*     BT_DBG("udev %p ifnum %d", udev, ifnum); */

        iface = udev->actconfig->interface[0];

@@ -934,10 +934,10 @@
        BT_DBG("%s", hdev->name);

        hci_usb_close(hdev);
-
+/*
        if (husb->isoc_iface)
                usb_driver_release_interface(&hci_usb_driver, 
husb->isoc_iface);
-
+*/
        if (hci_unregister_dev(hdev) < 0)
                BT_ERR("Can't unregister HCI device %s", hdev->name);
 }
----------------

I've also commented out a nonfunctional BT_DBG.

I haven't identified the real cause for the oops, but according to the 
documentation for usb_driver_release_interface() this call isn't necessary 
and is called automatically when a driver is disconnected.

According /proc/bus/usb/devices the interface is removed, even without a 
call to usb_driver_release_interface in disconnect().

Maybe this should be forwarded to one of the usb-devels for examing.
I think they will find the real cause for the oops much faster than I (I 
almost know nothing about the usb subsystem). ;)

Regards,

Alexander



-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] Patch for kernel oops while disconnecting devices (hci_usb.c)

[Marcel Holtmann]

Hi Alexander,

> I've a small patch for hci_usb.c (from 2.6-test11 and maybe below) which 
> removes the kernel oops when disconnecting a csr-usb-devices (in my case 
> with a button on my laptop).

what happens if you compile the HCI USB driver without SCO support?

Regards

Marcel




-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] Patch for kernel oops while disconnecting devices (hci_usb.c)

[Alexander Holler]

Marcel Holtmann wrote:
> Hi Alexander,
> 
>>I've a small patch for hci_usb.c (from 2.6-test11 and maybe below) which 
>>removes the kernel oops when disconnecting a csr-usb-devices (in my case 
>>with a button on my laptop).
> 
> what happens if you compile the HCI USB driver without SCO support?

The same as with this patch and SCO support, it works. Because that 
pointer on the lkml I've found the fix/workaround relativly fast.

Regards, Alexander




-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel

Re: [Bluez-devel] Patch for kernel oops while disconnecting devices (hci_usb.c)

[Marcel Holtmann]

Hi Alexander,

> > what happens if you compile the HCI USB driver without SCO support?
> 
> The same as with this patch and SCO support, it works. Because that 
> pointer on the lkml I've found the fix/workaround relativly fast.

I never have worked on USB with ISOC transfers, so I don't know what is
the correct way to fix it. I propose that you send a full detailed bug
report (with ksymoops etc.) to the USB developer mailing list and let us
see what they say.

BTW the BT_DBG is not non-functional ;)

Regards

Marcel




-------------------------------------------------------
This SF.net email is sponsored by: IBM Linux Tutorials.
Become an expert in LINUX or just sharpen your skills.  Sign up for IBM's
Free Linux Tutorials.  Learn everything from the bash shell to sys admin.
Click now! http://ads.osdn.com/?ad_id=1278&alloc_id=3371&op=click
_______________________________________________
Bluez-devel mailing list
Bluez-devel@lists.sourceforge.net
https://lists.sourceforge.net/lists/listinfo/bluez-devel