Re: bluetooth disabled with current 2.6.38-rc4

["Gustavo F. Padovan" <padovan@profusion.mobi>]

Hi Justin,

* Justin Mattock <justinmattock@gmail.com> [2011-02-15 19:26:12 -0800]:

> 
> On Feb 15, 2011, at 5:52 AM, Johan Hedberg wrote:
> 
> > Hi Justin,
> >
> > On Sun, Feb 13, 2011, Justin Mattock wrote:
> >> maybe I missed something, but my bluetooth is just not functioning  
> >> with
> >> 2.6.38-rc4(works with 2.6.37-rc4)
> >>
> >> I've done a bisect on this, but was pointed to:
> >> c0e45c1ca3162acb2e77b3d9e152ce6e7b6fa3f5
> >> but doesn't look correct to me
> >>
> >> here is what I am seeing with the bluetooth-applet etc..:
> >>
> >> working correctly:
> >> http://www.flickr.com/photos/44066293@N08/5443727238/
> >>
> >> not working:
> >> http://www.flickr.com/photos/44066293@N08/5443124859/
> >>
> >> my /var/log/daemon.log shows:
> >>
> >> Feb 13 17:12:22 Linux-2 acpid: 1 client rule loaded
> >> Feb 13 17:12:23 Linux-2 bluetoothd[1950]: HCI dev 0 registered
> >> Feb 13 17:12:23 Linux-2 bluetoothd[1950]: Listening for HCI events
> >> on hci0
> >> Feb 13 17:12:23 Linux-2 bluetoothd[1950]: HCI dev 0 up
> >> Feb 13 17:12:23 Linux-2 bluetoothd[1950]: Unable to find matching
> >> adapter
> >>
> >> I can try at another bisect, but might take some time.. let me know
> >> if there is something I can test
> >> or do.
> >
> > Are you sure this is a kernel problem? There was a similar issue with
> > BlueZ 4.86 or 4.87 which was already fixed. Could you try with 4.88?
> >
> > Johan
> 
> o.k. I built 4.88 and still had the issue show up, except this time no  
> icon(instead of the icon, that is not lit up)..
> 
> 
> anyways I did do another bisect and had better results(hopefully)this  
> time..:
> 
> 23bb57633df97ede067ea26f3cdc8a7ba2cd8109 is the first bad commit
> commit 23bb57633df97ede067ea26f3cdc8a7ba2cd8109
> Author: Johan Hedberg <johan.hedberg@nokia.com>
> Date:   Tue Dec 21 23:01:27 2010 +0200
> 
>     Bluetooth: Fix __hci_request synchronization for hci_open_dev
> 
>     The initialization function used by hci_open_dev (hci_init_req)  
> sends
>     many different HCI commands. The __hci_request function should only
>     return when all of these commands have completed (or a timeout  
> occurs).
>     Several of these commands cause hci_req_complete to be called which
>     causes __hci_request to return prematurely.
> 
>     This patch fixes the issue by adding a new hdev->req_last_cmd  
> variable
>     which is set during the initialization procedure. The  
> hci_req_complete
>     function will no longer mark the request as complete until the  
> command
>     matching hdev->req_last_cmd completes.
> 
>     Signed-off-by: Johan Hedberg <johan.hedberg@nokia.com>
>     Acked-by: Marcel Holtmann <marcel@holtmann.org>
>     Signed-off-by: Gustavo F. Padovan <padovan@profusion.mobi>
> 
> :040000 040000 e8d8ae5fab74b2ba73d0c52e7e09b30e5df8484e  
> 8d1409e8dc2206492cb86e8848c8c23aff182e2a M      include
> :040000 040000 07fdb5f6ceadcfde12c46a3b5c82937b7dac892c  
> 8f3b08fbbc9fcd078d0f5224b8bc818eeb072dc4 M      net

Can you please send the hcidump output? There should be a package for it in
your distribution, if not get the bluez-hcidump from git.kernel.org

-- 
Gustavo F. Padovan
http://profusion.mobi