Hi John,
A few more fixes to 3.6, here we have four important fix to the MGMT interface,
two from Johan Hedberg and Andrzej Kaczmarek. Andrei fixed a free on
uninitialized memory and I added support for Broadcom/Foxconn devices.
Please pull, or let me know of any problems. Thanks.
Gustavo
The following changes since commit f10723841e624c0726c70356b31d91befed01dd6:
libertas sdio: fix suspend when interface is down (2012-09-05 14:53:36 -0400)
are available in the git repository at:
git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth master
for you to fetch changes up to b6f04c364ef853d8baa8949be20ef708000bcb97:
Bluetooth: Fix freeing uninitialized delayed works (2012-09-08 17:27:52 -0300)
----------------------------------------------------------------
Andrei Emeltchenko (1):
Bluetooth: Fix freeing uninitialized delayed works
Andrzej Kaczmarek (2):
Bluetooth: mgmt: Fix enabling SSP while powered off
Bluetooth: mgmt: Fix enabling LE while powered off
Gustavo Padovan (1):
Bluetooth: Add USB_VENDOR_AND_INTERFACE_INFO() for Broadcom/Foxconn
Johan Hedberg (2):
Bluetooth: mgmt: Implement support for passkey notification
Bluetooth: Update management interface revision
drivers/bluetooth/btusb.c | 2 +-
include/net/bluetooth/hci.h | 18 ++++++++++++
include/net/bluetooth/hci_core.h | 5 ++++
include/net/bluetooth/mgmt.h | 7 +++++
net/bluetooth/hci_event.c | 67 ++++++++++++++++++++++++++++++++++++++++++
net/bluetooth/l2cap_core.c | 2 +-
net/bluetooth/mgmt.c | 35 +++++++++++++++++++++-
7 files changed, 133 insertions(+), 3 deletions(-)
Hi John,
* John W. Linville <[email protected]> [2012-09-12 14:13:10 -0400]:
> On Sat, Sep 08, 2012 at 06:59:56PM -0300, Gustavo Padovan wrote:
> > Hi John,
> >
> > A few more fixes to 3.6, here we have four important fix to the MGMT interface,
> > two from Johan Hedberg and Andrzej Kaczmarek. Andrei fixed a free on
> > uninitialized memory and I added support for Broadcom/Foxconn devices.
>
> "Bluetooth: mgmt: Implement support for passkey notification" doesn't
> look like a fix at all.
It is a fix in the sense that we were failing to work with some keyboards out
there that only support Bluetooth SSP KeyboardOnly IO capability and there are
few of those in the market now. But yes, we are not fix any crash, etc in the
kernel, not critical though. It is just improving compatibility of the
Bluetooth stack (which is important too). It's up to you merge this or not.
>
> "Bluetooth: Update management interface revision" is suspect. It kinda
> makes sense, but if it is so important than it should have been merged
> much earlier, no?
Not really, but we like to bump version when we do significative changes to
MGMT like the above.
>
> "Bluetooth: Add USB_VENDOR_AND_INTERFACE_INFO() for Broadcom/Foxconn"
> might normally be OK, but it is getting late enough in the cycle
> that I would prefer to delay any device ID patches until the next
> merge window.
Ok.
>
> The others look fine. Do you want me to merge them directly?
> Or would you rather respin your pull request?
You can pull them directly, I'll redo my tree once they get pulled.
Thanks, John.
Gustavo
On Sat, Sep 08, 2012 at 06:59:56PM -0300, Gustavo Padovan wrote:
> Hi John,
>
> A few more fixes to 3.6, here we have four important fix to the MGMT interface,
> two from Johan Hedberg and Andrzej Kaczmarek. Andrei fixed a free on
> uninitialized memory and I added support for Broadcom/Foxconn devices.
"Bluetooth: mgmt: Implement support for passkey notification" doesn't
look like a fix at all.
"Bluetooth: Update management interface revision" is suspect. It kinda
makes sense, but if it is so important than it should have been merged
much earlier, no?
"Bluetooth: Add USB_VENDOR_AND_INTERFACE_INFO() for Broadcom/Foxconn"
might normally be OK, but it is getting late enough in the cycle
that I would prefer to delay any device ID patches until the next
merge window.
The others look fine. Do you want me to merge them directly?
Or would you rather respin your pull request?
John
> Please pull, or let me know of any problems. Thanks.
>
> Gustavo
>
> The following changes since commit f10723841e624c0726c70356b31d91befed01dd6:
>
> libertas sdio: fix suspend when interface is down (2012-09-05 14:53:36 -0400)
>
> are available in the git repository at:
>
> git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth master
>
> for you to fetch changes up to b6f04c364ef853d8baa8949be20ef708000bcb97:
>
> Bluetooth: Fix freeing uninitialized delayed works (2012-09-08 17:27:52 -0300)
>
> ----------------------------------------------------------------
> Andrei Emeltchenko (1):
> Bluetooth: Fix freeing uninitialized delayed works
>
> Andrzej Kaczmarek (2):
> Bluetooth: mgmt: Fix enabling SSP while powered off
> Bluetooth: mgmt: Fix enabling LE while powered off
>
> Gustavo Padovan (1):
> Bluetooth: Add USB_VENDOR_AND_INTERFACE_INFO() for Broadcom/Foxconn
>
> Johan Hedberg (2):
> Bluetooth: mgmt: Implement support for passkey notification
> Bluetooth: Update management interface revision
>
> drivers/bluetooth/btusb.c | 2 +-
> include/net/bluetooth/hci.h | 18 ++++++++++++
> include/net/bluetooth/hci_core.h | 5 ++++
> include/net/bluetooth/mgmt.h | 7 +++++
> net/bluetooth/hci_event.c | 67 ++++++++++++++++++++++++++++++++++++++++++
> net/bluetooth/l2cap_core.c | 2 +-
> net/bluetooth/mgmt.c | 35 +++++++++++++++++++++-
> 7 files changed, 133 insertions(+), 3 deletions(-)
>
--
John W. Linville Someday the world will need a hero, and you
[email protected] might be all we have. Be ready.
Hi John,
On Wed, Sep 12, 2012, John W. Linville wrote:
> "Bluetooth: mgmt: Implement support for passkey notification" doesn't
> look like a fix at all.
This fixes pairing with keyboards that have a "KeyboardOnly" IO
capability. This used to work previously in a world where user space
spoke HCI directly to the Bluetooth controller but is broken with any
user space version using the recently introduced management interface.
> "Bluetooth: Update management interface revision" is suspect. It kinda
> makes sense, but if it is so important than it should have been merged
> much earlier, no?
This patch is due to the first one. We increment the revision every time
something gets added to the management interface. FWIW, I've tested that
pairing now works with keyboards of this kind with these patches.
Johan