Content-Type: text/plain;
        charset=utf-8
Mime-Version: 1.0 (Mac OS X Mail 12.2 \(3445.102.3\))
Subject: Re: [PATCH] btusb: Add shutdown routine for BTUSB_INTEL_NEW devices
From:   Marcel Holtmann <marcel@holtmann.org>
In-Reply-To: <CAMH2TCqtZM++VSboKbFB9htTCuP=8Z+8PmGDymdkSgrEXKr2Og@mail.gmail.com>
Date:   Mon, 28 Jan 2019 17:00:12 +0100
Cc:     Raghuram Hegde <raghuram.hegde@intel.com>,
        Bluez mailing list <linux-bluetooth@vger.kernel.org>,
        Chethan T N <chethan.tumkur.narayan@intel.com>
Content-Transfer-Encoding: 8BIT
Message-Id: <110D522B-D320-4839-A519-C8379C251314@holtmann.org>
References: <1548665625-23742-1-git-send-email-raghuram.hegde@intel.com>
 <91F174CE-3CBE-4D3B-B0C9-E350C01F3DDF@holtmann.org>
 <CAMH2TCqtZM++VSboKbFB9htTCuP=8Z+8PmGDymdkSgrEXKr2Og@mail.gmail.com>
To:     chethan tn <chethantn@gmail.com>
Sender: linux-bluetooth-owner@vger.kernel.org
Precedence: bulk

Hi Chethan,

>>> btusb_shutdown_intel routine shall reset the controller
>>> and stop all BT operation.
>>> Advantages:
>>>      1. Power save on the platform
>>>      2. Host and controller will be in Sync.
>>> 
>>> Signed-off-by: Raghuram Hegde <raghuram.hegde@intel.com>
>>> Signed-off-by: Chethan T N <chethan.tumkur.narayan@intel.com>
>>> ---
>>> drivers/bluetooth/btusb.c | 1 +
>>> 1 file changed, 1 insertion(+)
>>> 
>>> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
>>> index 5de0c2e59b97..66483ca3d870 100644
>>> --- a/drivers/bluetooth/btusb.c
>>> +++ b/drivers/bluetooth/btusb.c
>>> @@ -3139,6 +3139,7 @@ static int btusb_probe(struct usb_interface *intf,
>>>              hdev->manufacturer = 2;
>>>              hdev->send = btusb_send_frame_intel;
>>>              hdev->setup = btusb_setup_intel_new;
>>> +             hdev->shutdown = btusb_shutdown_intel;
>>>              hdev->hw_error = btintel_hw_error;
>>>              hdev->set_diag = btintel_set_diag;
>>>              hdev->set_bdaddr = btintel_set_bdaddr;
>> 
>> I assumed that this was only needed for the older ROM versions of the Intel controllers and not the newer RAM versions. I have been told they don’t inherit the LED issue that we tried to fix with this. Please read the comments in btusb_shutdown_intel and amend comments if needed and provide a detailed commit message.
> Yes you're absolutely right about the LED issue on the older ROM products.
> But in the recent day we have observed that in case BT
> operation(Inquiry/LE Scan) were triggered through the stack and
> further BT was turned off through "hciconfig hci0 down". In this case
> controller would active doing BT operation and consume power and also
> might cause race condition on the next BT on as the controller might
> try to push the events that were queued up before processing the reset
> command. So to make sure when BT is turned off either through stack or
> through command line thought it would be better approach to reset the
> controller(This is applicable for ROM or RAM products).

what about creating a btusb_shutdown_intel_new() that just sends the HCI Reset command. That should be enough to ensure that anything in the radio scheduler gets cancelled and cleaned up.

Regards

Marcel