[PATCH] Bluetooth: btusb: Only check needs_reset_resume DMI table for QCA rome chipsets

[PATCH] Bluetooth: btusb: Only check needs_reset_resume DMI table for QCA rome chipsets

[Hans de Goede]

Jeremy Cline correctly points out in rhbz#1514836 that a device where the
QCA rome chipset needs the USB_QUIRK_RESET_RESUME quirk, may also ship
with a different wifi/bt chipset in some configurations.

If that is the case then we are needlessly penalizing those other chipsets
with a reset-resume quirk, typically causing 0.4W extra power use because
this disables runtime-pm.

This commit moves the DMI table check to a btusb_check_needs_reset_resume()
helper (so that we can easily also call it for other chipsets) and calls
this new helper only for QCA_ROME chipsets for now.

BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
Cc: stable@vger.kernel.org
Cc: Jeremy Cline <jcline@redhat.com>
Suggested-by: Jeremy Cline <jcline@redhat.com>
Signed-off-by: Hans de Goede <hdegoede@redhat.com>
---
 drivers/bluetooth/btusb.c | 10 +++++++---
 1 file changed, 7 insertions(+), 3 deletions(-)

diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
index f064984c9ec0..15e7cdca6eb5 100644
--- a/drivers/bluetooth/btusb.c
+++ b/drivers/bluetooth/btusb.c
@@ -2863,6 +2863,12 @@ static int btusb_config_oob_wake(struct hci_dev *hdev)
 }
 #endif
 
+static void btusb_check_needs_reset_resume(struct usb_interface *intf)
+{
+	if (dmi_check_system(btusb_needs_reset_resume_table))
+		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
+}
+
 static int btusb_probe(struct usb_interface *intf,
 		       const struct usb_device_id *id)
 {
@@ -2985,9 +2991,6 @@ static int btusb_probe(struct usb_interface *intf,
 	hdev->send   = btusb_send_frame;
 	hdev->notify = btusb_notify;
 
-	if (dmi_check_system(btusb_needs_reset_resume_table))
-		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
-
 #ifdef CONFIG_PM
 	err = btusb_config_oob_wake(hdev);
 	if (err)
@@ -3076,6 +3079,7 @@ static int btusb_probe(struct usb_interface *intf,
 		data->setup_on_usb = btusb_setup_qca;
 		hdev->set_bdaddr = btusb_set_bdaddr_ath3012;
 		set_bit(HCI_QUIRK_SIMULTANEOUS_DISCOVERY, &hdev->quirks);
+		btusb_check_needs_reset_resume(intf);
 	}
 
 #ifdef CONFIG_BT_HCIBTUSB_RTL
-- 
2.17.0

Re: [PATCH] Bluetooth: btusb: Only check needs_reset_resume DMI table for QCA rome chipsets

[Jeremy Cline]

On 04/27/2018 05:26 AM, Hans de Goede wrote:
> Jeremy Cline correctly points out in rhbz#1514836 that a device where the
> QCA rome chipset needs the USB_QUIRK_RESET_RESUME quirk, may also ship
> with a different wifi/bt chipset in some configurations.
> 
> If that is the case then we are needlessly penalizing those other chipsets
> with a reset-resume quirk, typically causing 0.4W extra power use because
> this disables runtime-pm.
> 
> This commit moves the DMI table check to a btusb_check_needs_reset_resume()
> helper (so that we can easily also call it for other chipsets) and calls
> this new helper only for QCA_ROME chipsets for now.
> 
> BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
> Cc: stable@vger.kernel.org
> Cc: Jeremy Cline <jcline@redhat.com>
> Suggested-by: Jeremy Cline <jcline@redhat.com>
> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
> ---
>  drivers/bluetooth/btusb.c | 10 +++++++---
>  1 file changed, 7 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
> index f064984c9ec0..15e7cdca6eb5 100644
> --- a/drivers/bluetooth/btusb.c
> +++ b/drivers/bluetooth/btusb.c
> @@ -2863,6 +2863,12 @@ static int btusb_config_oob_wake(struct hci_dev *hdev)
>  }
>  #endif
>  
> +static void btusb_check_needs_reset_resume(struct usb_interface *intf)
> +{
> +	if (dmi_check_system(btusb_needs_reset_resume_table))
> +		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
> +}
> +
>  static int btusb_probe(struct usb_interface *intf,
>  		       const struct usb_device_id *id)
>  {
> @@ -2985,9 +2991,6 @@ static int btusb_probe(struct usb_interface *intf,
>  	hdev->send   = btusb_send_frame;
>  	hdev->notify = btusb_notify;
>  
> -	if (dmi_check_system(btusb_needs_reset_resume_table))
> -		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
> -
>  #ifdef CONFIG_PM
>  	err = btusb_config_oob_wake(hdev);
>  	if (err)
> @@ -3076,6 +3079,7 @@ static int btusb_probe(struct usb_interface *intf,
>  		data->setup_on_usb = btusb_setup_qca;
>  		hdev->set_bdaddr = btusb_set_bdaddr_ath3012;
>  		set_bit(HCI_QUIRK_SIMULTANEOUS_DISCOVERY, &hdev->quirks);
> +		btusb_check_needs_reset_resume(intf);

If we later need the quirk applied to an Intel chipset for a particular
DMI, this call would get added to the Intel block and then the quirk
would start getting applied to the Intel variety of the XPS 13 9360
again, right?

I have a feeling that's a rather large "if" and I don't have any idea
how likely it is. Is it even something worth worrying about?

Regards,
Jeremy

Re: [PATCH] Bluetooth: btusb: Only check needs_reset_resume DMI table for QCA rome chipsets

[Hans de Goede]

Hi,

On 04/27/2018 04:00 PM, Jeremy Cline wrote:
> On 04/27/2018 05:26 AM, Hans de Goede wrote:
>> Jeremy Cline correctly points out in rhbz#1514836 that a device where the
>> QCA rome chipset needs the USB_QUIRK_RESET_RESUME quirk, may also ship
>> with a different wifi/bt chipset in some configurations.
>>
>> If that is the case then we are needlessly penalizing those other chipsets
>> with a reset-resume quirk, typically causing 0.4W extra power use because
>> this disables runtime-pm.
>>
>> This commit moves the DMI table check to a btusb_check_needs_reset_resume()
>> helper (so that we can easily also call it for other chipsets) and calls
>> this new helper only for QCA_ROME chipsets for now.
>>
>> BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
>> Cc: stable@vger.kernel.org
>> Cc: Jeremy Cline <jcline@redhat.com>
>> Suggested-by: Jeremy Cline <jcline@redhat.com>
>> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
>> ---
>>   drivers/bluetooth/btusb.c | 10 +++++++---
>>   1 file changed, 7 insertions(+), 3 deletions(-)
>>
>> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
>> index f064984c9ec0..15e7cdca6eb5 100644
>> --- a/drivers/bluetooth/btusb.c
>> +++ b/drivers/bluetooth/btusb.c
>> @@ -2863,6 +2863,12 @@ static int btusb_config_oob_wake(struct hci_dev *hdev)
>>   }
>>   #endif
>>   
>> +static void btusb_check_needs_reset_resume(struct usb_interface *intf)
>> +{
>> +	if (dmi_check_system(btusb_needs_reset_resume_table))
>> +		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
>> +}
>> +
>>   static int btusb_probe(struct usb_interface *intf,
>>   		       const struct usb_device_id *id)
>>   {
>> @@ -2985,9 +2991,6 @@ static int btusb_probe(struct usb_interface *intf,
>>   	hdev->send   = btusb_send_frame;
>>   	hdev->notify = btusb_notify;
>>   
>> -	if (dmi_check_system(btusb_needs_reset_resume_table))
>> -		interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
>> -
>>   #ifdef CONFIG_PM
>>   	err = btusb_config_oob_wake(hdev);
>>   	if (err)
>> @@ -3076,6 +3079,7 @@ static int btusb_probe(struct usb_interface *intf,
>>   		data->setup_on_usb = btusb_setup_qca;
>>   		hdev->set_bdaddr = btusb_set_bdaddr_ath3012;
>>   		set_bit(HCI_QUIRK_SIMULTANEOUS_DISCOVERY, &hdev->quirks);
>> +		btusb_check_needs_reset_resume(intf);
> 
> If we later need the quirk applied to an Intel chipset for a particular
> DMI, this call would get added to the Intel block and then the quirk
> would start getting applied to the Intel variety of the XPS 13 9360
> again, right?
> 
> I have a feeling that's a rather large "if" and I don't have any idea
> how likely it is. Is it even something worth worrying about?

ATM that seems rather unlikely, but we are tracking the USB-ids to
which the DMI quirks belong (as comments) so we can later switch from
a dmi-id only check to a combined dmi + usb-id check if necessary.

Regards,

Hans

Re: [PATCH] Bluetooth: btusb: Only check needs_reset_resume DMI table for QCA rome chipsets

[Jeremy Cline]

On 04/27/2018 02:43 PM, Hans de Goede wrote:
> Hi,
> 
> On 04/27/2018 04:00 PM, Jeremy Cline wrote:
>> On 04/27/2018 05:26 AM, Hans de Goede wrote:
>>> Jeremy Cline correctly points out in rhbz#1514836 that a device where
>>> the
>>> QCA rome chipset needs the USB_QUIRK_RESET_RESUME quirk, may also ship
>>> with a different wifi/bt chipset in some configurations.
>>>
>>> If that is the case then we are needlessly penalizing those other
>>> chipsets
>>> with a reset-resume quirk, typically causing 0.4W extra power use
>>> because
>>> this disables runtime-pm.
>>>
>>> This commit moves the DMI table check to a
>>> btusb_check_needs_reset_resume()
>>> helper (so that we can easily also call it for other chipsets) and calls
>>> this new helper only for QCA_ROME chipsets for now.
>>>
>>> BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
>>> Cc: stable@vger.kernel.org
>>> Cc: Jeremy Cline <jcline@redhat.com>
>>> Suggested-by: Jeremy Cline <jcline@redhat.com>
>>> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
>>> ---
>>>   drivers/bluetooth/btusb.c | 10 +++++++---
>>>   1 file changed, 7 insertions(+), 3 deletions(-)
>>>
>>> diff --git a/drivers/bluetooth/btusb.c b/drivers/bluetooth/btusb.c
>>> index f064984c9ec0..15e7cdca6eb5 100644
>>> --- a/drivers/bluetooth/btusb.c
>>> +++ b/drivers/bluetooth/btusb.c
>>> @@ -2863,6 +2863,12 @@ static int btusb_config_oob_wake(struct
>>> hci_dev *hdev)
>>>   }
>>>   #endif
>>>   +static void btusb_check_needs_reset_resume(struct usb_interface
>>> *intf)
>>> +{
>>> +    if (dmi_check_system(btusb_needs_reset_resume_table))
>>> +        interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
>>> +}
>>> +
>>>   static int btusb_probe(struct usb_interface *intf,
>>>                  const struct usb_device_id *id)
>>>   {
>>> @@ -2985,9 +2991,6 @@ static int btusb_probe(struct usb_interface *intf,
>>>       hdev->send   = btusb_send_frame;
>>>       hdev->notify = btusb_notify;
>>>   -    if (dmi_check_system(btusb_needs_reset_resume_table))
>>> -        interface_to_usbdev(intf)->quirks |= USB_QUIRK_RESET_RESUME;
>>> -
>>>   #ifdef CONFIG_PM
>>>       err = btusb_config_oob_wake(hdev);
>>>       if (err)
>>> @@ -3076,6 +3079,7 @@ static int btusb_probe(struct usb_interface *intf,
>>>           data->setup_on_usb = btusb_setup_qca;
>>>           hdev->set_bdaddr = btusb_set_bdaddr_ath3012;
>>>           set_bit(HCI_QUIRK_SIMULTANEOUS_DISCOVERY, &hdev->quirks);
>>> +        btusb_check_needs_reset_resume(intf);
>>
>> If we later need the quirk applied to an Intel chipset for a particular
>> DMI, this call would get added to the Intel block and then the quirk
>> would start getting applied to the Intel variety of the XPS 13 9360
>> again, right?
>>
>> I have a feeling that's a rather large "if" and I don't have any idea
>> how likely it is. Is it even something worth worrying about?
> 
> ATM that seems rather unlikely, but we are tracking the USB-ids to
> which the DMI quirks belong (as comments) so we can later switch from
> a dmi-id only check to a combined dmi + usb-id check if necessary.

Makes sense, thanks.

Regards,
Jeremy

Re: [PATCH] Bluetooth: btusb: Only check needs_reset_resume DMI table for QCA rome chipsets

[Marcel Holtmann]

Hi Hans,

> Jeremy Cline correctly points out in rhbz#1514836 that a device where the
> QCA rome chipset needs the USB_QUIRK_RESET_RESUME quirk, may also ship
> with a different wifi/bt chipset in some configurations.
> 
> If that is the case then we are needlessly penalizing those other chipsets
> with a reset-resume quirk, typically causing 0.4W extra power use because
> this disables runtime-pm.
> 
> This commit moves the DMI table check to a btusb_check_needs_reset_resume()
> helper (so that we can easily also call it for other chipsets) and calls
> this new helper only for QCA_ROME chipsets for now.
> 
> BugLink: https://bugzilla.redhat.com/show_bug.cgi?id=1514836
> Cc: stable@vger.kernel.org
> Cc: Jeremy Cline <jcline@redhat.com>
> Suggested-by: Jeremy Cline <jcline@redhat.com>
> Signed-off-by: Hans de Goede <hdegoede@redhat.com>
> ---
> drivers/bluetooth/btusb.c | 10 +++++++---
> 1 file changed, 7 insertions(+), 3 deletions(-)

patch has been applied to bluetooth-stable tree.

Regards

Marcel