Re: [PATCH] Revert "Bluetooth: hci_qca: Set BDA quirk bit if fwnode exists in DT"

[Johan Hovold]

On Fri, Mar 29, 2024 at 12:55:40PM +0530, Janaki Ramaiah Thota wrote:
> On 3/28/2024 8:53 PM, Johan Hovold wrote:
> > On Thu, Mar 28, 2024 at 08:25:16PM +0530, Janaki Ramaiah Thota wrote:

> >> We made this change to configure the device which supports persistent
> >> memory for the BD-Address
> > 
> > Can you say something more about which devices support persistent
> > storage for the address? Is that all or just some of the chip variants?

> Most of the devices support persistent storage, and bd-address storage
> is chosen based on the OEM and Target.

Can you be more specific about which devices support it (or say which do
not)?

Is the address stored in some external eeprom or similar which the OEM
can choose to populate?

> >> So to make device functional in both scenarios we are adding a new
> >> property in dts file to distinguish persistent and non-persistent
> >> support of BD Address and set HCI_QUIRK_USE_BDADDR_PROPERTY bit
> >> accordingly
> > 
> > Depending on the answer to my questions above, you may be able to infer
> > this from the compatible string and/or you can read out the address from
> > the device and only set the quirk if it's set to the default address.
> > 
> > You should not need to add a new property for this.

> As per my understanding, altering the compatible string may cause duplicate
> configuration, right ?

If it's the same device and just a different configuration then we can't
use the compatible string for this.

It seems we need a patch like the below to address this. But please
provide some more details (e.g. answers to the questions above) so I can
determine what the end result should look like.

Johan