Return-Path: Message-ID: <4f436aae0701181559h63b7d509n5dc73c55353d8c79@mail.gmail.com> Date: Fri, 19 Jan 2007 01:59:09 +0200 From: "Choi Sonim" Sender: ohadbc@gmail.com To: "Marcel Holtmann" Subject: Re: [Bluez-devel] /sys/devices/platform/bluetooth Cc: "BlueZ development" In-Reply-To: <1169164343.6853.7.camel@violet> MIME-Version: 1.0 Content-Type: text/plain; charset=ISO-8859-1; format=flowed References: <4f436aae0701172208l7ac85a75tc592fa15d1b57f8@mail.gmail.com> <1169108019.6853.1.camel@violet> <4f436aae0701181523y55ad36bdu9645c786a252c04e@mail.gmail.com> <1169164343.6853.7.camel@violet> List-ID: Hi Marcel, On 1/19/07, Marcel Holtmann wrote: > Hi Choi, > > > > > Where is the code that does the platform registration to this device : > > > > /sys/devices/platform/bluetooth > > > > > > you find it in net/bluetooth/hci_sysfs.c in the kernel source code. > > > > What is the purpose of the bluetooth platform device (and bus and class..) ? > > > > I guess that bluetooth devices would prefer to register themselves as children > > of their true bus (like USB or PCI or even serial) so when the system goes > > to suspend, their suspend function will be called before their father's > > (the true hardware bus of our bluetooth device). > > if they do have a physical parent. Some devices like virtual HCI or > serial based chips don't have a physical parent. About the serial based chips - I'm sure they would like their suspend function to be called before the serial uart driver's suspend, so for them to set device parent to the serial device can be a benefit too, no ? About the virtual HCI - I really don't know what are them, can you please explain this ? Thank You Very very Much ! Choi. > > Regards > > Marcel > > >