linux-kernel - Re: Releasing bus devices?

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <CALCETrXvKM18sUpOGnSLoX-Sz-UAOcyNWK-xFtSK7zOcikm5Bw@mail.gmail.com>
Date:	Fri, 27 Nov 2015 09:27:09 -0800
From:	Andy Lutomirski <luto@...capital.net>
To:	Greg KH <gregkh@...uxfoundation.org>
Cc:	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: Releasing bus devices?

On Wed, Nov 25, 2015 at 8:37 AM, Greg KH <gregkh@...uxfoundation.org> wrote:
> On Wed, Nov 25, 2015 at 08:21:36AM -0800, Andy Lutomirski wrote:
>> I'm writing a bus driver, and I'm confused.
>
> Yeah, writing a new bus is a major pain, I keep meaning to work on the
> api every year or so when I have to write a new one, but then I just
> suffer through it and never get around to it :(
>
>> If I create a class and class devices, I can set a release method on
>> the class.  But if I create a bus, there's no bus_type.dev_release.
>> Am I really supposed to manually set .release on each struct device
>> that I enumerate?  If so, why?
>
> No, you shouldn't have to do that, the "struct device_type" that is
> assigned to your device has a .release callback, that is what will be
> called.  Make sure you set that for each device you create for your bus.
>
> hope this helps,
>

It does, thanks.

One more question: device_destroy seems to be the opposite of
device_create.  But device_destroy does put_device and
device_unregister, and device_unregister also does put_device.  That
suggests that device_create returns a device with *two* references.
How does that happen?

I'm having trouble reconciling this with the fact that device_register
followed immediately by device_unregister appears to release the
device.

Is there some reference counting subtlety I'm missing?

--Andy

-- 
Andy Lutomirski
AMA Capital Management, LLC
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/