linux-kernel - Re: [PATCH 2/2] gpiolib: Defer gpio device setup until after gpiolib initialization

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:	Tue, 5 Apr 2016 09:50:17 +1000
From:	Greg Ungerer <gregungerer@...tnet.com.au>
To:	Linus Walleij <linus.walleij@...aro.org>
Cc:	Guenter Roeck <linux@...ck-us.net>,
	"linux-gpio@...r.kernel.org" <linux-gpio@...r.kernel.org>,
	"linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
	Alexandre Courbot <gnurou@...il.com>,
	Greg Ungerer <gerg@...inux.org>
Subject: Re: [PATCH 2/2] gpiolib: Defer gpio device setup until after gpiolib
 initialization

Hi Linus,

On 01/04/16 18:16, Linus Walleij wrote:
> On Fri, Apr 1, 2016 at 3:33 AM, Greg Ungerer <gregungerer@...tnet.com.au> wrote:
>> On 01/04/16 10:53, Guenter Roeck wrote:
>
>>> Probably coldfire ?
>>>
>>> arch/m68k/coldfire/gpio.c:
>>
>> Yes, that is the one.
>>
>>> static struct bus_type mcfgpio_subsys = {
>>>          .name           = "gpio",
>>>        .dev_name       = "gpio",
>>> };
>>>
>>> No idea what to do about that. Can that bus be renamed ?
>>
>> Yeah, it could. But is it even necessary at all now?
>>
>> I am thinking I could remove the subsys_system_register(&mcfgpio_subsys, NULL)
>> call from that coldfire/gpio.c. Doing that certainly cleans up the
>> boot. There was nothing much under the old coldfire /sys/gpio other than
>> the standard devices/drivers/etc. And the new gpio api ofcourse has all
>> that as well.
>
> Please kill it if you can. Or are there userspaces for
> coldfire that use this? In that case we need some compatibility
> shim for them.

I don't see any reason we can't just kill it.
It never had any real links to anything under it, only
the usual set of bus sysfs nodes. So the new "/sys/bus/gpio"
has all that and more. So no need for a shim.


> I'm trying to pull all needed functionality into the proper gpiolib
> so we can get some order around the place...

Yep :-)
I'll prepare a patch to remove the coldfire gpio sys bus device.
It is pretty simple in this case.

Regards
Greg