linux-kernel - Re: [PATCH 1/1] gpio: add a driver for the Synopsys DesignWare APB GPIO block

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

Date:	Thu, 7 Nov 2013 15:06:04 -0600
From:	delicious quinoa <delicious.quinoa@...il.com>
To:	Sebastian Hesselbarth <sebastian.hesselbarth@...il.com>
Cc:	Jamie Iles <jamie@...ieiles.com>,
	linux-kernel <linux-kernel@...r.kernel.org>,
	linux-gpio@...r.kernel.org,
	Linus Walleij <linus.walleij@...ricsson.com>,
	"linux-doc@...r.kernel.org" <linux-doc@...r.kernel.org>,
	"devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
	Grant Likely <grant.likely@...retlab.ca>,
	Rob Herring <rob.herring@...xeda.com>,
	Steffen Trumtrar <s.trumtrar@...gutronix.de>,
	Heiko Stuebner <heiko@...ech.de>, Alan Tull <atull@...era.com>,
	Dinh Nguyen <dinguyen@...era.com>,
	Yves Vandervennet <rocket.yvanderv@...il.com>
Subject: Re: [PATCH 1/1] gpio: add a driver for the Synopsys DesignWare APB
 GPIO block

On Wed, Nov 6, 2013 at 5:44 PM, Sebastian Hesselbarth
<sebastian.hesselbarth@...il.com> wrote:
>>> +struct dwapb_gpio {
>>> +       struct  device          *dev;
>>> +       void __iomem            *regs;
>>> +       struct dwapb_gpio_port  *ports;
>>> +       unsigned int            nr_ports;
>>> +       struct irq_domain       *domain;
>>> +       int                     hwirq;
>>
>>
>> I'm not sure I fully understand what hwirq is in this context - is it
>> the IRQ line from the Synopsys block to the system interrupt controller?
>> If so I don't think this covers all configurations - the Picochip
>> devices for instance have each GPIO in port A as an individual IRQ going
>> to the VIC.
>
>
> Usually, hwirq is the interrupt as seen from this very device, i.e. in
> this case it is 0 up to 32 referencing the portA gpio line that
> triggered an interrupt. In this drivers context 'hwirq' above is
> actually the virtual irq number Linux made up to identify an interrupt
> coming from this IP block.
>
> Therefore, above should really be irq instead of hwirq - but with a
> more detailed review, I think it can be removed completely.
>
>> It looks here like hwirq is used for all of the interrupt registers so
>> only one GPIO interrupt is supported?
>
>
> Looking in _probe, currently only one upstream interrupt is requested.
> Jamie is right, that you should grab all interrupts - but you can forget
> their virtual number after registering the handler. There are helper
> functions to get it back in irq handler.
>

The IP configuration I have has only one irq line going to the gic,
(that is the "hwirq" here) so I had one hwirq (the gic) for each IP
instance.  The domain maps a Linux IRQ number for each gpio line.

I will need to rework this to also support both this (one irq per IP
block) and the picochip (one irq per gpio line coming out of the IP
block).

>>> +static void __exit dwapb_gpio_exit(void)
>>> +{
>>> +       platform_driver_unregister(&dwapb_gpio_driver);
>>> +}
>>> +module_exit(dwapb_gpio_exit);
>>
>>
>> We can replace the registration and unregistration with
>> module_platform_driver() now.
>
>
> +1 (again)

OK

Alan

>
> Sebastian
>
--
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/