linux-kernel - Re: [PATCH 3/3] gpio: regmap: Support a custom ->to

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]

Message-ID: <EP4LzL5PhvL2RLOodKu5K24zNSfLTjAi@localhost>
Date:   Mon, 04 Jul 2022 17:38:00 +0100
From:   Aidan MacDonald <aidanmacdonald.0x0@...il.com>
To:     Andy Shevchenko <andy.shevchenko@...il.com>
Cc:     Michael Walle <michael@...le.cc>,
        Linus Walleij <linus.walleij@...aro.org>,
        Bartosz Golaszewski <brgl@...ev.pl>,
        "open list:GPIO SUBSYSTEM" <linux-gpio@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH 3/3] gpio: regmap: Support a custom ->to_irq() hook

Andy Shevchenko <andy.shevchenko@...il.com> writes:

> On Sun, Jul 3, 2022 at 1:11 PM Aidan MacDonald
> <aidanmacdonald.0x0@...il.com> wrote:
>>
>> Some GPIO chips require a custom to_irq() callback for mapping
>> their IRQs, eg. because their interrupts come from a parent IRQ
>> chip where the GPIO offset doesn't map 1-to-1 with hwirq number.
>
> Don't they follow a hierarchical IRQ domain in that case?
>
> And to be honest after the commit ef38237444ce ("gpiolib: add a
> warning on gpiochip->to_irq defined") I have no idea how it works in
> your case and also I feel this patch is a wrong direction of
> development.

My own use case is an MFD device with a shared IRQ chip that is
used by other sub-drivers. This is a very common case that seems
to map onto ->to_irq() cleanly. Do we really need an IRQ domain?
What you're suggesting would be a 1-to-1 mapping from GPIO offset
to hwirq number in a virtual domain, then remapping to the real
hwirq number, which seems unnecessarily complicated when we can
just change the GPIO offset -> hwirq mapping.

The commit you mentioned is warning users of GPIOLIB_IRQCHIP when a
custom ->to_irq() method is overridden. That's not relevant here.
Using an IRQ domain also overrides ->to_irq() so I included a check
in this patch to ensure gpio-regmap chips are well-behaved.