linux-kernel - Re: [PATCH v1 4/5] gpio: max77620: Don't handle disabled interrupts

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: <d39caa8f-816c-5d4d-6f54-99baea3e0d5a@gmail.com>
Date:   Wed, 8 Jul 2020 13:54:47 +0300
From:   Dmitry Osipenko <digetx@...il.com>
To:     Andy Shevchenko <andy.shevchenko@...il.com>
Cc:     Thierry Reding <thierry.reding@...il.com>,
        Jonathan Hunter <jonathanh@...dia.com>,
        Laxman Dewangan <ldewangan@...dia.com>,
        Bartosz Golaszewski <bgolaszewski@...libre.com>,
        Linus Walleij <linus.walleij@...aro.org>,
        linux-tegra@...r.kernel.org,
        "open list:GPIO SUBSYSTEM" <linux-gpio@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v1 4/5] gpio: max77620: Don't handle disabled interrupts

08.07.2020 13:11, Andy Shevchenko пишет:
> On Wed, Jul 8, 2020 at 12:19 PM Dmitry Osipenko <digetx@...il.com> wrote:
>> 08.07.2020 11:46, Andy Shevchenko пишет:
>>> On Wed, Jul 8, 2020 at 11:29 AM Dmitry Osipenko <digetx@...il.com> wrote:
>>>>
>>>> Check whether GPIO IRQ is enabled before proceeding with handling the
>>>> interrupt request. The interrupt handler now returns IRQ_NONE if none
>>>> of interrupts were handled, which is usually a sign of a problem.
>>>
>>> ...
>>>
>>>> -       pending = value;
>>>> +       pending = value & gpio->irq_enb_mask;
>>>
>>>> +       if (!pending)
>>>> +               return IRQ_NONE;
>>>
>>> for_each_set_bit() should take care of it, no?
>>
>> Do you mean that the handle_nested_irq() takes care of handling
>> unrequested interrupts? Actually, looks like it cares. Alright, I'll
>> drop this patch since it should be unnecessary. Thank you for the comment!
> 
> I think it's still good to have reduced IRQs to handle by dropping not
> enabled ones, my comment was about the case when pending == 0. Sorry
> if it was unclear.

It should be unnecessary since we now see that the handle_nested_irq()
checks whether interrupt was requested and if it wasn't, then particular
GPIO interrupt will be treated as spurious [1]. The pending == 0
condition is an extreme case, I don't think that there is a need to
optimize it without any good reason.

[1] https://elixir.bootlin.com/linux/v5.8-rc3/source/kernel/irq/chip.c#L485

Hence it should be better to drop this patch.