linux-kernel - Re: [PATCH 4/5] iio: adc: xilinx: use devres for irq handling

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:   Fri, 30 Oct 2020 11:52:00 +0100
From:   Bartosz Golaszewski <brgl@...ev.pl>
To:     Jonathan Cameron <jic23@...nel.org>
Cc:     Lars-Peter Clausen <lars@...afoo.de>,
        Peter Meerwald-Stadler <pmeerw@...erw.net>,
        Michal Simek <michal.simek@...inx.com>,
        linux-iio <linux-iio@...r.kernel.org>,
        Linux ARM <linux-arm-kernel@...ts.infradead.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        Bartosz Golaszewski <bgolaszewski@...libre.com>
Subject: Re: [PATCH 4/5] iio: adc: xilinx: use devres for irq handling

On Thu, Oct 29, 2020 at 4:41 PM Jonathan Cameron <jic23@...nel.org> wrote:
>
> On Mon, 26 Oct 2020 14:36:08 +0100
> Bartosz Golaszewski <brgl@...ev.pl> wrote:
>
> > From: Bartosz Golaszewski <bgolaszewski@...libre.com>
> >
> > Further simplify the remove() callback and error paths in probe() by
> > using the managed variant of request_irq() as well as using a devm action
> > for cancelling the delayed work at driver detach.
> >
> > Signed-off-by: Bartosz Golaszewski <bgolaszewski@...libre.com>
>
> Again, this is potentially fine but I'd rather you cleaned up the ordering first
> rather than doing things in this order.
>
> The end result of the whole series looks like it will be correct, but that isn't
> so obvious for the intermediate patches on their own.
>
> Also, you end up with a lot of noise renaming gotos that then go away at the
> end.
>
> Jonathan
>

Hi Jonathan,

My two priorities for the ordering of this series were: correct
end-result and not breaking anything on the way. The latter
unfortunately gets in the way of cleaner looking intermediate patches.

I tried to not alter the ordering in which the resources are freed at
any step. As devres release callbacks are called *after* remove() and
in a reverse order to how they were registered, I needed to start from
the bottom of the remove() callback and convert the last operation,
then go upwards from there.

If I tried to do it from the top - I probably could remove labels
earlier and in a cleaner manner but it wouldn't guarantee
bisectability.

Bartosz