linux-kernel - Re: [PATCH 1/1] gpio: vf610: make irq

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: <86wn4jxjph.wl-maz@kernel.org>
Date:   Wed, 15 Feb 2023 11:16:10 +0000
From:   Marc Zyngier <maz@...nel.org>
To:     Linus Walleij <linus.walleij@...aro.org>
Cc:     Alexander Stein <alexander.stein@...tq-group.com>,
        Bartosz Golaszewski <brgl@...ev.pl>,
        linux-gpio@...r.kernel.org, linux-kernel@...r.kernel.org
Subject: Re: [PATCH 1/1] gpio: vf610: make irq_chip immutable

On Wed, 15 Feb 2023 10:19:28 +0000,
Linus Walleij <linus.walleij@...aro.org> wrote:
> 
> On Tue, Feb 14, 2023 at 8:36 AM Alexander Stein
> <alexander.stein@...tq-group.com> wrote:
> 
> > Since recently, the kernel is nagging about mutable irq_chips:
> >
> >     "not an immutable chip, please consider fixing it!"
> >
> > Drop the unneeded copy, flag it as IRQCHIP_IMMUTABLE, add the new
> > helper functions and call the appropriate gpiolib functions.
> >
> > Signed-off-by: Alexander Stein <alexander.stein@...tq-group.com>
> 
> Looks good to me, CC to Marc Z.

Looks wrong to me. This is missing the explicit callbacks into gpiolib
so that it knows what gets enabled/disabled on mask/unmask.

>
> Reviewed-by: Linus Walleij <linus.walleij@...aro.org>
> 
> We fixed quite a few of these now, Marc do you have an idea about
> how much we have left until we can make immutable the default?

I haven't tracked that, and making it the default would probably mean
getting rid of the code that patches the irqchip structures. I'd say
that once -rc1 is out, we replace the polite nag with something
nastier (WARN_ON() of some sort), and push that into -next.

Leave the warning in place for a couple of releases (until the next
LTS), and then drop the patching code. The not-so-nice part is that
that drivers that haven't been fixed will break silently. The good
side is that these drivers will not have been touched over 2 LTS
releases, and are thus most likely abandonware.

Thanks,

	M.

-- 
Without deviation from the norm, progress is not possible.