linux-kernel - Re: [PATCH v2 10/10] gpiolib: remove gpiochip_is

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: <ZWjN_oYS3JANyx7u@smile.fi.intel.com>
Date:   Thu, 30 Nov 2023 20:01:34 +0200
From:   Andy Shevchenko <andriy.shevchenko@...ux.intel.com>
To:     Bartosz Golaszewski <brgl@...ev.pl>
Cc:     Linus Walleij <linus.walleij@...aro.org>,
        linux-gpio@...r.kernel.org, linux-kernel@...r.kernel.org,
        Bartosz Golaszewski <bartosz.golaszewski@...aro.org>
Subject: Re: [PATCH v2 10/10] gpiolib: remove gpiochip_is_requested()

On Thu, Nov 30, 2023 at 06:46:20PM +0100, Bartosz Golaszewski wrote:
> On Thu, Nov 30, 2023 at 5:46 PM Andy Shevchenko
> <andriy.shevchenko@...ux.intel.com> wrote:
> > On Thu, Nov 30, 2023 at 02:46:30PM +0100, Bartosz Golaszewski wrote:

...

> > > -     cpy = kstrdup(label, GFP_KERNEL);
> > > -     if (!cpy)
> > > -             return ERR_PTR(-ENOMEM);
> > > +     scoped_guard(spinlock_irqsave, &gpio_lock) {
> > > +             if (!test_bit(FLAG_REQUESTED, &desc->flags))
> > > +                     return NULL;
> >
> > > +             cpy = kstrdup(desc->label, GFP_KERNEL);
> > > +             if (!cpy)
> > > +                     return ERR_PTR(-ENOMEM);
> >
> > You just introduced these lines earlier in the series, and here you moved
> > them again. With guard() instead it may be kept in a better shape.
> >
> 
> I wanted to limit the critical section to a minimum hence scoped
> variant. And this will go away as soon as we have a desc lock so it's
> temporary anyway. What matters to me is how the code looks when
> sending it to Torvalds. On the off chance that we don't get the
> locking rework merged in time for v6.8, I want this to at least be
> under the existing lock.

guard() here is equally scoped, no? And what's wrong with that when gets
to Torvalds? He accepted your guard() cases last time IIRC.

> > > +     }

-- 
With Best Regards,
Andy Shevchenko