Received-SPF: pass (google.com: domain of linux-kernel+bounces-6895-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.80.249 as permitted sender) client-ip=147.75.80.249;
Precedence: bulk
MIME-Version: 1.0
References: <20231220015106.16732-1-warthog618@gmail.com> <20231220015106.16732-5-warthog618@gmail.com>
 <CACRpkdZ5HzOxtbexQNE-A-bKhiUW1nHjvJQA_CCnmVXf+R6dbg@mail.gmail.com>
 <CAMRc=MfvKzOxPrmz1wmgWMwYUbNhWAjqoKOmcaggQntcDprLmQ@mail.gmail.com>
 <ZYLaayENrvL1Nh6H@rigel> <CAMRc=MfyCBpZ07SYfxMtug6FVYiKA0MRgvjMTOAzKiVLGdPM+w@mail.gmail.com>
 <ZYLczeiVDjd2cWQF@rigel>
In-Reply-To: <ZYLczeiVDjd2cWQF@rigel>
From: Bartosz Golaszewski <brgl@bgdev.pl>
Date: Wed, 20 Dec 2023 13:30:57 +0100
Message-ID: <CAMRc=MeXa5g6iQNYF4W+vGL+kgRTyVjFB-yXE_UBpuTnn2ZKng@mail.gmail.com>
Subject: Re: [PATCH 4/4] gpiolib: cdev: replace locking wrappers for
 gpio_device with guards
To: Kent Gibson <warthog618@gmail.com>
Cc: Linus Walleij <linus.walleij@linaro.org>, Peter Zijlstra <peterz@infradead.org>, 
	linux-kernel@vger.kernel.org, linux-gpio@vger.kernel.org, andy@kernel.org
Content-Type: text/plain; charset="UTF-8"
Content-Transfer-Encoding: quoted-printable

On Wed, Dec 20, 2023 at 1:23=E2=80=AFPM Kent Gibson <warthog618@gmail.com> =
wrote:
>
> On Wed, Dec 20, 2023 at 01:16:00PM +0100, Bartosz Golaszewski wrote:
> > On Wed, Dec 20, 2023 at 1:13=E2=80=AFPM Kent Gibson <warthog618@gmail.c=
om> wrote:
> > >
> > > On Wed, Dec 20, 2023 at 01:05:35PM +0100, Bartosz Golaszewski wrote:
> > > > On Wed, Dec 20, 2023 at 12:56=E2=80=AFPM Linus Walleij <linus.walle=
ij@linaro.org> wrote:
> > > > >
> > > > > (+PeterZ)
> > > > >
> > > > > On Wed, Dec 20, 2023 at 2:52=E2=80=AFAM Kent Gibson <warthog618@g=
mail.com> wrote:
> > > > >
> > > > > > Replace the wrapping functions that inhibit removal of the gpio=
_device
> > > > > > with equivalent guard macros.
> > > > > >
> > > > > > Signed-off-by: Kent Gibson <warthog618@gmail.com>
> > > > > (...)
> > > > > > +DEFINE_CLASS(_read_sem_guard,
> > > > > > +            struct rw_semaphore *,
> > > > > > +            up_read(_T),
> > > > > > +            ({
> > > > > > +               down_read(sem);
> > > > > > +               sem;
> > > > > > +            }),
> > > > > > +            struct rw_semaphore *sem);
> > > > >
> > > > > Isn't this so generic that it should be in <linux/cleanup.h>?
> > > > >
> > > > > Otherwise all the patches look good to me.
> > > > >
> > > >
> > > > We already have this:
> > > >
> > > > DEFINE_GUARD(rwsem_read, struct rw_semaphore *, down_read(_T), up_r=
ead(_T))
> > > > DEFINE_GUARD(rwsem_write, struct rw_semaphore *, down_write(_T), up=
_write(_T))
> > > >
> > > > DEFINE_FREE(up_read, struct rw_semaphore *, if (_T) up_read(_T))
> > > > DEFINE_FREE(up_write, struct rw_semaphore *, if (_T) up_write(_T))
> > > >
> > >
> > > Ah - in rwsem.h - I missed that.
> > >
> > > > This can surely be used here, right?
> > > >
> > >
> > > Don't see why not.
> > >
> > > I would still like to move the gpio_device specific macros to gpiolib=
.h,
> > > as they apply to the struct gpio_device defined there.
> >
> > Which ones? Because I'd rather use guard(rwsem_read)(&gdev->sem); than
> > some custom wrapper as this one's purpose is clearer.
> >
>
> It would be read and write guards for the gpio_device.
> cdev would only be using the read flavour.
> And possibly named something other than read/write as the purpose is to
> prevent (read) or allow (write) object removal.
>
> I though that would be clearer than having to reference gpiolib.h to see
> what gdev->sem covers, and allow you to change the locking
> mechanism later and not have to update cdev.
>

I still prefer open-coded guards here for clarity. I hope that with
SRCU in gpiolib.c, we'll get rid of locking in cdev entirely anyway.

Bart

> Cheers,
> Kent.
>