Received-SPF: pass (google.com: domain of linux-kernel+bounces-65945-linux.lists.archive=gmail.com@vger.kernel.org designates 147.75.199.223 as permitted sender) client-ip=147.75.199.223;
Date: Wed, 14 Feb 2024 12:44:09 -0800
From: "Paul E. McKenney" <paulmck@kernel.org>
To: Bartosz Golaszewski <brgl@bgdev.pl>
Cc: Linus Walleij <linus.walleij@linaro.org>,
	Kent Gibson <warthog618@gmail.com>, Alex Elder <elder@linaro.org>,
	Geert Uytterhoeven <geert+renesas@glider.be>,
	Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
	Wolfram Sang <wsa@the-dreams.de>, Mark Brown <broonie@kernel.org>,
	Dan Carpenter <dan.carpenter@linaro.org>,
	linux-gpio@vger.kernel.org, linux-kernel@vger.kernel.org,
	Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
Subject: Re: [PATCH v2 0/4] gpio: fix SRCU bugs
Message-ID: <ef420f58-cf5a-4bd3-adc4-b9ef651498ca@paulmck-laptop>
Reply-To: paulmck@kernel.org
References: <20240214084419.6194-1-brgl@bgdev.pl>
 <4a08c610-c249-4b56-b09c-09e1fd07f723@paulmck-laptop>
 <CAMRc=McXr_3OD6Uf+PKD-wLQfJKJSWes3fh_ZD9fnrPyF9GoSQ@mail.gmail.com>
Precedence: bulk
MIME-Version: 1.0
Content-Type: text/plain; charset=utf-8
Content-Disposition: inline
Content-Transfer-Encoding: 8bit
In-Reply-To: <CAMRc=McXr_3OD6Uf+PKD-wLQfJKJSWes3fh_ZD9fnrPyF9GoSQ@mail.gmail.com>

On Wed, Feb 14, 2024 at 08:08:33PM +0100, Bartosz Golaszewski wrote:
> On Wed, Feb 14, 2024 at 7:44 PM Paul E. McKenney <paulmck@kernel.org> wrote:
> >
> > On Wed, Feb 14, 2024 at 09:44:15AM +0100, Bartosz Golaszewski wrote:
> > > From: Bartosz Golaszewski <bartosz.golaszewski@linaro.org>
> > >
> > > Here are four fixes to some bugs in recent SRCU changes. The first one fixes
> > > an actual race condition. The other three just make lockdep happy.
> >
> > For 1/4-3/4:
> >
> > Acked-by: Paul E. McKenney <paulmck@kernel.org>
> >
> > For 4/4, you are playing with fire, but I will assume that you know what
> > you are doing.  ;-)
> 
> Up until this rework, this gdev->chip pointer could go from under any
> user at any point. Now we have this gpio_device wrapper that provides
> an entry point to using the chip safely while protected by the SRCU
> read lock. Anyone who is still accessing gpio_chip directly (and not
> being the GPIO provider themselves) is asking for trouble. There's
> however no point in spamming lockdep splats in this case. I may end up
> adding a warning to these routines.
> 
> Unfortunately, it's hard to fix 15 years of technical debt. :(

Indeed, life is sometimes hard!

							Thanx, Paul

> Thanks for the Acks.
> Bartosz
> 
> [snip]