Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <a9af5139b6b8eb687495ffae69d32acd305ac2f3.1633351482.git.geert+renesas@glider.be>
In-Reply-To: <a9af5139b6b8eb687495ffae69d32acd305ac2f3.1633351482.git.geert+renesas@glider.be>
From:   Andy Shevchenko <andy.shevchenko@gmail.com>
Date:   Mon, 4 Oct 2021 16:22:17 +0300
Message-ID: <CAHp75Vf-sRz2WMTa2SUfPr0LRnWYb=29WjcjB9rmV9Ty9WJ8zA@mail.gmail.com>
Subject: Re: [PATCH] gpio: aggregator: Wrap access to gpiochip_fwd.tmp[]
To:     Geert Uytterhoeven <geert+renesas@glider.be>
Cc:     Linus Walleij <linus.walleij@linaro.org>,
        Bartosz Golaszewski <bgolaszewski@baylibre.com>,
        Andy Shevchenko <andriy.shevchenko@linux.intel.com>,
        "open list:GPIO SUBSYSTEM" <linux-gpio@vger.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Mon, Oct 4, 2021 at 3:47 PM Geert Uytterhoeven
<geert+renesas@glider.be> wrote:
>
> The tmp[] member of the gpiochip_fwd structure is used to store both the
> temporary values bitmap and the desc pointers for operations on multiple
> GPIOs.  As both are arrays with sizes unknown at compile-time, accessing
> them requires offset calculations, which are currently duplicated in
> gpio_fwd_get_multiple() and gpio_fwd_set_multiple().
>
> Introduce (a) accessors for both arrays and (b) a macro to calculate the
> needed storage size.  This confines the layout of the tmp[] member into
> a single spot, to ease maintenance.

...

> +#define fwd_tmp_descs(fwd)     (void *)&(fwd)->tmp[BITS_TO_LONGS((fwd)->chip.ngpio)]
> +
> +#define fwd_tmp_size(ngpios)   (BITS_TO_LONGS((ngpios)) + (ngpios))

...

> -       fwd = devm_kzalloc(dev, struct_size(fwd, tmp,
> -                          BITS_TO_LONGS(ngpios) + ngpios), GFP_KERNEL);
> +       fwd = devm_kzalloc(dev, struct_size(fwd, tmp, fwd_tmp_size(ngpios)),
> +                          GFP_KERNEL);

Shouldn't we rather use devm_bitmap_zalloc() / bitmap_free()?

>         if (!fwd)
>                 return ERR_PTR(-ENOMEM);


-- 
With Best Regards,
Andy Shevchenko