linux-kernel - Re: [PATCH v2 2/2] gpio: mlxbf3: Support add_pin

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: <CAHp75VcEe-2+=2qXCFHp+cxN=nWcsc=oCVYUeinM_cMsPES+mQ@mail.gmail.com>
Date:   Thu, 17 Aug 2023 11:21:52 +0300
From:   Andy Shevchenko <andy.shevchenko@...il.com>
To:     Asmaa Mnebhi <asmaa@...dia.com>
Cc:     "linux-gpio@...r.kernel.org" <linux-gpio@...r.kernel.org>,
        "linus.walleij@...aro.org" <linus.walleij@...aro.org>,
        "bgolaszewski@...libre.com" <bgolaszewski@...libre.com>,
        "brgl@...ev.pl" <brgl@...ev.pl>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v2 2/2] gpio: mlxbf3: Support add_pin_ranges()

On Thu, Aug 17, 2023 at 12:28 AM Asmaa Mnebhi <asmaa@...dia.com> wrote:
>
>  > > v1->v2:
> > > - No changes.
> >
> > Is this correct?
> Ah my apologies, I added the wrong comment here. I put it in "v2 1/2".
>
> > > +static int mlxbf3_gpio_add_pin_ranges(struct gpio_chip *chip) {
> > > +       unsigned int id = 0;
> > > +       int ret;
> > > +
> > > +       if (chip->ngpio % MLXBF3_GPIO_MAX_PINS_PER_BLOCK)
> > > +               id = 1;
> >
> > This id calculation seems wrong to me as I said in v1 review.
> > Why do you think the above is what you want and not just working by luck?
>
> I would like to get the gpio block id which can only be 0 or 1 on BlueField-3 (only 2 gpio blocks, one with 32 gpio pins and one with 24 gpio pins).
> The above logic was an "easy" way for me to get the gpio block id. Then the pin_base for each gpio block is:
> pin_base = id * MLXBF3_GPIO_MAX_PINS_PER_BLOCK

It's fragile. Use a direct case switch for that, which will be more
explicit and robust (however still can fail for any new chip
revision/version where it might be a different GPIO layout).

-- 
With Best Regards,
Andy Shevchenko