linux-kernel - Re: [PATCH v2 2/3] gpio: Add gpio driver support for ThunderX and OCTEON-TX

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] [day] [month] [year] [list]

Date:   Wed, 11 Jan 2017 16:07:24 +0100
From:   Linus Walleij <linus.walleij@...aro.org>
To:     David Daney <ddaney@...iumnetworks.com>
Cc:     David Daney <ddaney.cavm@...il.com>,
        Alexandre Courbot <gnurou@...il.com>,
        Rob Herring <robh+dt@...nel.org>,
        Mark Rutland <mark.rutland@....com>,
        "linux-gpio@...r.kernel.org" <linux-gpio@...r.kernel.org>,
        "devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        David Daney <david.daney@...ium.com>,
        Marc Zyngier <marc.zyngier@....com>
Subject: Re: [PATCH v2 2/3] gpio: Add gpio driver support for ThunderX and OCTEON-TX

On Mon, Jan 9, 2017 at 9:02 PM, David Daney <ddaney@...iumnetworks.com> wrote:

>> if (test_bit(line, gpio->invert_mask))
>>    return !(read_bits & BIT(bank_bit));
>> else
>>    return !!(read_bits & BIT(bank_bit));
>>
>> OK maybe not much clearer but seems clearer to me.
>
> As I really dislike the "!!" idiom, would you settle for:
>
>  if (test_bit(line, gpio->invert_mask))
>     return (read_bits & BIT(bank_bit)) == 0;
>  else
>     return (read_bits & BIT(bank_bit)) != 0;

Not the biggest issue in the world. But I maintain a huge stack
of GPIO drivers and it drives me crazy that each one has to bear
the mark of the authors habits rather than mine.

>> I think this is overkill. Use hierarchical irqdomain.
>
> I will look into it.  I suspect it will require more lines of driver code to
> implement it than what I have here (that does actually work).

I understand. But at the same time, the kernel needs to have the
right idea of what it is dealing with here.

The generic IRQ handling code will take a shorter fastpath if
you are using hierarchical irqdomain (I think?) but I can't claim
to be an expert. When in doubt, consult Marc Z.

Yours,
Linus Walleij