linux-kernel - Re: [PATCH v2] drivers/irqchip: Enable IMX

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]

Date:   Fri, 14 Apr 2017 23:01:57 +0200 (CEST)
From:   Thomas Gleixner <tglx@...utronix.de>
To:     Andrey Smirnov <andrew.smirnov@...il.com>
cc:     linux-kernel@...r.kernel.org, yurovsky@...il.com,
        Jason Cooper <jason@...edaemon.net>,
        Marc Zyngier <marc.zyngier@....com>,
        Shawn Guo <shawnguo@...nel.org>,
        Tyler Baker <tyler.baker@...aro.org>
Subject: Re: [PATCH v2] drivers/irqchip: Enable IMX_GPCV2 on i.MX7 by
 default

On Fri, 14 Apr 2017, Andrey Smirnov wrote:

> i.MX7 has GPCv2 IP block and should be one of the main users of that
> driver. Unitl now, however, this driver was not being used by i.MX7
> code and GPCv2 block not specified in corresponding .dts(i). In light
> of the series [1] and problems that lack of the driver creates with it
> [2], enable this option when i.MX7 support is selected.

Please don't do this link dance. Explain in short precise words what this
is about. Links are only useful when there is information which is too big
to put into a change log or the mail thread contains useful context like a
sequence of steps which led to a patch.

In that case [1] links to a patch for a device tree blob. That's not really
useful to understand this patch. [2] is halfways useful, but you could
simply explain it directly.

Let me give you an example:

  i.MX7 contains a GPCv2 IP block, but it was not enabled so far because
  the required device tree parts are missing and no other code depends on
  it.

  The device tree parts and actual users are on the way to mainline, so
  it's required to select the GPCv2 driver for i.MX7.

That nicely gives the context (Paragraph 1) and the reason (P 2).

But, this patch is actually wrong. If the SoC or other code parts depend on
it then these parts should simply add a

   select IMX_GPCV2

statement to the relevant Kconfig. That avoids endless

 default y if FOO || (BLA && !MEH) || BRR .....

lines and enforces the enablement when the actual user is selected.

> [1] https://lkml.org/lkml/2017/4/13/327
> [2] https://lkml.org/lkml/2017/4/13/646

Please do not use lkml.org references. If you ever want to link to a mail
thread then please use http://lkml.kernel.org/r/$Message-Id. That is a
redirector to a known to be available mail archive.

Thanks,

	tglx