linux-kernel - Re: [PATCH v3 1/4] i2c: mux-pinctrl: Rework to honor disabled child nodes

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, 27 Mar 2015 22:08:06 +0100
From:	Sebastian Hesselbarth <sebastian.hesselbarth@...il.com>
To:	Wolfram Sang <wsa@...-dreams.de>
CC:	Stephen Warren <swarren@...dotorg.org>,
	Jason Cooper <jason@...edaemon.net>,
	Andrew Lunn <andrew@...n.ch>,
	Gregory Clement <gregory.clement@...e-electrons.com>,
	Gabriel Dobato <dobatog@...il.com>, linux-i2c@...r.kernel.org,
	devicetree@...r.kernel.org, linux-arm-kernel@...ts.infradead.org,
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH v3 1/4] i2c: mux-pinctrl: Rework to honor disabled child
 nodes

On 23.03.2015 19:32, Wolfram Sang wrote:
>> If modifying i2c-mux-pinctrl to respect the sub-bus status property is
>> such a big issue, I'd rather leave the driver as is and expose all
>> sub-busses to userspace.
>
> Well, dunno what 'big issue' is in your book :) What definately needs to
> be done is:

Wolfram,

I had a look at the code in question again and prepared some patches
to return ERR_PTRs from i2c_add_mux_adapter(), rework users to check
for IS_ERR() and finally skip i2c_add_adapter for the OF disabled case.

I have them compile-tested but I don't have any hardware available
right now.

Anyway, I still have some questions.

> * handle "status" at mux-core level, not mux-driver

I get that.. but:

> * that probably needs us to convert i2c_add_mux_adapter() to return
>    ERR_PTRs instead of NULL, so we can distinguish the "disabled" case

What do you want to return from i2c_add_mux_adapter() if you find an OF
disabled child node?

AFAIKS, there is no explicit errno for a disabled device (node) so all
I can imagine here is to return either the &priv->adap before actually
calling i2c_add_adapter on it or NULL now that we explicitly check for
errors.

> * that would mean to fix all existing users

If we choose to return NULL, those users who can deal with a
missing/disabled sub-bus on the mux can check with IS_ERR()
the others should check with IS_ERR_OR_NULL().

We could also choose to return some other errno (-ENODEV maybe)
but still we'd have to double-check that return value on i2c-mux-pinctrl
and the others too if we don't care that i2c_add_adapter() wasn't
called for a mux.

> That's all not groundbreaking stuff, but needs caution and thoroughness.
> There might be some gory details left, though...

As I said before, the intention was not disable a possible i2c bus that
can be dynamically added/removed on that specific Dove SBC/CM-A510
board but to have a single i2c-mux-pinctrl in the SoC dtsi just because
the SoC has the optional i2c bus muxing.

While thinking about it (and I still think of it as a 'big issue'
compared to the intention of the initial patch) I came to the
conclusion that I should maybe just go for a board-specific
i2c-mux-pinctrl node instead of adding it to the SoC dtsi. That will
also avoid doubled i2c busses on boards with just the default i2c
option.

Sebastian
--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/