linux-kernel - Re: regmap: mmio: regression in pre-v4.6-rc1

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]

Message-ID: <20160323120535.GK2566@sirena.org.uk>
Date:	Wed, 23 Mar 2016 12:05:35 +0000
From:	Mark Brown <broonie@...nel.org>
To:	Alexander Stein <alexander.stein@...tec-electronic.com>
Cc:	linux-kernel@...r.kernel.org
Subject: Re: regmap: mmio: regression in pre-v4.6-rc1

On Wed, Mar 23, 2016 at 12:50:53PM +0100, Alexander Stein wrote:

As I said previously please fix your mail client to word wrap within
paragraphs at something substantially less than 80 columns.  Doing this
makes your messages much easier to read and reply to.

> > > In regmap_mmio_gen_context ctx->reg_read is set to regmap_mmio_read32le and 
> > > ctx->reg_write to regmap_mmio_write32le respectively.

> > So how does that happen then?  We set these values if the bus is
> > default, little or native endian but if it's big endian we go into a
> > completely different case...

> Well, in regmap_mmio_gen_context config->reg_format_endian is still
> set to REGMAP_ENDIAN_DEFAULT. of_syscon_register sets
> config.val_format_endian (notice val_ instead of reg_) depending on
> "big-endian" (or "little-endian") property.  I'm kinda confused
> regarding reg_format_endian and val_format_endian. Dunno what should
> be set in which way.

Ah, I see.  That's definitely broken - if we're changing the value
format we should be checking the value.

Download attachment "signature.asc" of type "application/pgp-signature" (474 bytes)