linux-kernel - Re: [PATCH] tty: max310x: work around regmap->regcache data corruption

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, 1 Dec 2023 16:38:46 -0500
From:   Hugo Villeneuve <hugo@...ovil.com>
To:     Mark Brown <broonie@...nel.org>
Cc:     Jan Kundrát <jan.kundrat@...net.cz>,
        Cosmin Tanislav <cosmin.tanislav@...log.com>,
        linux-serial@...r.kernel.org,
        Andy Shevchenko <andy.shevchenko@...il.com>,
        linux-kernel@...r.kernel.org
Subject: Re: [PATCH] tty: max310x: work around regmap->regcache data
 corruption

On Fri, 1 Dec 2023 18:34:38 +0000
Mark Brown <broonie@...nel.org> wrote:

> On Fri, Dec 01, 2023 at 01:27:36PM -0500, Hugo Villeneuve wrote:
> 
> > it is funny, as I am preparing to send a patch for the sc16is7xx driver
> > to convert FIFO R/W to use the _noinc_ versions of regmap functions,
> > inspired by your patch 3f42b142ea11 ("serial: max310x: fix IO data
> > corruption in batched operations").
> 
> If you're working on that driver it'd also be good to update the current
> use of cache bypass for the enhanced features/interrupt identification
> register (and anything else in there, that did seem to be the only one)
> to use regmap ranges instead - that'd remove the need for the efr_lock
> and be a much more sensible/idiomatic use of the regmap APIs.

Hi Mark,
agreed, and I have already removed all cache bypass code (after some
fix for volatile registers)...

I will also look to remove the efr_lock, altough it has more
implications since this ship has some registers that share a common
address, and selected by bits in other registers, and I think this
is why there is this efr_lock.

I need to run more tests to make sure everything is ok, but so far
so good, and I should be submitting some of these patches soon.

Hugo Villeneuve.