linux-kernel - Re: [PATCHv2 6/6] ASoC: da7213: Add default clock handling

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: <20191128131341.GD4210@sirena.org.uk>
Date:   Thu, 28 Nov 2019 13:13:41 +0000
From:   Mark Brown <broonie@...nel.org>
To:     Adam Thomson <Adam.Thomson.Opensource@...semi.com>
Cc:     Sebastian Reichel <sebastian.reichel@...labora.com>,
        Support Opensource <Support.Opensource@...semi.com>,
        Liam Girdwood <lgirdwood@...il.com>,
        Jaroslav Kysela <perex@...ex.cz>,
        Takashi Iwai <tiwai@...e.com>,
        "alsa-devel@...a-project.org" <alsa-devel@...a-project.org>,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>,
        "kernel@...labora.com" <kernel@...labora.com>
Subject: Re: [PATCHv2 6/6] ASoC: da7213: Add default clock handling

On Wed, Nov 27, 2019 at 06:10:00PM +0000, Adam Thomson wrote:
> On 27 November 2019 16:41, Mark Brown wrote:

> > The PLL configuration specifies both input and output clock rates (as
> > well as an input clock source) so if it's got to configure the MCLK I'd
> > expect the driver to figure that out without needing the caller to
> > separately set the MCLK rate.

> Yes it does but the name of the function implies it's setting the codec's PLL,
> not the system clock, whereas the other function implies setting the system
> clock and not the PLL. Also generally you're only setting the sysclk once
> whereas you may want to configure and enable/disable the PLL more dynamically,
> at least for devices which do have a built-in PLL. Of course that could still be
> handled through the single PLL function call.

I wouldn't be so sure about only setting the SYSCLK once - if you've got
an input clock you can configure then you might for example want to
switch between 44.1kHz and 48kHz bases, and disable it or run it at very
low frequency when idle.  In some systems it's as dynamic as a PLL is.

> Just as an informational, what's the future for these two functions if
> essentially one is only really required and the other deemed redundant? I would
> just like to be clear so I'm not falling over things like this in the future,
> and wasting your time as well. Seems that the PLL call isn't part of simple
> generic card code so would the be deemed surplus to requirements some point down
> the line?

Things like simple-card are good for 90% of systems but I'm fairly sure
we'll never be able to handle more complex setups entirely
automatically.  What we *should* be doing over time is transitioning all
this clock stuff into the actual clock framework but that's a big, long
term thing which nobody is currently actually working on.

Download attachment "signature.asc" of type "application/pgp-signature" (489 bytes)