linux-kernel - Re: [PATCH] firmware: meson-sm: use generic compatible

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, 20 Oct 2017 10:30:53 +0200
From:   Jerome Brunet <jbrunet@...libre.com>
To:     Rob Herring <robh@...nel.org>, Kevin Hilman <khilman@...libre.com>
Cc:     Carlo Caione <carlo@...one.org>,
        "devicetree@...r.kernel.org" <devicetree@...r.kernel.org>,
        "linux-arm-kernel@...ts.infradead.org" 
        <linux-arm-kernel@...ts.infradead.org>,
        linux-amlogic@...ts.infradead.org,
        "linux-kernel@...r.kernel.org" <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH] firmware: meson-sm: use generic compatible

On Thu, 2017-10-19 at 16:18 -0500, Rob Herring wrote:
> On Thu, Oct 19, 2017 at 5:25 AM, Kevin Hilman <khilman@...libre.com> wrote:
> > Rob Herring <robh@...nel.org> writes:
> > 
> > > On Thu, Oct 12, 2017 at 03:47:43PM +0200, Jerome Brunet wrote:
> > > > The meson secure monitor seems to be compatible with more SoCs than
> > > > initially thought. Let's use the most generic compatible he have in
> > > > DT instead of the gxbb specific one
> > > > 
> > > > Signed-off-by: Jerome Brunet <jbrunet@...libre.com>
> > > > ---
> > > >  Documentation/devicetree/bindings/firmware/meson/meson_sm.txt | 4 ++--
> > > >  drivers/firmware/meson/meson_sm.c                             | 4 ++--
> > > >  2 files changed, 4 insertions(+), 4 deletions(-)
> > > 
> > > Seems like a pointless, not backwards compatible change to me.
> > 
> > I've verified that it's backwards compatible with existing upstream DTs.
> 
> Perhaps if you all are documenting only what the driver uses, not what
> the dts can have as Jerome said.
> 
> > > end, it's just a string to match on. Who cares what the string is.
> > 
> > As platform maintiner, I very much care what the strings are and I want
> > it to be coherent with the platform generic names, and I want the
> > SoC-specific strings to correspond to the actual SoC names.
> 
> The most specific compatible should be, absolutely. The fallbacks can
> be anything really. Ideally, they are the compatible string for the
> 1st SoC with "the same" compatible IP. Could be another vendor
> entirely even because mergers happen.

Then what's your problem with these patches again ?

I am just asking the driver to match the generic binding instead of the SoC
specific, because we are also using it on other SoC, as explain in the patch
comment. Does not seems that "pointless" to me.

Right now the driver match only on: vendor,soc-one
in dts, we have compatible = "vendor,family", "vendor,soc-one" 

but it is compatible with soc-two as well.
to match we would have to put "vendor,soc-one" as well which is a mess

By expressing correctly what the driver is compatible with, "vendor,family"
we can dts that makes sense for soc-two as well
compatible = "vendor,family", "vendor,soc-two"

Same goes for the other patches

> 
> Rob