netdev - Re: [PATCH v9 net-next 2/7] net: mdio: mscc-miim: add ability to be used in a non-mmio configuration

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: <20220611164520.GB848433@euler>
Date:   Sat, 11 Jun 2022 09:45:20 -0700
From:   Colin Foster <colin.foster@...advantage.com>
To:     Andy Shevchenko <andy.shevchenko@...il.com>
Cc:     devicetree <devicetree@...r.kernel.org>,
        Linux Kernel Mailing List <linux-kernel@...r.kernel.org>,
        netdev <netdev@...r.kernel.org>,
        linux-arm Mailing List <linux-arm-kernel@...ts.infradead.org>,
        "open list:GPIO SUBSYSTEM" <linux-gpio@...r.kernel.org>,
        Vladimir Oltean <vladimir.oltean@....com>,
        Lee Jones <lee.jones@...aro.org>,
        Rob Herring <robh+dt@...nel.org>,
        Krzysztof Kozlowski <krzysztof.kozlowski+dt@...aro.org>,
        Andrew Lunn <andrew@...n.ch>,
        Heiner Kallweit <hkallweit1@...il.com>,
        Russell King <linux@...linux.org.uk>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        Lars Povlsen <lars.povlsen@...rochip.com>,
        Steen Hegelund <Steen.Hegelund@...rochip.com>,
        Microchip Linux Driver Support <UNGLinuxDriver@...rochip.com>,
        Linus Walleij <linus.walleij@...aro.org>,
        Wolfram Sang <wsa@...nel.org>,
        Terry Bowman <terry.bowman@....com>
Subject: Re: [PATCH v9 net-next 2/7] net: mdio: mscc-miim: add ability to be
 used in a non-mmio configuration

Hi Andy,

On Sat, Jun 11, 2022 at 12:34:59PM +0200, Andy Shevchenko wrote:
> On Fri, Jun 10, 2022 at 7:57 PM Colin Foster
> <colin.foster@...advantage.com> wrote:
> >
> > There are a few Ocelot chips that contain the logic for this bus, but are
> > controlled externally. Specifically the VSC7511, 7512, 7513, and 7514. In
> > the externally controlled configurations these registers are not
> > memory-mapped.
> >
> > Add support for these non-memory-mapped configurations.
> 
> ...
> 
> > +       ocelot_platform_init_regmap_from_resource(pdev, 0, &mii_regmap, NULL,
> > +                                                 &mscc_miim_regmap_config);
> 
> This is a bit non-standard, why not to follow the previously used API
> design, i.e.
> 
> mii_regmap.map = ...
> 
> ?

I see your point. It looks like there's no reason to pass in &mii_regmap
and it can just be returned.

> 
> ...
> 
> > +       ocelot_platform_init_regmap_from_resource(pdev, 1, &phy_regmap, &res,
> > +                                                 &mscc_miim_phy_regmap_config);
> 
> Ditto.
> 
> Also here is the question how '_from_'  is aligned with '&res'. If
> it's _from_ a resource then the resource is already a pointer.

Yes, this is probably worth a second look. During v8 you noted that I
was repeating a lot of the same logic across several files, so I created
ocelot_platform_init_regmap_from_resource.

The "gotcha" is while most of those scenarios have a required resource,
the phy_regmap is optional - so a scenario where the resource doesn't
exist could be considered valid.

Would it make sense to make the init_regmap_from_resource function
return ERR_PTR(-ENOENT) if regs doesn't exist? That would clean up the
API quite a bit:

phy_regmap = ocelot_platform_init_regmap_from_resource(pdev, 1,
                                                       &map_config);
if (IS_ERR(phy_regmap) && phy_regmap != -ENOENT) {
        dev_err(); ...
}

It looks like none of the two functions that would get returned
otherwise (devm_regmap_init or devm_regmap_init_mmio) would return that
value...

> 
> ...
> 
> >         if (res) {
> > -               phy_regs = devm_ioremap_resource(dev, res);
> > -               if (IS_ERR(phy_regs)) {
> > -                       dev_err(dev, "Unable to map internal phy registers\n");
> > -                       return PTR_ERR(phy_regs);
> > -               }
> > -
> > -               phy_regmap = devm_regmap_init_mmio(dev, phy_regs,
> > -                                                  &mscc_miim_phy_regmap_config);
> >                 if (IS_ERR(phy_regmap)) {
> >                         dev_err(dev, "Unable to create phy register regmap\n");
> >                         return PTR_ERR(phy_regmap);
> >                 }
> 
> This looks weird. You check an error here instead of the API you
> called. It's a weird design, the rationale of which is doubtful and
> has to be at least explained.

I agree. With the changes I'm suggesting above this block of code would
become:

if (IS_ERR(phy_regmap)) {
        if (phy_regmap == -ENOENT) {
                phy_regmap = NULL;
        } else {
                dev_err(dev, "...");
                return PTR_ERR(phy_regmap);
        }
}

That seems easier to follow than the if(res) block...


Thanks for the feedback!

> 
> > +       } else {
> > +               phy_regmap = NULL;
> >         }
> 
> --
> With Best Regards,
> Andy Shevchenko