netdev - Re: [PATCH v2 net-next 4/4] net: dsa: validate that DT nodes of shared ports have the properties they need

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]

Message-ID: <2a159954-7175-b747-a53b-0282998eec90@gmail.com>
Date:   Fri, 29 Jul 2022 14:44:31 -0700
From:   Florian Fainelli <f.fainelli@...il.com>
To:     Marcin Wojtas <mw@...ihalf.com>
Cc:     Andrew Lunn <andrew@...n.ch>,
        Vladimir Oltean <vladimir.oltean@....com>,
        netdev <netdev@...r.kernel.org>,
        "David S. Miller" <davem@...emloft.net>,
        Eric Dumazet <edumazet@...gle.com>,
        Jakub Kicinski <kuba@...nel.org>,
        Paolo Abeni <pabeni@...hat.com>,
        Vivien Didelot <vivien.didelot@...il.com>,
        Oleksij Rempel <linux@...pel-privat.de>,
        Christian Marangi <ansuelsmth@...il.com>,
        John Crispin <john@...ozen.org>,
        Kurt Kanzenbach <kurt@...utronix.de>,
        Mans Rullgard <mans@...sr.com>,
        Arun Ramadoss <arun.ramadoss@...rochip.com>,
        Woojung Huh <woojung.huh@...rochip.com>,
        "UNGLinuxDriver@...rochip.com" <UNGLinuxDriver@...rochip.com>,
        Claudiu Manoil <claudiu.manoil@....com>,
        Alexandre Belloni <alexandre.belloni@...tlin.com>,
        George McCollister <george.mccollister@...il.com>,
        DENG Qingfang <dqfext@...il.com>,
        Sean Wang <sean.wang@...iatek.com>,
        Landen Chao <Landen.Chao@...iatek.com>,
        Matthias Brugger <matthias.bgg@...il.com>,
        Hauke Mehrtens <hauke@...ke-m.de>,
        Martin Blumenstingl <martin.blumenstingl@...glemail.com>,
        Aleksander Jan Bajkowski <olek2@...pl>,
        Alvin Šipraga <alsi@...g-olufsen.dk>,
        Luiz Angelo Daros de Luca <luizluca@...il.com>,
        Linus Walleij <linus.walleij@...aro.org>,
        Pawel Dembicki <paweldembicki@...il.com>,
        Clément Léger <clement.leger@...tlin.com>,
        Geert Uytterhoeven <geert+renesas@...der.be>,
        Russell King <rmk+kernel@...linux.org.uk>,
        Marek Behún <kabel@...nel.org>,
        Rob Herring <robh+dt@...nel.org>,
        Frank Rowand <frowand.list@...il.com>,
        Tomasz Nowicki <tn@...ihalf.com>,
        Grzegorz Jaszczyk <jaz@...ihalf.com>
Subject: Re: [PATCH v2 net-next 4/4] net: dsa: validate that DT nodes of
 shared ports have the properties they need

On 7/29/22 14:33, Marcin Wojtas wrote:
> pt., 29 lip 2022 o 23:24 Florian Fainelli <f.fainelli@...il.com> napisał(a):
>>
>> On 7/29/22 14:17, Andrew Lunn wrote:
>>>> What I propose is to enforce more strictly an update of DT description
>>>> with a specified timeline, abandoning 'camps' idea and driver-specific
>>>> contents in the generic code.
>>>
>>> Regressions are the problem. We are supposed to be backwards
>>> compatible with older DT blobs. If we now say old DT blobs are
>>> invalid, and refuse to probe, we cause a regression.
>>>
>>> For some of the in kernel DT files using the mv88e6xxx i can make a
>>> good guess at what the missing properties are. However, i'm bound to
>>> guess wrong at some point, and cause a regression. So we could change
>>> just those we can test. But at some point, the other blobs are going
>>> to fail the enforces checks and cause a regression anyway.
>>>
>>> And what about out of tree blobs? Probably OpenWRT have some. Do we
>>> want to cause them to regress?
>>
>> No, we do not want that, which is why Vladimir's approach IMHO is reasonable in that it acknowledges mistakes or shortcomings of the past into the present, and expects the future to be corrected and not repeat those same mistakes. The deprectiation window idea is all well and good in premise, however with such a large user base, I am not sure it is going to go very far unfortunately, nor that it will hinder our ability to have a more maintainable DSA framework TBH.
>>
>> BTW, OpenWrt does not typically ship DT blobs that stay frozen, all of the kernel, DTBs, root filesystem, and sometimes a recent u-boot copy will be updated at the same time because very rarely do the existing boot loader satisfy modern requirements (PSCI, etc.).
> 
> Initially, I thought that the idea is a probe failure (hence the camps
> to prevent that) - but it was clarified later, it's not the case.
> 
> I totally agree and I am all against breaking the backward
> compatibility (this is why I work on ACPI support that much :) ). The
> question is whether for existing deployments with 'broken' DT
> description we would be ok to introduce a dev_warn/WARN_ON message
> after a kernel update. That would be a case if the check is performed
> unconditionally - this way we can keep compat strings out of net/dsa.
> What do you think?

A warning seems fine and appropriate to give just the right amount of nudge to get things fixed, I would not as far as a full backtrace WARN() because those will definitively upset people's CI in a wayt that seems a bit over reacting. Anyway I do have my share of DT blobs to update.
-- 
Florian