netdev - Re: [PATCH 1/2] net: dsa: microchip: add KSZ9896 switch support

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]

Date:   Tue, 30 Aug 2022 09:50:30 +0200
From:   Romain Naour <romain.naour@...le.fr>
To:     Jakub Kicinski <kuba@...nel.org>
Cc:     Andrew Lunn <andrew@...n.ch>, netdev@...r.kernel.org,
        woojung.huh@...rochip.com, UNGLinuxDriver@...rochip.com,
        arun.ramadoss@...rochip.com, Romain Naour <romain.naour@....com>
Subject: Re: [PATCH 1/2] net: dsa: microchip: add KSZ9896 switch support

Hi,

Le 30/08/2022 à 07:07, Jakub Kicinski a écrit :
> On Mon, 29 Aug 2022 09:20:06 +0200 Romain Naour wrote:
>>> It's pretty common to use a different email server than what's
>>> on the From and S-o-b lines. You can drop the second S-o-b.  
>>
>> Thanks for clarification.
>>
>> Should resend the series now or wait for a review?
> 
> Resend, please.

ok.

> 
> Looking quickly at the code it's interesting to see the compatible
> already listed in the DT schema but the driver not supporting it. 
> Is this common in the embedded world?
> 
> That's orthogonal to your patch, I'm just curious.

I was surprised too about the compatible being listed in the DT schema but being
missing in the driver.

AFAIK, the DT schema was merged in 2019 (kernel 5.1) but the corresponding
driver patch didn't pass the review and no further patch was sent on the mailing
list for the KSZ9896.

It's true that the KSZ9896 is really close to the KSZ9897 (only one CPU port is
missing) but the driver reject this device due to a mismatch in the check
between the device id and the compatible. I tried to workaround this check but
the driver really need to know how many port are really present on the ethernet
switch otherwise there is a kernel oops.

Best regards,
Romain