linux-kernel - Re: [PATCH v3] Add two new configuration drivers for Microchip USB hubs

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:   Mon, 24 Aug 2020 12:31:40 +0200
From:   Christian Eggers <ceggers@...i.de>
To:     Richard Leitner <richard.leitner@...data.com>
CC:     Greg Kroah-Hartman <gregkh@...uxfoundation.org>,
        Rob Herring <robh@...nel.org>,
        Rob Herring <robh+dt@...nel.org>, <devicetree@...r.kernel.org>,
        <linux-usb@...r.kernel.org>, <linux-kernel@...r.kernel.org>
Subject: Re: [PATCH v3] Add two new configuration drivers for Microchip USB hubs

Hi Richard,

On Monday, 17 August 2020, 09:54:26 CEST, Richard Leitner wrote:
> Hi Christian,
> sorry for the late reply. My MUA somehow didn't show me that series
> earlier...
likewise... I was on holiday last week.

> I haven't looked into the patches in detail, but at a first glance it
> looks like a lot copy-n-paste.
> Have you thought about merging the (after your series) 3 hub drivers
> into one? Something like a "microchip i2c usb hub driver"?
> Would that be feasible for your point of view?
I'm not sure about the criteria for having separate drivers vs. a combined 
one.
As changing the driver usually requires testing with real hardware, keeping 
them separate may be easier. On the other hand, I already synced some changes 
from usb251xb into "my" drivers, so it is likely that such work will also have 
to done in future.

Rob Herring already suggested to create a common yaml document for the
device tree bindings. It would probably make sense to share the device tree 
code between our drivers. But I would like to keep the "hardware" side of the 
drivers independent, as there are subtle differences between the different hub 
series.

Compared to usb251x, the new drivers don't write the full configuration memory 
of the hub (the configuration memory space is much bigger than 8 bit and has 
many holes). They rely on the defaults after hardware reset on perform read-
modify-write for the properties set in the device tree.

If the usb251xb hubs could be programmed in a similar way, merging all drivers 
should be possible using a mic_usb_hub_ops struct with function pointers for 
applying the (common) device tree properties (e.g. set_product_id(), 
set_device_id(), ..., set_oc_delay(), set_manufacturer_string(), ...).

Best regards
Christian