linux-kernel - Re: FYI: i.MX8MP ISP (RKISP1) MI registers corruption: resolved

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: <175686096405.1542294.2802960782852730947@neptunite.rasen.tech>
Date: Wed, 03 Sep 2025 09:56:04 +0900
From: Paul Elder <paul.elder@...asonboard.com>
To: Adam Ford <aford173@...il.com>, Krzysztof Hałasa <khalasa@...p.pl>
Cc: Stefan Klug <stefan.klug@...asonboard.com>, Laurent Pinchart <laurent.pinchart@...asonboard.com>, Dafna Hirschfeld <dafna@...tmail.com>, Heiko Stuebner <heiko@...ech.de>, Jacopo Mondi <jacopo.mondi@...asonboard.com>, Ondrej Jirman <megi@....cz>, linux-media@...r.kernel.org, linux-rockchip@...ts.infradead.org, linux-arm-kernel@...ts.infradead.org, linux-kernel@...r.kernel.org
Subject: Re: FYI: i.MX8MP ISP (RKISP1) MI registers corruption: resolved

Hi Krzysztof,

Quoting Krzysztof Hałasa (2025-09-02 18:54:27)
> Hi,
> 
> summary:
> 
> I've done a few additional tests and it seems the MEDIA_AXI clock is the
> problem. Reducing it to 400 MHz while still running MEDIA_ISP at 500 MHz
> produces no errors.
> MEDIA_ISP at 400 MHz and MEDIA_AXI at 500 MHz produces errors, though
> (register address errors while reading and writing from/to ISP MI
> (memory interface) registers, only on the secondary ISP (isp1), and
> generally only while streaming data from the ISP).
> 
> What is driven by MEDIA_AXI clock root? MEDIAMIX: ISI, LCDIF, ISP, DWE.
> 
> According to both datasheets (industrial and commercial), MEDIA_AXI
> is limited to 400 MHz in normal mode and 500 MHz in overdrive mode.
> All my hardware is setup for overdrive mode, though (two manufacturers,
> both using the same PMIC setup).

Thanks for the investigation!

> 
> Since no hardware in the official Linux kernel tree (DT) uses the second
> ISP... Should we just add a warning to the imx8mp.dtsi and be done with
> it?

afaiu we can't yet use the ISI and ISP simultaneously, so the ISI is
enabled by default and you need overlays to enable the ISP, so
technically nothing upstream uses either ISP. To my knowledge, there is
hardware where you can use both though, such as the Debix SOM A
(imx8mp-debix-som-a-bmb-08.dts) [0].


Paul

[0] https://debix.io/hardware/debix-som-a-io-board.html

> Out of tree hardware using isp1 (csi1) obviously exists.
> -- 
> Krzysztof "Chris" Hałasa
> 
> Sieć Badawcza Łukasiewicz
> Przemysłowy Instytut Automatyki i Pomiarów PIAP
> Al. Jerozolimskie 202, 02-486 Warszawa