linux-kernel - Re: [NOT A REGRESSION] firmware: framebuffer-coreboot: duplicate device name "simple-framebuffer.0"

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 for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <CAODwPW8P+jcF0erUph5XyWoyQgLFbZWxEM6Ygi_LFCCTLmH89Q@mail.gmail.com>
Date: Thu, 12 Sep 2024 09:03:34 -0700
From: Julius Werner <jwerner@...omium.org>
To: Javier Martinez Canillas <javierm@...hat.com>
Cc: Brian Norris <briannorris@...omium.org>, Borislav Petkov <bp@...en8.de>, 
	Hugues Bruant <hugues.bruant@...il.com>, stable@...r.kernel.org, 
	regressions@...ts.linux.dev, linux-kernel@...r.kernel.org, 
	Fenghua Yu <fenghua.yu@...el.com>, Reinette Chatre <reinette.chatre@...el.com>, 
	Tony Luck <tony.luck@...el.com>, Tzung-Bi Shih <tzungbi@...nel.org>, 
	Julius Werner <jwerner@...omium.org>, chrome-platform@...ts.linux.dev, 
	Jani Nikula <jani.nikula@...ux.intel.com>, 
	Joonas Lahtinen <joonas.lahtinen@...ux.intel.com>, Rodrigo Vivi <rodrigo.vivi@...el.com>, 
	Tvrtko Ursulin <tursulin@...ulin.net>, intel-gfx@...ts.freedesktop.org, 
	dri-devel@...ts.freedesktop.org
Subject: Re: [NOT A REGRESSION] firmware: framebuffer-coreboot: duplicate
 device name "simple-framebuffer.0"

> On Coreboot platforms, a system framebuffer may be provided to the Linux
> kernel by filling a LB_TAG_FRAMEBUFFER entry in the Coreboot table. But
> it seems SeaBIOS payload can also provide a VGA mode in the boot params.
>
> [...]
>
> To prevent the issue, make the framebuffer_core driver to disable sysfb
> if there is system framebuffer data in the Coreboot table. That way only
> this driver will register a device and sysfb would not attempt to do it
> (or remove its registered device if was already executed before).

I wonder if the priority should be the other way around? coreboot's
framebuffer is generally only valid when coreboot exits to the payload
(e.g. SeaBIOS). Only if the payload doesn't touch the display
controller or if there is no payload and coreboot directly hands off
to a kernel does the kernel driver for LB_TAG_FRAMEBUFFER make sense.
But if there is some other framebuffer information passed to the
kernel from a firmware component running after coreboot, most likely
that one is more up to date and the framebuffer described by the
coreboot table doesn't work anymore (because the payload usually
doesn't modify the coreboot tables again, even if it changes hardware
state). So if there are two drivers fighting over which firmware
framebuffer description is the correct one, the coreboot driver should
probably give way.