linux-kernel - Re: [PATCH drm-next] drm/bochs: Add support for drm

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: <1961897d-cf42-4495-aec7-f4c20dfe7566@redhat.com>
Date: Tue, 15 Jul 2025 10:57:04 +0200
From: Jocelyn Falempe <jfalempe@...hat.com>
To: Askar Safin <safinaskar@...omail.com>, ryasuoka@...hat.com
Cc: airlied@...il.com, dri-devel@...ts.freedesktop.org, kraxel@...hat.com,
 linux-kernel@...r.kernel.org, maarten.lankhorst@...ux.intel.com,
 mripard@...nel.org, simona@...ll.ch, tzimmermann@...e.de,
 virtualization@...ts.linux.dev
Subject: Re: [PATCH drm-next] drm/bochs: Add support for drm_panic

On 14/07/2025 02:04, Askar Safin wrote:
> Are normal panics (i. e. not drm panics) still supposed to work with bochs?

"Normal panics", is just the console logs through fbcon. The problem is 
that this is not designed to work in a panic context, so on some driver 
it can work, but it's not reliable. Also depending on the panic source 
(Like if the panic occurs in IRQ context), you probably won't see anything.
I think the regression is likely because the driver switched to use a 
shadow buffer (ie: fbcon draws to a buffer in system memory, which is 
copied to the "VRAM" in a workqueue, and workqueues are disabled in a 
panic context). So there is no easy fix for the fbcon panic.

This patch series, only take care of DRM panic. You can get the same 
output as fbcon, by selecting CONFIG_DRM_PANIC_SCREEN"kmsg".

-- 

Jocelyn

> 
> If yes, then I want to point out that they, in fact, don't work since 2019.
> I. e. panics are not shown in Qemu if:
> 1) bochs is used
> 2) we use "normal" panics (not drm panics)
> 
> I already reported this here: https://lore.kernel.org/regressions/197f290e30b.eaadc7bc7913.7315623184036672946@zohomail.com/T/#u ,
> but nobody answered.
> 
> --
> Askar Safin
>