linux-kernel - [4.17 Regression]: x86/dma: Use DMA-direct

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>] [day] [month] [year] [list]

Message-ID: <CAP8WD_YG_qirbSoVtyYoWRhZeON6p2RbN06B_yGQk1pA3==UTg@mail.gmail.com>
Date:   Thu, 7 Feb 2019 08:00:14 -0500
From:   tedheadster <tedheadster@...il.com>
To:     Christoph Hellwig <hch@...radead.org>,
        Fenghua Yu <fenghua.yu@...el.com>
Cc:     Linux Kernel Mailing List <linux-kernel@...r.kernel.org>
Subject: [4.17 Regression]: x86/dma: Use DMA-direct

Christoph,
  in the 4.17 cycle you submitted the patch "x86/dma: Use DMA-direct".
This causes the VGA video on my laptop to stop working, but the system
continues to be up and functioning. Video works until about 3/4 the
way through boot, when I think it is loading a video driver, and then
video hangs. I could ssh to it and the keyboard would respond to
alt-sysrq commands.

Strangely, it started working again in the 4.20 cycle with the
addition of Fenghua's patch "x86/cpufeatures: Enumerate MOVDIRI
instruction". This is odd to me because it looks like a very simple
patch.

I only have CONFIG_VGA_CONSOLE, CONFIG_FRAMEBUFFER_CONSOLE and
CONFIG_FRAMEBUFFER_CONSOLE_DETECT_PRIMARY enabled, with no
hardware-specific framebuffer config option set.

lspci -nn of my VGA adapter gives the following:

00:02.0 VGA compatible controller [0300]: Intel Corporation
82852/855GM Integrated Graphics Device [8086:3582] (rev 02)

Your thoughts? How can I help debug this?

Reported-by: Matthew Whitehead <tedheadster@...il.com>