linux-kernel - Re: [PATCH] PCI: Disable D3cold on Asus B1400 PCI-NVMe bridge

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

Message-ID: <4478a4b0-fd17-4d5f-b11c-ba4a867bf92b@amd.com>
Date: Tue, 30 Jan 2024 13:10:09 -0600
From: Mario Limonciello <mario.limonciello@....com>
To: Daniel Drake <drake@...lessos.org>
Cc: tglx@...utronix.de, mingo@...hat.com, bp@...en8.de,
 dave.hansen@...ux.intel.com, x86@...nel.org, hpa@...or.com,
 linux-pci@...r.kernel.org, linux-kernel@...r.kernel.org,
 bhelgaas@...gle.com, david.e.box@...ux.intel.com,
 Jian-Hong Pan <jhp@...lessos.org>
Subject: Re: [PATCH] PCI: Disable D3cold on Asus B1400 PCI-NVMe bridge

On 1/30/2024 13:00, Daniel Drake wrote:
> On Tue, Jan 30, 2024 at 2:47 PM Mario Limonciello
> <mario.limonciello@....com> wrote:
>> Has there already been a check done whether any newer firmware is
>> available from ASUS that doesn't suffer the deficiency described below?
> 
> The latest firmware does flip StorageD3Enable to 0, which has the side
> effect of never putting the NVMe device or the parent bridge into
> D3cold.
> However, we have shipped hundreds of these devices with the original
> production BIOS version to first time computer users, so it is not
> feasible to ask the end user to upgrade. And there is no
> Linux-compatible online firmware update for this product range. Hence
> a Linux-level workaround for this issue would be highly valuable.

Sure; it makes sense to me why to put a workaround in Linux with your 
above explanation.

Perhaps it make sense to limit the quirk to only the DMI version strings 
of the initial production firmware?

> 
>> Is this the only problem blocking s2idle from working effectively on
>> this platform?  If so, I would think you want to just do the revert in
>> the same series if it's decided this patch needs to spin again for any
>> reason.
> 
> Yes these could be combined into the same series, with agreement from
> the drivers/acpi/ maintainers for the S3 revert.
> 
> Thanks
> Daniel