[PATCH v2] nvme: Add a module parameter for users to force simple suspend

[Limonciello, Mario]

[Public]



> -----Original Message-----
> From: Christoph Hellwig <hch at lst.de>
> Sent: Thursday, March 9, 2023 03:37
> To: Limonciello, Mario <Mario.Limonciello at amd.com>
> Cc: Keith Busch <kbusch at kernel.org>; Jens Axboe <axboe at fb.com>;
> Christoph Hellwig <hch at lst.de>; Sagi Grimberg <sagi at grimberg.me>;
> elvis.angelaccio at kde.org; linux-nvme at lists.infradead.org; linux-
> kernel at vger.kernel.org
> Subject: Re: [PATCH v2] nvme: Add a module parameter for users to force
> simple suspend
> 
> On Wed, Mar 01, 2023 at 02:33:01PM -0600, Mario Limonciello wrote:
> > Elvis Angelaccio reports that his HP laptop that has two SSDs takes
> > a long time to resume from suspend to idle and has low hardware sleep
> > residency.  In analyzing the logs and acpidump from the BIOS it's clear
> > that BIOS does set the StorageD3Enable _DSD property but it's only
> > set on one of the SSDs.
> 
> Please quirk the platform in the PCIe core instead of requiring the
> user to override manually in the nvme drivers that is not relevant.
> 

OK for this situation reported I'll quirk it another way.

> And as a representative for a CPU/platform vendor that is all behind
> this stupid crap please work with Microsoft and Intel to sort it out
> properly the firmware level.  It's a never ending pain caused on us
> that you guys caused for absolutely no reason.

The primary way to do this is via the _DSD, but for whatever reason
Microsoft also offers:
1) An allow list of old parts without the _DSD
2) A knob in the OS users can touch
3) A setting that OEMs can set in the factory

Any OEM that works with us on enabling Linux on their hardware I'll make
sure they have the _DSD for all the ports they use SSDs.