[PATCH v2] nvme: Favor D3cold for suspend if NVMe device supports it

[Liang, Prike]

[AMD Public Use]

According to BIOS guys the _PR3 should be always implemented for NVMe device on the onwards ASIC. This solution seems more simple and looks good to me.

Reviewed-by: Prike Liang <Prike.Liang at amd.com>

Thanks,
Prike
> -----Original Message-----
> From: Kai-Heng Feng <kai.heng.feng at canonical.com>
> Sent: Monday, April 19, 2021 2:58 PM
> To: Christoph Hellwig <hch at lst.de>
> Cc: Keith Busch <kbusch at kernel.org>; Jens Axboe <axboe at fb.com>; Sagi
> Grimberg <sagi at grimberg.me>; Deucher, Alexander
> <Alexander.Deucher at amd.com>; Liang, Prike <Prike.Liang at amd.com>; S-k,
> Shyam-sundar <Shyam-sundar.S-k at amd.com>; open list:NVM EXPRESS
> DRIVER <linux-nvme at lists.infradead.org>; open list <linux-
> kernel at vger.kernel.org>
> Subject: Re: [PATCH v2] nvme: Favor D3cold for suspend if NVMe device
> supports it
>
> On Mon, Apr 19, 2021 at 2:50 PM Christoph Hellwig <hch at lst.de> wrote:
> >
> > On Fri, Apr 16, 2021 at 05:13:44PM +0800, Kai-Heng Feng wrote:
> > > On AMD platforms that use s2idle, NVMe timeouts on s2idle resume,
> > > because their SMU FW may cut off NVMe power during sleep.
> >
> > We're already have a discussion on a proper quirk for thse broken
> > platforms on the linux-nvme list, please take part in that discussion.
>
> Thanks. I didn't notice v5 was sent the to mailing list.
> As of now, AMD folks are also reviewing this, and I believe this approach is
> less quirky.
>
> Kai-Heng