Re: [PATCH v2 2/2] mmc: Checking BKOPS status prior to Suspend

[Shawn Lin <shawn.lin@rock-chips.com>]

Hi Alex,

On 2017/2/5 23:13, Alex Lemberg wrote:
> Hi Shawn,
>
> On 2/5/17, 3:43 AM, "Shawn Lin" <shawn.lin@rock-chips.com> wrote:
>
>
>     >
>     > Hmm.
>     >
>     > Shouldn't we abort (return -EBUSY) also in the system PM suspend case
>     > and not only for runtime PM suspend?
>
>     I would rather we don't abort runtime PM, but do it for system PM
>     instead. What we do for runtime PM is just for saving power for
>     our hosts including manipulating clock and genpd etc.. Thus we don't
>     touch any power-supply for eMMC. If so, it could still be possible
>     for eMMC firmware to work on doing bkops in rpm context.
>
> If I understand correctly, you are suggesting do not send SLEEP command
> on Runtime Suspend?

yes.

> If so, and if mmc host does not touch any power supply for eMMC,
> the device can continue doing Auto BKOPS during Runtime Suspend.
> The question is why the Sleep command is sent now,
> in the original implementation?
> I assume it is sent for saving a power from the device side as well?

right, it was invented with the cap of MMC_CAP_AGGRESSIVE_PM, so some 
platforms ask for the aggressive power policy will send sleep command
to the cards in rpm context. So I assume they care power-saving more
than others.


>
> N�����r��y���b�X��ǧv�^�)޺{.n�+����{��g"��^n�r���z�\x1a��h����&��\x1e�G���h�\x03(�階�ݢj"��\x1a�^[m�����z�ޖ���f���h���~�mml==
>