On Wed, 2022-04-20 at 16:34 +0000, Limonciello, Mario wrote:
> [AMD Official Use Only]
>
>
>
> > -----Original Message-----
> > From: David E. Box <david.e.box@xxxxxxxxxxxxxxx>
> > Sent: Wednesday, April 20, 2022 10:56
> > To: hdegoede@xxxxxxxxxx; david.e.box@xxxxxxxxxxxxxxx;
> > markgross@xxxxxxxxxx; platform-driver-x86@xxxxxxxxxxxxxxx; linux-
> > kernel@xxxxxxxxxxxxxxx
> > Subject: [PATCH 2/3] platform/x86/intel/sdsi: Poll on ready bit for writes
> >
> > Due to change in firmware flow, update mailbox writes to poll on ready bit
> > instead of run_busy bit. This change makes the polling method consistent
> > for both writes and reads, which also uses the ready bit.
>
> Does this need some sort of guard on the behavior based on the firmware
> version you are running on or are these all pre-production still?
Firmware updates would include this change, but a guard isn't necessary. Polling
on run_busy would still work for the current ABI. Since currently mailbox writes
are single packet only, both ready and run_busy update at the same time. But it
would not work for future changes where there will be multi packet writes. In
that case, ready indicates completion of the packet while run_busy indicates
completion of the entire message.
David
>
> > Fixes: 2546c6000430 ("platform/x86: Add Intel Software Defined Silicon
> > driver")
> > Signed-off-by: David E. Box <david.e.box@xxxxxxxxxxxxxxx>
> > ---
> > drivers/platform/x86/intel/sdsi.c | 4 ++--
> > 1 file changed, 2 insertions(+), 2 deletions(-)
> >
> > diff --git a/drivers/platform/x86/intel/sdsi.c
> > b/drivers/platform/x86/intel/sdsi.c
> > index 11f211402479..89729fed030c 100644
> > --- a/drivers/platform/x86/intel/sdsi.c
> > +++ b/drivers/platform/x86/intel/sdsi.c
> > @@ -245,8 +245,8 @@ static int sdsi_mbox_cmd_write(struct sdsi_priv *priv,
> > struct sdsi_mbox_info *in
> > FIELD_PREP(CTRL_PACKET_SIZE, info->size);
> > writeq(control, priv->control_addr);
> >
> > - /* Poll on run_busy bit */
> > - ret = readq_poll_timeout(priv->control_addr, control, !(control &
> > CTRL_RUN_BUSY),
> > + /* Poll on ready bit */
> > + ret = readq_poll_timeout(priv->control_addr, control, control &
> > CTRL_READY,
> > MBOX_POLLING_PERIOD_US,
> > MBOX_TIMEOUT_US);
> >
> > if (ret)
> > --
> > 2.25.1
