RE: [PATCH] drm/i915/display: Wait for PHY readiness not needed for disabling sequence
> -Original Message-
> From: Deak, Imre
> Sent: Tuesday, December 12, 2023 12:08 PM
> To: Kahola, Mika
> Cc: intel-gfx@lists.freedesktop.org
> Subject: Re: [PATCH] drm/i915/display: Wait for PHY readiness not needed for
> disabling sequence
>
> On Tue, Dec 12, 2023 at 11:55:20AM +0200, Mika Kahola wrote:
> > When going through the disconnection flow we don't need to wait for
> > PHY readiness and hence we can skip the wait part. For disabling the
> > function returns false as an indicator that the power is not enabled.
> > After all, we are not even using the return value when Type-C is
> > disconnecting.
> >
> > BSpec: 65380
> >
> > Signed-off-by: Mika Kahola
> > ---
> > drivers/gpu/drm/i915/display/intel_tc.c | 3 +++
> > 1 file changed, 3 insertions(+)
> >
> > diff --git a/drivers/gpu/drm/i915/display/intel_tc.c
> > b/drivers/gpu/drm/i915/display/intel_tc.c
> > index f64d348a969e..79ec17fa3edd 100644
> > --- a/drivers/gpu/drm/i915/display/intel_tc.c
> > +++ b/drivers/gpu/drm/i915/display/intel_tc.c
> > @@ -1030,6 +1030,9 @@ static bool
> > xelpdp_tc_phy_enable_tcss_power(struct intel_tc_port *tc, bool enabl
> >
> > __xelpdp_tc_phy_enable_tcss_power(tc, enable);
> >
> > + if (!enable)
> > + return false;
>
> Waiting for xelpdp_tc_phy_wait_for_tcss_power(tc, false); is also needed in
> the above case.
Ok. I would still keep this disabling part separate as if I combine this with
existing if() the code looks a bit busy for me.
Thanks!
Mika
>
> > +
> > if ((!tc_phy_wait_for_ready(tc) ||
> > !xelpdp_tc_phy_wait_for_tcss_power(tc, enable)) &&
> > !drm_WARN_ON(>drm, tc->mode == TC_PORT_LEGACY)) {
> > --
> > 2.34.1
> >
Re: [PATCH] drm/i915/display: Wait for PHY readiness not needed for disabling sequence
On Tue, Dec 12, 2023 at 11:55:20AM +0200, Mika Kahola wrote:
> When going through the disconnection flow we don't need to wait for PHY
> readiness and hence we can skip the wait part. For disabling the function
> returns false as an indicator that the power is not enabled. After all,
> we are not even using the return value when Type-C is disconnecting.
>
> BSpec: 65380
>
> Signed-off-by: Mika Kahola
> ---
> drivers/gpu/drm/i915/display/intel_tc.c | 3 +++
> 1 file changed, 3 insertions(+)
>
> diff --git a/drivers/gpu/drm/i915/display/intel_tc.c
> b/drivers/gpu/drm/i915/display/intel_tc.c
> index f64d348a969e..79ec17fa3edd 100644
> --- a/drivers/gpu/drm/i915/display/intel_tc.c
> +++ b/drivers/gpu/drm/i915/display/intel_tc.c
> @@ -1030,6 +1030,9 @@ static bool xelpdp_tc_phy_enable_tcss_power(struct
> intel_tc_port *tc, bool enabl
>
> __xelpdp_tc_phy_enable_tcss_power(tc, enable);
>
> + if (!enable)
> + return false;
Waiting for xelpdp_tc_phy_wait_for_tcss_power(tc, false); is also needed
in the above case.
> +
> if ((!tc_phy_wait_for_ready(tc) ||
>!xelpdp_tc_phy_wait_for_tcss_power(tc, enable)) &&
> !drm_WARN_ON(>drm, tc->mode == TC_PORT_LEGACY)) {
> --
> 2.34.1
>
[PATCH] drm/i915/display: Wait for PHY readiness not needed for disabling sequence
When going through the disconnection flow we don't need to wait for PHY
readiness and hence we can skip the wait part. For disabling the function
returns false as an indicator that the power is not enabled. After all,
we are not even using the return value when Type-C is disconnecting.
BSpec: 65380
Signed-off-by: Mika Kahola
---
drivers/gpu/drm/i915/display/intel_tc.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/gpu/drm/i915/display/intel_tc.c
b/drivers/gpu/drm/i915/display/intel_tc.c
index f64d348a969e..79ec17fa3edd 100644
--- a/drivers/gpu/drm/i915/display/intel_tc.c
+++ b/drivers/gpu/drm/i915/display/intel_tc.c
@@ -1030,6 +1030,9 @@ static bool xelpdp_tc_phy_enable_tcss_power(struct
intel_tc_port *tc, bool enabl
__xelpdp_tc_phy_enable_tcss_power(tc, enable);
+ if (!enable)
+ return false;
+
if ((!tc_phy_wait_for_ready(tc) ||
!xelpdp_tc_phy_wait_for_tcss_power(tc, enable)) &&
!drm_WARN_ON(>drm, tc->mode == TC_PORT_LEGACY)) {
--
2.34.1
