2020-07-09 09:19:21

by Rakesh Pillai

[permalink] [raw]
Subject: [PATCH] arm64: dts: qcom: sc7180: Add missing properties for Wifi node

The wlan firmware memory is statically mapped in
the Trusted Firmware, hence the wlan driver does
not need to map/unmap this region dynamically.

Hence add the property to indicate the wlan driver
to not map/unamp the firmware memory region
dynamically.

Also add the chain1 voltage supply for wlan.

Signed-off-by: Rakesh Pillai <[email protected]>
---
This patch is created on top of the change by
Douglas Anderson.
https://lkml.org/lkml/2020/6/25/817

Also the dt-bindings for the chain1 voltage supply
is added by the below patch series:
https://patchwork.kernel.org/project/linux-wireless/list/?series=309137
---
arch/arm64/boot/dts/qcom/sc7180-idp.dts | 2 ++
1 file changed, 2 insertions(+)

diff --git a/arch/arm64/boot/dts/qcom/sc7180-idp.dts b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
index 472f7f4..4c64bc1 100644
--- a/arch/arm64/boot/dts/qcom/sc7180-idp.dts
+++ b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
@@ -391,10 +391,12 @@

&wifi {
status = "okay";
+ qcom,msa-fixed-perm;
vdd-0.8-cx-mx-supply = <&vreg_l9a_0p6>;
vdd-1.8-xo-supply = <&vreg_l1c_1p8>;
vdd-1.3-rfa-supply = <&vreg_l2c_1p3>;
vdd-3.3-ch0-supply = <&vreg_l10c_3p3>;
+ vdd-3.3-ch1-supply = <&vreg_l11c_3p3>;
wifi-firmware {
iommus = <&apps_smmu 0xc2 0x1>;
};
--
2.7.4


2020-07-09 20:07:24

by Doug Anderson

[permalink] [raw]
Subject: Re: [PATCH] arm64: dts: qcom: sc7180: Add missing properties for Wifi node

Hi,

On Thu, Jul 9, 2020 at 2:18 AM Rakesh Pillai <[email protected]> wrote:
>
> The wlan firmware memory is statically mapped in
> the Trusted Firmware, hence the wlan driver does
> not need to map/unmap this region dynamically.
>
> Hence add the property to indicate the wlan driver
> to not map/unamp the firmware memory region
> dynamically.
>
> Also add the chain1 voltage supply for wlan.
>
> Signed-off-by: Rakesh Pillai <[email protected]>
> ---
> This patch is created on top of the change by
> Douglas Anderson.
> https://lkml.org/lkml/2020/6/25/817
>
> Also the dt-bindings for the chain1 voltage supply
> is added by the below patch series:
> https://patchwork.kernel.org/project/linux-wireless/list/?series=309137
> ---
> arch/arm64/boot/dts/qcom/sc7180-idp.dts | 2 ++
> 1 file changed, 2 insertions(+)
>
> diff --git a/arch/arm64/boot/dts/qcom/sc7180-idp.dts b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
> index 472f7f4..4c64bc1 100644
> --- a/arch/arm64/boot/dts/qcom/sc7180-idp.dts
> +++ b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
> @@ -391,10 +391,12 @@
>
> &wifi {
> status = "okay";
> + qcom,msa-fixed-perm;

At one point in time I thought +Sibi said that this wouldn't be needed
once the firmware was fixed. ...afterwards you said that it was
needed for SSR (subsystem reset). Would be good to get confirmation
from Sibi that this matches his understanding.


> vdd-0.8-cx-mx-supply = <&vreg_l9a_0p6>;
> vdd-1.8-xo-supply = <&vreg_l1c_1p8>;
> vdd-1.3-rfa-supply = <&vreg_l2c_1p3>;
> vdd-3.3-ch0-supply = <&vreg_l10c_3p3>;
> + vdd-3.3-ch1-supply = <&vreg_l11c_3p3>;
> wifi-firmware {
> iommus = <&apps_smmu 0xc2 0x1>;
> };

Other than the one question this looks good to me.

-Doug

2020-07-11 03:38:20

by Rakesh Pillai

[permalink] [raw]
Subject: RE: [PATCH] arm64: dts: qcom: sc7180: Add missing properties for Wifi node



> -----Original Message-----
> From: Doug Anderson <[email protected]>
> Sent: Friday, July 10, 2020 1:36 AM
> To: Rakesh Pillai <[email protected]>
> Cc: open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS
> <[email protected]>; Evan Green <[email protected]>;
> Andy Gross <[email protected]>; Bjorn Andersson
> <[email protected]>; Rob Herring <[email protected]>; linux-
> arm-msm <[email protected]>; LKML <linux-
> [email protected]>; Sibi Sankar <[email protected]>
> Subject: Re: [PATCH] arm64: dts: qcom: sc7180: Add missing properties for
> Wifi node
>
> Hi,
>
> On Thu, Jul 9, 2020 at 2:18 AM Rakesh Pillai <[email protected]> wrote:
> >
> > The wlan firmware memory is statically mapped in
> > the Trusted Firmware, hence the wlan driver does
> > not need to map/unmap this region dynamically.
> >
> > Hence add the property to indicate the wlan driver
> > to not map/unamp the firmware memory region
> > dynamically.
> >
> > Also add the chain1 voltage supply for wlan.
> >
> > Signed-off-by: Rakesh Pillai <[email protected]>
> > ---
> > This patch is created on top of the change by
> > Douglas Anderson.
> > https://lkml.org/lkml/2020/6/25/817
> >
> > Also the dt-bindings for the chain1 voltage supply
> > is added by the below patch series:
> > https://patchwork.kernel.org/project/linux-wireless/list/?series=309137
> > ---
> > arch/arm64/boot/dts/qcom/sc7180-idp.dts | 2 ++
> > 1 file changed, 2 insertions(+)
> >
> > diff --git a/arch/arm64/boot/dts/qcom/sc7180-idp.dts
> b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
> > index 472f7f4..4c64bc1 100644
> > --- a/arch/arm64/boot/dts/qcom/sc7180-idp.dts
> > +++ b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
> > @@ -391,10 +391,12 @@
> >
> > &wifi {
> > status = "okay";
> > + qcom,msa-fixed-perm;
>
> At one point in time I thought +Sibi said that this wouldn't be needed
> once the firmware was fixed. ...afterwards you said that it was
> needed for SSR (subsystem reset). Would be good to get confirmation
> from Sibi that this matches his understanding.

Hi Doug,

This is now needed as the firmware memory mapping was moved to Trusted firmware.
This region is now statically mapped to avoid access from driver.

>
>
> > vdd-0.8-cx-mx-supply = <&vreg_l9a_0p6>;
> > vdd-1.8-xo-supply = <&vreg_l1c_1p8>;
> > vdd-1.3-rfa-supply = <&vreg_l2c_1p3>;
> > vdd-3.3-ch0-supply = <&vreg_l10c_3p3>;
> > + vdd-3.3-ch1-supply = <&vreg_l11c_3p3>;
> > wifi-firmware {
> > iommus = <&apps_smmu 0xc2 0x1>;
> > };
>
> Other than the one question this looks good to me.
>
> -Doug

2020-07-11 04:59:40

by Sibi Sankar

[permalink] [raw]
Subject: Re: [PATCH] arm64: dts: qcom: sc7180: Add missing properties for Wifi node

On 2020-07-11 09:07, Rakesh Pillai wrote:
>> -----Original Message-----
>> From: Doug Anderson <[email protected]>
>> Sent: Friday, July 10, 2020 1:36 AM
>> To: Rakesh Pillai <[email protected]>
>> Cc: open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS
>> <[email protected]>; Evan Green <[email protected]>;
>> Andy Gross <[email protected]>; Bjorn Andersson
>> <[email protected]>; Rob Herring <[email protected]>; linux-
>> arm-msm <[email protected]>; LKML <linux-
>> [email protected]>; Sibi Sankar <[email protected]>
>> Subject: Re: [PATCH] arm64: dts: qcom: sc7180: Add missing properties
>> for
>> Wifi node
>>
>> Hi,
>>
>> On Thu, Jul 9, 2020 at 2:18 AM Rakesh Pillai <[email protected]>
>> wrote:
>> >
>> > The wlan firmware memory is statically mapped in
>> > the Trusted Firmware, hence the wlan driver does
>> > not need to map/unmap this region dynamically.
>> >
>> > Hence add the property to indicate the wlan driver
>> > to not map/unamp the firmware memory region
>> > dynamically.
>> >
>> > Also add the chain1 voltage supply for wlan.
>> >
>> > Signed-off-by: Rakesh Pillai <[email protected]>
>> > ---
>> > This patch is created on top of the change by
>> > Douglas Anderson.
>> > https://lkml.org/lkml/2020/6/25/817
>> >
>> > Also the dt-bindings for the chain1 voltage supply
>> > is added by the below patch series:
>> > https://patchwork.kernel.org/project/linux-wireless/list/?series=309137
>> > ---
>> > arch/arm64/boot/dts/qcom/sc7180-idp.dts | 2 ++
>> > 1 file changed, 2 insertions(+)
>> >
>> > diff --git a/arch/arm64/boot/dts/qcom/sc7180-idp.dts
>> b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
>> > index 472f7f4..4c64bc1 100644
>> > --- a/arch/arm64/boot/dts/qcom/sc7180-idp.dts
>> > +++ b/arch/arm64/boot/dts/qcom/sc7180-idp.dts
>> > @@ -391,10 +391,12 @@
>> >
>> > &wifi {
>> > status = "okay";
>> > + qcom,msa-fixed-perm;

Should we include ^^ property
in the base dts? Since I don't
foresee any platform skipping
qcom,msa-fixed-perm property.

>>
>> At one point in time I thought +Sibi said that this wouldn't be needed
>> once the firmware was fixed. ...afterwards you said that it was
>> needed for SSR (subsystem reset). Would be good to get confirmation
>> from Sibi that this matches his understanding.

Yes all of ^^ happened and yes
we now need qcom,msa-fixed-perm
since the wlan_fw_mem permission
assignment in now handled in ATF.

>
> Hi Doug,
>
> This is now needed as the firmware memory mapping was moved to Trusted
> firmware.
> This region is now statically mapped to avoid access from driver.
>
>>
>>
>> > vdd-0.8-cx-mx-supply = <&vreg_l9a_0p6>;
>> > vdd-1.8-xo-supply = <&vreg_l1c_1p8>;
>> > vdd-1.3-rfa-supply = <&vreg_l2c_1p3>;
>> > vdd-3.3-ch0-supply = <&vreg_l10c_3p3>;
>> > + vdd-3.3-ch1-supply = <&vreg_l11c_3p3>;
>> > wifi-firmware {
>> > iommus = <&apps_smmu 0xc2 0x1>;
>> > };
>>
>> Other than the one question this looks good to me.
>>
>> -Doug

--
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project.