McASP AFIFOs are not necessary with UDMA-P/BCDMA as there is buffering
on the DMA IP. Drop these for better audio latency.

Fixes: 316b80246b16 ("arm64: dts: ti: add verdin am62")
Signed-off-by: Jai Luthra <j-luthra@ti.com>
---
 arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
index 2038c5e04639..27e5220e1dc7 100644
--- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
+++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
@@ -1364,8 +1364,8 @@ &mcasp0 {
 	       0 0 0 0
 	>;
 	tdm-slots = <2>;
-	rx-num-evt = <32>;
-	tx-num-evt = <32>;
+	rx-num-evt = <0>;
+	tx-num-evt = <0>;
 	#sound-dai-cells = <0>;
 	status = "disabled";
 };
@@ -1382,8 +1382,8 @@ &mcasp1 {
 	       0 0 0 0
 	>;
 	tdm-slots = <2>;
-	rx-num-evt = <32>;
-	tx-num-evt = <32>;
+	rx-num-evt = <0>;
+	tx-num-evt = <0>;
 	#sound-dai-cells = <0>;
 	status = "disabled";
 };

-- 
2.43.0

Hello Jai, thanks for the patch.

On Tue, Jun 04, 2024 at 03:11:06PM +0530, Jai Luthra wrote:
> McASP AFIFOs are not necessary with UDMA-P/BCDMA as there is buffering
> on the DMA IP. Drop these for better audio latency.
> 
> Fixes: 316b80246b16 ("arm64: dts: ti: add verdin am62")
> Signed-off-by: Jai Luthra <j-luthra@ti.com>
> ---
>  arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 8 ++++----
>  1 file changed, 4 insertions(+), 4 deletions(-)
> 
> diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> index 2038c5e04639..27e5220e1dc7 100644
> --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> @@ -1364,8 +1364,8 @@ &mcasp0 {
>  	       0 0 0 0
>  	>;
>  	tdm-slots = <2>;
> -	rx-num-evt = <32>;
> -	tx-num-evt = <32>;
> +	rx-num-evt = <0>;
> +	tx-num-evt = <0>;

From my understanding of the dt-binding having these properties set to
<0> is equivalent to not having those properties at all.
My suggestion would be to just remove those, therefore.

With that said, in any case, whatever you do my suggested change or not,

Acked-by: Francesco Dolcini <francesco.dolcini@toradex.com>

Francesco

Hi Francesco,

Thanks for the review.

On Jun 05, 2024 at 10:47:13 +0200, Francesco Dolcini wrote:
> Hello Jai, thanks for the patch.
> 
> On Tue, Jun 04, 2024 at 03:11:06PM +0530, Jai Luthra wrote:
> > McASP AFIFOs are not necessary with UDMA-P/BCDMA as there is buffering
> > on the DMA IP. Drop these for better audio latency.
> > 
> > Fixes: 316b80246b16 ("arm64: dts: ti: add verdin am62")
> > Signed-off-by: Jai Luthra <j-luthra@ti.com>
> > ---
> >  arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 8 ++++----
> >  1 file changed, 4 insertions(+), 4 deletions(-)
> > 
> > diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > index 2038c5e04639..27e5220e1dc7 100644
> > --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > @@ -1364,8 +1364,8 @@ &mcasp0 {
> >  	       0 0 0 0
> >  	>;
> >  	tdm-slots = <2>;
> > -	rx-num-evt = <32>;
> > -	tx-num-evt = <32>;
> > +	rx-num-evt = <0>;
> > +	tx-num-evt = <0>;
> 
> From my understanding of the dt-binding having these properties set to
> <0> is equivalent to not having those properties at all.
> My suggestion would be to just remove those, therefore.
> 

I was following the existing convention of setting it to 0 for 
disabling, as done in k3-j721e-common-proc-board.dts

IMO it is better to keep it explicit, but I don't feel strongly against 
dropping it either (as long as we do the same for j721e). Will defer the 
decision to the tree maintainers.

> With that said, in any case, whatever you do my suggested change or 
> not,
> 
> Acked-by: Francesco Dolcini <francesco.dolcini@toradex.com>
> 
> Francesco
> 

-- 
Thanks,
Jai

GPG Fingerprint: 4DE0 D818 E5D5 75E8 D45A AFC5 43DE 91F9 249A 7145

On 14:28-20240605, Jai Luthra wrote:
> Hi Francesco,
> 
> Thanks for the review.
> 
> On Jun 05, 2024 at 10:47:13 +0200, Francesco Dolcini wrote:
> > Hello Jai, thanks for the patch.
> > 
> > On Tue, Jun 04, 2024 at 03:11:06PM +0530, Jai Luthra wrote:
> > > McASP AFIFOs are not necessary with UDMA-P/BCDMA as there is buffering
> > > on the DMA IP. Drop these for better audio latency.
> > > 
> > > Fixes: 316b80246b16 ("arm64: dts: ti: add verdin am62")
> > > Signed-off-by: Jai Luthra <j-luthra@ti.com>
> > > ---
> > >  arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi | 8 ++++----
> > >  1 file changed, 4 insertions(+), 4 deletions(-)
> > > 
> > > diff --git a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > > index 2038c5e04639..27e5220e1dc7 100644
> > > --- a/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > > +++ b/arch/arm64/boot/dts/ti/k3-am62-verdin.dtsi
> > > @@ -1364,8 +1364,8 @@ &mcasp0 {
> > >  	       0 0 0 0
> > >  	>;
> > >  	tdm-slots = <2>;
> > > -	rx-num-evt = <32>;
> > > -	tx-num-evt = <32>;
> > > +	rx-num-evt = <0>;
> > > +	tx-num-evt = <0>;
> > 
> > From my understanding of the dt-binding having these properties set to
> > <0> is equivalent to not having those properties at all.
> > My suggestion would be to just remove those, therefore.
> > 
> 
> I was following the existing convention of setting it to 0 for 
> disabling, as done in k3-j721e-common-proc-board.dts
> 
> IMO it is better to keep it explicit, but I don't feel strongly against 
> dropping it either (as long as we do the same for j721e). Will defer the 
> decision to the tree maintainers.
> 

If they are equivalent, then just drop them.

-- 
Regards,
Nishanth Menon
Key (0xDDB5849D1736249D) / Fingerprint: F8A2 8693 54EB 8232 17A3  1A34 DDB5 849D 1736 249D