The MT7622 device tree never bothered to specify the number of virtual DMA
channels for the HSDMA controller, always falling back to the default value of
3. Make this value explicit, in order to avoid the following dmesg notification:

mtk_hsdma 1b007000.dma-controller: Using 3 as missing dma-requests property

Signed-off-by: Rui Salvaterra <rsalvaterra@gmail.com>
---
 arch/arm64/boot/dts/mediatek/mt7622.dtsi | 1 +
 1 file changed, 1 insertion(+)

diff --git a/arch/arm64/boot/dts/mediatek/mt7622.dtsi b/arch/arm64/boot/dts/mediatek/mt7622.dtsi
index 3d6eaf6dd078..5551f004945b 100644
--- a/arch/arm64/boot/dts/mediatek/mt7622.dtsi
+++ b/arch/arm64/boot/dts/mediatek/mt7622.dtsi
@@ -924,6 +924,7 @@ hsdma: dma-controller@1b007000 {
 		clock-names = "hsdma";
 		power-domains = <&scpsys MT7622_POWER_DOMAIN_ETHSYS>;
 		#dma-cells = <1>;
+		dma-requests = <3>;
 	};
 
 	eth: ethernet@1b100000 {
-- 
2.36.0

On 29/04/2022 10:42, Rui Salvaterra wrote:
> The MT7622 device tree never bothered to specify the number of virtual DMA
> channels for the HSDMA controller, always falling back to the default value of
> 3. Make this value explicit, in order to avoid the following dmesg notification:
> 
> mtk_hsdma 1b007000.dma-controller: Using 3 as missing dma-requests property
> 
> Signed-off-by: Rui Salvaterra <rsalvaterra@gmail.com>

Applied, thanks!

> ---
>   arch/arm64/boot/dts/mediatek/mt7622.dtsi | 1 +
>   1 file changed, 1 insertion(+)
> 
> diff --git a/arch/arm64/boot/dts/mediatek/mt7622.dtsi b/arch/arm64/boot/dts/mediatek/mt7622.dtsi
> index 3d6eaf6dd078..5551f004945b 100644
> --- a/arch/arm64/boot/dts/mediatek/mt7622.dtsi
> +++ b/arch/arm64/boot/dts/mediatek/mt7622.dtsi
> @@ -924,6 +924,7 @@ hsdma: dma-controller@1b007000 {
>   		clock-names = "hsdma";
>   		power-domains = <&scpsys MT7622_POWER_DOMAIN_ETHSYS>;
>   		#dma-cells = <1>;
> +		dma-requests = <3>;
>   	};
>   
>   	eth: ethernet@1b100000 {

Hi, Matthias,

On Fri, 29 Apr 2022 at 12:57, Matthias Brugger <matthias.bgg@gmail.com> wrote:
>
> Applied, thanks!

Would this and the cache topology patch qualify as stable material?
Intuitively, I'd say yes, but I'm not really sure.

Thanks,
Rui

Hi Rui,

On 29/04/2022 15:05, Rui Salvaterra wrote:
> Hi, Matthias,
> 
> On Fri, 29 Apr 2022 at 12:57, Matthias Brugger <matthias.bgg@gmail.com> wrote:
>>
>> Applied, thanks!
> 
> Would this and the cache topology patch qualify as stable material?
> Intuitively, I'd say yes, but I'm not really sure.
> 

my understanding is that these fix some warning, but no real bug in the sense 
something does not work without them. So no I don't think they are stable material.

Regards,
Matthias