It has been reported on IRC and in KernelCI boot tests, this change breaks
internal PHY support on the Amlogic G12A/SM1 Based boards.
We suspect the added signal to reset more than the Ethernet MAC but also
the MDIO/(RG)MII mux used to redirect the MAC signals to the internal PHY.
This reverts commit f3362f0c18174a1f334a419ab7d567a36bd1b3f3 while we find
and acceptable solution to cleanly reset the Ethernet MAC.
Reported-by: Corentin Labbe <[email protected]>
Acked-by: Jérôme Brunet <[email protected]>
Signed-off-by: Neil Armstrong <[email protected]>
---
Hi Kevin,
This has been reported to also break on 5.10, when this lands on 5.11, I'll send another
patch for 5.10 because meson-axg.dtsi needs a conflict resolution on 5.11.
Jerome gave me his ack off-list.
Neil
arch/arm64/boot/dts/amlogic/meson-axg.dtsi | 2 --
arch/arm64/boot/dts/amlogic/meson-g12-common.dtsi | 2 --
arch/arm64/boot/dts/amlogic/meson-gx.dtsi | 3 ---
3 files changed, 7 deletions(-)
diff --git a/arch/arm64/boot/dts/amlogic/meson-axg.dtsi b/arch/arm64/boot/dts/amlogic/meson-axg.dtsi
index ba1c6dfdc4b6..d945c84ab697 100644
--- a/arch/arm64/boot/dts/amlogic/meson-axg.dtsi
+++ b/arch/arm64/boot/dts/amlogic/meson-axg.dtsi
@@ -280,8 +280,6 @@ ethmac: ethernet@ff3f0000 {
"timing-adjustment";
rx-fifo-depth = <4096>;
tx-fifo-depth = <2048>;
- resets = <&reset RESET_ETHERNET>;
- reset-names = "stmmaceth";
power-domains = <&pwrc PWRC_AXG_ETHERNET_MEM_ID>;
status = "disabled";
};
diff --git a/arch/arm64/boot/dts/amlogic/meson-g12-common.dtsi b/arch/arm64/boot/dts/amlogic/meson-g12-common.dtsi
index 9c90d562ada1..851eeb5e9de1 100644
--- a/arch/arm64/boot/dts/amlogic/meson-g12-common.dtsi
+++ b/arch/arm64/boot/dts/amlogic/meson-g12-common.dtsi
@@ -224,8 +224,6 @@ ethmac: ethernet@ff3f0000 {
"timing-adjustment";
rx-fifo-depth = <4096>;
tx-fifo-depth = <2048>;
- resets = <&reset RESET_ETHERNET>;
- reset-names = "stmmaceth";
status = "disabled";
mdio0: mdio {
diff --git a/arch/arm64/boot/dts/amlogic/meson-gx.dtsi b/arch/arm64/boot/dts/amlogic/meson-gx.dtsi
index 726b91d3a905..0edd137151f8 100644
--- a/arch/arm64/boot/dts/amlogic/meson-gx.dtsi
+++ b/arch/arm64/boot/dts/amlogic/meson-gx.dtsi
@@ -13,7 +13,6 @@
#include <dt-bindings/interrupt-controller/irq.h>
#include <dt-bindings/interrupt-controller/arm-gic.h>
#include <dt-bindings/power/meson-gxbb-power.h>
-#include <dt-bindings/reset/amlogic,meson-gxbb-reset.h>
#include <dt-bindings/thermal/thermal.h>
/ {
@@ -576,8 +575,6 @@ ethmac: ethernet@c9410000 {
interrupt-names = "macirq";
rx-fifo-depth = <4096>;
tx-fifo-depth = <2048>;
- resets = <&reset RESET_ETHERNET>;
- reset-names = "stmmaceth";
power-domains = <&pwrc PWRC_GXBB_ETHERNET_MEM_ID>;
status = "disabled";
};
--
2.25.1
On Tue, 26 Jan 2021 09:09:51 +0100, Neil Armstrong wrote:
> It has been reported on IRC and in KernelCI boot tests, this change breaks
> internal PHY support on the Amlogic G12A/SM1 Based boards.
>
> We suspect the added signal to reset more than the Ethernet MAC but also
> the MDIO/(RG)MII mux used to redirect the MAC signals to the internal PHY.
>
> This reverts commit f3362f0c18174a1f334a419ab7d567a36bd1b3f3 while we find
> and acceptable solution to cleanly reset the Ethernet MAC.
Applied, thanks!
[1/1] Revert "arm64: dts: amlogic: add missing ethernet reset ID"
commit: 19f6fe976a61f9afc289b062b7ef67f99b72e7b9
Best regards,
--
Kevin Hilman <[email protected]>
Hi Neil,
Neil Armstrong <[email protected]> writes:
> It has been reported on IRC and in KernelCI boot tests, this change breaks
> internal PHY support on the Amlogic G12A/SM1 Based boards.
>
> We suspect the added signal to reset more than the Ethernet MAC but also
> the MDIO/(RG)MII mux used to redirect the MAC signals to the internal PHY.
>
> This reverts commit f3362f0c18174a1f334a419ab7d567a36bd1b3f3 while we find
> and acceptable solution to cleanly reset the Ethernet MAC.
>
> Reported-by: Corentin Labbe <[email protected]>
> Acked-by: Jérôme Brunet <[email protected]>
> Signed-off-by: Neil Armstrong <[email protected]>
> ---
> Hi Kevin,
>
> This has been reported to also break on 5.10, when this lands on 5.11, I'll send another
> patch for 5.10 because meson-axg.dtsi needs a conflict resolution on 5.11.
Looks like this never got submitted to v5.10. I just discovered it that
v5.10 in SEI610 (internal PHY) is broken.
Could you submit this to v5.10 stable also please?
Thanks,
Kevin
On 03/03/2021 00:28, Kevin Hilman wrote:
> Hi Neil,
>
> Neil Armstrong <[email protected]> writes:
>
>> It has been reported on IRC and in KernelCI boot tests, this change breaks
>> internal PHY support on the Amlogic G12A/SM1 Based boards.
>>
>> We suspect the added signal to reset more than the Ethernet MAC but also
>> the MDIO/(RG)MII mux used to redirect the MAC signals to the internal PHY.
>>
>> This reverts commit f3362f0c18174a1f334a419ab7d567a36bd1b3f3 while we find
>> and acceptable solution to cleanly reset the Ethernet MAC.
>>
>> Reported-by: Corentin Labbe <[email protected]>
>> Acked-by: Jérôme Brunet <[email protected]>
>> Signed-off-by: Neil Armstrong <[email protected]>
>> ---
>> Hi Kevin,
>>
>> This has been reported to also break on 5.10, when this lands on 5.11, I'll send another
>> patch for 5.10 because meson-axg.dtsi needs a conflict resolution on 5.11.
>
> Looks like this never got submitted to v5.10. I just discovered it that
> v5.10 in SEI610 (internal PHY) is broken.
>
> Could you submit this to v5.10 stable also please?
Sure, I'll prepare this.
Neil
>
> Thanks,
>
> Kevin
>
>
>
>
>