Thread (6 messages) 6 messages, 3 authors, 2025-03-02

Re: [PATCH v1] arm64: dts: ti: k3-am62p: Enable AUDIO_REFCLKx

From: Nishanth Menon <nm@ti.com>
Date: 2025-02-28 13:53:43
Also in: linux-devicetree, lkml

On 13:43-20250228, Francesco Dolcini wrote:
On Fri, Feb 07, 2025 at 09:33:01AM +0100, Francesco Dolcini wrote:
quoted
On Thu, Feb 06, 2025 at 07:42:39PM -0600, Nishanth Menon wrote:
quoted
On 16:39-20250206, Francesco Dolcini wrote:
quoted
From: Francesco Dolcini <redacted>

On AM62P-based SoCs the AUDIO_REFCLKx clocks can be used as an input to
external peripherals when configured through CTRL_MMR, so add the
clock nodes.

Link: http://downloads.ti.com/tisci/esd/latest/5_soc_doc/am62px/clocks.html
Signed-off-by: Francesco Dolcini <redacted>
---
 arch/arm64/boot/dts/ti/k3-am62p-main.dtsi | 20 ++++++++++++++++++++
 1 file changed, 20 insertions(+)
diff --git a/arch/arm64/boot/dts/ti/k3-am62p-main.dtsi b/arch/arm64/boot/dts/ti/k3-am62p-main.dtsi
index 420c77c8e9e5..4b47b0774330 100644
--- a/arch/arm64/boot/dts/ti/k3-am62p-main.dtsi
+++ b/arch/arm64/boot/dts/ti/k3-am62p-main.dtsi
@@ -42,6 +42,26 @@ &inta_main_dmss {
 	ti,interrupt-ranges = <5 69 35>;
 };
 
+&main_conf {
	Why not add it to main_conf section it self in the file?
The reason is that main_conf is defined in k3-am62p-j722s-common-main.dtsi,
that is shared between am62p and j722s.

On j722s the audio refclk is added in k3-j722s-main.dtsi the same way as I did
here, so I cannot move this to k3-am62p-j722s-common-main.dtsi without updating
also k3-j722s-main.dtsi.

I looked into the differences of j722s and am62p, and from my understanding,
from the audio refclk point of view, they are identical (same IP, same reg, same
clocks and same IDs), so this should naturally be moved to
k3-am62p-j722s-common-main.dtsi as you are suggesting.

... however, for some reason I am not aware of, on k3-j722s-main.dtsi a different
parent clock is used, and I cannot understand the reason. The actual parent clocks
in this patch are just the same we already have everywhere apart on j722s. I tried
to look at the history of this and it seems that on the TI downstream kernel branch
this is defined in the board dts file (!) and this confused me even more.

So, not wanting to break stuff I was not able to understand I came up with this
proposal.

An alternative could be to override the "unexpected" clocks from
k3-j722s-main.dtsi to the board dts file, and have the "standard" clocks, as
proposed in this patch and already used on all the other AM62 variants, in
k3-am62p-j722s-common-main.dtsi.

+Jayesh that is the author of this specific change in k3-j722s-main.dtsi.
Jayesh, Nishanth, any comment on this?

Should I proceed with this last option ?

   An alternative could be to override the "unexpected" clocks from
   k3-j722s-main.dtsi to the board dts file, and have the "standard" clocks, as
   proposed in this patch and already used on all the other AM62 variants, in
   k3-am62p-j722s-common-main.dtsi.
no, with your explanation it is clear that your patch matches with the
strategy we are using currently.

-- 
Regards,
Nishanth Menon
Key (0xDDB5849D1736249D) / Fingerprint: F8A2 8693 54EB 8232 17A3  1A34 DDB5 849D 1736 249D
Keyboard shortcuts
hback out one level
jnext message in thread
kprevious message in thread
ldrill in
Escclose help / fold thread tree
?toggle this help