Re: [PATCH 1/3] arm64: dts: S32G3: Introduce device tree for S32G-VNP-RDB3

On Fri, Mar 15, 2024 at 03:41:14PM -0700, Stephen Boyd wrote:

Quoting Wadim Mueller (2024-03-15 15:27:47)

quoted

diff --git a/include/dt-bindings/clock/nxp,s32-scmi-clock.h b/include/dt-bindings/clock/nxp,s32-scmi-clock.h
new file mode 100644
index 000000000000..240022c1f109
--- /dev/null
+++ b/include/dt-bindings/clock/nxp,s32-scmi-clock.h

@@ -0,0 +1,158 @@
+/* SPDX-License-Identifier: BSD-3-Clause */
+
+#ifndef __DT_BINDINGS_NXP_SCMI_CLOCK_S32_H
+#define __DT_BINDINGS_NXP_SCMI_CLOCK_S32_H
+
+#define S32_SCMI_COMPLEX_CLK           0xFFFFFFFFU
+#define S32_SCMI_NOT_IMPLEMENTED_CLK   0xFFFFFFFEU
+
+#define S32_SCMI_CLK_BASE_ID           0U
+#define S32_SCMI_CLK(N)                        ((N) + S32_SCMI_CLK_BASE_ID)
+#define S32_PLAT_SCMI_CLK(N)           ((N) + S32_SCMI_PLAT_CLK_BASE_ID)

I think we've been recommending that scmi clk consumers simply list the
number instead of making up defines for them.

yes, that makes sense. Was not aware that this is the recommendation. I
will get rid of this file completely, and instead hard code the numbers
into the DT. Thanks for the hint.

quoted

+
+#define S32_SCMI_CLK_VERSION_MAJOR     (1)
+#define S32_SCMI_CLK_VERSION_MINOR     (0)

Why is this part of the dt binding?

yes, this is stupid, you are right. Simply missed to delete it! As the
file will be removed, it will be gone on the next spin. Thanks.