On Sat, Feb 28, 2026 at 08:56:04AM +0800, Peng Fan wrote:

On Fri, Feb 27, 2026 at 03:32:16PM +0000, Cristian Marussi wrote:

quoted

Use the Clock protocol layer determine_rate logic to calculate the closest
rate that can be supported by a specific clock.

No functional change.

Cc: Brian Masney <bmasney@redhat.com>
Cc: Michael Turquette <mturquette@baylibre.com>
Cc: Stephen Boyd <sboyd@kernel.org>
Cc: linux-clk@vger.kernel.org
Signed-off-by: Cristian Marussi <cristian.marussi@arm.com>
---
Note that the calculation logic in the protocol layer is exactly the same
as it wes here.

@Brian I suppose once your CLK_ROUNDING_FW_MANAGED sereis is merged I can flag
such SCMI clocks.

Per my reading of Brain's thread, if ->determine_rate exists,
->determine_rate() will be used.

 	} else if (core->ops->determine_rate) {
 		return core->ops->determine_rate(core->hw, req);
+	} else if (clk_is_rounding_fw_managed(core)) {
+		return 0;

So unless update scmi_clk_determine_rate() to something:
--------
if (clk & CLK_ROUNDING_FW_MANAGED)
	return 0;

return scmi_proto_clk_ops->determine_rate(clk->ph, clk->id, &req->rate);
--------

It maybe better to update Brain's patch to move clk_is_rounding_fw_managed()
above the check of core->ops->determine_rate().

The clk framework has some basic sanity checks in place that are called
during device probe to ensure that various ops are configured properly. I
could add a check that if CLK_ROUNDING_FW_MANAGED [*] is set, and a
determine_rate() op is set, then it gives an error.

[*] Note: I am tentatively planning to rename that to CLK_ROUNDING_NOOP
in v2 in about a week.

Brian