On 02/20, Joachim Eastwood wrote:

+static int lpc18xx_creg_clk_probe(struct platform_device *pdev)
+{
+	struct device_node *np = pdev->dev.of_node;
+	struct regmap *syscon;
+
+	syscon = syscon_node_to_regmap(np->parent);
+	if (IS_ERR(syscon)) {
+		dev_err(&pdev->dev, "syscon lookup failed\n");
+		return PTR_ERR(syscon);
+	}
+
+	clk_creg[CREG_CLK_1KHZ] =
+		clk_register_creg_clk(&clk_creg_clocks[CREG_CLK_1KHZ],
+				      &clk_creg_clocks[CREG_CLK_32KHZ].name,
+				      syscon);
+
+	of_clk_add_provider(np, of_clk_src_onecell_get, &clk_base_data);

If we use the same clk_base_data both times then why can't we
just insert the 1KHz clk into the clk_creg array when the
platform device probes? That would be racy if something was
trying to read the array when it's being modified, so the
alternative non-racy solution would be to make two clk_base_data
structures, one in CLK_OF_DECLARE() and one in platform device
probe.

Otherwise this looks like what I was envisioning.

-- 
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum,
a Linux Foundation Collaborative Project
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html