On 2/5/2019 3:06 PM, Stephen Boyd wrote:

Quoting Jeffrey Hugo (2019-01-30 08:36:11)
[..]

quoted

+};
+MODULE_DEVICE_TABLE(of, mmcc_msm8998_match_table);
+
+static int mmcc_msm8998_probe(struct platform_device *pdev)
+{
+       struct device *dev = &pdev->dev;
+       int i, ret;
+       struct regmap *regmap;
+       struct clk *c;
+
+       c = devm_clk_get(dev, "xo");
+       if (!IS_ERR(c))
+               dummy_xo_init.parent_names =
+                                       (const char *[]){__clk_get_name(c)};
+       else
+               return PTR_ERR(c);
+
+       c = devm_clk_get(dev, "gpll0");
+       if (!IS_ERR(c))
+               dummy_gpll0_init.parent_names =
+                                       (const char *[]){__clk_get_name(c)};
+       else
+               return PTR_ERR(c);

Urgh, my parent mapping series can't come be merged fast enough it
seems!

Sorry, I'm not familiar with the context.  Something I should incorporate?

quoted

+
+       regmap = qcom_cc_map(pdev, &mmcc_msm8998_desc);
+       if (IS_ERR(regmap))
+               return PTR_ERR(regmap);
+
+       for (i = 0; i < ARRAY_SIZE(mmcc_msm8998_hws); i++) {
+               ret = devm_clk_hw_register(dev, mmcc_msm8998_hws[i]);

I think we should make this common to the qcom_cc_desc now. Can you
throw in an optional list of clk_hw pointers that also get registered as
its own patch and update the existing drivers that are all doing this?

Yeah, can do.

quoted

+               if (ret)
+                       return ret;
+       }
+
+       return qcom_cc_really_probe(pdev, &mmcc_msm8998_desc, regmap);
+}

-- 
Jeffrey Hugo
Qualcomm Datacenter Technologies as an affiliate of Qualcomm 
Technologies, Inc.
Qualcomm Technologies, Inc. is a member of the
Code Aurora Forum, a Linux Foundation Collaborative Project.