On 6/8/19 3:19 PM, Geert Uytterhoeven wrote:

Hi Leonard,

On Sat, Jun 8, 2019 at 9:26 AM Leonard Crestez [off-list ref] wrote:

quoted

On 6/7/19 10:05 PM, Stephen Boyd wrote:

quoted

Quoting Leonard Crestez (2019-05-24 01:25:25)

quoted

quoted

@@ -3040,10 +3051,11 @@ static void clk_debug_create_one(struct clk_core *core, struct dentry *pdentry)
          debugfs_create_u32("clk_enable_count", 0444, root, &core->enable_count);
          debugfs_create_u32("clk_protect_count", 0444, root, &core->protect_count);
          debugfs_create_u32("clk_notifier_count", 0444, root, &core->notifier_count);
          debugfs_create_file("clk_duty_cycle", 0444, root, core,
                              &clk_duty_cycle_fops);
+       debugfs_create_file("clk_parent", 0444, root, core, &current_parent_fops);

Shouldn't we skip creation of this file if core->num_parents == 0? So
put this under the if condition below?

It's still useful to determine clk tree structure from debugfs fields,
otherwise you'd have to extract by parsing other files.

Would you hide clk_rate for fixed-rate? I'd rather have everything
available for uniformity, even if it's otherwise constant at runtime.

Unless I' missing something, there's a big difference here: all clocks
have a rate, but not all clocks have a parent.

Sorry, I got confused and thought that condition checked for muxes so 
the parent of intermediate clk nodes will be hidden. But intermediate 
nodes have num_parents == 1.

Will fix in v2.

--
Regards,
Leonard

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel