Hello,

On Fri, 25 Jul 2014 09:29:09 -0500, Rob Herring wrote:

> > Any idea on how can we get some temporary solution for 3.17 as we didn't
> > conclude anything yet on bindings ?
>
> A temporary solution would have to be NOT in DT because once you add
> something into DT you are stuck with it for some time. You could
> simply support independent clocks by looking at the platform type, but
> that is still risky since you may want to define the OPP in DT
> differently for the 2 cases.
>
> The other problem with temporary solutions is once they are accepted,
> people loose motivation to create the permanent solution. "Can you
> take this now and I'll fix the issues later" is a red flag to
> maintainers.

On the Marvell Armada XP side of things, the OPPs are registered
dynamically because they change from one board to the other. The only
thing that is in the DT for cpufreq-generic is the clock-latency
parameter.

Thanks,

Thomas
--
Thomas Petazzoni, CTO, Free Electrons
Embedded Linux, Kernel and Android engineering
http://free-electrons.com

2014-07-28 14:01:43

by Grant Likely

[permalink] [raw]

Subject: Re: [PATCH 02/14] clk: Create of_clk_shared_by_cpus()

On Tue, 1 Jul 2014 22:02:31 +0530, Viresh Kumar <[email protected]> wrote:
> Create a new routine of_clk_shared_by_cpus() that finds if clock lines are
> shared between two CPUs. This is verified by comparing "clocks" property from
> CPU's DT node.
>
> Returns 1 if clock line is shared between them, 0 if clock isn't shared and
> return appropriate errors in case nodes/properties are missing.
>
> Cc: Mike Turquette <[email protected]>
> Signed-off-by: Viresh Kumar <[email protected]>
> ---
> drivers/clk/clk.c | 56 +++++++++++++++++++++++++++++++++++++++++++++++++++++
> include/linux/clk.h | 6 ++++++
> 2 files changed, 62 insertions(+)
>
> diff --git a/drivers/clk/clk.c b/drivers/clk/clk.c
> index 8b73ede..497735c 100644
> --- a/drivers/clk/clk.c
> +++ b/drivers/clk/clk.c
> @@ -10,6 +10,7 @@
> */
>
> #include <linux/clk-private.h>
> +#include <linux/cpu.h>
> #include <linux/module.h>
> #include <linux/mutex.h>
> #include <linux/spinlock.h>
> @@ -2528,6 +2529,61 @@ const char *of_clk_get_parent_name(struct device_node *np, int index)
> }
> EXPORT_SYMBOL_GPL(of_clk_get_parent_name);
>
> +/**
> + * of_clk_shared_by_cpus - Finds if clock line is shared between CPUs
> + * @cpu1, cpu2: CPU numbers
> + *
> + * Finds if clock lines are shared by two CPUs. This is verified by comparing
> + * "clocks" property from CPU's DT node.
> + *
> + * Returns 1 if clock line is shared between them, 0 if clock isn't shared.
> + * Return appropriate errors in case some requirements aren't met.
> + */
> +int of_clk_shared_by_cpus(int cpu1, int cpu2)
> +{
> + struct device *cpu1_dev, *cpu2_dev;
> + struct device_node *np1, *np2;
> + int ret;
> +
> + cpu1_dev = get_cpu_device(cpu1);
> + if (!cpu1_dev) {
> + pr_err("%s: failed to get cpu_dev for cpu%d\n", __func__, cpu1);
> + return -ENODEV;
> + }
> +
> + cpu2_dev = get_cpu_device(cpu2);
> + if (!cpu2_dev) {
> + pr_err("%s: failed to get cpu_dev for cpu%d\n", __func__, cpu2);
> + return -ENODEV;
> + }
> +
> + np1 = of_node_get(cpu1_dev->of_node);
> + if (!np1) {
> + pr_err("%s: failed to find of_node for cpu%d\n", __func__,
> + cpu1);
> + return -ENODEV;
> + }
> +
> + np2 = of_node_get(cpu2_dev->of_node);
> + if (!np2) {
> + pr_err("%s: failed to find of_node for cpu%d\n", __func__,
> + cpu2);
> + ret = -ENODEV;
> + goto put_np1;
> + }
> +
> + /* Match "clocks" property */
> + ret = of_property_match(np1, np2, "clocks");

This looks naïve. It is entirely possible for different clock specifiers
to resolve to the same clock line, or for there to be multple clocks in
the clocks property. This looks like a buggy way to do it. The only
reliable way to determine if two clocks resolve to the same thing is to
resolve the references with the clock controller.

g.