Skip to contentSkip to conversation

clk: Mark a fwnode as initialized when using CLK_OF_DECLARE() macro

PATCH Applied Saravana Kannan 4 participants 9 messages Mar 2, 2023 – Mar 8, 2023

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

PATCH v2 1/1 clk: Mark a fwnode as initialized when using CLK_OF_DECLARE() macro

We already mark fwnodes as initialized when they are registered as clock providers. We do this so that fw_devlink can tell when a clock driver doesn't use the driver core framework to probe/initialize its device. This ensures fw_devlink doesn't block the consumers of such a clock provider indefinitely.

However, some users of CLK_OF_DECLARE() macros don't use the same node that matches the macro as the node for the clock provider, but they initialize the entire node. To cover these cases, also mark the nodes that match the macros as initialized when the init callback function is called.

An example of this is "stericsson,u8500-clks" that's handled using CLK_OF_DECLARE() and looks something like this:

clocks {
	compatible = "stericsson,u8500-clks";

	prcmu_clk: prcmu-clock {
		#clock-cells = <1>;
	};

	prcc_pclk: prcc-periph-clock {
		#clock-cells = <2>;
	};

	prcc_kclk: prcc-kernel-clock {
		#clock-cells = <2>;
	};

	prcc_reset: prcc-reset-controller {
		#reset-cells = <2>;
	};
	...
	...
};

This patch makes sure that "clocks" is marked as initialized so that fw_devlink knows that all nodes under it have been initialized.

If the driver creates struct devices for some of the subnodes, fw_devlink is smart enough to know to wait for those devices to probe. So, no special handling is required for those cases.

---
 include/linux/clk-provider.h | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
include/linux/clk-provider.h +7 -1
index 842e72a5348f..c9f5276006a0 100644
@@ -1363,7 +1363,13 @@ struct clk_hw_onecell_data {
 	struct clk_hw *hws[];
 };
 
-#define CLK_OF_DECLARE(name, compat, fn) OF_DECLARE_1(clk, name, compat, fn)
+#define CLK_OF_DECLARE(name, compat, fn) \
+	static void __init name##_of_clk_init_declare(struct device_node *np) \
+	{								\
+		fn(np);							\
+		fwnode_dev_initialized(of_fwnode_handle(np), true);	\
+	}								\
+	OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
 
 /*
  * Use this macro when you have a driver that requires two initialization
Signature · 1 line
2.39.2.722.g9855ee24e9-goog
#1 lore raw
Quoted 45 lines · Saravana Kannan

We already mark fwnodes as initialized when they are registered as clock providers. We do this so that fw_devlink can tell when a clock driver doesn't use the driver core framework to probe/initialize its device. This ensures fw_devlink doesn't block the consumers of such a clock provider indefinitely.

However, some users of CLK_OF_DECLARE() macros don't use the same node that matches the macro as the node for the clock provider, but they initialize the entire node. To cover these cases, also mark the nodes that match the macros as initialized when the init callback function is called.

An example of this is "stericsson,u8500-clks" that's handled using CLK_OF_DECLARE() and looks something like this:

clocks {
        compatible = "stericsson,u8500-clks";

        prcmu_clk: prcmu-clock {
                #clock-cells = <1>;
        };

        prcc_pclk: prcc-periph-clock {
                #clock-cells = <2>;
        };

        prcc_kclk: prcc-kernel-clock {
                #clock-cells = <2>;
        };

        prcc_reset: prcc-reset-controller {
                #reset-cells = <2>;
        };
        ...
        ...
};

This patch makes sure that "clocks" is marked as initialized so that fw_devlink knows that all nodes under it have been initialized.

If the driver creates struct devices for some of the subnodes, fw_devlink is smart enough to know to wait for those devices to probe. So, no special handling is required for those cases.

Quoted 5 lines

Linus, I changed the code a bit and I think this should work too. Can you give it a shot please?

-Saravana

Quoted 26 lines
---
 include/linux/clk-provider.h | 8 +++++++-
 1 file changed, 7 insertions(+), 1 deletion(-)
include/linux/clk-provider.h +7 -2
index 842e72a5348f..c9f5276006a0 100644
@@ -1363,7 +1363,13 @@ struct clk_hw_onecell_data {
        struct clk_hw *hws[];
 };
 
-#define CLK_OF_DECLARE(name, compat, fn) OF_DECLARE_1(clk, name, compat, fn)
+#define CLK_OF_DECLARE(name, compat, fn) \
+       static void __init name##_of_clk_init_declare(struct device_node *np) \
+       {                                                               \
+               fn(np);                                                 \
+               fwnode_dev_initialized(of_fwnode_handle(np), true);     \
+       }                                                               \
+       OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
 
 /*
  * Use this macro when you have a driver that requires two initialization
--

2.39.2.722.g9855ee24e9-goog

Tested-by Reviewed-by on v2 1/1
#2 lore raw
In reply to Saravana Kannan
Quoted 44 lines · Saravana Kannan

We already mark fwnodes as initialized when they are registered as clock providers. We do this so that fw_devlink can tell when a clock driver doesn't use the driver core framework to probe/initialize its device. This ensures fw_devlink doesn't block the consumers of such a clock provider indefinitely.

However, some users of CLK_OF_DECLARE() macros don't use the same node that matches the macro as the node for the clock provider, but they initialize the entire node. To cover these cases, also mark the nodes that match the macros as initialized when the init callback function is called.

An example of this is "stericsson,u8500-clks" that's handled using CLK_OF_DECLARE() and looks something like this:

clocks {
        compatible = "stericsson,u8500-clks";

        prcmu_clk: prcmu-clock {
                #clock-cells = <1>;
        };

        prcc_pclk: prcc-periph-clock {
                #clock-cells = <2>;
        };

        prcc_kclk: prcc-kernel-clock {
                #clock-cells = <2>;
        };

        prcc_reset: prcc-reset-controller {
                #reset-cells = <2>;
        };
        ...
        ...
};

This patch makes sure that "clocks" is marked as initialized so that fw_devlink knows that all nodes under it have been initialized.

If the driver creates struct devices for some of the subnodes, fw_devlink is smart enough to know to wait for those devices to probe. So, no special handling is required for those cases.

Quoted 5 lines

Works like a charm on U8500!

Yours,
Linus Walleij

#3 lore raw
Quoted 53 lines · Linus Walleij
Quoted 44 lines · Saravana Kannan

We already mark fwnodes as initialized when they are registered as clock providers. We do this so that fw_devlink can tell when a clock driver doesn't use the driver core framework to probe/initialize its device. This ensures fw_devlink doesn't block the consumers of such a clock provider indefinitely.

However, some users of CLK_OF_DECLARE() macros don't use the same node that matches the macro as the node for the clock provider, but they initialize the entire node. To cover these cases, also mark the nodes that match the macros as initialized when the init callback function is called.

An example of this is "stericsson,u8500-clks" that's handled using CLK_OF_DECLARE() and looks something like this:

clocks {
        compatible = "stericsson,u8500-clks";

        prcmu_clk: prcmu-clock {
                #clock-cells = <1>;
        };

        prcc_pclk: prcc-periph-clock {
                #clock-cells = <2>;
        };

        prcc_kclk: prcc-kernel-clock {
                #clock-cells = <2>;
        };

        prcc_reset: prcc-reset-controller {
                #reset-cells = <2>;
        };
        ...
        ...
};

This patch makes sure that "clocks" is marked as initialized so that fw_devlink knows that all nodes under it have been initialized.

If the driver creates struct devices for some of the subnodes, fw_devlink is smart enough to know to wait for those devices to probe. So, no special handling is required for those cases.

Quoted 5 lines
Quoted 3 lines

Works like a charm on U8500!

Stephen, Does this look good? Can we have Greg pull this in?

-Saravana

#5 lore raw
Quoted 11 lines · Stephen Boyd

Quoting Saravana Kannan (2023-03-03 13:24:00)

Quoted 8 lines · Linus Walleij
Quoted 4 lines

Works like a charm on U8500!

Stephen, Does this look good? Can we have Greg pull this in?

I picked it up for fixes. Will send it off later this week.

This change causes a build regression:

drivers/clk/mvebu/kirkwood.c:358:1: error: expected identifier or '(' CLK_OF_DECLARE(98dx1135_clk, "marvell,mv98dx1135-core-clock",
^
include/linux/clk-provider.h:1367:21: note: expanded from macro 'CLK_OF_DECLARE'

          static void __init name##_of_clk_init_declare(struct device_node *np) \
                             ^
  <scratch space>:124:1: note: expanded from here
  98dx1135_clk_of_clk_init_declare
  ^
  drivers/clk/mvebu/kirkwood.c:358:1: error: invalid digit 'd' in decimal constant
  include/linux/clk-provider.h:1372:34: note: expanded from macro 'CLK_OF_DECLARE'
          OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
                                          ^
  <scratch space>:125:3: note: expanded from here
  98dx1135_clk_of_clk_init_declare
    ^
  drivers/clk/mvebu/kirkwood.c:358:1: error: invalid digit 'd' in decimal constant
  include/linux/clk-provider.h:1372:34: note: expanded from macro 'CLK_OF_DECLARE'
          OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
                                          ^
  <scratch space>:125:3: note: expanded from here
  98dx1135_clk_of_clk_init_declare
    ^
  drivers/clk/mvebu/kirkwood.c:358:1: error: invalid digit 'd' in decimal constant
  include/linux/clk-provider.h:1372:34: note: expanded from macro 'CLK_OF_DECLARE'
          OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
                                          ^
  <scratch space>:125:3: note: expanded from here
  98dx1135_clk_of_clk_init_declare
    ^

4 errors generated.

The clock's name starts with a number, resulting in an invalid C function name. Are clock names ABI? Should this be fixed by adjusting the name like so or adjusting where the name appears in the function defined by CLK_OF_DECLARE()?

drivers/clk/mvebu/kirkwood.c +1 -1
index 8bc893df4736..5d0a7e3bfde5 100644
@@ -355,5 +355,5 @@ CLK_OF_DECLARE(kirkwood_clk, "marvell,kirkwood-core-clock",
 	       kirkwood_clk_init);
 CLK_OF_DECLARE(mv88f6180_clk, "marvell,mv88f6180-core-clock",
 	       kirkwood_clk_init);
-CLK_OF_DECLARE(98dx1135_clk, "marvell,mv98dx1135-core-clock",
+CLK_OF_DECLARE(mv98dx1135_clk, "marvell,mv98dx1135-core-clock",
 	       kirkwood_clk_init);

Cheers,
Nathan

#6 lore raw
Quoted 46 lines · Nathan Chancellor
Quoted 11 lines · Stephen Boyd

Quoting Saravana Kannan (2023-03-03 13:24:00)

Quoted 8 lines · Linus Walleij
Quoted 4 lines

Works like a charm on U8500!

Stephen, Does this look good? Can we have Greg pull this in?

I picked it up for fixes. Will send it off later this week.

This change causes a build regression:

drivers/clk/mvebu/kirkwood.c:358:1: error: expected identifier or '(' CLK_OF_DECLARE(98dx1135_clk, "marvell,mv98dx1135-core-clock",
^
include/linux/clk-provider.h:1367:21: note: expanded from macro 'CLK_OF_DECLARE'

          static void __init name##_of_clk_init_declare(struct device_node *np) \
                             ^
  <scratch space>:124:1: note: expanded from here
  98dx1135_clk_of_clk_init_declare
  ^
  drivers/clk/mvebu/kirkwood.c:358:1: error: invalid digit 'd' in decimal constant
  include/linux/clk-provider.h:1372:34: note: expanded from macro 'CLK_OF_DECLARE'
          OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
                                          ^
  <scratch space>:125:3: note: expanded from here
  98dx1135_clk_of_clk_init_declare
    ^
  drivers/clk/mvebu/kirkwood.c:358:1: error: invalid digit 'd' in decimal constant
  include/linux/clk-provider.h:1372:34: note: expanded from macro 'CLK_OF_DECLARE'
          OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
                                          ^
  <scratch space>:125:3: note: expanded from here
  98dx1135_clk_of_clk_init_declare
    ^
  drivers/clk/mvebu/kirkwood.c:358:1: error: invalid digit 'd' in decimal constant
  include/linux/clk-provider.h:1372:34: note: expanded from macro 'CLK_OF_DECLARE'
          OF_DECLARE_1(clk, name, compat, name##_of_clk_init_declare)
                                          ^
  <scratch space>:125:3: note: expanded from here
  98dx1135_clk_of_clk_init_declare
    ^
Quoted 7 lines

4 errors generated.

The clock's name starts with a number, resulting in an invalid C function name. Are clock names ABI? Should this be fixed by adjusting the name like so or adjusting where the name appears in the function defined by CLK_OF_DECLARE()?

Not sure if the name matters, but it's just simpler to fix the macro than add churn to all those drivers. Can you send a patch that just adds __ in the front of the function name in the macro please?

Thanks,
Saravana

Quoted 14 lines
drivers/clk/mvebu/kirkwood.c +1 -1
index 8bc893df4736..5d0a7e3bfde5 100644
@@ -355,5 +355,5 @@ CLK_OF_DECLARE(kirkwood_clk, "marvell,kirkwood-core-clock",
               kirkwood_clk_init);
 CLK_OF_DECLARE(mv88f6180_clk, "marvell,mv88f6180-core-clock",
               kirkwood_clk_init);
-CLK_OF_DECLARE(98dx1135_clk, "marvell,mv98dx1135-core-clock",
+CLK_OF_DECLARE(mv98dx1135_clk, "marvell,mv98dx1135-core-clock",
               kirkwood_clk_init);

Cheers,
Nathan

maintainer Applied
#7 lore raw
In reply to Saravana Kannan

Quoting Saravana Kannan (2023-03-01 17:46:38)

Quoted 44 lines

We already mark fwnodes as initialized when they are registered as clock providers. We do this so that fw_devlink can tell when a clock driver doesn't use the driver core framework to probe/initialize its device. This ensures fw_devlink doesn't block the consumers of such a clock provider indefinitely.

However, some users of CLK_OF_DECLARE() macros don't use the same node that matches the macro as the node for the clock provider, but they initialize the entire node. To cover these cases, also mark the nodes that match the macros as initialized when the init callback function is called.

An example of this is "stericsson,u8500-clks" that's handled using CLK_OF_DECLARE() and looks something like this:

clocks {
        compatible = "stericsson,u8500-clks";

        prcmu_clk: prcmu-clock {
                #clock-cells = <1>;
        };

        prcc_pclk: prcc-periph-clock {
                #clock-cells = <2>;
        };

        prcc_kclk: prcc-kernel-clock {
                #clock-cells = <2>;
        };

        prcc_reset: prcc-reset-controller {
                #reset-cells = <2>;
        };
        ...
        ...
};

This patch makes sure that "clocks" is marked as initialized so that fw_devlink knows that all nodes under it have been initialized.

If the driver creates struct devices for some of the subnodes, fw_devlink is smart enough to know to wait for those devices to probe. So, no special handling is required for those cases.

Quoted 6 lines

---

Applied to clk-fixes

#8 lore raw
Quoted 46 lines · Stephen Boyd

Quoting Saravana Kannan (2023-03-01 17:46:38)

Quoted 44 lines

We already mark fwnodes as initialized when they are registered as clock providers. We do this so that fw_devlink can tell when a clock driver doesn't use the driver core framework to probe/initialize its device. This ensures fw_devlink doesn't block the consumers of such a clock provider indefinitely.

However, some users of CLK_OF_DECLARE() macros don't use the same node that matches the macro as the node for the clock provider, but they initialize the entire node. To cover these cases, also mark the nodes that match the macros as initialized when the init callback function is called.

An example of this is "stericsson,u8500-clks" that's handled using CLK_OF_DECLARE() and looks something like this:

clocks {
        compatible = "stericsson,u8500-clks";

        prcmu_clk: prcmu-clock {
                #clock-cells = <1>;
        };

        prcc_pclk: prcc-periph-clock {
                #clock-cells = <2>;
        };

        prcc_kclk: prcc-kernel-clock {
                #clock-cells = <2>;
        };

        prcc_reset: prcc-reset-controller {
                #reset-cells = <2>;
        };
        ...
        ...
};

This patch makes sure that "clocks" is marked as initialized so that fw_devlink knows that all nodes under it have been initialized.

If the driver creates struct devices for some of the subnodes, fw_devlink is smart enough to know to wait for those devices to probe. So, no special handling is required for those cases.

Quoted 8 lines
Quoted 6 lines

---

Applied to clk-fixes

Thanks!

-Saravana