Skip to contentSkip to conversation

soc: fsl: add Platform PM driver QorIQ platforms

PATCH In review Ran Wang 5 participants 22 messages Aug 31, 2018 – Sep 11, 2018

Subsystems (2)
lore
Share & tools
lore

Ran Wang

author
#0 lore raw

PATCH 1/3 soc: fsl: add Platform PM driver QorIQ platforms

This driver is to provide a independent framework for PM service provider and consumer to configure system level wake up feature. For example, RCPM driver could register a callback function on this platform first, and Flex timer driver who want to enable timer wake up feature, will call generic API provided by this platform driver, and then it will trigger RCPM driver to do it. The benefit is to isolate the user and service, such as flex timer driver will not have to know the implement details of wakeup function it require. Besides, it is also easy for service side to upgrade its logic when design is changed and remain user side unchanged.

---
 drivers/soc/fsl/Kconfig   |   14 +++++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/plat_pm.c |  144 +++++++++++++++++++++++++++++++++++++++++++++
 include/soc/fsl/plat_pm.h |   22 +++++++
 4 files changed, 181 insertions(+), 0 deletions(-)
 create mode 100644 drivers/soc/fsl/plat_pm.c
 create mode 100644 include/soc/fsl/plat_pm.h
drivers/soc/fsl/Kconfig +14 -0
index 7a9fb9b..6517412 100644
@@ -16,3 +16,17 @@ config FSL_GUTS
 	  Initially only reading SVR and registering soc device are supported.
 	  Other guts accesses, such as reading RCW, should eventually be moved
 	  into this driver as well.
+
+config FSL_PLAT_PM
+	bool "Freescale platform PM framework"
+	help
+	  This driver is to provide a independent framework for PM service
+	  provider and consumer to configure system level wake up feature. For
+	  example, RCPM driver could register a callback function on this
+	  platform first, and Flex timer driver who want to enable timer wake
+	  up feature, will call generic API provided by this platform driver,
+	  and then it will trigger RCPM driver to do it. The benefit is to
+	  isolate the user and service, such as  flex timer driver will not
+	  have to know the implement details of wakeup function it require.
+	  Besides, it is also easy for service side to upgrade its logic when
+	  design changed and remain user side unchanged.
drivers/soc/fsl/Makefile +1 -0
index 44b3beb..8f9db23 100644
@@ -6,3 +6,4 @@ obj-$(CONFIG_FSL_DPAA)                 += qbman/
 obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
+obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
drivers/soc/fsl/plat_pm.c new file +144 -0

144 lines not on this page · raw message

include/soc/fsl/plat_pm.h new file +22 -0
new file mode 100644
index 0000000..bbe151e
@@ -0,0 +1,22 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.h - Freescale platform PM Header
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,
+
+#ifndef __FSL_PLAT_PM_H
+#define __FSL_PLAT_PM_H
+
+typedef int (*fsl_plat_pm_handle)(struct device *dev, bool flag,
+		void *handle_priv);
+
+int register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv);
+int deregister_fsl_platform_wakeup_source(void *handle_priv);
+int fsl_platform_wakeup_config(struct device *dev, bool flag);
+int fsl_platform_wakeup_enable(struct device *dev);
+int fsl_platform_wakeup_disable(struct device *dev);
+
+#endif	// __FSL_PLAT_PM_H
Signature · 1 line
1.7.1
author
#1 lore raw

PATCH 2/3 Documentation: dt: binding: fsl: update property description for RCPM

Add property 'big-endian' and supportted IP's configuration info. Remove property 'fsl,#rcpm-wakeup-cell'.

---
 Documentation/devicetree/bindings/soc/fsl/rcpm.txt |   42 ++++++-------------
 1 files changed, 13 insertions(+), 29 deletions(-)
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +13 -29
index e284e4e..7fc630a 100644
@@ -5,8 +5,6 @@ and power management.
 
 Required properites:
   - reg : Offset and length of the register set of the RCPM block.
-  - fsl,#rcpm-wakeup-cells : The number of IPPDEXPCR register cells in the
-	fsl,rcpm-wakeup property.
   - compatible : Must contain a chip-specific RCPM block compatible string
 	and (if applicable) may contain a chassis-version RCPM compatible
 	string. Chip-specific strings are of the form "fsl,<chip>-rcpm",
@@ -27,37 +25,23 @@ Chassis Version		Example Chips
 ---------------		-------------------------------
 1.0				p4080, p5020, p5040, p2041, p3041
 2.0				t4240, b4860, b4420
-2.1				t1040, ls1021
+2.1				t1040, ls1021, ls1012
+
+Optional properties:
+ - big-endian : Indicate RCPM registers is big-endian. A RCPM node
+   that doesn't have this property will be regarded as little-endian.
+ - <property 'compatible' string of consumer device> : This string
+   is referred by RCPM driver to judge if the consumer (such as flex timer)
+   is able to be regards as wakeup source or not, such as 'fsl,ls1012a-ftm'.
+   Further, this property will carry the bit mask info to control
+   coresponding wake up source.
 
 Example:
 The RCPM node for T4240:
 	rcpm: global-utilities at e2000 {
 		compatible = "fsl,t4240-rcpm", "fsl,qoriq-rcpm-2.0";
 		reg = <0xe2000 0x1000>;
-		fsl,#rcpm-wakeup-cells = <2>;
-	};
-
-* Freescale RCPM Wakeup Source Device Tree Bindings
--------------------------------------------
-Required fsl,rcpm-wakeup property should be added to a device node if the device
-can be used as a wakeup source.
-
-  - fsl,rcpm-wakeup: Consists of a phandle to the rcpm node and the IPPDEXPCR
-	register cells. The number of IPPDEXPCR register cells is defined in
-	"fsl,#rcpm-wakeup-cells" in the rcpm node. The first register cell is
-	the bit mask that should be set in IPPDEXPCR0, and the second register
-	cell is for IPPDEXPCR1, and so on.
-
-	Note: IPPDEXPCR(IP Powerdown Exception Control Register) provides a
-	mechanism for keeping certain blocks awake during STANDBY and MEM, in
-	order to use them as wake-up sources.
-
-Example:
-	lpuart0: serial at 2950000 {
-		compatible = "fsl,ls1021a-lpuart";
-		reg = <0x0 0x2950000 0x0 0x1000>;
-		interrupts = <GIC_SPI 80 IRQ_TYPE_LEVEL_HIGH>;
-		clocks = <&sysclk>;
-		clock-names = "ipg";
-		fsl,rcpm-wakeup = <&rcpm 0x0 0x40000000>;
+		big-endian;
+		fsl,ls1012a-ftm = <0x20000>;
+		fsl,pfe = <0xf0000020>;
 	};
Signature · 1 line
1.7.1
maintainer
#2 lore raw

PATCH 2/3 Documentation: dt: binding: fsl: update property description for RCPM

Quoted 2 lines · Ran Wang

Add property 'big-endian' and supportted IP's configuration info. Remove property 'fsl,#rcpm-wakeup-cell'.

"dt-bindings: soc: ..." for the subject

It is obvious reading the diff you are removing fsl,#rcpm-wakeup-cell. What is not obvious is why? The commit msg should answer that.

You also are mixing several things in this patch like adding ls1012 which you don't mention. Please split.

Quoted 5 lines
---
 Documentation/devicetree/bindings/soc/fsl/rcpm.txt |   42 ++++++-------------
 1 files changed, 13 insertions(+), 29 deletions(-)
author
#3 lore raw

RE: [PATCH 2/3] Documentation: dt: binding: fsl: update property description for RCPM

Hi Rob

Quoted 10 lines

-----Original Message-----
From: Rob Herring <[email protected]>
Sent: Tuesday, September 04, 2018 09:25
To: Ran Wang <[email protected]>
Cc: Leo Li <[email protected]>; Mark Rutland <[email protected]>; linuxppc-dev at lists.ozlabs.org; linux-arm-kernel at lists.infradead.org; devicetree at vger.kernel.org; linux-kernel at vger.kernel.org
Subject: Re: [PATCH 2/3] Documentation: dt: binding: fsl: update property description for RCPM

Quoted 8 lines
Quoted 2 lines · Ran Wang

Add property 'big-endian' and supportted IP's configuration info. Remove property 'fsl,#rcpm-wakeup-cell'.

"dt-bindings: soc: ..." for the subject

It is obvious reading the diff you are removing fsl,#rcpm-wakeup-cell. What is not obvious is why? The commit msg should answer that.

Sure, I will add this in next version patch.

Quoted 2 lines

You also are mixing several things in this patch like adding ls1012 which you don't mention. Please split.

Got it, will split them and add more information in next version.

Ran

Quoted 6 lines
Quoted 4 lines
---
 Documentation/devicetree/bindings/soc/fsl/rcpm.txt |   42 ++++++-----------

--

Quoted 1 line
 1 files changed, 13 insertions(+), 29 deletions(-)
#4 lore raw

Re: [PATCH 2/3] Documentation: dt: binding: fsl: update property description for RCPM

In reply to Ran Wang

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 32

Quoted 3 lines · Ran Wang
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +3 -0
+Optional properties:
+ - big-endian : Indicate RCPM registers is big-endian. A RCPM node
+   that doesn't have this property will be regarded as little-endian.

You've just broken all the existing powerpc device trees that are big-endian and have no big-endian property.

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 37

Quoted 6 lines

+ - <property 'compatible' string of consumer device> : This string

(quoted diff) +2 -0
+   is referred by RCPM driver to judge if the consumer (such as flex timer)
+   is able to be regards as wakeup source or not, such as 'fsl,ls1012a-

ftm'.

Documentation/devicetree/bindings/soc/fsl/rcpm.txt +2 -0
+   Further, this property will carry the bit mask info to control
+   coresponding wake up source.

What will you do if there are multiple instances of a device with the same compatible, and different wakeup bits? Plus, it's an awkward design in general, and you don't describe what the value actually means (bits in which register?). What was wrong with the existing binding? Alternatively, use the clock bindings.

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 46

Quoted 12 lines
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +3 -9
-
-Example:
-	lpuart0: serial at 2950000 {
-		compatible = "fsl,ls1021a-lpuart";
-		reg = <0x0 0x2950000 0x0 0x1000>;
-		interrupts = <GIC_SPI 80 IRQ_TYPE_LEVEL_HIGH>;
-		clocks = <&sysclk>;
-		clock-names = "ipg";
-		fsl,rcpm-wakeup = <&rcpm 0x0 0x40000000>;
+		big-endian;
+		fsl,ls1012a-ftm = <0x20000>;
+		fsl,pfe = <0xf0000020>;

fsl,pfe is not documented.

-Scott

author
#5 lore raw

RE: [PATCH 2/3] Documentation: dt: binding: fsl: update property description for RCPM

Hi Scott,

Quoted 8 lines · Scott Wood

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 32

Quoted 3 lines · Ran Wang
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +3 -0
+Optional properties:
+ - big-endian : Indicate RCPM registers is big-endian. A RCPM node
+   that doesn't have this property will be regarded as little-endian.

You've just broken all the existing powerpc device trees that are big-endian and have no big-endian property.

Yes, powerpc RCPM driver (arch/powerpc/sysdev/fsl_rcpm.c) will not refer to big-endian. However, I think if Layerscape RCPM driver use different compatible id (such as ' fsl,qoriq-rcpm-2.2'), it might be safe. Is that OK?

Quoted 10 lines

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 37

Quoted 7 lines

+ - <property 'compatible' string of consumer device> : This string

(quoted diff) +3 -0
+   is referred by RCPM driver to judge if the consumer (such as flex timer)
+   is able to be regards as wakeup source or not, such as
+ 'fsl,ls1012a-

ftm'.

Documentation/devicetree/bindings/soc/fsl/rcpm.txt +2 -0
+   Further, this property will carry the bit mask info to control
+   coresponding wake up source.

What will you do if there are multiple instances of a device with the same compatible, and different wakeup bits?

You got me! This is a problem in current version. Well, we have to decouple wake up source IP and RCPM driver. That's why I add a plat_pm driver to prevent wake up IP knowing any information of RCPM. So in current context, one idea come to me is to redesign property ' fsl,ls1012a-ftm = <0x20000>;', change to 'fsl,ls1012a-ftm = <&ftm0 0x20000>;' What do you say? And could you tell me which API I can use to check if that device's name is ftm0 (for example we want to find a node ftm0: ftm at 29d000)?

Quoted 3 lines

Plus, it's an awkward design in
general, and you don't describe what the value actually means (bits in which register?).

Yes, I admit my design looks ugly and not flexible and extensible enough. However, for above reason, do you have any good idea, please? :)

As to the register information, I can explain here in details (will add to commit message of next version): There is a RCPM HW block which has register named IPPDEXPCR. It controls whether prevent certain IP (such as timer, usb, etc) from entering low power mode when system suspended. So it's necessary to program it if we want one of those IP work as a wakeup source. However, different Layerscape SoCs have different bit vs. IP mapping layout. So I have to record this information in device tree and fetched by RCPM driver directly.

Do I need to list all SoC's mapping information in this binding doc for reference?

Quoted 1 line

What was wrong with the existing binding?

There was one version of RCPM patch which requiring property 'fsl,#rcpm-wakeup-cells' but was not accepted by upstream finally. Now we will no need it any longer due to new design allow case of multiple RCPM devices existing case.

Quoted 1 line

Alternatively, use the clock bindings.

Sorry I didn't get your point.

Quoted 6 lines
Quoted 6 lines
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +0 -6
-
-Example:
-	lpuart0: serial at 2950000 {
-		compatible = "fsl,ls1021a-lpuart";
-		reg = <0x0 0x2950000 0x0 0x1000>;
-		interrupts = <GIC_SPI 80 IRQ_TYPE_LEVEL_HIGH>;
Quoted 8 lines

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 46

Quoted 6 lines
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +3 -3
-		clocks = <&sysclk>;
-		clock-names = "ipg";
-		fsl,rcpm-wakeup = <&rcpm 0x0 0x40000000>;
+		big-endian;
+		fsl,ls1012a-ftm = <0x20000>;
+		fsl,pfe = <0xf0000020>;

fsl,pfe is not documented.

pfe patch is not upstream yet, will remove it.

Regards,
Ran

Quoted 1 line

-Scott

#6 lore raw

Re: [PATCH 2/3] Documentation: dt: binding: fsl: update property description for RCPM

Quoted 10 lines · Ran Wang

Hi Scott,

Quoted 6 lines · Scott Wood
Quoted 3 lines · Ran Wang
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +3 -0
+Optional properties:
+ - big-endian : Indicate RCPM registers is big-endian. A RCPM node
+   that doesn't have this property will be regarded as little-endian.
Quoted 6 lines
Quoted 2 lines

You've just broken all the existing powerpc device trees that are big-endian and have no big-endian property.

Yes, powerpc RCPM driver (arch/powerpc/sysdev/fsl_rcpm.c) will not refer to big-endian. However, I think if Layerscape RCPM driver use different compatible id (such as ' fsl,qoriq-rcpm-2.2'), it might be safe. Is that OK?

For Layerscape Chassis(gen3) based chips, the Reference Manual named the power management IP as COP_PMU instead of RCPM which makes sense to me as the register map is also quite different from the reg map of RCPM. So I think it is reasonable to create a new binding and driver for the Chassis Gen3 based PMU. However, the arch/powerpc/sysdev/fsl_rcpm.c driver probably should be moved to drivers/soc/fsl, as the Gen2.x Chassis and RCPM IP are also used in some non-PowerPC Layerscape SoCs. From what I know, all the RCPM based IP are all big endian, so there is no need to add this property for the old binding.

Quoted 19 lines
Quoted 11 lines

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 37

Quoted 8 lines

+ - <property 'compatible' string of consumer device> : This string

(quoted diff) +3 -0
+   is referred by RCPM driver to judge if the consumer (such as flex timer)
+   is able to be regards as wakeup source or not, such as
+ 'fsl,ls1012a-

ftm'.

Documentation/devicetree/bindings/soc/fsl/rcpm.txt +2 -0
+   Further, this property will carry the bit mask info to control
+   coresponding wake up source.

What will you do if there are multiple instances of a device with the same compatible, and different wakeup bits?

You got me! This is a problem in current version. Well, we have to decouple wake up source IP and RCPM driver. That's why I add a plat_pm driver to prevent wake up IP knowing any information of RCPM. So in current context, one idea come to me is to redesign property ' fsl,ls1012a-ftm = <0x20000>;', change to 'fsl,ls1012a-ftm = <&ftm0 0x20000>;' What do you say? And could you tell me which API I can use to check if that device's name is ftm0 (for example we want to find a node ftm0: ftm at 29d000)?

Quoted 6 lines
Quoted 3 lines

Plus, it's an awkward design in
general, and you don't describe what the value actually means (bits in which register?).

Yes, I admit my design looks ugly and not flexible and extensible enough. However, for above reason, do you have any good idea, please? :)

Why do you have to move the wakeup information from device nodes to the RCPM/PMU node? For information related to both on-chip device and SoC integration, the information normally goes into the node of on-chip device instead of the integration IP's device node. Existing example like: interrupt, clock, memory map, and etc. It is much cleaner than listing all the interrupts in the interrupt controller's node, right?

Quoted 43 lines

As to the register information, I can explain here in details (will add to commit message of next version): There is a RCPM HW block which has register named IPPDEXPCR. It controls whether prevent certain IP (such as timer, usb, etc) from entering low power mode when system suspended. So it's necessary to program it if we want one of those IP work as a wakeup source. However, different Layerscape SoCs have different bit vs. IP mapping layout. So I have to record this information in device tree and fetched by RCPM driver directly.

Do I need to list all SoC's mapping information in this binding doc for reference?

Quoted 1 line

What was wrong with the existing binding?

There was one version of RCPM patch which requiring property 'fsl,#rcpm-wakeup-cells' but was not accepted by upstream finally. Now we will no need it any longer due to new design allow case of multiple RCPM devices existing case.

Quoted 1 line

Alternatively, use the clock bindings.

Sorry I didn't get your point.

Quoted 6 lines
Quoted 6 lines
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +0 -6
-
-Example:
-   lpuart0: serial at 2950000 {
-           compatible = "fsl,ls1021a-lpuart";
-           reg = <0x0 0x2950000 0x0 0x1000>;
-           interrupts = <GIC_SPI 80 IRQ_TYPE_LEVEL_HIGH>;
Quoted 8 lines

Documentation/devicetree/bindings/soc/fsl/rcpm.txt Chassis Version Example Chips · line 46

Quoted 6 lines
Documentation/devicetree/bindings/soc/fsl/rcpm.txt +3 -3
-           clocks = <&sysclk>;
-           clock-names = "ipg";
-           fsl,rcpm-wakeup = <&rcpm 0x0 0x40000000>;
+           big-endian;
+           fsl,ls1012a-ftm = <0x20000>;
+           fsl,pfe = <0xf0000020>;

fsl,pfe is not documented.

pfe patch is not upstream yet, will remove it.

Regards,
Ran

Quoted 1 line

-Scott

author
#7 lore raw

PATCH 3/3 soc: fsl: add RCPM driver

In reply to Ran Wang

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)
 create mode 100644 drivers/soc/fsl/ls-rcpm.c
drivers/soc/fsl/Kconfig +6 -0
index 6517412..882330d 100644
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
 	  have to know the implement details of wakeup function it require.
 	  Besides, it is also easy for service side to upgrade its logic when
 	  design changed and remain user side unchanged.
+
+config LS_RCPM
+	bool "Freescale RCPM support"
+	depends on (FSL_PLAT_PM)
+	help
+	  This feature is to enable specified wakeup source for system sleep.
drivers/soc/fsl/Makefile +1 -0
index 8f9db23..43ff71a 100644
@@ -7,3 +7,4 @@ obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
 obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
+obj-$(CONFIG_LS_RCPM)		+= ls-rcpm.o
drivers/soc/fsl/ls-rcpm.c new file +153 -0

153 lines not on this page · raw message

Signature · 1 line
1.7.1
#8 lore raw

Re: [PATCH 3/3] soc: fsl: add RCPM driver

Please change your comments style.

Quoted 89 lines · Ran Wang

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)
 create mode 100644 drivers/soc/fsl/ls-rcpm.c
drivers/soc/fsl/Kconfig +6 -0
index 6517412..882330d 100644
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
 	  have to know the implement details of wakeup function it require.
 	  Besides, it is also easy for service side to upgrade its logic when
 	  design changed and remain user side unchanged.
+
+config LS_RCPM
+	bool "Freescale RCPM support"
+	depends on (FSL_PLAT_PM)
+	help
+	  This feature is to enable specified wakeup source for system sleep.
drivers/soc/fsl/Makefile +1 -0
index 8f9db23..43ff71a 100644
@@ -7,3 +7,4 @@ obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
 obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
+obj-$(CONFIG_LS_RCPM)		+= ls-rcpm.o
drivers/soc/fsl/ls-rcpm.c new file +43 -0
new file mode 100644
index 0000000..b0feb88
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.c - Freescale Layerscape RCPM driver
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_address.h>
+#include <linux/slab.h>
+#include <soc/fsl/plat_pm.h>
+
+#define MAX_COMPATIBLE_NUM	10
+
+struct rcpm_t {
+	struct device *dev;
+	void __iomem *ippdexpcr_addr;
+	bool big_endian;	/* Big/Little endian of RCPM module */
+};
+
+// rcpm_handle - Configure RCPM reg according to wake up source request
+// @user_dev: pointer to user's device struct
+// @flag: to enable(true) or disable(false) wakeup source
+// @handle_priv: pointer to struct rcpm_t instance
+//
+// Return 0 on success other negative errno
+static int rcpm_handle(struct device *user_dev, bool flag, void *handle_priv)
+{
+	struct rcpm_t *rcpm;
+	bool big_endian;
+	const char  *dev_compatible_array[MAX_COMPATIBLE_NUM];
+	void __iomem *ippdexpcr_addr;
+	u32 ippdexpcr;
+	u32 set_bit;
+	int ret, num, i;
+
+	rcpm = handle_priv;
+	big_endian = rcpm->big_endian;
+	ippdexpcr_addr = rcpm->ippdexpcr_addr;
+

drivers/soc/fsl/ls-rcpm.c line 53

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+	num = device_property_read_string_array(user_dev, "compatible",
+			dev_compatible_array, MAX_COMPATIBLE_NUM);
+	if (num < 0)
+		return num;
+
+	for (i = 0; i < num; i++) {
+		if (!device_property_present(rcpm->dev,
+					dev_compatible_array[i]))
+			continue;
+		else {

Remove this else.

drivers/soc/fsl/ls-rcpm.c line 60

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +7 -0
+			ret = device_property_read_u32(rcpm->dev,
+					dev_compatible_array[i], &set_bit);
+			if (ret)
+				return ret;
+
+			if (!device_property_present(rcpm->dev,
+						dev_compatible_array[i]))

This has been checked. Continue ? or return ENODEV?

drivers/soc/fsl/ls-rcpm.c line 62

Quoted 2 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
+				return -ENODEV;
+			else {

Remove this else.

Quoted 9 lines
drivers/soc/fsl/ls-rcpm.c new file +9 -0
+				ret = device_property_read_u32(rcpm->dev,
+						dev_compatible_array[i], &set_bit);
+				if (ret)
+					return ret;
+
+				if (big_endian)
+					ippdexpcr = ioread32be(ippdexpcr_addr);
+				else
+					ippdexpcr = ioread32(ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c line 81

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+
+				if (flag)
+					ippdexpcr |= set_bit;
+				else
+					ippdexpcr &= ~set_bit;
+
+				if (big_endian) {
+					iowrite32be(ippdexpcr, ippdexpcr_addr);
+					ippdexpcr = ioread32be(ippdexpcr_addr);
+				} else

if (x) {
....
....
} else {

}

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+					iowrite32(ippdexpcr, ippdexpcr_addr);
+
+				return 0;
+			}
+		}
+	}
+
+	return -ENODEV;
+}
+

drivers/soc/fsl/ls-rcpm.c line 101

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+static int ls_rcpm_probe(struct platform_device *pdev)
+{
+	struct resource *r;
+	struct rcpm_t *rcpm;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!r)
+		return -ENODEV;
+
+	rcpm = kmalloc(sizeof(*rcpm), GFP_KERNEL);

kzalloc is better.

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+	if (!rcpm)
+		return -ENOMEM;
+
+	rcpm->big_endian = device_property_read_bool(&pdev->dev, "big-endian");
+
+	rcpm->ippdexpcr_addr = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(rcpm->ippdexpcr_addr))
+		return PTR_ERR(rcpm->ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c line 119

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+
+	rcpm->dev = &pdev->dev;
+	platform_set_drvdata(pdev, rcpm);
+
+	return register_fsl_platform_wakeup_source(rcpm_handle, rcpm);
+}
+
+static int ls_rcpm_remove(struct platform_device *pdev)
+{
+	struct rcpm_t	*rcpm;

Not need a table.

Cheers,
-Dongsheng

Quoted 34 lines
drivers/soc/fsl/ls-rcpm.c new file +34 -0
+
+	rcpm = platform_get_drvdata(pdev);
+	deregister_fsl_platform_wakeup_source(rcpm);
+	kfree(rcpm);
+
+	return 0;
+}
+
+static const struct of_device_id ls_rcpm_of_match[] = {
+	{ .compatible = "fsl,qoriq-rcpm-2.1", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, ls_rcpm_of_match);
+
+static struct platform_driver ls_rcpm_driver = {
+	.driver = {
+		.name = "ls-rcpm",
+		.of_match_table = ls_rcpm_of_match,
+	},
+	.probe = ls_rcpm_probe,
+	.remove = ls_rcpm_remove,
+};
+
+static int __init ls_rcpm_init(void)
+{
+	return platform_driver_register(&ls_rcpm_driver);
+}
+subsys_initcall(ls_rcpm_init);
+
+static void __exit ls_rcpm_exit(void)
+{
+	platform_driver_unregister(&ls_rcpm_driver);
+}
+module_exit(ls_rcpm_exit);
#9 lore raw

Re: [PATCH 3/3] soc: fsl: add RCPM driver

Quoted 2 lines · Dongsheng

Please change your comments style.

Although this doesn't get into the Linux kernel coding style documentation yet, Linus seems changed his mind to prefer // than /*
*/ comment style now. https://lkml.org/lkml/2017/11/25/133 So the
// style should be acceptable for now.

Quoted 34 lines · Ran Wang
Quoted 32 lines

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153 +++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)
 create mode 100644 drivers/soc/fsl/ls-rcpm.c
drivers/soc/fsl/Kconfig +6 -0
index 6517412..882330d 100644
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
        have to know the implement details of wakeup function it require.
        Besides, it is also easy for service side to upgrade its logic when
        design changed and remain user side unchanged.
+
+config LS_RCPM
+     bool "Freescale RCPM support"
+     depends on (FSL_PLAT_PM)
+     help
+       This feature is to enable specified wakeup source for system sleep.
drivers/soc/fsl/Makefile +0 -0
Quoted 8 lines

drivers/soc/fsl/Makefile obj-$(CONFIG_QUICC_ENGINE) += qe/ · line 10

Quoted 8 lines

index 8f9db23..43ff71a 100644

drivers/soc/fsl/Makefile +1 -0
@@ -7,3 +7,4 @@ obj-$(CONFIG_QUICC_ENGINE)            += qe/
 obj-$(CONFIG_CPM)                    += qe/
 obj-$(CONFIG_FSL_GUTS)                       += guts.o
 obj-$(CONFIG_FSL_PLAT_PM)    += plat_pm.o
+obj-$(CONFIG_LS_RCPM)                += ls-rcpm.o

Probably use "_" instead of "-" for alignment.

Quoted 9 lines

drivers/soc/fsl/ls-rcpm.c line 3

Quoted 9 lines
drivers/soc/fsl/ls-rcpm.c new file +3 -0
new file mode 100644
index 0000000..b0feb88
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.c - Freescale Layerscape RCPM driver

The file name here is not the same as the real file name.

Quoted 4 lines

drivers/soc/fsl/ls-rcpm.c line 7

Quoted 4 lines
drivers/soc/fsl/ls-rcpm.c new file +4 -0
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,

Where do you need the comma in the end?

Quoted 14 lines
Quoted 14 lines
drivers/soc/fsl/ls-rcpm.c new file +14 -0
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_address.h>
+#include <linux/slab.h>
+#include <soc/fsl/plat_pm.h>
+
+#define MAX_COMPATIBLE_NUM   10
+
+struct rcpm_t {
+     struct device *dev;
+     void __iomem *ippdexpcr_addr;
+     bool big_endian;        /* Big/Little endian of RCPM module */
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 29

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+};
+
+// rcpm_handle - Configure RCPM reg according to wake up source request
+// @user_dev: pointer to user's device struct
+// @flag: to enable(true) or disable(false) wakeup source
+// @handle_priv: pointer to struct rcpm_t instance
+//
+// Return 0 on success other negative errno

Although Linus preferred this // comment style. I'm not sure if this will be handled correctly by the kernel-doc compiler.
https://www.kernel.org/doc/html/v4.18/doc-guide/kernel-doc.html

Quoted 141 lines
Quoted 14 lines
drivers/soc/fsl/ls-rcpm.c new file +14 -0
+static int rcpm_handle(struct device *user_dev, bool flag, void *handle_priv)
+{
+     struct rcpm_t *rcpm;
+     bool big_endian;
+     const char  *dev_compatible_array[MAX_COMPATIBLE_NUM];
+     void __iomem *ippdexpcr_addr;
+     u32 ippdexpcr;
+     u32 set_bit;
+     int ret, num, i;
+
+     rcpm = handle_priv;
+     big_endian = rcpm->big_endian;
+     ippdexpcr_addr = rcpm->ippdexpcr_addr;
+

drivers/soc/fsl/ls-rcpm.c line 53

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+     num = device_property_read_string_array(user_dev, "compatible",
+                     dev_compatible_array, MAX_COMPATIBLE_NUM);
+     if (num < 0)
+             return num;
+
+     for (i = 0; i < num; i++) {
+             if (!device_property_present(rcpm->dev,
+                                     dev_compatible_array[i]))
+                     continue;
+             else {

Remove this else.

drivers/soc/fsl/ls-rcpm.c line 60

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +7 -0
+                     ret = device_property_read_u32(rcpm->dev,
+                                     dev_compatible_array[i], &set_bit);
+                     if (ret)
+                             return ret;
+
+                     if (!device_property_present(rcpm->dev,
+                                             dev_compatible_array[i]))

This has been checked. Continue ? or return ENODEV?

drivers/soc/fsl/ls-rcpm.c line 62

Quoted 2 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
+                             return -ENODEV;
+                     else {

Remove this else.

Quoted 9 lines
drivers/soc/fsl/ls-rcpm.c new file +9 -0
+                             ret = device_property_read_u32(rcpm->dev,
+                                             dev_compatible_array[i], &set_bit);
+                             if (ret)
+                                     return ret;
+
+                             if (big_endian)
+                                     ippdexpcr = ioread32be(ippdexpcr_addr);
+                             else
+                                     ippdexpcr = ioread32(ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c line 81

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+
+                             if (flag)
+                                     ippdexpcr |= set_bit;
+                             else
+                                     ippdexpcr &= ~set_bit;
+
+                             if (big_endian) {
+                                     iowrite32be(ippdexpcr, ippdexpcr_addr);
+                                     ippdexpcr = ioread32be(ippdexpcr_addr);
+                             } else

if (x) {
....
....
} else {

}

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+                                     iowrite32(ippdexpcr, ippdexpcr_addr);
+
+                             return 0;
+                     }
+             }
+     }
+
+     return -ENODEV;
+}
+

drivers/soc/fsl/ls-rcpm.c line 101

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+static int ls_rcpm_probe(struct platform_device *pdev)
+{
+     struct resource *r;
+     struct rcpm_t *rcpm;
+
+     r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+     if (!r)
+             return -ENODEV;
+
+     rcpm = kmalloc(sizeof(*rcpm), GFP_KERNEL);

kzalloc is better.

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+     if (!rcpm)
+             return -ENOMEM;
+
+     rcpm->big_endian = device_property_read_bool(&pdev->dev, "big-endian");
+
+     rcpm->ippdexpcr_addr = devm_ioremap_resource(&pdev->dev, r);
+     if (IS_ERR(rcpm->ippdexpcr_addr))
+             return PTR_ERR(rcpm->ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c line 119

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+
+     rcpm->dev = &pdev->dev;
+     platform_set_drvdata(pdev, rcpm);
+
+     return register_fsl_platform_wakeup_source(rcpm_handle, rcpm);
+}
+
+static int ls_rcpm_remove(struct platform_device *pdev)
+{
+     struct rcpm_t   *rcpm;

Not need a table.

Cheers,
-Dongsheng

Quoted 34 lines
drivers/soc/fsl/ls-rcpm.c new file +34 -0
+
+     rcpm = platform_get_drvdata(pdev);
+     deregister_fsl_platform_wakeup_source(rcpm);
+     kfree(rcpm);
+
+     return 0;
+}
+
+static const struct of_device_id ls_rcpm_of_match[] = {
+     { .compatible = "fsl,qoriq-rcpm-2.1", },
+     {}
+};
+MODULE_DEVICE_TABLE(of, ls_rcpm_of_match);
+
+static struct platform_driver ls_rcpm_driver = {
+     .driver = {
+             .name = "ls-rcpm",
+             .of_match_table = ls_rcpm_of_match,
+     },
+     .probe = ls_rcpm_probe,
+     .remove = ls_rcpm_remove,
+};
+
+static int __init ls_rcpm_init(void)
+{
+     return platform_driver_register(&ls_rcpm_driver);
+}
+subsys_initcall(ls_rcpm_init);
+
+static void __exit ls_rcpm_exit(void)
+{
+     platform_driver_unregister(&ls_rcpm_driver);
+}
+module_exit(ls_rcpm_exit);
author
#10 lore raw

RE: [PATCH 3/3] soc: fsl: add RCPM driver

Hi Leo,

Quoted 65 lines · On September 05, 2018 at 11:22 Yang Li wrote:

-----Original Message-----
From: Li Yang <[email protected]>
Sent: Wednesday, September 05, 2018 11:22
To: dongsheng.wang at hxt-semitech.com
Cc: Ran Wang <[email protected]>; Rob Herring <[email protected]>; Mark Rutland <[email protected]>; open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS <[email protected]>; linuxppc- dev <[email protected]>; lkml <[email protected]>; moderated list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE <linux-arm- kernel at lists.infradead.org>
Subject: Re: [PATCH 3/3] soc: fsl: add RCPM driver

Quoted 2 lines · Dongsheng

Please change your comments style.

Although this doesn't get into the Linux kernel coding style documentation yet, Linus seems changed his mind to prefer // than /*
*/ comment style now.
https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flkml
.org%2Flkml%2F2017%2F11%2F25%2F133&amp;data=02%7C01%7Cran.wang_ 1%40nxp.com%7Cc0d88e6690384e02b95108d612dec235%7C686ea1d3bc2b4c 6fa92cd99c5c301635%7C0%7C0%7C636717145285126200&amp;sdata=JIoCZp WhRyW76EqgSflfTDA1f0gMQGKa%2FcbvSc5CO%2Fw%3D&amp;reserved=0
So the
// style should be acceptable for now.

Quoted 37 lines · Ran Wang
Quoted 35 lines

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153
+++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/ls-rcpm.c

drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

6517412..882330d 100644

drivers/soc/fsl/Kconfig +6 -0
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
        have to know the implement details of wakeup function it require.
        Besides, it is also easy for service side to upgrade its logic when
        design changed and remain user side unchanged.
+
+config LS_RCPM
+     bool "Freescale RCPM support"
+     depends on (FSL_PLAT_PM)
+     help
+       This feature is to enable specified wakeup source for system sleep.
drivers/soc/fsl/Makefile +0 -1
index 8f9db23..43ff71a 100644
--- a/drivers/soc/fsl/Makefile
Quoted 8 lines
Quoted 6 lines

drivers/soc/fsl/Makefile obj-$(CONFIG_QUICC_ENGINE) += qe/ · line 10

Quoted 6 lines

+++ b/drivers/soc/fsl/Makefile

drivers/soc/fsl/Makefile +1 -0
@@ -7,3 +7,4 @@ obj-$(CONFIG_QUICC_ENGINE)            += qe/
 obj-$(CONFIG_CPM)                    += qe/
 obj-$(CONFIG_FSL_GUTS)                       += guts.o
 obj-$(CONFIG_FSL_PLAT_PM)    += plat_pm.o
+obj-$(CONFIG_LS_RCPM)                += ls-rcpm.o

Probably use "_" instead of "-" for alignment.

OK, will update in next version

Quoted 9 lines
Quoted 7 lines

drivers/soc/fsl/ls-rcpm.c line 2

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
new file mode 100644 index 0000000..b0feb88
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0 // // plat_pm.c - Freescale
+Layerscape RCPM driver

The file name here is not the same as the real file name.

Got it, will correct it.

Quoted 6 lines
Quoted 4 lines

drivers/soc/fsl/ls-rcpm.c line 7

Quoted 4 lines
drivers/soc/fsl/ls-rcpm.c new file +4 -0
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,

Where do you need the comma in the end?

My bad, will remove comma in next version.

Quoted 22 lines
Quoted 13 lines
Quoted 13 lines
drivers/soc/fsl/ls-rcpm.c new file +13 -0
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_address.h>
+#include <linux/slab.h>
+#include <soc/fsl/plat_pm.h>
+
+#define MAX_COMPATIBLE_NUM   10
+
+struct rcpm_t {
+     struct device *dev;
+     void __iomem *ippdexpcr_addr;
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 27

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+     bool big_endian;        /* Big/Little endian of RCPM module */
+};
+
+// rcpm_handle - Configure RCPM reg according to wake up source
+request // @user_dev: pointer to user's device struct // @flag: to
+enable(true) or disable(false) wakeup source // @handle_priv:
+pointer to struct rcpm_t instance // // Return 0 on success other
+negative errno
Quoted 8 lines

Although Linus preferred this // comment style. I'm not sure if this will be handled correctly by the kernel-doc compiler.
https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fww
w.kernel.org%2Fdoc%2Fhtml%2Fv4.18%2Fdoc-guide%2Fkernel- doc.html&amp;data=02%7C01%7Cran.wang_1%40nxp.com%7Cc0d88e669038 4e02b95108d612dec235%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0 %7C636717145285126200&amp;sdata=H7GkUNOLVG%2FCcZESzhtHBeHCbO9 %2FK4k9EdH30Cxq2%2BM%3D&amp;reserved=0

So, do you think I need to change all comment style back to '/* ... */' ? Actually I feel a little bit confused here.

Regards,
Ran

Quoted 141 lines
Quoted 141 lines
Quoted 14 lines
drivers/soc/fsl/ls-rcpm.c new file +14 -0
+static int rcpm_handle(struct device *user_dev, bool flag, void
+*handle_priv) {
+     struct rcpm_t *rcpm;
+     bool big_endian;
+     const char  *dev_compatible_array[MAX_COMPATIBLE_NUM];
+     void __iomem *ippdexpcr_addr;
+     u32 ippdexpcr;
+     u32 set_bit;
+     int ret, num, i;
+
+     rcpm = handle_priv;
+     big_endian = rcpm->big_endian;
+     ippdexpcr_addr = rcpm->ippdexpcr_addr;
+

drivers/soc/fsl/ls-rcpm.c line 53

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+     num = device_property_read_string_array(user_dev, "compatible",
+                     dev_compatible_array, MAX_COMPATIBLE_NUM);
+     if (num < 0)
+             return num;
+
+     for (i = 0; i < num; i++) {
+             if (!device_property_present(rcpm->dev,
+                                     dev_compatible_array[i]))
+                     continue;
+             else {

Remove this else.

drivers/soc/fsl/ls-rcpm.c

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+                     ret = device_property_read_u32(rcpm->dev,
+                                     dev_compatible_array[i], &set_bit);
+                     if (ret)
+                             return ret;
+
+                     if (!device_property_present(rcpm->dev,
+
+ dev_compatible_array[i]))

This has been checked. Continue ? or return ENODEV?

drivers/soc/fsl/ls-rcpm.c line 62

Quoted 2 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
+                             return -ENODEV;
+                     else {

Remove this else.

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+                             ret = device_property_read_u32(rcpm->dev,
+                                             dev_compatible_array[i], &set_bit);
+                             if (ret)
+                                     return ret;
+
+                             if (big_endian)
+                                     ippdexpcr = ioread32be(ippdexpcr_addr);
+                             else
+                                     ippdexpcr =
+ ioread32(ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c line 81

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+
+                             if (flag)
+                                     ippdexpcr |= set_bit;
+                             else
+                                     ippdexpcr &= ~set_bit;
+
+                             if (big_endian) {
+                                     iowrite32be(ippdexpcr, ippdexpcr_addr);
+                                     ippdexpcr = ioread32be(ippdexpcr_addr);
+                             } else

if (x) {
....
....
} else {

}

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+                                     iowrite32(ippdexpcr,
+ ippdexpcr_addr);
+
+                             return 0;
+                     }
+             }
+     }
+
+     return -ENODEV;
+}

drivers/soc/fsl/ls-rcpm.c line 101

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+
+static int ls_rcpm_probe(struct platform_device *pdev) {
+     struct resource *r;
+     struct rcpm_t *rcpm;
+
+     r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+     if (!r)
+             return -ENODEV;
+
+     rcpm = kmalloc(sizeof(*rcpm), GFP_KERNEL);

kzalloc is better.

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+     if (!rcpm)
+             return -ENOMEM;
+
+     rcpm->big_endian = device_property_read_bool(&pdev->dev,
+ "big-endian");
+
+     rcpm->ippdexpcr_addr = devm_ioremap_resource(&pdev->dev, r);
+     if (IS_ERR(rcpm->ippdexpcr_addr))

drivers/soc/fsl/ls-rcpm.c line 119

Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+             return PTR_ERR(rcpm->ippdexpcr_addr);
+
+     rcpm->dev = &pdev->dev;
+     platform_set_drvdata(pdev, rcpm);
+
+     return register_fsl_platform_wakeup_source(rcpm_handle, rcpm);
+}
+
+static int ls_rcpm_remove(struct platform_device *pdev) {
+     struct rcpm_t   *rcpm;

Not need a table.

Cheers,
-Dongsheng

Quoted 32 lines
drivers/soc/fsl/ls-rcpm.c new file +32 -0
+
+     rcpm = platform_get_drvdata(pdev);
+     deregister_fsl_platform_wakeup_source(rcpm);
+     kfree(rcpm);
+
+     return 0;
+}
+
+static const struct of_device_id ls_rcpm_of_match[] = {
+     { .compatible = "fsl,qoriq-rcpm-2.1", },
+     {}
+};
+MODULE_DEVICE_TABLE(of, ls_rcpm_of_match);
+
+static struct platform_driver ls_rcpm_driver = {
+     .driver = {
+             .name = "ls-rcpm",
+             .of_match_table = ls_rcpm_of_match,
+     },
+     .probe = ls_rcpm_probe,
+     .remove = ls_rcpm_remove,
+};
+
+static int __init ls_rcpm_init(void) {
+     return platform_driver_register(&ls_rcpm_driver);
+}
+subsys_initcall(ls_rcpm_init);
+
+static void __exit ls_rcpm_exit(void) {
+     platform_driver_unregister(&ls_rcpm_driver);
+}
+module_exit(ls_rcpm_exit);
#11 lore raw

Re: [PATCH 3/3] soc: fsl: add RCPM driver

Quoted 132 lines · Ran Wang

Hi Leo,

Quoted 65 lines · On September 05, 2018 at 11:22 Yang Li wrote:

-----Original Message-----
From: Li Yang <[email protected]>
Sent: Wednesday, September 05, 2018 11:22
To: dongsheng.wang at hxt-semitech.com
Cc: Ran Wang <[email protected]>; Rob Herring <[email protected]>; Mark Rutland <[email protected]>; open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS <[email protected]>; linuxppc- dev <[email protected]>; lkml <[email protected]>; moderated list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE <linux-arm- kernel at lists.infradead.org>
Subject: Re: [PATCH 3/3] soc: fsl: add RCPM driver

Quoted 2 lines · Dongsheng

Please change your comments style.

Although this doesn't get into the Linux kernel coding style documentation yet, Linus seems changed his mind to prefer // than /*
*/ comment style now.
https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Flkml
.org%2Flkml%2F2017%2F11%2F25%2F133&amp;data=02%7C01%7Cran.wang_ 1%40nxp.com%7Cc0d88e6690384e02b95108d612dec235%7C686ea1d3bc2b4c 6fa92cd99c5c301635%7C0%7C0%7C636717145285126200&amp;sdata=JIoCZp WhRyW76EqgSflfTDA1f0gMQGKa%2FcbvSc5CO%2Fw%3D&amp;reserved=0
So the
// style should be acceptable for now.

Quoted 37 lines · Ran Wang
Quoted 35 lines

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153
+++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/ls-rcpm.c

drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

6517412..882330d 100644

drivers/soc/fsl/Kconfig +6 -0
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
        have to know the implement details of wakeup function it require.
        Besides, it is also easy for service side to upgrade its logic when
        design changed and remain user side unchanged.
+
+config LS_RCPM
+     bool "Freescale RCPM support"
+     depends on (FSL_PLAT_PM)
+     help
+       This feature is to enable specified wakeup source for system sleep.
drivers/soc/fsl/Makefile +0 -1
index 8f9db23..43ff71a 100644
--- a/drivers/soc/fsl/Makefile
Quoted 8 lines
Quoted 6 lines

drivers/soc/fsl/Makefile obj-$(CONFIG_QUICC_ENGINE) += qe/ · line 10

Quoted 6 lines

+++ b/drivers/soc/fsl/Makefile

drivers/soc/fsl/Makefile +1 -0
@@ -7,3 +7,4 @@ obj-$(CONFIG_QUICC_ENGINE)            += qe/
 obj-$(CONFIG_CPM)                    += qe/
 obj-$(CONFIG_FSL_GUTS)                       += guts.o
 obj-$(CONFIG_FSL_PLAT_PM)    += plat_pm.o
+obj-$(CONFIG_LS_RCPM)                += ls-rcpm.o

Probably use "_" instead of "-" for alignment.

OK, will update in next version

Quoted 9 lines
Quoted 7 lines

drivers/soc/fsl/ls-rcpm.c line 2

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
new file mode 100644 index 0000000..b0feb88
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0 // // plat_pm.c - Freescale
+Layerscape RCPM driver

The file name here is not the same as the real file name.

Got it, will correct it.

Quoted 6 lines
Quoted 4 lines

drivers/soc/fsl/ls-rcpm.c line 7

Quoted 4 lines
drivers/soc/fsl/ls-rcpm.c new file +4 -0
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,

Where do you need the comma in the end?

My bad, will remove comma in next version.

Quoted 22 lines
Quoted 13 lines
Quoted 13 lines
drivers/soc/fsl/ls-rcpm.c new file +13 -0
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_address.h>
+#include <linux/slab.h>
+#include <soc/fsl/plat_pm.h>
+
+#define MAX_COMPATIBLE_NUM   10
+
+struct rcpm_t {
+     struct device *dev;
+     void __iomem *ippdexpcr_addr;
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 27

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+     bool big_endian;        /* Big/Little endian of RCPM module */
+};
+
+// rcpm_handle - Configure RCPM reg according to wake up source
+request // @user_dev: pointer to user's device struct // @flag: to
+enable(true) or disable(false) wakeup source // @handle_priv:
+pointer to struct rcpm_t instance // // Return 0 on success other
+negative errno
Quoted 8 lines

Although Linus preferred this // comment style. I'm not sure if this will be handled correctly by the kernel-doc compiler.
https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fww
w.kernel.org%2Fdoc%2Fhtml%2Fv4.18%2Fdoc-guide%2Fkernel- doc.html&amp;data=02%7C01%7Cran.wang_1%40nxp.com%7Cc0d88e669038 4e02b95108d612dec235%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0 %7C636717145285126200&amp;sdata=H7GkUNOLVG%2FCcZESzhtHBeHCbO9 %2FK4k9EdH30Cxq2%2BM%3D&amp;reserved=0

Quoted 2 lines

So, do you think I need to change all comment style back to '/* ... */' ? Actually I feel a little bit confused here.

I think Linus's comment about // comment style applies to normal code comment. But kernel-doc comment is a special kind of code comment that needs to meet certain requirements. People can use the scripts/kernel-doc tool to generate readable API documents from the source code. It looks like you wanted to make the function description aligned with the kernel-doc format, but kernel-doc specifically requires to use the /* */ style(at least for now).

Regards,
Leo

author
#12 lore raw

RE: [PATCH 3/3] soc: fsl: add RCPM driver

Hi Leo,

Quoted 153 lines · Yang Li
Quoted 9 lines · Ran Wang

Hi Leo,

Quoted 5 lines · On September 05, 2018 at 11:22 Yang Li wrote:

-----Original Message-----
From: Li Yang <[email protected]>
Sent: Wednesday, September 05, 2018 11:22
To: dongsheng.wang at hxt-semitech.com
Cc: Ran Wang <[email protected]>; Rob Herring

<[email protected]>;

Quoted 8 lines
Quoted 8 lines

Mark Rutland <[email protected]>; open list:OPEN FIRMWARE AND FLATTENED DEVICE TREE BINDINGS <[email protected]>; linuxppc- dev <[email protected]>; lkml <[email protected]>; moderated list:ARM/FREESCALE IMX / MXC ARM ARCHITECTURE <linux-arm- [email protected]>
Subject: Re: [PATCH 3/3] soc: fsl: add RCPM driver

On Tue, Sep 4, 2018 at 9:58 PM Wang, Dongsheng

<dongsheng.wang@hxt-

Quoted 9 lines
Quoted 9 lines
Quoted 2 lines · semitech.com> wrote:

Please change your comments style.

Although this doesn't get into the Linux kernel coding style documentation yet, Linus seems changed his mind to prefer // than /* */ comment style now.
https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fl

kml .org%2Flkml%2F2017%2F11%2F25%2F133&amp;data=02%7C01%7Cran.w ang_

Quoted 1 line
Quoted 1 line

1%40nxp.com%7Cc0d88e6690384e02b95108d612dec235%7C686ea1d3bc2b4c

Quoted 1 line
Quoted 1 line

6fa92cd99c5c301635%7C0%7C0%7C636717145285126200&amp;sdata=JIoCZp

Quoted 1 line
Quoted 1 line

WhRyW76EqgSflfTDA1f0gMQGKa%2FcbvSc5CO%2Fw%3D&amp;reserved=0

Quoted 8 lines
Quoted 8 lines

So the
// style should be acceptable for now.

Quoted 5 lines · Ran Wang
Quoted 3 lines

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as

wakeup source control.

Quoted 15 lines
Quoted 15 lines
Quoted 15 lines
Quoted 15 lines

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153
+++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/ls-rcpm.c

Quoted 6 lines
Quoted 6 lines
Quoted 6 lines

drivers/soc/fsl/Kconfig config FSL_PLAT_PM · line 30

Quoted 6 lines
drivers/soc/fsl/Kconfig +0 -0
index 6517412..882330d 100644
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
        have to know the implement details of wakeup function it

require.

Quoted 8 lines
Quoted 8 lines
Quoted 8 lines

drivers/soc/fsl/Kconfig config FSL_PLAT_PM

Quoted 8 lines
drivers/soc/fsl/Kconfig +6 -0
        Besides, it is also easy for service side to upgrade its logic when
        design changed and remain user side unchanged.
+
+config LS_RCPM
+     bool "Freescale RCPM support"
+     depends on (FSL_PLAT_PM)
+     help
+       This feature is to enable specified wakeup source for system

sleep.

Quoted 27 lines
Quoted 11 lines
Quoted 9 lines

drivers/soc/fsl/Makefile obj-$(CONFIG_QUICC_ENGINE) += qe/ · line 10

Quoted 9 lines
drivers/soc/fsl/Makefile +1 -0
index 8f9db23..43ff71a 100644
@@ -7,3 +7,4 @@ obj-$(CONFIG_QUICC_ENGINE)            += qe/
 obj-$(CONFIG_CPM)                    += qe/
 obj-$(CONFIG_FSL_GUTS)                       += guts.o
 obj-$(CONFIG_FSL_PLAT_PM)    += plat_pm.o
+obj-$(CONFIG_LS_RCPM)                += ls-rcpm.o

Probably use "_" instead of "-" for alignment.

OK, will update in next version

Quoted 10 lines
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 2

Quoted 8 lines
(quoted diff) +0 -0

b/drivers/soc/fsl/ls-rcpm.c new file mode 100644 index 0000000..b0feb88

drivers/soc/fsl/ls-rcpm.c new file +2 -0
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0 // // plat_pm.c - Freescale
+Layerscape RCPM driver

The file name here is not the same as the real file name.

Got it, will correct it.

Quoted 34 lines
Quoted 6 lines
Quoted 4 lines

drivers/soc/fsl/ls-rcpm.c line 7

Quoted 4 lines
drivers/soc/fsl/ls-rcpm.c new file +4 -0
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,

Where do you need the comma in the end?

My bad, will remove comma in next version.

Quoted 25 lines
Quoted 13 lines
Quoted 13 lines
drivers/soc/fsl/ls-rcpm.c new file +13 -0
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h> #include
+<linux/of_address.h> #include <linux/slab.h> #include
+<soc/fsl/plat_pm.h>
+
+#define MAX_COMPATIBLE_NUM   10
+
+struct rcpm_t {
+     struct device *dev;
+     void __iomem *ippdexpcr_addr;
+     bool big_endian;        /* Big/Little endian of RCPM module */
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 27

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+};
+
+// rcpm_handle - Configure RCPM reg according to wake up source
+request // @user_dev: pointer to user's device struct // @flag:
+to
+enable(true) or disable(false) wakeup source // @handle_priv:
+pointer to struct rcpm_t instance // // Return 0 on success
+other negative errno

Although Linus preferred this // comment style. I'm not sure if this will be handled correctly by the kernel-doc compiler.

https://emea01.safelinks.protection.outlook.com/?url=https%3A%2F%2Fw

Quoted 3 lines
Quoted 3 lines

w w.kernel.org%2Fdoc%2Fhtml%2Fv4.18%2Fdoc-guide%2Fkernel-

doc.html&amp;data=02%7C01%7Cran.wang_1%40nxp.com%7Cc0d88e669038

Quoted 1 line
Quoted 1 line

4e02b95108d612dec235%7C686ea1d3bc2b4c6fa92cd99c5c301635%7C0%7C0

Quoted 1 line
Quoted 1 line

%7C636717145285126200&amp;sdata=H7GkUNOLVG%2FCcZESzhtHBeHC

bO9

Quoted 4 lines
Quoted 1 line

%2FK4k9EdH30Cxq2%2BM%3D&amp;reserved=0

So, do you think I need to change all comment style back to '/* ... */' ? Actually I feel a little bit confused here.

Quoted 7 lines

I think Linus's comment about // comment style applies to normal code comment. But kernel-doc comment is a special kind of code comment that needs to meet certain requirements. People can use the scripts/kernel-doc tool to generate readable API documents from the source code. It looks like you wanted to make the function description aligned with the kernel-doc format, but kernel-doc specifically requires to use the /* */ style(at least for now).

OK, will change style back to /* */.

Regards,
Ran

Quoted 2 lines

Regards,
Leo

author
#13 lore raw

RE: [PATCH 3/3] soc: fsl: add RCPM driver

In reply to Wang, Dongsheng

Hi Dongsheng,

Quoted 91 lines · Dongsheng Wang

Please change your comments style.

Quoted 87 lines · Ran Wang

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153
+++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/ls-rcpm.c

drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

6517412..882330d 100644

drivers/soc/fsl/Kconfig +6 -0
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
 	  have to know the implement details of wakeup function it require.
 	  Besides, it is also easy for service side to upgrade its logic when
 	  design changed and remain user side unchanged.
+
+config LS_RCPM
+	bool "Freescale RCPM support"
+	depends on (FSL_PLAT_PM)
+	help
+	  This feature is to enable specified wakeup source for system sleep.
drivers/soc/fsl/Makefile index renamed from drivers/soc/fsl/Makefile +0 -0

8f9db23..43ff71a 100644

drivers/soc/fsl/Makefile +1 -0
@@ -7,3 +7,4 @@ obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
 obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
+obj-$(CONFIG_LS_RCPM)		+= ls-rcpm.o
drivers/soc/fsl/ls-rcpm.c new renamed from drivers/soc/fsl/ls-rcpm.c +0 -0

file mode 100644 index 0000000..b0feb88

drivers/soc/fsl/ls-rcpm.c new file +41 -0
@@ -0,0 +1,153 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.c - Freescale Layerscape RCPM driver // // Copyright 2018
+NXP // // Author: Ran Wang <[email protected]>,
+
+#include <linux/init.h>
+#include <linux/module.h>
+#include <linux/platform_device.h>
+#include <linux/of_address.h>
+#include <linux/slab.h>
+#include <soc/fsl/plat_pm.h>
+
+#define MAX_COMPATIBLE_NUM	10
+
+struct rcpm_t {
+	struct device *dev;
+	void __iomem *ippdexpcr_addr;
+	bool big_endian;	/* Big/Little endian of RCPM module */
+};
+
+// rcpm_handle - Configure RCPM reg according to wake up source
+request // @user_dev: pointer to user's device struct // @flag: to
+enable(true) or disable(false) wakeup source // @handle_priv: pointer
+to struct rcpm_t instance // // Return 0 on success other negative
+errno static int rcpm_handle(struct device *user_dev, bool flag, void
+*handle_priv) {
+	struct rcpm_t *rcpm;
+	bool big_endian;
+	const char  *dev_compatible_array[MAX_COMPATIBLE_NUM];
+	void __iomem *ippdexpcr_addr;
+	u32 ippdexpcr;
+	u32 set_bit;
+	int ret, num, i;
+
+	rcpm = handle_priv;
+	big_endian = rcpm->big_endian;
+	ippdexpcr_addr = rcpm->ippdexpcr_addr;
+
+	num = device_property_read_string_array(user_dev, "compatible",
+			dev_compatible_array, MAX_COMPATIBLE_NUM);
+	if (num < 0)
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 53

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +7 -0
+		return num;
+
+	for (i = 0; i < num; i++) {
+		if (!device_property_present(rcpm->dev,
+					dev_compatible_array[i]))
+			continue;
+		else {

Remove this else.

Got it! Will update in next version.

Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 60

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +7 -0
+			ret = device_property_read_u32(rcpm->dev,
+					dev_compatible_array[i], &set_bit);
+			if (ret)
+				return ret;
+
+			if (!device_property_present(rcpm->dev,
+						dev_compatible_array[i]))

This has been checked. Continue ? or return ENODEV?

Yes, this checking is not necessary, will remove in next version

Quoted 3 lines

drivers/soc/fsl/ls-rcpm.c line 62

Quoted 2 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
+				return -ENODEV;
+			else {

Remove this else.

OK

Quoted 23 lines

drivers/soc/fsl/ls-rcpm.c line 59

Quoted 2 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
+				ret = device_property_read_u32(rcpm->dev,
+						dev_compatible_array[i],

&set_bit);

drivers/soc/fsl/ls-rcpm.c

Quoted 5 lines
drivers/soc/fsl/ls-rcpm.c new file +5 -0
+				if (ret)
+					return ret;
+
+				if (big_endian)
+					ippdexpcr =

ioread32be(ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c line 66

Quoted 2 lines
drivers/soc/fsl/ls-rcpm.c new file +2 -0
+				else
+					ippdexpcr =

ioread32(ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c

Quoted 8 lines
drivers/soc/fsl/ls-rcpm.c new file +8 -0
+
+				if (flag)
+					ippdexpcr |= set_bit;
+				else
+					ippdexpcr &= ~set_bit;
+
+				if (big_endian) {
+					iowrite32be(ippdexpcr,

ippdexpcr_addr);

drivers/soc/fsl/ls-rcpm.c line 75

Quoted 1 line
drivers/soc/fsl/ls-rcpm.c new file +1 -0
+					ippdexpcr =

ioread32be(ippdexpcr_addr);

Quoted 7 lines

drivers/soc/fsl/ls-rcpm.c line 81

Quoted 1 line
drivers/soc/fsl/ls-rcpm.c new file +1 -0
+				} else

if (x) {
....
....
} else {

}

Got it!

Quoted 12 lines
Quoted 12 lines
drivers/soc/fsl/ls-rcpm.c new file +12 -0
+					iowrite32(ippdexpcr, ippdexpcr_addr);
+
+				return 0;
+			}
+		}
+	}
+
+	return -ENODEV;
+}
+
+static int ls_rcpm_probe(struct platform_device *pdev) {
+	struct resource *r;
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 101

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +7 -0
+	struct rcpm_t *rcpm;
+
+	r = platform_get_resource(pdev, IORESOURCE_MEM, 0);
+	if (!r)
+		return -ENODEV;
+
+	rcpm = kmalloc(sizeof(*rcpm), GFP_KERNEL);

kzalloc is better.

OK

Quoted 10 lines
Quoted 10 lines
drivers/soc/fsl/ls-rcpm.c new file +10 -0
+	if (!rcpm)
+		return -ENOMEM;
+
+	rcpm->big_endian = device_property_read_bool(&pdev->dev,
+"big-endian");
+
+	rcpm->ippdexpcr_addr = devm_ioremap_resource(&pdev->dev, r);
+	if (IS_ERR(rcpm->ippdexpcr_addr))
+		return PTR_ERR(rcpm->ippdexpcr_addr);
+
Quoted 8 lines

drivers/soc/fsl/ls-rcpm.c line 119

Quoted 7 lines
drivers/soc/fsl/ls-rcpm.c new file +7 -0
+	rcpm->dev = &pdev->dev;
+	platform_set_drvdata(pdev, rcpm);
+
+	return register_fsl_platform_wakeup_source(rcpm_handle, rcpm); }
+
+static int ls_rcpm_remove(struct platform_device *pdev) {
+	struct rcpm_t	*rcpm;

Not need a table.

OK, thanks for your careful review.

Regards,
Ran

Quoted 37 lines

Cheers,
-Dongsheng

Quoted 33 lines
drivers/soc/fsl/ls-rcpm.c new file +33 -0
+
+	rcpm = platform_get_drvdata(pdev);
+	deregister_fsl_platform_wakeup_source(rcpm);
+	kfree(rcpm);
+
+	return 0;
+}
+
+static const struct of_device_id ls_rcpm_of_match[] = {
+	{ .compatible = "fsl,qoriq-rcpm-2.1", },
+	{}
+};
+MODULE_DEVICE_TABLE(of, ls_rcpm_of_match);
+
+static struct platform_driver ls_rcpm_driver = {
+	.driver = {
+		.name = "ls-rcpm",
+		.of_match_table = ls_rcpm_of_match,
+	},
+	.probe = ls_rcpm_probe,
+	.remove = ls_rcpm_remove,
+};
+
+static int __init ls_rcpm_init(void)
+{
+	return platform_driver_register(&ls_rcpm_driver);
+}
+subsys_initcall(ls_rcpm_init);
+
+static void __exit ls_rcpm_exit(void) {
+	platform_driver_unregister(&ls_rcpm_driver);
+}
+module_exit(ls_rcpm_exit);
#14 lore raw

Re: [PATCH 3/3] soc: fsl: add RCPM driver

In reply to Ran Wang
Quoted 10 lines · Ran Wang

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

Quoted 7 lines
---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153
+++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)
 create mode 100644 drivers/soc/fsl/ls-rcpm.c

Is there a reason why this is LS-specific, or could it be used with PPC RCPM blocks?

drivers/soc/fsl/Kconfig config FSL_PLAT_PM · line 36

Quoted 13 lines
drivers/soc/fsl/Kconfig +0 -0
index 6517412..882330d 100644
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
 	  have to know the implement details of wakeup function it require.
 	  Besides, it is also easy for service side to upgrade its logic

when

drivers/soc/fsl/Kconfig +4 -0
 	  design changed and remain user side unchanged.
+
+config LS_RCPM
+	bool "Freescale RCPM support"
+	depends on (FSL_PLAT_PM)

Why is this parenthesized?

-Scott

author
#15 lore raw

RE: [PATCH 3/3] soc: fsl: add RCPM driver

Hi Scott,

Quoted 20 lines · Scott Wood
Quoted 10 lines · Ran Wang

The NXP's QorIQ Processors based on ARM Core have RCPM module (Run Control and Power Management), which performs all device-level tasks associated with power management such as wakeup source control.

This driver depends on FSL platform PM driver framework which help to isolate user and PM service provider (such as RCPM driver).

Quoted 7 lines
---
 drivers/soc/fsl/Kconfig   |    6 ++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/ls-rcpm.c |  153
+++++++++++++++++++++++++++++++++++++++++++++
 3 files changed, 160 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/ls-rcpm.c

Quoted 2 lines

Is there a reason why this is LS-specific, or could it be used with PPC RCPM blocks?

They have different SW arch design on low power operation: PPC RCPM driver has taken care of most things of suspend enter & exit. And LS RCPM driver will only handle wakeup source configure and left rest work to system firmware to do. So you might be aware that LS RCPM will only get call whenever plat_pm driver API is called rather than system suspend begin where.

Quoted 16 lines
Quoted 9 lines
drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

6517412..882330d 100644

drivers/soc/fsl/Kconfig +0 -0
@@ -30,3 +30,9 @@ config FSL_PLAT_PM
 	  have to know the implement details of wakeup function it require.
 	  Besides, it is also easy for service side to upgrade its logic

when

drivers/soc/fsl/Kconfig config FSL_PLAT_PM · line 36

Quoted 5 lines
drivers/soc/fsl/Kconfig +4 -0
 	  design changed and remain user side unchanged.
+
+config LS_RCPM
+	bool "Freescale RCPM support"
+	depends on (FSL_PLAT_PM)

Why is this parenthesized?

Because we'd like to decouple RCPM driver and its user.
Benefit is that will allow user doesn't have to know who will serve it for some PM features (such as wake up source control), and provide some kind of flexibility when either RCMP or user driver evolve in the future. So I add a plat_pm driver to prevent wake up IP knowing any information of RCPM.

Regards,
Ran

Quoted 2 lines

-Scott

#16 lore raw
In reply to Ran Wang

Please change your comments style.

Quoted 147 lines · Ran Wang

This driver is to provide a independent framework for PM service provider and consumer to configure system level wake up feature. For example, RCPM driver could register a callback function on this platform first, and Flex timer driver who want to enable timer wake up feature, will call generic API provided by this platform driver, and then it will trigger RCPM driver to do it. The benefit is to isolate the user and service, such as flex timer driver will not have to know the implement details of wakeup function it require. Besides, it is also easy for service side to upgrade its logic when design is changed and remain user side unchanged.

---
 drivers/soc/fsl/Kconfig   |   14 +++++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/plat_pm.c |  144 +++++++++++++++++++++++++++++++++++++++++++++
 include/soc/fsl/plat_pm.h |   22 +++++++
 4 files changed, 181 insertions(+), 0 deletions(-)
 create mode 100644 drivers/soc/fsl/plat_pm.c
 create mode 100644 include/soc/fsl/plat_pm.h
drivers/soc/fsl/Kconfig +14 -0
index 7a9fb9b..6517412 100644
@@ -16,3 +16,17 @@ config FSL_GUTS
 	  Initially only reading SVR and registering soc device are supported.
 	  Other guts accesses, such as reading RCW, should eventually be moved
 	  into this driver as well.
+
+config FSL_PLAT_PM
+	bool "Freescale platform PM framework"
+	help
+	  This driver is to provide a independent framework for PM service
+	  provider and consumer to configure system level wake up feature. For
+	  example, RCPM driver could register a callback function on this
+	  platform first, and Flex timer driver who want to enable timer wake
+	  up feature, will call generic API provided by this platform driver,
+	  and then it will trigger RCPM driver to do it. The benefit is to
+	  isolate the user and service, such as  flex timer driver will not
+	  have to know the implement details of wakeup function it require.
+	  Besides, it is also easy for service side to upgrade its logic when
+	  design changed and remain user side unchanged.
drivers/soc/fsl/Makefile +1 -0
index 44b3beb..8f9db23 100644
@@ -6,3 +6,4 @@ obj-$(CONFIG_FSL_DPAA)                 += qbman/
 obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
+obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
drivers/soc/fsl/plat_pm.c new file +89 -0
new file mode 100644
index 0000000..19ea14e
@@ -0,0 +1,144 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.c - Freescale platform PM framework
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <soc/fsl/plat_pm.h>
+
+
+struct plat_pm_t {
+	struct list_head node;
+	fsl_plat_pm_handle handle;
+	void *handle_priv;
+	spinlock_t	lock;
+};
+
+static struct plat_pm_t plat_pm;
+
+// register_fsl_platform_wakeup_source - Register callback function to plat_pm
+// @handle: Pointer to handle PM feature requirement
+// @handle_priv: Handler specific data struct
+//
+// Return 0 on success other negative errno
+int register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv)
+{
+	struct plat_pm_t *p;
+	unsigned long	flags;
+
+	if (!handle) {
+		pr_err("FSL plat_pm: Handler invalid, reject\n");
+		return -EINVAL;
+	}
+
+	p = kmalloc(sizeof(*p), GFP_KERNEL);
+	if (!p)
+		return -ENOMEM;
+
+	p->handle = handle;
+	p->handle_priv = handle_priv;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_add_tail(&p->node, &plat_pm.node);
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(register_fsl_platform_wakeup_source);
+
+// Deregister_fsl_platform_wakeup_source - deregister callback function
+// @handle_priv: Handler specific data struct
+//
+// Return 0 on success other negative errno
+int deregister_fsl_platform_wakeup_source(void *handle_priv)
+{
+	struct plat_pm_t *p, *tmp;
+	unsigned long	flags;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry_safe(p, tmp, &plat_pm.node, node) {
+		if (p->handle_priv == handle_priv) {
+			list_del(&p->node);
+			kfree(p);
+		}
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(deregister_fsl_platform_wakeup_source);
+
+// fsl_platform_wakeup_config - Configure wakeup source by calling handlers
+// @dev: pointer to user's device struct
+// @flag: to tell enable or disable wakeup source
+//
+// Return 0 on success other negative errno
+int fsl_platform_wakeup_config(struct device *dev, bool flag)
+{
+	struct plat_pm_t *p;
+	int ret;
+	bool success_handled;
+	unsigned long	flags;
+

drivers/soc/fsl/plat_pm.c line 99

Quoted 10 lines
drivers/soc/fsl/plat_pm.c new file +10 -0
+	success_handled = false;
+
+	// Will consider success if at least one callback return 0.
+	// Also, rest handles still get oppertunity to be executed
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry(p, &plat_pm.node, node) {
+		if (p->handle) {
+			ret = p->handle(dev, flag, p->handle_priv);
+			if (!ret)
+				success_handled = true;

Miss a break?

drivers/soc/fsl/plat_pm.c line 101

Quoted 2 lines
drivers/soc/fsl/plat_pm.c new file +2 -0
+			else if (ret != -ENODEV) {
+				pr_err("FSL plat_pm: Failed to config wakeup source:%d\n", ret);

Please unlock before return.

drivers/soc/fsl/plat_pm.c line 112

Quoted 11 lines
drivers/soc/fsl/plat_pm.c new file +11 -0
+				return ret;
+			}
+		} else
+			pr_warn("FSL plat_pm: Invalid handler detected, skip\n");
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	if (success_handled == false) {
+		pr_err("FSL plat_pm: Cannot find the matchhed handler for wakeup source config\n");
+		return -ENODEV;
+	}

Add this into the loop.

Quoted 60 lines
drivers/soc/fsl/plat_pm.c new file +32 -0
+
+	return 0;
+}
+
+// fsl_platform_wakeup_enable - Enable wakeup source
+// @dev: pointer to user's device struct
+//
+// Return 0 on success other negative errno
+int fsl_platform_wakeup_enable(struct device *dev)
+{
+	return fsl_platform_wakeup_config(dev, true);
+}
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_enable);
+
+// fsl_platform_wakeup_disable - Disable wakeup source
+// @dev: pointer to user's device struct
+//
+// Return 0 on success other negative errno
+int fsl_platform_wakeup_disable(struct device *dev)
+{
+	return fsl_platform_wakeup_config(dev, false);
+}
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_disable);
+
+static int __init fsl_plat_pm_init(void)
+{
+	spin_lock_init(&plat_pm.lock);
+	INIT_LIST_HEAD(&plat_pm.node);
+	return 0;
+}
+
+core_initcall(fsl_plat_pm_init);
include/soc/fsl/plat_pm.h new file +22 -0
new file mode 100644
index 0000000..bbe151e
@@ -0,0 +1,22 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.h - Freescale platform PM Header
+//
+// Copyright 2018 NXP
+//
+// Author: Ran Wang <[email protected]>,
+
+#ifndef __FSL_PLAT_PM_H
+#define __FSL_PLAT_PM_H
+
+typedef int (*fsl_plat_pm_handle)(struct device *dev, bool flag,
+		void *handle_priv);
+
+int register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv);
+int deregister_fsl_platform_wakeup_source(void *handle_priv);
+int fsl_platform_wakeup_config(struct device *dev, bool flag);
+int fsl_platform_wakeup_enable(struct device *dev);
+int fsl_platform_wakeup_disable(struct device *dev);
+
+#endif	// __FSL_PLAT_PM_H
author
#17 lore raw

Hi Dongsheng

Quoted 149 lines · Dongsheng Wang

Please change your comments style.

Quoted 11 lines · Ran Wang

This driver is to provide a independent framework for PM service provider and consumer to configure system level wake up feature. For example, RCPM driver could register a callback function on this platform first, and Flex timer driver who want to enable timer wake up feature, will call generic API provided by this platform driver, and then it will trigger RCPM driver to do it. The benefit is to isolate the user and service, such as flex timer driver will not have to know the implement details of wakeup function it require. Besides, it is also easy for service side to upgrade its logic when design is changed and remain user side unchanged.

Quoted 5 lines
---
 drivers/soc/fsl/Kconfig   |   14 +++++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/plat_pm.c |  144

+++++++++++++++++++++++++++++++++++++++++++++

drivers/soc/fsl/Kconfig config FSL_GUTS · line 17

Quoted 12 lines
 include/soc/fsl/plat_pm.h |   22 +++++++
 4 files changed, 181 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/plat_pm.c create mode 100644 include/soc/fsl/plat_pm.h

drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

7a9fb9b..6517412 100644

drivers/soc/fsl/Kconfig +0 -0
@@ -16,3 +16,17 @@ config FSL_GUTS
 	  Initially only reading SVR and registering soc device are supported.
 	  Other guts accesses, such as reading RCW, should eventually be

moved

drivers/soc/fsl/Kconfig config FSL_GUTS

Quoted 7 lines
drivers/soc/fsl/Kconfig +6 -0
 	  into this driver as well.
+
+config FSL_PLAT_PM
+	bool "Freescale platform PM framework"
+	help
+	  This driver is to provide a independent framework for PM service
+	  provider and consumer to configure system level wake up feature.

For

Quoted 106 lines
drivers/soc/fsl/Kconfig +8 -0
+	  example, RCPM driver could register a callback function on this
+	  platform first, and Flex timer driver who want to enable timer wake
+	  up feature, will call generic API provided by this platform driver,
+	  and then it will trigger RCPM driver to do it. The benefit is to
+	  isolate the user and service, such as  flex timer driver will not
+	  have to know the implement details of wakeup function it require.
+	  Besides, it is also easy for service side to upgrade its logic when
+	  design changed and remain user side unchanged.
drivers/soc/fsl/Makefile index renamed from drivers/soc/fsl/Makefile +0 -0

44b3beb..8f9db23 100644

drivers/soc/fsl/Makefile +1 -0
@@ -6,3 +6,4 @@ obj-$(CONFIG_FSL_DPAA)                 += qbman/
 obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
+obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
drivers/soc/fsl/plat_pm.c new renamed from drivers/soc/fsl/plat_pm.c +0 -0

file mode 100644 index 0000000..19ea14e

drivers/soc/fsl/plat_pm.c new file +84 -0
@@ -0,0 +1,144 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.c - Freescale platform PM framework // // Copyright 2018
+NXP // // Author: Ran Wang <[email protected]>,
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <soc/fsl/plat_pm.h>
+
+
+struct plat_pm_t {
+	struct list_head node;
+	fsl_plat_pm_handle handle;
+	void *handle_priv;
+	spinlock_t	lock;
+};
+
+static struct plat_pm_t plat_pm;
+
+// register_fsl_platform_wakeup_source - Register callback function
+to plat_pm // @handle: Pointer to handle PM feature requirement //
+ at handle_priv: Handler specific data struct // // Return 0 on success
+other negative errno int
+register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv)
+{
+	struct plat_pm_t *p;
+	unsigned long	flags;
+
+	if (!handle) {
+		pr_err("FSL plat_pm: Handler invalid, reject\n");
+		return -EINVAL;
+	}
+
+	p = kmalloc(sizeof(*p), GFP_KERNEL);
+	if (!p)
+		return -ENOMEM;
+
+	p->handle = handle;
+	p->handle_priv = handle_priv;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_add_tail(&p->node, &plat_pm.node);
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(register_fsl_platform_wakeup_source);
+
+// Deregister_fsl_platform_wakeup_source - deregister callback
+function // @handle_priv: Handler specific data struct // // Return 0
+on success other negative errno int
+deregister_fsl_platform_wakeup_source(void *handle_priv) {
+	struct plat_pm_t *p, *tmp;
+	unsigned long	flags;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry_safe(p, tmp, &plat_pm.node, node) {
+		if (p->handle_priv == handle_priv) {
+			list_del(&p->node);
+			kfree(p);
+		}
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(deregister_fsl_platform_wakeup_source);
+
+// fsl_platform_wakeup_config - Configure wakeup source by calling
+handlers // @dev: pointer to user's device struct // @flag: to tell
+enable or disable wakeup source // // Return 0 on success other
+negative errno int fsl_platform_wakeup_config(struct device *dev,
+bool flag) {
+	struct plat_pm_t *p;
+	int ret;
+	bool success_handled;
+	unsigned long	flags;
+
+	success_handled = false;
+
+	// Will consider success if at least one callback return 0.
Quoted 8 lines

drivers/soc/fsl/plat_pm.c line 99

Quoted 7 lines
drivers/soc/fsl/plat_pm.c new file +7 -0
+	// Also, rest handles still get oppertunity to be executed
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry(p, &plat_pm.node, node) {
+		if (p->handle) {
+			ret = p->handle(dev, flag, p->handle_priv);
+			if (!ret)
+				success_handled = true;

Miss a break?

Actually my idea is to allow more than one registered handler to handle this
request, so I define a flag rather than return to indicated if there is at least one handler successfully do it. This design might give more flexibility to framework when running.

Quoted 4 lines

drivers/soc/fsl/plat_pm.c line 93

Quoted 2 lines
drivers/soc/fsl/plat_pm.c new file +2 -0
+			else if (ret != -ENODEV) {
+				pr_err("FSL plat_pm: Failed to config wakeup

source:%d\n", ret);
Please unlock before return.

Yes, will fix it in next version, thanks for pointing out!

Quoted 11 lines

drivers/soc/fsl/plat_pm.c

Quoted 4 lines
drivers/soc/fsl/plat_pm.c new file +4 -0
+				return ret;
+			}
+		} else
+			pr_warn("FSL plat_pm: Invalid handler detected,

skip\n");

drivers/soc/fsl/plat_pm.c

Quoted 5 lines
drivers/soc/fsl/plat_pm.c new file +5 -0
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	if (success_handled == false) {
+		pr_err("FSL plat_pm: Cannot find the matchhed handler for

wakeup source config\n");

Quoted 3 lines

drivers/soc/fsl/plat_pm.c line 104

Quoted 2 lines
drivers/soc/fsl/plat_pm.c new file +2 -0
+		return -ENODEV;
+	}

Add this into the loop.

My design is that if the 1st handler return -ENODEV to indicated this device it doesn't support, then the framework will continue try 2nd handler...

So I think it is needed to place this checking out of loop, what do you say?

Regards,
Ran

Quoted 48 lines
Quoted 47 lines
drivers/soc/fsl/plat_pm.c new file +23 -0
+
+	return 0;
+}
+
+// fsl_platform_wakeup_enable - Enable wakeup source // @dev: pointer
+to user's device struct // // Return 0 on success other negative
+errno int fsl_platform_wakeup_enable(struct device *dev) {
+	return fsl_platform_wakeup_config(dev, true); }
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_enable);
+
+// fsl_platform_wakeup_disable - Disable wakeup source // @dev:
+pointer to user's device struct // // Return 0 on success other
+negative errno int fsl_platform_wakeup_disable(struct device *dev) {
+	return fsl_platform_wakeup_config(dev, false); }
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_disable);
+
+static int __init fsl_plat_pm_init(void) {
+	spin_lock_init(&plat_pm.lock);
+	INIT_LIST_HEAD(&plat_pm.node);
+	return 0;
+}
+
+core_initcall(fsl_plat_pm_init);
include/soc/fsl/plat_pm.h new renamed from include/soc/fsl/plat_pm.h +0 -0

file mode 100644 index 0000000..bbe151e

include/soc/fsl/plat_pm.h new file +19 -0
@@ -0,0 +1,22 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.h - Freescale platform PM Header // // Copyright 2018 NXP
+// // Author: Ran Wang <[email protected]>,
+
+#ifndef __FSL_PLAT_PM_H
+#define __FSL_PLAT_PM_H
+
+typedef int (*fsl_plat_pm_handle)(struct device *dev, bool flag,
+		void *handle_priv);
+
+int register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv);
+int deregister_fsl_platform_wakeup_source(void *handle_priv); int
+fsl_platform_wakeup_config(struct device *dev, bool flag); int
+fsl_platform_wakeup_enable(struct device *dev); int
+fsl_platform_wakeup_disable(struct device *dev);
+
+#endif	// __FSL_PLAT_PM_H
#18 lore raw
Quoted 159 lines · Ran Wang

Hi Dongsheng

Quoted 149 lines · Dongsheng Wang

Please change your comments style.

Quoted 11 lines · Ran Wang

This driver is to provide a independent framework for PM service provider and consumer to configure system level wake up feature. For example, RCPM driver could register a callback function on this platform first, and Flex timer driver who want to enable timer wake up feature, will call generic API provided by this platform driver, and then it will trigger RCPM driver to do it. The benefit is to isolate the user and service, such as flex timer driver will not have to know the implement details of wakeup function it require. Besides, it is also easy for service side to upgrade its logic when design is changed and remain user side unchanged.

Quoted 5 lines
---
 drivers/soc/fsl/Kconfig   |   14 +++++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/plat_pm.c |  144

+++++++++++++++++++++++++++++++++++++++++++++

drivers/soc/fsl/Kconfig config FSL_GUTS · line 17

Quoted 12 lines
 include/soc/fsl/plat_pm.h |   22 +++++++
 4 files changed, 181 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/plat_pm.c create mode 100644 include/soc/fsl/plat_pm.h

drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

7a9fb9b..6517412 100644

drivers/soc/fsl/Kconfig +0 -0
@@ -16,3 +16,17 @@ config FSL_GUTS
 	  Initially only reading SVR and registering soc device are supported.
 	  Other guts accesses, such as reading RCW, should eventually be

moved

drivers/soc/fsl/Kconfig config FSL_GUTS

Quoted 7 lines
drivers/soc/fsl/Kconfig +6 -0
 	  into this driver as well.
+
+config FSL_PLAT_PM
+	bool "Freescale platform PM framework"
+	help
+	  This driver is to provide a independent framework for PM service
+	  provider and consumer to configure system level wake up feature.

For

Quoted 106 lines
drivers/soc/fsl/Kconfig +8 -0
+	  example, RCPM driver could register a callback function on this
+	  platform first, and Flex timer driver who want to enable timer wake
+	  up feature, will call generic API provided by this platform driver,
+	  and then it will trigger RCPM driver to do it. The benefit is to
+	  isolate the user and service, such as  flex timer driver will not
+	  have to know the implement details of wakeup function it require.
+	  Besides, it is also easy for service side to upgrade its logic when
+	  design changed and remain user side unchanged.
drivers/soc/fsl/Makefile index renamed from drivers/soc/fsl/Makefile +0 -0

44b3beb..8f9db23 100644

drivers/soc/fsl/Makefile +1 -0
@@ -6,3 +6,4 @@ obj-$(CONFIG_FSL_DPAA)                 += qbman/
 obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
+obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
drivers/soc/fsl/plat_pm.c new renamed from drivers/soc/fsl/plat_pm.c +0 -0

file mode 100644 index 0000000..19ea14e

drivers/soc/fsl/plat_pm.c new file +84 -0
@@ -0,0 +1,144 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.c - Freescale platform PM framework // // Copyright 2018
+NXP // // Author: Ran Wang <[email protected]>,
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <soc/fsl/plat_pm.h>
+
+
+struct plat_pm_t {
+	struct list_head node;
+	fsl_plat_pm_handle handle;
+	void *handle_priv;
+	spinlock_t	lock;
+};
+
+static struct plat_pm_t plat_pm;
+
+// register_fsl_platform_wakeup_source - Register callback function
+to plat_pm // @handle: Pointer to handle PM feature requirement //
+ at handle_priv: Handler specific data struct // // Return 0 on success
+other negative errno int
+register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv)
+{
+	struct plat_pm_t *p;
+	unsigned long	flags;
+
+	if (!handle) {
+		pr_err("FSL plat_pm: Handler invalid, reject\n");
+		return -EINVAL;
+	}
+
+	p = kmalloc(sizeof(*p), GFP_KERNEL);
+	if (!p)
+		return -ENOMEM;
+
+	p->handle = handle;
+	p->handle_priv = handle_priv;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_add_tail(&p->node, &plat_pm.node);
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(register_fsl_platform_wakeup_source);
+
+// Deregister_fsl_platform_wakeup_source - deregister callback
+function // @handle_priv: Handler specific data struct // // Return 0
+on success other negative errno int
+deregister_fsl_platform_wakeup_source(void *handle_priv) {
+	struct plat_pm_t *p, *tmp;
+	unsigned long	flags;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry_safe(p, tmp, &plat_pm.node, node) {
+		if (p->handle_priv == handle_priv) {
+			list_del(&p->node);
+			kfree(p);
+		}
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(deregister_fsl_platform_wakeup_source);
+
+// fsl_platform_wakeup_config - Configure wakeup source by calling
+handlers // @dev: pointer to user's device struct // @flag: to tell
+enable or disable wakeup source // // Return 0 on success other
+negative errno int fsl_platform_wakeup_config(struct device *dev,
+bool flag) {
+	struct plat_pm_t *p;
+	int ret;
+	bool success_handled;
+	unsigned long	flags;
+
+	success_handled = false;
+
+	// Will consider success if at least one callback return 0.
Quoted 8 lines

drivers/soc/fsl/plat_pm.c line 99

Quoted 7 lines
drivers/soc/fsl/plat_pm.c new file +7 -0
+	// Also, rest handles still get oppertunity to be executed
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry(p, &plat_pm.node, node) {
+		if (p->handle) {
+			ret = p->handle(dev, flag, p->handle_priv);
+			if (!ret)
+				success_handled = true;

Miss a break?

Quoted 3 lines

Actually my idea is to allow more than one registered handler to handle this
request, so I define a flag rather than return to indicated if there is at least one handler successfully do it. This design might give more flexibility to framework when running.

There is only one flag(success_handled) here, how did know which handler failed?

BTW, I don't think we need this flag. We can only use the return value.

Cheers,
Dongsheng

Quoted 75 lines
Quoted 4 lines

drivers/soc/fsl/plat_pm.c line 93

Quoted 2 lines
drivers/soc/fsl/plat_pm.c new file +2 -0
+			else if (ret != -ENODEV) {
+				pr_err("FSL plat_pm: Failed to config wakeup

source:%d\n", ret);
Please unlock before return.

Yes, will fix it in next version, thanks for pointing out!

Quoted 11 lines

drivers/soc/fsl/plat_pm.c

Quoted 4 lines
drivers/soc/fsl/plat_pm.c new file +4 -0
+				return ret;
+			}
+		} else
+			pr_warn("FSL plat_pm: Invalid handler detected,

skip\n");

drivers/soc/fsl/plat_pm.c

Quoted 5 lines
drivers/soc/fsl/plat_pm.c new file +5 -0
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	if (success_handled == false) {
+		pr_err("FSL plat_pm: Cannot find the matchhed handler for

wakeup source config\n");

Quoted 3 lines

drivers/soc/fsl/plat_pm.c line 104

Quoted 2 lines
drivers/soc/fsl/plat_pm.c new file +2 -0
+		return -ENODEV;
+	}

Add this into the loop.

My design is that if the 1st handler return -ENODEV to indicated this device it doesn't support, then the framework will continue try 2nd handler...

So I think it is needed to place this checking out of loop, what do you say?

Regards,
Ran

Quoted 47 lines
Quoted 47 lines
drivers/soc/fsl/plat_pm.c new file +23 -0
+
+	return 0;
+}
+
+// fsl_platform_wakeup_enable - Enable wakeup source // @dev: pointer
+to user's device struct // // Return 0 on success other negative
+errno int fsl_platform_wakeup_enable(struct device *dev) {
+	return fsl_platform_wakeup_config(dev, true); }
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_enable);
+
+// fsl_platform_wakeup_disable - Disable wakeup source // @dev:
+pointer to user's device struct // // Return 0 on success other
+negative errno int fsl_platform_wakeup_disable(struct device *dev) {
+	return fsl_platform_wakeup_config(dev, false); }
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_disable);
+
+static int __init fsl_plat_pm_init(void) {
+	spin_lock_init(&plat_pm.lock);
+	INIT_LIST_HEAD(&plat_pm.node);
+	return 0;
+}
+
+core_initcall(fsl_plat_pm_init);
include/soc/fsl/plat_pm.h new renamed from include/soc/fsl/plat_pm.h +0 -0

file mode 100644 index 0000000..bbe151e

include/soc/fsl/plat_pm.h new file +19 -0
@@ -0,0 +1,22 @@
+// SPDX-License-Identifier: GPL-2.0
+//
+// plat_pm.h - Freescale platform PM Header // // Copyright 2018 NXP
+// // Author: Ran Wang <[email protected]>,
+
+#ifndef __FSL_PLAT_PM_H
+#define __FSL_PLAT_PM_H
+
+typedef int (*fsl_plat_pm_handle)(struct device *dev, bool flag,
+		void *handle_priv);
+
+int register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv);
+int deregister_fsl_platform_wakeup_source(void *handle_priv); int
+fsl_platform_wakeup_config(struct device *dev, bool flag); int
+fsl_platform_wakeup_enable(struct device *dev); int
+fsl_platform_wakeup_disable(struct device *dev);
+
+#endif	// __FSL_PLAT_PM_H
author
#19 lore raw

Hi Dongsheng,

Quoted 161 lines · Dongsheng Wang
Quoted 159 lines · Ran Wang

Hi Dongsheng

Quoted 157 lines · Dongsheng Wang

Please change your comments style.

Quoted 11 lines · Ran Wang

This driver is to provide a independent framework for PM service provider and consumer to configure system level wake up feature. For example, RCPM driver could register a callback function on this platform first, and Flex timer driver who want to enable timer wake up feature, will call generic API provided by this platform driver, and then it will trigger RCPM driver to do it. The benefit is to isolate the user and service, such as flex timer driver will not have to know the implement details of wakeup function it require. Besides, it is also easy for service side to upgrade its logic when design is changed and remain user side unchanged.

Quoted 5 lines
---
 drivers/soc/fsl/Kconfig   |   14 +++++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/plat_pm.c |  144

+++++++++++++++++++++++++++++++++++++++++++++

drivers/soc/fsl/Kconfig config FSL_GUTS · line 17

Quoted 12 lines
 include/soc/fsl/plat_pm.h |   22 +++++++
 4 files changed, 181 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/plat_pm.c create mode 100644 include/soc/fsl/plat_pm.h

drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

7a9fb9b..6517412 100644

drivers/soc/fsl/Kconfig +0 -0
@@ -16,3 +16,17 @@ config FSL_GUTS
 	  Initially only reading SVR and registering soc device are supported.
 	  Other guts accesses, such as reading RCW, should eventually be

moved

drivers/soc/fsl/Kconfig config FSL_GUTS

Quoted 7 lines
drivers/soc/fsl/Kconfig +6 -0
 	  into this driver as well.
+
+config FSL_PLAT_PM
+	bool "Freescale platform PM framework"
+	help
+	  This driver is to provide a independent framework for PM service
+	  provider and consumer to configure system level wake up feature.

For

Quoted 103 lines
drivers/soc/fsl/Kconfig +8 -0
+	  example, RCPM driver could register a callback function on this
+	  platform first, and Flex timer driver who want to enable timer wake
+	  up feature, will call generic API provided by this platform driver,
+	  and then it will trigger RCPM driver to do it. The benefit is to
+	  isolate the user and service, such as  flex timer driver will not
+	  have to know the implement details of wakeup function it require.
+	  Besides, it is also easy for service side to upgrade its logic when
+	  design changed and remain user side unchanged.
drivers/soc/fsl/Makefile +0 -0

index
44b3beb..8f9db23 100644

drivers/soc/fsl/Makefile +1 -0
@@ -6,3 +6,4 @@ obj-$(CONFIG_FSL_DPAA)                 += qbman/
 obj-$(CONFIG_QUICC_ENGINE)		+= qe/
 obj-$(CONFIG_CPM)			+= qe/
 obj-$(CONFIG_FSL_GUTS)			+= guts.o
+obj-$(CONFIG_FSL_PLAT_PM)	+= plat_pm.o
drivers/soc/fsl/plat_pm.c new file +80 -0
new file mode 100644 index 0000000..19ea14e
@@ -0,0 +1,144 @@
+// SPDX-License-Identifier: GPL-2.0 // // plat_pm.c - Freescale
+platform PM framework // // Copyright 2018 NXP // // Author: Ran
+Wang <[email protected]>,
+
+#include <linux/kernel.h>
+#include <linux/device.h>
+#include <linux/list.h>
+#include <linux/slab.h>
+#include <linux/err.h>
+#include <soc/fsl/plat_pm.h>
+
+
+struct plat_pm_t {
+	struct list_head node;
+	fsl_plat_pm_handle handle;
+	void *handle_priv;
+	spinlock_t	lock;
+};
+
+static struct plat_pm_t plat_pm;
+
+// register_fsl_platform_wakeup_source - Register callback function
+to plat_pm // @handle: Pointer to handle PM feature requirement //
+ at handle_priv: Handler specific data struct // // Return 0 on
+success other negative errno int
+register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv)
+{
+	struct plat_pm_t *p;
+	unsigned long	flags;
+
+	if (!handle) {
+		pr_err("FSL plat_pm: Handler invalid, reject\n");
+		return -EINVAL;
+	}
+
+	p = kmalloc(sizeof(*p), GFP_KERNEL);
+	if (!p)
+		return -ENOMEM;
+
+	p->handle = handle;
+	p->handle_priv = handle_priv;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_add_tail(&p->node, &plat_pm.node);
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	return 0;
+}
+EXPORT_SYMBOL_GPL(register_fsl_platform_wakeup_source);
+
+// Deregister_fsl_platform_wakeup_source - deregister callback
+function // @handle_priv: Handler specific data struct // // Return
+0 on success other negative errno int
+deregister_fsl_platform_wakeup_source(void *handle_priv) {
+	struct plat_pm_t *p, *tmp;
+	unsigned long	flags;
+
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry_safe(p, tmp, &plat_pm.node, node) {
+		if (p->handle_priv == handle_priv) {
+			list_del(&p->node);
+			kfree(p);
+		}
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+	return 0;
+}
+EXPORT_SYMBOL_GPL(deregister_fsl_platform_wakeup_source);
+
+// fsl_platform_wakeup_config - Configure wakeup source by calling
+handlers // @dev: pointer to user's device struct // @flag: to tell
+enable or disable wakeup source // // Return 0 on success other
+negative errno int fsl_platform_wakeup_config(struct device *dev,
+bool flag) {
+	struct plat_pm_t *p;
+	int ret;
+	bool success_handled;
+	unsigned long	flags;
+

drivers/soc/fsl/plat_pm.c line 90

Quoted 10 lines
drivers/soc/fsl/plat_pm.c new file +10 -0
+	success_handled = false;
+
+	// Will consider success if at least one callback return 0.
+	// Also, rest handles still get oppertunity to be executed
+	spin_lock_irqsave(&plat_pm.lock, flags);
+	list_for_each_entry(p, &plat_pm.node, node) {
+		if (p->handle) {
+			ret = p->handle(dev, flag, p->handle_priv);
+			if (!ret)
+				success_handled = true;

Miss a break?

Quoted 8 lines
Quoted 3 lines

Actually my idea is to allow more than one registered handler to handle this request, so I define a flag rather than return to indicated if there is at least one handler successfully do it. This design

might give more flexibility to framework when running.
There is only one flag(success_handled) here, how did know which handler failed?

BTW, I don't think we need this flag. We can only use the return value.

Well, the plat_pm driver will not handle most errors returned by registered handlers, except -NODEV. For -NODEV, plat_pm driver consider that handler cannot support this request and will go to call next one.

Besides, actually it doesn't restrict that request can be served by only one handler. So I add that flag to cover the case of more than one handler can successfully support and others might return -NODEV.

Regards,
Ran

Quoted 78 lines

Cheers,
Dongsheng

Quoted 17 lines
Quoted 4 lines

drivers/soc/fsl/plat_pm.c line 92

Quoted 2 lines
drivers/soc/fsl/plat_pm.c new file +2 -0
+			else if (ret != -ENODEV) {
+				pr_err("FSL plat_pm: Failed to config wakeup

source:%d\n", ret);
Please unlock before return.

Yes, will fix it in next version, thanks for pointing out!

Quoted 11 lines

drivers/soc/fsl/plat_pm.c

Quoted 4 lines
drivers/soc/fsl/plat_pm.c new file +4 -0
+				return ret;
+			}
+		} else
+			pr_warn("FSL plat_pm: Invalid handler detected,

skip\n");

drivers/soc/fsl/plat_pm.c

Quoted 5 lines
drivers/soc/fsl/plat_pm.c new file +5 -0
+	}
+	spin_unlock_irqrestore(&plat_pm.lock, flags);
+
+	if (success_handled == false) {
+		pr_err("FSL plat_pm: Cannot find the matchhed handler for

wakeup source config\n");

Quoted 5 lines
Quoted 3 lines

drivers/soc/fsl/plat_pm.c line 103

Quoted 2 lines
drivers/soc/fsl/plat_pm.c new file +2 -0
+		return -ENODEV;
+	}

Add this into the loop.

My design is that if the 1st handler return -ENODEV to indicated this device it doesn't support, then the framework will continue try 2nd

handler...

Quoted 52 lines

So I think it is needed to place this checking out of loop, what do you say?

Regards,
Ran

Quoted 46 lines
Quoted 46 lines
drivers/soc/fsl/plat_pm.c new file +23 -0
+
+	return 0;
+}
+
+// fsl_platform_wakeup_enable - Enable wakeup source // @dev:
+pointer to user's device struct // // Return 0 on success other
+negative errno int fsl_platform_wakeup_enable(struct device *dev) {
+	return fsl_platform_wakeup_config(dev, true); }
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_enable);
+
+// fsl_platform_wakeup_disable - Disable wakeup source // @dev:
+pointer to user's device struct // // Return 0 on success other
+negative errno int fsl_platform_wakeup_disable(struct device *dev) {
+	return fsl_platform_wakeup_config(dev, false); }
+EXPORT_SYMBOL_GPL(fsl_platform_wakeup_disable);
+
+static int __init fsl_plat_pm_init(void) {
+	spin_lock_init(&plat_pm.lock);
+	INIT_LIST_HEAD(&plat_pm.node);
+	return 0;
+}
+
+core_initcall(fsl_plat_pm_init);
include/soc/fsl/plat_pm.h new file +18 -0
new file mode 100644 index 0000000..bbe151e
@@ -0,0 +1,22 @@
+// SPDX-License-Identifier: GPL-2.0 // // plat_pm.h - Freescale
+platform PM Header // // Copyright 2018 NXP // // Author: Ran Wang
+<[email protected]>,
+
+#ifndef __FSL_PLAT_PM_H
+#define __FSL_PLAT_PM_H
+
+typedef int (*fsl_plat_pm_handle)(struct device *dev, bool flag,
+		void *handle_priv);
+
+int register_fsl_platform_wakeup_source(fsl_plat_pm_handle handle,
+		void *handle_priv);
+int deregister_fsl_platform_wakeup_source(void *handle_priv); int
+fsl_platform_wakeup_config(struct device *dev, bool flag); int
+fsl_platform_wakeup_enable(struct device *dev); int
+fsl_platform_wakeup_disable(struct device *dev);
+
+#endif	// __FSL_PLAT_PM_H
#20 lore raw
In reply to Ran Wang
Quoted 28 lines · Ran Wang

This driver is to provide a independent framework for PM service provider and consumer to configure system level wake up feature. For example, RCPM driver could register a callback function on this platform first, and Flex timer driver who want to enable timer wake up feature, will call generic API provided by this platform driver, and then it will trigger RCPM driver to do it. The benefit is to isolate the user and service, such as flex timer driver will not have to know the implement details of wakeup function it require. Besides, it is also easy for service side to upgrade its logic when design is changed and remain user side unchanged.

---
 drivers/soc/fsl/Kconfig   |   14 +++++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/plat_pm.c |  144
+++++++++++++++++++++++++++++++++++++++++++++
 include/soc/fsl/plat_pm.h |   22 +++++++
 4 files changed, 181 insertions(+), 0 deletions(-)
 create mode 100644 drivers/soc/fsl/plat_pm.c
 create mode 100644 include/soc/fsl/plat_pm.h
drivers/soc/fsl/Kconfig +0 -0
index 7a9fb9b..6517412 100644
@@ -16,3 +16,17 @@ config FSL_GUTS
 	  Initially only reading SVR and registering soc device are
Quoted 4 lines

supported.
Other guts accesses, such as reading RCW, should eventually be
moved
into this driver as well.

+

drivers/soc/fsl/Kconfig config FSL_GUTS · line 21

Quoted 2 lines
drivers/soc/fsl/Kconfig +2 -0
+config FSL_PLAT_PM
+	bool "Freescale platform PM framework"

This name seems to be simultaneously too generic (for something that is likely intended only for use with certain Freescale/NXP chip families) and too specific (for something that seems to be general infrastructure with no real hardware dependencies).

What specific problems with Linux's generic wakeup infrastructure is this trying to solve, and why would those problems not be better solved there?

Also, you should CC linux-pm on these patches.

-Scott

author
#21 lore raw

Hi Scott,

Quoted 35 lines · Scott Wood
Quoted 28 lines · Ran Wang

This driver is to provide a independent framework for PM service provider and consumer to configure system level wake up feature. For example, RCPM driver could register a callback function on this platform first, and Flex timer driver who want to enable timer wake up feature, will call generic API provided by this platform driver, and then it will trigger RCPM driver to do it. The benefit is to isolate the user and service, such as flex timer driver will not have to know the implement details of wakeup function it require. Besides, it is also easy for service side to upgrade its logic when design is changed and remain user side unchanged.

---
 drivers/soc/fsl/Kconfig   |   14 +++++
 drivers/soc/fsl/Makefile  |    1 +
 drivers/soc/fsl/plat_pm.c |  144
+++++++++++++++++++++++++++++++++++++++++++++
 include/soc/fsl/plat_pm.h |   22 +++++++
 4 files changed, 181 insertions(+), 0 deletions(-)  create mode

100644 drivers/soc/fsl/plat_pm.c create mode 100644 include/soc/fsl/plat_pm.h

drivers/soc/fsl/Kconfig index renamed from drivers/soc/fsl/Kconfig +0 -0

7a9fb9b..6517412 100644

drivers/soc/fsl/Kconfig +0 -0
@@ -16,3 +16,17 @@ config FSL_GUTS
 	  Initially only reading SVR and registering soc device are
Quoted 4 lines

supported.
Other guts accesses, such as reading RCW, should eventually be
moved
into this driver as well.

+

Quoted 7 lines

drivers/soc/fsl/Kconfig config FSL_GUTS · line 21

Quoted 2 lines
drivers/soc/fsl/Kconfig +2 -0
+config FSL_PLAT_PM
+	bool "Freescale platform PM framework"

This name seems to be simultaneously too generic (for something that is likely intended only for use with certain Freescale/NXP chip families) and too specific (for something that seems to be general infrastructure with no real hardware dependencies).

Yes, this driver has no real HW dependencies at all. But we'd like to introduce it to help providing more flexibility & generic on FSL PM feature configure (so far we have RCPM on system wakeup source control). I think it's good
for driver/IP porting among different SoC in the future. As to the name, do you have better suggestion?

Quoted 2 lines

What specific problems with Linux's generic wakeup infrastructure is this trying to solve, and why would those problems not be better solved there?

Actually, I am not sure if generic wakeup infrastructure have this kind of PM feature (keep specific IP alive during system suspend, could you please show me?). And I think it is not common requirement, so I decide to put it in FSL folder.

Quoted 1 line

Also, you should CC linux-pm on these patches.

Yes, thanks for suggestion

Regards,
Ran

Quoted 1 line

-Scott