Skip to contentSkip to conversation

Add new reg_update_bits() support

PATCH Superseded Baolin Wang 3 participants 12 messages Apr 17, 2020 – Apr 24, 2020

Subsystems (3)

v1 of 4 Superseded Latest is v4 →

lore
Share & tools
lore

Baolin Wang

author maintainer
#0 lore raw

cover Add new reg_update_bits() support

The Spreadtrum platform uses a special set/clear method to update registers' bits, thus this patch set registers a physical regmap bus into syscon core to support this feature instead of using the MMIO bus, which is not a physical regmap bus.

Any comments are welcome. Thanks.

Changes from RFC v2:
- Drop regmap change, which was applied by Mark.
- Add more information about how to use set/clear.
- Add checking to ensure the platform is compatible with using a new physical regmap bus.

Changes from RFC v1:
- Add new helper to registers a physical regmap bus instead of using the MMIO bus.

Baolin Wang (2):
mfd: syscon: Support physical regmap bus
soc: sprd: Add Spreadtrum special bits updating support

 drivers/mfd/syscon.c           | 23 +++++++++-
 drivers/soc/Kconfig            |  1 +
 drivers/soc/Makefile           |  1 +
 drivers/soc/sprd/Kconfig       | 16 +++++++
 drivers/soc/sprd/Makefile      |  2 +
 drivers/soc/sprd/sprd_syscon.c | 82 ++++++++++++++++++++++++++++++++++
 include/linux/mfd/syscon.h     |  7 +++
 7 files changed, 130 insertions(+), 2 deletions(-)
 create mode 100644 drivers/soc/sprd/Kconfig
 create mode 100644 drivers/soc/sprd/Makefile
 create mode 100644 drivers/soc/sprd/sprd_syscon.c
Signature · 1 line
2.17.1
author maintainer
#1 lore raw

PATCH 1/2 mfd: syscon: Support physical regmap bus

Some platforms such as Spreadtrum platform, define a special method to update bits of the registers instead of reading and writing, which means we should use a physical regmap bus to define the reg_update_bits() operation instead of the MMIO regmap bus.

Thus add a new helper for the syscon driver to allow to register a physical regmap bus to support this new requirement.

---
 drivers/mfd/syscon.c       | 23 +++++++++++++++++++++--
 include/linux/mfd/syscon.h |  7 +++++++
 2 files changed, 28 insertions(+), 2 deletions(-)
drivers/mfd/syscon.c +21 -2
index 3a97816d0cba..92bfe87038ca 100644
@@ -24,6 +24,7 @@
 #include <linux/slab.h>
 
 static struct platform_driver syscon_driver;
+static struct regmap_bus *syscon_phy_regmap_bus;
 
 static DEFINE_SPINLOCK(syscon_list_slock);
 static LIST_HEAD(syscon_list);
@@ -106,14 +107,25 @@ static struct syscon *of_syscon_register(struct device_node *np, bool check_clk)
 	syscon_config.val_bits = reg_io_width * 8;
 	syscon_config.max_register = resource_size(&res) - reg_io_width;
 
-	regmap = regmap_init_mmio(NULL, base, &syscon_config);
+	 /*
+	  * The Spreadtrum syscon need register a real physical regmap bus
+	  * with new atomic bits updating operation instead of using
+	  * read-modify-write.
+	  */
+	if (IS_ENABLED(CONFIG_ARCH_SPRD) &&
+	    of_device_is_compatible(np, "sprd,atomic-syscon") &&
+	    syscon_phy_regmap_bus)
+		regmap = regmap_init(NULL, syscon_phy_regmap_bus, base,
+				     &syscon_config);
+	else
+		regmap = regmap_init_mmio(NULL, base, &syscon_config);
 	if (IS_ERR(regmap)) {
 		pr_err("regmap init failed\n");
 		ret = PTR_ERR(regmap);
 		goto err_regmap;
 	}
 
-	if (check_clk) {
+	if (check_clk && !syscon_phy_regmap_bus) {
 		clk = of_clk_get(np, 0);
 		if (IS_ERR(clk)) {
 			ret = PTR_ERR(clk);
@@ -172,6 +184,13 @@ static struct regmap *device_node_get_regmap(struct device_node *np,
 	return syscon->regmap;
 }
 
+void syscon_register_phys_regmap_bus(struct regmap_bus *phy_regmap_bus)
+{
+	if (phy_regmap_bus)
+		syscon_phy_regmap_bus = phy_regmap_bus;
+}
+EXPORT_SYMBOL_GPL(syscon_register_phys_regmap_bus);
+
 struct regmap *device_node_to_regmap(struct device_node *np)
 {
 	return device_node_get_regmap(np, false);
include/linux/mfd/syscon.h +7 -0
index 7f20e9b502a5..fbc2a2f0f414 100644
@@ -13,6 +13,7 @@
 
 #include <linux/err.h>
 #include <linux/errno.h>
+#include <linux/regmap.h>
 
 struct device_node;
 
@@ -28,6 +29,7 @@ extern struct regmap *syscon_regmap_lookup_by_phandle_args(
 					const char *property,
 					int arg_count,
 					unsigned int *out_args);
+extern void syscon_register_phys_regmap_bus(struct regmap_bus *phy_regmap_bus);
 #else
 static inline struct regmap *device_node_to_regmap(struct device_node *np)
 {
@@ -59,6 +61,11 @@ static inline struct regmap *syscon_regmap_lookup_by_phandle_args(
 {
 	return ERR_PTR(-ENOTSUPP);
 }
+
+static inline void
+syscon_register_phys_regmap_bus(struct regmap_bus *phy_regmap_bus)
+{
+}
 #endif
 
 #endif /* __LINUX_MFD_SYSCON_H__ */
Signature · 1 line
2.17.1
#2 lore raw

Re: [PATCH 1/2] mfd: syscon: Support physical regmap bus

Quoted 28 lines · Baolin Wang

Some platforms such as Spreadtrum platform, define a special method to update bits of the registers instead of reading and writing, which means we should use a physical regmap bus to define the reg_update_bits() operation instead of the MMIO regmap bus.

Thus add a new helper for the syscon driver to allow to register a physical regmap bus to support this new requirement.

---
 drivers/mfd/syscon.c       | 23 +++++++++++++++++++++--
 include/linux/mfd/syscon.h |  7 +++++++
 2 files changed, 28 insertions(+), 2 deletions(-)
drivers/mfd/syscon.c +1 -0
index 3a97816d0cba..92bfe87038ca 100644
@@ -24,6 +24,7 @@
 #include <linux/slab.h>
 
 static struct platform_driver syscon_driver;
+static struct regmap_bus *syscon_phy_regmap_bus;
 
 static DEFINE_SPINLOCK(syscon_list_slock);
 static LIST_HEAD(syscon_list);
@@ -106,14 +107,25 @@ static struct syscon *of_syscon_register(struct device_node *np, bool check_clk)
 	syscon_config.val_bits = reg_io_width * 8;

drivers/mfd/syscon.c static struct syscon *of_syscon_register(struct device_node *np, bool check_clk) · line 116

Quoted 10 lines
drivers/mfd/syscon.c +7 -1
 	syscon_config.max_register = resource_size(&res) - reg_io_width;
 
-	regmap = regmap_init_mmio(NULL, base, &syscon_config);
+	 /*
+	  * The Spreadtrum syscon need register a real physical regmap bus
+	  * with new atomic bits updating operation instead of using
+	  * read-modify-write.
+	  */
+	if (IS_ENABLED(CONFIG_ARCH_SPRD) &&
+	    of_device_is_compatible(np, "sprd,atomic-syscon") &&

Please find a more generic way of supporting your use-case. This is a generic driver, and as such I am vehemently against adding any sort of vendor specific code in here.

Quoted 5 lines
drivers/mfd/syscon.c +5 -0
+	    syscon_phy_regmap_bus)
+		regmap = regmap_init(NULL, syscon_phy_regmap_bus, base,
+				     &syscon_config);
+	else
+		regmap = regmap_init_mmio(NULL, base, &syscon_config);
Signature · 4 lines
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
maintainer
#3 lore raw

Re: [PATCH 1/2] mfd: syscon: Support physical regmap bus

Quoted 5 lines · Lee Jones
Quoted 4 lines · Baolin Wang
drivers/mfd/syscon.c +0 -0
@@ -106,14 +107,25 @@ static struct syscon *of_syscon_register(struct device_node *np, bool check_clk)
      syscon_config.val_bits = reg_io_width * 8;
      syscon_config.max_register = resource_size(&res) - reg_io_width;
 
Quoted 12 lines

drivers/mfd/syscon.c static struct syscon *of_syscon_register(struct device_node *np, bool check_clk) · line 116

Quoted 8 lines
drivers/mfd/syscon.c +7 -1
-     regmap = regmap_init_mmio(NULL, base, &syscon_config);
+      /*
+       * The Spreadtrum syscon need register a real physical regmap bus
+       * with new atomic bits updating operation instead of using
+       * read-modify-write.
+       */
+     if (IS_ENABLED(CONFIG_ARCH_SPRD) &&
+         of_device_is_compatible(np, "sprd,atomic-syscon") &&

Please find a more generic way of supporting your use-case. This is a generic driver, and as such I am vehemently against adding any sort of vendor specific code in here.

I suggested doing it this way, as all alternatives seemed worse than this.

Arnd

#4 lore raw

Re: [PATCH 1/2] mfd: syscon: Support physical regmap bus

Quoted 19 lines · Arnd Bergmann
Quoted 5 lines · Lee Jones
Quoted 4 lines · Baolin Wang
drivers/mfd/syscon.c +0 -0
@@ -106,14 +107,25 @@ static struct syscon *of_syscon_register(struct device_node *np, bool check_clk)
      syscon_config.val_bits = reg_io_width * 8;
      syscon_config.max_register = resource_size(&res) - reg_io_width;
 
Quoted 12 lines

drivers/mfd/syscon.c static struct syscon *of_syscon_register(struct device_node *np, bool check_clk) · line 116

Quoted 8 lines
drivers/mfd/syscon.c +7 -1
-     regmap = regmap_init_mmio(NULL, base, &syscon_config);
+      /*
+       * The Spreadtrum syscon need register a real physical regmap bus
+       * with new atomic bits updating operation instead of using
+       * read-modify-write.
+       */
+     if (IS_ENABLED(CONFIG_ARCH_SPRD) &&
+         of_device_is_compatible(np, "sprd,atomic-syscon") &&

Please find a more generic way of supporting your use-case. This is a generic driver, and as such I am vehemently against adding any sort of vendor specific code in here.

Quoted 1 line

I suggested doing it this way, as all alternatives seemed worse than this.

If we're using a registration function (could probably be swapped out for or accompanied by a Device Tree property) anyway, then why conduct the vendor platform checks?

Signature · 4 lines
Lee Jones [李琼斯]
Linaro Services Technical Lead
Linaro.org │ Open source software for ARM SoCs
Follow Linaro: Facebook | Twitter | Blog
author maintainer
#5 lore raw

Re: [PATCH 1/2] mfd: syscon: Support physical regmap bus

Quoted 22 lines · Lee Jones
Quoted 19 lines · Arnd Bergmann
Quoted 17 lines · Lee Jones

drivers/mfd/syscon.c static struct syscon *of_syscon_register(struct device_node *np, bool check_clk) · line 116

Quoted 12 lines · Baolin Wang
drivers/mfd/syscon.c +7 -1
@@ -106,14 +107,25 @@ static struct syscon *of_syscon_register(struct device_node *np, bool check_clk)
      syscon_config.val_bits = reg_io_width * 8;
      syscon_config.max_register = resource_size(&res) - reg_io_width;
 
-     regmap = regmap_init_mmio(NULL, base, &syscon_config);
+      /*
+       * The Spreadtrum syscon need register a real physical regmap bus
+       * with new atomic bits updating operation instead of using
+       * read-modify-write.
+       */
+     if (IS_ENABLED(CONFIG_ARCH_SPRD) &&
+         of_device_is_compatible(np, "sprd,atomic-syscon") &&

Please find a more generic way of supporting your use-case. This is a generic driver, and as such I am vehemently against adding any sort of vendor specific code in here.

Quoted 5 lines
Quoted 1 line

I suggested doing it this way, as all alternatives seemed worse than this.

If we're using a registration function (could probably be swapped out for or accompanied by a Device Tree property) anyway, then why conduct the vendor platform checks?

Actually I've send out the v3 patch according to Arnd's suggestion. In v3 patch, I removed the registration function, but we agreed that adding the vendor specific support in the syscon driver seems a better way than others.

Signature · 1 line
Baolin Wang
author maintainer
#6 lore raw

Re: [PATCH 1/2] mfd: syscon: Support physical regmap bus

Quoted 26 lines · Baolin Wang
Quoted 24 lines · Lee Jones
Quoted 20 lines · Arnd Bergmann
Quoted 5 lines · Lee Jones
Quoted 4 lines · Baolin Wang
drivers/mfd/syscon.c +0 -0
@@ -106,14 +107,25 @@ static struct syscon *of_syscon_register(struct device_node *np, bool check_clk)
      syscon_config.val_bits = reg_io_width * 8;
      syscon_config.max_register = resource_size(&res) - reg_io_width;
 
Quoted 12 lines

drivers/mfd/syscon.c static struct syscon *of_syscon_register(struct device_node *np, bool check_clk) · line 116

Quoted 8 lines
drivers/mfd/syscon.c +7 -1
-     regmap = regmap_init_mmio(NULL, base, &syscon_config);
+      /*
+       * The Spreadtrum syscon need register a real physical regmap bus
+       * with new atomic bits updating operation instead of using
+       * read-modify-write.
+       */
+     if (IS_ENABLED(CONFIG_ARCH_SPRD) &&
+         of_device_is_compatible(np, "sprd,atomic-syscon") &&

Please find a more generic way of supporting your use-case. This is a generic driver, and as such I am vehemently against adding any sort of vendor specific code in here.

I suggested doing it this way, as all alternatives seemed worse than this.

Quoted 8 lines
Quoted 3 lines

If we're using a registration function (could probably be swapped out for or accompanied by a Device Tree property) anyway, then why conduct the vendor platform checks?

Actually I've send out the v3 patch according to Arnd's suggestion. In v3 patch, I removed the registration function, but we agreed that adding the vendor specific support in the syscon driver seems a better way than others.

Sorry, I forgot the link:
https://lore.kernel.org/patchwork/patch/1228160/

Signature · 1 line
Baolin Wang
author maintainer
#7 lore raw

PATCH 2/2 soc: sprd: Add Spreadtrum special bits updating support

In reply to Baolin Wang

The spreadtrum platform uses a special set/clear method to update registers' bits, which can remove the race of updating the global registers between the multiple subsystems. Thus we can register a physical regmap bus into syscon core to support this.

---
 drivers/soc/Kconfig            |  1 +
 drivers/soc/Makefile           |  1 +
 drivers/soc/sprd/Kconfig       | 16 +++++++
 drivers/soc/sprd/Makefile      |  2 +
 drivers/soc/sprd/sprd_syscon.c | 82 ++++++++++++++++++++++++++++++++++
 5 files changed, 102 insertions(+)
 create mode 100644 drivers/soc/sprd/Kconfig
 create mode 100644 drivers/soc/sprd/Makefile
 create mode 100644 drivers/soc/sprd/sprd_syscon.c
drivers/soc/Kconfig +1 -0
index 425ab6f7e375..8cfbf2dc518d 100644
@@ -23,5 +23,6 @@ source "drivers/soc/versatile/Kconfig"
 source "drivers/soc/xilinx/Kconfig"
 source "drivers/soc/zte/Kconfig"
 source "drivers/soc/kendryte/Kconfig"
+source "drivers/soc/sprd/Kconfig"
 
 endmenu
drivers/soc/Makefile +1 -0
index 36452bed86ef..7d156a6dd536 100644
@@ -29,3 +29,4 @@ obj-$(CONFIG_PLAT_VERSATILE)	+= versatile/
 obj-y				+= xilinx/
 obj-$(CONFIG_ARCH_ZX)		+= zte/
 obj-$(CONFIG_SOC_KENDRYTE)	+= kendryte/
+obj-$(CONFIG_ARCH_SPRD)		+= sprd/
drivers/soc/sprd/Kconfig new file +16 -0
new file mode 100644
index 000000000000..38d1f5971a28
@@ -0,0 +1,16 @@
+# SPDX-License-Identifier: GPL-2.0-only
+#
+# SPRD SoC drivers
+#
+
+menu "Spreadtrum SoC drivers"
+	depends on ARCH_SPRD || COMPILE_TEST
+
+config SPRD_SYSCON
+	tristate "Spreadtrum syscon support"
+	depends on ARCH_SPRD || COMPILE_TEST
+	help
+	  Say yes here to add support for the Spreadtrum syscon driver,
+	  which is used to implement the atomic method of bits updating.
+
+endmenu
drivers/soc/sprd/Makefile new file +2 -0
new file mode 100644
index 000000000000..4d7715553cf6
@@ -0,0 +1,2 @@
+# SPDX-License-Identifier: GPL-2.0-only
+obj-$(CONFIG_SPRD_SYSCON) += sprd_syscon.o
drivers/soc/sprd/sprd_syscon.c new file +82 -0

82 lines not on this page · raw message

Signature · 1 line
2.17.1
maintainer
#8 lore raw

Re: [PATCH 2/2] soc: sprd: Add Spreadtrum special bits updating support

Quoted 6 lines · Baolin Wang

+ *
+ * Note: there is a potential risk when users want to set and clear bits
+ * at the same time, since the set/clear method will always do bits setting
+ * before bits clearing, which may cause some unexpected results if the
+ * operation sequence is strict. Thus we recommend that do not set and
+ * clear bits at the same time if you are not sure the results.

Would it make sense to have a

WARN_ONCE(set && clk, "%s: non-atomic update", __func__);

in the code to check for this?

drivers/soc/sprd/sprd_syscon.c line 78

Quoted 7 lines
drivers/soc/sprd/sprd_syscon.c new file +7 -0
+static int sprd_syscon_init(void)
+{
+       syscon_register_phys_regmap_bus(&sprd_syscon_regmap);
+
+       return 0;
+}
+core_initcall_sync(sprd_syscon_init);

This no longer breaks at runtime based on the changes in the other patch, but I still don't like how you have to manually load this module on spreadtrum platforms.

What I meant to suggest in my previous reply was to add the regmap_bus instance into drivers/mfd/syscon.c itself.

Arnd

author maintainer
#9 lore raw

Re: [PATCH 2/2] soc: sprd: Add Spreadtrum special bits updating support

Quoted 9 lines · Arnd Bergmann
Quoted 6 lines · Baolin Wang

+ *
+ * Note: there is a potential risk when users want to set and clear bits
+ * at the same time, since the set/clear method will always do bits setting
+ * before bits clearing, which may cause some unexpected results if the
+ * operation sequence is strict. Thus we recommend that do not set and
+ * clear bits at the same time if you are not sure the results.

Quoted 5 lines

Would it make sense to have a

WARN_ONCE(set && clk, "%s: non-atomic update", __func__);

in the code to check for this?

Yes, will add.

Quoted 13 lines

drivers/soc/sprd/sprd_syscon.c line 78

Quoted 8 lines
drivers/soc/sprd/sprd_syscon.c new file +7 -0
+static int sprd_syscon_init(void)
+{
+       syscon_register_phys_regmap_bus(&sprd_syscon_regmap);
+
+       return 0;
+}
+core_initcall_sync(sprd_syscon_init);

This no longer breaks at runtime based on the changes in the other patch, but I still don't like how you have to manually load this module on spreadtrum platforms.

Quoted 2 lines

What I meant to suggest in my previous reply was to add the regmap_bus instance into drivers/mfd/syscon.c itself.

Sorry, I misunderstood your meaning before, but what you suggested will add some vendor-specific things into the common syscon driver, if other platforms have different update bits method, we will add another vendor-specific regmap bus into the syscon.c, which will make syscon.c more complicated.

But if you still prefer to add these vendor-specific things into the syscon.c, I will follow your suggestion in next version. Thanks.

Signature · 1 line
Baolin Wang
maintainer
#10 lore raw

Re: [PATCH 2/2] soc: sprd: Add Spreadtrum special bits updating support

Quoted 14 lines · Baolin Wang
Quoted 13 lines · Arnd Bergmann

drivers/soc/sprd/sprd_syscon.c line 78

Quoted 7 lines · Baolin Wang
drivers/soc/sprd/sprd_syscon.c new file +7 -0
+static int sprd_syscon_init(void)
+{
+       syscon_register_phys_regmap_bus(&sprd_syscon_regmap);
+
+       return 0;
+}
+core_initcall_sync(sprd_syscon_init);

This no longer breaks at runtime based on the changes in the other patch, but I still don't like how you have to manually load this module on spreadtrum platforms.

Quoted 8 lines
Quoted 2 lines

What I meant to suggest in my previous reply was to add the regmap_bus instance into drivers/mfd/syscon.c itself.

Sorry, I misunderstood your meaning before, but what you suggested will add some vendor-specific things into the common syscon driver, if other platforms have different update bits method, we will add another vendor-specific regmap bus into the syscon.c, which will make syscon.c more complicated.

I think we can always deal with this once it gets too complex, as long as the DT binding allows it.

Quoted 2 lines

But if you still prefer to add these vendor-specific things into the syscon.c, I will follow your suggestion in next version. Thanks.

Yes, please do.

Arnd

author maintainer
#11 lore raw

Re: [PATCH 2/2] soc: sprd: Add Spreadtrum special bits updating support

Quoted 25 lines · Arnd Bergmann
Quoted 14 lines · Baolin Wang
Quoted 13 lines · Arnd Bergmann

drivers/soc/sprd/sprd_syscon.c line 78

Quoted 7 lines · Baolin Wang
drivers/soc/sprd/sprd_syscon.c new file +7 -0
+static int sprd_syscon_init(void)
+{
+       syscon_register_phys_regmap_bus(&sprd_syscon_regmap);
+
+       return 0;
+}
+core_initcall_sync(sprd_syscon_init);

This no longer breaks at runtime based on the changes in the other patch, but I still don't like how you have to manually load this module on spreadtrum platforms.

Quoted 8 lines
Quoted 2 lines

What I meant to suggest in my previous reply was to add the regmap_bus instance into drivers/mfd/syscon.c itself.

Sorry, I misunderstood your meaning before, but what you suggested will add some vendor-specific things into the common syscon driver, if other platforms have different update bits method, we will add another vendor-specific regmap bus into the syscon.c, which will make syscon.c more complicated.

Quoted 7 lines

I think we can always deal with this once it gets too complex, as long as the DT binding allows it.

Quoted 2 lines

But if you still prefer to add these vendor-specific things into the syscon.c, I will follow your suggestion in next version. Thanks.

Yes, please do.

Sure. Thanks.

Signature · 1 line
Baolin Wang