Skip to contentSkip to conversation

ARM: mvebu: minor improvements

PATCH New Thomas Petazzoni 4 participants 11 messages Dec 23, 2013 – Dec 25, 2013

Subsystems (4)
lore
Share & tools
lore
#0 lore raw

cover ARM: mvebu: minor improvements

Jason, Gregory, Sebastian, Andrew,

Here are three patches making minor improvements to the mvebu code base.

* The first one moves an 370/XP specific declaration to armada-370-xp.h, in preparation for the future addition of the support for other SOCs in mach-mvebu/. It should go to mvebu/soc.

* The second one fixes the length of the registers of the Armada XP PMSU. It should go to mvebu/dt. There is no need to push this as a fix, as it *reduces* the length of the register area, so it is really only a cosmetic improvement.

* The third one makes a function static. It should to to mvebu/soc.

Thanks!

Thomas

Thomas Petazzoni (3):
ARM: mvebu: move ARMADA_XP_MAX_CPUS to armada-370-xp.h
ARM: mvebu: fix register length for Armada XP PMSU
ARM: mvebu: make armada_370_xp_pmsu_init() static

 arch/arm/boot/dts/armada-xp.dtsi    | 2 +-
 arch/arm/mach-mvebu/armada-370-xp.h | 2 ++
 arch/arm/mach-mvebu/common.h        | 3 ---
 arch/arm/mach-mvebu/pmsu.c          | 2 +-
 4 files changed, 4 insertions(+), 5 deletions(-)
Signature · 1 line
1.8.3.2
#1 lore raw

PATCH 1/3 ARM: mvebu: move ARMADA_XP_MAX_CPUS to armada-370-xp.h

The ARMADA_XP_MAX_CPUS definition was in common.h, which as its name says, is common to all mvebu SoCs. It is more logical to have this XP specific definition in the already existing armada-370-xp.h header file, especially in preparation to the addition of the support for other SOCs in mach-mvebu.

---
 arch/arm/mach-mvebu/armada-370-xp.h | 2 ++
 arch/arm/mach-mvebu/common.h        | 2 --
 2 files changed, 2 insertions(+), 2 deletions(-)
arch/arm/mach-mvebu/armada-370-xp.h +2 -0
index c612b2c..4ed534d 100644
@@ -18,6 +18,8 @@
 #ifdef CONFIG_SMP
 #include <linux/cpumask.h>
 
+#define ARMADA_XP_MAX_CPUS 4
+
 void armada_mpic_send_doorbell(const struct cpumask *mask, unsigned int irq);
 void armada_xp_mpic_smp_cpu_init(void);
 #endif
arch/arm/mach-mvebu/common.h +0 -2
index e366010..f7ad45e 100644
@@ -15,8 +15,6 @@
 #ifndef __ARCH_MVEBU_COMMON_H
 #define __ARCH_MVEBU_COMMON_H
 
-#define ARMADA_XP_MAX_CPUS 4
-
 #include <linux/reboot.h>
 
 void mvebu_restart(enum reboot_mode mode, const char *cmd);
Signature · 1 line
1.8.3.2
#2 lore raw

PATCH 1/3 ARM: mvebu: move ARMADA_XP_MAX_CPUS to armada-370-xp.h

Hi Thomas,

Quoted 7 lines · Thomas Petazzoni

The ARMADA_XP_MAX_CPUS definition was in common.h, which as its name says, is common to all mvebu SoCs. It is more logical to have this XP specific definition in the already existing armada-370-xp.h header file, especially in preparation to the addition of the support for other SOCs in mach-mvebu.

Thanks,

Gregory

Quoted 32 lines
---
 arch/arm/mach-mvebu/armada-370-xp.h | 2 ++
 arch/arm/mach-mvebu/common.h        | 2 --
 2 files changed, 2 insertions(+), 2 deletions(-)
arch/arm/mach-mvebu/armada-370-xp.h +2 -0
index c612b2c..4ed534d 100644
@@ -18,6 +18,8 @@
 #ifdef CONFIG_SMP
 #include <linux/cpumask.h>
 
+#define ARMADA_XP_MAX_CPUS 4
+
 void armada_mpic_send_doorbell(const struct cpumask *mask, unsigned int irq);
 void armada_xp_mpic_smp_cpu_init(void);
 #endif
arch/arm/mach-mvebu/common.h +0 -2
index e366010..f7ad45e 100644
@@ -15,8 +15,6 @@
 #ifndef __ARCH_MVEBU_COMMON_H
 #define __ARCH_MVEBU_COMMON_H
 
-#define ARMADA_XP_MAX_CPUS 4
-
 #include <linux/reboot.h>
 
 void mvebu_restart(enum reboot_mode mode, const char *cmd);
Signature · 4 lines
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
Applied
#3 lore raw

PATCH 1/3 ARM: mvebu: move ARMADA_XP_MAX_CPUS to armada-370-xp.h

In reply to Thomas Petazzoni
Quoted 11 lines · Thomas Petazzoni

The ARMADA_XP_MAX_CPUS definition was in common.h, which as its name says, is common to all mvebu SoCs. It is more logical to have this XP specific definition in the already existing armada-370-xp.h header file, especially in preparation to the addition of the support for other SOCs in mach-mvebu.

---
 arch/arm/mach-mvebu/armada-370-xp.h | 2 ++
 arch/arm/mach-mvebu/common.h        | 2 --
 2 files changed, 2 insertions(+), 2 deletions(-)

Applied to mvebu/soc

thx,

Jason.

#4 lore raw

PATCH 2/3 ARM: mvebu: fix register length for Armada XP PMSU

In reply to Thomas Petazzoni

The per-CPU PMSU registers documented in the datasheet start at 0x22100 and the last register for CPU3 is at 0x22428. However, the DT informations use <0x22100 0x430>, which makes the region end at 0x22530 and not 0x22430.

Moreover, looking at the datasheet, we can see that the registers for CPU0 start at 0x22100, for CPU1 at 0x22200, for CPU2 at 0x22300 and for CPU3 at 0x22400. It seems clear that 0x100 bytes of registers have been used per CPU.

Therefore, this commit reduces the length of the PMSU per-CPU register area from the incorrect 0x430 bytes to a more logical 0x400 bytes.

---
 arch/arm/boot/dts/armada-xp.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/boot/dts/armada-xp.dtsi +1 -1
index 281c644..aa49f17 100644
@@ -48,7 +48,7 @@
 
 			armada-370-xp-pmsu at 22000 {
 				compatible = "marvell,armada-370-xp-pmsu";
-				reg = <0x22100 0x430>, <0x20800 0x20>;
+				reg = <0x22100 0x400>, <0x20800 0x20>;
 			};
 
 			serial at 12200 {
Signature · 1 line
1.8.3.2
#5 lore raw

PATCH 2/3 ARM: mvebu: fix register length for Armada XP PMSU

Hi Thomas, Jason, Sebastian, Andrew,

Quoted 10 lines · Thomas Petazzoni

The per-CPU PMSU registers documented in the datasheet start at 0x22100 and the last register for CPU3 is at 0x22428. However, the DT informations use <0x22100 0x430>, which makes the region end at 0x22530 and not 0x22430.

Moreover, looking at the datasheet, we can see that the registers for CPU0 start at 0x22100, for CPU1 at 0x22200, for CPU2 at 0x22300 and for CPU3 at 0x22400. It seems clear that 0x100 bytes of registers have been used per CPU.

Quoted 4 lines

Therefore, this commit reduces the length of the PMSU per-CPU register area from the incorrect 0x430 bytes to a more logical 0x400 bytes.

Thanks,

Gregory

Quoted 18 lines
---
 arch/arm/boot/dts/armada-xp.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/arm/boot/dts/armada-xp.dtsi +1 -1
index 281c644..aa49f17 100644
@@ -48,7 +48,7 @@
 
 			armada-370-xp-pmsu at 22000 {
 				compatible = "marvell,armada-370-xp-pmsu";
-				reg = <0x22100 0x430>, <0x20800 0x20>;
+				reg = <0x22100 0x400>, <0x20800 0x20>;
 			};
 
 			serial at 12200 {
Signature · 4 lines
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
Applied
#6 lore raw

PATCH 2/3 ARM: mvebu: fix register length for Armada XP PMSU

In reply to Thomas Petazzoni
Quoted 10 lines · Thomas Petazzoni

The per-CPU PMSU registers documented in the datasheet start at 0x22100 and the last register for CPU3 is at 0x22428. However, the DT informations use <0x22100 0x430>, which makes the region end at 0x22530 and not 0x22430.

Moreover, looking at the datasheet, we can see that the registers for CPU0 start at 0x22100, for CPU1 at 0x22200, for CPU2 at 0x22300 and for CPU3 at 0x22400. It seems clear that 0x100 bytes of registers have been used per CPU.

Quoted 7 lines

Therefore, this commit reduces the length of the PMSU per-CPU register area from the incorrect 0x430 bytes to a more logical 0x400 bytes.

---
 arch/arm/boot/dts/armada-xp.dtsi | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

Applied to mvebu/dt after a little reshuffling due to the dtsi node sorting.

thx,

Jason.

#7 lore raw

PATCH 3/3 ARM: mvebu: make armada_370_xp_pmsu_init() static

In reply to Thomas Petazzoni

The armada_370_xp_pmsu_init() function is called as an early_initcall(). Therefore, there is no need to export this function, and we can make it static.

---
 arch/arm/mach-mvebu/common.h | 1 -
 arch/arm/mach-mvebu/pmsu.c   | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)
arch/arm/mach-mvebu/common.h +0 -1
index f7ad45e..8de509a 100644
@@ -24,7 +24,6 @@ void armada_370_xp_handle_irq(struct pt_regs *regs);
 
 void armada_xp_cpu_die(unsigned int cpu);
 int armada_370_xp_coherency_init(void);
-int armada_370_xp_pmsu_init(void);
 void armada_xp_secondary_startup(void);
 extern struct smp_operations armada_xp_smp_ops;
 #endif
arch/arm/mach-mvebu/pmsu.c +1 -1
index 27fc4f0..f614727 100644
@@ -58,7 +58,7 @@ int armada_xp_boot_cpu(unsigned int cpu_id, void *boot_addr)
 }
 #endif
 
-int __init armada_370_xp_pmsu_init(void)
+static int __init armada_370_xp_pmsu_init(void)
 {
 	struct device_node *np;
 
Signature · 1 line
1.8.3.2
#8 lore raw

PATCH 3/3 ARM: mvebu: make armada_370_xp_pmsu_init() static

Hi Thomas,

Quoted 5 lines · Thomas Petazzoni

The armada_370_xp_pmsu_init() function is called as an early_initcall(). Therefore, there is no need to export this function, and we can make it static.

Thanks,

Gregory

Quoted 31 lines
---
 arch/arm/mach-mvebu/common.h | 1 -
 arch/arm/mach-mvebu/pmsu.c   | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)
arch/arm/mach-mvebu/common.h +0 -1
index f7ad45e..8de509a 100644
@@ -24,7 +24,6 @@ void armada_370_xp_handle_irq(struct pt_regs *regs);
 
 void armada_xp_cpu_die(unsigned int cpu);
 int armada_370_xp_coherency_init(void);
-int armada_370_xp_pmsu_init(void);
 void armada_xp_secondary_startup(void);
 extern struct smp_operations armada_xp_smp_ops;
 #endif
arch/arm/mach-mvebu/pmsu.c +1 -1
index 27fc4f0..f614727 100644
@@ -58,7 +58,7 @@ int armada_xp_boot_cpu(unsigned int cpu_id, void *boot_addr)
 }
 #endif
 
-int __init armada_370_xp_pmsu_init(void)
+static int __init armada_370_xp_pmsu_init(void)
 {
 	struct device_node *np;
 
Signature · 4 lines
Gregory Clement, Free Electrons
Kernel, drivers, real-time and embedded Linux
development, consulting, training and support.
http://free-electrons.com
#9 lore raw

PATCH 3/3 ARM: mvebu: make armada_370_xp_pmsu_init() static

In reply to Thomas Petazzoni

Thomas,

Quoted 9 lines · Thomas Petazzoni

The armada_370_xp_pmsu_init() function is called as an early_initcall(). Therefore, there is no need to export this function, and we can make it static.

---
 arch/arm/mach-mvebu/common.h | 1 -
 arch/arm/mach-mvebu/pmsu.c   | 2 +-
 2 files changed, 1 insertion(+), 2 deletions(-)

Thanks, but this was already cleaned up up by:

b12634e343be ARM: mvebu: fix some sparse warnings

in mvebu/soc (it was actually HEAD up until a few moments ago ;-)

Ho, ho, ho,

Jason.

maintainer
#10 lore raw

cover ARM: mvebu: minor improvements

In reply to Thomas Petazzoni
Quoted 9 lines · Thomas Petazzoni

Jason, Gregory, Sebastian, Andrew,

Here are three patches making minor improvements to the mvebu code base.

* The first one moves an 370/XP specific declaration to armada-370-xp.h, in preparation for the future addition of the support for other SOCs in mach-mvebu/. It should go to mvebu/soc.

Quoted 8 lines

* The second one fixes the length of the registers of the Armada XP PMSU. It should go to mvebu/dt. There is no need to push this as a fix, as it *reduces* the length of the register area, so it is really only a cosmetic improvement.

* The third one makes a function static. It should to to mvebu/soc.

Thanks!

Hi Thomas

All look sensible to me,

Andrew