Skip to contentSkip to conversation

vexpress: Initialise the sysregs before setting up the clocks

PATCH New Catalin Marinas 3 participants 5 messages May 1, 2014

Subsystems (3)
Common clk Arm/versatile express platform Arm integrator, versatile and realview support
lore
Share & tools
lore

Catalin Marinas

author
#0 lore raw

PATCH 1/1 vexpress: Initialise the sysregs before setting up the clocks

Following arm64 commit bc3ee18a7a57 (arm64: init: Move of_clk_init to time_init()), vexpress_osc_of_setup() is called via of_clk_init() long before initcalls are issued. Initialising the vexpress oscillators requires the vespress sysregs to be already initialised, so this patch adds an explicit call to vexpress_sysreg_of_early_init() in vexpress oscillator setup function. In addition, vexpress_sysreg_init() is changed to a device_initcall() as the core_initcall() is no longer required.

Signed-off-by Catalin Marinas <[email protected]>
Tested-by Will Deacon <[email protected]>
Acked-by Will Deacon <[email protected]>
Tested-by Pawel Moll <[email protected]>
Acked-by Pawel Moll <[email protected]>
Cc Mike Turquette <[email protected]>

---

Mike, I'd like to get this in for 3.15 and I'm happy to push it via the arm64 tree. Are you ok with this? Thanks.

 drivers/clk/versatile/clk-vexpress-osc.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/clk/versatile/clk-vexpress-osc.c +2 -0
index a535c7bf8574..422391242b39 100644
@@ -100,6 +100,8 @@ void __init vexpress_osc_of_setup(struct device_node *node)
 	struct clk *clk;
 	u32 range[2];
 
+	vexpress_sysreg_of_early_init();
+
 	osc = kzalloc(sizeof(*osc), GFP_KERNEL);
 	if (!osc)
 		return;
#1 lore raw

PATCH 1/1 vexpress: Initialise the sysregs before setting up the clocks

On Thu, 2014-05-01 at 15:05 +0100, Catalin Marinas wrote:
[...]

Quoted 1 line

In addition, vexpress_sysreg_init() is changed to a device_initcall()

Not in this patch it isn't ;-)

Quoted 35 lines

as the core_initcall() is no longer
required.

Signed-off-by Catalin Marinas <[email protected]>
Tested-by Will Deacon <[email protected]>
Acked-by Will Deacon <[email protected]>
Tested-by Pawel Moll <[email protected]>
Acked-by Pawel Moll <[email protected]>
Cc Mike Turquette <[email protected]>

---

Mike, I'd like to get this in for 3.15 and I'm happy to push it via the arm64 tree. Are you ok with this? Thanks.

 drivers/clk/versatile/clk-vexpress-osc.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/clk/versatile/clk-vexpress-osc.c +2 -0
index a535c7bf8574..422391242b39 100644
@@ -100,6 +100,8 @@ void __init vexpress_osc_of_setup(struct device_node *node)
 	struct clk *clk;
 	u32 range[2];
 
+	vexpress_sysreg_of_early_init();
+
 	osc = kzalloc(sizeof(*osc), GFP_KERNEL);
 	if (!osc)
 		return;

_______________________________________________
linux-arm-kernel mailing list
linux-arm-kernel at lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-arm-kernel

author
#2 lore raw

PATCH 1/1 vexpress: Initialise the sysregs before setting up the clocks

Quoted 5 lines · Jon Medhurst (Tixy)

On Thu, 2014-05-01 at 15:05 +0100, Catalin Marinas wrote:
[...]

Quoted 1 line

In addition, vexpress_sysreg_init() is changed to a device_initcall()

Not in this patch it isn't ;-)

Indeed, Pawel pinged me in private already. It was there initially but it's not really justified for an -rc fix.

Signature · 1 line
Catalin
#3 lore raw

PATCH 1/1 vexpress: Initialise the sysregs before setting up the clocks

In reply to Catalin Marinas

Hello.

Quoted 8 lines · Catalin Marinas

Following arm64 commit bc3ee18a7a57 (arm64: init: Move of_clk_init to time_init()), vexpress_osc_of_setup() is called via of_clk_init() long before initcalls are issued. Initialising the vexpress oscillators requires the vespress sysregs to be already initialised, so this patch adds an explicit call to vexpress_sysreg_of_early_init() in vexpress oscillator setup function. In addition, vexpress_sysreg_init() is changed to a device_initcall() as the core_initcall() is no longer required.

I don't see that last change in this patch...

Quoted 19 lines
Signed-off-by Catalin Marinas <[email protected]>
Tested-by Will Deacon <[email protected]>
Acked-by Will Deacon <[email protected]>
Tested-by Pawel Moll <[email protected]>
Acked-by Pawel Moll <[email protected]>
Cc Mike Turquette <[email protected]>

---

Mike, I'd like to get this in for 3.15 and I'm happy to push it via the arm64 tree. Are you ok with this? Thanks.

  drivers/clk/versatile/clk-vexpress-osc.c | 2 ++
  1 file changed, 2 insertions(+)
drivers/clk/versatile/clk-vexpress-osc.c +0 -0
index a535c7bf8574..422391242b39 100644

drivers/clk/versatile/clk-vexpress-osc.c void __init vexpress_osc_of_setup(struct device_node *node) · line 107

Quoted 8 lines
drivers/clk/versatile/clk-vexpress-osc.c +2 -0
@@ -100,6 +100,8 @@ void __init vexpress_osc_of_setup(struct device_node *node)
  	struct clk *clk;
  	u32 range[2];
 
+	vexpress_sysreg_of_early_init();
+
  	osc = kzalloc(sizeof(*osc), GFP_KERNEL);
  	if (!osc)
  		return;

WBR, Sergei

author
#4 lore raw

PATCH 1/1 vexpress: Initialise the sysregs before setting up the clocks

Quoted 11 lines · Sergei Shtylyov
Quoted 8 lines · Catalin Marinas

Following arm64 commit bc3ee18a7a57 (arm64: init: Move of_clk_init to time_init()), vexpress_osc_of_setup() is called via of_clk_init() long before initcalls are issued. Initialising the vexpress oscillators requires the vespress sysregs to be already initialised, so this patch adds an explicit call to vexpress_sysreg_of_early_init() in vexpress oscillator setup function. In addition, vexpress_sysreg_init() is changed to a device_initcall() as the core_initcall() is no longer required.

Quoted 1 line

I don't see that last change in this patch?

Please see my reply to Tixy. I dropped the change but forgot the log.

Catalin