Skip to contentSkip to conversation

s3c-hsudc powerdomain using generic power domains

PATCH New Heiko Stübner 3 participants 19 messages Jan 8, 2012 – Feb 1, 2012

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

cover s3c-hsudc powerdomain using generic power domains

The S3C2443/S3C2416/S3C2450 contain a power domain for the usb phy. Up until now the s3c-hsudc was writing directly to the PWRCFG register in the (un)init_phy functions to control it.

This patchset introduces a power domain for the usbphy and hooks the hsudc-device to it. The new runtime_pm calls then handle the enabling and disabling of the power domain on udc start and stop.

As a result another arch dependency is gone from the driver.

This set applies cleanly to linux-next from 2012-01-06 and is tested on S3C2416 hardware.

Heiko Stuebner (3):
s3c-hsudc: Use helper functions instead of generic container_of
s3c-hsudc: add basic runtime_pm calls
S3C2443: add power domain for usb phy

 arch/arm/mach-s3c2416/Kconfig             |    1 +
 arch/arm/mach-s3c2443/Kconfig             |    1 +
 arch/arm/plat-s3c24xx/Kconfig             |    7 +++
 arch/arm/plat-s3c24xx/Makefile            |    1 +
 arch/arm/plat-s3c24xx/s3c2443-pm-common.c |   65 +++++++++++++++++++++++++++++
 drivers/usb/gadget/s3c-hsudc.c            |   25 ++++++-----
 6 files changed, 89 insertions(+), 11 deletions(-)
 create mode 100644 arch/arm/plat-s3c24xx/s3c2443-pm-common.c
Signature · 1 line
1.7.2.3
#1 lore raw

PATCH 1/3 s3c-hsudc: Use helper functions instead of generic container_of

The helper functions were definied but never used until now.

---
 drivers/usb/gadget/s3c-hsudc.c |   10 +++++-----
 1 files changed, 5 insertions(+), 5 deletions(-)
drivers/usb/gadget/s3c-hsudc.c +5 -5
index df8661d..41b3a58 100644
@@ -759,7 +759,7 @@ static int s3c_hsudc_ep_enable(struct usb_ep *_ep,
 	unsigned long flags;
 	u32 ecr = 0;
 
-	hsep = container_of(_ep, struct s3c_hsudc_ep, ep);
+	hsep = our_ep(_ep);
 	if (!_ep || !desc || hsep->desc || _ep->name == ep0name
 		|| desc->bDescriptorType != USB_DT_ENDPOINT
 		|| hsep->bEndpointAddress != desc->bEndpointAddress
@@ -853,7 +853,7 @@ static void s3c_hsudc_free_request(struct usb_ep *ep, struct usb_request *_req)
 {
 	struct s3c_hsudc_req *hsreq;
 
-	hsreq = container_of(_req, struct s3c_hsudc_req, req);
+	hsreq = our_req(_req);
 	WARN_ON(!list_empty(&hsreq->queue));
 	kfree(hsreq);
 }
@@ -876,12 +876,12 @@ static int s3c_hsudc_queue(struct usb_ep *_ep, struct usb_request *_req,
 	u32 offset;
 	u32 csr;
 
-	hsreq = container_of(_req, struct s3c_hsudc_req, req);
+	hsreq = our_req(_req);
 	if ((!_req || !_req->complete || !_req->buf ||
 		!list_empty(&hsreq->queue)))
 		return -EINVAL;
 
-	hsep = container_of(_ep, struct s3c_hsudc_ep, ep);
+	hsep = our_ep(_ep);
 	hsudc = hsep->dev;
 	if (!hsudc->driver || hsudc->gadget.speed == USB_SPEED_UNKNOWN)
 		return -ESHUTDOWN;
@@ -935,7 +935,7 @@ static int s3c_hsudc_dequeue(struct usb_ep *_ep, struct usb_request *_req)
 	struct s3c_hsudc_req *hsreq;
 	unsigned long flags;
 
-	hsep = container_of(_ep, struct s3c_hsudc_ep, ep);
+	hsep = our_ep(_ep);
 	if (!_ep || hsep->ep.name == ep0name)
 		return -EINVAL;
 
Signature · 1 line
1.7.2.3
#3 lore raw

PATCH 2/3 s3c-hsudc: add basic runtime_pm calls

In reply to Heiko Stübner

This will enable the system to check for activity of the usb gadget and also in a later patch to control the usbphy power-domain. When handling the power domain there, it will be possible to remove another reference to architecture code.

---
 drivers/usb/gadget/s3c-hsudc.c |    9 +++++++++
 1 files changed, 9 insertions(+), 0 deletions(-)
drivers/usb/gadget/s3c-hsudc.c +9 -0
index 41b3a58..5e87293 100644
@@ -30,6 +30,7 @@
 #include <linux/prefetch.h>
 #include <linux/platform_data/s3c-hsudc.h>
 #include <linux/regulator/consumer.h>
+#include <linux/pm_runtime.h>
 
 #include <mach/regs-s3c2443-clock.h>
 
@@ -1178,6 +1179,9 @@ static int s3c_hsudc_start(struct usb_gadget *gadget,
 	dev_info(hsudc->dev, "bound driver %s\n", driver->driver.name);
 
 	s3c_hsudc_reconfig(hsudc);
+
+	pm_runtime_get_sync(hsudc->dev);
+
 	s3c_hsudc_init_phy();
 	if (hsudc->pd->gpio_init)
 		hsudc->pd->gpio_init();
@@ -1208,6 +1212,9 @@ static int s3c_hsudc_stop(struct usb_gadget *gadget,
 	hsudc->gadget.dev.driver = NULL;
 	hsudc->gadget.speed = USB_SPEED_UNKNOWN;
 	s3c_hsudc_uninit_phy();
+
+	pm_runtime_put(hsudc->dev);
+
 	if (hsudc->pd->gpio_uninit)
 		hsudc->pd->gpio_uninit();
 	s3c_hsudc_stop_activity(hsudc);
@@ -1362,6 +1369,8 @@ static int __devinit s3c_hsudc_probe(struct platform_device *pdev)
 	if (ret)
 		goto err_add_udc;
 
+	pm_runtime_enable(dev);
+
 	return 0;
 err_add_udc:
 	device_unregister(&hsudc->gadget.dev);
Signature · 1 line
1.7.2.3
#5 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

In reply to Heiko Stübner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

---
 arch/arm/mach-s3c2416/Kconfig             |    1 +
 arch/arm/mach-s3c2443/Kconfig             |    1 +
 arch/arm/plat-s3c24xx/Kconfig             |    7 +++
 arch/arm/plat-s3c24xx/Makefile            |    1 +
 arch/arm/plat-s3c24xx/s3c2443-pm-common.c |   65 +++++++++++++++++++++++++++++
 drivers/usb/gadget/s3c-hsudc.c            |    6 ---
 6 files changed, 75 insertions(+), 6 deletions(-)
 create mode 100644 arch/arm/plat-s3c24xx/s3c2443-pm-common.c
arch/arm/mach-s3c2416/Kconfig +1 -0
index 84c7b03..240aef7 100644
@@ -15,6 +15,7 @@ config CPU_S3C2416
 	select CPU_LLSERIAL_S3C2440
 	select SAMSUNG_CLKSRC
 	select S3C2443_CLOCK
+	select S3C2443_PM_COMMON
 	help
 	  Support for the S3C2416 SoC from the S3C24XX line
 
arch/arm/mach-s3c2443/Kconfig +1 -0
index 8814031..3bff23f 100644
@@ -10,6 +10,7 @@ config CPU_S3C2443
 	select CPU_LLSERIAL_S3C2440
 	select SAMSUNG_CLKSRC
 	select S3C2443_CLOCK
+	select S3C2443_PM_COMMON
 	help
 	  Support for the S3C2443 SoC from the S3C24XX line
 
arch/arm/plat-s3c24xx/Kconfig +7 -0
index d8973ac..0a9f37c 100644
@@ -50,6 +50,13 @@ config S3C2443_CLOCK
 	  Clock code for the S3C2443 and similar processors, which includes
 	  the S3C2416 and S3C2450.
 
+config S3C2443_PM_COMMON
+	bool
+	select PM_GENERIC_DOMAINS
+	help
+	  Common power management code for the S3C2443 and similar processors,
+	  which include the S3C2416 and S3C2450.
+
 config S3C24XX_DCLK
 	bool
 	help
arch/arm/plat-s3c24xx/Makefile +1 -0
index b2b0112..300f77c 100644
@@ -29,6 +29,7 @@ obj-$(CONFIG_PM)		+= irq-pm.o
 obj-$(CONFIG_PM)		+= sleep.o
 obj-$(CONFIG_S3C2410_CLOCK)	+= s3c2410-clock.o
 obj-$(CONFIG_S3C2443_CLOCK)	+= s3c2443-clock.o
+obj-$(CONFIG_S3C2443_PM_COMMON)	+= s3c2443-pm-common.o
 obj-$(CONFIG_S3C2410_DMA)	+= dma.o
 obj-$(CONFIG_S3C2410_IOTIMING)	+= s3c2410-iotiming.o
 obj-$(CONFIG_S3C2412_IOTIMING)	+= s3c2412-iotiming.o
arch/arm/plat-s3c24xx/s3c2443-pm-common.c new file +65 -0

65 lines not on this page · raw message

drivers/usb/gadget/s3c-hsudc.c +0 -6
index 5e87293..9a81ad3 100644
@@ -195,9 +195,6 @@ static void s3c_hsudc_init_phy(void)
 {
 	u32 cfg;
 
-	cfg = readl(S3C2443_PWRCFG) | S3C2443_PWRCFG_USBPHY;
-	writel(cfg, S3C2443_PWRCFG);
-
 	cfg = readl(S3C2443_URSTCON);
 	cfg |= (S3C2443_URSTCON_FUNCRST | S3C2443_URSTCON_PHYRST);
 	writel(cfg, S3C2443_URSTCON);
@@ -229,9 +226,6 @@ static void s3c_hsudc_uninit_phy(void)
 {
 	u32 cfg;
 
-	cfg = readl(S3C2443_PWRCFG) & ~S3C2443_PWRCFG_USBPHY;
-	writel(cfg, S3C2443_PWRCFG);
-
 	writel(S3C2443_PHYPWR_FSUSPEND, S3C2443_PHYPWR);
 
 	cfg = readl(S3C2443_UCLKCON) & ~S3C2443_UCLKCON_FUNC_CLKEN;
Signature · 1 line
1.7.2.3
Applied
#2 lore raw

PATCH 1/3 s3c-hsudc: Use helper functions instead of generic container_of

In reply to Heiko Stübner
Quoted 3 lines · Heiko St?bner

The helper functions were definied but never used until now.

applied, thanks

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/b9abfc67/attachment.sig>
Applied
#4 lore raw

PATCH 2/3 s3c-hsudc: add basic runtime_pm calls

In reply to Heiko Stübner
Quoted 6 lines · Heiko St?bner

This will enable the system to check for activity of the usb gadget and also in a later patch to control the usbphy power-domain. When handling the power domain there, it will be possible to remove another reference to architecture code.

applied, thanks

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/675c6db5/attachment.sig>
#6 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

In reply to Heiko Stübner

Hi,

Quoted 7 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that case, it can be carried with s3c's tree:

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/17509f4c/attachment.sig>
#7 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 6 lines

Hi,

Quoted 3 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

Quoted 7 lines
Quoted 4 lines

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

Heiko

#8 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Quoted 12 lines · Heiko St?bner

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 9 lines

Hi,

Quoted 6 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

I believe this doesn't really depend on previous patches (??) In that case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/18c832e2/attachment-0001.sig>
#9 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 15 lines · Heiko St?bner
Quoted 14 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 11 lines

Hi,

Quoted 7 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

I believe this doesn't really depend on previous patches (??) In that

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

ok, I will pester Kukjin Kim more about it to get the needed ack :-) .

If changes are requested to this third patch should I only resubmit this one or all of them again - i.e. will you still keep the first two patches?

Heiko

#10 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Quoted 22 lines · Heiko St?bner

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 21 lines · Heiko St?bner
Quoted 19 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 6 lines

Hi,

Quoted 3 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

Quoted 8 lines
Quoted 4 lines

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

Quoted 6 lines
Quoted 1 line

then I will need proper acks to carry this patch.

ok, I will pester Kukjin Kim more about it to get the needed ack :-) .

If changes are requested to this third patch should I only resubmit this one or all of them again - i.e. will you still keep the first two patches?

first two are already applied so no need to resend them :-)

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/ca419423/attachment.sig>
#11 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

In reply to Felipe Balbi

sorry for mail spam :-)

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 15 lines · Heiko St?bner
Quoted 14 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 11 lines

Hi,

Quoted 7 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

Quoted 7 lines
Quoted 5 lines
Quoted 3 lines

I believe this doesn't really depend on previous patches (??) In that

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

another possibility would be to have patch 2 and 3 get in through the samsung tree, as the majority of the change lies on this side (i.e. introducing a new file in plat-s3c24xx)

Heiko

#12 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Quoted 20 lines · Heiko St?bner

sorry for mail spam :-)

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 17 lines · Heiko St?bner
Quoted 16 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 12 lines

Hi,

Quoted 7 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that

Quoted 8 lines
Quoted 5 lines
Quoted 3 lines
Quoted 1 line

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

another possibility would be to have patch 2 and 3 get in through the samsung tree, as the majority of the change lies on this side (i.e. introducing a new file in plat-s3c24xx)

yet another one, would be to defer patch 3/3 by one merge window ;-)

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/fe17dcad/attachment-0001.sig>
#13 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Am Dienstag, 24. Januar 2012, 11:15:11 schrieb Felipe Balbi:

Quoted 27 lines · Heiko St?bner
Quoted 26 lines

sorry for mail spam :-)

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 23 lines · Heiko St?bner
Quoted 21 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 9 lines

Hi,

Quoted 6 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

Quoted 6 lines
Quoted 1 line

I believe this doesn't really depend on previous patches (??) In that

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

Quoted 7 lines
Quoted 5 lines
Quoted 1 line

then I will need proper acks to carry this patch.

another possibility would be to have patch 2 and 3 get in through the samsung tree, as the majority of the change lies on this side (i.e. introducing a new file in plat-s3c24xx)

yet another one, would be to defer patch 3/3 by one merge window ;-)

also fine by me ;-)

But let's see first what Kgene says.

#14 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Quoted 35 lines · Heiko St?bner

Am Dienstag, 24. Januar 2012, 11:15:11 schrieb Felipe Balbi:

Quoted 33 lines · Heiko St?bner
Quoted 31 lines

sorry for mail spam :-)

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 18 lines · Heiko St?bner
Quoted 17 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 13 lines

Hi,

Quoted 7 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

another possibility would be to have patch 2 and 3 get in through the samsung tree, as the majority of the change lies on this side (i.e. introducing a new file in plat-s3c24xx)

Quoted 5 lines
Quoted 1 line

yet another one, would be to defer patch 3/3 by one merge window ;-)

also fine by me ;-)

But let's see first what Kgene says.

Hi all,

Sorry for join lately on this.

I think we need to implement more pm domain for s3c24xx as well usbphy. So if Felipe is ok, I'd like to handle Heiko's 2nd and 3rd patches in my tree. Of course, I need Felipe's ack before that.
BTW, frankly, I'm not sure we can finish supporting pm domain for all of s3c24xx for v3.4 :)

And, Heiko, I don't know why its name is -pm-common.c, how about just -pm.c like others? Yes, when all of s3c24xx can be supported on that, just s3c24xx-pm.c would be better. But right now, s3c2443-pm.c is ok or we need to think about that again. Because as I said, I'm doing merge s3c24xx directories and it's almost done.
And the form of subject would be 'ARM: S3C24XX: ...', actually there were discussions about that.

Thanks.

Best regards,
Kgene.
--
Kukjin Kim <[email protected]>, Senior Engineer, SW Solution Development Team, Samsung Electronics Co., Ltd.

#15 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Hi,

Quoted 44 lines · Kukjin Kim
Quoted 35 lines · Heiko St?bner

Am Dienstag, 24. Januar 2012, 11:15:11 schrieb Felipe Balbi:

Quoted 33 lines · Heiko St?bner
Quoted 31 lines

sorry for mail spam :-)

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 18 lines · Heiko St?bner
Quoted 17 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 13 lines

Hi,

Quoted 7 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

another possibility would be to have patch 2 and 3 get in through the samsung tree, as the majority of the change lies on this side (i.e. introducing a new file in plat-s3c24xx)

Quoted 5 lines
Quoted 1 line

yet another one, would be to defer patch 3/3 by one merge window ;-)

also fine by me ;-)

But let's see first what Kgene says.

Hi all,

Quoted 7 lines

Sorry for join lately on this.

I think we need to implement more pm domain for s3c24xx as well usbphy. So if Felipe is ok, I'd like to handle Heiko's 2nd and 3rd patches in my tree. Of course, I need Felipe's ack before that. BTW, frankly, I'm not sure we can finish supporting pm domain for all of s3c24xx for v3.4 :)

In that case, how about we take things slow ? I can carry pm-runtime patch (2/3) and by 3.5 or 3.6 you will only need to touch arch/arm/-related code.

If you guarantee that it won't make it in 3.4, I'm ok carrying pm-runtime patch as it won't break anything anyway (I suppose).

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/d9c08cdb/attachment.sig>
#16 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Am Dienstag, 24. Januar 2012, 14:07:56 schrieb Felipe Balbi:

Quoted 55 lines

Hi,

Quoted 44 lines · Kukjin Kim
Quoted 35 lines · Heiko St?bner

Am Dienstag, 24. Januar 2012, 11:15:11 schrieb Felipe Balbi:

Quoted 34 lines · Heiko St?bner
Quoted 32 lines

sorry for mail spam :-)

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 19 lines · Heiko St?bner
Quoted 18 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 14 lines

Hi,

Quoted 8 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

another possibility would be to have patch 2 and 3 get in through the samsung tree, as the majority of the change lies on this side (i.e. introducing a new file in plat-s3c24xx)

Quoted 5 lines
Quoted 1 line

yet another one, would be to defer patch 3/3 by one merge window ;-)

also fine by me ;-)

But let's see first what Kgene says.

Hi all,

Quoted 7 lines

Sorry for join lately on this.

I think we need to implement more pm domain for s3c24xx as well usbphy. So if Felipe is ok, I'd like to handle Heiko's 2nd and 3rd patches in my tree. Of course, I need Felipe's ack before that. BTW, frankly, I'm not sure we can finish supporting pm domain for all of s3c24xx for v3.4 :)

Quoted 6 lines

In that case, how about we take things slow ? I can carry pm-runtime patch (2/3) and by 3.5 or 3.6 you will only need to touch arch/arm/-related code.

If you guarantee that it won't make it in 3.4, I'm ok carrying pm-runtime patch as it won't break anything anyway (I suppose).

Seems reasonable to me. The pm-runtime patch does not break anything - it simply does nothing without either some runtime-pm functions or the power domain work.

So I would second that we handle the power domain stuff after Kgene got the mach directories merged for 3.4, i.e. drop the patch (3/3) for now and I will resubmit it when the time comes.

Heiko

#17 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

Quoted 63 lines · Heiko St?bner

Am Dienstag, 24. Januar 2012, 14:07:56 schrieb Felipe Balbi:

Quoted 55 lines

Hi,

Quoted 44 lines · Kukjin Kim
Quoted 35 lines · Heiko St?bner

Am Dienstag, 24. Januar 2012, 11:15:11 schrieb Felipe Balbi:

Quoted 34 lines · Heiko St?bner
Quoted 32 lines

sorry for mail spam :-)

Am Dienstag, 24. Januar 2012, 10:51:31 schrieb Felipe Balbi:

Quoted 19 lines · Heiko St?bner
Quoted 18 lines

Hi Felipe,

Am Dienstag, 24. Januar 2012, 10:38:32 schrieb Felipe Balbi:

Quoted 14 lines

Hi,

Quoted 8 lines · Heiko St?bner

The phy-power-handling is common to S3C2443/2416/2450, so introduce a s3c2443-pm-common.c to handle this for all of them.

This makes it possible to remove the raw write to the PWRCFG-register from the s3c-hsudc driver.

I believe this doesn't really depend on previous patches (??) In that

Quoted 6 lines
Quoted 4 lines
Quoted 1 line

case, it can be carried with s3c's tree:

it does depend on the previous patch [2/3], as the device power domain is controlled through the runtime-pm subsystem whose calls were added in it.

then I will need proper acks to carry this patch.

another possibility would be to have patch 2 and 3 get in through the samsung tree, as the majority of the change lies on this side (i.e. introducing a new file in plat-s3c24xx)

Quoted 5 lines
Quoted 1 line

yet another one, would be to defer patch 3/3 by one merge window ;-)

also fine by me ;-)

But let's see first what Kgene says.

Hi all,

Quoted 7 lines

Sorry for join lately on this.

I think we need to implement more pm domain for s3c24xx as well usbphy. So if Felipe is ok, I'd like to handle Heiko's 2nd and 3rd patches in my tree. Of course, I need Felipe's ack before that. BTW, frankly, I'm not sure we can finish supporting pm domain for all of s3c24xx for v3.4 :)

Quoted 6 lines

In that case, how about we take things slow ? I can carry pm-runtime patch (2/3) and by 3.5 or 3.6 you will only need to touch arch/arm/-related code.

If you guarantee that it won't make it in 3.4, I'm ok carrying pm-runtime patch as it won't break anything anyway (I suppose).

Quoted 7 lines

Seems reasonable to me. The pm-runtime patch does not break anything - it simply does nothing without either some runtime-pm functions or the power domain work.

So I would second that we handle the power domain stuff after Kgene got the mach directories merged for 3.4, i.e. drop the patch (3/3) for now and I will resubmit it when the time comes.

sounds great to me :-) Thanks guys

Signature · 8 lines
balbi
-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 836 bytes
Desc: Digital signature
URL: <http://lists.infradead.org/pipermail/linux-arm-kernel/attachments/20120124/904bea41/attachment-0001.sig>
#18 lore raw

PATCH 3/3 S3C2443: add power domain for usb phy

In reply to Kukjin Kim

Hi Kgene,

Am Dienstag 24 Januar 2012, 14:01:33 schrieb Kukjin Kim:

Quoted 1 line

Because as I said, I'm doing merge s3c24xx directories and it's almost done.

any news of this or can I help in some way?

I'm asking because I would like to put stuff on top of your merge (for example support for spi-s3c64xx for the S3C2416) for 3.4 .

Thanks
Heiko