ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

PATCH Superseded Tarun Kanti DebBarma 6 participants 17 messages Apr 20, 2012 – Jul 6, 2012

Subsystems (5)

Arm (32-bit) Arm sub-architectures OMAP2+ support OMAP clock framework support +1 more subsystems

lore

Share & tools

lore

Tarun Kanti DebBarma

author

2012-04-20 12:39 UTC #0 lore raw

v2 cover ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

The devm API usage in probe() simplifies error handling operation. Since iclk is not used in the driver it is removed from wherever not needed.
Corrected the timer fck name mis-match between clock44xx_data.c and omap_hwmod_44xx_data.c.
Added omap_hwmod_get_main_clk() API. There is no more need to construct clock names using sprintf() to be used in clk_get() during initialization.

Reference:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
Commit: e816b57a337ea3b755de72bec38c10c864f23015 (Linux 3.4-rc3)

Series is available here for reference:
git://gitorious.org/~tarunkanti/omap-sw-develoment/tarunkantis-linux-omap-dev dmtimer_cleanup_for_3.5

Tested on following platforms:
OMAP5, OMAP4430SDP, OMAP3430SDP, OMAP2430SDP. Could not test on OMAP2420 due to unavailability of board.

v2:
- Use devm_request_and_ioremap() instead of request_mem_region() and ioremap()
- Add omap_hwmod_get_main_clk() API
- Reverted changes of clock names from OMAP2 and OMAP3 platforms

Cc Cousson, Benoit <[email protected]>

Cc Paul Walmsley <[email protected]>

Cc Tony Lindgren <[email protected]>

Cc Kevin Hilman <[email protected]>

Cc Rajendra Nayak <[email protected]>

Cc Santosh Shilimkar <[email protected]>

Tarun Kanti DebBarma (4):
ARM: OMAP: dmtimer: use devm_ API and do some cleanup in probe()
ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API
ARM: OMAP2+: dmtimer: cleanup iclk usage
ARM: OMAP2+: dmtimer: cleanup fclk usage

 arch/arm/mach-omap2/clock44xx_data.c         |   33 ++++++-----------
 arch/arm/mach-omap2/omap_hwmod.c             |   15 ++++++++
 arch/arm/mach-omap2/timer.c                  |   10 +-----
 arch/arm/plat-omap/dmtimer.c                 |   51 ++++++++------------------
 arch/arm/plat-omap/include/plat/dmtimer.h    |    2 +-
 arch/arm/plat-omap/include/plat/omap_hwmod.h |    2 +
 6 files changed, 46 insertions(+), 67 deletions(-)

Tarun Kanti DebBarma

author

2012-04-20 12:39 UTC #1 lore raw

PATCH v2 1/4 ARM: OMAP: dmtimer: use devm_ API and do some cleanup in probe()

Replace the regular kzalloc and ioremap with the devm_ equivalent to simplify error handling. We don't need kree() any more in omap_dm_timer_remove().

Also added *dev* pointer to reference pdev->dev which makes the usage shorter in code.

Cc Cousson, Benoit <[email protected]>

Cc Paul Walmsley <[email protected]>

Cc Tony Lindgren <[email protected]>

Cc Kevin Hilman <[email protected]>

Cc Rajendra Nayak <[email protected]>

Cc Santosh Shilimkar <[email protected]>

Signed-off-by Tarun Kanti DebBarma <[email protected]>

---
 arch/arm/plat-omap/dmtimer.c |   51 +++++++++++++----------------------------
 1 files changed, 16 insertions(+), 35 deletions(-)

arch/arm/plat-omap/dmtimer.c +16 ~~-35~~

index 652139c..8a8768b 100644
@@ -37,7 +37,7 @@
 
 #include <linux/module.h>
 #include <linux/io.h>
-#include <linux/slab.h>
+#include <linux/device.h>
 #include <linux/err.h>
 #include <linux/pm_runtime.h>

@@ -628,49 +628,39 @@ EXPORT_SYMBOL_GPL(omap_dm_timers_active);
  */
 static int __devinit omap_dm_timer_probe(struct platform_device *pdev)
 {
-	int ret;
 	unsigned long flags;
 	struct omap_dm_timer *timer;
-	struct resource *mem, *irq, *ioarea;
+	struct resource *mem, *irq;
+	struct device *dev = &pdev->dev;
 	struct dmtimer_platform_data *pdata = pdev->dev.platform_data;
 
 	if (!pdata) {
-		dev_err(&pdev->dev, "%s: no platform data.\n", __func__);
+		dev_err(dev, "%s: no platform data.\n", __func__);
 		return -ENODEV;
 	}
 
 	irq = platform_get_resource(pdev, IORESOURCE_IRQ, 0);
 	if (unlikely(!irq)) {
-		dev_err(&pdev->dev, "%s: no IRQ resource.\n", __func__);
+		dev_err(dev, "%s: no IRQ resource.\n", __func__);
 		return -ENODEV;
 	}
 
 	mem = platform_get_resource(pdev, IORESOURCE_MEM, 0);
 	if (unlikely(!mem)) {
-		dev_err(&pdev->dev, "%s: no memory resource.\n", __func__);
+		dev_err(dev, "%s: no memory resource.\n", __func__);
 		return -ENODEV;
 	}
 
-	ioarea = request_mem_region(mem->start, resource_size(mem),
-			pdev->name);
-	if (!ioarea) {
-		dev_err(&pdev->dev, "%s: region already claimed.\n", __func__);
-		return -EBUSY;
-	}
-
-	timer = kzalloc(sizeof(struct omap_dm_timer), GFP_KERNEL);
+	timer = devm_kzalloc(dev, sizeof(struct omap_dm_timer), GFP_KERNEL);
 	if (!timer) {
-		dev_err(&pdev->dev, "%s: no memory for omap_dm_timer.\n",
-			__func__);
-		ret = -ENOMEM;
-		goto err_free_ioregion;
+		dev_err(dev, "%s: memory alloc failed!\n", __func__);
+		return  -ENOMEM;
 	}
 
-	timer->io_base = ioremap(mem->start, resource_size(mem));
+	timer->io_base = devm_request_and_ioremap(dev, mem);
 	if (!timer->io_base) {
-		dev_err(&pdev->dev, "%s: ioremap failed.\n", __func__);
-		ret = -ENOMEM;
-		goto err_free_mem;
+		dev_err(dev, "%s: region already claimed.\n", __func__);
+		return -ENOMEM;
 	}
 
 	timer->id = pdev->id;

@@ -682,12 +672,12 @@ static int __devinit omap_dm_timer_probe(struct platform_device *pdev)
 
 	/* Skip pm_runtime_enable for OMAP1 */
 	if (!pdata->needs_manual_reset) {
-		pm_runtime_enable(&pdev->dev);
-		pm_runtime_irq_safe(&pdev->dev);
+		pm_runtime_enable(dev);
+		pm_runtime_irq_safe(dev);
 	}
 
 	if (!timer->reserved) {
-		pm_runtime_get_sync(&pdev->dev);
+		pm_runtime_get_sync(dev);
 		__omap_dm_timer_init_regs(timer);
 		pm_runtime_put(&pdev->dev);
 	}

@@ -697,17 +687,9 @@ static int __devinit omap_dm_timer_probe(struct platform_device *pdev)
 	list_add_tail(&timer->node, &omap_timer_list);
 	spin_unlock_irqrestore(&dm_timer_lock, flags);
 
-	dev_dbg(&pdev->dev, "Device Probed.\n");
+	dev_dbg(dev, "Device Probed.\n");
 
 	return 0;
-
-err_free_mem:
-	kfree(timer);
-
-err_free_ioregion:
-	release_mem_region(mem->start, resource_size(mem));
-
-	return ret;
 }
 
 /**

@@ -728,7 +710,6 @@ static int __devexit omap_dm_timer_remove(struct platform_device *pdev)
 	list_for_each_entry(timer, &omap_timer_list, node)
 		if (timer->pdev->id == pdev->id) {
 			list_del(&timer->node);
-			kfree(timer);
 			ret = 0;
 			break;
 		}

Signature · 1 line

1.7.0.4

Tarun Kanti DebBarma

author

2012-04-20 12:39 UTC #2 lore raw

PATCH v2 2/4 ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API

In reply to Tarun Kanti DebBarma

Add an API to get main clock name associated with a given @oh. This will avoid the need to construct fclk names during early initialization in order to get fclk handle using clk_get().

Cc Cousson, Benoit <[email protected]>

Cc Paul Walmsley <[email protected]>

Cc Tony Lindgren <[email protected]>

Cc Kevin Hilman <[email protected]>

Cc Rajendra Nayak <[email protected]>

Cc Santosh Shilimkar <[email protected]>

Signed-off-by Tarun Kanti DebBarma <[email protected]>

---
 arch/arm/mach-omap2/omap_hwmod.c             |   15 +++++++++++++++
 arch/arm/plat-omap/include/plat/omap_hwmod.h |    2 ++
 2 files changed, 17 insertions(+), 0 deletions(-)

arch/arm/mach-omap2/omap_hwmod.c +15 -0

index 2c27fdb..2c6214d 100644
@@ -2775,3 +2775,18 @@ int omap_hwmod_pad_route_irq(struct omap_hwmod *oh, int pad_idx, int irq_idx)
 
 	return 0;
 }
+
+/**
+ * omap_hwmod_get_main_clk - get pointer to main clock name
+ * @oh: struct omap_hwmod *
+ *
+ * Returns the main clock name assocated with @oh upon success,
+ * or NULL if @oh is NULL.
+ */
+const char *omap_hwmod_get_main_clk(struct omap_hwmod *oh)
+{
+	if (!oh)
+		return NULL;
+
+	return oh->main_clk;
+}

arch/arm/plat-omap/include/plat/omap_hwmod.h +2 -0

index 8070145..f3c0697 100644
@@ -608,6 +608,8 @@ int omap_hwmod_no_setup_reset(struct omap_hwmod *oh);
 
 int omap_hwmod_pad_route_irq(struct omap_hwmod *oh, int pad_idx, int irq_idx);
 
+const char *omap_hwmod_get_main_clk(struct omap_hwmod *oh);
+
 /*
  * Chip variant-specific hwmod init routines - XXX should be converted
  * to use initcalls once the initial boot ordering is straightened out

Signature · 1 line

1.7.0.4

Cousson, Benoit

2012-04-20 15:10 UTC #3 lore raw

PATCH v2 2/4 ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API

Hi Tarun,

Thanks for this series, it will help us cleaning a little bit more the legacy stuff we have inside our data files.

Quoted 4 lines · Tarun Kanti DebBarma

Add an API to get main clock name associated with a given @oh. This will avoid the need to construct fclk names during early initialization in order to get fclk handle using clk_get().

Quoted 8 lines

Cc Cousson, Benoit<[email protected]>
Cc Paul Walmsley<[email protected]>
Cc Tony Lindgren<[email protected]>
Cc Kevin Hilman<[email protected]>
Cc Rajendra Nayak<[email protected]>
Cc Santosh Shilimkar<[email protected]>
Signed-off-by Tarun Kanti DebBarma<[email protected]>

Acked-by Benoit Cousson <[email protected]>

Regards,
Benoit

Quoted 41 lines

---
  arch/arm/mach-omap2/omap_hwmod.c             |   15 +++++++++++++++
  arch/arm/plat-omap/include/plat/omap_hwmod.h |    2 ++
  2 files changed, 17 insertions(+), 0 deletions(-)

arch/arm/mach-omap2/omap_hwmod.c +15 -0

index 2c27fdb..2c6214d 100644
@@ -2775,3 +2775,18 @@ int omap_hwmod_pad_route_irq(struct omap_hwmod *oh, int pad_idx, int irq_idx)
 
  	return 0;
  }
+
+/**
+ * omap_hwmod_get_main_clk - get pointer to main clock name
+ * @oh: struct omap_hwmod *
+ *
+ * Returns the main clock name assocated with @oh upon success,
+ * or NULL if @oh is NULL.
+ */
+const char *omap_hwmod_get_main_clk(struct omap_hwmod *oh)
+{
+	if (!oh)
+		return NULL;
+
+	return oh->main_clk;
+}

arch/arm/plat-omap/include/plat/omap_hwmod.h +2 -0

index 8070145..f3c0697 100644
@@ -608,6 +608,8 @@ int omap_hwmod_no_setup_reset(struct omap_hwmod *oh);
 
  int omap_hwmod_pad_route_irq(struct omap_hwmod *oh, int pad_idx, int irq_idx);
 
+const char *omap_hwmod_get_main_clk(struct omap_hwmod *oh);
+
  /*
   * Chip variant-specific hwmod init routines - XXX should be converted
   * to use initcalls once the initial boot ordering is straightened out

Paul Walmsley

maintainer

2012-06-18 16:23 UTC #4 lore raw

PATCH v2 2/4 ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API

In reply to Tarun Kanti DebBarma

Hi,

Quoted 4 lines · Tarun Kanti DebBarma

Add an API to get main clock name associated with a given @oh. This will avoid the need to construct fclk names during early initialization in order to get fclk handle using clk_get().

Quoted 8 lines

Cc Cousson, Benoit <[email protected]>
Cc Paul Walmsley <[email protected]>
Cc Tony Lindgren <[email protected]>
Cc Kevin Hilman <[email protected]>
Cc Rajendra Nayak <[email protected]>
Cc Santosh Shilimkar <[email protected]>
Signed-off-by Tarun Kanti DebBarma <[email protected]>

Thanks, queued for 3.6 with Beno?t's ack.

- Paul

Paul Walmsley

maintainer

2012-07-04 15:39 UTC #5 lore raw

PATCH v2 2/4 ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API

In reply to Tarun Kanti DebBarma

Quoted 3 lines · Tarun Kanti DebBarma

Add an API to get main clock name associated with a given @oh. This will avoid the need to construct fclk names during early initialization in order to get fclk handle using clk_get().

Thanks, queued.

- Paul

Tarun Kanti DebBarma

author

2012-04-20 12:39 UTC #6 lore raw

PATCH v2 3/4 ARM: OMAP2+: dmtimer: cleanup iclk usage

In reply to Tarun Kanti DebBarma

We do not use iclk anywhere in the dmtimer driver and so removing it. Hence removing the timer iclk entries from OMAP4 clkdev table as well.

Cc Cousson, Benoit <[email protected]>

Cc Paul Walmsley <[email protected]>

Cc Tony Lindgren <[email protected]>

Cc Kevin Hilman <[email protected]>

Cc Rajendra Nayak <[email protected]>

Cc Santosh Shilimkar <[email protected]>

Signed-off-by Tarun Kanti DebBarma <[email protected]>

---
 arch/arm/mach-omap2/clock44xx_data.c      |   11 -----------
 arch/arm/mach-omap2/timer.c               |    7 -------
 arch/arm/plat-omap/include/plat/dmtimer.h |    2 +-
 3 files changed, 1 insertions(+), 19 deletions(-)

arch/arm/mach-omap2/clock44xx_data.c +0 ~~-11~~

index fa6ea65..2172f66 100644
@@ -3355,17 +3355,6 @@ static struct omap_clk omap44xx_clks[] = {
 	CLK(NULL,	"auxclk5_ck",			&auxclk5_ck,	CK_443X),
 	CLK(NULL,	"auxclkreq5_ck",		&auxclkreq5_ck,	CK_443X),
 	CLK(NULL,	"gpmc_ck",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt1_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt2_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt3_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt4_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt5_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt6_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt7_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt8_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt9_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt10_ick",			&dummy_ck,	CK_443X),
-	CLK(NULL,	"gpt11_ick",			&dummy_ck,	CK_443X),
 	CLK("omap_i2c.1",	"ick",				&dummy_ck,	CK_443X),
 	CLK("omap_i2c.2",	"ick",				&dummy_ck,	CK_443X),
 	CLK("omap_i2c.3",	"ick",				&dummy_ck,	CK_443X),

arch/arm/mach-omap2/timer.c +0 -7

index c512bac..5d7a0ee 100644
@@ -169,13 +169,6 @@ static int __init omap_dm_timer_init_one(struct omap_dm_timer *timer,
 	if (IS_ERR(timer->fclk))
 		return -ENODEV;
 
-	sprintf(name, "gpt%d_ick", gptimer_id);
-	timer->iclk = clk_get(NULL, name);
-	if (IS_ERR(timer->iclk)) {
-		clk_put(timer->fclk);
-		return -ENODEV;
-	}
-
 	omap_hwmod_enable(oh);
 
 	sys_timer_reserved |= (1 << (gptimer_id - 1));

arch/arm/plat-omap/include/plat/dmtimer.h +1 -1

index 9418f00..be2b8c4 100644
@@ -259,7 +259,7 @@ struct omap_dm_timer {
 	unsigned long phys_base;
 	int id;
 	int irq;
-	struct clk *iclk, *fclk;
+	struct clk *fclk;
 
 	void __iomem	*io_base;
 	void __iomem	*sys_stat;	/* TISTAT timer status */

Signature · 1 line

1.7.0.4

Kevin Hilman

2012-04-23 17:48 UTC #7 lore raw

PATCH v2 3/4 ARM: OMAP2+: dmtimer: cleanup iclk usage

Quoted 2 lines · Tarun Kanti DebBarma

We do not use iclk anywhere in the dmtimer driver and so removing it. Hence removing the timer iclk entries from OMAP4 clkdev table as well.

What about the OMAP2/3 clkdev tables? Can't those be removed as well?

Comment applies to PATCH 4/4 also.

Kevin

DebBarma, Tarun Kanti

author

2012-04-23 19:36 UTC #8 lore raw

PATCH v2 3/4 ARM: OMAP2+: dmtimer: cleanup iclk usage

Quoted 6 lines · Kevin Hilman

Quoted 2 lines · Tarun Kanti DebBarma
We do not use iclk anywhere in the dmtimer driver and so removing it. Hence removing the timer iclk entries from OMAP4 clkdev table as well.
What about the OMAP2/3 clkdev tables? ?Can't those be removed as well?

No, they are being used. In fact I removed them in the beginning only to discover they are needed.

Quoted 2 lines

Comment applies to PATCH 4/4 also.

You mean removing iclk from clkdev table? That is already done in this patch!! --
Tarun

Quoted 2 lines

Kevin

Kevin Hilman

2012-04-24 00:34 UTC #9 lore raw

PATCH v2 3/4 ARM: OMAP2+: dmtimer: cleanup iclk usage

Quoted 9 lines · Tarun Kanti

Quoted 6 lines · Kevin Hilman
Quoted 2 lines · Tarun Kanti DebBarma
We do not use iclk anywhere in the dmtimer driver and so removing it. Hence removing the timer iclk entries from OMAP4 clkdev table as well.
What about the OMAP2/3 clkdev tables? Can't those be removed as well?
No, they are being used. In fact I removed them in the beginning only to discover they are needed.

Maybe PATCH 0/x could summarize why they're still needed for OMAP2/3.

Quoted 3 lines

Quoted 2 lines
Comment applies to PATCH 4/4 also.
You mean removing iclk from clkdev table? That is already done in this patch!!

I meant why can't OMAP2/3 data be removed in PATCH 4/4 also, which only touches OMAP4.

Kevin
--
To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

Tarun Kanti DebBarma

author

2012-04-20 12:39 UTC #10 lore raw

PATCH v2 4/4 ARM: OMAP2+: dmtimer: cleanup fclk usage

In reply to Tarun Kanti DebBarma

With omap_hwmod_get_main_clk() now available, this can be passed to clk_get() to extract the fclk and thus avoid construction of fclk name. Corrected the timer fck name mis-match between clock44xx_data.c and omap_hwmod_44xx_data.c. For other platforms this is already taken care.

Cc Cousson, Benoit <[email protected]>

Cc Paul Walmsley <[email protected]>

Cc Tony Lindgren <[email protected]>

Cc Kevin Hilman <[email protected]>

Cc Rajendra Nayak <[email protected]>

Cc Santosh Shilimkar <[email protected]>

Signed-off-by Tarun Kanti DebBarma <[email protected]>

---
 arch/arm/mach-omap2/clock44xx_data.c |   22 +++++++++++-----------
 arch/arm/mach-omap2/timer.c          |    3 +--
 2 files changed, 12 insertions(+), 13 deletions(-)

arch/arm/mach-omap2/clock44xx_data.c +11 ~~-11~~

index 2172f66..17f3bcd 100644
@@ -3294,17 +3294,17 @@ static struct omap_clk omap44xx_clks[] = {
 	CLK(NULL,	"smartreflex_core_fck",		&smartreflex_core_fck,	CK_443X),
 	CLK(NULL,	"smartreflex_iva_fck",		&smartreflex_iva_fck,	CK_443X),
 	CLK(NULL,	"smartreflex_mpu_fck",		&smartreflex_mpu_fck,	CK_443X),
-	CLK(NULL,	"gpt1_fck",			&timer1_fck,	CK_443X),
-	CLK(NULL,	"gpt10_fck",			&timer10_fck,	CK_443X),
-	CLK(NULL,	"gpt11_fck",			&timer11_fck,	CK_443X),
-	CLK(NULL,	"gpt2_fck",			&timer2_fck,	CK_443X),
-	CLK(NULL,	"gpt3_fck",			&timer3_fck,	CK_443X),
-	CLK(NULL,	"gpt4_fck",			&timer4_fck,	CK_443X),
-	CLK(NULL,	"gpt5_fck",			&timer5_fck,	CK_443X),
-	CLK(NULL,	"gpt6_fck",			&timer6_fck,	CK_443X),
-	CLK(NULL,	"gpt7_fck",			&timer7_fck,	CK_443X),
-	CLK(NULL,	"gpt8_fck",			&timer8_fck,	CK_443X),
-	CLK(NULL,	"gpt9_fck",			&timer9_fck,	CK_443X),
+	CLK(NULL,	"timer1_fck",			&timer1_fck,	CK_443X),
+	CLK(NULL,	"timer10_fck",			&timer10_fck,	CK_443X),
+	CLK(NULL,	"timer11_fck",			&timer11_fck,	CK_443X),
+	CLK(NULL,	"timer2_fck",			&timer2_fck,	CK_443X),
+	CLK(NULL,	"timer3_fck",			&timer3_fck,	CK_443X),
+	CLK(NULL,	"timer4_fck",			&timer4_fck,	CK_443X),
+	CLK(NULL,	"timer5_fck",			&timer5_fck,	CK_443X),
+	CLK(NULL,	"timer6_fck",			&timer6_fck,	CK_443X),
+	CLK(NULL,	"timer7_fck",			&timer7_fck,	CK_443X),
+	CLK(NULL,	"timer8_fck",			&timer8_fck,	CK_443X),
+	CLK(NULL,	"timer9_fck",			&timer9_fck,	CK_443X),
 	CLK(NULL,	"uart1_fck",			&uart1_fck,	CK_443X),
 	CLK(NULL,	"uart2_fck",			&uart2_fck,	CK_443X),
 	CLK(NULL,	"uart3_fck",			&uart3_fck,	CK_443X),

arch/arm/mach-omap2/timer.c +1 -2

index 5d7a0ee..80c9667 100644
@@ -164,8 +164,7 @@ static int __init omap_dm_timer_init_one(struct omap_dm_timer *timer,
 		return -ENXIO;
 
 	/* After the dmtimer is using hwmod these clocks won't be needed */
-	sprintf(name, "gpt%d_fck", gptimer_id);
-	timer->fclk = clk_get(NULL, name);
+	timer->fclk = clk_get(NULL, omap_hwmod_get_main_clk(oh));
 	if (IS_ERR(timer->fclk))
 		return -ENODEV;

Signature · 1 line

1.7.0.4

DebBarma, Tarun Kanti

author

2012-06-14 10:54 UTC #11 lore raw

v2 cover ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

In reply to Tarun Kanti DebBarma

Quoted 31 lines · Tarun Kanti DebBarma

The devm API usage in probe() simplifies error handling operation. Since iclk is not used in the driver it is removed from wherever not needed.
Corrected the timer fck name mis-match between clock44xx_data.c and omap_hwmod_44xx_data.c.
Added omap_hwmod_get_main_clk() API. There is no more need to construct clock names using sprintf() to be used in clk_get() during initialization.
Reference:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
Commit: e816b57a337ea3b755de72bec38c10c864f23015 (Linux 3.4-rc3)
Series is available here for reference:
git://gitorious.org/~tarunkanti/omap-sw-develoment/tarunkantis-linux-omap-dev dmtimer_cleanup_for_3.5
Tested on following platforms:
OMAP5, OMAP4430SDP, OMAP3430SDP, OMAP2430SDP. Could not test on OMAP2420 due to unavailability of board.
v2:
- Use devm_request_and_ioremap() instead of request_mem_region() and ioremap()
- Add omap_hwmod_get_main_clk() API
- Reverted changes of clock names from OMAP2 and OMAP3 platforms
Cc Cousson, Benoit <[email protected]>
Cc Paul Walmsley <[email protected]>
Cc Tony Lindgren <[email protected]>
Cc Kevin Hilman <[email protected]>
Cc Rajendra Nayak <[email protected]>
Cc Santosh Shilimkar <[email protected]>

Quoted 5 lines

Tarun Kanti DebBarma (4):
?ARM: OMAP: dmtimer: use devm_ API and do some cleanup in probe() ?ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API ?ARM: OMAP2+: dmtimer: cleanup iclk usage
?ARM: OMAP2+: dmtimer: cleanup fclk usage

[ping]
Only one of the patches in the series (ARM: OMAP2+: dmtimer: cleanup iclk usage) has been taken. I just applied the remaining patches on top of Linux 3.5-rc2 and tested on OMAP3 and OMAP4. Can the remaining patches be taken as well? --
Tarun

Quoted 9 lines

?arch/arm/mach-omap2/clock44xx_data.c ? ? ? ? | ? 33 ++++++-----------
?arch/arm/mach-omap2/omap_hwmod.c ? ? ? ? ? ? | ? 15 ++++++++
?arch/arm/mach-omap2/timer.c ? ? ? ? ? ? ? ? ?| ? 10 +-----
?arch/arm/plat-omap/dmtimer.c ? ? ? ? ? ? ? ? | ? 51 ++++++++------------------
?arch/arm/plat-omap/include/plat/dmtimer.h ? ?| ? ?2 +-
?arch/arm/plat-omap/include/plat/omap_hwmod.h | ? ?2 +

?6 files changed, 46 insertions(+), 67 deletions(-)

Shilimkar, Santosh

2012-06-14 10:58 UTC #12 lore raw

v2 cover ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

Tony,

Quoted 38 lines · Tarun Kanti

Quoted 31 lines · Tarun Kanti DebBarma
The devm API usage in probe() simplifies error handling operation. Since iclk is not used in the driver it is removed from wherever not needed.
Corrected the timer fck name mis-match between clock44xx_data.c and omap_hwmod_44xx_data.c.
Added omap_hwmod_get_main_clk() API. There is no more need to construct clock names using sprintf() to be used in clk_get() during initialization.
Reference:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
Commit: e816b57a337ea3b755de72bec38c10c864f23015 (Linux 3.4-rc3)
Series is available here for reference:
git://gitorious.org/~tarunkanti/omap-sw-develoment/tarunkantis-linux-omap-dev dmtimer_cleanup_for_3.5
Tested on following platforms:
OMAP5, OMAP4430SDP, OMAP3430SDP, OMAP2430SDP. Could not test on OMAP2420 due to unavailability of board.
v2:
- Use devm_request_and_ioremap() instead of request_mem_region() and ioremap()
- Add omap_hwmod_get_main_clk() API
- Reverted changes of clock names from OMAP2 and OMAP3 platforms
Cc Cousson, Benoit <[email protected]>
Cc Paul Walmsley <[email protected]>
Cc Tony Lindgren <[email protected]>
Cc Kevin Hilman <[email protected]>
Cc Rajendra Nayak <[email protected]>
Cc Santosh Shilimkar <[email protected]>
Quoted 5 lines
Tarun Kanti DebBarma (4):
?ARM: OMAP: dmtimer: use devm_ API and do some cleanup in probe() ?ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API ?ARM: OMAP2+: dmtimer: cleanup iclk usage
?ARM: OMAP2+: dmtimer: cleanup fclk usage

Quoted 4 lines

[ping]
Only one of the patches in the series (ARM: OMAP2+: dmtimer: cleanup iclk usage) has been taken. I just applied the remaining patches on top of Linux 3.5-rc2 and tested on OMAP3 and OMAP4. Can the remaining patches be taken as well?

Just to let you know that this cleanup is essential for Benoit's OMAP4 hwmod cleanup
and OMAP5 minimal support series.

Regards
santosh

DebBarma, Tarun Kanti

author

2012-07-05 04:02 UTC #13 lore raw

v2 cover ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

Hi Tony,

Quoted 42 lines · Santosh

Tony,
Quoted 38 lines · Tarun Kanti
Quoted 36 lines · Tarun Kanti DebBarma
The devm API usage in probe() simplifies error handling operation. Since iclk is not used in the driver it is removed from wherever not needed.
Corrected the timer fck name mis-match between clock44xx_data.c and omap_hwmod_44xx_data.c.
Added omap_hwmod_get_main_clk() API. There is no more need to construct clock names using sprintf() to be used in clk_get() during initialization.
Reference:
git://git.kernel.org/pub/scm/linux/kernel/git/torvalds/linux-2.6.git
Commit: e816b57a337ea3b755de72bec38c10c864f23015 (Linux 3.4-rc3)
Series is available here for reference:
git://gitorious.org/~tarunkanti/omap-sw-develoment/tarunkantis-linux-omap-dev dmtimer_cleanup_for_3.5
Tested on following platforms:
OMAP5, OMAP4430SDP, OMAP3430SDP, OMAP2430SDP. Could not test on OMAP2420 due to unavailability of board.
v2:
- Use devm_request_and_ioremap() instead of request_mem_region() and ioremap()
- Add omap_hwmod_get_main_clk() API
- Reverted changes of clock names from OMAP2 and OMAP3 platforms
Cc Cousson, Benoit <[email protected]>
Cc Paul Walmsley <[email protected]>
Cc Tony Lindgren <[email protected]>
Cc Kevin Hilman <[email protected]>
Cc Rajendra Nayak <[email protected]>
Cc Santosh Shilimkar <[email protected]>
Tarun Kanti DebBarma (4):
ARM: OMAP: dmtimer: use devm_ API and do some cleanup in probe()
ARM: OMAP2+: hwmod: add omap_hwmod_get_main_clk() API
ARM: OMAP2+: dmtimer: cleanup iclk usage
ARM: OMAP2+: dmtimer: cleanup fclk usage

Quoted 8 lines

Quoted 4 lines
[ping]
Only one of the patches in the series (ARM: OMAP2+: dmtimer: cleanup iclk usage) has been taken. I just applied the remaining patches on top of Linux 3.5-rc2 and tested on OMAP3 and OMAP4. Can the remaining patches be taken as well?
Just to let you know that this cleanup is essential for Benoit's OMAP4 hwmod cleanup
and OMAP5 minimal support series.

This is a gentle reminder in case you have missed somehow!! Let me know if anything else needs to be done. Thanks.
--
Tarun

Tony Lindgren

maintainer

2012-07-05 09:28 UTC #14 lore raw

v2 cover ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

Quoted 8 lines · DebBarma, Tarun Kanti

Quoted 4 lines · Santosh
Just to let you know that this cleanup is essential for Benoit's OMAP4 hwmod cleanup
and OMAP5 minimal support series.
This is a gentle reminder in case you have missed somehow!! Let me know if anything else needs to be done.

Now with the hwmod patch 2/4 pulled into cleanup-part2 from Paul, I've also applied patch 1/4 there. Patches 3 and 4 have still some pending comments from Kevin it seems?

Also at least patch 3 won't apply, so please check Kevin's comments and repost patches 3 and 4 on top of cleanup-part2 at commit 74dd9ec62.

Regards,

Tony

DebBarma, Tarun Kanti

author

2012-07-05 11:41 UTC #15 lore raw

v2 cover ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

Hi Tony,

Quoted 7 lines · Tony Lindgren

Quoted 6 lines · DebBarma, Tarun Kanti
Quoted 4 lines · Santosh
Just to let you know that this cleanup is essential for Benoit's OMAP4 hwmod cleanup
and OMAP5 minimal support series.

Quoted 6 lines

Quoted 2 lines
This is a gentle reminder in case you have missed somehow!! Let me know if anything else needs to be done.
Now with the hwmod patch 2/4 pulled into cleanup-part2 from Paul, I've also applied patch 1/4 there. Patches 3 and 4 have still some pending comments from Kevin it seems?

Yes....

Quoted 3 lines

Also at least patch 3 won't apply, so please check Kevin's comments and repost patches 3 and 4 on top of cleanup-part2 at commit 74dd9ec62.

Patch 3: Comment was about adding summary in patch 0/x why we have not removed iclk in OMAP2/3 platform. In other words there was no specific changes needed on patch 3 as such. BTW, patch 3 is already merged.

Patch 4:
--- [Kevin's comment] ---

Quoted 3 lines

Quoted 2 lines
Comment applies to PATCH 4/4 also.
You mean removing iclk from clkdev table? That is already done in this patch!!

I meant why can't OMAP2/3 data be removed in PATCH 4/4 also, which only touches OMAP4.
--- [end] -----------------------
In patch 4 the only change done in the clock table was to correct the name mis-match
of functional clock names for OMAP4 between clock data and hwmod data. For OMAP2/3 the names are already matching. As the comment was given in patch 3 where iclk entries got removed for OMAP4, he must have expected similar cleanup is done for fclk in patch 4. But as I said, only the name mis-match is corrected.
In summary, there is no specific changes needed in the patch.

I will rebase on top of cleanup-part2 are repost patch 4. --
Tarun

Quoted 4 lines

Regards,
Tony

Tony Lindgren

maintainer

2012-07-06 07:09 UTC #16 lore raw

v2 cover ARM: OMAP2+: dmtimer: cleanup related to devm API and clk usage

Quoted 4 lines · DebBarma, Tarun Kanti

Patch 3: Comment was about adding summary in patch 0/x why we have not removed iclk in OMAP2/3 platform. In other words there was no specific changes needed on patch 3 as such. BTW, patch 3 is already merged.

...

Quoted 1 line

I will rebase on top of cleanup-part2 are repost patch 4.

OK thanks!

Tony