Skip to contentSkip to conversation

i2c-designware-baytrail: describe magic numbers

PATCH In review Andy Shevchenko 3 participants 17 messages Feb 10, 2015 – Mar 7, 2015 7 messages not in the archive

Subsystems (3)
lore
Share & tools
lore
#0 lore raw

PATCH 1/5 i2c-designware-baytrail: describe magic numbers

The patch converts hardcoded numerical constants to a named ones.

While here, align the variable name in get_sem() and reset_semaphore().

---
 drivers/i2c/busses/i2c-designware-baytrail.c | 16 +++++++++-------
 1 file changed, 9 insertions(+), 7 deletions(-)
drivers/i2c/busses/i2c-designware-baytrail.c +9 -7
index 5f1ff4c..e9cb355 100644
@@ -22,22 +22,24 @@
 
 #define SEMAPHORE_TIMEOUT	100
 #define PUNIT_SEMAPHORE		0x7
+#define PUNIT_SEMAPHORE_BIT	BIT(0)
+#define PUNIT_SEMAPHORE_ACQUIRE	BIT(1)
 
 static unsigned long acquired;
 
 static int get_sem(struct device *dev, u32 *sem)
 {
-	u32 reg_val;
+	u32 data;
 	int ret;
 
 	ret = iosf_mbi_read(BT_MBI_UNIT_PMC, BT_MBI_BUNIT_READ, PUNIT_SEMAPHORE,
-			    &reg_val);
+				&data);
 	if (ret) {
 		dev_err(dev, "iosf failed to read punit semaphore\n");
 		return ret;
 	}
 
-	*sem = reg_val & 0x1;
+	*sem = data & PUNIT_SEMAPHORE_BIT;
 
 	return 0;
 }
@@ -52,9 +54,9 @@ static void reset_semaphore(struct device *dev)
 		return;
 	}
 
-	data = data & 0xfffffffe;
+	data &= ~PUNIT_SEMAPHORE_BIT;
 	if (iosf_mbi_write(BT_MBI_UNIT_PMC, BT_MBI_BUNIT_WRITE,
-				 PUNIT_SEMAPHORE, data))
+				PUNIT_SEMAPHORE, data))
 		dev_err(dev, "iosf failed to reset punit semaphore during write\n");
 }
 
@@ -70,9 +72,9 @@ int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	if (!dev->acquire_lock)
 		return 0;
 
-	/* host driver writes 0x2 to side band semaphore register */
+	/* host driver writes to side band semaphore register */
 	ret = iosf_mbi_write(BT_MBI_UNIT_PMC, BT_MBI_BUNIT_WRITE,
-				 PUNIT_SEMAPHORE, 0x2);
+				PUNIT_SEMAPHORE, PUNIT_SEMAPHORE_ACQUIRE);
 	if (ret) {
 		dev_err(dev->dev, "iosf punit semaphore request failed\n");
 		return ret;
Signature · 1 line
2.1.4

A message that is not in the archive (sent off-list or not yet received): 1423587970-19136-1-git-send-email-andriy.shevchenko-VuQAYsv1563Yd54FQh9/[email protected]

#2 lore raw

PATCH 2/5 i2c-designware-baytrail: fix typo in error path

It seems we have same message for different return values in get_sem() and baytrail_i2c_acquire(). I suspect this is just a typo, so this patch fixes it.

---
 drivers/i2c/busses/i2c-designware-baytrail.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/i2c/busses/i2c-designware-baytrail.c +2 -2
index e9cb355..9b67655 100644
@@ -99,8 +99,8 @@ int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	reset_semaphore(dev->dev);
 
 	ret = iosf_mbi_read(BT_MBI_UNIT_PMC, BT_MBI_BUNIT_READ,
-		PUNIT_SEMAPHORE, &sem);
-	if (!ret)
+				PUNIT_SEMAPHORE, &sem);
+	if (ret)
 		dev_err(dev->dev, "iosf failed to read punit semaphore\n");
 	else
 		dev_err(dev->dev, "PUNIT SEM: %d\n", sem);
Signature · 1 line
2.1.4

A message that is not in the archive (sent off-list or not yet received): 1423587970-19136-2-git-send-email-andriy.shevchenko-VuQAYsv1563Yd54FQh9/[email protected]

Acked-by
#4 lore raw

Re: [PATCH v1 2/5] i2c-designware-baytrail: fix typo in error path

Quoted 4 lines · Andy Shevchenko

It seems we have same message for different return values in get_sem() and baytrail_i2c_acquire(). I suspect this is just a typo, so this patch fixes it.

Thanks for spotting this. Timeouts are infrequent and I didn't notice the error here.

Dave

Quoted 22 lines
---
 drivers/i2c/busses/i2c-designware-baytrail.c | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)
drivers/i2c/busses/i2c-designware-baytrail.c +2 -2
index e9cb355..9b67655 100644
@@ -99,8 +99,8 @@ int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	reset_semaphore(dev->dev);
 
 	ret = iosf_mbi_read(BT_MBI_UNIT_PMC, BT_MBI_BUNIT_READ,
-		PUNIT_SEMAPHORE, &sem);
-	if (!ret)
+				PUNIT_SEMAPHORE, &sem);
+	if (ret)
 		dev_err(dev->dev, "iosf failed to read punit semaphore\n");
 	else
 		dev_err(dev->dev, "PUNIT SEM: %d\n", sem);
Signature · 1 line
2.1.4
#5 lore raw

Re: [PATCH v1 2/5] i2c-designware-baytrail: fix typo in error path

In reply to a message not in the archive

drivers/i2c/busses/i2c-designware-baytrail.c int baytrail_i2c_acquire(struct dw_i2c_dev *dev) · line 106

Quoted 6 lines
drivers/i2c/busses/i2c-designware-baytrail.c +2 -0
+				PUNIT_SEMAPHORE, &sem);
+	if (ret)
 		dev_err(dev->dev, "iosf failed to read punit semaphore\n");
 	else
 		dev_err(dev->dev, "PUNIT SEM: %d\n", sem);

Shouldn't the latter be a dev_dbg?

#6 lore raw

Re: [PATCH v1 2/5] i2c-designware-baytrail: fix typo in error path

Quoted 7 lines · Wolfram Sang

drivers/i2c/busses/i2c-designware-baytrail.c int baytrail_i2c_acquire(struct dw_i2c_dev *dev) · line 106

Quoted 5 lines
drivers/i2c/busses/i2c-designware-baytrail.c +2 -0
+				PUNIT_SEMAPHORE, &sem);
+	if (ret)
 		dev_err(dev->dev, "iosf failed to read punit semaphore\n");
 	else
 		dev_err(dev->dev, "PUNIT SEM: %d\n", sem);

Shouldn't the latter be a dev_dbg?

For me it seems not. Here is error patch and we have already in error recovery, so, intention to see if the semaphore becomes alive after reset. Am I right, David?

Signature · 2 lines
Andy Shevchenko <[email protected]>
Intel Finland Oy

A message that is not in the archive (sent off-list or not yet received): 1423673991.31903.535.camel-VuQAYsv1563Yd54FQh9/[email protected]

#8 lore raw

Re: [PATCH v1 2/5] i2c-designware-baytrail: fix typo in error path

Quoted 9 lines · Andy Shevchenko
Quoted 7 lines · Wolfram Sang

drivers/i2c/busses/i2c-designware-baytrail.c int baytrail_i2c_acquire(struct dw_i2c_dev *dev) · line 106

Quoted 5 lines
drivers/i2c/busses/i2c-designware-baytrail.c +2 -0
+				PUNIT_SEMAPHORE, &sem);
+	if (ret)
 		dev_err(dev->dev, "iosf failed to read punit semaphore\n");
 	else
 		dev_err(dev->dev, "PUNIT SEM: %d\n", sem);

Shouldn't the latter be a dev_dbg?

Quoted 3 lines

For me it seems not. Here is error patch and we have already in error recovery, so, intention to see if the semaphore becomes alive after reset. Am I right, David?

Yes. We've timed out by this section of code so we want to verify the semaphore was reset. Failure to read the semaphore though is a separate error as well.

Dave

Quoted 6 lines
Signature · 3 lines
-- 
Andy Shevchenko <[email protected]>
Intel Finland Oy
#9 lore raw

Re: [PATCH v1 2/5] i2c-designware-baytrail: fix typo in error path

In reply to Wolfram Sang
Quoted 7 lines · Wolfram Sang

drivers/i2c/busses/i2c-designware-baytrail.c int baytrail_i2c_acquire(struct dw_i2c_dev *dev) · line 106

Quoted 5 lines
drivers/i2c/busses/i2c-designware-baytrail.c +2 -0
+				PUNIT_SEMAPHORE, &sem);
+	if (ret)
 		dev_err(dev->dev, "iosf failed to read punit semaphore\n");
 	else
 		dev_err(dev->dev, "PUNIT SEM: %d\n", sem);

Shouldn't the latter be a dev_dbg?

Wolfram, do we agree on this as states in the patch? Maybe you have more comments, questions?

Otherwise, should I rebase this on top of 4.0-rc1 or you will be okay with current version?

Signature · 2 lines
Andy Shevchenko <[email protected]>
Intel Finland Oy

A message that is not in the archive (sent off-list or not yet received): 1424696051.14897.13.camel-VuQAYsv1563Yd54FQh9/[email protected]

#11 lore raw

Re: [PATCH v1 2/5] i2c-designware-baytrail: fix typo in error path

Quoted 5 lines
Quoted 2 lines

Shouldn't the latter be a dev_dbg?

Wolfram, do we agree on this as states in the patch? Maybe you have more comments, questions?

Yes, it is fine this way.

Quoted 2 lines

Otherwise, should I rebase this on top of 4.0-rc1 or you will be okay with current version?

Well, if it needs rebasing, please do. If not, please say so, too :)

#12 lore raw

Re: [PATCH v1 2/5] i2c-designware-baytrail: fix typo in error path

Quoted 11 lines · Wolfram Sang
Quoted 4 lines
Quoted 1 line

Shouldn't the latter be a dev_dbg?

Wolfram, do we agree on this as states in the patch? Maybe you have more comments, questions?

Yes, it is fine this way.

Quoted 2 lines

Otherwise, should I rebase this on top of 4.0-rc1 or you will be okay with current version?

Well, if it needs rebasing, please do. If not, please say so, too :)

It applies clearly on top of current linux-next without any changes from my side. So, no rebase is needed.

Signature · 2 lines
Andy Shevchenko <[email protected]>
Intel Finland Oy
#13 lore raw

PATCH 3/5 i2c-designware-baytrail: fix sparse warnings

In reply to a message not in the archive

There is no need to export functions that are used as the callbacks in the struct dw_i2c_dev. Otherwise we get the following warnings:

drivers/i2c/busses/i2c-designware-baytrail.c:63:5: warning: symbol 'baytrail_i2c_acquire' was not declared. Should it be static? drivers/i2c/busses/i2c-designware-baytrail.c:114:6: warning: symbol 'baytrail_i2c_release' was not declared. Should it be static?

While here, do few indentation fixes, remove i2c_dw_eval_lock_support() from functions exported to the modules and redundant assignment of local sem variable.

---
 drivers/i2c/busses/i2c-designware-baytrail.c | 12 +++++-------
 1 file changed, 5 insertions(+), 7 deletions(-)
drivers/i2c/busses/i2c-designware-baytrail.c +5 -7
index 9b67655..d334744 100644
@@ -17,7 +17,9 @@
 #include <linux/acpi.h>
 #include <linux/i2c.h>
 #include <linux/interrupt.h>
+
 #include <asm/iosf_mbi.h>
+
 #include "i2c-designware-core.h"
 
 #define SEMAPHORE_TIMEOUT	100
@@ -60,9 +62,9 @@ static void reset_semaphore(struct device *dev)
 		dev_err(dev, "iosf failed to reset punit semaphore during write\n");
 }
 
-int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
+static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 {
-	u32 sem = 0;
+	u32 sem;
 	int ret;
 	unsigned long start, end;
 
@@ -109,9 +111,8 @@ int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 
 	return -ETIMEDOUT;
 }
-EXPORT_SYMBOL(baytrail_i2c_acquire);
 
-void baytrail_i2c_release(struct dw_i2c_dev *dev)
+static void baytrail_i2c_release(struct dw_i2c_dev *dev)
 {
 	if (!dev || !dev->dev)
 		return;
@@ -123,7 +124,6 @@ void baytrail_i2c_release(struct dw_i2c_dev *dev)
 	dev_dbg(dev->dev, "punit semaphore held for %ums\n",
 		jiffies_to_msecs(jiffies - acquired));
 }
-EXPORT_SYMBOL(baytrail_i2c_release);
 
 int i2c_dw_eval_lock_support(struct dw_i2c_dev *dev)
 {
@@ -139,7 +139,6 @@ int i2c_dw_eval_lock_support(struct dw_i2c_dev *dev)
 		return 0;
 
 	status = acpi_evaluate_integer(handle, "_SEM", NULL, &shared_host);
-
 	if (ACPI_FAILURE(status))
 		return 0;
 
@@ -155,7 +154,6 @@ int i2c_dw_eval_lock_support(struct dw_i2c_dev *dev)
 
 	return 0;
 }
-EXPORT_SYMBOL(i2c_dw_eval_lock_support);
 
 MODULE_AUTHOR("David E. Box <david.e.box-VuQAYsv1563Yd54FQh9/[email protected]>");
 MODULE_DESCRIPTION("Baytrail I2C Semaphore driver");
Signature · 1 line
2.1.4

A message that is not in the archive (sent off-list or not yet received): 1423587970-19136-3-git-send-email-andriy.shevchenko-VuQAYsv1563Yd54FQh9/[email protected]

Acked-by
#15 lore raw

Re: [PATCH v1 3/5] i2c-designware-baytrail: fix sparse warnings

Quoted 11 lines · Andy Shevchenko

There is no need to export functions that are used as the callbacks in the struct dw_i2c_dev. Otherwise we get the following warnings:

drivers/i2c/busses/i2c-designware-baytrail.c:63:5: warning: symbol 'baytrail_i2c_acquire' was not declared. Should it be static? drivers/i2c/busses/i2c-designware-baytrail.c:114:6: warning: symbol 'baytrail_i2c_release' was not declared. Should it be static?

While here, do few indentation fixes, remove i2c_dw_eval_lock_support() from functions exported to the modules and redundant assignment of local sem variable.

Yep. These remained from original versions of the patch that did use them as callbacks in struct dw_i2c_dev. Thanks.

#16 lore raw

PATCH 4/5 i2c-designware-baytrail: cross-check lock functions

In reply to a message not in the archive

It seems the idea behind the cross-check is to prevent acquire semaphore when there is no release callback and vice versa. Thus, patch fixes a typo.

---
 drivers/i2c/busses/i2c-designware-baytrail.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/i2c/busses/i2c-designware-baytrail.c +1 -1
index d334744..036d9bdc0 100644
@@ -71,7 +71,7 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	if (!dev || !dev->dev)
 		return -ENODEV;
 
-	if (!dev->acquire_lock)
+	if (!dev->release_lock)
 		return 0;
 
 	/* host driver writes to side band semaphore register */
Signature · 1 line
2.1.4

A message that is not in the archive (sent off-list or not yet received): 1423587970-19136-4-git-send-email-andriy.shevchenko-VuQAYsv1563Yd54FQh9/[email protected]

Acked-by
#18 lore raw

Re: [PATCH v1 4/5] i2c-designware-baytrail: cross-check lock functions

Quoted 4 lines · Andy Shevchenko

It seems the idea behind the cross-check is to prevent acquire semaphore when there is no release callback and vice versa. Thus, patch fixes a typo.

Quoted 20 lines
---
 drivers/i2c/busses/i2c-designware-baytrail.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/i2c/busses/i2c-designware-baytrail.c +1 -1
index d334744..036d9bdc0 100644
@@ -71,7 +71,7 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	if (!dev || !dev->dev)
 		return -ENODEV;
 
-	if (!dev->acquire_lock)
+	if (!dev->release_lock)
 		return 0;
 
 	/* host driver writes to side band semaphore register */
Signature · 1 line
2.1.4
#19 lore raw

PATCH 5/5 i2c-designware-baytrail: baytrail_i2c_acquire() might sleep

In reply to a message not in the archive

This patch marks baytrail_i2c_acquire() that it might sleep. Also it chages while-loop to do-while and, though it is matter of taste, gives a chance to check one more time before report a timeout.

---
 drivers/i2c/busses/i2c-designware-baytrail.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/i2c/busses/i2c-designware-baytrail.c +4 -2
index 036d9bdc0..7d7ae97 100644
@@ -68,6 +68,8 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	int ret;
 	unsigned long start, end;
 
+	might_sleep();
+
 	if (!dev || !dev->dev)
 		return -ENODEV;
 
@@ -85,7 +87,7 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	/* host driver waits for bit 0 to be set in semaphore register */
 	start = jiffies;
 	end = start + msecs_to_jiffies(SEMAPHORE_TIMEOUT);
-	while (!time_after(jiffies, end)) {
+	do {
 		ret = get_sem(dev->dev, &sem);
 		if (!ret && sem) {
 			acquired = jiffies;
@@ -95,7 +97,7 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 		}
 
 		usleep_range(1000, 2000);
-	}
+	} while (time_before(jiffies, end));
 
 	dev_err(dev->dev, "punit semaphore timed out, resetting\n");
 	reset_semaphore(dev->dev);
Signature · 1 line
2.1.4

A message that is not in the archive (sent off-list or not yet received): 1423587970-19136-5-git-send-email-andriy.shevchenko-VuQAYsv1563Yd54FQh9/[email protected]

Acked-by
#21 lore raw

Re: [PATCH v1 5/5] i2c-designware-baytrail: baytrail_i2c_acquire() might sleep

Quoted 5 lines · Andy Shevchenko

This patch marks baytrail_i2c_acquire() that it might sleep. Also it chages while-loop to do-while and, though it is matter of taste, gives a chance to check one more time before report a timeout.

Quoted 38 lines
---
 drivers/i2c/busses/i2c-designware-baytrail.c | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/i2c/busses/i2c-designware-baytrail.c +4 -2
index 036d9bdc0..7d7ae97 100644
@@ -68,6 +68,8 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	int ret;
 	unsigned long start, end;
 
+	might_sleep();
+
 	if (!dev || !dev->dev)
 		return -ENODEV;
 
@@ -85,7 +87,7 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 	/* host driver waits for bit 0 to be set in semaphore register */
 	start = jiffies;
 	end = start + msecs_to_jiffies(SEMAPHORE_TIMEOUT);
-	while (!time_after(jiffies, end)) {
+	do {
 		ret = get_sem(dev->dev, &sem);
 		if (!ret && sem) {
 			acquired = jiffies;
@@ -95,7 +97,7 @@ static int baytrail_i2c_acquire(struct dw_i2c_dev *dev)
 		}
 
 		usleep_range(1000, 2000);
-	}
+	} while (time_before(jiffies, end));
 
 	dev_err(dev->dev, "punit semaphore timed out, resetting\n");
 	reset_semaphore(dev->dev);
Signature · 1 line
2.1.4