Skip to contentSkip to conversation

i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

PATCH Superseded Ismael Luceno Cortes 3 participants 24 messages Mar 6, 2019 – Mar 7, 2019

Subsystems (2)

v3 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

PATCH v3 1/6 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

Setting up the delay only needs to be done once; move it to rcar_i2c_set_speed so it's done at initialization time.

---
 drivers/i2c/rcar_i2c.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/i2c/rcar_i2c.c +6 -3
index 10b0f8bad4..a57f72a228 100644
@@ -116,9 +116,7 @@ static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
 	writel(0, priv->base + RCAR_I2C_ICMSR);
 	writel(priv->icccr, priv->base + RCAR_I2C_ICCCR);
 
-	if (priv->type == RCAR_I2C_TYPE_GEN3)
-		writel(RCAR_I2C_ICFBSCR_TCYC17, priv->base + RCAR_I2C_ICFBSCR);
-
+	/* Wait for the bus */
 	ret = wait_for_bit_le32(priv->base + RCAR_I2C_ICMCR,
 				RCAR_I2C_ICMCR_FSDA, false, 2, true);
 	if (ret) {
@@ -304,6 +302,11 @@ scgd_find:
 	priv->icccr = (scgd << RCAR_I2C_ICCCR_SCGD_OFF) | cdf;
 	writel(priv->icccr, priv->base + RCAR_I2C_ICCCR);
 
+	if (priv->type == RCAR_I2C_TYPE_GEN3) {
+		/* Set SCL/SDA delay */
+		writel(RCAR_I2C_ICFBSCR_TCYC17, priv->base + RCAR_I2C_ICFBSCR);
+	}
+
 	return 0;
 }
 

base-commit: f08023c07d826fbc8e62fdd3367961b2f0b06844
prerequisite-patch-id: 9e5b0458bc15640eb483ccad91dbe85150f9f7be

Signature · 1 line
2.19.1
#1 lore raw

PATCH v3 2/6 i2c: rcar_i2c: Add comments about registers & values

Document the meaning of macros related to registers and values to be written to them.

---
 drivers/i2c/rcar_i2c.c | 47 +++++++++++++++++++++++-------------------
 1 file changed, 26 insertions(+), 21 deletions(-)
drivers/i2c/rcar_i2c.c +26 -21
index a57f72a228..d7b27204cb 100644
@@ -18,35 +18,40 @@
 #include <asm/io.h>
 #include <wait_bit.h>
 
-#define RCAR_I2C_ICSCR			0x00
-#define RCAR_I2C_ICMCR			0x04
-#define RCAR_I2C_ICMCR_MDBS		BIT(7)
-#define RCAR_I2C_ICMCR_FSCL		BIT(6)
-#define RCAR_I2C_ICMCR_FSDA		BIT(5)
-#define RCAR_I2C_ICMCR_OBPC		BIT(4)
-#define RCAR_I2C_ICMCR_MIE		BIT(3)
+#define RCAR_I2C_ICSCR			0x00 /* slave ctrl */
+#define RCAR_I2C_ICMCR			0x04 /* master ctrl */
+#define RCAR_I2C_ICMCR_MDBS		BIT(7) /* non-fifo mode switch */
+#define RCAR_I2C_ICMCR_FSCL		BIT(6) /* override SCL pin */
+#define RCAR_I2C_ICMCR_FSDA		BIT(5) /* override SDA pin */
+#define RCAR_I2C_ICMCR_OBPC		BIT(4) /* override pins */
+#define RCAR_I2C_ICMCR_MIE		BIT(3) /* master if enable */
 #define RCAR_I2C_ICMCR_TSBE		BIT(2)
-#define RCAR_I2C_ICMCR_FSB		BIT(1)
-#define RCAR_I2C_ICMCR_ESG		BIT(0)
-#define RCAR_I2C_ICSSR			0x08
-#define RCAR_I2C_ICMSR			0x0c
+#define RCAR_I2C_ICMCR_FSB		BIT(1) /* force stop bit */
+#define RCAR_I2C_ICMCR_ESG		BIT(0) /* enable start bit gen */
+#define RCAR_I2C_ICSSR			0x08 /* slave status */
+#define RCAR_I2C_ICMSR			0x0c /* master status */
 #define RCAR_I2C_ICMSR_MASK		0x7f
-#define RCAR_I2C_ICMSR_MNR		BIT(6)
-#define RCAR_I2C_ICMSR_MAL		BIT(5)
-#define RCAR_I2C_ICMSR_MST		BIT(4)
+#define RCAR_I2C_ICMSR_MNR		BIT(6) /* Nack */
+#define RCAR_I2C_ICMSR_MAL		BIT(5) /* Arbitration lost */
+#define RCAR_I2C_ICMSR_MST		BIT(4) /* Stop */
 #define RCAR_I2C_ICMSR_MDE		BIT(3)
 #define RCAR_I2C_ICMSR_MDT		BIT(2)
 #define RCAR_I2C_ICMSR_MDR		BIT(1)
 #define RCAR_I2C_ICMSR_MAT		BIT(0)
-#define RCAR_I2C_ICSIER			0x10
-#define RCAR_I2C_ICMIER			0x14
-#define RCAR_I2C_ICCCR			0x18
+#define RCAR_I2C_ICSIER			0x10 /* slave irq enable */
+#define RCAR_I2C_ICMIER			0x14 /* master irq enable */
+#define RCAR_I2C_ICCCR			0x18 /* clock dividers */
 #define RCAR_I2C_ICCCR_SCGD_OFF		3
-#define RCAR_I2C_ICSAR			0x1c
-#define RCAR_I2C_ICMAR			0x20
-#define RCAR_I2C_ICRXD_ICTXD		0x24
+#define RCAR_I2C_ICSAR			0x1c /* slave address */
+#define RCAR_I2C_ICMAR			0x20 /* master address */
+#define RCAR_I2C_ICRXD_ICTXD		0x24 /* data port */
+/*
+ * First Bit Setup Cycle (Gen3).
+ * Defines 1st bit delay between SDA and SCL.
+ */
 #define RCAR_I2C_ICFBSCR		0x38
-#define RCAR_I2C_ICFBSCR_TCYC17		0x0f
+#define RCAR_I2C_ICFBSCR_TCYC17		0x0f /* 17*Tcyc */
+
 
 enum rcar_i2c_type {
 	RCAR_I2C_TYPE_GEN2,
Signature · 1 line
2.19.1
#3 lore raw

PATCH v3 2/6 i2c: rcar_i2c: Add comments about registers & values

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

Document the meaning of macros related to registers and values to be written to them.

---
  drivers/i2c/rcar_i2c.c | 47 +++++++++++++++++++++++-------------------
  1 file changed, 26 insertions(+), 21 deletions(-)

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#4 lore raw

PATCH v3 3/6 i2c: rcar_i2c: Fix sending of slave addresses

In reply to Ismael Luceno Cortes

Do the reset before clearing the MSR, otherwise it may result in a read or write operation instead if the start condition is repeated.

---
 drivers/i2c/rcar_i2c.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/i2c/rcar_i2c.c +3 -1
index d7b27204cb..a88fbcf34f 100644
@@ -134,9 +134,11 @@ static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
 	}
 
 	writel((chip << 1) | read, priv->base + RCAR_I2C_ICMAR);
-	writel(0, priv->base + RCAR_I2C_ICMSR);
+	/* Reset */
 	writel(RCAR_I2C_ICMCR_MDBS | RCAR_I2C_ICMCR_MIE | RCAR_I2C_ICMCR_ESG,
 	       priv->base + RCAR_I2C_ICMCR);
+	/* Clear Status */
+	writel(0, priv->base + RCAR_I2C_ICMSR);
 
 	ret = wait_for_bit_le32(priv->base + RCAR_I2C_ICMSR, mask,
 				true, 100, true);
Signature · 1 line
2.19.1
#6 lore raw

PATCH v3 3/6 i2c: rcar_i2c: Fix sending of slave addresses

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 8 lines

Do the reset before clearing the MSR, otherwise it may result in a read or write operation instead if the start condition is repeated.

---
  drivers/i2c/rcar_i2c.c | 4 +++-
  1 file changed, 3 insertions(+), 1 deletion(-)

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#7 lore raw

PATCH v3 4/6 i2c: rcar_i2c: Don't mask errors with EREMOTEIO at rcar_i2c_xfer

In reply to Ismael Luceno Cortes

Fix rcar_i2c_xfer return value, previously it was always returning -EREMOTEIO when dealing with errors from calls to the read/write functions.

---
 drivers/i2c/rcar_i2c.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/i2c/rcar_i2c.c +1 -1
index a88fbcf34f..9223eaecbf 100644
@@ -221,7 +221,7 @@ static int rcar_i2c_xfer(struct udevice *dev, struct i2c_msg *msg, int nmsgs)
 			ret = rcar_i2c_write_common(dev, msg);
 
 		if (ret)
-			return -EREMOTEIO;
+			return ret;
 	}
 
 	return ret;
Signature · 1 line
2.19.1
#8 lore raw

PATCH v3 4/6 i2c: rcar_i2c: Don't mask errors with EREMOTEIO at rcar_i2c_xfer

Quoted 5 lines · Ismael Luceno Cortes

Fix rcar_i2c_xfer return value, previously it was always returning -EREMOTEIO when dealing with errors from calls to the read/write functions.

Quoted 18 lines
---
 drivers/i2c/rcar_i2c.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
drivers/i2c/rcar_i2c.c +1 -1
index a88fbcf34f..9223eaecbf 100644
@@ -221,7 +221,7 @@ static int rcar_i2c_xfer(struct udevice *dev, struct i2c_msg *msg, int nmsgs)
 			ret = rcar_i2c_write_common(dev, msg);
 
 		if (ret)
-			return -EREMOTEIO;
+			return ret;
 	}
 
 	return ret;
Signature · 2 lines
Best regards,
Marek Vasut
#9 lore raw

PATCH v3 4/6 i2c: rcar_i2c: Don't mask errors with EREMOTEIO at rcar_i2c_xfer

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 8 lines

Fix rcar_i2c_xfer return value, previously it was always returning -EREMOTEIO when dealing with errors from calls to the read/write functions.

---
  drivers/i2c/rcar_i2c.c | 2 +-
  1 file changed, 1 insertion(+), 1 deletion(-)

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#10 lore raw

PATCH v3 5/6 i2c: rcar_i2c: Set the slave address from rcar_i2c_xfer

In reply to Ismael Luceno Cortes

It needs to be done for both reads and writes, so do it at rcar_i2c_xfer to avoid duplication.

---
 drivers/i2c/rcar_i2c.c | 14 +++++---------
 1 file changed, 5 insertions(+), 9 deletions(-)
drivers/i2c/rcar_i2c.c +5 -9
index 9223eaecbf..78ef8acacc 100644
@@ -158,10 +158,6 @@ static int rcar_i2c_read_common(struct udevice *dev, struct i2c_msg *msg)
 	u32 icmcr = RCAR_I2C_ICMCR_MDBS | RCAR_I2C_ICMCR_MIE;
 	int i, ret = -EREMOTEIO;
 
-	ret = rcar_i2c_set_addr(dev, msg->addr, 1);
-	if (ret)
-		return ret;
-
 	for (i = 0; i < msg->len; i++) {
 		if (msg->len - 1 == i)
 			icmcr |= RCAR_I2C_ICMCR_FSB;
@@ -188,10 +184,6 @@ static int rcar_i2c_write_common(struct udevice *dev, struct i2c_msg *msg)
 	u32 icmcr = RCAR_I2C_ICMCR_MDBS | RCAR_I2C_ICMCR_MIE;
 	int i, ret = -EREMOTEIO;
 
-	ret = rcar_i2c_set_addr(dev, msg->addr, 0);
-	if (ret)
-		return ret;
-
 	for (i = 0; i < msg->len; i++) {
 		writel(msg->buf[i], priv->base + RCAR_I2C_ICRXD_ICTXD);
 		writel(icmcr, priv->base + RCAR_I2C_ICMCR);
@@ -215,6 +207,10 @@ static int rcar_i2c_xfer(struct udevice *dev, struct i2c_msg *msg, int nmsgs)
 	int ret;
 
 	for (; nmsgs > 0; nmsgs--, msg++) {
+		ret = rcar_i2c_set_addr(dev, msg->addr, 1);
+		if (ret)
+			return ret;
+
 		if (msg->flags & I2C_M_RD)
 			ret = rcar_i2c_read_common(dev, msg);
 		else
@@ -224,7 +220,7 @@ static int rcar_i2c_xfer(struct udevice *dev, struct i2c_msg *msg, int nmsgs)
 			return ret;
 	}
 
-	return ret;
+	return 0;
 }
 
 static int rcar_i2c_probe_chip(struct udevice *dev, uint addr, uint flags)
Signature · 1 line
2.19.1
#12 lore raw

PATCH v3 5/6 i2c: rcar_i2c: Set the slave address from rcar_i2c_xfer

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

It needs to be done for both reads and writes, so do it at rcar_i2c_xfer to avoid duplication.

---
  drivers/i2c/rcar_i2c.c | 14 +++++---------
  1 file changed, 5 insertions(+), 9 deletions(-)

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#13 lore raw

PATCH v3 6/6 i2c: rcar_i2c: Move FSDA check to rcar_i2c_recover

In reply to Ismael Luceno Cortes

Cosmetic change. Any call to the recover function would need to do the same check afterwards, so it's sensible to make it part of the function.

---
 drivers/i2c/rcar_i2c.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/i2c/rcar_i2c.c +6 -4
index 78ef8acacc..8fe0aadffa 100644
@@ -81,12 +81,13 @@ static int rcar_i2c_finish(struct udevice *dev)
 	return ret;
 }
 
-static void rcar_i2c_recover(struct udevice *dev)
+static int rcar_i2c_recover(struct udevice *dev)
 {
 	struct rcar_i2c_priv *priv = dev_get_priv(dev);
 	u32 mcr = RCAR_I2C_ICMCR_MDBS | RCAR_I2C_ICMCR_OBPC;
 	u32 mcra = mcr | RCAR_I2C_ICMCR_FSDA;
 	int i;
+	u32 mstat;
 
 	/* Send 9 SCL pulses */
 	for (i = 0; i < 9; i++) {
@@ -106,6 +107,9 @@ static void rcar_i2c_recover(struct udevice *dev)
 	udelay(5);
 	writel(mcra | RCAR_I2C_ICMCR_FSCL, priv->base + RCAR_I2C_ICMCR);
 	udelay(5);
+
+	mstat = readl(priv->base + RCAR_I2C_ICMSR);
+	return mstat & RCAR_I2C_ICMCR_FSDA ? -EBUSY : 0;
 }
 
 static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
@@ -125,9 +129,7 @@ static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
 	ret = wait_for_bit_le32(priv->base + RCAR_I2C_ICMCR,
 				RCAR_I2C_ICMCR_FSDA, false, 2, true);
 	if (ret) {
-		rcar_i2c_recover(dev);
-		val = readl(priv->base + RCAR_I2C_ICMSR);
-		if (val & RCAR_I2C_ICMCR_FSDA) {
+		if (rcar_i2c_recover(dev)) {
 			dev_err(dev, "Bus busy, aborting\n");
 			return ret;
 		}
Signature · 1 line
2.19.1
#14 lore raw

PATCH v3 6/6 i2c: rcar_i2c: Move FSDA check to rcar_i2c_recover

Quoted 4 lines · Ismael Luceno Cortes

Cosmetic change. Any call to the recover function would need to do the same check afterwards, so it's sensible to make it part of the function.

Quoted 10 lines
---
 drivers/i2c/rcar_i2c.c | 10 ++++++----
 1 file changed, 6 insertions(+), 4 deletions(-)
drivers/i2c/rcar_i2c.c +0 -0
index 78ef8acacc..8fe0aadffa 100644
@@ -81,12 +81,13 @@ static int rcar_i2c_finish(struct udevice *dev)
 	return ret;

drivers/i2c/rcar_i2c.c static int rcar_i2c_finish(struct udevice *dev) · line 90

Quoted 10 lines
drivers/i2c/rcar_i2c.c +2 -1
 }
 
-static void rcar_i2c_recover(struct udevice *dev)
+static int rcar_i2c_recover(struct udevice *dev)
 {
 	struct rcar_i2c_priv *priv = dev_get_priv(dev);
 	u32 mcr = RCAR_I2C_ICMCR_MDBS | RCAR_I2C_ICMCR_OBPC;
 	u32 mcra = mcr | RCAR_I2C_ICMCR_FSDA;
 	int i;
+	u32 mstat;

Nit: would be nice to keep these variables sorted in reverse xmas tree.

Signature · 2 lines
Best regards,
Marek Vasut
#15 lore raw

PATCH v3 6/6 i2c: rcar_i2c: Move FSDA check to rcar_i2c_recover

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

Cosmetic change. Any call to the recover function would need to do the same check afterwards, so it's sensible to make it part of the function.

---
  drivers/i2c/rcar_i2c.c | 10 ++++++----
  1 file changed, 6 insertions(+), 4 deletions(-)

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#16 lore raw

PATCH v3 6/6 i2c: rcar_i2c: Move FSDA check to rcar_i2c_recover

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 38 lines

Cosmetic change. Any call to the recover function would need to do the same check afterwards, so it's sensible to make it part of the function.

---
  drivers/i2c/rcar_i2c.c | 10 ++++++----
  1 file changed, 6 insertions(+), 4 deletions(-)
drivers/i2c/rcar_i2c.c +5 -1
index 78ef8acacc..8fe0aadffa 100644
@@ -81,12 +81,13 @@ static int rcar_i2c_finish(struct udevice *dev)
  	return ret;
  }
 
-static void rcar_i2c_recover(struct udevice *dev)
+static int rcar_i2c_recover(struct udevice *dev)
  {
  	struct rcar_i2c_priv *priv = dev_get_priv(dev);
  	u32 mcr = RCAR_I2C_ICMCR_MDBS | RCAR_I2C_ICMCR_OBPC;
  	u32 mcra = mcr | RCAR_I2C_ICMCR_FSDA;
  	int i;
+	u32 mstat;
 
  	/* Send 9 SCL pulses */
  	for (i = 0; i < 9; i++) {
@@ -106,6 +107,9 @@ static void rcar_i2c_recover(struct udevice *dev)
  	udelay(5);
  	writel(mcra | RCAR_I2C_ICMCR_FSCL, priv->base + RCAR_I2C_ICMCR);
  	udelay(5);
+
+	mstat = readl(priv->base + RCAR_I2C_ICMSR);
+	return mstat & RCAR_I2C_ICMCR_FSDA ? -EBUSY : 0;
  }
 
  static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)

drivers/i2c/rcar_i2c.c static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read) · line 135

Quoted 10 lines
drivers/i2c/rcar_i2c.c +1 -3
@@ -125,9 +129,7 @@ static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
  	ret = wait_for_bit_le32(priv->base + RCAR_I2C_ICMCR,
  				RCAR_I2C_ICMCR_FSDA, false, 2, true);
  	if (ret) {
-		rcar_i2c_recover(dev);
-		val = readl(priv->base + RCAR_I2C_ICMSR);
-		if (val & RCAR_I2C_ICMCR_FSDA) {
+		if (rcar_i2c_recover(dev)) {
  			dev_err(dev, "Bus busy, aborting\n");
  			return ret;
  		}

This change brings travis build to fail, see:

https://travis-ci.org/hsdenx/u-boot-i2c/jobs/502984991#L1265

On which hardware have you tested this patchset?

The fix is easy:

$ git diff
drivers/i2c/rcar_i2c.c +0 -1
index 8fe0aadffa..2ebae349ed 100644
@@ -117,7 +117,6 @@ static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
         struct rcar_i2c_priv *priv = dev_get_priv(dev);
         u32 mask = RCAR_I2C_ICMSR_MAT |
                    (read ? RCAR_I2C_ICMSR_MDR : RCAR_I2C_ICMSR_MDE);
-       u32 val;
         int ret;
 
         writel(0, priv->base + RCAR_I2C_ICMIER);

$

but I am unsure if you ever tested the changes...

So, please, send a v4, Thanks!

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#18 lore raw

PATCH v3 1/6 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

In reply to Ismael Luceno Cortes
Quoted 4 lines · Ismael Luceno Cortes

Setting up the delay only needs to be done once; move it to rcar_i2c_set_speed so it's done at initialization time.

Quoted 35 lines
---
 drivers/i2c/rcar_i2c.c | 9 ++++++---
 1 file changed, 6 insertions(+), 3 deletions(-)
drivers/i2c/rcar_i2c.c +6 -3
index 10b0f8bad4..a57f72a228 100644
@@ -116,9 +116,7 @@ static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
 	writel(0, priv->base + RCAR_I2C_ICMSR);
 	writel(priv->icccr, priv->base + RCAR_I2C_ICCCR);
 
-	if (priv->type == RCAR_I2C_TYPE_GEN3)
-		writel(RCAR_I2C_ICFBSCR_TCYC17, priv->base + RCAR_I2C_ICFBSCR);
-
+	/* Wait for the bus */
 	ret = wait_for_bit_le32(priv->base + RCAR_I2C_ICMCR,
 				RCAR_I2C_ICMCR_FSDA, false, 2, true);
 	if (ret) {
@@ -304,6 +302,11 @@ scgd_find:
 	priv->icccr = (scgd << RCAR_I2C_ICCCR_SCGD_OFF) | cdf;
 	writel(priv->icccr, priv->base + RCAR_I2C_ICCCR);
 
+	if (priv->type == RCAR_I2C_TYPE_GEN3) {
+		/* Set SCL/SDA delay */
+		writel(RCAR_I2C_ICFBSCR_TCYC17, priv->base + RCAR_I2C_ICFBSCR);
+	}
+
 	return 0;
 }
 

base-commit: f08023c07d826fbc8e62fdd3367961b2f0b06844
prerequisite-patch-id: 9e5b0458bc15640eb483ccad91dbe85150f9f7be

Signature · 2 lines
Best regards,
Marek Vasut
#19 lore raw

PATCH v3 1/6 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

Setting up the delay only needs to be done once; move it to rcar_i2c_set_speed so it's done at initialization time.

---
  drivers/i2c/rcar_i2c.c | 9 ++++++---
  1 file changed, 6 insertions(+), 3 deletions(-)

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#20 lore raw

PATCH v3 1/6 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

In reply to Ismael Luceno Cortes

Hello Ismael,

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

Setting up the delay only needs to be done once; move it to rcar_i2c_set_speed so it's done at initialization time.

---
  drivers/i2c/rcar_i2c.c | 9 ++++++---
  1 file changed, 6 insertions(+), 3 deletions(-)

Your patch does not apply to

base-commit: f08023c07d826fbc8e62fdd3367961b2f0b06844

I get based on f08023c07d826fbc8e62fdd3367961b2f0b06844

Applying: i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed
error: sha1 information is lacking or useless (drivers/i2c/rcar_i2c.c).
error: could not build fake ancestor
Patch failed at 0001 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed Use 'git am --show-current-patch' to see the failed patch
When you have resolved this problem, run "git am --continue".
If you prefer to skip this patch, run "git am --skip" instead.
To restore the original branch and stop patching, run "git am --abort".

My base:

pollux:u-boot hs [temp] $ glg
* f08023c07d - (HEAD -> temp, tag: v2019.04-rc3, origin/master, origin/HEAD, master) Prepare v2019.04-rc4 (2 days ago) <Tom Rini>
* e102f74d65 - Merge branch 'master' of git://git.denx.de/u-boot-sh (2 days ago) <Tom Rini>
|\
| * 65eef78cdb - pinctrl: renesas: r8a77990: Reivse USB ID pin name (3 days ago) <Hiroyuki Yokoyama>

How does this work for you ?

Quoted 5 lines
drivers/i2c/rcar_i2c.c +0 -0
index 10b0f8bad4..a57f72a228 100644

drivers/i2c/rcar_i2c.c static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read) · line 122

Quoted 10 lines
drivers/i2c/rcar_i2c.c +1 -3
@@ -116,9 +116,7 @@ static int rcar_i2c_set_addr(struct udevice *dev, u8 chip, u8 read)
  	writel(0, priv->base + RCAR_I2C_ICMSR);
  	writel(priv->icccr, priv->base + RCAR_I2C_ICCCR);
 
-	if (priv->type == RCAR_I2C_TYPE_GEN3)
-		writel(RCAR_I2C_ICFBSCR_TCYC17, priv->base + RCAR_I2C_ICFBSCR);
-
+	/* Wait for the bus */
  	ret = wait_for_bit_le32(priv->base + RCAR_I2C_ICMCR,
  				RCAR_I2C_ICMCR_FSDA, false, 2, true);
  	if (ret) {

I do not find RCAR_I2C_TYPE_GEN3 in current mainline code ... please check!

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#21 lore raw

PATCH v3 1/6 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

Quoted 1 line · Heiko Schocher

Hello Ismael,

Hi,

Quoted 37 lines

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

Setting up the delay only needs to be done once; move it to rcar_i2c_set_speed so it's done at initialization time.

---
  drivers/i2c/rcar_i2c.c | 9 ++++++---

  1 file changed, 6 insertions(+), 3 deletions(-)

Your patch does not apply to

base-commit: f08023c07d826fbc8e62fdd3367961b2f0b06844

I get based on f08023c07d826fbc8e62fdd3367961b2f0b06844

Applying: i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed
error: sha1 information is lacking or useless (drivers/i2c/rcar_i2c.c).
error: could not build fake ancestor
Patch failed at 0001 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed
Use 'git am --show-current-patch' to see the failed patch When you have resolved this problem, run "git am --continue". If you prefer to skip this patch, run "git am --skip" instead. To restore the original branch and stop patching, run "git am --abort".

My base:

pollux:u-boot hs [temp] $ glg
* f08023c07d - (HEAD -> temp, tag: v2019.04-rc3, origin/master, origin/HEAD, master) Prepare v2019.04-rc4 (2 days ago) <Tom Rini> *   e102f74d65 - Merge branch 'master' of git://git.denx.de/u-boot-sh (2 days ago) <Tom Rini>
|\
| * 65eef78cdb - pinctrl: renesas: r8a77990: Reivse USB ID pin name (3 days ago) <Hiroyuki Yokoyama>

Quoted 1 line

How does this work for you ?

This series applies in top of

[PATCH] i2c: rcar_i2c: Add Gen3 SoC support

[...]

Signature · 2 lines
Best regards,
Marek Vasut
#22 lore raw

PATCH v3 1/6 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

Hello Marek,

Am 07.03.2019 um 09:04 schrieb Marek Vasut:

Quoted 42 lines · Heiko Schocher
Quoted 1 line

Hello Ismael,

Hi,

Quoted 37 lines

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

Setting up the delay only needs to be done once; move it to rcar_i2c_set_speed so it's done at initialization time.

---
   drivers/i2c/rcar_i2c.c | 9 ++++++---
   1 file changed, 6 insertions(+), 3 deletions(-)

Your patch does not apply to

base-commit: f08023c07d826fbc8e62fdd3367961b2f0b06844

I get based on f08023c07d826fbc8e62fdd3367961b2f0b06844

Applying: i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed
error: sha1 information is lacking or useless (drivers/i2c/rcar_i2c.c).
error: could not build fake ancestor
Patch failed at 0001 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed
Use 'git am --show-current-patch' to see the failed patch When you have resolved this problem, run "git am --continue". If you prefer to skip this patch, run "git am --skip" instead. To restore the original branch and stop patching, run "git am --abort".

My base:

pollux:u-boot hs [temp] $ glg
* f08023c07d - (HEAD -> temp, tag: v2019.04-rc3, origin/master, origin/HEAD, master) Prepare v2019.04-rc4 (2 days ago) <Tom Rini> *   e102f74d65 - Merge branch 'master' of git://git.denx.de/u-boot-sh (2 days ago) <Tom Rini>
|\
| * 65eef78cdb - pinctrl: renesas: r8a77990: Reivse USB ID pin name (3 days ago) <Hiroyuki Yokoyama>

Quoted 5 lines
Quoted 1 line

How does this work for you ?

This series applies in top of

[PATCH] i2c: rcar_i2c: Add Gen3 SoC support

Ah! Thanks for the hint, I missed this...

Hmm... I thought the patchseries from Ismael as bugfix patch, so I thought to pick it up for 2019.04 ... but based on the this patch, is it okay to pick up this stuff, when merge window is open?

bye,
Heiko

Signature · 3 lines
DENX Software Engineering GmbH,      Managing Director: Wolfgang Denk
HRB 165235 Munich, Office: Kirchenstr.5, D-82194 Groebenzell, Germany
Phone: +49-8142-66989-52   Fax: +49-8142-66989-80   Email: hs at denx.de
#23 lore raw

PATCH v3 1/6 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed

Quoted 1 line · Heiko Schocher

Hello Marek,

Hi,

Quoted 47 lines

Am 07.03.2019 um 09:04 schrieb Marek Vasut:

Quoted 46 lines · Heiko Schocher
Quoted 1 line

Hello Ismael,

Hi,

Quoted 37 lines

Am 06.03.2019 um 11:46 schrieb Ismael Luceno Cortes:

Quoted 7 lines

Setting up the delay only needs to be done once; move it to rcar_i2c_set_speed so it's done at initialization time.

---
   drivers/i2c/rcar_i2c.c | 9 ++++++---

   1 file changed, 6 insertions(+), 3 deletions(-)

Your patch does not apply to

base-commit: f08023c07d826fbc8e62fdd3367961b2f0b06844

I get based on f08023c07d826fbc8e62fdd3367961b2f0b06844

Applying: i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed
error: sha1 information is lacking or useless (drivers/i2c/rcar_i2c.c).
error: could not build fake ancestor
Patch failed at 0001 i2c: rcar_i2c: Setup SCL/SDA delay at rcar_i2c_set_speed
Use 'git am --show-current-patch' to see the failed patch When you have resolved this problem, run "git am --continue". If you prefer to skip this patch, run "git am --skip" instead. To restore the original branch and stop patching, run "git am --abort".

My base:

pollux:u-boot hs [temp] $ glg
* f08023c07d - (HEAD -> temp, tag: v2019.04-rc3, origin/master, origin/HEAD, master) Prepare v2019.04-rc4 (2 days ago) <Tom Rini> *   e102f74d65 - Merge branch 'master' of git://git.denx.de/u-boot-sh (2 days ago) <Tom Rini>
|\
| * 65eef78cdb - pinctrl: renesas: r8a77990: Reivse USB ID pin name (3 days ago) <Hiroyuki Yokoyama>

Quoted 1 line

How does this work for you ?

This series applies in top of

Quoted 7 lines
Quoted 1 line

[PATCH] i2c: rcar_i2c: Add Gen3 SoC support

Ah! Thanks for the hint, I missed this...

Hmm... I thought the patchseries from Ismael as bugfix patch, so I thought to pick it up for 2019.04 ... but based on the this patch, is it okay to pick up this stuff, when merge window is open?

The Gen3 support is technically a bugfix too, since it's mostly fixing build warnings on ARM64 . It should be fine to put it into this release.

Signature · 2 lines
Best regards,
Marek Vasut