diff mbox series

[V1,2/2] i2c: iproc: add slave pec support

Message ID 20200716080836.2279-3-rayagonda.kokatanur@broadcom.com (mailing list archive)
State New, archived
Headers show
Series add PEC support on slave side | expand

Commit Message

Rayagonda Kokatanur July 16, 2020, 8:08 a.m. UTC
Iproc supports PEC computation and checking in both Master
and Slave mode.

This patch adds support for PEC in slave mode.

Signed-off-by: Rayagonda Kokatanur <rayagonda.kokatanur@broadcom.com>
---
 drivers/i2c/busses/i2c-bcm-iproc.c | 50 +++++++++++++++++++++++++++---
 1 file changed, 46 insertions(+), 4 deletions(-)

Comments

Andy Shevchenko July 16, 2020, 10:14 a.m. UTC | #1
On Thu, Jul 16, 2020 at 11:14 AM Rayagonda Kokatanur
<rayagonda.kokatanur@broadcom.com> wrote:
>
> Iproc supports PEC computation and checking in both Master
> and Slave mode.
>
> This patch adds support for PEC in slave mode.

...

> -#define S_RX_PEC_ERR_SHIFT           29
> +#define S_RX_PEC_ERR_SHIFT           28
> +#define S_RX_PEC_ERR_MASK            0x3
> +#define S_RX_PEC_ERR                 0x1

This needs to be explained in the commit message, in particular why
this change makes no regression.

...

> +static int bcm_iproc_smbus_check_slave_pec(struct bcm_iproc_i2c_dev *iproc_i2c,
> +                                          u32 val)
> +{
> +       u8 err_status;

> +       int ret = 0;

Completely redundant variable.

> +       if (!iproc_i2c->en_s_pec)
> +               return ret;

return 0;

> +       err_status = (u8)((val >> S_RX_PEC_ERR_SHIFT) & S_RX_PEC_ERR_MASK);

Why casting?

> +       if (err_status == S_RX_PEC_ERR) {
> +               dev_err(iproc_i2c->device, "Slave PEC error\n");

> +               ret = -EBADMSG;

return ...

> +       }
> +
> +       return ret;

return 0;

> +}

...

> +                       if (rx_status == I2C_SLAVE_RX_END) {
> +                               int ret;
> +
> +                               ret = bcm_iproc_smbus_check_slave_pec(iproc_i2c,
> +                                                                     val);

One line looks better.

> +                               if (!ret)

Why not positive conditional?

> +                                       i2c_slave_event(iproc_i2c->slave,
> +                                                       I2C_SLAVE_STOP, &value);
> +                               else
> +                                       i2c_slave_event(iproc_i2c->slave,
> +                                                       I2C_SLAVE_PEC_ERR,
> +                                                       &value);
> +                       }
Rayagonda Kokatanur July 16, 2020, 5:19 p.m. UTC | #2
Hi Andy,

On Thu, Jul 16, 2020 at 3:44 PM Andy Shevchenko
<andy.shevchenko@gmail.com> wrote:
>
> On Thu, Jul 16, 2020 at 11:14 AM Rayagonda Kokatanur
> <rayagonda.kokatanur@broadcom.com> wrote:
> >
> > Iproc supports PEC computation and checking in both Master
> > and Slave mode.
> >
> > This patch adds support for PEC in slave mode.
>
> ...
>
> > -#define S_RX_PEC_ERR_SHIFT           29
> > +#define S_RX_PEC_ERR_SHIFT           28
> > +#define S_RX_PEC_ERR_MASK            0x3
> > +#define S_RX_PEC_ERR                 0x1
>
> This needs to be explained in the commit message, in particular why
> this change makes no regression.

I didn't get what do you mean by "no regression", please elaborate.

>
> ...
>
> > +static int bcm_iproc_smbus_check_slave_pec(struct bcm_iproc_i2c_dev *iproc_i2c,
> > +                                          u32 val)
> > +{
> > +       u8 err_status;
>
> > +       int ret = 0;
>
> Completely redundant variable.
>
> > +       if (!iproc_i2c->en_s_pec)
> > +               return ret;
>
> return 0;
>
> > +       err_status = (u8)((val >> S_RX_PEC_ERR_SHIFT) & S_RX_PEC_ERR_MASK);
>
> Why casting?
>
> > +       if (err_status == S_RX_PEC_ERR) {
> > +               dev_err(iproc_i2c->device, "Slave PEC error\n");
>
> > +               ret = -EBADMSG;
>
> return ...
>
> > +       }
> > +
> > +       return ret;
>
> return 0;
>
> > +}
>
> ...
>
> > +                       if (rx_status == I2C_SLAVE_RX_END) {
> > +                               int ret;
> > +
> > +                               ret = bcm_iproc_smbus_check_slave_pec(iproc_i2c,
> > +                                                                     val);
>
> One line looks better.

Yes, but to have 80 char per line, I have to do this.
>
> > +                               if (!ret)
>
> Why not positive conditional?

Thank you for your review.
Will fix all above.

Best regards,
Rayagonda

>
> > +                                       i2c_slave_event(iproc_i2c->slave,
> > +                                                       I2C_SLAVE_STOP, &value);
> > +                               else
> > +                                       i2c_slave_event(iproc_i2c->slave,
> > +                                                       I2C_SLAVE_PEC_ERR,
> > +                                                       &value);
> > +                       }
>
> --
> With Best Regards,
> Andy Shevchenko
Andy Shevchenko July 16, 2020, 6:26 p.m. UTC | #3
On Thu, Jul 16, 2020 at 10:49:14PM +0530, Rayagonda Kokatanur wrote:
> On Thu, Jul 16, 2020 at 3:44 PM Andy Shevchenko
> <andy.shevchenko@gmail.com> wrote:
> > On Thu, Jul 16, 2020 at 11:14 AM Rayagonda Kokatanur
> > <rayagonda.kokatanur@broadcom.com> wrote:

...
> > > -#define S_RX_PEC_ERR_SHIFT           29
> > > +#define S_RX_PEC_ERR_SHIFT           28

> > This needs to be explained in the commit message, in particular why
> > this change makes no regression.
> 
> I didn't get what do you mean by "no regression", please elaborate.

The definition above has been changed. The point is you have to point out in
the commit message why it's okay and makes no regression.
For example, "..._SHIFT is changed to ... according to documentation. Since
there was no user of it no regression will be made." Provide proper text, b/c
I have no idea what is exactly the reason of the change and if it's indeed used
to have no users.

...

> > > +                               ret = bcm_iproc_smbus_check_slave_pec(iproc_i2c,
> > > +                                                                     val);
> >
> > One line looks better.
> 
> Yes, but to have 80 char per line, I have to do this.

We have more, but even if you stick with 80 the above is harder to get than if it is one line.
diff mbox series

Patch

diff --git a/drivers/i2c/busses/i2c-bcm-iproc.c b/drivers/i2c/busses/i2c-bcm-iproc.c
index 8a3c98866fb7..51c8b165bb5e 100644
--- a/drivers/i2c/busses/i2c-bcm-iproc.c
+++ b/drivers/i2c/busses/i2c-bcm-iproc.c
@@ -93,6 +93,7 @@ 
 #define S_CMD_STATUS_MASK            0x07
 #define S_CMD_STATUS_SUCCESS         0x0
 #define S_CMD_STATUS_TIMEOUT         0x5
+#define S_CMD_PEC_SHIFT              8
 
 #define IE_OFFSET                    0x38
 #define IE_M_RX_FIFO_FULL_SHIFT      31
@@ -138,7 +139,9 @@ 
 #define S_RX_OFFSET                  0x4c
 #define S_RX_STATUS_SHIFT            30
 #define S_RX_STATUS_MASK             0x03
-#define S_RX_PEC_ERR_SHIFT           29
+#define S_RX_PEC_ERR_SHIFT           28
+#define S_RX_PEC_ERR_MASK            0x3
+#define S_RX_PEC_ERR                 0x1
 #define S_RX_DATA_SHIFT              0
 #define S_RX_DATA_MASK               0xff
 
@@ -205,6 +208,8 @@  struct bcm_iproc_i2c_dev {
 	/* bytes that have been read */
 	unsigned int rx_bytes;
 	unsigned int thld_bytes;
+
+	bool en_s_pec;
 };
 
 /*
@@ -321,6 +326,24 @@  static void bcm_iproc_i2c_check_slave_status(
 	}
 }
 
+static int bcm_iproc_smbus_check_slave_pec(struct bcm_iproc_i2c_dev *iproc_i2c,
+					   u32 val)
+{
+	u8 err_status;
+	int ret = 0;
+
+	if (!iproc_i2c->en_s_pec)
+		return ret;
+
+	err_status = (u8)((val >> S_RX_PEC_ERR_SHIFT) & S_RX_PEC_ERR_MASK);
+	if (err_status == S_RX_PEC_ERR) {
+		dev_err(iproc_i2c->device, "Slave PEC error\n");
+		ret = -EBADMSG;
+	}
+
+	return ret;
+}
+
 static bool bcm_iproc_i2c_slave_isr(struct bcm_iproc_i2c_dev *iproc_i2c,
 				    u32 status)
 {
@@ -347,6 +370,8 @@  static bool bcm_iproc_i2c_slave_isr(struct bcm_iproc_i2c_dev *iproc_i2c,
 			iproc_i2c_wr_reg(iproc_i2c, S_TX_OFFSET, value);
 
 			val = BIT(S_CMD_START_BUSY_SHIFT);
+			if (iproc_i2c->en_s_pec)
+				val |= BIT(S_CMD_PEC_SHIFT);
 			iproc_i2c_wr_reg(iproc_i2c, S_CMD_OFFSET, val);
 
 			/*
@@ -361,9 +386,19 @@  static bool bcm_iproc_i2c_slave_isr(struct bcm_iproc_i2c_dev *iproc_i2c,
 			value = (u8)((val >> S_RX_DATA_SHIFT) & S_RX_DATA_MASK);
 			i2c_slave_event(iproc_i2c->slave,
 					I2C_SLAVE_WRITE_RECEIVED, &value);
-			if (rx_status == I2C_SLAVE_RX_END)
-				i2c_slave_event(iproc_i2c->slave,
-						I2C_SLAVE_STOP, &value);
+			if (rx_status == I2C_SLAVE_RX_END) {
+				int ret;
+
+				ret = bcm_iproc_smbus_check_slave_pec(iproc_i2c,
+								      val);
+				if (!ret)
+					i2c_slave_event(iproc_i2c->slave,
+							I2C_SLAVE_STOP, &value);
+				else
+					i2c_slave_event(iproc_i2c->slave,
+							I2C_SLAVE_PEC_ERR,
+							&value);
+			}
 		}
 	} else if (status & BIT(IS_S_TX_UNDERRUN_SHIFT)) {
 		/* Master read other than start */
@@ -372,6 +407,8 @@  static bool bcm_iproc_i2c_slave_isr(struct bcm_iproc_i2c_dev *iproc_i2c,
 
 		iproc_i2c_wr_reg(iproc_i2c, S_TX_OFFSET, value);
 		val = BIT(S_CMD_START_BUSY_SHIFT);
+		if (iproc_i2c->en_s_pec)
+			val |= BIT(S_CMD_PEC_SHIFT);
 		iproc_i2c_wr_reg(iproc_i2c, S_CMD_OFFSET, val);
 	}
 
@@ -1065,6 +1102,11 @@  static int bcm_iproc_i2c_reg_slave(struct i2c_client *slave)
 	if (slave->flags & I2C_CLIENT_TEN)
 		return -EAFNOSUPPORT;
 
+	/* Enable partial slave HW PEC support if requested by the client */
+	iproc_i2c->en_s_pec = !!(slave->flags & I2C_CLIENT_PEC);
+	if (iproc_i2c->en_s_pec)
+		dev_info(iproc_i2c->device, "Enable PEC\n");
+
 	iproc_i2c->slave = slave;
 	bcm_iproc_i2c_slave_init(iproc_i2c, false);
 	return 0;