i2c-algo-bit: Fix NAK/ARB comments

Update comments and logging on return path for byte writes.  NAK is
an error, to be reported or optionally ignored.  Timeouts are always
errors.  Lost arbitration is not currently handled, so don't even list
it as an option in the error message.

Don't return bogus EFAULT code for inappropriate NAK; EIO is better,
there is no bad userspace address in question.

Signed-off-by: David Brownell <dbrownell@users.sourceforge.net>
Signed-off-by: Jean Delvare <khali@linux-fr.org>

1 files changed, 20 insertions, 4 deletions

diff --git a/drivers/i2c/algos/i2c-algo-bit.c b/drivers/i2c/algos/i2c-algo-bit.c
index 260c5d70b0b4..35812823787b 100644
--- a/drivers/i2c/algos/i2c-algo-bit.c
+++ b/drivers/i2c/algos/i2c-algo-bit.c
@@ -357,10 +357,26 @@ static int sendbytes(struct i2c_adapter *i2c_adap, struct i2c_msg *msg)
 			count--;
 			temp++;
 			wrcount++;
-		} else { /* arbitration or no acknowledge */
-			dev_err(&i2c_adap->dev, "sendbytes: error - bailout.\n");
-			return (retval<0)? retval : -EFAULT;
-			        /* got a better one ?? */
+
+		/* A slave NAKing the master means the slave didn't like
+		 * something about the data it saw.  For example, maybe
+		 * the SMBus PEC was wrong.
+		 */
+		} else if (retval == 0) {
+			dev_err(&i2c_adap->dev, "sendbytes: NAK bailout.\n");
+			return -EIO;
+
+		/* Timeout; or (someday) lost arbitration
+		 *
+		 * FIXME Lost ARB implies retrying the transaction from
+		 * the first message, after the "winning" master issues
+		 * its STOP.  As a rule, upper layer code has no reason
+		 * to know or care about this ... it is *NOT* an error.
+		 */
+		} else {
+			dev_err(&i2c_adap->dev, "sendbytes: error %d\n",
+					retval);
+			return retval;
 		}
 	}
 	return wrcount;