[PATCH 12/40] hwmon: (lm90) Only re-read registers if volatile

From: Guenter Roeck
Date: Wed May 25 2022 - 09:59:34 EST


When reading 16-bit volatile registers, the code uses a trick to
determine if a temperature is consistent: It reads the high part
of the register twice. If the values are the same, the code assumes
that the reading is consistent. If the value differs, the code
re-reads the second register as well and assumes that it now has
correct values.

This is only necessary for volatile registers. Add a parameter to
lm90_read16() to indicate if the register is volatile to avoid the
extra overhead for non-volatile registers.

Signed-off-by: Guenter Roeck <linux@xxxxxxxxxxxx>
---
drivers/hwmon/lm90.c | 41 +++++++++++++++++++++++------------------
1 file changed, 23 insertions(+), 18 deletions(-)

diff --git a/drivers/hwmon/lm90.c b/drivers/hwmon/lm90.c
index acb9ca3b99b0..b20be0cb28b5 100644
--- a/drivers/hwmon/lm90.c
+++ b/drivers/hwmon/lm90.c
@@ -602,29 +602,34 @@ static int lm90_write_reg(struct i2c_client *client, u8 reg, u8 val)
return i2c_smbus_write_byte_data(client, lm90_write_reg_addr(reg), val);
}

-static int lm90_read16(struct i2c_client *client, u8 regh, u8 regl)
+static int lm90_read16(struct i2c_client *client, u8 regh, u8 regl,
+ bool is_volatile)
{
int oldh, newh, l;

- /*
- * There is a trick here. We have to read two registers to have the
- * sensor temperature, but we have to beware a conversion could occur
- * between the readings. The datasheet says we should either use
- * the one-shot conversion register, which we don't want to do
- * (disables hardware monitoring) or monitor the busy bit, which is
- * impossible (we can't read the values and monitor that bit at the
- * exact same time). So the solution used here is to read the high
- * byte once, then the low byte, then the high byte again. If the new
- * high byte matches the old one, then we have a valid reading. Else
- * we have to read the low byte again, and now we believe we have a
- * correct reading.
- */
oldh = lm90_read_reg(client, regh);
if (oldh < 0)
return oldh;
l = lm90_read_reg(client, regl);
if (l < 0)
return l;
+
+ if (!is_volatile)
+ return (oldh << 8) | l;
+
+ /*
+ * For volatile registers we have to use a trick.
+ * We have to read two registers to have the sensor temperature,
+ * but we have to beware a conversion could occur between the
+ * readings. The datasheet says we should either use
+ * the one-shot conversion register, which we don't want to do
+ * (disables hardware monitoring) or monitor the busy bit, which is
+ * impossible (we can't read the values and monitor that bit at the
+ * exact same time). So the solution used here is to read the high
+ * the high byte again. If the new high byte matches the old one,
+ * then we have a valid reading. Otherwise we have to read the low
+ * byte again, and now we believe we have a correct reading.
+ */
newh = lm90_read_reg(client, regh);
if (newh < 0)
return newh;
@@ -766,7 +771,7 @@ static int lm90_update_limits(struct device *dev)

if (data->flags & LM90_HAVE_OFFSET) {
val = lm90_read16(client, LM90_REG_REMOTE_OFFSH,
- LM90_REG_REMOTE_OFFSL);
+ LM90_REG_REMOTE_OFFSL, false);
if (val < 0)
return val;
data->temp11[REMOTE_OFFSET] = val;
@@ -999,7 +1004,7 @@ static int lm90_update_device(struct device *dev)

if (data->reg_local_ext) {
val = lm90_read16(client, LM90_REG_LOCAL_TEMP,
- data->reg_local_ext);
+ data->reg_local_ext, true);
if (val < 0)
return val;
data->temp11[LOCAL_TEMP] = val;
@@ -1010,7 +1015,7 @@ static int lm90_update_device(struct device *dev)
data->temp11[LOCAL_TEMP] = val << 8;
}
val = lm90_read16(client, LM90_REG_REMOTE_TEMPH,
- LM90_REG_REMOTE_TEMPL);
+ LM90_REG_REMOTE_TEMPL, true);
if (val < 0)
return val;
data->temp11[REMOTE_TEMP] = val;
@@ -1021,7 +1026,7 @@ static int lm90_update_device(struct device *dev)
return val;

val = lm90_read16(client, LM90_REG_REMOTE_TEMPH,
- LM90_REG_REMOTE_TEMPL);
+ LM90_REG_REMOTE_TEMPL, true);
if (val < 0) {
lm90_select_remote_channel(data, 0);
return val;
--
2.35.1