All of lore.kernel.org
 help / color / mirror / Atom feed
From: Guenter Roeck <linux@roeck-us.net>
To: linux-hwmon@vger.kernel.org
Cc: linux-kernel@vger.kernel.org, Jean Delvare <jdelvare@suse.com>,
	Slawomir Stepien <sst@poczta.fm>,
	Guenter Roeck <linux@roeck-us.net>
Subject: [PATCH 12/40] hwmon: (lm90) Only re-read registers if volatile
Date: Wed, 25 May 2022 06:57:30 -0700	[thread overview]
Message-ID: <20220525135758.2944744-13-linux@roeck-us.net> (raw)
In-Reply-To: <20220525135758.2944744-1-linux@roeck-us.net>

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@roeck-us.net>
---
 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


  parent reply	other threads:[~2022-05-25 13:58 UTC|newest]

Thread overview: 44+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-05-25 13:57 [PATCH 00/40] hwmon: (lm90) Various improvements to lm90 driver Guenter Roeck
2022-05-25 13:57 ` [PATCH 01/40] hwmon: (lm90) Generate sysfs and udev events for all alarms Guenter Roeck
2022-05-25 13:57 ` [PATCH 02/40] hwmon: (lm90) Rework alarm/status handling Guenter Roeck
2022-05-25 13:57 ` [PATCH 03/40] hwmon: (lm90) Reorder include files in alphabetical order Guenter Roeck
2022-05-25 13:57 ` [PATCH 04/40] hwmon: (lm90) Reorder chip enumeration to be " Guenter Roeck
2022-05-25 13:57 ` [PATCH 05/40] hwmon: (lm90) Use BIT macro Guenter Roeck
2022-05-25 13:57 ` [PATCH 06/40] hwmon: (lm90) Move status register bit shifts to compile time Guenter Roeck
2022-05-25 13:57 ` [PATCH 07/40] hwmon: (lm90) Stop using R_/W_ register prefix Guenter Roeck
2022-05-25 13:57 ` [PATCH 08/40] hwmon: (lm90) Improve PEC support Guenter Roeck
2022-05-25 13:57 ` [PATCH 09/40] hwmon: (lm90) Add partial PEC support for ADT7461 Guenter Roeck
2022-05-25 13:57 ` [PATCH 10/40] hwmon: (lm90) Enable full PEC support for ADT7461A Guenter Roeck
2022-05-25 13:57 ` [PATCH 11/40] hwmon: (lm90) Add support for unsigned and signed temperatures Guenter Roeck
2022-05-25 13:57 ` Guenter Roeck [this message]
2022-05-25 13:57 ` [PATCH 13/40] hwmon: (lm90) Support multiple temperature resolutions Guenter Roeck
2022-05-26  7:12   ` Slawomir Stepien
2022-05-26 14:32     ` Guenter Roeck
2022-05-25 13:57 ` [PATCH 14/40] hwmon: (lm90) Use single flag to indicate extended temperature support Guenter Roeck
2022-05-25 13:57 ` [PATCH 15/40] hwmon: (lm90) Rework detect function Guenter Roeck
2022-05-25 13:57 ` [PATCH 16/40] hwmon: (lm90) Add support for additional chip revision of NCT1008 Guenter Roeck
2022-05-25 13:57 ` [PATCH 17/40] hwmon: (lm90) Fix/Add detection of G781-1 Guenter Roeck
2022-05-25 13:57 ` [PATCH 18/40] hwmon: (lm90) Add flag to indicate 'alarms' attribute support Guenter Roeck
2022-05-25 13:57 ` [PATCH 19/40] hwmon: (lm90) Add explicit support for MAX6648/MAX6692 Guenter Roeck
2022-05-25 13:57 ` [PATCH 20/40] hwmon: (lm90) Add support for ADT7481, ADT7482, and ADT7483 Guenter Roeck
2022-05-27  5:08   ` Slawomir Stepien
2022-05-25 13:57 ` [PATCH 21/40] hwmon: (lm90) Strengthen chip detection for ADM1032, ADT7461(A), and NCT1008 Guenter Roeck
2022-05-25 13:57 ` [PATCH 22/40] hwmon: (lm90) Add support for MAX6690 Guenter Roeck
2022-05-25 13:57 ` [PATCH 23/40] hwmon: (lm90) Add flag to indicate support for minimum temperature limits Guenter Roeck
2022-05-25 13:57 ` [PATCH 24/40] hwmon: (lm90) Add flag to indicate conversion rate support Guenter Roeck
2022-05-25 13:57 ` [PATCH 25/40] hwmon: (lm90) Add support for MAX6642 Guenter Roeck
2022-05-25 13:57 ` [PATCH 26/40] hwmon: (lm90) Let lm90_read16() handle 8-bit read operations Guenter Roeck
2022-05-25 13:57 ` [PATCH 27/40] hwmon: (lm90) Introduce 16-bit register write function Guenter Roeck
2022-05-25 13:57 ` [PATCH 28/40] hwmon: (lm90) Support MAX1617 and LM84 Guenter Roeck
2022-05-25 13:57 ` [PATCH 29/40] hwmon: (lm90) Add support for ADM1021, ADM1021A, and ADM1023 Guenter Roeck
2022-05-25 13:57 ` [PATCH 30/40] hwmon: (lm90) Add remaining chips supported by adm1021 driver Guenter Roeck
2022-05-25 13:57 ` [PATCH 31/40] hwmon: (lm90) Combine lm86 and lm90 configuration Guenter Roeck
2022-05-25 13:57 ` [PATCH 32/40] hwmon: (lm90) Add explicit support for NCT210 Guenter Roeck
2022-05-25 13:57 ` [PATCH 33/40] hwmon: (lm90) Add support for ON Semiconductor NCT214 and NCT72 Guenter Roeck
2022-05-25 13:57 ` [PATCH 34/40] hwmon: (lm90) Add support for ON Semiconductor NCT218 Guenter Roeck
2022-05-25 13:57 ` [PATCH 35/40] hwmon: (lm90) Add support for ADT7421 Guenter Roeck
2022-05-25 13:57 ` [PATCH 36/40] hwmon: (lm90) Only disable alerts if not already disabled Guenter Roeck
2022-05-25 13:57 ` [PATCH 37/40] hwmon: (lm90) Add explicit support for ADM1020 Guenter Roeck
2022-05-25 13:57 ` [PATCH 38/40] hwmon: (lm90) Add support and detection of Philips/NXP NE1618 Guenter Roeck
2022-05-25 13:57 ` [PATCH 39/40] hwmon: (lm90) Add table with supported Analog/ONSEMI devices Guenter Roeck
2022-05-25 13:57 ` [PATCH 40/40] hwmon: (lm90) Support temp_samples attribute Guenter Roeck

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20220525135758.2944744-13-linux@roeck-us.net \
    --to=linux@roeck-us.net \
    --cc=jdelvare@suse.com \
    --cc=linux-hwmon@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=sst@poczta.fm \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is an external index of several public inboxes,
see mirroring instructions on how to clone and mirror
all data and code used by this external index.