hwmon: tmp401: support for TMP435 and fix for a probe issue
cover hwmon: tmp401: support for TMP435 and fix for a probe issue
This series adds support for the TI TMP435 chip and fixes an issue similar to the one we observed for ina2xx - ignoring the write errors when probing the device.
Bartosz Golaszewski (1):
hwmon: tmp401: bail-out from tmp401_probe() in case of write errors
Patrick Titiano (1):
hwmon: tmp401: add support for TI TMP435
drivers/hwmon/tmp401.c | 32 ++++++++++++++++++++++++-------- 1 file changed, 24 insertions(+), 8 deletions(-)
Signature · 7 lines
2.1.3 _______________________________________________ lm-sensors mailing list [email protected] http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
PATCH 1/2 hwmon: tmp401: add support for TI TMP435
From: Patrick Titiano <[email protected]>
--- drivers/hwmon/tmp401.c | 16 ++++++++++++++-- 1 file changed, 14 insertions(+), 2 deletions(-)
drivers/hwmon/tmp401.c +14 -2
@@ -8,6 +8,9 @@* Cleanup and support for TMP431 and TMP432 by Guenter Roeck * Copyright (c) 2013 Guenter Roeck <[email protected]> * +* Support for TMP435 by Patrick Titiano +* Copyright (c) 2014 Patrick Titiano <[email protected]> +* * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or
@@ -46,7 +49,7 @@/* Addresses to scan */ static const unsigned short normal_i2c[] = {0x4c ,0x4d ,0x4e ,I2C_CLIENT_END }; -enum chips { tmp401, tmp411, tmp431, tmp432 }; +enum chips { tmp401, tmp411, tmp431, tmp432, tmp435 };/* * The TMP401 registers, note some registers have different addresses for
@@ -136,6 +139,7 @@ static const u8 TMP432_STATUS_REG[] = {#define TMP411C_DEVICE_ID 0x10#define TMP431_DEVICE_ID 0x31#define TMP432_DEVICE_ID 0x32 +#define TMP435_DEVICE_ID 0x35/* * Driver data (common to all clients)
@@ -146,6 +150,7 @@ static const struct i2c_device_id tmp401_id[] = { { "tmp411", tmp411 }, { "tmp431", tmp431 }, { "tmp432", tmp432 }, + { "tmp435", tmp435 }, { } };MODULE_DEVICE_TABLE (i2c, tmp401_id);
@@ -684,6 +689,11 @@ static int tmp401_detect(struct i2c_client *client,return -ENODEV ; kind = tmp432;break ; +case TMP435_DEVICE_ID : +if (client->addr !=0x4c ) +return -ENODEV ; + kind = tmp435; +break ;default :return -ENODEV ; }
@@ -705,7 +715,9 @@ static int tmp401_detect(struct i2c_client *client,static int tmp401_probe (struct i2c_client *client,const struct i2c_device_id *id) { -const char *names[] = {"TMP401" ,"TMP411" ,"TMP431" ,"TMP432" }; +static const char *names[] = { +"TMP401" ,"TMP411" ,"TMP431" ,"TMP432" ,"TMP435" + };struct device *dev = &client->dev;struct device *hwmon_dev;struct tmp401_data *data;
Signature · 7 lines
2.1.3 _______________________________________________ lm-sensors mailing list [email protected] http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
Guenter Roeck
maintainerRe: [PATCH 1/2] hwmon: tmp401: add support for TI TMP435
Quoted 6 lines · Bartosz Golaszewski
From: Patrick Titiano <[email protected]>
Signed-off-by Patrick Titiano <[email protected]>Signed-off-by Bartosz Golaszewski <[email protected]>--- drivers/hwmon/tmp401.c | 16 ++++++++++++++--
Please also update Kconfig and Documentation/hwmon/tmp401.
Quoted 26 lines
1 file changed, 14 insertions(+), 2 deletions(-)drivers/hwmon/tmp401.c +4
-1@@ -8,6 +8,9 @@* Cleanup and support for TMP431 and TMP432 by Guenter Roeck * Copyright (c) 2013 Guenter Roeck <[email protected]> * +* Support for TMP435 by Patrick Titiano +* Copyright (c) 2014 Patrick Titiano <[email protected]> +* * This program is free software; you can redistribute it and/or modify * it under the terms of the GNU General Public License as published by * the Free Software Foundation; either version 2 of the License, or @@ -46,7 +49,7 @@/* Addresses to scan */ static const unsigned short normal_i2c[] = {0x4c ,0x4d ,0x4e ,I2C_CLIENT_END }; -enum chips { tmp401, tmp411, tmp431, tmp432 }; +enum chips { tmp401, tmp411, tmp431, tmp432, tmp435 };/* * The TMP401 registers, note some registers have different addresses for
drivers/hwmon/tmp401.c static const u8 TMP432_STATUS_REG[] = { · line 142
Quoted 4 lines
drivers/hwmon/tmp401.c +1
-0@@ -136,6 +139,7 @@ static const u8 TMP432_STATUS_REG[] = {#define TMP411C_DEVICE_ID 0x10#define TMP431_DEVICE_ID 0x31#define TMP432_DEVICE_ID 0x32 +#define TMP435_DEVICE_ID 0x35
The datasheet claims otherwise (0x31, specifically). Is this based on real chip information ?
Can you send me the output from i2cdump for the chip ?
Thanks,
Guenter
Quoted 25 lines
drivers/hwmon/tmp401.c +6
-0/* * Driver data (common to all clients) @@ -146,6 +150,7 @@ static const struct i2c_device_id tmp401_id[] = { { "tmp411", tmp411 }, { "tmp431", tmp431 }, { "tmp432", tmp432 }, + { "tmp435", tmp435 }, { } };MODULE_DEVICE_TABLE (i2c, tmp401_id);@@ -684,6 +689,11 @@ static int tmp401_detect(struct i2c_client *client,return -ENODEV ; kind = tmp432;break ; +case TMP435_DEVICE_ID : +if (client->addr !=0x4c ) +return -ENODEV ; + kind = tmp435; +break ;default :return -ENODEV ; }
drivers/hwmon/tmp401.c static int tmp401_detect(struct i2c_client *client, · line 723
Quoted 10 lines
drivers/hwmon/tmp401.c +3
-1@@ -705,7 +715,9 @@ static int tmp401_detect(struct i2c_client *client,static int tmp401_probe (struct i2c_client *client,const struct i2c_device_id *id) { -const char *names[] = {"TMP401" ,"TMP411" ,"TMP431" ,"TMP432" }; +static const char *names[] = { +"TMP401" ,"TMP411" ,"TMP431" ,"TMP432" ,"TMP435" + };struct device *dev = &client->dev;struct device *hwmon_dev;struct tmp401_data *data;
_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
Re: [PATCH 1/2] hwmon: tmp401: add support for TI TMP435
2014-12-04 15:19 GMT+01:00 Guenter Roeck <[email protected]>:
Quoted 4 lines
The datasheet claims otherwise (
0x31, specifically). Is this based on real chip information ?
Can you send me the output from i2cdump for the chip ?
This is what I'm reading:
# i2cdump -y 1 0x4c
No size specified (using byte-data access)
0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef
00: 1a 1b 00 00 05 55 00 55 00 00 05 55 00 55 00 0d ??..?U.U..?U.U.?
10: 70 00 00 00 00 90 00 00 00 55 1c XX XX XX XX 00 p....?...U?XXXX.
20: 55 0a 70 XX XX 0f XX XX XX XX XX XX XX XX XX XX U?pXX?XXXXXXXXXX
30: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
40: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
50: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
60: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
70: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
80: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
90: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
a0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
b0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
c0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
d0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
e0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX
f0: XX XX XX XX XX XX XX XX XX XX XX XX 00 35 55 35 XXXXXXXXXXXX.5U5# i2cget -y 1 0x4c 0xff b 0x35
Bart
_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
Re: [lm-sensors] [PATCH 1/2] hwmon: tmp401: add support for TI TMP435
Quoted 28 lines · Bartosz Golaszewski
2014-12-04 15:19 GMT+01:00 Guenter Roeck <[email protected]>:
Quoted 4 lines
The datasheet claims otherwise (
0x31, specifically). Is this based on real chip information ?
Can you send me the output from i2cdump for the chip ?This is what I'm reading:
# i2cdump -y 1 0x4c No size specified (using byte-data access) 0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef 00: 1a 1b 00 00 05 55 00 55 00 00 05 55 00 55 00 0d ??..?U.U..?U.U.? 10: 70 00 00 00 00 90 00 00 00 55 1c XX XX XX XX 00 p....?...U?XXXX. 20: 55 0a 70 XX XX 0f XX XX XX XX XX XX XX XX XX XX U?pXX?XXXXXXXXXX 30: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 40: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 50: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 60: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 70: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 80: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 90: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX a0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX b0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX c0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX d0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX e0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX f0: XX XX XX XX XX XX XX XX XX XX XX XX 00 35 55 35 XXXXXXXXXXXX.5U5
Quoted 2 lines
# i2cget -y 1 0x4c 0xff b 0x35
That makes sense for a chip named TMP435. Looks like TI forgot to update the value when copying the datasheet from TMP431 to TMP435.
Thanks for the dump, by the way :)
Signature · 7 lines
Jean Delvare SUSE L3 Support _______________________________________________ lm-sensors mailing list [email protected] http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
Guenter Roeck
maintainerRe: [lm-sensors] [PATCH 1/2] hwmon: tmp401: add support for TI TMP435
Quoted 29 lines · Jean Delvare
Quoted 28 lines · Bartosz Golaszewski
2014-12-04 15:19 GMT+01:00 Guenter Roeck <[email protected]>:
Quoted 4 lines
The datasheet claims otherwise (
0x31, specifically). Is this based on real chip information ?
Can you send me the output from i2cdump for the chip ?This is what I'm reading:
# i2cdump -y 1 0x4c No size specified (using byte-data access) 0 1 2 3 4 5 6 7 8 9 a b c d e f 0123456789abcdef 00: 1a 1b 00 00 05 55 00 55 00 00 05 55 00 55 00 0d ??..?U.U..?U.U.? 10: 70 00 00 00 00 90 00 00 00 55 1c XX XX XX XX 00 p....?...U?XXXX. 20: 55 0a 70 XX XX 0f XX XX XX XX XX XX XX XX XX XX U?pXX?XXXXXXXXXX 30: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 40: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 50: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 60: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 70: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 80: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX 90: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX a0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX b0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX c0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX d0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX e0: XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XX XXXXXXXXXXXXXXXX f0: XX XX XX XX XX XX XX XX XX XX XX XX 00 35 55 35 XXXXXXXXXXXX.5U5
Quoted 6 lines
Quoted 2 lines
# i2cget -y 1 0x4c 0xff b 0x35That makes sense for a chip named TMP435. Looks like TI forgot to update the value when copying the datasheet from TMP431 to TMP435.
Guess so.
I'll make the necessary updates to sensors-detect and the web page.
Quoted 2 lines
Thanks for the dump, by the way :)
Same here .. that also helps for writing module test scripts.
Thanks,
Guenter
_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
PATCH 2/2 hwmon: tmp401: bail-out from tmp401_probe() in case of write errors
In reply to Bartosz GolaszewskiThe return value of i2c_smbus_read_byte_data() is checked in tmp401_init_client(), but only a warning is printed and the device is registered anyway. This leads to devices being registered even if they cannot be physically detected.
Bail-out from probe in case of write errors and notify the user.
--- drivers/hwmon/tmp401.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-)
drivers/hwmon/tmp401.c +10 -6
@@ -618,8 +618,8 @@ static const struct attribute_group tmp432_group = {* Begin non sysfs callback code (aka Real code) */ -static void tmp401_init_client (struct tmp401_data *data, -struct i2c_client *client) +static int tmp401_init_client (struct tmp401_data *data, +struct i2c_client *client) {int config, config_orig;
@@ -630,8 +630,8 @@ static void tmp401_init_client(struct tmp401_data *data,/* Start conversions (disable shutdown if necessary) */ config =i2c_smbus_read_byte_data (client,TMP401_CONFIG_READ );if (config <0 ) { -dev_warn (&client->dev,"Initialization failed!\n" ); -return ; +dev_err (&client->dev,"Initialization failed!" ); +return config; } config_orig = config;
@@ -639,6 +639,8 @@ static void tmp401_init_client(struct tmp401_data *data,if (config != config_orig)i2c_smbus_write_byte_data (client,TMP401_CONFIG_WRITE , config); + +return 0 ; } static int tmp401_detect(struct i2c_client *client,
@@ -721,7 +723,7 @@ static int tmp401_probe(struct i2c_client *client,struct device *dev = &client->dev;struct device *hwmon_dev;struct tmp401_data *data; -int groups =0 ; +int groups =0 , status; data =devm_kzalloc (dev,sizeof (struct tmp401_data ),GFP_KERNEL );if (!data)
@@ -732,7 +734,9 @@ static int tmp401_probe(struct i2c_client *client, data->kind = id->driver_data;/* Initialize the TMP401 chip */ -tmp401_init_client (data, client); + status =tmp401_init_client (data, client); +if (status <0 ) +return status;/* Register sysfs hooks */ data->groups[groups++] = &tmp401_group;
Signature · 7 lines
2.1.3 _______________________________________________ lm-sensors mailing list [email protected] http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
Guenter Roeck
maintainerRe: [PATCH 2/2] hwmon: tmp401: bail-out from tmp401_probe() in case of write errors
Quoted 39 lines · Bartosz Golaszewski
The return value of
i2c_smbus_read_byte_data()is checked intmp401_init_client(), but only a warning is printed and the device is registered anyway. This leads to devices being registered even if they cannot be physically detected.Bail-out from probe in case of write errors and notify the user.
Signed-off-by Bartosz Golaszewski <[email protected]>--- drivers/hwmon/tmp401.c | 16 ++++++++++------ 1 file changed, 10 insertions(+), 6 deletions(-)drivers/hwmon/tmp401.c +4
-4@@ -618,8 +618,8 @@ static const struct attribute_group tmp432_group = {* Begin non sysfs callback code (aka Real code) */ -static void tmp401_init_client (struct tmp401_data *data, -struct i2c_client *client) +static int tmp401_init_client (struct tmp401_data *data, +struct i2c_client *client) {int config, config_orig;@@ -630,8 +630,8 @@ static void tmp401_init_client(struct tmp401_data *data,/* Start conversions (disable shutdown if necessary) */ config =i2c_smbus_read_byte_data (client,TMP401_CONFIG_READ );if (config <0 ) { -dev_warn (&client->dev,"Initialization failed!\n" ); -return ; +dev_err (&client->dev,"Initialization failed!" ); +return config; } config_orig = config;
drivers/hwmon/tmp401.c static void tmp401_init_client(struct tmp401_data *data, · line 641
Quoted 3 lines
drivers/hwmon/tmp401.c +0
-0@@ -639,6 +639,8 @@ static void tmp401_init_client(struct tmp401_data *data,if (config != config_orig)i2c_smbus_write_byte_data (client,TMP401_CONFIG_WRITE , config);
No error check and return here ?
Guenter
Quoted 16 lines
drivers/hwmon/tmp401.c +3
-1+ +return 0 ; } static int tmp401_detect(struct i2c_client *client,@@ -721,7 +723,7 @@ static int tmp401_probe(struct i2c_client *client,struct device *dev = &client->dev;struct device *hwmon_dev;struct tmp401_data *data; -int groups =0 ; +int groups =0 , status; data =devm_kzalloc (dev,sizeof (struct tmp401_data ),GFP_KERNEL );if (!data)
drivers/hwmon/tmp401.c static int tmp401_probe(struct i2c_client *client, · line 742
Quoted 10 lines
drivers/hwmon/tmp401.c +3
-1@@ -732,7 +734,9 @@ static int tmp401_probe(struct i2c_client *client, data->kind = id->driver_data;/* Initialize the TMP401 chip */ -tmp401_init_client (data, client); + status =tmp401_init_client (data, client); +if (status <0 ) +return status;/* Register sysfs hooks */ data->groups[groups++] = &tmp401_group;
_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors