Add a thermal_read_mode_check helper function to make the code
simpler during init.
This helps particularly when the new TPEC_12 mode is added in
the next patch.
Suggested-by: Ilpo Jarvinen <[email protected]>
Signed-off-by: Vishnu Sankar <[email protected]>
---
drivers/platform/x86/thinkpad_acpi.c | 136 +++++++++++++--------------
1 file changed, 66 insertions(+), 70 deletions(-)
diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c
index c4895e9bc714..2428c8bd0fa2 100644
--- a/drivers/platform/x86/thinkpad_acpi.c
+++ b/drivers/platform/x86/thinkpad_acpi.c
@@ -6147,6 +6147,71 @@ struct ibm_thermal_sensors_struct {
static enum thermal_access_mode thermal_read_mode;
static bool thermal_use_labels;
+/* Function to check thermal read mode */
+static enum thermal_access_mode thermal_read_mode_check(void)
+{
+ u8 t, ta1, ta2, ver = 0;
+ int i;
+
+ if (thinkpad_id.ec_model) {
+ /*
+ * Direct EC access mode: sensors at registers
+ * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
+ * non-implemented, thermal sensors return 0x80 when
+ * not available
+ * The above rule is unfortunately flawed. This has been seen with
+ * 0xC2 (power supply ID) causing thermal control problems.
+ * The EC version can be determined by offset 0xEF and at least for
+ * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
+ * are not thermal registers.
+ */
+ if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
+ pr_warn("Thinkpad ACPI EC unable to access EC version\n");
+
+ ta1 = ta2 = 0;
+ for (i = 0; i < 8; i++) {
+ if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
+ ta1 |= t;
+ } else {
+ ta1 = 0;
+ break;
+ }
+ if (ver < 3) {
+ if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
+ ta2 |= t;
+ } else {
+ ta1 = 0;
+ break;
+ }
+ }
+ }
+
+ if (ta1 == 0) {
+ pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
+ return TPACPI_THERMAL_NONE;
+ }
+
+ if (ver >= 3) {
+ thermal_use_labels = true;
+ return TPACPI_THERMAL_TPEC_8;
+ }
+
+ return (ta2 != 0) ? TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
+ }
+
+ if (acpi_evalf(ec_handle, NULL, "TMP7", "qv")) {
+ if (tpacpi_is_ibm() &&
+ acpi_evalf(ec_handle, NULL, "UPDT", "qv"))
+ /* 600e/x, 770e, 770x */
+ return TPACPI_THERMAL_ACPI_UPDT;
+ /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
+ return TPACPI_THERMAL_ACPI_TMP07;
+ }
+
+ /* temperatures not supported on 570, G4x, R30, R31, R32 */
+ return TPACPI_THERMAL_NONE;
+}
+
/* idx is zero-based */
static int thermal_get_sensor(int idx, s32 *value)
{
@@ -6375,78 +6440,9 @@ static const struct attribute_group temp_label_attr_group = {
static int __init thermal_init(struct ibm_init_struct *iibm)
{
- u8 t, ta1, ta2, ver = 0;
- int i;
- int acpi_tmp7;
-
vdbg_printk(TPACPI_DBG_INIT, "initializing thermal subdriver\n");
- acpi_tmp7 = acpi_evalf(ec_handle, NULL, "TMP7", "qv");
-
- if (thinkpad_id.ec_model) {
- /*
- * Direct EC access mode: sensors at registers
- * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
- * non-implemented, thermal sensors return 0x80 when
- * not available
- * The above rule is unfortunately flawed. This has been seen with
- * 0xC2 (power supply ID) causing thermal control problems.
- * The EC version can be determined by offset 0xEF and at least for
- * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
- * are not thermal registers.
- */
- if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
- pr_warn("Thinkpad ACPI EC unable to access EC version\n");
-
- ta1 = ta2 = 0;
- for (i = 0; i < 8; i++) {
- if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
- ta1 |= t;
- } else {
- ta1 = 0;
- break;
- }
- if (ver < 3) {
- if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
- ta2 |= t;
- } else {
- ta1 = 0;
- break;
- }
- }
- }
- if (ta1 == 0) {
- /* This is sheer paranoia, but we handle it anyway */
- if (acpi_tmp7) {
- pr_err("ThinkPad ACPI EC access misbehaving, falling back to ACPI TMPx access mode\n");
- thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
- } else {
- pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
- thermal_read_mode = TPACPI_THERMAL_NONE;
- }
- } else {
- if (ver >= 3) {
- thermal_read_mode = TPACPI_THERMAL_TPEC_8;
- thermal_use_labels = true;
- } else {
- thermal_read_mode =
- (ta2 != 0) ?
- TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
- }
- }
- } else if (acpi_tmp7) {
- if (tpacpi_is_ibm() &&
- acpi_evalf(ec_handle, NULL, "UPDT", "qv")) {
- /* 600e/x, 770e, 770x */
- thermal_read_mode = TPACPI_THERMAL_ACPI_UPDT;
- } else {
- /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
- thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
- }
- } else {
- /* temperatures not supported on 570, G4x, R30, R31, R32 */
- thermal_read_mode = TPACPI_THERMAL_NONE;
- }
+ thermal_read_mode = thermal_read_mode_check();
vdbg_printk(TPACPI_DBG_INIT, "thermal is %s, mode %d\n",
str_supported(thermal_read_mode != TPACPI_THERMAL_NONE),
--
2.34.1
Added non-standard thermal register support for some ThinkPads.
Some of the Thinkpads use a non-standard ECFW which use different
thermal register addresses.
This is a fix to correct the wrong temperature reporting on
those systems.
Tested on Lenovo ThinkPad L13 Yoga Gen2
Suggested-by: Mark Pearson <[email protected]>
Signed-off-by: Vishnu Sankar <[email protected]>
---
-Improvements as requested.
-Improved the readability in case TPACPI_THERMAL_TPEC_12.
-idx < 8 from idx idx <=7 to match idx = 8
-KILO used from linux/units.h instead of 1000.
---
---
drivers/platform/x86/thinkpad_acpi.c | 74 +++++++++++++++++++++++++---
1 file changed, 68 insertions(+), 6 deletions(-)
diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c
index 2428c8bd0fa2..9be114572f17 100644
--- a/drivers/platform/x86/thinkpad_acpi.c
+++ b/drivers/platform/x86/thinkpad_acpi.c
@@ -69,6 +69,7 @@
#include <linux/sysfs.h>
#include <linux/types.h>
#include <linux/uaccess.h>
+#include <linux/units.h>
#include <linux/workqueue.h>
#include <acpi/battery.h>
@@ -6126,12 +6127,15 @@ enum thermal_access_mode {
TPACPI_THERMAL_ACPI_TMP07, /* Use ACPI TMP0-7 */
TPACPI_THERMAL_ACPI_UPDT, /* Use ACPI TMP0-7 with UPDT */
TPACPI_THERMAL_TPEC_8, /* Use ACPI EC regs, 8 sensors */
+ TPACPI_THERMAL_TPEC_12, /* Use ACPI EC regs, 12 sensors */
TPACPI_THERMAL_TPEC_16, /* Use ACPI EC regs, 16 sensors */
};
enum { /* TPACPI_THERMAL_TPEC_* */
TP_EC_THERMAL_TMP0 = 0x78, /* ACPI EC regs TMP 0..7 */
TP_EC_THERMAL_TMP8 = 0xC0, /* ACPI EC regs TMP 8..15 */
+ TP_EC_THERMAL_TMP0_NS = 0xA8, /* ACPI EC Non-Standard regs TMP 0..7 */
+ TP_EC_THERMAL_TMP8_NS = 0xB8, /* ACPI EC Non-standard regs TMP 8..11 */
TP_EC_FUNCREV = 0xEF, /* ACPI EC Functional revision */
TP_EC_THERMAL_TMP_NA = -128, /* ACPI EC sensor not available */
@@ -6144,8 +6148,22 @@ struct ibm_thermal_sensors_struct {
s32 temp[TPACPI_MAX_THERMAL_SENSORS];
};
+static const struct tpacpi_quirk thermal_quirk_table[] __initconst = {
+ /* Non-standard address for thermal registers on some ThinkPads */
+ TPACPI_Q_LNV3('R', '1', 'F', true), /* L13 Yoga Gen 2 */
+ TPACPI_Q_LNV3('N', '2', 'U', true), /* X13 Yoga Gen 2*/
+ TPACPI_Q_LNV3('R', '0', 'R', true), /* L380 */
+ TPACPI_Q_LNV3('R', '1', '5', true), /* L13 Yoga Gen 1*/
+ TPACPI_Q_LNV3('R', '1', '0', true), /* L390 */
+ TPACPI_Q_LNV3('N', '2', 'L', true), /* X13 Yoga Gen 1*/
+ TPACPI_Q_LNV3('R', '0', 'T', true), /* 11e Gen5 GL*/
+ TPACPI_Q_LNV3('R', '1', 'D', true), /* 11e Gen5 GL-R*/
+ TPACPI_Q_LNV3('R', '0', 'V', true), /* 11e Gen5 KL-Y*/
+};
+
static enum thermal_access_mode thermal_read_mode;
static bool thermal_use_labels;
+static bool thermal_with_ns_address; /*Non-standard thermal reg address*/
/* Function to check thermal read mode */
static enum thermal_access_mode thermal_read_mode_check(void)
@@ -6168,6 +6186,16 @@ static enum thermal_access_mode thermal_read_mode_check(void)
if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
pr_warn("Thinkpad ACPI EC unable to access EC version\n");
+ /* Quirks to check non-standard EC */
+ thermal_with_ns_address = tpacpi_check_quirks(thermal_quirk_table,
+ ARRAY_SIZE(thermal_quirk_table));
+
+ /* Support for Thinkpads with non-standard address */
+ if (thermal_with_ns_address) {
+ pr_info("ECFW with non-standard thermal registers found\n");
+ return TPACPI_THERMAL_TPEC_12;
+ }
+
ta1 = ta2 = 0;
for (i = 0; i < 8; i++) {
if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
@@ -6239,6 +6267,20 @@ static int thermal_get_sensor(int idx, s32 *value)
}
break;
+ /* The Non-standard EC uses 12 Thermal areas */
+ case TPACPI_THERMAL_TPEC_12:
+ if (idx >= 12)
+ return -EINVAL;
+
+ t = idx < 8 ? TP_EC_THERMAL_TMP0_NS + idx :
+ TP_EC_THERMAL_TMP8_NS + (idx - 8);
+
+ if (!acpi_ec_read(t, &tmp))
+ return -EIO;
+
+ *value = tmp * KILO;
+ return 0;
+
case TPACPI_THERMAL_ACPI_UPDT:
if (idx <= 7) {
snprintf(tmpi, sizeof(tmpi), "TMP%c", '0' + idx);
@@ -6284,6 +6326,8 @@ static int thermal_get_sensors(struct ibm_thermal_sensors_struct *s)
if (thermal_read_mode == TPACPI_THERMAL_TPEC_16)
n = 16;
+ else if (thermal_read_mode == TPACPI_THERMAL_TPEC_12)
+ n = 12;
for (i = 0 ; i < n; i++) {
res = thermal_get_sensor(i, &s->temp[i]);
@@ -6382,18 +6426,36 @@ static struct attribute *thermal_temp_input_attr[] = {
NULL
};
+#define to_dev_attr(_attr) container_of(_attr, struct device_attribute, attr)
+
static umode_t thermal_attr_is_visible(struct kobject *kobj,
struct attribute *attr, int n)
{
- if (thermal_read_mode == TPACPI_THERMAL_NONE)
+ struct device_attribute *dev_attr = to_dev_attr(attr);
+ struct sensor_device_attribute *sensor_attr =
+ to_sensor_dev_attr(dev_attr);
+
+ int idx = sensor_attr->index;
+
+ switch (thermal_read_mode) {
+ case TPACPI_THERMAL_NONE:
return 0;
- if (attr == THERMAL_ATTRS(8) || attr == THERMAL_ATTRS(9) ||
- attr == THERMAL_ATTRS(10) || attr == THERMAL_ATTRS(11) ||
- attr == THERMAL_ATTRS(12) || attr == THERMAL_ATTRS(13) ||
- attr == THERMAL_ATTRS(14) || attr == THERMAL_ATTRS(15)) {
- if (thermal_read_mode != TPACPI_THERMAL_TPEC_16)
+ case TPACPI_THERMAL_ACPI_TMP07:
+ case TPACPI_THERMAL_ACPI_UPDT:
+ case TPACPI_THERMAL_TPEC_8:
+ if (idx >= 8)
return 0;
+ break;
+
+ case TPACPI_THERMAL_TPEC_12:
+ if (idx >= 12)
+ return 0;
+ break;
+
+ default:
+ break;
+
}
return attr->mode;
--
2.34.1
On Wed, 14 Feb 2024, Vishnu Sankar wrote:
Thanks for doing this, it's major improvement to readability already as
is, and even more of after the second patch!!
> Add a thermal_read_mode_check helper function to make the code
thermal_read_mode_check()
remove "function" as it's obvious.
> simpler during init.
> This helps particularly when the new TPEC_12 mode is added in
> the next patch.
Flow the paragraph normally without the premature line break.
> Suggested-by: Ilpo Jarvinen <[email protected]>
This is not my email address, please use
Suggested-by: Ilpo J?rvinen <[email protected]>
> Signed-off-by: Vishnu Sankar <[email protected]>
> ---
> drivers/platform/x86/thinkpad_acpi.c | 136 +++++++++++++--------------
> 1 file changed, 66 insertions(+), 70 deletions(-)
>
> diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c
> index c4895e9bc714..2428c8bd0fa2 100644
> --- a/drivers/platform/x86/thinkpad_acpi.c
> +++ b/drivers/platform/x86/thinkpad_acpi.c
> @@ -6147,6 +6147,71 @@ struct ibm_thermal_sensors_struct {
> static enum thermal_access_mode thermal_read_mode;
> static bool thermal_use_labels;
>
> +/* Function to check thermal read mode */
> +static enum thermal_access_mode thermal_read_mode_check(void)
> +{
> + u8 t, ta1, ta2, ver = 0;
> + int i;
> +
> + if (thinkpad_id.ec_model) {
> + /*
> + * Direct EC access mode: sensors at registers
> + * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
Remove the double space, one is enough in kernel comments.
> + * non-implemented, thermal sensors return 0x80 when
> + * not available
Add the missing . please.
Perhaps add a empty line here to make this two paragraphs.
> + * The above rule is unfortunately flawed. This has been seen with
> + * 0xC2 (power supply ID) causing thermal control problems.
> + * The EC version can be determined by offset 0xEF and at least for
> + * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
> + * are not thermal registers.
> + */
While the patch touches this, this entire comment should be reflowed
properly for 80 columns.
> + if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
> + pr_warn("Thinkpad ACPI EC unable to access EC version\n");
> +
> + ta1 = ta2 = 0;
> + for (i = 0; i < 8; i++) {
> + if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
> + ta1 |= t;
> + } else {
> + ta1 = 0;
> + break;
> + }
> + if (ver < 3) {
> + if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
> + ta2 |= t;
> + } else {
> + ta1 = 0;
> + break;
> + }
> + }
> + }
> +
> + if (ta1 == 0) {
> + pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
> + return TPACPI_THERMAL_NONE;
> + }
> +
> + if (ver >= 3) {
> + thermal_use_labels = true;
> + return TPACPI_THERMAL_TPEC_8;
> + }
> +
> + return (ta2 != 0) ? TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
> + }
> +
> + if (acpi_evalf(ec_handle, NULL, "TMP7", "qv")) {
> + if (tpacpi_is_ibm() &&
> + acpi_evalf(ec_handle, NULL, "UPDT", "qv"))
Single line and keep the braces please (I know it will go >80 chars but no
important info is lost).
> + /* 600e/x, 770e, 770x */
> + return TPACPI_THERMAL_ACPI_UPDT;
> + /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
> + return TPACPI_THERMAL_ACPI_TMP07;
> + }
> +
> + /* temperatures not supported on 570, G4x, R30, R31, R32 */
> + return TPACPI_THERMAL_NONE;
> +}
> +
> /* idx is zero-based */
> static int thermal_get_sensor(int idx, s32 *value)
> {
> @@ -6375,78 +6440,9 @@ static const struct attribute_group temp_label_attr_group = {
>
> static int __init thermal_init(struct ibm_init_struct *iibm)
> {
> - u8 t, ta1, ta2, ver = 0;
> - int i;
> - int acpi_tmp7;
> -
> vdbg_printk(TPACPI_DBG_INIT, "initializing thermal subdriver\n");
>
> - acpi_tmp7 = acpi_evalf(ec_handle, NULL, "TMP7", "qv");
> -
> - if (thinkpad_id.ec_model) {
> - /*
> - * Direct EC access mode: sensors at registers
> - * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
> - * non-implemented, thermal sensors return 0x80 when
> - * not available
> - * The above rule is unfortunately flawed. This has been seen with
> - * 0xC2 (power supply ID) causing thermal control problems.
> - * The EC version can be determined by offset 0xEF and at least for
> - * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
> - * are not thermal registers.
> - */
> - if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
> - pr_warn("Thinkpad ACPI EC unable to access EC version\n");
> -
> - ta1 = ta2 = 0;
> - for (i = 0; i < 8; i++) {
> - if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
> - ta1 |= t;
> - } else {
> - ta1 = 0;
> - break;
> - }
> - if (ver < 3) {
> - if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
> - ta2 |= t;
> - } else {
> - ta1 = 0;
> - break;
> - }
> - }
> - }
> - if (ta1 == 0) {
> - /* This is sheer paranoia, but we handle it anyway */
> - if (acpi_tmp7) {
> - pr_err("ThinkPad ACPI EC access misbehaving, falling back to ACPI TMPx access mode\n");
> - thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
Eh, where did this go in the new helper?
--
i.
> - } else {
> - pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
> - thermal_read_mode = TPACPI_THERMAL_NONE;
> - }
> - } else {
> - if (ver >= 3) {
> - thermal_read_mode = TPACPI_THERMAL_TPEC_8;
> - thermal_use_labels = true;
> - } else {
> - thermal_read_mode =
> - (ta2 != 0) ?
> - TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
> - }
> - }
> - } else if (acpi_tmp7) {
> - if (tpacpi_is_ibm() &&
> - acpi_evalf(ec_handle, NULL, "UPDT", "qv")) {
> - /* 600e/x, 770e, 770x */
> - thermal_read_mode = TPACPI_THERMAL_ACPI_UPDT;
> - } else {
> - /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
> - thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
> - }
> - } else {
> - /* temperatures not supported on 570, G4x, R30, R31, R32 */
> - thermal_read_mode = TPACPI_THERMAL_NONE;
> - }
> + thermal_read_mode = thermal_read_mode_check();
>
> vdbg_printk(TPACPI_DBG_INIT, "thermal is %s, mode %d\n",
> str_supported(thermal_read_mode != TPACPI_THERMAL_NONE),
>
On Wed, 14 Feb 2024, Vishnu Sankar wrote:
> Added non-standard thermal register support for some ThinkPads.
>
> Some of the Thinkpads use a non-standard ECFW which use different
> thermal register addresses.
> This is a fix to correct the wrong temperature reporting on
> those systems.
>
> Tested on Lenovo ThinkPad L13 Yoga Gen2
>
> Suggested-by: Mark Pearson <[email protected]>
> Signed-off-by: Vishnu Sankar <[email protected]>
> ---
> -Improvements as requested.
> -Improved the readability in case TPACPI_THERMAL_TPEC_12.
> -idx < 8 from idx idx <=7 to match idx = 8
> -KILO used from linux/units.h instead of 1000.
> static enum thermal_access_mode thermal_read_mode;
> static bool thermal_use_labels;
> +static bool thermal_with_ns_address; /*Non-standard thermal reg address*/
Comment is missing spaces.
> @@ -6239,6 +6267,20 @@ static int thermal_get_sensor(int idx, s32 *value)
> }
> break;
>
> + /* The Non-standard EC uses 12 Thermal areas */
> + case TPACPI_THERMAL_TPEC_12:
> + if (idx >= 12)
> + return -EINVAL;
> +
> + t = idx < 8 ? TP_EC_THERMAL_TMP0_NS + idx :
> + TP_EC_THERMAL_TMP8_NS + (idx - 8);
> +
> + if (!acpi_ec_read(t, &tmp))
> + return -EIO;
> +
> + *value = tmp * KILO;
Hmm, MILLI would be much more approriate here? But if this relates to
degrees, there is MILLIDEGREE_PER_DEGREE?
Reviewed-by: Ilpo J?rvinen <[email protected]>
--
i.
On Wed, 14 Feb 2024, Vishnu Sankar wrote:
Thanks for doing this, it's major improvement to readability already as
is, and even more of after the second patch!!
> Add a thermal_read_mode_check helper function to make the code
thermal_read_mode_check()
remove "function" as it's obvious.
> simpler during init.
> This helps particularly when the new TPEC_12 mode is added in
> the next patch.
Flow the paragraph normally without the premature line break.
> Suggested-by: Ilpo Jarvinen <[email protected]>
This is not my email address, please use
Suggested-by: Ilpo J?rvinen <[email protected]>
> Signed-off-by: Vishnu Sankar <[email protected]>
> ---
> drivers/platform/x86/thinkpad_acpi.c | 136 +++++++++++++--------------
> 1 file changed, 66 insertions(+), 70 deletions(-)
>
> diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c
> index c4895e9bc714..2428c8bd0fa2 100644
> --- a/drivers/platform/x86/thinkpad_acpi.c
> +++ b/drivers/platform/x86/thinkpad_acpi.c
> @@ -6147,6 +6147,71 @@ struct ibm_thermal_sensors_struct {
> static enum thermal_access_mode thermal_read_mode;
> static bool thermal_use_labels;
>
> +/* Function to check thermal read mode */
> +static enum thermal_access_mode thermal_read_mode_check(void)
> +{
> + u8 t, ta1, ta2, ver = 0;
> + int i;
> +
> + if (thinkpad_id.ec_model) {
> + /*
> + * Direct EC access mode: sensors at registers
> + * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
While this comment is touches, lets make minor improvements to it.
Remove the double space, they are not needed in kernel comments.
> + * non-implemented, thermal sensors return 0x80 when
> + * not available
Please add the missing .
Perhaps add an empty line here to break the comment into two paragraphs?
> + * The above rule is unfortunately flawed. This has been seen with
> + * 0xC2 (power supply ID) causing thermal control problems.
> + * The EC version can be determined by offset 0xEF and at least for
> + * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
> + * are not thermal registers.
Please reflow the entire comment to 80 chars.
> + */
> + if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
> + pr_warn("Thinkpad ACPI EC unable to access EC version\n");
> +
> + ta1 = ta2 = 0;
> + for (i = 0; i < 8; i++) {
> + if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
> + ta1 |= t;
> + } else {
> + ta1 = 0;
> + break;
> + }
> + if (ver < 3) {
> + if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
> + ta2 |= t;
> + } else {
> + ta1 = 0;
> + break;
> + }
> + }
> + }
> +
> + if (ta1 == 0) {
> + pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
> + return TPACPI_THERMAL_NONE;
> + }
> +
> + if (ver >= 3) {
> + thermal_use_labels = true;
> + return TPACPI_THERMAL_TPEC_8;
> + }
> +
> + return (ta2 != 0) ? TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
> + }
> +
> + if (acpi_evalf(ec_handle, NULL, "TMP7", "qv")) {
> + if (tpacpi_is_ibm() &&
> + acpi_evalf(ec_handle, NULL, "UPDT", "qv"))
Put this to single line and retain the braces please (I know it will be
>80 char but nothing important is lost).
> + /* 600e/x, 770e, 770x */
> + return TPACPI_THERMAL_ACPI_UPDT;
> + /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
> + return TPACPI_THERMAL_ACPI_TMP07;
> + }
> +
> + /* temperatures not supported on 570, G4x, R30, R31, R32 */
> + return TPACPI_THERMAL_NONE;
> +}
> +
> /* idx is zero-based */
> static int thermal_get_sensor(int idx, s32 *value)
> {
> @@ -6375,78 +6440,9 @@ static const struct attribute_group temp_label_attr_group = {
>
> static int __init thermal_init(struct ibm_init_struct *iibm)
> {
> - u8 t, ta1, ta2, ver = 0;
> - int i;
> - int acpi_tmp7;
> -
> vdbg_printk(TPACPI_DBG_INIT, "initializing thermal subdriver\n");
>
> - acpi_tmp7 = acpi_evalf(ec_handle, NULL, "TMP7", "qv");
> -
> - if (thinkpad_id.ec_model) {
> - /*
> - * Direct EC access mode: sensors at registers
> - * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
> - * non-implemented, thermal sensors return 0x80 when
> - * not available
> - * The above rule is unfortunately flawed. This has been seen with
> - * 0xC2 (power supply ID) causing thermal control problems.
> - * The EC version can be determined by offset 0xEF and at least for
> - * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
> - * are not thermal registers.
> - */
> - if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
> - pr_warn("Thinkpad ACPI EC unable to access EC version\n");
> -
> - ta1 = ta2 = 0;
> - for (i = 0; i < 8; i++) {
> - if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
> - ta1 |= t;
> - } else {
> - ta1 = 0;
> - break;
> - }
> - if (ver < 3) {
> - if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
> - ta2 |= t;
> - } else {
> - ta1 = 0;
> - break;
> - }
> - }
> - }
> - if (ta1 == 0) {
> - /* This is sheer paranoia, but we handle it anyway */
> - if (acpi_tmp7) {
> - pr_err("ThinkPad ACPI EC access misbehaving, falling back to ACPI TMPx access mode\n");
> - thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
Eh, where did this go in the new helper?
--
i.
> - } else {
> - pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
> - thermal_read_mode = TPACPI_THERMAL_NONE;
> - }
> - } else {
> - if (ver >= 3) {
> - thermal_read_mode = TPACPI_THERMAL_TPEC_8;
> - thermal_use_labels = true;
> - } else {
> - thermal_read_mode =
> - (ta2 != 0) ?
> - TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
> - }
> - }
> - } else if (acpi_tmp7) {
> - if (tpacpi_is_ibm() &&
> - acpi_evalf(ec_handle, NULL, "UPDT", "qv")) {
> - /* 600e/x, 770e, 770x */
> - thermal_read_mode = TPACPI_THERMAL_ACPI_UPDT;
> - } else {
> - /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
> - thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
> - }
> - } else {
> - /* temperatures not supported on 570, G4x, R30, R31, R32 */
> - thermal_read_mode = TPACPI_THERMAL_NONE;
> - }
> + thermal_read_mode = thermal_read_mode_check();
>
> vdbg_printk(TPACPI_DBG_INIT, "thermal is %s, mode %d\n",
> str_supported(thermal_read_mode != TPACPI_THERMAL_NONE),
>
Ilpo,
Thanks a lot for the review.
On Wed, Feb 14, 2024 at 7:14 PM Ilpo Järvinen
<[email protected]> wrote:
>
> On Wed, 14 Feb 2024, Vishnu Sankar wrote:
>
> Thanks for doing this, it's major improvement to readability already as
> is, and even more of after the second patch!!
>
> > Add a thermal_read_mode_check helper function to make the code
>
> thermal_read_mode_check()
>
> remove "function" as it's obvious.
Acked
>
> > simpler during init.
> > This helps particularly when the new TPEC_12 mode is added in
> > the next patch.
>
> Flow the paragraph normally without the premature line break.
Acked.
>
> > Suggested-by: Ilpo Jarvinen <[email protected]>
>
> This is not my email address, please use
>
> Suggested-by: Ilpo Järvinen <[email protected]>
Sorry for this.
Will correct email address.
>
> > Signed-off-by: Vishnu Sankar <[email protected]>
> > ---
> > drivers/platform/x86/thinkpad_acpi.c | 136 +++++++++++++--------------
> > 1 file changed, 66 insertions(+), 70 deletions(-)
> >
> > diff --git a/drivers/platform/x86/thinkpad_acpi.c b/drivers/platform/x86/thinkpad_acpi.c
> > index c4895e9bc714..2428c8bd0fa2 100644
> > --- a/drivers/platform/x86/thinkpad_acpi.c
> > +++ b/drivers/platform/x86/thinkpad_acpi.c
> > @@ -6147,6 +6147,71 @@ struct ibm_thermal_sensors_struct {
> > static enum thermal_access_mode thermal_read_mode;
> > static bool thermal_use_labels;
> >
> > +/* Function to check thermal read mode */
> > +static enum thermal_access_mode thermal_read_mode_check(void)
> > +{
> > + u8 t, ta1, ta2, ver = 0;
> > + int i;
> > +
> > + if (thinkpad_id.ec_model) {
> > + /*
> > + * Direct EC access mode: sensors at registers
> > + * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
>
> Remove the double space, one is enough in kernel comments.
Acked
>
> > + * non-implemented, thermal sensors return 0x80 when
> > + * not available
>
> Add the missing . please.
>
> Perhaps add a empty line here to make this two paragraphs.
Acked
>
> > + * The above rule is unfortunately flawed. This has been seen with
> > + * 0xC2 (power supply ID) causing thermal control problems.
> > + * The EC version can be determined by offset 0xEF and at least for
> > + * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
> > + * are not thermal registers.
> > + */
>
> While the patch touches this, this entire comment should be reflowed
> properly for 80 columns.
Will try re-writing the comments block for 80 columns.
>
> > + if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
> > + pr_warn("Thinkpad ACPI EC unable to access EC version\n");
> > +
> > + ta1 = ta2 = 0;
> > + for (i = 0; i < 8; i++) {
> > + if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
> > + ta1 |= t;
> > + } else {
> > + ta1 = 0;
> > + break;
> > + }
> > + if (ver < 3) {
> > + if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
> > + ta2 |= t;
> > + } else {
> > + ta1 = 0;
> > + break;
> > + }
> > + }
> > + }
> > +
> > + if (ta1 == 0) {
> > + pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
> > + return TPACPI_THERMAL_NONE;
> > + }
> > +
> > + if (ver >= 3) {
> > + thermal_use_labels = true;
> > + return TPACPI_THERMAL_TPEC_8;
> > + }
> > +
> > + return (ta2 != 0) ? TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
> > + }
> > +
> > + if (acpi_evalf(ec_handle, NULL, "TMP7", "qv")) {
> > + if (tpacpi_is_ibm() &&
> > + acpi_evalf(ec_handle, NULL, "UPDT", "qv"))
>
> Single line and keep the braces please (I know it will go >80 chars but no
> important info is lost).
Acked.
Will change this.
>
> > + /* 600e/x, 770e, 770x */
> > + return TPACPI_THERMAL_ACPI_UPDT;
> > + /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
> > + return TPACPI_THERMAL_ACPI_TMP07;
> > + }
> > +
> > + /* temperatures not supported on 570, G4x, R30, R31, R32 */
> > + return TPACPI_THERMAL_NONE;
> > +}
> > +
> > /* idx is zero-based */
> > static int thermal_get_sensor(int idx, s32 *value)
> > {
> > @@ -6375,78 +6440,9 @@ static const struct attribute_group temp_label_attr_group = {
> >
> > static int __init thermal_init(struct ibm_init_struct *iibm)
> > {
> > - u8 t, ta1, ta2, ver = 0;
> > - int i;
> > - int acpi_tmp7;
> > -
> > vdbg_printk(TPACPI_DBG_INIT, "initializing thermal subdriver\n");
> >
> > - acpi_tmp7 = acpi_evalf(ec_handle, NULL, "TMP7", "qv");
> > -
> > - if (thinkpad_id.ec_model) {
> > - /*
> > - * Direct EC access mode: sensors at registers
> > - * 0x78-0x7F, 0xC0-0xC7. Registers return 0x00 for
> > - * non-implemented, thermal sensors return 0x80 when
> > - * not available
> > - * The above rule is unfortunately flawed. This has been seen with
> > - * 0xC2 (power supply ID) causing thermal control problems.
> > - * The EC version can be determined by offset 0xEF and at least for
> > - * version 3 the Lenovo firmware team confirmed that registers 0xC0-0xC7
> > - * are not thermal registers.
> > - */
> > - if (!acpi_ec_read(TP_EC_FUNCREV, &ver))
> > - pr_warn("Thinkpad ACPI EC unable to access EC version\n");
> > -
> > - ta1 = ta2 = 0;
> > - for (i = 0; i < 8; i++) {
> > - if (acpi_ec_read(TP_EC_THERMAL_TMP0 + i, &t)) {
> > - ta1 |= t;
> > - } else {
> > - ta1 = 0;
> > - break;
> > - }
> > - if (ver < 3) {
> > - if (acpi_ec_read(TP_EC_THERMAL_TMP8 + i, &t)) {
> > - ta2 |= t;
> > - } else {
> > - ta1 = 0;
> > - break;
> > - }
> > - }
> > - }
> > - if (ta1 == 0) {
> > - /* This is sheer paranoia, but we handle it anyway */
> > - if (acpi_tmp7) {
> > - pr_err("ThinkPad ACPI EC access misbehaving, falling back to ACPI TMPx access mode\n");
> > - thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
>
> Eh, where did this go in the new helper?
Sorry.
This will be added in V2.
>
> --
> i.
>
> > - } else {
> > - pr_err("ThinkPad ACPI EC access misbehaving, disabling thermal sensors access\n");
> > - thermal_read_mode = TPACPI_THERMAL_NONE;
> > - }
> > - } else {
> > - if (ver >= 3) {
> > - thermal_read_mode = TPACPI_THERMAL_TPEC_8;
> > - thermal_use_labels = true;
> > - } else {
> > - thermal_read_mode =
> > - (ta2 != 0) ?
> > - TPACPI_THERMAL_TPEC_16 : TPACPI_THERMAL_TPEC_8;
> > - }
> > - }
> > - } else if (acpi_tmp7) {
> > - if (tpacpi_is_ibm() &&
> > - acpi_evalf(ec_handle, NULL, "UPDT", "qv")) {
> > - /* 600e/x, 770e, 770x */
> > - thermal_read_mode = TPACPI_THERMAL_ACPI_UPDT;
> > - } else {
> > - /* IBM/LENOVO DSDT EC.TMPx access, max 8 sensors */
> > - thermal_read_mode = TPACPI_THERMAL_ACPI_TMP07;
> > - }
> > - } else {
> > - /* temperatures not supported on 570, G4x, R30, R31, R32 */
> > - thermal_read_mode = TPACPI_THERMAL_NONE;
> > - }
> > + thermal_read_mode = thermal_read_mode_check();
> >
> > vdbg_printk(TPACPI_DBG_INIT, "thermal is %s, mode %d\n",
> > str_supported(thermal_read_mode != TPACPI_THERMAL_NONE),
> >
--
Regards,
Vishnu Sankar
+817015150407 (Japan)
Ilpo,
Thanks for the review.
On Wed, Feb 14, 2024 at 7:23 PM Ilpo Järvinen
<[email protected]> wrote:
>
> On Wed, 14 Feb 2024, Vishnu Sankar wrote:
>
> > Added non-standard thermal register support for some ThinkPads.
> >
> > Some of the Thinkpads use a non-standard ECFW which uses different
> > thermal register addresses.
> > This is a fix to correct the wrong temperature reporting on
> > those systems.
> >
> > Tested on Lenovo ThinkPad L13 Yoga Gen2
> >
> > Suggested-by: Mark Pearson <[email protected]>
> > Signed-off-by: Vishnu Sankar <[email protected]>
> > ---
> > -Improvements as requested.
> > -Improved the readability in case TPACPI_THERMAL_TPEC_12.
> > -idx < 8 from idx idx <=7 to match idx = 8
> > -KILO used from linux/units.h instead of 1000.
>
> > static enum thermal_access_mode thermal_read_mode;
> > static bool thermal_use_labels;
> > +static bool thermal_with_ns_address; /*Non-standard thermal reg address*/
>
> Comment is missing spaces.
Acked.
>
> > @@ -6239,6 +6267,20 @@ static int thermal_get_sensor(int idx, s32 *value)
> > }
> > break;
> >
> > + /* The Non-standard EC uses 12 Thermal areas */
> > + case TPACPI_THERMAL_TPEC_12:
> > + if (idx >= 12)
> > + return -EINVAL;
> > +
> > + t = idx < 8 ? TP_EC_THERMAL_TMP0_NS + idx :
> > + TP_EC_THERMAL_TMP8_NS + (idx - 8);
> > +
> > + if (!acpi_ec_read(t, &tmp))
> > + return -EIO;
> > +
> > + *value = tmp * KILO;
>
> Hmm, MILLI would be much more approriate here? But if this relates to
> degrees, there is MILLIDEGREE_PER_DEGREE?
Will Change to MILLIDEGREE_PER_DEGREE from KILO
>
> Reviewed-by: Ilpo Järvinen <[email protected]>
>
> --
> i.
--
Regards,
Vishnu Sankar
+817015150407 (Japan)