Skip to contentSkip to conversation

hwmon: (coretemp) Atom CPUs don't support TjMax; no warning needed

PATCH In review Guenter Roeck 2 participants 4 messages Jul 1, 2013 – Jul 3, 2013

Subsystems (2)
lore
Share & tools
lore

Guenter Roeck

author maintainer
#0 lore raw

PATCH 1/1 hwmon: (coretemp) Atom CPUs don't support TjMax; no warning needed

Display warning "Unable to read TjMax from CPU x" only if the CPU is supposed to support it. This is not the case for the various Atom CPUs.

---
 drivers/hwmon/coretemp.c |   14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
drivers/hwmon/coretemp.c +13 -1
index ade35cf..be58da1 100644
@@ -317,6 +317,18 @@ static int __cpuinit adjust_tjmax(struct cpuinfo_x86 *c, u32 id,
 	return tjmax;
 }
 
+static bool cpu_has_tjmax(struct cpuinfo_x86 *c)
+{
+	u8 model = c->x86_model;
+
+	return model > 0xe &&
+	       model != 0x1c &&
+	       model != 0x26 &&
+	       model != 0x27 &&
+	       model != 0x35 &&
+	       model != 0x36;
+}
+
 static int __cpuinit get_tjmax(struct cpuinfo_x86 *c, u32 id,
 			       struct device *dev)
 {
@@ -330,7 +342,7 @@ static int __cpuinit get_tjmax(struct cpuinfo_x86 *c, u32 id,
 	 */
 	err = rdmsr_safe_on_cpu(id, MSR_IA32_TEMPERATURE_TARGET, &eax, &edx);
 	if (err) {
-		if (c->x86_model > 0xe && c->x86_model != 0x1c)
+		if (cpu_has_tjmax(c))
 			dev_warn(dev, "Unable to read TjMax from CPU %u\n", id);
 	} else {
 		val = (eax >> 16) & 0xff;
Signature · 7 lines
1.7.9.7


_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors
#1 lore raw

Hi Guenter,

Quoted 27 lines

-----Original Message-----
From: [email protected] [mailto:lm-sensors-bounces@lm- sensors.org] On Behalf Of Guenter Roeck
Sent: Monday, July 01, 2013 9:46 PM
To: lm-sensors
Cc: Yu, Fenghua
Subject: [lm-sensors] [PATCH] hwmon: (coretemp) Atom CPUs don't support TjMax; no warning needed

Display warning "Unable to read TjMax from CPU x" only if the CPU is supposed to support it. This is not the case for the various Atom CPUs.

---
 drivers/hwmon/coretemp.c |   14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
drivers/hwmon/coretemp.c +1 -0
index ade35cf..be58da1 100644
@@ -317,6 +317,18 @@ static int __cpuinit adjust_tjmax(struct cpuinfo_x86 *c,
 u32 id,
 	return tjmax;
 }
 
+static bool cpu_has_tjmax(struct cpuinfo_x86 *c)

drivers/hwmon/coretemp.c static int __cpuinit adjust_tjmax(struct cpuinfo_x86 *c, · line 331

Quoted 10 lines
drivers/hwmon/coretemp.c +10 -0
+{
+	u8 model = c->x86_model;
+
+	return model > 0xe &&
+	       model != 0x1c &&
+	       model != 0x26 &&
+	       model != 0x27 &&
+	       model != 0x35 &&
+	       model != 0x36;
+}

Thanks for the patch.
Please add model ids: 0x4a and 0x37.

Thanks,
Durga

Quoted 20 lines
drivers/hwmon/coretemp.c +1 -0
+
 static int __cpuinit get_tjmax(struct cpuinfo_x86 *c, u32 id,
 			       struct device *dev)

{

drivers/hwmon/coretemp.c +1 -1
@@ -330,7 +342,7 @@ static int __cpuinit get_tjmax(struct cpuinfo_x86 *c, u32
 id,
 	 */
 	err = rdmsr_safe_on_cpu(id, MSR_IA32_TEMPERATURE_TARGET, &eax,
 &edx);
 	if (err) {
-		if (c->x86_model > 0xe && c->x86_model != 0x1c)
+		if (cpu_has_tjmax(c))
 			dev_warn(dev, "Unable to read TjMax from CPU %u\n",

id);

drivers/hwmon/coretemp.c +0 -1
 	} else {
 		val = (eax >> 16) & 0xff;
--

1.7.9.7

Quoted 4 lines

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

author maintainer
#2 lore raw
In reply to Guenter Roeck
Quoted 28 lines · Durgadoss

Hi Guenter,

Quoted 26 lines

-----Original Message-----
From: [email protected] [mailto:lm-sensors-bounces@lm- sensors.org] On Behalf Of Guenter Roeck
Sent: Monday, July 01, 2013 9:46 PM
To: lm-sensors
Cc: Yu, Fenghua
Subject: [lm-sensors] [PATCH] hwmon: (coretemp) Atom CPUs don't support TjMax; no warning needed

Display warning "Unable to read TjMax from CPU x" only if the CPU is supposed to support it. This is not the case for the various Atom CPUs.

---
 drivers/hwmon/coretemp.c |   14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
drivers/hwmon/coretemp.c +0 -0
index ade35cf..be58da1 100644
@@ -317,6 +317,18 @@ static int __cpuinit adjust_tjmax(struct cpuinfo_x86 *c,
 u32 id,
 	return tjmax;
 }
 
Quoted 15 lines

drivers/hwmon/coretemp.c static int __cpuinit adjust_tjmax(struct cpuinfo_x86 *c, u32 id, · line 330

Quoted 11 lines
drivers/hwmon/coretemp.c +11 -0
+static bool cpu_has_tjmax(struct cpuinfo_x86 *c)
+{
+	u8 model = c->x86_model;
+
+	return model > 0xe &&
+	       model != 0x1c &&
+	       model != 0x26 &&
+	       model != 0x27 &&
+	       model != 0x35 &&
+	       model != 0x36;
+}

Thanks for the patch.
Please add model ids: 0x4a and 0x37.

Hi Durga,

are those for the new Atom CPUs (S1220/S1240/S1260/S1269/S1279/S1289) ?

I am still trying to figure out how to identify those CPUs to add TjMax information. All I could find in the manual was to use the PCI root bridge ID for S1220/S1240/S1260; the datasheets for the others are not published yet.

This means we might have to go back to using the PCI ID to identify some of the chips. On the upside, pci_get_bus_and_slot() can now be called even if PCI is disabled, the method is more efficient, and we could use the same mechanism to detect the CE41x0 CPUs, so that might not be too bad.

If you have any additional information which you can share (means to detect the CPUs, and Tjmax for the S1269/S1279/S1289), please let me know.

Thanks,
Guenter

Quoted 23 lines

Thanks,
Durga

Quoted 20 lines
drivers/hwmon/coretemp.c +1 -0
+
 static int __cpuinit get_tjmax(struct cpuinfo_x86 *c, u32 id,
 			       struct device *dev)

{

drivers/hwmon/coretemp.c +1 -1
@@ -330,7 +342,7 @@ static int __cpuinit get_tjmax(struct cpuinfo_x86 *c, u32
 id,
 	 */
 	err = rdmsr_safe_on_cpu(id, MSR_IA32_TEMPERATURE_TARGET, &eax,
 &edx);
 	if (err) {
-		if (c->x86_model > 0xe && c->x86_model != 0x1c)
+		if (cpu_has_tjmax(c))
 			dev_warn(dev, "Unable to read TjMax from CPU %u\n",

id);

drivers/hwmon/coretemp.c +0 -1
 	} else {
 		val = (eax >> 16) & 0xff;
--

1.7.9.7

Quoted 5 lines
Quoted 4 lines

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors

author maintainer
#3 lore raw
In reply to Guenter Roeck
Quoted 52 lines · Guenter Roeck
Quoted 28 lines · Durgadoss

Hi Guenter,

Quoted 26 lines

-----Original Message-----
From: [email protected] [mailto:lm-sensors-bounces@lm- sensors.org] On Behalf Of Guenter Roeck
Sent: Monday, July 01, 2013 9:46 PM
To: lm-sensors
Cc: Yu, Fenghua
Subject: [lm-sensors] [PATCH] hwmon: (coretemp) Atom CPUs don't support TjMax; no warning needed

Display warning "Unable to read TjMax from CPU x" only if the CPU is supposed to support it. This is not the case for the various Atom CPUs.

---
 drivers/hwmon/coretemp.c |   14 +++++++++++++-
 1 file changed, 13 insertions(+), 1 deletion(-)
drivers/hwmon/coretemp.c +0 -0
index ade35cf..be58da1 100644
@@ -317,6 +317,18 @@ static int __cpuinit adjust_tjmax(struct cpuinfo_x86 *c,
 u32 id,
 	return tjmax;
 }
 
Quoted 15 lines

drivers/hwmon/coretemp.c static int __cpuinit adjust_tjmax(struct cpuinfo_x86 *c, u32 id, · line 330

Quoted 11 lines
drivers/hwmon/coretemp.c +11 -0
+static bool cpu_has_tjmax(struct cpuinfo_x86 *c)
+{
+	u8 model = c->x86_model;
+
+	return model > 0xe &&
+	       model != 0x1c &&
+	       model != 0x26 &&
+	       model != 0x27 &&
+	       model != 0x35 &&
+	       model != 0x36;
+}

Thanks for the patch.
Please add model ids: 0x4a and 0x37.

Hi Durga,

are those for the new Atom CPUs (S1220/S1240/S1260/S1269/S1279/S1289) ?

I am still trying to figure out how to identify those CPUs to add TjMax information. All I could find in the manual was to use the PCI root bridge ID for S1220/S1240/S1260; the datasheets for the others are not published yet.

Quoted 7 lines

This means we might have to go back to using the PCI ID to identify some of the chips. On the upside, pci_get_bus_and_slot() can now be called even if PCI is disabled, the method is more efficient, and we could use the same mechanism to detect the CE41x0 CPUs, so that might not be too bad.

If you have any additional information which you can share (means to detect the CPUs, and Tjmax for the S1269/S1279/S1289), please let me know.

To add to all this, I got my hands on one of the new Rangely Atoms. CPU model
number is 77 (0x4d), the chip reports as "CPU 4000" (no Atom in there),
and coretemp doesn't work at all. Or, rather, it reports nonsensical temperatures and temperature limits. Guess we'll have some work ahead of us.

Guenter

_______________________________________________
lm-sensors mailing list
[email protected]
http://lists.lm-sensors.org/mailman/listinfo/lm-sensors