Skip to contentSkip to conversation

x86 intel power: Initialize MSR_IA32_ENERGY_PERF_BIAS

PATCH In review Len Brown 9 participants 25 messages Mar 30, 2011 – Aug 18, 2011 x86

lore
Share & tools
lore

Len Brown

author
#0 lore raw

PATCH 1/1 x86 intel power: Initialize MSR_IA32_ENERGY_PERF_BIAS

From: Len Brown <[email protected]>

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility is available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   14 ++++++++++++++
 2 files changed, 17 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..91fedd9 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +14 -0
index d16c2c5..48cca4a 100644
@@ -448,6 +448,20 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
Signature · 1 line
1.7.4.2.406.gbe91
#1 lore raw
Quoted 51 lines · Len Brown

From: Len Brown <[email protected]>

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility is available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   14 ++++++++++++++
 2 files changed, 17 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..91fedd9 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +4 -0
index d16c2c5..48cca4a 100644
@@ -448,6 +448,20 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time

arch/x86/kernel/cpu/intel.c static void __cpuinit init_intel(struct cpuinfo_x86 *c) · line 464

Quoted 10 lines
arch/x86/kernel/cpu/intel.c +10 -0
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}

Dunno, this patch appears to silently modify the system to be slower than it was before under Linux.

Won't people report this as a regression if this change reduces performance for them?

They wont be able to see your comments in the code and in the changelog either, when this happens to them. They might look into /proc/cpuinfo and see 'epb' there but it wont tell them anything. They wont know about a utility available in tools/power/x86/ either.

So this patch has 'future trouble' written all over it i'm afraid.

Thanks,

Ingo

#2 lore raw
In reply to Len Brown
Quoted 51 lines · Len Brown

From: Len Brown <[email protected]>

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility is available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   14 ++++++++++++++
 2 files changed, 17 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..91fedd9 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +4 -0
index d16c2c5..48cca4a 100644
@@ -448,6 +448,20 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time

arch/x86/kernel/cpu/intel.c static void __cpuinit init_intel(struct cpuinfo_x86 *c) · line 464

Quoted 10 lines
arch/x86/kernel/cpu/intel.c +10 -0
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}

Dunno, this patch appears to silently modify the system to be slower than it was before under Linux.

Won't people report this as a regression if this change reduces performance for them?

They wont be able to see your comments in the code and in the changelog either, when this happens to them. They might look into /proc/cpuinfo and see 'epb' there but it wont tell them anything. They wont know about a utility available in tools/power/x86/ either.

So this patch has 'future trouble' written all over it i'm afraid.

Thanks,

Ingo

author
#3 lore raw
Quoted 26 lines
Quoted 26 lines

From: Len Brown <[email protected]>

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility is available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   14 ++++++++++++++
 2 files changed, 17 insertions(+), 0 deletions(-)

...

Quoted 11 lines

Dunno, this patch appears to silently modify the system to be slower than it was before under Linux.

Won't people report this as a regression if this change reduces performance for them?

They wont be able to see your comments in the code and in the changelog either, when this happens to them. They might look into /proc/cpuinfo and see 'epb' there but it wont tell them anything. They wont know about a utility available in tools/power/x86/ either.

This patch makes no change to the epb feature indicator /proc/cpuinfo.

Quoted 1 line

So this patch has 'future trouble' written all over it i'm afraid.

EPB is limited to SNB and later.
So the installed base as yet is small.
(it also exists on WSM-EP, but doesn't do so much there) EPB will have a more significant effect on future hardware.

Linux currently trails competing operating systems in energy efficiency on SNB due to this setting, and Linux will trail competing operating systems even more on future hardware if this default is not fixed.

Will it be possible to measure a performance difference between "performance" and "normal"? Yes, it will be possible. Will 99.9% of users notice? Nope. More likely they'll notice the the power savings that are disabled in "performance" mode.

I should have called it "benchmark" mode instead of "performance" mode...

thanks,
Len Brown, Intel Open Source Technology Center

author
#4 lore raw
In reply to Ingo Molnar
Quoted 26 lines
Quoted 26 lines

From: Len Brown <[email protected]>

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility is available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   14 ++++++++++++++
 2 files changed, 17 insertions(+), 0 deletions(-)

...

Quoted 11 lines

Dunno, this patch appears to silently modify the system to be slower than it was before under Linux.

Won't people report this as a regression if this change reduces performance for them?

They wont be able to see your comments in the code and in the changelog either, when this happens to them. They might look into /proc/cpuinfo and see 'epb' there but it wont tell them anything. They wont know about a utility available in tools/power/x86/ either.

This patch makes no change to the epb feature indicator /proc/cpuinfo.

Quoted 1 line

So this patch has 'future trouble' written all over it i'm afraid.

EPB is limited to SNB and later.
So the installed base as yet is small.
(it also exists on WSM-EP, but doesn't do so much there) EPB will have a more significant effect on future hardware.

Linux currently trails competing operating systems in energy efficiency on SNB due to this setting, and Linux will trail competing operating systems even more on future hardware if this default is not fixed.

Will it be possible to measure a performance difference between "performance" and "normal"? Yes, it will be possible. Will 99.9% of users notice? Nope. More likely they'll notice the the power savings that are disabled in "performance" mode.

I should have called it "benchmark" mode instead of "performance" mode...

thanks,
Len Brown, Intel Open Source Technology Center

#5 lore raw
Quoted 34 lines · Len Brown
Quoted 26 lines
Quoted 26 lines

From: Len Brown <[email protected]>

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility is available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   14 ++++++++++++++
 2 files changed, 17 insertions(+), 0 deletions(-)

...

Quoted 7 lines

Dunno, this patch appears to silently modify the system to be slower than it was before under Linux.

Won't people report this as a regression if this change reduces performance for them?

Quoted 7 lines
Quoted 4 lines

They wont be able to see your comments in the code and in the changelog either, when this happens to them. They might look into /proc/cpuinfo and see 'epb' there but it wont tell them anything. They wont know about a utility available in tools/power/x86/ either.

This patch makes no change to the epb feature indicator /proc/cpuinfo.

I know. I reacted to this bit in the changelog:

Quoted 1 line
Quoted 1 line
Quoted 1 line

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo,

Printing the existence of a CPU feature does nothing to inform users.

Quoted 12 lines
Quoted 1 line

So this patch has 'future trouble' written all over it i'm afraid.

EPB is limited to SNB and later.
So the installed base as yet is small.
(it also exists on WSM-EP, but doesn't do so much there) EPB will have a more significant effect on future hardware.

Linux currently trails competing operating systems in energy efficiency on SNB due to this setting, and Linux will trail competing operating systems even more on future hardware if this default is not fixed.

Quoted 6 lines

Will it be possible to measure a performance difference between "performance" and "normal"? Yes, it will be possible. Will 99.9% of users notice? Nope. More likely they'll notice the the power savings that are disabled in "performance" mode.

I should have called it "benchmark" mode instead of "performance" mode...

That's all fair but does not address the concerns i raised. A silent change during bootup is asking for trouble.

So how about informing users, how about making it non-silent? An informative printk that also mentions the power configuration tool, etc. This solves the concerns i mentioned.

Thanks,

Ingo

#6 lore raw
In reply to Len Brown
Quoted 34 lines · Len Brown
Quoted 26 lines
Quoted 26 lines

From: Len Brown <[email protected]>

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility is available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   14 ++++++++++++++
 2 files changed, 17 insertions(+), 0 deletions(-)

...

Quoted 7 lines

Dunno, this patch appears to silently modify the system to be slower than it was before under Linux.

Won't people report this as a regression if this change reduces performance for them?

Quoted 7 lines
Quoted 4 lines

They wont be able to see your comments in the code and in the changelog either, when this happens to them. They might look into /proc/cpuinfo and see 'epb' there but it wont tell them anything. They wont know about a utility available in tools/power/x86/ either.

This patch makes no change to the epb feature indicator /proc/cpuinfo.

I know. I reacted to this bit in the changelog:

Quoted 1 line
Quoted 1 line
Quoted 1 line

Since 2.6.35 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo,

Printing the existence of a CPU feature does nothing to inform users.

Quoted 12 lines
Quoted 1 line

So this patch has 'future trouble' written all over it i'm afraid.

EPB is limited to SNB and later.
So the installed base as yet is small.
(it also exists on WSM-EP, but doesn't do so much there) EPB will have a more significant effect on future hardware.

Linux currently trails competing operating systems in energy efficiency on SNB due to this setting, and Linux will trail competing operating systems even more on future hardware if this default is not fixed.

Quoted 6 lines

Will it be possible to measure a performance difference between "performance" and "normal"? Yes, it will be possible. Will 99.9% of users notice? Nope. More likely they'll notice the the power savings that are disabled in "performance" mode.

I should have called it "benchmark" mode instead of "performance" mode...

That's all fair but does not address the concerns i raised. A silent change during bootup is asking for trouble.

So how about informing users, how about making it non-silent? An informative printk that also mentions the power configuration tool, etc. This solves the concerns i mentioned.

Thanks,

Ingo

author
#7 lore raw
Quoted 4 lines

So how about informing users, how about making it non-silent? An informative printk that also mentions the power configuration tool, etc. This solves the concerns i mentioned.

Something like this?

                rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
                if ((epb & 0xF) == 0) {
                        printk_once(KERN_WARN, "x86: updated energy_perf_bias"
                                " to 'normal' from 'performance'\n"
                                "You can view and update epb via utility,"
                                " such as x86_energy_perf_policy(8)\n");
                        epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
                        wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
                }
#8 lore raw

Ack. Let's just do this. Ingo?

Linus

Quoted 18 lines · Len Brown
Quoted 4 lines

So how about informing users, how about making it non-silent? An informative printk that also mentions the power configuration tool, etc. This solves the concerns i mentioned.

Something like this?

               rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
               if ((epb & 0xF) == 0) {
                       printk_once(KERN_WARN, "x86: updated energy_perf_bias" " to 'normal' from 'performance'\n" "You can view and update epb via utility," " such as x86_energy_perf_policy(8)\n"); epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL; wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
               }

reviewer
#9 lore raw
Quoted 3 lines · Linus Torvalds

Ack. Let's just do this. Ingo?

Linus

Ingo is travelling this week, but this seems to have converged.

-hpa

author
#11 lore raw

PATCH v2 1/1 x86 intel power: Initialize MSR_IA32_ENERGY_PERF_BIAS

From: Len Brown <[email protected]>

Since 2.6.36 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility has been available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR, presumably because this is handled by high-volume shrink-wrap operating systems...

Linux distros, on the other hand, do not yet invoke x86_energy_perf_policy(8). As a result, WSM-EP, SNB, and later hardware from Intel will run in its default hardware power-on state (performance), which assumes that users care for performance at all costs and not for energy efficiency. While that is fine for performance benchmarks, the hardware's intended default operating point is "normal" mode...

Initialize the MSR to the "normal" by default during kernel boot.

x86_energy_perf_policy(8) is available to change the default after boot, should the user have a different preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   18 ++++++++++++++++++
 2 files changed, 21 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..91fedd9 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +18 -0
index d16c2c5..7c1ca07 100644
@@ -448,6 +448,24 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			printk_once(KERN_WARNING, "x86: updated energy_perf_bias"
+				" to 'normal' from 'performance'\n"
+				"You can view and update epb via utility,"
+				" such as x86_energy_perf_policy(8)\n");
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
Signature · 1 line
1.7.6.134.gcf13f
author
#12 lore raw

PATCH v2 1/1 x86 intel power: Initialize MSR_IA32_ENERGY_PERF_BIAS

In reply to H. Peter Anvin

From: Len Brown <[email protected]>

Since 2.6.36 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility has been available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR, presumably because this is handled by high-volume shrink-wrap operating systems...

Linux distros, on the other hand, do not yet invoke x86_energy_perf_policy(8). As a result, WSM-EP, SNB, and later hardware from Intel will run in its default hardware power-on state (performance), which assumes that users care for performance at all costs and not for energy efficiency. While that is fine for performance benchmarks, the hardware's intended default operating point is "normal" mode...

Initialize the MSR to the "normal" by default during kernel boot.

x86_energy_perf_policy(8) is available to change the default after boot, should the user have a different preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   18 ++++++++++++++++++
 2 files changed, 21 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..91fedd9 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +18 -0
index d16c2c5..7c1ca07 100644
@@ -448,6 +448,24 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			printk_once(KERN_WARNING, "x86: updated energy_perf_bias"
+				" to 'normal' from 'performance'\n"
+				"You can view and update epb via utility,"
+				" such as x86_energy_perf_policy(8)\n");
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
Signature · 1 line
1.7.6.134.gcf13f
Acked-by on v2 1/1
#13 lore raw
Quoted 16 lines · Len Brown

From: Len Brown <[email protected]>

Since 2.6.36 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility has been available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR, presumably because this is handled by high-volume shrink-wrap operating systems...

Linux distros, on the other hand, do not yet invoke x86_energy_perf_policy(8). As a result, WSM-EP, SNB, and later hardware from Intel will run in its default hardware power-on state (performance), which assumes that users care for performance at all costs and not for energy efficiency. While that is fine for performance benchmarks, the hardware's intended default operating point is "normal" mode...

Quoted 7 lines

Initialize the MSR to the "normal" by default during kernel boot.

x86_energy_perf_policy(8) is available to change the default after boot, should the user have a different preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
Acked-by Rafael J. Wysocki <[email protected]>
Quoted 49 lines
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   18 ++++++++++++++++++
 2 files changed, 21 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..91fedd9 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +18 -0
index d16c2c5..7c1ca07 100644
@@ -448,6 +448,24 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			printk_once(KERN_WARNING, "x86: updated energy_perf_bias"
+				" to 'normal' from 'performance'\n"
+				"You can view and update epb via utility,"
+				" such as x86_energy_perf_policy(8)\n");
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
Acked-by on v2 1/1
#14 lore raw
In reply to Len Brown
Quoted 16 lines · Len Brown

From: Len Brown <[email protected]>

Since 2.6.36 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility has been available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR, presumably because this is handled by high-volume shrink-wrap operating systems...

Linux distros, on the other hand, do not yet invoke x86_energy_perf_policy(8). As a result, WSM-EP, SNB, and later hardware from Intel will run in its default hardware power-on state (performance), which assumes that users care for performance at all costs and not for energy efficiency. While that is fine for performance benchmarks, the hardware's intended default operating point is "normal" mode...

Quoted 7 lines

Initialize the MSR to the "normal" by default during kernel boot.

x86_energy_perf_policy(8) is available to change the default after boot, should the user have a different preference.

cc [email protected]
Signed-off-by Len Brown <[email protected]>
Acked-by Rafael J. Wysocki <[email protected]>
Quoted 49 lines
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   18 ++++++++++++++++++
 2 files changed, 21 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..91fedd9 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +18 -0
index d16c2c5..7c1ca07 100644
@@ -448,6 +448,24 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			printk_once(KERN_WARNING, "x86: updated energy_perf_bias"
+				" to 'normal' from 'performance'\n"
+				"You can view and update epb via utility,"
+				" such as x86_energy_perf_policy(8)\n");
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
#15 lore raw

[tip:x86/cpu] x86, intel, power: Initialize MSR_IA32_ENERGY_PERF_BIAS

In reply to Len Brown
Commit-ID:  abe48b108247e9b90b4c6739662a2e5c765ed114
Gitweb:     http://git.kernel.org/tip/abe48b108247e9b90b4c6739662a2e5c765ed114
Author:     Len Brown <[email protected]>
AuthorDate: Thu, 14 Jul 2011 00:53:24 -0400
Committer:  H. Peter Anvin <[email protected]>
CommitDate: Thu, 14 Jul 2011 12:13:42 -0700

x86, intel, power: Initialize MSR_IA32_ENERGY_PERF_BIAS

Since 2.6.36 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility has been available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR, presumably because this is handled by high-volume shrink-wrap operating systems...

Linux distros, on the other hand, do not yet invoke x86_energy_perf_policy(8). As a result, WSM-EP, SNB, and later hardware from Intel will run in its default hardware power-on state (performance), which assumes that users care for performance at all costs and not for energy efficiency. While that is fine for performance benchmarks, the hardware's intended default operating point is "normal" mode...

Initialize the MSR to the "normal" by default during kernel boot.

x86_energy_perf_policy(8) is available to change the default after boot, should the user have a different preference.

Signed-off-by Len Brown <[email protected]>
Link http://lkml.kernel.org/r/alpine.LFD.2.02.1107140051020.18606@x980 open here
Acked-by Rafael J. Wysocki <[email protected]>
Signed-off-by H. Peter Anvin <[email protected]>
Cc <[email protected]>
---
 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   18 ++++++++++++++++++
 2 files changed, 21 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 485b4f1..23a9d89 100644
@@ -259,6 +259,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSWAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +18 -0
index 1edf5ba..da0d779 100644
@@ -456,6 +456,24 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == 0) {
+			printk_once(KERN_WARNING, "x86: updated energy_perf_bias"
+				" to 'normal' from 'performance'\n"
+				"You can view and update epb via utility,"
+				" such as x86_energy_perf_policy(8)\n");
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
#16 lore raw

[tip:x86/cpu] x86, msr: Fix typo in ENERGY_PERF_BIAS_POWERSAVE

Commit-ID:  4bb82178f5cb074783aaeaa06f9f840c67af7707
Gitweb:     http://git.kernel.org/tip/4bb82178f5cb074783aaeaa06f9f840c67af7707
Author:     H. Peter Anvin <[email protected]>
AuthorDate: Thu, 14 Jul 2011 14:58:44 -0700
Committer:  H. Peter Anvin <[email protected]>
CommitDate: Thu, 14 Jul 2011 14:58:44 -0700

x86, msr: Fix typo in ENERGY_PERF_BIAS_POWERSAVE

Fix a trivial typo in the name of the constant ENERGY_PERF_BIAS_POWERSAVE. This didn't cause trouble because this constant is not currently used for anything.

Signed-off-by H. Peter Anvin <[email protected]>
Cc Len Brown <[email protected]>
Link http://lkml.kernel.org/r/[email protected] open here
---
 arch/x86/include/asm/msr-index.h |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)
arch/x86/include/asm/msr-index.h +1 -1
index 23a9d89..d96bdb2 100644
@@ -261,7 +261,7 @@
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
 #define ENERGY_PERF_BIAS_PERFORMANCE	0
 #define ENERGY_PERF_BIAS_NORMAL		6
-#define ENERGY_PERF_BIAS_POWERSWAVE	15
+#define ENERGY_PERF_BIAS_POWERSAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
author
#17 lore raw

PATCH v3 1/1 x86 intel power: Initialize MSR_IA32_ENERGY_PERF_BIAS

In reply to Len Brown

From: Len Brown <[email protected]>

Since 2.6.36 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility has been available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR, presumably because this is handled by high-volume shrink-wrap operating systems...

Linux distros, on the other hand, do not yet invoke x86_energy_perf_policy(8). As a result, WSM-EP, SNB, and later hardware from Intel will run in its default hardware power-on state (performance), which assumes that users care for performance at all costs and not for energy efficiency. While that is fine for performance benchmarks, the hardware's intended default operating point is "normal" mode...

Initialize the MSR to the "normal" by default during kernel boot.

x86_energy_perf_policy(8) is available to change the default after boot, should the user have a different preference.

cc [email protected]
Acked-by Rafael J. Wysocki <[email protected]>
Signed-off-by Len Brown <[email protected]>

---
v3: fix #define typo in header and ',' typo in printk shorten printk to fix in 80 columns

 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   18 ++++++++++++++++++
 2 files changed, 21 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..55a11e0 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +18 -0
index d16c2c5..24cba78 100644
@@ -448,6 +448,24 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == ENERGY_PERF_BIAS_PERFORMANCE) {
+			printk_once(KERN_WARNING "ENERGY_PERF_BIAS:"
+				" Set to 'normal', was 'performance'\n"
+				"ENERGY_PERF_BIAS: View and update with"
+				" x86_energy_perf_policy(8)\n");
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
Signature · 1 line
1.7.6.134.gcf13f
author
#18 lore raw

PATCH v3 1/1 x86 intel power: Initialize MSR_IA32_ENERGY_PERF_BIAS

In reply to Len Brown

From: Len Brown <[email protected]>

Since 2.6.36 (23016bf0d25), Linux prints the existence of "epb" in /proc/cpuinfo, Since 2.6.38 (d5532ee7b40), the x86_energy_perf_policy(8) utility has been available in-tree to update MSR_IA32_ENERGY_PERF_BIAS.

However, the typical BIOS fails to initialize the MSR, presumably because this is handled by high-volume shrink-wrap operating systems...

Linux distros, on the other hand, do not yet invoke x86_energy_perf_policy(8). As a result, WSM-EP, SNB, and later hardware from Intel will run in its default hardware power-on state (performance), which assumes that users care for performance at all costs and not for energy efficiency. While that is fine for performance benchmarks, the hardware's intended default operating point is "normal" mode...

Initialize the MSR to the "normal" by default during kernel boot.

x86_energy_perf_policy(8) is available to change the default after boot, should the user have a different preference.

cc [email protected]
Acked-by Rafael J. Wysocki <[email protected]>
Signed-off-by Len Brown <[email protected]>

---
v3: fix #define typo in header and ',' typo in printk shorten printk to fix in 80 columns

 arch/x86/include/asm/msr-index.h |    3 +++
 arch/x86/kernel/cpu/intel.c      |   18 ++++++++++++++++++
 2 files changed, 21 insertions(+), 0 deletions(-)
arch/x86/include/asm/msr-index.h +3 -0
index 43a18c7..55a11e0 100644
@@ -250,6 +250,9 @@
 #define MSR_IA32_TEMPERATURE_TARGET	0x000001a2
 
 #define MSR_IA32_ENERGY_PERF_BIAS	0x000001b0
+#define ENERGY_PERF_BIAS_PERFORMANCE	0
+#define ENERGY_PERF_BIAS_NORMAL		6
+#define ENERGY_PERF_BIAS_POWERSAVE	15
 
 #define MSR_IA32_PACKAGE_THERM_STATUS		0x000001b1
arch/x86/kernel/cpu/intel.c +18 -0
index d16c2c5..24cba78 100644
@@ -448,6 +448,24 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 
 	if (cpu_has(c, X86_FEATURE_VMX))
 		detect_vmx_virtcap(c);
+
+	/*
+	 * Initialize MSR_IA32_ENERGY_PERF_BIAS if BIOS did not.
+	 * x86_energy_perf_policy(8) is available to change it at run-time
+	 */
+	if (cpu_has(c, X86_FEATURE_EPB)) {
+		u64 epb;
+
+		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		if ((epb & 0xF) == ENERGY_PERF_BIAS_PERFORMANCE) {
+			printk_once(KERN_WARNING "ENERGY_PERF_BIAS:"
+				" Set to 'normal', was 'performance'\n"
+				"ENERGY_PERF_BIAS: View and update with"
+				" x86_energy_perf_policy(8)\n");
+			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
+			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32
Signature · 1 line
1.7.6.134.gcf13f
#19 lore raw

[tip:x86/cpu] x86, intel, power: Correct the MSR_IA32_ENERGY_PERF_BIAS message

Commit-ID:  17edf2d79f1ea6dfdb4c444801d928953b9f98d6
Gitweb:     http://git.kernel.org/tip/17edf2d79f1ea6dfdb4c444801d928953b9f98d6
Author:     Len Brown <[email protected]>
AuthorDate: Fri, 15 Jul 2011 17:37:15 -0400
Committer:  H. Peter Anvin <[email protected]>
CommitDate: Fri, 15 Jul 2011 15:13:55 -0700

x86, intel, power: Correct the MSR_IA32_ENERGY_PERF_BIAS message

Fix the printk_once() so that it actually prints (didn't print before due to a stray comma.)

[ hpa: changed to an incremental patch and adjusted the description
accordingly. ]

Signed-off-by Len Brown <[email protected]>
Link http://lkml.kernel.org/r/alpine.LFD.2.02.1107151732480.18606@x980 open here
Cc <[email protected]>
Signed-off-by H. Peter Anvin <[email protected]>
---
 arch/x86/kernel/cpu/intel.c |   10 +++++-----
 1 files changed, 5 insertions(+), 5 deletions(-)
arch/x86/kernel/cpu/intel.c +5 -5
index da0d779..ed6086e 100644
@@ -465,11 +465,11 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 		u64 epb;
 
 		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
-		if ((epb & 0xF) == 0) {
-			printk_once(KERN_WARNING, "x86: updated energy_perf_bias"
-				" to 'normal' from 'performance'\n"
-				"You can view and update epb via utility,"
-				" such as x86_energy_perf_policy(8)\n");
+		if ((epb & 0xF) == ENERGY_PERF_BIAS_PERFORMANCE) {
+			printk_once(KERN_WARNING "ENERGY_PERF_BIAS:"
+				" Set to 'normal', was 'performance'\n"
+				"ENERGY_PERF_BIAS: View and update with"
+				" x86_energy_perf_policy(8)\n");
 			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
 			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
 		}
Reviewed-by
#20 lore raw

Re: [tip:x86/cpu] x86, intel, power: Correct the MSR_IA32_ENERGY_PERF_BIAS message

Quoted 12 lines · tip-bot for Len Brown
Commit-ID:  17edf2d79f1ea6dfdb4c444801d928953b9f98d6
Gitweb:     http://git.kernel.org/tip/17edf2d79f1ea6dfdb4c444801d928953b9f98d6
Author:     Len Brown <[email protected]>
AuthorDate: Fri, 15 Jul 2011 17:37:15 -0400
Committer:  H. Peter Anvin <[email protected]>
CommitDate: Fri, 15 Jul 2011 15:13:55 -0700

x86, intel, power: Correct the MSR_IA32_ENERGY_PERF_BIAS message

Fix the printk_once() so that it actually prints (didn't print before due to a stray comma.)

Quoted 6 lines

[ hpa: changed to an incremental patch and adjusted the description
accordingly. ]

Signed-off-by Len Brown <[email protected]>
Link http://lkml.kernel.org/r/alpine.LFD.2.02.1107151732480.18606@x980 open here
Cc <[email protected]>

This "table" should be stable :)

Cc <[email protected]>
Reviewed-by Christoph Fritz <[email protected]>

Thanks,
Christoph

Quoted 31 lines
Signed-off-by H. Peter Anvin <[email protected]>
---
 arch/x86/kernel/cpu/intel.c |   10 +++++-----
 1 files changed, 5 insertions(+), 5 deletions(-)
arch/x86/kernel/cpu/intel.c +5 -6
index da0d779..ed6086e 100644
@@ -465,11 +465,11 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 		u64 epb;
 
 		rdmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
-		if ((epb & 0xF) == 0) {
-			printk_once(KERN_WARNING, "x86: updated energy_perf_bias"
-				" to 'normal' from 'performance'\n"
-				"You can view and update epb via utility,"
-				" such as x86_energy_perf_policy(8)\n");
+		if ((epb & 0xF) == ENERGY_PERF_BIAS_PERFORMANCE) {
+			printk_once(KERN_WARNING "ENERGY_PERF_BIAS:"
+				" Set to 'normal', was 'performance'\n"
+				"ENERGY_PERF_BIAS: View and update with"
+				" x86_energy_perf_policy(8)\n");
 			epb = (epb & ~0xF) | ENERGY_PERF_BIAS_NORMAL;
 			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
 		}
--

To unsubscribe from this list: send the line "unsubscribe linux-kernel" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/

#21 lore raw
In reply to Len Brown
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]

...

Quoted 2 lines

Will it be possible to measure a performance difference between "performance" and "normal"? Yes, it will be possible.

Ok. So... what "serious bug" does this fix? You really need to use cc:
stable less. Tweaking performance/power ratio is _not_ serious bug.

Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
#22 lore raw
In reply to Len Brown
Quoted 14 lines
Quoted 14 lines
Quoted 14 lines

However, the typical BIOS fails to initialize the MSR,
and the typical Linux distro neglects to invoke x86_energy_perf_policy(8).

The result is that some modern hardware is running in hardware default, which is "performance" mode, rather than the intended design default of "normal" mode.

Initialize the MSR to the "normal" setting during kernel boot.

Of course, x86_energy_perf_policy(8) is available to change the default after boot, should the user have a policy preference.

cc [email protected]

...

Quoted 2 lines

Will it be possible to measure a performance difference between "performance" and "normal"? Yes, it will be possible.

Ok. So... what "serious bug" does this fix? You really need to use cc:
stable less. Tweaking performance/power ratio is _not_ serious bug.

Pavel

Signature · 2 lines
(english) http://www.livejournal.com/~pavelmachek
(cesky, pictures) http://atrey.karlin.mff.cuni.cz/~pavel/picture/horses/blog.html
author
#23 lore raw
Quoted 2 lines

Ok. So... what "serious bug" does this fix? You really need to use cc:
stable less. Tweaking performance/power ratio is _not_ serious bug.

While the performance difference may not be significant, the energy difference may be.

Some people think it is serious when Linux has worse out-of-box energy efficiency than Windows on the same hardware.

Some people think that it is serious when their Linux distribution has worse energy efficiency than competing Linux distributions.

Greg has given me a hard time for not cc'ing stable _enough_. I guess the folks that answer the stable mail get to decide...

cheers,
-Len

author
#24 lore raw
In reply to Pavel Machek
Quoted 2 lines

Ok. So... what "serious bug" does this fix? You really need to use cc:
stable less. Tweaking performance/power ratio is _not_ serious bug.

While the performance difference may not be significant, the energy difference may be.

Some people think it is serious when Linux has worse out-of-box energy efficiency than Windows on the same hardware.

Some people think that it is serious when their Linux distribution has worse energy efficiency than competing Linux distributions.

Greg has given me a hard time for not cc'ing stable _enough_. I guess the folks that answer the stable mail get to decide...

cheers,
-Len