Forcibly enable some MISC_ENABLE features on Intel

PATCH Superseded Andy Lutomirski 3 participants 9 messages Aug 6, 2011 – Aug 9, 2011

Subsystems (2)

x86 ACPI

lore

Share & tools

lore

Andy Lutomirski

author

2011-08-06 11:42 UTC #0 lore raw

v2 cover Forcibly enable some MISC_ENABLE features on Intel

Intel allows BIOS or the OS to enable or disable some CPU fueatures via IA32_MISC_ENABLE. I have machines that don't enable fast strings or monitor/mwait in BIOS, so do it on bootup instead.

The Intel SDM volume 3, appendix B.1 says that the OS should not touch the monitor enable bit if SSE3 is not present, which presumably means that the OS may touch that bit if SSE3 is present. In any case, these patches seem to work.

Changes from v1:
- Display FW_WARN messages.
- Don't change the kmemcheck message.
- Improve the fast string comment.
- Improve the changelogs.

Andy Lutomirski (2):
x86: Enable fast strings on Intel if BIOS hasn't already
x86: Enable monitor/mwait on Intel if BIOS hasn't already

 arch/x86/kernel/cpu/intel.c |   50 ++++++++++++++++++++++++++++++++++++++++--
 1 files changed, 47 insertions(+), 3 deletions(-)

Signature · 1 line

1.7.6

Andy Lutomirski

author

2011-08-06 11:42 UTC #1 lore raw

PATCH v2 1/2 x86: Enable fast strings on Intel if BIOS hasn't already

Intel SDM volume 3A, 8.4.2 says:

Software can disable fast-string operation by clearing the fast-string-enable bit (bit 0) of IA32_MISC_ENABLE MSR. However, Intel recomments that system software always enable fast-string operation.

The Intel DQ67SW board (with latest BIOS) disables fast string operations if TXT is enabled. A Lenovo X220 disables it regardless of TXT setting. I doubt I'm the only person with a dumb BIOS like this.

Signed-off-by Andy Lutomirski <[email protected]>

---
 arch/x86/kernel/cpu/intel.c |   25 ++++++++++++++++++++++---
 1 files changed, 22 insertions(+), 3 deletions(-)

arch/x86/kernel/cpu/intel.c +22 -3

index ed6086e..c80ab41 100644
@@ -30,6 +30,7 @@
 static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
 {
 	u64 misc_enable;
+	bool allow_fast_string = true;
 
 	/* Unmask CPUID levels if masked: */
 	if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {

@@ -118,8 +119,9 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
 	 * (model 2) with the same problem.
 	 */
 	if (c->x86 == 15) {
-		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+		allow_fast_string = false;
 
+		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
 		if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) {
 			printk(KERN_INFO "kmemcheck: Disabling fast string operations\n");

@@ -130,11 +132,28 @@ static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
 #endif
 
 	/*
-	 * If fast string is not enabled in IA32_MISC_ENABLE for any reason,
-	 * clear the fast string and enhanced fast string CPU capabilities.
+	 * If BIOS didn't enable fast string operation, try to enable
+	 * it ourselves.  If that fails, then clear the fast string
+	 * and enhanced fast string CPU capabilities.
 	 */
 	if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
 		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+		if (allow_fast_string &&
+		    !(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) {
+			misc_enable |= MSR_IA32_MISC_ENABLE_FAST_STRING;
+			wrmsr_safe(MSR_IA32_MISC_ENABLE, (u32)misc_enable,
+				   (u32)(misc_enable >> 32));
+
+			/* Re-read to make sure it stuck. */
+			rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+			if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)
+				printk(KERN_WARNING FW_WARN "CPU #%d: "
+				       "IA32_MISC_ENABLE.FAST_STRING_ENABLE "
+				       "was not set", c->cpu_index);
+		}
+
 		if (!(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) {
 			printk(KERN_INFO "Disabled fast string operations\n");
 			setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);

Signature · 1 line

1.7.6

Yu, Fenghua

2011-08-06 19:34 UTC #2 lore raw

RE: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already

Quoted 70 lines

-----Original Message-----
From: Andy Lutomirski [mailto:[email protected]]
Sent: Saturday, August 06, 2011 4:43 AM
To: [email protected]; [email protected]
Cc: Yu, Fenghua; Matthew Garrett; Len Brown; linux- [email protected]; Ingo Molnar; Andy Lutomirski
Subject: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Intel SDM volume 3A, 8.4.2 says:
Software can disable fast-string operation by clearing the fast-string-enable bit (bit 0) of IA32_MISC_ENABLE MSR. However, Intel recomments that system software always enable fast-string operation.
The Intel DQ67SW board (with latest BIOS) disables fast string operations if TXT is enabled. A Lenovo X220 disables it regardless of TXT setting. I doubt I'm the only person with a dumb BIOS like this.
Signed-off-by Andy Lutomirski <[email protected]>
---
 arch/x86/kernel/cpu/intel.c |   25 ++++++++++++++++++++++---
 1 files changed, 22 insertions(+), 3 deletions(-)
arch/x86/kernel/cpu/intel.c +3 -1
index ed6086e..c80ab41 100644
@@ -30,6 +30,7 @@
 static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
 {
 	u64 misc_enable;
+	bool allow_fast_string = true;
 
 	/* Unmask CPUID levels if masked: */
 	if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
@@ -118,8 +119,9 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
 	 * (model 2) with the same problem.
 	 */
 	if (c->x86 == 15) {
-		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+		allow_fast_string = false;
 
+		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
 		if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) {
 			printk(KERN_INFO "kmemcheck: Disabling fast string
operations\n");
arch/x86/kernel/cpu/intel.c +0 -2
@@ -130,11 +132,28 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
 #endif
 
 	/*
-	 * If fast string is not enabled in IA32_MISC_ENABLE for any
 reason,
-	 * clear the fast string and enhanced fast string CPU
capabilities.
arch/x86/kernel/cpu/intel.c +7 -0
+	 * If BIOS didn't enable fast string operation, try to enable
+	 * it ourselves.  If that fails, then clear the fast string
+	 * and enhanced fast string CPU capabilities.
 	 */
 	if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
 		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+		if (allow_fast_string &&
+		    !(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) {
+			misc_enable |= MSR_IA32_MISC_ENABLE_FAST_STRING;

arch/x86/kernel/cpu/intel.c static void __cpuinit early_init_intel(struct cpuinfo_x86 *c) · line 154

Quoted 10 lines

arch/x86/kernel/cpu/intel.c +10 -0

+			wrmsr_safe(MSR_IA32_MISC_ENABLE, (u32)misc_enable,
+				   (u32)(misc_enable >> 32));
+
+			/* Re-read to make sure it stuck. */
+			rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+			if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)
+				printk(KERN_WARNING FW_WARN "CPU #%d: "
+				       "IA32_MISC_ENABLE.FAST_STRING_ENABLE "
+				       "was not set", c->cpu_index);

This printk is redundant because the same info is dumped in the below printk. Plus it's not firmware's issue if we can not set fast string bit 0 in MISC_ENABLE register. So I don't think FW_WARN is right.

I would suggest to remove this printk (and the if () statement) and change KERN_INFO to KERN_WARNING in the following printk.

Quoted 8 lines

arch/x86/kernel/cpu/intel.c +2 -0
+		}
+
 		if (!(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) {
 			printk(KERN_INFO "Disabled fast string
operations\n");
setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);
--
1.7.6

Andrew Lutomirski

author

2011-08-06 23:32 UTC #3 lore raw

Re: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already

Quoted 73 lines · Fenghua

Quoted 73 lines
-----Original Message-----
From: Andy Lutomirski [mailto:[email protected]]
Sent: Saturday, August 06, 2011 4:43 AM
To: [email protected]; [email protected]
Cc: Yu, Fenghua; Matthew Garrett; Len Brown; linux- [email protected]; Ingo Molnar; Andy Lutomirski
Subject: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Intel SDM volume 3A, 8.4.2 says:
Software can disable fast-string operation by clearing the fast-string-enable bit (bit 0) of IA32_MISC_ENABLE MSR. However, Intel recomments that system software always enable fast-string operation.
The Intel DQ67SW board (with latest BIOS) disables fast string operations if TXT is enabled. A Lenovo X220 disables it regardless of TXT setting. I doubt I'm the only person with a dumb BIOS like this.
Signed-off-by Andy Lutomirski <[email protected]>
---
arch/x86/kernel/cpu/intel.c | 25 ++++++++++++++++++++++---
 1 files changed, 22 insertions(+), 3 deletions(-)
arch/x86/kernel/cpu/intel.c +3 -1
index ed6086e..c80ab41 100644
@@ -30,6 +30,7 @@
  static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
  {
       u64 misc_enable;
+     bool allow_fast_string = true;
 
       /* Unmask CPUID levels if masked: */
       if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
@@ -118,8 +119,9 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
        * (model 2) with the same problem.
        */
       if (c->x86 == 15) {
-             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+             allow_fast_string = false;
 
+             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
               if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) {
printk(KERN_INFO "kmemcheck: Disabling fast string operations\n");
arch/x86/kernel/cpu/intel.c +0 -2
@@ -130,11 +132,28 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
  #endif
 
       /*
-      * If fast string is not enabled in IA32_MISC_ENABLE for any
 reason,
-      * clear the fast string and enhanced fast string CPU
capabilities.
+ * If BIOS didn't enable fast string operation, try to enable
+ * it ourselves. If that fails, then clear the fast string
+ * and enhanced fast string CPU capabilities.
*/
if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
arch/x86/kernel/cpu/intel.c +7 -0
+
+             if (allow_fast_string &&
+                 !(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) {
+                     misc_enable |= MSR_IA32_MISC_ENABLE_FAST_STRING;
+                     wrmsr_safe(MSR_IA32_MISC_ENABLE, (u32)misc_enable,
+                                (u32)(misc_enable >> 32));
+

Quoted 8 lines

Quoted 7 lines
+ /* Re-read to make sure it stuck. */
+ rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); +
+ if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)
+ printk(KERN_WARNING FW_WARN "CPU #%d: "
+ "IA32_MISC_ENABLE.FAST_STRING_ENABLE "
+ "was not set", c->cpu_index);
This printk is redundant because the same info is dumped in the below printk. Plus it's not firmware's issue if we can not set fast string bit 0 in MISC_ENABLE register. So I don't think FW_WARN is right.

Huh? This is the success path -- the patch prints the warning if flipping the bit worked...

Quoted 9 lines

I would suggest to remove this printk (and the if () statement) and change KERN_INFO to KERN_WARNING in the following printk.
Quoted 6 lines
+ }
+
if (!(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)) { printk(KERN_INFO "Disabled fast string operations\n");
setup_clear_cpu_cap(X86_FEATURE_REP_GOOD);

...and this (which could be more clear) is displayed if flipping the bit did not work, in which case the problem isn't BIOS's fault.

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

Yu, Fenghua

2011-08-07 23:45 UTC #4 lore raw

RE: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already

Quoted 98 lines

-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Andrew Lutomirski
Sent: Saturday, August 06, 2011 4:32 PM
To: Yu, Fenghua
Cc: [email protected]; [email protected]; Matthew Garrett; Len Brown; [email protected]; Ingo Molnar
Subject: Re: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Quoted 21 lines · Fenghua
Quoted 21 lines
-----Original Message-----
From: Andy Lutomirski [mailto:[email protected]]
Sent: Saturday, August 06, 2011 4:43 AM
To: [email protected]; [email protected]
Cc: Yu, Fenghua; Matthew Garrett; Len Brown; linux- [email protected]; Ingo Molnar; Andy Lutomirski
Subject: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Intel SDM volume 3A, 8.4.2 says:
Software can disable fast-string operation by clearing the fast-string-enable bit (bit 0) of IA32_MISC_ENABLE MSR. However, Intel recomments that system software always enable fast-string operation.
The Intel DQ67SW board (with latest BIOS) disables fast string operations if TXT is enabled. A Lenovo X220 disables it regardless of TXT setting. I doubt I'm the only person with a dumb BIOS like this.
Quoted 6 lines
Review
Quoted 6 lines
Signed-off-by Andy Lutomirski <[email protected]>
---
arch/x86/kernel/cpu/intel.c | 25 ++++++++++++++++++++++---
 1 files changed, 22 insertions(+), 3 deletions(-)
(quoted diff) +0 -0
b/arch/x86/kernel/cpu/intel.c
Quoted 19 lines
Quoted 19 lines
index ed6086e..c80ab41 100644
arch/x86/kernel/cpu/intel.c +2 -1
@@ -30,6 +30,7 @@
  static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
  {
       u64 misc_enable;
+     bool allow_fast_string = true;
 
       /* Unmask CPUID levels if masked: */
       if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
@@ -118,8 +119,9 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
        * (model 2) with the same problem.
        */
       if (c->x86 == 15) {
-             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+             allow_fast_string = false;
 
Quoted 3 lines
Quoted 3 lines
arch/x86/kernel/cpu/intel.c +1 -0
+             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
               if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) {
printk(KERN_INFO "kmemcheck: Disabling fast
string
Quoted 6 lines
Quoted 6 lines
operations\n");
arch/x86/kernel/cpu/intel.c +0 -0
@@ -130,11 +132,28 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
  #endif
 
Quoted 14 lines
Quoted 6 lines
arch/x86/kernel/cpu/intel.c +0 -2
       /*
-      * If fast string is not enabled in IA32_MISC_ENABLE for any
 reason,
-      * clear the fast string and enhanced fast string CPU
capabilities.
+ * If BIOS didn't enable fast string operation, try to enable
Quoted 8 lines
+ * it ourselves. If that fails, then clear the fast string
+ * and enhanced fast string CPU capabilities.
*/
if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); +
+ if (allow_fast_string &&
+ !(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING))
{
Quoted 1 line
Quoted 1 line
+ misc_enable |=
MSR_IA32_MISC_ENABLE_FAST_STRING;
Quoted 1 line
Quoted 1 line
+ wrmsr_safe(MSR_IA32_MISC_ENABLE,
(u32)misc_enable,
Quoted 6 lines
arch/x86/kernel/cpu/intel.c static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
Quoted 6 lines
+ (u32)(misc_enable >> 32));
arch/x86/kernel/cpu/intel.c +5 -0
+
+                     /* Re-read to make sure it stuck. */
+                     rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+                     if (misc_enable &
MSR_IA32_MISC_ENABLE_FAST_STRING)
Quoted 1 line
Quoted 1 line
+ printk(KERN_WARNING FW_WARN "CPU #%d:
"
Quoted 1 line
Quoted 1 line
+

Quoted 8 lines

"IA32_MISC_ENABLE.FAST_STRING_ENABLE "
Quoted 2 lines
Quoted 1 line
+ "was not set", c->cpu_index);
This printk is redundant because the same info is dumped in the below
printk. Plus it's not firmware's issue if we can not set fast string bit 0 in MISC_ENABLE register. So I don't think FW_WARN is right.
Huh? This is the success path -- the patch prints the warning if flipping the bit worked...

Ok. I see.

Still I would suggest to remove this printk. This info will be printed on every single CPU if BIOS doesn't enable fast string. This could flood the log buffer on big machines or many threads machines. And the info doesn't really provide useful message because we enable fast string in OS anyway. Reporting BIOS issue by this info probably will be viewed as low priority or ignored.

The following info is more important because it reports an abnormal behavior.

Thanks.

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

Ingo Molnar

2011-08-09 15:01 UTC #5 lore raw

Re: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already

Quoted 104 lines · Yu, Fenghua

Quoted 104 lines
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Andrew Lutomirski
Sent: Saturday, August 06, 2011 4:32 PM
To: Yu, Fenghua
Cc: [email protected]; [email protected]; Matthew Garrett; Len Brown; [email protected]; Ingo Molnar
Subject: Re: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Quoted 21 lines · Fenghua
Quoted 21 lines
-----Original Message-----
From: Andy Lutomirski [mailto:[email protected]]
Sent: Saturday, August 06, 2011 4:43 AM
To: [email protected]; [email protected]
Cc: Yu, Fenghua; Matthew Garrett; Len Brown; linux- [email protected]; Ingo Molnar; Andy Lutomirski
Subject: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Intel SDM volume 3A, 8.4.2 says:
Software can disable fast-string operation by clearing the fast-string-enable bit (bit 0) of IA32_MISC_ENABLE MSR. However, Intel recomments that system software always enable fast-string operation.
The Intel DQ67SW board (with latest BIOS) disables fast string operations if TXT is enabled. A Lenovo X220 disables it regardless of TXT setting. I doubt I'm the only person with a dumb BIOS like this.
Quoted 6 lines
Review
Quoted 6 lines
Signed-off-by Andy Lutomirski <[email protected]>
---
arch/x86/kernel/cpu/intel.c | 25 ++++++++++++++++++++++---
 1 files changed, 22 insertions(+), 3 deletions(-)
(quoted diff) +0 -0
b/arch/x86/kernel/cpu/intel.c
Quoted 19 lines
Quoted 19 lines
index ed6086e..c80ab41 100644
arch/x86/kernel/cpu/intel.c +2 -1
@@ -30,6 +30,7 @@
  static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
  {
       u64 misc_enable;
+     bool allow_fast_string = true;
 
       /* Unmask CPUID levels if masked: */
       if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
@@ -118,8 +119,9 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
        * (model 2) with the same problem.
        */
       if (c->x86 == 15) {
-             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+             allow_fast_string = false;
 
Quoted 3 lines
Quoted 3 lines
arch/x86/kernel/cpu/intel.c +1 -0
+             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
               if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) {
printk(KERN_INFO "kmemcheck: Disabling fast
string
Quoted 6 lines
Quoted 6 lines
operations\n");
arch/x86/kernel/cpu/intel.c +0 -0
@@ -130,11 +132,28 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
  #endif
 
Quoted 14 lines
Quoted 6 lines
arch/x86/kernel/cpu/intel.c +0 -2
       /*
-      * If fast string is not enabled in IA32_MISC_ENABLE for any
 reason,
-      * clear the fast string and enhanced fast string CPU
capabilities.
+ * If BIOS didn't enable fast string operation, try to enable
Quoted 8 lines
+ * it ourselves. If that fails, then clear the fast string
+ * and enhanced fast string CPU capabilities.
*/
if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); +
+ if (allow_fast_string &&
+ !(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING))
{
Quoted 1 line
Quoted 1 line
+ misc_enable |=
MSR_IA32_MISC_ENABLE_FAST_STRING;
Quoted 1 line
Quoted 1 line
+ wrmsr_safe(MSR_IA32_MISC_ENABLE,
(u32)misc_enable,
Quoted 6 lines
arch/x86/kernel/cpu/intel.c static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
Quoted 6 lines
+ (u32)(misc_enable >> 32));
arch/x86/kernel/cpu/intel.c +5 -0
+
+                     /* Re-read to make sure it stuck. */
+                     rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+                     if (misc_enable &
MSR_IA32_MISC_ENABLE_FAST_STRING)
Quoted 1 line
Quoted 1 line
+ printk(KERN_WARNING FW_WARN "CPU #%d:
"
Quoted 1 line
Quoted 1 line
+
"IA32_MISC_ENABLE.FAST_STRING_ENABLE "
Quoted 2 lines
Quoted 1 line
+ "was not set", c->cpu_index);
This printk is redundant because the same info is dumped in the below
printk. Plus it's not firmware's issue if we can not set fast string bit 0 in MISC_ENABLE register. So I don't think FW_WARN is right.

Quoted 7 lines

Quoted 2 lines
Huh? This is the success path -- the patch prints the warning if flipping the bit worked...
Ok. I see.
Still I would suggest to remove this printk. This info will be printed on every single CPU if BIOS doesn't enable fast string.

just do printk_once(). We want to inform the user that something's going on.

Thanks,

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

Andrew Lutomirski

author

2011-08-09 15:13 UTC #6 lore raw

Re: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already

Quoted 110 lines · Ingo Molnar

Quoted 107 lines · Yu, Fenghua
Quoted 106 lines
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Andrew Lutomirski
Sent: Saturday, August 06, 2011 4:32 PM
To: Yu, Fenghua
Cc: [email protected]; [email protected]; Matthew Garrett; Len Brown; [email protected]; Ingo Molnar
Subject: Re: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Quoted 21 lines · Fenghua
Quoted 21 lines
-----Original Message-----
From: Andy Lutomirski [mailto:[email protected]]
Sent: Saturday, August 06, 2011 4:43 AM
To: [email protected]; [email protected]
Cc: Yu, Fenghua; Matthew Garrett; Len Brown; linux- [email protected]; Ingo Molnar; Andy Lutomirski
Subject: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Intel SDM volume 3A, 8.4.2 says:
Software can disable fast-string operation by clearing the fast-string-enable bit (bit 0) of IA32_MISC_ENABLE MSR. However, Intel recomments that system software always enable fast-string operation.
The Intel DQ67SW board (with latest BIOS) disables fast string operations if TXT is enabled. A Lenovo X220 disables it regardless of TXT setting. I doubt I'm the only person with a dumb BIOS like this.
Quoted 6 lines
Review
Quoted 6 lines
Signed-off-by Andy Lutomirski <[email protected]>
---
arch/x86/kernel/cpu/intel.c | 25 ++++++++++++++++++++++---
 1 files changed, 22 insertions(+), 3 deletions(-)
(quoted diff) +0 -0
b/arch/x86/kernel/cpu/intel.c
Quoted 19 lines
Quoted 19 lines
index ed6086e..c80ab41 100644
arch/x86/kernel/cpu/intel.c +2 -1
@@ -30,6 +30,7 @@
  static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
  {
       u64 misc_enable;
+     bool allow_fast_string = true;
 
       /* Unmask CPUID levels if masked: */
       if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
@@ -118,8 +119,9 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
        * (model 2) with the same problem.
        */
       if (c->x86 == 15) {
-             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+             allow_fast_string = false;
 
Quoted 3 lines
Quoted 3 lines
arch/x86/kernel/cpu/intel.c +1 -0
+             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
               if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING) {
printk(KERN_INFO "kmemcheck: Disabling fast
string
Quoted 6 lines
Quoted 6 lines
operations\n");
arch/x86/kernel/cpu/intel.c +0 -0
@@ -130,11 +132,28 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
  #endif
 
Quoted 14 lines
Quoted 6 lines
arch/x86/kernel/cpu/intel.c +0 -2
       /*
-      * If fast string is not enabled in IA32_MISC_ENABLE for any
 reason,
-      * clear the fast string and enhanced fast string CPU
capabilities.
+ * If BIOS didn't enable fast string operation, try to enable
Quoted 8 lines
+ * it ourselves. If that fails, then clear the fast string
+ * and enhanced fast string CPU capabilities.
*/
if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); +
+ if (allow_fast_string &&
+ !(misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING))
{
Quoted 1 line
Quoted 1 line
+ misc_enable |=
MSR_IA32_MISC_ENABLE_FAST_STRING;
Quoted 1 line
Quoted 1 line
+ wrmsr_safe(MSR_IA32_MISC_ENABLE,
(u32)misc_enable,
Quoted 6 lines
arch/x86/kernel/cpu/intel.c static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
Quoted 6 lines
+ (u32)(misc_enable >> 32));
arch/x86/kernel/cpu/intel.c +5 -0
+
+                     /* Re-read to make sure it stuck. */
+                     rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+                     if (misc_enable &
MSR_IA32_MISC_ENABLE_FAST_STRING)
Quoted 1 line
Quoted 1 line
+ printk(KERN_WARNING FW_WARN "CPU #%d:
"
Quoted 1 line
Quoted 1 line
+
"IA32_MISC_ENABLE.FAST_STRING_ENABLE "
Quoted 2 lines
Quoted 1 line
+ "was not set", c->cpu_index);
This printk is redundant because the same info is dumped in the below
printk. Plus it's not firmware's issue if we can not set fast string bit 0 in MISC_ENABLE register. So I don't think FW_WARN is right.
Huh? This is the success path -- the patch prints the warning if flipping the bit worked...

Quoted 7 lines

Quoted 4 lines
Ok. I see.
Still I would suggest to remove this printk. This info will be printed on every single CPU if BIOS doesn't enable fast string.
just do printk_once(). We want to inform the user that something's going on.

Already there in v3.

--Andy

Quoted 5 lines

Thanks,
Ingo

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

Yu, Fenghua

2011-08-08 18:43 UTC #7 lore raw

RE: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already

In reply to Andrew Lutomirski

Quoted 121 lines

-----Original Message-----
From: Yu, Fenghua
Sent: Sunday, August 07, 2011 4:45 PM
To: 'Andrew Lutomirski'
Cc: [email protected]; [email protected]; Matthew Garrett; Len Brown; [email protected]; Ingo Molnar
Subject: RE: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Quoted 6 lines
-----Original Message-----
From: [email protected] [mailto:[email protected]] On Behalf Of Andrew Lutomirski
Sent: Saturday, August 06, 2011 4:32 PM
To: Yu, Fenghua
Cc: [email protected]; [email protected]; Matthew Garrett;
Len
Quoted 4 lines
Brown; [email protected]; Ingo Molnar
Subject: Re: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Quoted 20 lines
Quoted 18 lines · Fenghua
Quoted 18 lines
-----Original Message-----
From: Andy Lutomirski [mailto:[email protected]]
Sent: Saturday, August 06, 2011 4:43 AM
To: [email protected]; [email protected]
Cc: Yu, Fenghua; Matthew Garrett; Len Brown; linux- [email protected]; Ingo Molnar; Andy Lutomirski
Subject: [PATCH v2 1/2] x86: Enable fast strings on Intel if BIOS hasn't already
Intel SDM volume 3A, 8.4.2 says:
Software can disable fast-string operation by clearing the fast-string-enable bit (bit 0) of IA32_MISC_ENABLE MSR. However, Intel recomments that system software always enable fast-string operation.
The Intel DQ67SW board (with latest BIOS) disables fast string operations if TXT is enabled. A Lenovo X220 disables it
regardless
Quoted 29 lines
Quoted 9 lines
Review
Quoted 9 lines
of TXT setting. I doubt I'm the only person with a dumb BIOS like this.
Signed-off-by Andy Lutomirski <[email protected]>
---
arch/x86/kernel/cpu/intel.c | 25 ++++++++++++++++++++++---
 1 files changed, 22 insertions(+), 3 deletions(-)
(quoted diff) +0 -0
b/arch/x86/kernel/cpu/intel.c
Quoted 19 lines
Quoted 19 lines
index ed6086e..c80ab41 100644
arch/x86/kernel/cpu/intel.c +2 -1
@@ -30,6 +30,7 @@
  static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
  {
       u64 misc_enable;
+     bool allow_fast_string = true;
 
       /* Unmask CPUID levels if masked: */
       if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) {
@@ -118,8 +119,9 @@ static void __cpuinit early_init_intel(struct
 cpuinfo_x86 *c)
        * (model 2) with the same problem.
        */
       if (c->x86 == 15) {
-             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+             allow_fast_string = false;
 
Quoted 2 lines
Quoted 2 lines
arch/x86/kernel/cpu/intel.c static void __cpuinit early_init_intel(struct · line 126
Quoted 2 lines
arch/x86/kernel/cpu/intel.c +1 -0
+             rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
               if (misc_enable & MSR_IA32_MISC_ENABLE_FAST_STRING)
{
Quoted 5 lines
Quoted 1 line
Quoted 1 line
printk(KERN_INFO "kmemcheck: Disabling fast
string
Quoted 3 lines
arch/x86/kernel/cpu/intel.c static void __cpuinit
Quoted 3 lines
operations\n");
arch/x86/kernel/cpu/intel.c +0 -0
@@ -130,11 +132,28 @@ static void __cpuinit
early_init_intel(struct
Quoted 9 lines
Quoted 9 lines
Quoted 9 lines
arch/x86/kernel/cpu/intel.c +0 -2
 cpuinfo_x86 *c)
  #endif
 
       /*
-      * If fast string is not enabled in IA32_MISC_ENABLE for any
 reason,
-      * clear the fast string and enhanced fast string CPU
capabilities.
+ * If BIOS didn't enable fast string operation, try to
enable
Quoted 8 lines
Quoted 8 lines
Quoted 8 lines
+ * it ourselves. If that fails, then clear the fast string
+ * and enhanced fast string CPU capabilities.
*/
if (c->x86 > 6 || (c->x86 == 6 && c->x86_model >= 0xd)) { rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable); +
+ if (allow_fast_string &&
+ !(misc_enable &
MSR_IA32_MISC_ENABLE_FAST_STRING))
Quoted 4 lines
{
Quoted 1 line
Quoted 1 line
+ misc_enable |=
MSR_IA32_MISC_ENABLE_FAST_STRING;
Quoted 1 line
Quoted 1 line
+ wrmsr_safe(MSR_IA32_MISC_ENABLE,
Quoted 9 lines
(u32)misc_enable,
Quoted 6 lines
arch/x86/kernel/cpu/intel.c static void __cpuinit early_init_intel(struct cpuinfo_x86 *c)
Quoted 6 lines
+ (u32)(misc_enable >> 32));
arch/x86/kernel/cpu/intel.c +5 -0
+
+                     /* Re-read to make sure it stuck. */
+                     rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+
+                     if (misc_enable &
MSR_IA32_MISC_ENABLE_FAST_STRING)
Quoted 1 line
Quoted 1 line
+ printk(KERN_WARNING FW_WARN "CPU
#%d:
Quoted 5 lines
"
Quoted 1 line
Quoted 1 line
+
"IA32_MISC_ENABLE.FAST_STRING_ENABLE "
Quoted 2 lines
Quoted 1 line
+ "was not set", c->cpu_index);
This printk is redundant because the same info is dumped in the
below
Quoted 3 lines
printk. Plus it's not firmware's issue if we can not set fast string bit 0 in MISC_ENABLE register. So I don't think FW_WARN is right.

Quoted 8 lines

Quoted 2 lines
Huh? This is the success path -- the patch prints the warning if flipping the bit worked...
Ok. I see.
Still I would suggest to remove this printk. This info will be printed on every single CPU if BIOS doesn't enable fast string. This could flood the log buffer on big machines or many threads machines. And the

On the bottom line, you need to use printk_once to reduce repeated message on each CPU.

Thanks.

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

Andy Lutomirski

author

2011-08-06 11:43 UTC #8 lore raw

PATCH v2 2/2 x86: Enable monitor/mwait on Intel if BIOS hasn't already

In reply to Andy Lutomirski

My Intel DQ67SW (latest BIOS) disables monitor/mwait on the boot CPU if TXT is enabled. We're lucky that the system works at all, since the feature is still enabled on other CPUs.

The obvious fix is to just re-enable it ourselves.

Signed-off-by Andy Lutomirski <[email protected]>

---
 arch/x86/kernel/cpu/intel.c |   25 +++++++++++++++++++++++++
 1 files changed, 25 insertions(+), 0 deletions(-)

arch/x86/kernel/cpu/intel.c +25 -0

index c80ab41..5a0150c 100644
@@ -493,6 +493,31 @@ static void __cpuinit init_intel(struct cpuinfo_x86 *c)
 			wrmsrl(MSR_IA32_ENERGY_PERF_BIAS, epb);
 		}
 	}
+
+	/* Enable monitor/mwait if BIOS didn't do it for us. */
+	if (!cpu_has(c, X86_FEATURE_MWAIT) && cpu_has(c, X86_FEATURE_XMM3)
+	    && c->x86 >= 6 && !(c->x86 == 6 && c->x86_model < 0x1c)
+	    && !(c->x86 == 0xf && c->x86_model < 3)) {
+		u64 misc_enable;
+		rdmsrl(MSR_IA32_MISC_ENABLE, misc_enable);
+		misc_enable |= MSR_IA32_MISC_ENABLE_MWAIT;
+
+		/*
+		 * Some non-SSE3 cpus will #GP.  We check for that,
+		 * but it can't hurt to be safe.
+		 */
+		wrmsr_safe(MSR_IA32_MISC_ENABLE, (u32)misc_enable,
+			   (u32)(misc_enable >> 32));
+
+		/* Re-read monitor capability. */
+		if (cpuid_ecx(1) & 0x8) {
+			set_cpu_cap(c, X86_FEATURE_MWAIT);
+
+			printk(KERN_WARNING FW_WARN "CPU #%d: "
+			       "IA32_MISC_ENABLE.ENABLE_MONITOR_FSM "
+			       "was not set\n", c->cpu_index);
+		}
+	}
 }
 
 #ifdef CONFIG_X86_32

Signature · 1 line

1.7.6