Skip to contentSkip to conversation

x86/cpu/intel: Fix the constant_tsc model check for Pentium 4

PATCH Superseded Suchit Karunakaran 2 participants 3 messages Aug 4, 2025 – Aug 5, 2025

Subsystems (2)

v4 of 5 Superseded Latest is v5 →

lore
Share & tools
lore
#0 lore raw

PATCH v4 1/1 x86/cpu/intel: Fix the constant_tsc model check for Pentium 4

Pentium 4's which are INTEL_P4_PRESCOTT (model 0x03) and later have a constant TSC. This was correctly captured until commit fadb6f569b10 ("x86/cpu/intel: Limit the non-architectural constant_tsc model checks"). In that commit, an error was introduced while selecting the last P4 model (0x06) as the upper bound. Model 0x06 was transposed to INTEL_P4_WILLAMETTE, which is just plain wrong. That was presumably a simple typo, probably just copying and pasting the wrong P4 model. Fix the constant TSC logic to cover all later P4 models. End at INTEL_P4_CEDARMILL which accurately corresponds to the last P4 model.

Fixes fadb6f569b10 ("x86/cpu/intel: Limit the non-architectural constant_tsc model checks")
Cc <[email protected]> # v6.15

Changes since v3:
- Refined changelog

Changes since v2:
- Improve commit message

Changes since v1:
- Fix incorrect logic

---
 arch/x86/kernel/cpu/intel.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/cpu/intel.c +1 -1
index 076eaa41b8c8..6f5bd5dbc249 100644
@@ -262,7 +262,7 @@ static void early_init_intel(struct cpuinfo_x86 *c)
 	if (c->x86_power & (1 << 8)) {
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 		set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
-	} else if ((c->x86_vfm >= INTEL_P4_PRESCOTT && c->x86_vfm <= INTEL_P4_WILLAMETTE) ||
+	} else if ((c->x86_vfm >=  INTEL_P4_PRESCOTT && c->x86_vfm <= INTEL_P4_CEDARMILL) ||
 		   (c->x86_vfm >= INTEL_CORE_YONAH  && c->x86_vfm <= INTEL_IVYBRIDGE)) {
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 	}
Signature · 1 line
2.50.1
#1 lore raw

Hi Suchit

I would strongly suggest spending more time understanding reviewer feedback and incorporating it before sending another version. Most of the comments below are a repeat from previous reviews by multiple folks.

Quoted 9 lines · Suchit Karunakaran

Pentium 4's which are INTEL_P4_PRESCOTT (model 0x03) and later have a constant TSC. This was correctly captured until commit fadb6f569b10 ("x86/cpu/intel: Limit the non-architectural constant_tsc model checks"). In that commit, an error was introduced while selecting the last P4 model (0x06) as the upper bound. Model 0x06 was transposed to INTEL_P4_WILLAMETTE, which is just plain wrong. That was presumably a simple typo, probably just copying and pasting the wrong P4 model. Fix the constant TSC logic to cover all later P4 models. End at INTEL_P4_CEDARMILL which accurately corresponds to the last P4 model.

Please use proper spacing and line breaks. Posting this as a single chunk makes it very hard to read.

Quoted 6 lines
Fixes fadb6f569b10 ("x86/cpu/intel: Limit the non-architectural constant_tsc model checks")
Cc <[email protected]> # v6.15

No blank lines between these 3 statements.

Please take all review comments seriously, even if they are nits.

https://lore.kernel.org/lkml/2025073013-stimulus-snowdrift-d28c@gregkh/

Quoted 9 lines

Changes since v3:
- Refined changelog

Changes since v2:
- Improve commit message

Changes since v1:
- Fix incorrect logic

Patch-to-patch changes go below the --- line.

You have been provided the same feedback by other folks as well.

https://lore.kernel.org/lkml/[email protected]/

https://lore.kernel.org/lkml/[email protected]/

If you are not sure, please look at other submissions to the mailing list that do this.

Quoted 9 lines
---
 arch/x86/kernel/cpu/intel.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/cpu/intel.c +0 -0
index 076eaa41b8c8..6f5bd5dbc249 100644

arch/x86/kernel/cpu/intel.c static void early_init_intel(struct cpuinfo_x86 *c) · line 266

Quoted 6 lines
arch/x86/kernel/cpu/intel.c +1 -1
@@ -262,7 +262,7 @@ static void early_init_intel(struct cpuinfo_x86 *c)
 	if (c->x86_power & (1 << 8)) {
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 		set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
-	} else if ((c->x86_vfm >= INTEL_P4_PRESCOTT && c->x86_vfm <= INTEL_P4_WILLAMETTE) ||
+	} else if ((c->x86_vfm >=  INTEL_P4_PRESCOTT && c->x86_vfm <= INTEL_P4_CEDARMILL) ||
 		   (c->x86_vfm >= INTEL_CORE_YONAH  && c->x86_vfm <= INTEL_IVYBRIDGE)) {

Again, this changes the previous alignment. You do not need the extra space before INTEL_P4_PRESCOTT. Avoiding that would keep both the lines aligned.

https://lore.kernel.org/lkml/[email protected]/

It's acceptable to ask clarifying questions, but disregarding review comments frustrates reviewers and inclined to ignore future submissions.

Quoted 2 lines
arch/x86/kernel/cpu/intel.c +0 -0
 		set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
 	}
#2 lore raw
Quoted 77 lines · Sohil Mehta

Hi Suchit

I would strongly suggest spending more time understanding reviewer feedback and incorporating it before sending another version. Most of the comments below are a repeat from previous reviews by multiple folks.

Quoted 9 lines · Suchit Karunakaran

Pentium 4's which are INTEL_P4_PRESCOTT (model 0x03) and later have a constant TSC. This was correctly captured until commit fadb6f569b10 ("x86/cpu/intel: Limit the non-architectural constant_tsc model checks"). In that commit, an error was introduced while selecting the last P4 model (0x06) as the upper bound. Model 0x06 was transposed to INTEL_P4_WILLAMETTE, which is just plain wrong. That was presumably a simple typo, probably just copying and pasting the wrong P4 model. Fix the constant TSC logic to cover all later P4 models. End at INTEL_P4_CEDARMILL which accurately corresponds to the last P4 model.

Please use proper spacing and line breaks. Posting this as a single chunk makes it very hard to read.

Quoted 6 lines
Fixes fadb6f569b10 ("x86/cpu/intel: Limit the non-architectural constant_tsc model checks")
Cc <[email protected]> # v6.15

No blank lines between these 3 statements.

Please take all review comments seriously, even if they are nits.

https://lore.kernel.org/lkml/2025073013-stimulus-snowdrift-d28c@gregkh/

Quoted 9 lines

Changes since v3:
- Refined changelog

Changes since v2:
- Improve commit message

Changes since v1:
- Fix incorrect logic

Patch-to-patch changes go below the --- line.

You have been provided the same feedback by other folks as well.

https://lore.kernel.org/lkml/[email protected]/

https://lore.kernel.org/lkml/[email protected]/

If you are not sure, please look at other submissions to the mailing list that do this.

Quoted 9 lines
---
 arch/x86/kernel/cpu/intel.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
arch/x86/kernel/cpu/intel.c +0 -0
index 076eaa41b8c8..6f5bd5dbc249 100644

arch/x86/kernel/cpu/intel.c static void early_init_intel(struct cpuinfo_x86 *c) · line 266

Quoted 6 lines
arch/x86/kernel/cpu/intel.c +1 -1
@@ -262,7 +262,7 @@ static void early_init_intel(struct cpuinfo_x86 *c)
      if (c->x86_power & (1 << 8)) {
              set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
              set_cpu_cap(c, X86_FEATURE_NONSTOP_TSC);
-     } else if ((c->x86_vfm >= INTEL_P4_PRESCOTT && c->x86_vfm <= INTEL_P4_WILLAMETTE) ||
+     } else if ((c->x86_vfm >=  INTEL_P4_PRESCOTT && c->x86_vfm <= INTEL_P4_CEDARMILL) ||
                 (c->x86_vfm >= INTEL_CORE_YONAH  && c->x86_vfm <= INTEL_IVYBRIDGE)) {

Again, this changes the previous alignment. You do not need the extra space before INTEL_P4_PRESCOTT. Avoiding that would keep both the lines aligned.

Quoted 9 lines

https://lore.kernel.org/lkml/[email protected]/

It's acceptable to ask clarifying questions, but disregarding review comments frustrates reviewers and inclined to ignore future submissions.

Quoted 2 lines
arch/x86/kernel/cpu/intel.c +0 -0
              set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
      }

Hi Sohil,
Thank you very much for your time and the detailed feedback. I sincerely apologize for the repeated mistakes in my submission. It was never my intention to disregard any of the review comments. I was just focused on the commit description and, as a result, was careless in properly incorporating the prior suggestions and formatting requirements. I understand how frustrating this can be, and I truly appreciate the patience shown by you and others throughout the review process. I will take greater care moving forward to ensure that all feedback is fully addressed before resubmitting.
Thank you again for your time and guidance.