Re: [PATCH v3] x86/cpu/intel: Fix the constant_tsc model check for Pentium 4s

[Dave Hansen]

On 7/29/25 21:26, Suchit Karunakaran wrote:
> The logic to synthesize constant_tsc for Pentium 4s (Family 15) is
> wrong. Since INTEL_P4_PRESCOTT is numerically greater than
> INTEL_P4_WILLAMETTE, the logic always results in false and never sets
> X86_FEATURE_CONSTANT_TSC for any Pentium 4 model.
> The error was introduced while replacing the x86_model check with a VFM
> one. The original check was as follows:
>         if ((c->x86 == 0xf && c->x86_model >= 0x03) ||
>                 (c->x86 == 0x6 && c->x86_model >= 0x0e))
>                 set_cpu_cap(c, X86_FEATURE_CONSTANT_TSC);
> 
> Fix the logic to cover all Pentium 4 models from Prescott (model 3) to
> Cedarmill (model 6) which is the last model released in Family 15.

Could we have a slightly different changelog, please? The fact that the
logic results in the bit never getting set for P4's is IMNHO immaterial.
This looks like a plain and simple typo, not a logical error on the
patch author's part.

How about this as a changelog?

--

Pentium 4's which are INTEL_P4_PRESCOTT (mode 0x03) and later have a
constant TSC. This was correctly captured until fadb6f569b10
("x86/cpu/intel: Limit the non-architectural constant_tsc model
checks"). In that commit, the model was transposed from 0x03 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 is the last P4 model.