Skip to content

SCHED_MIC: regression test / benchmark for hybrid placement #393

@laffer1

Description

@laffer1

Follow-up to #346 / SCHED_MIC (#385).

#346's acceptance criteria include a simple test/benchmark (CPU affinity / placement verification). None was committed; Intel detection was only validated ad hoc on real Alder Lake hardware (Core i7-1260P: 8 P-thread class 1, 8 E-core class 2).

Goal: add a committed test that (a) checks kern.sched.core_class classification against expected per-SKU counts and (b) verifies fill order (P -> E -> SMT sibling -> LP, and AMD cache-CCD packing with prefer_compute toggling). Could live under tests/ and/or as a tools script.

Relevant code: kern.sched.core_class and the tunables in sys/kern/sched_mic.c.

Metadata

Metadata

Assignees

No one assigned

    Labels

    enhancementNew feature or request

    Type

    No type
    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions