test__switch_tracking(9)

💅 test__switch_tracking — the kernel API manual, but make it cute 🎀

okay so bestie, welcome to test__switch_tracking(9) — yes that's a section 9 page, yes that means kernel API, and yes we're gonna get through it together 💖

✨ NAME ✨

test__switch_tracking — it's a test. and what's it testing? using sched_switch and tracking events. that's the whole vibe check 🦋

🌸 SYNOPSIS 🌸

int test__switch_tracking (struct test_suite *test , int subtest );

like, look at her. two params, returns an int, absolutely iconic 👑

🫶 ARGUMENTS 🫶

• test — honestly? -- undescribed --. the docs did not come through for us here, no cap 😭
• subtest — also -- undescribed --. same story, we're on our own with this one 🙈

ngl the manpage really said "figure it out" for both of these and i respect the honesty honestly 💀

🔥 DESCRIPTION 🔥

okay so here's the main character energy ✨

This function implements a test — a real one — that checks that sched_switch events and tracking events can be recorded for a workload. and "workload" here means the current process, just so we're clear 🧚

How's it doing that? With the evsel->core.system_wide flag and the evsel->tracking flag — those two, respectively, matching up with the two kinds of events in order, iykyk 💅

and there are other events hanging around in there that are sometimes enabled and sometimes disabled, which is honestly so real of them 🌈

so if you were wondering "can these events actually get recorded for the current process?" — that's the question, and this test is the answer. slay 🔥

⭐ SEE ALSO ⭐

Kernel file ./tools/perf/tests/switch-tracking.c — that's where the actual source lives, go look, it's giving main character 🎀

---

and that's the page bestie 🫶 straight from the ./tools/perf/tests source, dated October 2026 — catch you in the next manpage, mwah 💜✨💅