parse_ids(9)

💅 parse_ids(9)

okay so bestie, real talk — this is a Kernel API Manual page for parse_ids, and it's honestly giving "the group chat needs to know how metric IDs turn into events." The man page name is parse_ids and it lives in section 9, iykyk. 🎀

---

✨ NAME ✨

parse_ids — Build the event string for the ids and parse them creating an evlist. The encoded metric_ids are decoded.

Like, that's the whole personality: you hand it some encoded IDs, it builds the event string, parses them, and hands you back an evlist with everything decoded. Slay. 💅

---

🌸 SYNOPSIS 🌸

int parse_ids (bool metric_no_merge , bool fake_pmu , struct
expr_parse_ctx *ids , const char *modifier , bool group_events , const
bool tool_events[TOOL_PMU__EVENT_MAX] , struct evlist **out_evlist ,
const char *filter_pmu , bool cputype_filter );

Notice the return type: <cyan>int</cyan>. Not void, not vibes — an actual int. ✨

---

🦋 ARGUMENTS 🦋

okay so literally every single one of these matters, no skipping, periodt:

• metric_no_merge — is metric sharing explicitly disabled. fr fr, it's a yes/no situation.
• fake_pmu — use a fake PMU when testing metrics not supported by the current CPU. the "pretend the hardware is cooler than it is" flag 💀
• ids — the event identifiers parsed from a metric. this is the struct expr_parse_ctx * main character.
• modifier — any modifiers added to the events. just a const char *, doing its little thing.
• group_events — should events be placed in a weak group. note the word weak — that's the original's word, not me being shady 🫶
• tool_events — entries set true if the tool event of index could be present in the overall list of metrics. its type is const bool tool_events[TOOL_PMU__EVENT_MAX], and yes that array bound is TOOL_PMU__EVENT_MAX, exactly like that, don't touch it.
• out_evlist — the created list of events. it's a struct evlist **, so it's the pointer-to-pointer moment. iconic double-star energy ⭐
• filter_pmu — -- undescribed -- 😭 the original man page literally left this one blank, so I'm not gonna invent lore for it. that's the whole description. bestie, it's just const char *filter_pmu.
• cputype_filter — -- undescribed -- 🙈 same situation. the source page gives us nothing but the type, bool cputype_filter, and honestly? we respect the mystery.

not me being honest about the undescribed ones instead of making stuff up. no cap, that's the rule 💖

---

👑 SEE ALSO 👑

you will be reading these next, bestie, they're all in Kernel file ./tools/perf/util/metricgroup.c:

• setup_metric_events(9)
• resolve_metric(9)
• __add_metric(9)
• metric_list_cmp(9)
• default_metricgroup_cmp(9)
• metricgroup__add_metric(9)
• metricgroup__add_metric_list(9)
• find_tool_events(9)
• build_combined_expr_ctx(9)

that's the whole friend group, go say hi 💜

---

and that's the page!! 🍓 this little gem comes from ./tools/perf/util and it's dated October 2026, in the Kernel API Manual — catch you in the next man page, bestie, mwah 🧚✨💅