psched_ratecfg_precompute__(9)

🦋 psched_ratecfg_precompute__(9) 💖

okay so welcome to the Kernel API Manual, bestie — this one is all about pre-computing reciprocal division values, and it’s lowkey iconic ✨

---

💅 NAME 💅

psched_ratecfg_precompute__ — Pre-compute values for reciprocal division

---

✨ SYNOPSIS ✨

void psched_ratecfg_precompute__ (u64 rate , u32 *mult , u8 *shift );

---

🎀 ARGUMENTS 🎀

• rate — Rate to compute reciprocal division values of
• mult — Multiplier for reciprocal division
• shift — Shift for reciprocal division

---

🌸 DESCRIPTION 🌸

The multiplier and shift for reciprocal division by rate are stored in mult and shift. That’s the whole vibe 💫

The deal here is to replace a divide by a reciprocal one in the fast path — and a reciprocal divide is literally a multiply and a shift. Slay.

The normal formula would be:

time_in_ns = (NSEC_PER_SEC * len) / rate_bps

So we compute mult/shift to use instead:

time_in_ns = (len * mult) >> shift;

We try to get the highest possible mult value for accuracy, but we have to make sure no overflows will ever happen. No cap, overflow is not cute 😭🔥

Also: reciprocal_value is not used here — it doesn’t handle 64-bit values. Periodt.

---

👑 SEE ALSO 👑

• Kernel file ./net/sched/sch_generic.c
• netif_carrier_on(9)
• netif_carrier_off(9)
• netif_carrier_event(9)
• dev_deactivate_many(9)

---

okay bye bestie 🫶 — source/version: ./net/sched, dated October 2026, and this whole page is psched_ratecfg_precompute__(9) 💖✨