calculate_alt_copy_time_us(9)

💅 calculate_alt_copy_time_us(9) — the Kernel API Manual glow-up ✨

okay bestie, so this one lives in the calculate_alt_copy_time_us(9) corner of the kernel API manual, and honestly? it's a lil math gremlin and i respect it 🫶

---

✨ NAME ✨

calculate_alt_copy_time_us — crunches out a tight upper bound for the copy time across, like, all the planes and streams 🔥

---

🎀 SYNOPSIS 🎀

unsigned int calculate_alt_copy_time_us (const struct
dml2_core_calcs_calculate_alternate_params *p );

---

🌸 ARGUMENTS 🌸

• p — bestie… the manpage literally says -- undescribed -- 😭 so that's all we got. it's the const struct dml2_core_calcs_calculate_alternate_params * from the synopsis. no cap, that's the whole description.

---

🦋 DESCRIPTION 🦋

okay so this function's whole job is calculating a tight upper bound for the copy time over all planes and streams. and when we say "tight"? we mean it 💅 — it does the math assuming SVP0 and SVP1 cannot both overlap with prefetch (and prefetch is the one that potentially has the "most" amount of bytes to copy per dst line, so yeah, that's the spicy assumption).

the other iconic thing: it uses the exact NoOfDPP as calculated by DML — like, it does not take an input num_dpp param, she computes it herself 👑. why? because we want the copy time calculation to be precise, so we can minimize FW latency / overhead. main character energy fr fr.

---

🌈 SEE ALSO 🌈

• Kernel file — ./drivers/gpu/drm/amd/display/dc/dml2_0/dml21/src/dml2_core/dml2_core_dcn6_calcs_dchub.c
• get_plane_params(9)
• compute_pre_rec_first_hdl(9)
• calculate_copy_swaths(9)
• calculate_ub_copy_size_per_plane_per_dpp_per_svp(9)
• calculate_swath_params(9)

---

okay and before i go — the source path on this one is ./drivers/gpu/drm/amd/display/..., dated October 2026, and yes it's still calculate_alt_copy_time_us(9) at the bottom, we love a consistent footer 🍓✨ ty for coming to my ted talk, go read the kernel file, ily 💜🧚