drm_sched_e...add_gpu_time(9)

📦 version 7.3.1, in nixpkgs unstable

💖 drm_sched_e...add_gpu_time(9) 💖

okay bestie, so we're in the Kernel API Manual and today's main character is a function with a name so long it needs a snack break in the middle (drm_sched_e... — yes that ellipsis is literally part of it, iykyk). Buckle up 🎀

✨ NAME ✨

drm_sched_entity_stats_job_add_gpu_time — the one that accounts a job's execution time to its entity. it's giving "timekeeper energy" fr fr 🕰️

🌸 SYNOPSIS 🌸

ktime_t drm_sched_entity_stats_job_add_gpu_time (struct drm_sched_job *job );

look at that signature. ktime_t in, ktime_t out, one struct drm_sched_job *job because we keep it simple and slay 💅

🦋 ARGUMENTS 🦋

  • job — the scheduler job we're accounting for. that's it. that's the whole argument. she's low maintenance 👑

🧚 DESCRIPTION 🧚

so basically, this function accounts the execution time of job to its respective entity stats object. like, the job's time gets filed under the right entity's little stats bookkeeping situation. no cap, that's the whole vibe — the entity gets the credit for the time its job spent doing stuff ✨📊

💜 RETURN 💜

returns the job's real duration in micro seconds — as a ktime_t, which is iconic of it. so yes, you get the actual number, in microseconds, no fake numbers here, periodt 🫶

🌈 SEE ALSO 🌈

the kernel file that lives here, bestie:

./drivers/gpu/drm/scheduler/sched_entity.c

and the whole friend group (all (9) because they're kernel-internal queens 👑):

🫶 that's the whole page, no filler, no fluff — every flag-shaped detail intact because we do not skip the boring parts in this group chat 💖

and that's a wrap, bestie 🎀 this lil' page came straight out of ./drivers/gpu/drm/scheduler, dated October 2026, and she's filed under drm_sched_e...add_gpu_time(9) in the Kernel API Manual. mwah 💋✨

📦 where this page comes from