xe_devcoredump(9)

💅 xe_devcoredump(9) 💅

okay so this is a page from the Kernel API Manual and it's literally about the moment everything goes 💀 — welcome to xe_devcoredump, bestie 🫶

---

✨ NAME ✨

xe_devcoredump — take the required snapshots and initialize coredump device. no cap, that's the whole vibe 🎀

---

🌸 SYNOPSIS 🌸

void xe_devcoredump (struct xe_exec_queue *q , struct xe_sched_job *job
, const char *fmt , ... );

---

🦋 ARGUMENTS 🦋

• q — the faulty xe_exec_queue, like, the exact spot where the issue was detected 👀
• job — the faulty xe_sched_job, again, where the issue was detected 🔥
• fmt — a printf format + args, and you use it to describe the reason for the core dump. spill the tea fr fr 🍓
• ... — variable arguments 💫

---

👑 DESCRIPTION 👑

so here's the deal: this function should be called at the crash time, within the serialized gt_reset. that's not optional energy, that's the requirement 🚨

and like — it gets skipped if we still have the core dump device available with the information of the 'first' snapshot. so the first one wins, periodt 💅

---

🌈 SEE ALSO 🌈

• Kernel file ./drivers/gpu/drm/xe/xe_devcoredump.c
• Xe device coredump(9)
• xe_devcoredump_read(9)
• xe_print_blob_ascii85(9)

---

and that's the whole page bestie, go forth and snapshot ✨

shoutout to the source/version it lives under, ./drivers/gpu/drm/xe, dated October 2026 — that's the cute little sign-off, mwah 💖🧚