amdgpu_gem_...andles_ioctl(9)

💅 amdgpu_gem_list_handles_ioctl — Kernel API Manual ✨

okay bestie, so this is a kernel API page, and the star of the show is amdgpu_gem_list_handles_ioctl — it's literally the function that gets you the tea on a process' buffer objects. like, all of them. no cap. 🫶

---

🌸 NAME 🌸

amdgpu_gem_list_handles_ioctl — get information about a process' buffer objects. that's it, that's the whole vibe check 💖

---

🦋 SYNOPSIS 🦋

int amdgpu_gem_list_handles_ioctl (struct drm_device *dev , void *data ,
struct drm_file *filp );

like, look at her. she's an int, she takes three args, she's iconic 😭

---

🎀 ARGUMENTS 🎀

• dev — drm device pointer. that's the device bestie, the whole reason we're here 💫
• data — a drm_amdgpu_gem_list_handles. this is where the actual list situation lives, keep an eye on her 👀
• filp — drm file pointer. the file, the vibe, the context, y'know? 🌈

---

💜 DESCRIPTION 💜

so here's how it works, and honestly it's giving "two-way street" energy:

num_entries gets set as an input to the size of the entries array. that's you telling the kernel how much room you've got 🎀

then num_entries comes back as output as the number of bos in the process. so the kernel's like "okay here's the real count, bestie" ✨

and here's the important part, don't skip this: if that number is larger than the size of the array, the ioctl must be retried. literally, no shortcuts. call it again with a bigger array or you're not getting the whole picture 💅

---

👑 RETURN 👑

• 0 for success 🎉
• -errno for errors 😭

that's the whole return situation. simple, clean, slay.

---

⭐ SEE ALSO ⭐

• Kernel file ./drivers/gpu/drm/amd/amdgpu/amdgpu_gem.c
• amdgpu_gem_timeout(9)
• amdgpu_gem_va_update_vm(9)

go read these, they're the friend group 🫶

---

💫 sign-off 💫

this page lives at ./drivers/gpu/drm/amd/amdgpu (source/version), dated October 2026 — so yeah, she's got a timestamp and she's proud of it. anyway bestie, that's the whole manpage, go forth and list those handles 💖✨🎀🦋🌸💅