dax_set_ops(9)

💅 dax_set_ops — like, the kernel API manual, bestie

Okay so real talk, this is a page from the Kernel API Manual and it's all about dax_set_ops(9). Buckle up, it's short but iconic 🎀

---

✨ NAME ✨

dax_set_ops — set the dax_operations for a dax_device, no cap 🫶

---

🌸 SYNOPSIS 🌸

int dax_set_ops (struct dax_device *dax_dev , const struct
       dax_operations *ops );

---

🦋 ARGUMENTS 🦋

• dax_dev — the dax_device you're configuring, like it's getting a glow-up ✨
• ops — the operations to set. and yes, it may be NULL to clear, so you can absolutely just. undo it 💅

---

💖 DESCRIPTION 💖

So here's the vibe: this lets drivers set the dax_operations after the dax_device has already been allocated. Which is kind of a big deal, because sometimes the device gets created before the driver that needs specific ops is even bound — fsdev_dax binding to a dev_dax created by hmem is the example given, iykyk 😌

Now the important part, pay attention bestie:

• When you're setting non-NULL ops, it fails if ops are already set — returns -EBUSY. No double-dipping. 🚫
• When you're clearing ops (NULL), it always succeeds. Slay, zero drama 💀✨

Fr fr, don't try to overwrite existing ops — the kernel will hit you with -EBUSY and honestly? It's fair.

---

🔥 RETURN 🔥

• 0 on success 🎉
• -EBUSY if ops already set 😭

---

🌈 SEE ALSO 🌈

• Kernel file ./drivers/dax/super.c
• struct dax_device(9)
• fs_dax_get_by_bdev(9)
• fs_put_dax(9)
• fs_dax_get(9)
• dax_direct_access(9)
• dax_holder(9)
• inode_dax(9)

Go read them, they're all main character energy too 📚👑

---

And that's the whole page, bestie — brought to you by the fine folks in ./drivers/dax, dated October 2026 💜 Mark your calendars, this manual page ate and left no crumbs 🍓✨