ceph_start_io_direct(9)

✨ NAME ✨

ceph_start_io_direct — declare the file is being used for direct i/o 💅

Okay bestie, welcome to the Kernel API Manual page for ceph_start_io_direct(9) — manual name folded in right here, cute and compliant 🦋. This function is lowkey giving kernel-internals main character energy.

---

💅 SYNOPSIS 💅

int ceph_start_io_direct (struct inode *inode );

---

🎀 ARGUMENTS 🎀

• inode — file inode. That’s the whole argument list, no extra crumbs 🍓.

---

🔥 DESCRIPTION 🔥

So ceph_start_io_direct is the function you call to declare that a direct I/O operation is about to start, and it makes sure we block all buffered I/O. Periodt.

On exit, the function ensures that the CEPH_I_ODIRECT flag is set, and holds a shared lock on inode->i_rwsem to ensure that the flag cannot be changed. No cap.

In practice, this means direct I/O operations are allowed to execute in parallel, thanks to the shared lock, whereas buffered I/O operations need to wait to grab an exclusive lock in order to clear CEPH_I_ODIRECT. The way that works is honestly kind of iconic.

Also note: buffered writes and truncates both take a write lock on inode->i_rwsem, meaning those are serialised w.r.t. O_DIRECT. 💀

---

🌈 SEE ALSO 🌈

Kernel file ./fs/ceph/io.c

• ceph_start_io_read(9)
• ceph_end_io_read(9)
• ceph_start_io_write(9)
• ceph_end_io_write(9)
• ceph_end_io_direct(9)

---

Catch you in the source tree, bestie 🫶 — this page is from the ./fs/ceph source/version, dated October 2026, and the manual tag is ceph_start_io_direct(9) ✨💖