netfs_start_io_direct(9)

🌸 NAME 🌸

okay so bestie, welcome to the Kernel API Manual page for netfs_start_io_direct — yes, netfs_start_io_direct(9) if we’re being formal — and it’s literally the function that declares the file is being used for direct i/o. 💅

✨ SYNOPSIS ✨

int netfs_start_io_direct (struct inode *inode );

🎀 ARGUMENTS 🎀

• inode — file inode. that’s the whole argument, bestie. 🦋

💖 DESCRIPTION 💖

okay so netfs_start_io_direct declares that a direct I/O operation is about to start, and it makes sure we block all buffered I/O. no cap, that’s the vibe. 🛑

On exit, the function makes sure the NETFS_ICTX_ODIRECT flag is set, and it holds a shared lock on inode->i_rwsem to ensure that the flag cannot be changed. periodt. 🔒

In practice, this means direct I/O operations are allowed to execute in parallel — thank you, shared lock — whereas buffered I/O operations need to wait to grab an exclusive lock in order to clear NETFS_ICTX_ODIRECT. the way that locking situation works is honestly iconic. ✨

also pls note: buffered writes and truncates both take a write lock on inode->i_rwsem, meaning that those are serialised w.r.t. O_DIRECT. so yeah, they are not out here running wild together. 💀

🌈 SEE ALSO 🌈

Kernel file ./fs/netfs/locking.c — plus these besties, iykyk:

• netfs_start_io_read(9)
• netfs_end_io_read(9)
• netfs_start_io_write(9)
• netfs_end_io_write(9)
• netfs_end_io_direct(9)
• netfs_wb_begin(9)

cute little sign-off: this page comes from ./fs/netfs, it’s dated October 2026, and it’s all about netfs_start_io_direct(9). mwah 💋✨