__audit_inode_child(9)

πŸ“¦ version 7.3.1, in nixpkgs unstable

✨ __audit_inode_child β€” Kernel API Manual ✨

okay so bestie, you're about to meet the little audit helper that handles the children of filesystem objects πŸ’… like, the parent gets all the attention normally, but this one is for the created/removed objects themselves. main character energy fr fr πŸŽ€


🌸 NAME 🌸

__audit_inode_child β€” collect inode info for created/removed objects πŸ‘‘


πŸ¦‹ SYNOPSIS πŸ¦‹

void __audit_inode_child (struct inode *parent , const struct dentry *dentry , const unsigned char type );

πŸ’– ARGUMENTS πŸ’–

  • parent β€” the inode of the dentry's parent 🌈 like, the parent's inode, got it
  • dentry β€” the dentry being audited πŸŽ€ the actual thing we're looking at
  • type β€” an AUDIT_TYPE_* value that we're looking for ✨

πŸ”₯ DESCRIPTION πŸ”₯

For syscalls that create or remove filesystem objects, audit_inode can only collect info for the filesystem object's parent β€” lowkey sad, I know 😭 This call right here updates the audit context with the child's information, so the child gets its moment too πŸ’…βœ¨

Syscalls that create a new filesystem object must be hooked after the object is created 🫢 periodt, no cap.

Syscalls that remove a filesystem object must be hooked prior β€” like, before β€” so that the target inode can be captured during unsuccessful attempts πŸ‘€πŸ’€ yes bestie, even the failed attempts matter, don't skip that part.


🧚 SEE ALSO 🧚

Kernel file ./kernel/auditsc.c and literally the whole friend group:


πŸ’œ ngl bestie, this is a kernel-internal audit thing β€” if you're reading this page, you're already deep in the ./kernel directory, so slay, you're exactly where you need to be 🫢

and that's the whole vibe check 🌈 this page lives in ./kernel, dated October 2026, and it's called __audit_inode_child(9) β€” iconic, no notes πŸ’…βœ¨πŸŽ€

πŸ“¦ where this page comes from