INJECT.F2FS(8)

💾 inject.f2fs — like, the manual page, but make it make sense 💅

okay bestie, so this is inject.f2fs, and it lives in section 8 of the System Manager's Manual (yes, the big scary admin one — don't be intimidated, you've got this 🫶).

---

✨ NAME ✨

inject.f2fs — inject a Linux F2FS file system

---

🌸 SYNOPSIS 🌸

inject.f2fs [ options ] device

---

🦋 DESCRIPTION 🦋

So inject.f2fs is the tool you use to modify metadata — or data, like a directory entry — inside an f2fs file system image, offline, and super flexibly. That's the whole vibe. It's giving surgeon energy, but for file systems. 🔥

---

🎀 OPTIONS 🎀

• -d debug level [default: 0] — sets how much debugging goodness you want. Specify the level of debugging options, that's it, that's the deal.
• -V — prints the version number and then dips. Bye! 👋
• --mb member name — specifies the member name inside a struct that's getting injected. Basically: which field are we poking at, bestie.
• --val new value — the new value to set, but only if mb is a number. Numbers only, no cap.
• --str new string — the new string to set, if mb happens to be a string. Different vibe entirely.
• --idx slot index — tells it which slot gets injected when mb is an array. Pick your fighter.
• --nid nid — specifies which nid is getting injected.
• --blk blkaddr — specifies which blkaddr is getting injected.
• --sb 0 or 1 or 2 — injects the super block. The argument tells it which sb pack gets injected, and 0 chooses the current valid sb automatically (iconic, honestly — it just knows). The available mb of sb are:
  • magic — magic number.
  • s_stop_reason — s_stop_reason array.
  • s_errors — s_errors array.
  • feature — features.
  • devs.path — path in devs array.
• --cp 0 or 1 or 2 — injects the checkpoint. The argument says which cp pack gets injected, and 0 chooses the current valid cp automatically. The available mb of cp are:
  • checkpoint_ver — checkpoint version.
  • ckpt_flags — checkpoint flags.
  • cur_node_segno — cur_node_segno array.
  • cur_node_blkoff — cur_node_blkoff array.
  • cur_data_segno — cur_data_segno array.
  • cur_data_blkoff — cur_data_blkoff array.
  • next_blkaddr — fsync dnodes.
  • alloc_type — alloc_type array.
  • crc — crc checksum.
  • elapsed_time — elapsed mount time.
• --nat 0 or 1 or 2 — injects the nat entry specified by nid. The argument says which nat pack gets injected, and 0 chooses the current valid nat automatically. The available mb of nat are:
  • version — nat entry version.
  • ino — nat entry ino.
  • block_addr — nat entry block_addr.
• --sit 0 or 1 or 2 — injects the sit entry specified by blk. The argument says which sit pack gets injected, and 0 chooses the current valid sit automatically. The available mb of sit are:
  • vblocks — sit entry vblocks.
  • valid_map — sit entry valid_map.
  • mtime — sit entry mtime.
• --ssa — injects the summary block or summary entry specified by blk. The available mb of ssa are:
  • entry_type — summary block footer entry_type.
  • check_sum — summary block footer check_sum.
  • nid — summary entry nid.
  • version — summary entry version.
  • ofs_in_node — summary entry ofs_in_node.
• --node — injects the node block specified by nid. The available mb of node are:
  • nid — node footer nid.
  • ino — node footer ino.
  • flag — node footer flag.
  • cp_ver — node footer cp_ver.
  • next_blkaddr — node footer next_blkaddr.
  • i_mode — inode i_mode.
  • i_advise — inode i_advise.
  • i_inline — inode i_inline.
  • i_links — inode i_links.
  • i_size — inode i_size.
  • i_blocks — inode i_blocks.
  • i_xattr_nid — inode i_xattr_nid.
  • i_ext.fofs — inode i_ext.fofs.
  • i_ext.blk_addr — inode i_ext.blk_addr.
  • i_ext.len — inode i_ext.len.
  • i_extra_isize — inode i_extra_isize.
  • i_inline_xattr_size — inode i_inline_xattr_size.
  • i_inode_checksum — inode i_inode_checksum.
  • i_compr_blocks — inode i_compr_blocks.
  • i_addr — inode i_addr array specified by idx.
  • i_nid — inode i_nid array specified by idx.
  • addr — {in}direct node nid/addr array specified by idx.
• --dent — injects the dentry block or dir entry specified by nid.
  • --dots 1 or 2 — this one means the "." or ".." directory entry of nid is going to be injected. Cute, right? 🎀

The available mb of dent are:

• d_bitmap — dentry block d_bitmap.
• d_hash — dentry hash.
• d_ino — dentry ino.
• d_ftype — dentry ftype.
• filename — dentry filename, and the corresponding d_hash and namelen are updated implicitly. Sneaky but helpful. ✨

• --dry-run — do not really inject. Look but don't touch, bestie. 🙈

⚠️ heads up: this is an offline image-editing tool, and it is genuinely touching raw metadata — so --dry-run exists for a reason. Use it. 👑

---

👑 AUTHOR 👑

This version of inject.f2fs has been written by Sheng Yong <shengyong@oppo.com>. Shoutout to the author, fr fr. 🫶

---

🌈 AVAILABILITY 🌈

inject.f2fs is available from git://git.kernel.org/pub/scm/linux/kernel/git/jaegeuk/f2fs-tools.git. Go get it, it's right there. 💜

---

💖 SEE ALSO 💖

mkfs.f2fs(8), fsck.f2fs(8), dump.f2fs(8), defrag.f2fs(8), resize.f2fs(8), sload.f2fs(8), defrag.f2fs(8).

(yes, defrag.f2fs(8) is listed twice — that's the original's energy, not mine, I'm just the messenger 😭)

---

okay that's the whole page, no notes. 💅 this one's from f2fs-tools, printed under System Manager's Manual — mwah, slay, go inject something responsibly. 🦋✨🎀