linux-ext4 - ext4: Question about directory entry minor hash usage (documentation error?)

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [thread-next>] [day] [month] [year] [list]

Message-ID: <51a89ced-228d-4fd0-9613-3b4d027d9162@mailbox.org>
Date: Wed, 24 Sep 2025 18:04:08 +0200
From: Zeno Endemann <zeno.endemann@...lbox.org>
To: linux-ext4@...r.kernel.org
Subject: ext4: Question about directory entry minor hash usage (documentation
 error?)

Hello,

The documentation of hash tree directories claims that interior nodes are
"indexed by a minor hash".

However from my current understanding of the code, it seems to me the node
splitting works somewhat like regular B-trees, and there is no re-sorting
with a minor hash going on. The minor hash doesn't influence at all the
on-disk data structure, and is only used for sorting in a kernel internal
rb-tree. Is this correct? If so, I could offer to write up a patch for the
documentation.

As a side question, I was wondering a bit why the kernel differentiates
between htree-indexed dirs and others when simply iterating over it (as in
e.g. ext4_readdir), and what the point of that rb-tree there is, i.e. why one
would want to iterate over the entries in hash tree order.

Thanks and cheers,
Zeno Endemann