linux-kernel - Re: [PATCH v8 00/15] futex: Add support task local hash maps.

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

Message-ID: <20250205084658.iUmkP3_Z@linutronix.de>
Date: Wed, 5 Feb 2025 09:46:58 +0100
From: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To: Peter Zijlstra <peterz@...radead.org>
Cc: linux-kernel@...r.kernel.org,
	André Almeida <andrealmeid@...lia.com>,
	Darren Hart <dvhart@...radead.org>,
	Davidlohr Bueso <dave@...olabs.net>, Ingo Molnar <mingo@...hat.com>,
	Juri Lelli <juri.lelli@...hat.com>,
	Thomas Gleixner <tglx@...utronix.de>,
	Valentin Schneider <vschneid@...hat.com>,
	Waiman Long <longman@...hat.com>
Subject: Re: [PATCH v8 00/15] futex: Add support task local hash maps.

On 2025-02-04 16:14:05 [+0100], Peter Zijlstra wrote:
> On Mon, Feb 03, 2025 at 02:59:20PM +0100, Sebastian Andrzej Siewior wrote:
> >   futex: Decrease the waiter count before the unlock operation.
> >   futex: Prepare for reference counting of the process private hash end
> >     of operation.
> >   futex: Re-evaluate the hash bucket after dropping the lock
> >   futex: Introduce futex_get_locked_hb().
> 
> I must admit to not being a fan of these patches. I find them very hard
> to review.
> 
> In part because while you go stick _put() on various things, there is no
> corresponding _get(), things like futex_hash() become an implicit get
> -- this took a while to figure out.

There is one get in futex_hash() and multiple puts depending on the
context usually after dropping the hash bucket lock.

> I tried tying the whole get/put to the hb variable itself, using the
> cleanup stuff. The patch is pretty massive because fairly large chunks
> of code got re-indented, but perhaps the final result is clearer, not
> sure.

… Let me find some time and dig through this.

Sebastian