linux-kernel - Re: [PATCH v5 09/14] futex: Allow to re-allocate the private hash bucket.

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: Windows password security audit tool. GUI, reports in PDF.

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

Message-ID: <20241217145814.euTL-GEf@linutronix.de>
Date: Tue, 17 Dec 2024 15:58:14 +0100
From: Sebastian Andrzej Siewior <bigeasy@...utronix.de>
To: linux-kernel@...r.kernel.org
Cc: 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>,
	Peter Zijlstra <peterz@...radead.org>,
	Thomas Gleixner <tglx@...utronix.de>,
	Valentin Schneider <vschneid@...hat.com>,
	Waiman Long <longman@...hat.com>
Subject: Re: [PATCH v5 09/14] futex: Allow to re-allocate the private hash
 bucket.

On 2024-12-16 00:00:13 [+0100], To linux-kernel@...r.kernel.org wrote:
> @@ -154,6 +303,17 @@ struct futex_hash_bucket *futex_hash(union futex_key *key)
>  
>  void futex_hash_put(struct futex_hash_bucket *hb)
>  {
> +	struct futex_hash_bucket_private *hb_p;
> +
> +	if (hb->hb_slot == 0)
> +		return;
> +	hb_p = container_of(hb, struct futex_hash_bucket_private,
> +			    queues[hb->hb_slot - 1]);
> +
> +	if (!rcuref_put(&hb_p->users))
> +		return;
> +
> +	futex_assign_new_hb(NULL);

I have reworked two things:
- I have to remove futex_assign_new_hb() here and let the next
  rcuref_get() perform the assignment instead. The reason is that the
  caller can change its task state (futex_wait_queue()) and blocking on
  the mutex will reset it. Instead of delaying it after the schedule()
  it looks simpler to delay it to the next rcuref_get() where it has to
  be handled anyway.

- The logic in __futex_assign_new_hb() didn't deal properly with delayed
  assigns.

>  }

Sebastian