Date: Sat, 2 Apr 2016 18:26:32 +0200
From: Peter Zijlstra <peterz@infradead.org>
To: Thomas Gleixner <tglx@linutronix.de>
Cc: LKML <linux-kernel@vger.kernel.org>,
        Sebastian Andrzej Siewior <bigeasy@linutronix.de>,
        Darren Hart <darren@dvhart.com>, Ingo Molnar <mingo@kernel.org>,
        Michael Kerrisk <mtk.manpages@googlemail.com>,
        Davidlohr Bueso <dave@stgolabs.net>, Chris Mason <clm@fb.com>,
        "Carlos O'Donell" <carlos@redhat.com>,
        Torvald Riegel <triegel@redhat.com>,
        Eric Dumazet <edumazet@google.com>
Subject: Re: [RFC patch 4/7] futex: Add support for attached futexes
Message-ID: <20160402162632.GQ3448@twins.programming.kicks-ass.net>
References: <20160402095108.894519835@linutronix.de>
 <20160402110035.753145539@linutronix.de>
MIME-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <20160402110035.753145539@linutronix.de>
User-Agent: Mutt/1.5.21 (2012-12-30)
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 1064
Lines: 34

On Sat, Apr 02, 2016 at 11:09:18AM -0000, Thomas Gleixner wrote:
> +/**
> + * hash_local_futex - Return the hash bucket in the task local cache
> + * @uaddr:	The user space address of the futex
> + * @prime:	The prime number for the modulo operation
> + *
> + * That's a primitive hash function, but it turned out to be the most
> + * efficient one for the task local cache as we don't have anything to
> + * mix in like we have for the global hash.
> + */
> +static inline unsigned int
> +hash_local_futex(void __user *uaddr, unsigned int prime)
> +{
> +	return ((unsigned long) uaddr) % prime;
> +}

> +static unsigned int hash_prime(unsigned int size)
> +{
> +	switch(size) {
> +	case   16:
> +	default:   return 13;
> +	case   32: return 31;
> +	case   64: return 61;
> +	case  128: return 127;
> +	case  256: return 251;
> +	case  512: return 509;
> +	case 1024: return 1021;
> +	case 2048: return 2039;
> +	case 4096: return 4093;
> +	}
> +}

One wonders what's wrong with include/linux/hash.h ? It appears to me
hash_ptr() is pretty much what you want, no?