Re: [PATCH v2] mm: per-thread vma caching

From: Peter Zijlstra <peterz@infradead.org>
To: Davidlohr Bueso <davidlohr@hp.com>
Cc: Andrew Morton <akpm@linux-foundation.org>,
	Ingo Molnar <mingo@kernel.org>,
	Linus Torvalds <torvalds@linux-foundation.org>,
	Michel Lespinasse <walken@google.com>,
	Mel Gorman <mgorman@suse.de>, Rik van Riel <riel@redhat.com>,
	KOSAKI Motohiro <kosaki.motohiro@gmail.com>,
	aswin@hp.com, scott.norton@hp.com, linux-mm@kvack.org,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH v2] mm: per-thread vma caching
Date: Tue, 25 Feb 2014 20:46:49 +0100	[thread overview]
Message-ID: <20140225194649.GV6835@laptop.programming.kicks-ass.net> (raw)
In-Reply-To: <1393353454.2577.42.camel@buesod1.americas.hpqcorp.net>

On Tue, Feb 25, 2014 at 10:37:34AM -0800, Davidlohr Bueso wrote:
> On Tue, 2014-02-25 at 19:35 +0100, Peter Zijlstra wrote:
> > On Tue, Feb 25, 2014 at 10:16:46AM -0800, Davidlohr Bueso wrote:
> > > +void vmacache_update(struct mm_struct *mm, unsigned long addr,
> > > +		     struct vm_area_struct *newvma)
> > > +{
> > > +	/*
> > > +	 * Hash based on the page number. Provides a good
> > > +	 * hit rate for workloads with good locality and
> > > +	 * those with random accesses as well.
> > > +	 */
> > > +	int idx = (addr >> PAGE_SHIFT) & 3;
> > 
> >  % VMACACHE_SIZE
> > 
> > perhaps? GCC should turn that into a mask for all sensible values I
> > would think.
> > 
> > Barring that I think something like:
> > 
> > #define VMACACHE_BITS	2
> > #define VMACACHE_SIZE	(1U << VMACACHE_BITS)
> > #define VMACACHE_MASK	(VMACACHE_SIZE - 1)
> 
> Hmm all that seems like an overkill.

If GCC does the right thing with % VMACACHE_SIZE it gets rid of an ugly
constant. But the 3 VMACACHE_* things are 'better' in that its
impossible to set VMACACHE_SIZE to silly values.