Date: Thu, 11 Sep 2003 11:24:30 +0530
From: Ravikiran G Thirumalai <kiran@in.ibm.com>
To: Robert Love <rml@tech9.net>
Cc: linux-kernel@vger.kernel.org
Subject: Re: How reliable is SLAB_HWCACHE_ALIGN?
Message-ID: <20030911055428.GA1140@llm08.in.ibm.com>
References: <20030910081654.GA1129@llm08.in.ibm.com> <1063208464.700.35.camel@localhost>
Mime-Version: 1.0
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
In-Reply-To: <1063208464.700.35.camel@localhost>
User-Agent: Mutt/1.4i
Sender: linux-kernel-owner@vger.kernel.org
Content-Length: 835
Lines: 22

On Wed, Sep 10, 2003 at 11:41:04AM -0400, Robert Love wrote:
> On Wed, 2003-09-10 at 04:16, Ravikiran G Thirumalai wrote:
> 
> > Am I missing something or can there really be two objects on the same 
> > cacheline even when SLAB_HWCACHE_ALIGN is specified?
> 
> No, you are right.
> 
> If your object _must_ be cache aligned, use SLAB_MUST_HWCACHE_ALIGN.
> 

WOW!!!
Looking at the code though, SLAB_MUST_HWCACHE_ALIGN is never considered
by kmem_cache_create or kmem_cache_alloc. So, right now, there is no way of 
getting objects which are _guaranteed_ to be cacheline aligned!!!(?)

Kiran
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/