From mboxrd@z Thu Jan 1 00:00:00 1970 From: Andi Kleen Date: Mon, 18 Oct 2004 21:06:59 +0000 Subject: Re: [PATCH] General purpose zeroed page slab Message-Id: <20041018210658.GA8203@wotan.suse.de> List-Id: References: <20041014180427.GA7973@wotan.suse.de> <20041018184210.GI16153@parcelfarce.linux.theplanet.co.uk> In-Reply-To: <20041018184210.GI16153@parcelfarce.linux.theplanet.co.uk> MIME-Version: 1.0 Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: 7bit To: Matthew Wilcox Cc: "Martin K. Petersen" , Andi Kleen , linux-kernel@vger.kernel.org, linux-ia64@vger.kernel.org, akpm@osdl.org, tony.luck@intel.com > It's probably worth doing this with a static cachep in slab.c and only > exposing a get_zeroed_page() / free_zeroed_page() interface, with the > latter doing the memset to 0. Putting a memset in there would be dumb because the mm cleanup already zeroes the page tables. My dirty bitmap proposal would make that faster however, same as copy_page_range et.al. > I disagree with Andi over the dumbness > of zeroing the whole page. That makes it cache-hot, which is what you > want from a page you allocate from slab. It's already cache hot from the page table free and you only want one cache line in it cache hot, not the whole page. -Andi