From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mail-oi0-f51.google.com (mail-oi0-f51.google.com [209.85.218.51]) by kanga.kvack.org (Postfix) with ESMTP id EBDA76B0070 for ; Tue, 5 May 2015 21:21:23 -0400 (EDT) Received: by oica37 with SMTP id a37so160995821oic.0 for ; Tue, 05 May 2015 18:21:23 -0700 (PDT) Received: from g4t3426.houston.hp.com (g4t3426.houston.hp.com. [15.201.208.54]) by mx.google.com with ESMTPS id z17si11129936oep.68.2015.05.05.18.21.23 for (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128/128); Tue, 05 May 2015 18:21:23 -0700 (PDT) Message-ID: <55496C8F.606@hp.com> Date: Tue, 05 May 2015 21:21:19 -0400 From: Waiman Long MIME-Version: 1.0 Subject: Re: [PATCH 0/13] Parallel struct page initialisation v4 References: <1430231830-7702-1-git-send-email-mgorman@suse.de> <554030D1.8080509@hp.com> <5543F802.9090504@hp.com> <554415B1.2050702@hp.com> <20150504143046.9404c572486caf71bdef0676@linux-foundation.org> <20150505104514.GC2462@suse.de> <20150505130255.49ff76bbf0a3b32d884ab2ce@linux-foundation.org> In-Reply-To: <20150505130255.49ff76bbf0a3b32d884ab2ce@linux-foundation.org> Content-Type: text/plain; charset=ISO-8859-1; format=flowed Content-Transfer-Encoding: 7bit Sender: owner-linux-mm@kvack.org List-ID: To: Andrew Morton Cc: Mel Gorman , Nathan Zimmer , Dave Hansen , Scott Norton , Daniel J Blueman , Linux-MM , LKML On 05/05/2015 04:02 PM, Andrew Morton wrote: > On Tue, 5 May 2015 11:45:14 +0100 Mel Gorman wrote: > >> On Mon, May 04, 2015 at 02:30:46PM -0700, Andrew Morton wrote: >>>> Before the patch, the boot time from elilo prompt to ssh login was 694s. >>>> After the patch, the boot up time was 346s, a saving of 348s (about 50%). >>> Having to guesstimate the amount of memory which is needed for a >>> successful boot will be painful. Any number we choose will be wrong >>> 99% of the time. >>> >>> If the kswapd threads have started, all we need to do is to wait: take >>> a little nap in the allocator's page==NULL slowpath. >>> >>> I'm not seeing any reason why we can't start kswapd much earlier - >>> right at the start of do_basic_setup()? >> It doesn't even have to be kswapd, it just should be a thread pinned to >> a done. The difficulty is that dealing with the system hashes means the >> initialisation has to happen before vfs_caches_init_early() when there is >> no scheduler. > I bet we can run vfs_caches_init_early() after sched_init(). Might > need a few little fixups. > >> Those allocations could be delayed further but then there is >> the possibility that the allocations would not be contiguous and they'd >> have to rely on CMA to make the attempt. That potentially alters the >> performance of the large system hashes at run time. > hm, why. If the kswapd threads are running and busily creating free > pages then alloc_pages(order=10) can detect this situation and stall > for a while, waiting for kswapd to create an order-10 page. > > Alternatively, the page allocator can go off and synchronously > initialize some pageframes itself. Keep doing that until the > allocation attempt succeeds. > > Such an approach is much more robust than trying to predict how much > memory will be needed. > Most of those hash tables are allocated before smp_boot. In UP mode, you can't have another thread initializing memory. So we really need to preallocate enough for those tables. Cheers, Longman -- To unsubscribe, send a message with 'unsubscribe linux-mm' in the body to majordomo@kvack.org. For more info on Linux MM, see: http://www.linux-mm.org/ . Don't email: email@kvack.org