Return-Path: Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1751497AbdC0J1n (ORCPT ); Mon, 27 Mar 2017 05:27:43 -0400 Received: from mga01.intel.com ([192.55.52.88]:46957 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1751407AbdC0J1g (ORCPT ); Mon, 27 Mar 2017 05:27:36 -0400 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.36,231,1486454400"; d="scan'208";a="65434158" Message-ID: <1490606744.3661.7.camel@linux.intel.com> Subject: Re: [PATCH] x86/boot: Support uncompressed kernel From: Chao Peng To: Sebastian Andrzej Siewior Cc: Sergey Senozhatsky , Yinghai Lu , "H. Peter Anvin" , Thomas Gleixner , Ingo Molnar , the arch/x86 maintainers , Michal Marek , Kees Cook , Baoquan He , "H.J. Lu" , Paul Bolle , Masahiro Yamada , Borislav Petkov , Andrew Morton , Arnd Bergmann , Petr Mladek , "David S. Miller" , "Paul E. McKenney" , Andy Lutomirski , Thomas Garnier , Nicolas Pitre , Tejun Heo , Daniel Mack , Helge Deller , Rik van Riel , Linux Kernel Mailing List , linux-kbuild Date: Mon, 27 Mar 2017 17:25:44 +0800 In-Reply-To: <20170327075817.bb7wda62fcuvfa4b@linutronix.de> References: <1490273467-97948-1-git-send-email-chao.p.peng@linux.intel.com> <20170323153117.GA6356@tigerII.localdomain> <1490333740.2925.7.camel@linux.intel.com> <20170327075817.bb7wda62fcuvfa4b@linutronix.de> Content-Type: text/plain; charset="UTF-8" X-Mailer: Evolution 3.20.5 (3.20.5-1.fc24) Mime-Version: 1.0 Content-Transfer-Encoding: 8bit Sender: linux-kernel-owner@vger.kernel.org List-ID: X-Mailing-List: linux-kernel@vger.kernel.org Content-Length: 1010 Lines: 31 On Mon, 2017-03-27 at 09:58 +0200, Sebastian Andrzej Siewior wrote: > On 2017-03-24 13:35:40 [+0800], Chao Peng wrote: > > > > > > > > > > > > > > > > > > > > >     kernel               kernel size    time in > > > > > decompress_kernel > > > > >     compressed (gzip)    3.3M           53ms > > > > >     uncompressed         14M            3ms > > > > > > Exactly, LZ4 is the fastest. It takes 16ms to complete the > > decompression. Still sounds a little longer when compared to > > uncompressed kernel. > > Are we seriously talking here about one-time improvement of 13ms > boot time? The usage model for us is to lunch kernel in virtual machine and there will be thousands of instances lunched and shutdowned/re-lunched frequently, so every single million-second helps. And 13ms means 20% improvement to our existing optimization (the other part besides decompression is optimized to ~40ms). Chao > > > > > Chao > > Sebastian