From mboxrd@z Thu Jan 1 00:00:00 1970 From: Tvrtko Ursulin Subject: Re: [PATCH] drm/i915: Use hash tables for the command parser Date: Thu, 08 May 2014 14:42:16 +0100 Message-ID: <536B89B8.9070303@linux.intel.com> References: <1398698528-25256-1-git-send-email-bradley.d.volkin@intel.com> Mime-Version: 1.0 Content-Type: text/plain; charset="us-ascii"; Format="flowed" Content-Transfer-Encoding: 7bit Return-path: Received: from mga02.intel.com (mga02.intel.com [134.134.136.20]) by gabe.freedesktop.org (Postfix) with ESMTP id 134916EE6A for ; Thu, 8 May 2014 06:42:28 -0700 (PDT) In-Reply-To: <1398698528-25256-1-git-send-email-bradley.d.volkin@intel.com> List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: intel-gfx-bounces@lists.freedesktop.org Sender: "Intel-gfx" To: bradley.d.volkin@intel.com, intel-gfx@lists.freedesktop.org List-Id: intel-gfx@lists.freedesktop.org On 04/28/2014 04:22 PM, bradley.d.volkin@intel.com wrote: > From: Brad Volkin > > For clients that submit large batch buffers the command parser has > a substantial impact on performance. On my HSW ULT system performance > drops as much as ~20% on some tests. Most of the time is spent in the > command lookup code. Converting that from the current naive search to > a hash table lookup reduces the performance impact by as much as ~10%. I have another question - what about register lookups, do they come up in profiling as significant? Regards, Tvrtko