From: Mark Lord Subject: Re: [PATCH 2/2] Add batched discard support for ext4. Date: Wed, 21 Apr 2010 17:59:07 -0400 Message-ID: <4BCF752B.6050502@teksavvy.com> References: <1271674527-2977-1-git-send-email-lczerner@redhat.com> <4BCE6243.5010209@teksavvy.com> <4BCE66C5.3060906@redhat.com> <4BCF4C53.3010608@redhat.com> <4BCF67A9.2040902@redhat.com> <4BCF6831.7080506@redhat.com> Mime-Version: 1.0 Content-Type: text/plain; charset=UTF-8; format=flowed Content-Transfer-Encoding: 7bit Cc: Ric Wheeler , sandeen@redhat.com, Eric Sandeen , Jeff Moyer , Lukas Czerner , linux-ext4@vger.kernel.org, Edward Shishkin , Christoph Hellwig , James Bottomley To: Greg Freemyer Return-path: Received: from ironport2-out.teksavvy.com ([206.248.154.181]:17160 "EHLO ironport2-out.pppoe.ca" rhost-flags-OK-OK-OK-FAIL) by vger.kernel.org with ESMTP id S1753323Ab0DUV7T (ORCPT ); Wed, 21 Apr 2010 17:59:19 -0400 In-Reply-To: Sender: linux-ext4-owner@vger.kernel.org List-ID: On 21/04/10 05:47 PM, Greg Freemyer wrote: > > The ATA-8 spec. supports vectored trims and requires a minimum of 255 > sectors worth of range payload be supported. That equates to a single > trim being able to trim thousands of ranges in one command. > > Mark Lord has benchmarked in found a vectored trim to be drastically > faster than calling trim individually for each of those ranges. .. Does anyone have an Intel-based SSD they could spare? :) Rumour has it they they do not comply with the ATA-8 spec for TRIM, in that they don't support more than one "sector" of range data. The SSDs I have here all support much more than that. Still, even if we just do a single 512-byte payload of TRIM ranges, it would be a huge win. Cheers