From owner-freebsd-bugs@FreeBSD.ORG Fri Jan 24 21:41:48 2014 Return-Path: Delivered-To: freebsd-bugs@smarthost.ysv.freebsd.org Received: from mx1.freebsd.org (mx1.freebsd.org [IPv6:2001:1900:2254:206a::19:1]) (using TLSv1 with cipher ADH-AES256-SHA (256/256 bits)) (No client certificate requested) by hub.freebsd.org (Postfix) with ESMTPS id 47BCCA2B; Fri, 24 Jan 2014 21:41:48 +0000 (UTC) Received: from freefall.freebsd.org (freefall.freebsd.org [IPv6:2001:1900:2254:206c::16:87]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by mx1.freebsd.org (Postfix) with ESMTPS id 32D421B2F; Fri, 24 Jan 2014 21:41:48 +0000 (UTC) Received: from freefall.freebsd.org (localhost [127.0.0.1]) by freefall.freebsd.org (8.14.7/8.14.7) with ESMTP id s0OLfmaU077529; Fri, 24 Jan 2014 21:41:48 GMT (envelope-from jmg@freefall.freebsd.org) Received: (from jmg@localhost) by freefall.freebsd.org (8.14.7/8.14.7/Submit) id s0OLfljp077528; Fri, 24 Jan 2014 21:41:47 GMT (envelope-from jmg) Date: Fri, 24 Jan 2014 21:41:47 GMT Message-Id: <201401242141.s0OLfljp077528@freefall.freebsd.org> To: chargen@gmail.com, jmg@FreeBSD.org, freebsd-bugs@FreeBSD.org From: jmg@FreeBSD.org Subject: Re: kern/120270: [crypto] [patch] AES-192 and AES-256 support for HW-accellerators ie. Hifn 795x / and OpenSSL engine 0.9.8e to _default_ use Cryptodev X-BeenThere: freebsd-bugs@freebsd.org X-Mailman-Version: 2.1.17 Precedence: list List-Id: Bug reports List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , X-List-Received-Date: Fri, 24 Jan 2014 21:41:48 -0000 Synopsis: [crypto] [patch] AES-192 and AES-256 support for HW-accellerators ie. Hifn 795x / and OpenSSL engine 0.9.8e to _default_ use Cryptodev State-Changed-From-To: open->feedback State-Changed-By: jmg State-Changed-When: Fri Jan 24 21:35:22 UTC 2014 State-Changed-Why: Is this patch still applicable? I recently tried aes modes w/ OpenSSL (we've updated to 1.0.1 in 10-R and HEAD) and it will default to using cryptodev over the internal AES-NI implementation if cryptodev module is loaded... Also, you cannot just use openssl speed aes to test, as it will not use the normal path.. you have to use the -evp flag to force it to go through the normal device selection path... When using the -evp flag you should see very short CPU times, like: Doing aes-128-cbc for 3s on 16 size blocks: 910409 aes-128-cbc's in 0.26s instead of 3.00s as that is a measure of CPU time not real time, and w/ an accelerator, the accelerator is doing the work, not the local process... http://www.freebsd.org/cgi/query-pr.cgi?pr=120270