From owner-freebsd-stable Mon Feb 24 12:22:27 2003 Delivered-To: freebsd-stable@freebsd.org Received: from mx1.FreeBSD.org (mx1.freebsd.org [216.136.204.125]) by hub.freebsd.org (Postfix) with ESMTP id B6E5837B401 for ; Mon, 24 Feb 2003 12:22:24 -0800 (PST) Received: from ebb.errno.com (ebb.errno.com [66.127.85.87]) by mx1.FreeBSD.org (Postfix) with ESMTP id 0339243FB1 for ; Mon, 24 Feb 2003 12:22:24 -0800 (PST) (envelope-from sam@errno.com) Received: from melange (melange.errno.com [66.127.85.82]) (authenticated bits=0) by ebb.errno.com (8.12.5/8.12.1) with ESMTP id h1OKMNnN070631 (version=TLSv1/SSLv3 cipher=RC4-MD5 bits=128 verify=NO); Mon, 24 Feb 2003 12:22:23 -0800 (PST)?g (envelope-from sam@errno.com)œ X-Authentication-Warning: ebb.errno.com: Host melange.errno.com [66.127.85.82] claimed to be melange Message-ID: <1f4f01c2dc42$70a8d420$52557f42@errno.com> From: "Sam Leffler" To: , "Mike Tancsa" References: <5.2.0.9.0.20030224145408.05edb800@marble.sentex.ca> Subject: Re: HEADS UP: OpenSSL 0.9.7 in -STABLE Date: Mon, 24 Feb 2003 12:22:22 -0800 Organization: Errno Consulting MIME-Version: 1.0 Content-Type: text/plain; charset="iso-8859-1" Content-Transfer-Encoding: 7bit X-Priority: 3 X-MSMail-Priority: Normal X-Mailer: Microsoft Outlook Express 5.50.4807.1700 X-MimeOLE: Produced By Microsoft MimeOLE V5.50.4807.1700 Sender: owner-freebsd-stable@FreeBSD.ORG Precedence: bulk List-ID: List-Archive: (Web Archive) List-Help: (List Instructions) List-Subscribe: List-Unsubscribe: X-Loop: FreeBSD.ORG > Hi all, > Is anyone trying to use a hardware accelerator in STABLE ? > It's running on several of my production machines. > dmesg shows > > hifn0 mem 0xfa000000-0xfa000fff,0xfa800000-0xfa800fff irq 12 at device 15.0 > on pci0 > hifn0: Hifn 7951, rev 0, 128KB sram, 193 sessions > > Yet, when doing some ssh tests, I see approximately the same load average > when blasting across a stream of data via ssh (using 3des) with the card > and without. I would have thought there would be some noticeable > difference? I have a large backup server that processes numerous dumps > piped via ssh. I thought this would be a good test box to see how > effective the card might be. > > I have > # ls -l /dev/crypto > crw-rw-rw- 1 root operator 70, 0 Feb 24 11:16 /dev/crypto > > and I have > device crypto > device hifn > in the kernel. > > However, when I look through the NEWS file, there does not seem to be > specific mention of the HiFn ? Is there anything special I need to do in > order to build the lib to use the hardware accelerator ? Also the man page > for the HiFn states, > I don't know what the NEWS file. If you use gdb -k to look at /dev/mem then you can dump the cryptostats block to verify that operations are going through. Other ways to verify are the cryptotest program (which I forgot to MFC from current but you can find in /usr/src/tools/tools/crypto) and openssl: openssl speed -elapsed -evp des3 Oh and in the tools area is a little program to dump cryptostats w/o gdb. Note however that the 7951 is not fast and if you have a reasonably fast CPU you may not notice it's presence. > The Hifn 7951 and Hifn 7811 will also supply data to the kernel random(4) > subsystem. > > Do I need to tell rndcontrol about it ? > No. If the device is there then it automatically feeds pure entropy to the RNG. > > I do see the following additional kernel mibs > > # sysctl -x -A | grep -i cryp > kern.usercrypto: 1 > kern.userasymcrypto: 1 > kern.cryptodevallowsoft: 0 > kern.crypto_stats: Format:S,cryptostats Length:144 > Dump:0x000000000000000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000000000000000000000000000000000000000000000000 0000000000000000000000000000000000000000000000000000000000000000000 > cryptodesc: 64, 0, 0, 0, 0 > cryptop: 64, 0, 0, 0, 0 > debug.crypto_timing: 0 > > Also, on a similar note, should not the FAST_IPSEC be documented or at > least mentioned in LINT ? > Yes, on my TODO list (gotta get re approval now though). Sam To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-stable" in the body of the message