Date: Sun, 15 Aug 2004 17:04:45 +0900 (JST) From: Norikatsu Shigemura <nork@FreeBSD.org> To: Sam Leffler <sam@errno.com> Cc: freebsd-stable@FreeBSD.org Subject: Re: OpenSSL with hifn(4) (cryptodev) Message-ID: <200408150804.i7F84jCB010066@sakura.ninth-nine.com> In-Reply-To: <200408030902.59349.sam@errno.com> References: <20040803214133.31dd9b03.nork@FreeBSD.org> <200408030902.59349.sam@errno.com>
index | next in thread | previous in thread | raw e-mail
[-- Attachment #1 --] sam, Mike Tancsa, Doug White! Thank you for hints! On Tue, 3 Aug 2004 09:02:59 -0700 Sam Leffler <sam@errno.com> wrote: > On Tuesday 03 August 2004 05:41 am, Norikatsu Shigemura wrote: > > Hi sam! > > I have two Soekris vpn1401 crypto accelerator cards. I installed > > these to 4-stable machine and 5-current machine. > ... > > I confirmed `openssl speed -engine cryptodev', but it looks not > > works. Because 1st: same speed (before/after install it), 2nd: CPU > > loadavg is always high. So I consider that openssl didn't use > > cryptodev. Do you have any idea? > Look in /usr/src/tools/tools/crypto for the cryptostats and hifnstats > programs; they will tell you if the h/w is operating correctly. I and my friends, Naoki Fukaumi, inverstigated about this behavior. As the result, we confirmed that h/w accerator is good works but some limited. 1. `openssl speed' is not so good:-(. openssl speed -evp aes128(/des/3des) is good. I saw *Giant and crydev in top(1). 2. /usr/src/tools/tools/crypto/cryptotest.c clarified the problem. According to cryptotest(I tested ./cyrptotest -z 1000), hifn(4) (=vpn1401) supports des_cbc, 3des_cbc, aes_cbc, aes192_cbc, aes256_cbc, md5_hmac and sha1_hmac. (Of course, I saw ones in top(1)) 3. I read /usr/src/crypto/openssl/crypto/engine/hw_cryptodev.c. Accoring to it, cryptodev engine supports des_cbc, 3des_cbc, aes_cbc, blf_cbc, cast5_cbc, skipjack_cbc(?), sha1_hmac, ripemd160_hmac, md5_kpdk(?), sha1_kpdk(?), md5 and sha1(?). However, we can use these cifers by cryptodev_usable_ciphers, but cannot use these digests by cryptodev_usable_digests, in hw_cryptodev.c. According to comments: * XXXX just disable all digests for now, because it sucks. * we need a better way to decide this - i.e. I may not * want digests on slow cards like hifn on fast machines, * but might want them on slow or loaded machines, etc. * will also want them when using crypto cards that don't * suck moose gonads - would be nice to be able to decide something * as reasonable default without having hackery that's card dependent. * of course, the default should probably be just do everything, * with perhaps a sysctl to turn algoritms off (or have them off * by default) on cards that generally suck like the hifn. Hum..... By union set, so we can use only des_cbc, 3des_cbc and aes_cbc. [SEE ALSO] BenchMark1: openssl speed -elapsed -evp aes128 openssl speed -elapsed -evp des3 >>In 5-current with WITNESS aes-128-cbc 33.20k 211.17k 1184.66k 2574.12k 4918.85k des-ede3-cbc 70.37k 315.16k 901.09k 1643.15k 6840.56k >>In 5-current w/o WITNESS aes-128-cbc 324.79k 1264.01k 4650.77k 13378.57k 22098.25k des-ede3-cbc 324.65k 1278.52k 4645.58k 13392.40k 22017.54k >>In 4-stable aes-128-cbc 462.81k 1795.23k 6329.75k 16686.62k 29833.64k des-ede3-cbc 463.48k 1757.60k 1889.31k 16679.92k 29766.37k >>In 5-current w/o WITNESS w/o hifn(4) (PentiumIII-M 1.0GHz x1) aes-128-cbc 17732.99k 19308.65k 23740.17k 25805.46k 25179.36k des-ede3-cbc 7347.27k 5895.96k 7762.44k 7755.75k 7824.37k And also, I attached results of `./cryptotest -z 1000'. [-- Attachment #2 --] 0.125 sec, 2000 des crypts, 8 bytes, 127895 byte/sec, 1.0 Mb/sec 0.120 sec, 2000 des crypts, 16 bytes, 265668 byte/sec, 2.0 Mb/sec 0.123 sec, 2000 des crypts, 32 bytes, 520304 byte/sec, 4.0 Mb/sec 0.126 sec, 2000 des crypts, 64 bytes, 1017658 byte/sec, 7.8 Mb/sec 0.128 sec, 2000 des crypts, 128 bytes, 1999844 byte/sec, 15.3 Mb/sec 0.136 sec, 2000 des crypts, 256 bytes, 3752180 byte/sec, 28.6 Mb/sec 0.151 sec, 2000 des crypts, 512 bytes, 6760906 byte/sec, 51.6 Mb/sec 0.185 sec, 2000 des crypts, 1024 bytes, 11095340 byte/sec, 84.7 Mb/sec 0.260 sec, 2000 des crypts, 2048 bytes, 15726748 byte/sec, 120.0 Mb/sec 0.398 sec, 2000 des crypts, 4096 bytes, 20564262 byte/sec, 156.9 Mb/sec 0.713 sec, 2000 des crypts, 8192 bytes, 22993958 byte/sec, 175.4 Mb/sec 0.120 sec, 2000 3des crypts, 8 bytes, 133036 byte/sec, 1.0 Mb/sec 0.121 sec, 2000 3des crypts, 16 bytes, 264362 byte/sec, 2.0 Mb/sec 0.122 sec, 2000 3des crypts, 32 bytes, 525952 byte/sec, 4.0 Mb/sec 0.125 sec, 2000 3des crypts, 64 bytes, 1024754 byte/sec, 7.8 Mb/sec 0.127 sec, 2000 3des crypts, 128 bytes, 2011567 byte/sec, 15.3 Mb/sec 0.139 sec, 2000 3des crypts, 256 bytes, 3673332 byte/sec, 28.0 Mb/sec 0.151 sec, 2000 3des crypts, 512 bytes, 6783793 byte/sec, 51.8 Mb/sec 0.183 sec, 2000 3des crypts, 1024 bytes, 11201540 byte/sec, 85.5 Mb/sec 0.255 sec, 2000 3des crypts, 2048 bytes, 16079392 byte/sec, 122.7 Mb/sec 0.398 sec, 2000 3des crypts, 4096 bytes, 20590520 byte/sec, 157.1 Mb/sec 0.712 sec, 2000 3des crypts, 8192 bytes, 23011559 byte/sec, 175.6 Mb/sec 0.121 sec, 2000 aes crypts, 16 bytes, 265131 byte/sec, 2.0 Mb/sec 0.121 sec, 2000 aes crypts, 32 bytes, 529683 byte/sec, 4.0 Mb/sec 0.128 sec, 2000 aes crypts, 64 bytes, 1003402 byte/sec, 7.7 Mb/sec 0.126 sec, 2000 aes crypts, 128 bytes, 2024660 byte/sec, 15.4 Mb/sec 0.135 sec, 2000 aes crypts, 256 bytes, 3786534 byte/sec, 28.9 Mb/sec 0.150 sec, 2000 aes crypts, 512 bytes, 6821119 byte/sec, 52.0 Mb/sec 0.184 sec, 2000 aes crypts, 1024 bytes, 11152680 byte/sec, 85.1 Mb/sec 0.253 sec, 2000 aes crypts, 2048 bytes, 16165826 byte/sec, 123.3 Mb/sec 0.397 sec, 2000 aes crypts, 4096 bytes, 20618349 byte/sec, 157.3 Mb/sec 0.715 sec, 2000 aes crypts, 8192 bytes, 22919045 byte/sec, 174.9 Mb/sec 0.121 sec, 2000 aes192 crypts, 16 bytes, 263702 byte/sec, 2.0 Mb/sec 0.121 sec, 2000 aes192 crypts, 32 bytes, 528157 byte/sec, 4.0 Mb/sec 0.124 sec, 2000 aes192 crypts, 64 bytes, 1029899 byte/sec, 7.9 Mb/sec 0.126 sec, 2000 aes192 crypts, 128 bytes, 2033166 byte/sec, 15.5 Mb/sec 0.134 sec, 2000 aes192 crypts, 256 bytes, 3812020 byte/sec, 29.1 Mb/sec 0.150 sec, 2000 aes192 crypts, 512 bytes, 6819166 byte/sec, 52.0 Mb/sec 0.184 sec, 2000 aes192 crypts, 1024 bytes, 11143334 byte/sec, 85.0 Mb/sec 0.253 sec, 2000 aes192 crypts, 2048 bytes, 16213178 byte/sec, 123.7 Mb/sec 0.397 sec, 2000 aes192 crypts, 4096 bytes, 20609738 byte/sec, 157.2 Mb/sec 0.715 sec, 2000 aes192 crypts, 8192 bytes, 22900977 byte/sec, 174.7 Mb/sec 0.120 sec, 2000 aes256 crypts, 16 bytes, 265891 byte/sec, 2.0 Mb/sec 0.121 sec, 2000 aes256 crypts, 32 bytes, 528249 byte/sec, 4.0 Mb/sec 0.130 sec, 2000 aes256 crypts, 64 bytes, 982612 byte/sec, 7.5 Mb/sec 0.126 sec, 2000 aes256 crypts, 128 bytes, 2036887 byte/sec, 15.5 Mb/sec 0.135 sec, 2000 aes256 crypts, 256 bytes, 3804966 byte/sec, 29.0 Mb/sec 0.150 sec, 2000 aes256 crypts, 512 bytes, 6834321 byte/sec, 52.1 Mb/sec 0.183 sec, 2000 aes256 crypts, 1024 bytes, 11197804 byte/sec, 85.4 Mb/sec 0.252 sec, 2000 aes256 crypts, 2048 bytes, 16242946 byte/sec, 123.9 Mb/sec 0.399 sec, 2000 aes256 crypts, 4096 bytes, 20546571 byte/sec, 156.8 Mb/sec 0.712 sec, 2000 aes256 crypts, 8192 bytes, 23001641 byte/sec, 175.5 Mb/sec 0.061 sec, 1000 md5 crypts, 8 bytes, 130187 byte/sec, 1.0 Mb/sec 0.061 sec, 1000 md5 crypts, 16 bytes, 262188 byte/sec, 2.0 Mb/sec 0.060 sec, 1000 md5 crypts, 32 bytes, 530495 byte/sec, 4.0 Mb/sec 0.062 sec, 1000 md5 crypts, 64 bytes, 1037580 byte/sec, 7.9 Mb/sec 0.063 sec, 1000 md5 crypts, 128 bytes, 2032520 byte/sec, 15.5 Mb/sec 0.067 sec, 1000 md5 crypts, 256 bytes, 3848640 byte/sec, 29.4 Mb/sec 0.074 sec, 1000 md5 crypts, 512 bytes, 6908929 byte/sec, 52.7 Mb/sec 0.091 sec, 1000 md5 crypts, 1024 bytes, 11241259 byte/sec, 85.8 Mb/sec 0.124 sec, 1000 md5 crypts, 2048 bytes, 16492056 byte/sec, 125.8 Mb/sec 0.195 sec, 1000 md5 crypts, 4096 bytes, 20960734 byte/sec, 159.9 Mb/sec 0.352 sec, 1000 md5 crypts, 8192 bytes, 23274777 byte/sec, 177.6 Mb/sec 0.060 sec, 1000 sha1 crypts, 8 bytes, 132319 byte/sec, 1.0 Mb/sec 0.063 sec, 1000 sha1 crypts, 16 bytes, 252207 byte/sec, 1.9 Mb/sec 0.061 sec, 1000 sha1 crypts, 32 bytes, 527244 byte/sec, 4.0 Mb/sec 0.062 sec, 1000 sha1 crypts, 64 bytes, 1038843 byte/sec, 7.9 Mb/sec 0.064 sec, 1000 sha1 crypts, 128 bytes, 1990885 byte/sec, 15.2 Mb/sec 0.068 sec, 1000 sha1 crypts, 256 bytes, 3746908 byte/sec, 28.6 Mb/sec 0.075 sec, 1000 sha1 crypts, 512 bytes, 6813222 byte/sec, 52.0 Mb/sec 0.092 sec, 1000 sha1 crypts, 1024 bytes, 11184290 byte/sec, 85.3 Mb/sec 0.124 sec, 1000 sha1 crypts, 2048 bytes, 16464743 byte/sec, 125.6 Mb/sec 0.196 sec, 1000 sha1 crypts, 4096 bytes, 20918771 byte/sec, 159.6 Mb/sec 0.352 sec, 1000 sha1 crypts, 8192 bytes, 23274182 byte/sec, 177.6 Mb/sec [-- Attachment #3 --] 0.101 sec, 2000 des crypts, 8 bytes, 159107 byte/sec, 1.2 Mb/sec 0.098 sec, 2000 des crypts, 16 bytes, 327242 byte/sec, 2.5 Mb/sec 0.099 sec, 2000 des crypts, 32 bytes, 648548 byte/sec, 4.9 Mb/sec 0.100 sec, 2000 des crypts, 64 bytes, 1285838 byte/sec, 9.8 Mb/sec 0.103 sec, 2000 des crypts, 128 bytes, 2484303 byte/sec, 19.0 Mb/sec 0.110 sec, 2000 des crypts, 256 bytes, 4646646 byte/sec, 35.5 Mb/sec 0.125 sec, 2000 des crypts, 512 bytes, 8210128 byte/sec, 62.6 Mb/sec 0.153 sec, 2000 des crypts, 1024 bytes, 13393149 byte/sec, 102.2 Mb/sec 0.212 sec, 2000 des crypts, 2048 bytes, 19327501 byte/sec, 147.5 Mb/sec 0.336 sec, 2000 des crypts, 4096 bytes, 24382186 byte/sec, 186.0 Mb/sec 0.667 sec, 2000 des crypts, 8192 bytes, 24576283 byte/sec, 187.5 Mb/sec 0.098 sec, 2000 3des crypts, 8 bytes, 163718 byte/sec, 1.2 Mb/sec 0.098 sec, 2000 3des crypts, 16 bytes, 327429 byte/sec, 2.5 Mb/sec 0.099 sec, 2000 3des crypts, 32 bytes, 649634 byte/sec, 5.0 Mb/sec 0.100 sec, 2000 3des crypts, 64 bytes, 1283890 byte/sec, 9.8 Mb/sec 0.103 sec, 2000 3des crypts, 128 bytes, 2484496 byte/sec, 19.0 Mb/sec 0.110 sec, 2000 3des crypts, 256 bytes, 4643949 byte/sec, 35.4 Mb/sec 0.124 sec, 2000 3des crypts, 512 bytes, 8251543 byte/sec, 63.0 Mb/sec 0.153 sec, 2000 3des crypts, 1024 bytes, 13361343 byte/sec, 101.9 Mb/sec 0.213 sec, 2000 3des crypts, 2048 bytes, 19244955 byte/sec, 146.8 Mb/sec 0.340 sec, 2000 3des crypts, 4096 bytes, 24123989 byte/sec, 184.1 Mb/sec 0.667 sec, 2000 3des crypts, 8192 bytes, 24556502 byte/sec, 187.4 Mb/sec 0.098 sec, 2000 aes crypts, 16 bytes, 326574 byte/sec, 2.5 Mb/sec 0.100 sec, 2000 aes crypts, 32 bytes, 638302 byte/sec, 4.9 Mb/sec 0.100 sec, 2000 aes crypts, 64 bytes, 1281115 byte/sec, 9.8 Mb/sec 0.103 sec, 2000 aes crypts, 128 bytes, 2481438 byte/sec, 18.9 Mb/sec 0.110 sec, 2000 aes crypts, 256 bytes, 4645676 byte/sec, 35.4 Mb/sec 0.124 sec, 2000 aes crypts, 512 bytes, 8228270 byte/sec, 62.8 Mb/sec 0.154 sec, 2000 aes crypts, 1024 bytes, 13319199 byte/sec, 101.6 Mb/sec 0.212 sec, 2000 aes crypts, 2048 bytes, 19284097 byte/sec, 147.1 Mb/sec 0.339 sec, 2000 aes crypts, 4096 bytes, 24166118 byte/sec, 184.4 Mb/sec 0.667 sec, 2000 aes crypts, 8192 bytes, 24552307 byte/sec, 187.3 Mb/sec 0.098 sec, 2000 aes192 crypts, 16 bytes, 326771 byte/sec, 2.5 Mb/sec 0.100 sec, 2000 aes192 crypts, 32 bytes, 640474 byte/sec, 4.9 Mb/sec 0.100 sec, 2000 aes192 crypts, 64 bytes, 1285902 byte/sec, 9.8 Mb/sec 0.103 sec, 2000 aes192 crypts, 128 bytes, 2477595 byte/sec, 18.9 Mb/sec 0.110 sec, 2000 aes192 crypts, 256 bytes, 4642475 byte/sec, 35.4 Mb/sec 0.124 sec, 2000 aes192 crypts, 512 bytes, 8250546 byte/sec, 62.9 Mb/sec 0.154 sec, 2000 aes192 crypts, 1024 bytes, 13326393 byte/sec, 101.7 Mb/sec 0.212 sec, 2000 aes192 crypts, 2048 bytes, 19280375 byte/sec, 147.1 Mb/sec 0.340 sec, 2000 aes192 crypts, 4096 bytes, 24114828 byte/sec, 184.0 Mb/sec 0.666 sec, 2000 aes192 crypts, 8192 bytes, 24606586 byte/sec, 187.7 Mb/sec 0.098 sec, 2000 aes256 crypts, 16 bytes, 326667 byte/sec, 2.5 Mb/sec 0.100 sec, 2000 aes256 crypts, 32 bytes, 640038 byte/sec, 4.9 Mb/sec 0.100 sec, 2000 aes256 crypts, 64 bytes, 1280499 byte/sec, 9.8 Mb/sec 0.103 sec, 2000 aes256 crypts, 128 bytes, 2482015 byte/sec, 18.9 Mb/sec 0.110 sec, 2000 aes256 crypts, 256 bytes, 4646140 byte/sec, 35.4 Mb/sec 0.124 sec, 2000 aes256 crypts, 512 bytes, 8232173 byte/sec, 62.8 Mb/sec 0.153 sec, 2000 aes256 crypts, 1024 bytes, 13370503 byte/sec, 102.0 Mb/sec 0.213 sec, 2000 aes256 crypts, 2048 bytes, 19272392 byte/sec, 147.0 Mb/sec 0.340 sec, 2000 aes256 crypts, 4096 bytes, 24100072 byte/sec, 183.9 Mb/sec 0.666 sec, 2000 aes256 crypts, 8192 bytes, 24614312 byte/sec, 187.8 Mb/sec 0.048 sec, 1000 md5 crypts, 8 bytes, 167410 byte/sec, 1.3 Mb/sec 0.048 sec, 1000 md5 crypts, 16 bytes, 335359 byte/sec, 2.6 Mb/sec 0.048 sec, 1000 md5 crypts, 32 bytes, 669428 byte/sec, 5.1 Mb/sec 0.050 sec, 1000 md5 crypts, 64 bytes, 1279872 byte/sec, 9.8 Mb/sec 0.052 sec, 1000 md5 crypts, 128 bytes, 2444241 byte/sec, 18.6 Mb/sec 0.054 sec, 1000 md5 crypts, 256 bytes, 4708739 byte/sec, 35.9 Mb/sec 0.062 sec, 1000 md5 crypts, 512 bytes, 8270737 byte/sec, 63.1 Mb/sec 0.077 sec, 1000 md5 crypts, 1024 bytes, 13352632 byte/sec, 101.9 Mb/sec 0.105 sec, 1000 md5 crypts, 2048 bytes, 19491953 byte/sec, 148.7 Mb/sec 0.162 sec, 1000 md5 crypts, 4096 bytes, 25216551 byte/sec, 192.4 Mb/sec 0.329 sec, 1000 md5 crypts, 8192 bytes, 24927882 byte/sec, 190.2 Mb/sec 0.049 sec, 1000 sha1 crypts, 8 bytes, 162433 byte/sec, 1.2 Mb/sec 0.049 sec, 1000 sha1 crypts, 16 bytes, 325799 byte/sec, 2.5 Mb/sec 0.049 sec, 1000 sha1 crypts, 32 bytes, 659413 byte/sec, 5.0 Mb/sec 0.052 sec, 1000 sha1 crypts, 64 bytes, 1225373 byte/sec, 9.3 Mb/sec 0.052 sec, 1000 sha1 crypts, 128 bytes, 2473956 byte/sec, 18.9 Mb/sec 0.055 sec, 1000 sha1 crypts, 256 bytes, 4622524 byte/sec, 35.3 Mb/sec 0.062 sec, 1000 sha1 crypts, 512 bytes, 8281440 byte/sec, 63.2 Mb/sec 0.077 sec, 1000 sha1 crypts, 1024 bytes, 13222799 byte/sec, 100.9 Mb/sec 0.106 sec, 1000 sha1 crypts, 2048 bytes, 19245950 byte/sec, 146.8 Mb/sec 0.165 sec, 1000 sha1 crypts, 4096 bytes, 24767502 byte/sec, 189.0 Mb/sec 0.328 sec, 1000 sha1 crypts, 8192 bytes, 24942910 byte/sec, 190.3 Mb/sechelp
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200408150804.i7F84jCB010066>