Hello,
On one of my machines, the results of comparing SSLeay-0.9.0b against
OpenSSL-0.9.2b are disappointing:
(have a look at the md5, hmac, sha1, rmd160, rc5/32 and cast
result lines in the following tables:
performance has dropped by a factor of 2.5 ... 4!!!
Is there any recent modification which could explain these results?
The same compiler and platform was used, and on the (3 processor
machine) I ran these tests sequentially first, and parallel on
2 processors to eliminate workload glitches: the results are
identical within statistical range.
SSLeay 0.9.0b 29-Jun-1998
built on Wed Feb 24 17:23:25 MET 1999
options:bn(32,32) md2(int) rc4(idx,char) des(idx,cisc,4,long) idea(int)
blowfish(idx)
C flags not available
The 'numbers' are in 1000s of bytes per second processed.
type 8 bytes 64 bytes 256 bytes 1024 bytes 8192
bytes
md2 5959.75 16.53k 22.39k 24.35k
25.10k
mdc2 11.26k 11.92k 11.98k 12.03k
12.01k
md5 113.70k 450.62k 689.13k 789.35k
824.77k
hmac(md5) 48.62k 299.29k 699.56k 1045.03k
1219.15k
sha1 69.84k 318.53k 558.73k 690.60k
737.01k
rmd160 48.36k 204.54k 340.13k 406.83k
431.74k
rc4 220.41k 263.02k 268.63k 270.65k
271.22k
des cbc 56.22k 62.38k 63.22k 63.44k
63.44k
des ede3 21.18k 22.07k 22.15k 22.20k
21.39k
idea cbc 93.88k 114.27k 116.54k 117.18k
117.42k
rc2 cbc 50.94k 55.90k 56.35k 56.58k
56.59k
rc5-32/12 cbc 270.62k 381.33k 397.84k 402.18k
404.06k
blowfish cbc 102.12k 129.45k 134.05k 134.81k
134.70k
cast cbc 253.60k 350.96k 362.55k 369.12k
369.61k
sign verify sign/s verify/s
rsa 512 bits 0.7638s 0.0920s 1.3 10.9
rsa 1024 bits 5.1950s 0.3411s 0.2 2.9
rsa 2048 bits 38.3400s 1.3175s 0.0 0.8
sign verify sign/s verify/s
dsa 512 bits 0.8809s 1.7233s 1.1 0.6
dsa 1024 bits 3.2333s 6.5100s 0.3 0.2
dsa 2048 bits 12.4500s 25.2200s 0.1 0.0
- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - -
OpenSSL 0.9.2b 22 Mar 1999
built on: Wed Apr 14 11:04:25 MET DST 1999
options:bn(32,32) md2(int) rc4(idx,char) des(idx,cisc,4,long) idea(int)
blowfish(idx)
compiler: c89 -XLLML -XLLMK -XL -DB_ENDIAN -DTERMIOS -DCHARSET_EBCDIC
The 'numbers' are in 1000s of bytes per second processed.
type 8 bytes 64 bytes 256 bytes 1024 bytes 8192
bytes
md2 5878.26 16.14k 21.55k 23.56k
24.17k
mdc2 10.92k 11.55k 11.64k 11.64k
11.57k
md5 47.99k 160.02k 219.06k 241.82k
247.41k
hmac(md5) 23.67k 139.11k 300.43k 422.36k
478.32k
sha1 27.88k 136.72k 250.19k 315.27k
342.02k
rmd160 15.21k 63.27k 101.88k 120.61k
126.88k
rc4 221.34k 263.63k 268.28k 270.41k
271.24k
des cbc 55.78k 61.83k 62.45k 62.54k
62.60k
des ede3 21.17k 22.07k 22.18k 21.63k
21.99k
idea cbc 94.91k 113.75k 116.27k 116.95k
116.63k
rc2 cbc 50.80k 55.98k 56.53k 56.70k
56.59k
rc5-32/12 cbc 102.12k 125.08k 127.91k 124.01k
127.31k
blowfish cbc 104.60k 129.92k 133.64k 134.36k
134.70k
cast cbc 80.88k 93.64k 96.17k 96.52k
96.54k
sign verify sign/s verify/s
rsa 512 bits 0.8855s 0.0999s 1.1 10.0
rsa 1024 bits 5.7550s 0.3461s 0.2 2.9
rsa 2048 bits 39.1400s 1.2362s 0.0 0.8
sign verify sign/s verify/s
dsa 512 bits 0.9600s 1.1733s 1.0 0.9
dsa 1024 bits 3.2900s 3.9933s 0.3 0.3
dsa 2048 bits 11.5900s 14.2100s 0.1 0.1
Martin
--
<[EMAIL PROTECTED]> | Siemens Information and
Phone: +49-89-636-46021 | Communication Products
FAX: +49-89-636-47816 | 81730 Munich, Germany
______________________________________________________________________
OpenSSL Project http://www.openssl.org
Development Mailing List [EMAIL PROTECTED]
Automated List Manager [EMAIL PROTECTED]
