[speedstep] testers wanted

[Jochen Gensch]

Bruno Ducrot wrote:

dev.ichist.0.speedstep: 0
NB ~:openssl speed sha1
To get the most accurate results, try to run this
program when this computer is idle.
Doing sha1 for 3s on 16 size blocks: 713113 sha1's in 2.97s
Doing sha1 for 3s on 64 size blocks: 530042 sha1's in 2.98s
Doing sha1 for 3s on 256 size blocks: 352461 sha1's in 2.98s
Doing sha1 for 3s on 1024 size blocks: 149551 sha1's in 2.97s
Doing sha1 for 3s on 8192 size blocks: 23666 sha1's in 2.98s
OpenSSL 0.9.7d 17 Mar 2004
built on: Sun Sep 19 20:01:24 CEST 2004
options:bn(64,32) md2(int) rc4(idx,int) des(ptr,risc1,16,long) 
aes(partial) blowfish(idx)
compiler: cc
available timing options: USE_TOD HZ=128 [sysconf value]
timing function used: getrusage
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 
bytes
sha1              3837.34k    11373.23k    30244.12k    51526.49k 
65003.29k


> Could you please do something like that:
> sysctl dev.ichist.0.speedstep=1

SU NB /home/incmc:sysctl dev.ichist.0.speedstep=1
dev.ichist.0.speedstep: 0 -> 1
SU NB /home/incmc:openssl speed sha1
To get the most accurate results, try to run this
program when this computer is idle.
Doing sha1 for 3s on 16 size blocks: 474089 sha1's in 2.97s
Doing sha1 for 3s on 64 size blocks: 351283 sha1's in 2.96s
Doing sha1 for 3s on 256 size blocks: 234112 sha1's in 2.97s
Doing sha1 for 3s on 1024 size blocks: 99731 sha1's in 2.97s
Doing sha1 for 3s on 8192 size blocks: 15716 sha1's in 2.96s
OpenSSL 0.9.7d 17 Mar 2004
built on: Sun Sep 19 20:01:24 CEST 2004
options:bn(64,32) md2(int) rc4(idx,int) des(ptr,risc1,16,long) 
aes(partial) blowfish(idx)
compiler: cc
available timing options: USE_TOD HZ=128 [sysconf value]
timing function used: getrusage
The 'numbers' are in 1000s of bytes per second processed.
type             16 bytes     64 bytes    256 bytes   1024 bytes   8192 
bytes
sha1              2557.64k     7595.62k    20148.14k    34333.95k 
43423.69k


Guess it works... now this needs to be made dynamic :-)
Isn't there someone else, who is working on cpufreq (njl or so)?

Jochen