Kernels benchmarked: 2.0.31pre2, 2.1.46, 2.1.46+pentium memcpy
2.0.31pre2...
BYTE UNIX Benchmarks (Version 4.0.1)
System -- Linux pc23.lmh.ox.ac.uk 2.0.30 #4 Mon Jun 16 02:30:13 BST 1997 i586 unknown
Start Benchmark Run: Tue Jul 29 20:06:18 BST 1997
0 interactive users.
8:06pm up 23 min, 0 users, load average: 0.15, 0.51, 0.52
lrwxrwxrwx 1 root root 4 Jun 15 22:54 /bin/sh -> bash
/bin/sh: symbolic link to bash
/dev/hdc7 495714 286323 183790 61% /home
Execl Throughput 127.3 lps (29 secs, 3 samples)
File Read 1024 bufsize 2000 maxblocks 34302.0 KBps (30 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 37422.0 KBps (30 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 19799.0 KBps (30 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 21116.0 KBps (30 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 15375.0 KBps (30 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks 8541.0 KBps (30 secs, 3 samples)
File Read 4096 bufsize 8000 maxblocks 41450.0 KBps (30 secs, 3 samples)
File Write 4096 bufsize 8000 maxblocks 40266.0 KBps (30 secs, 3 samples)
File Copy 4096 bufsize 8000 maxblocks 22325.0 KBps (30 secs, 3 samples)
Pipe Throughput 71753.2 lps (10 secs, 10 samples)
Pipe-based Context Switching 24440.1 lps (10 secs, 10 samples)
Process Creation 876.2 lps (30 secs, 3 samples)
System Call Overhead 76802.9 lps (10 secs, 10 samples)
INDEX VALUES
TEST BASELINE RESULT INDEX
Execl Throughput 43.0 127.3 29.6
File Copy 1024 bufsize 2000 maxblocks 3960.0 19799.0 50.0
File Copy 256 bufsize 500 maxblocks 1655.0 8541.0 51.6
File Copy 4096 bufsize 8000 maxblocks 5800.0 22325.0 38.5
Pipe Throughput 12440.0 71753.2 57.7
Pipe-based Context Switching 4000.0 24440.1 61.1
Process Creation 126.0 876.2 69.5
System Call Overhead 15000.0 76802.9 51.2
=========
FINAL SCORE 49.6
2.1.46 (plain)
BYTE UNIX Benchmarks (Version 4.0.1)
System -- Linux pc23.lmh.ox.ac.uk 2.1.46 #3 Wed Jul 23 23:18:14 BST 1997 i586 unknown
Start Benchmark Run: Tue Jul 29 21:40:02 BST 1997
0 interactive users.
9:40pm up 1 min, 0 users, load average: 0.13, 0.05, 0.01
lrwxrwxrwx 1 root root 4 Jun 15 22:54 /bin/sh -> bash
/bin/sh: symbolic link to bash
/dev/hdc7 495714 286332 183781 61% /home
Execl Throughput 136.1 lps (29 secs, 3 samples)
File Read 1024 bufsize 2000 maxblocks 37894.0 KBps (30 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 40621.0 KBps (30 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 21153.0 KBps (30 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 28660.0 KBps (30 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 17127.0 KBps (30 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks 9716.0 KBps (30 secs, 3 samples)
File Read 4096 bufsize 8000 maxblocks 42653.0 KBps (30 secs, 3 samples)
File Write 4096 bufsize 8000 maxblocks 43199.0 KBps (30 secs, 3 samples)
File Copy 4096 bufsize 8000 maxblocks 23483.0 KBps (30 secs, 3 samples)
Pipe Throughput 101159.2 lps (10 secs, 10 samples)
Pipe-based Context Switching 32738.5 lps (10 secs, 10 samples)
Process Creation 934.6 lps (30 secs, 3 samples)
System Call Overhead 103402.4 lps (10 secs, 10 samples)
INDEX VALUES
TEST BASELINE RESULT INDEX
Execl Throughput 43.0 136.1 31.7
File Copy 1024 bufsize 2000 maxblocks 3960.0 21153.0 53.4
File Copy 256 bufsize 500 maxblocks 1655.0 9716.0 58.7
File Copy 4096 bufsize 8000 maxblocks 5800.0 23483.0 40.5
Pipe Throughput 12440.0 101159.2 81.3
Pipe-based Context Switching 4000.0 32738.5 81.8
Process Creation 126.0 934.6 74.2
System Call Overhead 15000.0 103402.4 68.9
=========
FINAL SCORE 58.5
And 2.1.46 with pentium memcpy....
BYTE UNIX Benchmarks (Version 4.0.1)
System -- Linux pc23.lmh.ox.ac.uk 2.1.46 #5 Thu Jul 24 20:57:32 BST 1997 i586 unknown
Start Benchmark Run: Tue Jul 29 22:12:15 BST 1997
0 interactive users.
10:12pm up 1 min, 0 users, load average: 0.06, 0.03, 0.01
lrwxrwxrwx 1 root root 4 Jun 15 22:54 /bin/sh -> bash
/bin/sh: symbolic link to bash
/dev/hdc7 495714 286336 183777 61% /home
Execl Throughput 141.7 lps (29 secs, 3 samples)
File Read 1024 bufsize 2000 maxblocks 37025.0 KBps (30 secs, 3 samples)
File Write 1024 bufsize 2000 maxblocks 16316.0 KBps (30 secs, 3 samples)
File Copy 1024 bufsize 2000 maxblocks 18393.0 KBps (30 secs, 3 samples)
File Read 256 bufsize 500 maxblocks 29744.0 KBps (30 secs, 3 samples)
File Write 256 bufsize 500 maxblocks 16740.0 KBps (30 secs, 3 samples)
File Copy 256 bufsize 500 maxblocks 9957.0 KBps (30 secs, 3 samples)
File Read 4096 bufsize 8000 maxblocks 40028.0 KBps (30 secs, 3 samples)
File Write 4096 bufsize 8000 maxblocks 36177.0 KBps (30 secs, 3 samples)
File Copy 4096 bufsize 8000 maxblocks 17584.0 KBps (30 secs, 3 samples)
Pipe Throughput 92532.3 lps (10 secs, 10 samples)
Pipe-based Context Switching 31053.9 lps (10 secs, 10 samples)
Process Creation 1062.5 lps (30 secs, 3 samples)
System Call Overhead 96129.0 lps (10 secs, 10 samples)
INDEX VALUES
TEST BASELINE RESULT INDEX
Execl Throughput 43.0 141.7 33.0
File Copy 1024 bufsize 2000 maxblocks 3960.0 18393.0 46.4
File Copy 256 bufsize 500 maxblocks 1655.0 9957.0 60.2
File Copy 4096 bufsize 8000 maxblocks 5800.0 17584.0 30.3
Pipe Throughput 12440.0 92532.3 74.4
Pipe-based Context Switching 4000.0 31053.9 77.6
Process Creation 126.0 1062.5 84.3
System Call Overhead 15000.0 96129.0 64.1
=========
FINAL SCORE 55.3
Interesting, eh?
It seemed to be suggested that the faster your RAM, he better the patch
will do for you. I have 60ns EDO ram in a high performance motherboard +
P100. The motherboard 'automatically' configures RAM timings and runs mine
at x2222 read/x2222 write.
Chris