pp2.0.1-#4 results

Michael Griffith (grif@cs.ucr.edu)
26 Sep 1996 15:22:17 GMT


Here are some results I have seen on a 486DX50 with 16M.

Working drivers:

drivers/net/smc-ultra.c

Non-working drivers:

drivers/scsi/seagate.c
drivers/block/xd.c
drivers/block/cmd640.c

lmbench 1.0 results:

Processor, Processes - times in microseconds
--------------------------------------------
Host OS Mhz Null Null Simple /bin/sh Mmap 2-proc 8-proc
Syscall Process Process Process lat ctxsw ctxsw
--------- ------------- ---- ------- ------- ------- ------- ---- ------ ------
grief.1 Linux 2.0.21 50 12 3K 24K 61K 354 21 32
grief.2 Linux 2.1.0 50 11 3K 25K 64K 266 23 34

*Local* Communication latencies in microseconds
-----------------------------------------------
Host OS Pipe UDP RPC/ TCP RPC/
UDP TCP
--------- ------------- ------- ------- ------- ------- -------
grief.1 Linux 2.0.21 103 612 1155 814 1667
grief.2 Linux 2.1.0 127 629 1330 907 1883

*Local* Communication bandwidths in megabytes/second
----------------------------------------------------
Host OS Pipe TCP File Mmap Bcopy Bcopy Mem Mem
reread reread (libc) (hand) read write
--------- ------------- ---- ---- ------ ------ ------ ------ ---- -----
grief.1 Linux 2.0.21 12 5 11 13 11 11 15 38
grief.2 Linux 2.1.0 8 3 11 14 12 11 15 38

The OS Pipe and TCP numbers are alittle distrubing. I'm going to
re-run to double check.

-- 
Michael A. Griffith (grif@cs.ucr.edu) | http://www.cs.ucr.edu/~grif/
Department of Computer Science        | "My freedom of speech implies
University of California, Riverside   | your freedom to be offended."