12/04/88

Here is Reinhold Weicker's DHRYSTONE 2.1 benchmark, slightly modified
by me for instrumentation purposes only.  This is an integer CPU
benchmark.  The differences between 2.0 and 2.1 are minor, and
I believe that 2.0 and 2.1 results are comparable.

For purists, I've enclosed context diffs to turn what I'm posting
as 2.1 back into what Reinhold sent me as 2.1.  The changes
are in pure2_1.dif.  As you can see, it is only the instrumentation
that is slightly different.

Both the C and the Pascal versions are here.  The Ada version I
have is marked preliminary (not to be posted).

I've enclosed a new submission form (note new address for mailings).
Please deluge this mailbox (..!uunet!pcrat!dry2) with your results.
I'll summarize and repost when the dust clears.  Please do not
assume that I will pull posted results off of the net (I won't, its
too much work).

I've attempted to include a Makefile for UNIX and Microsoft C (with ndmake).
Pay particular attention to the HZ parameter, even though your power may
be 50 or 60 hertz, your computer may not be.  You may have to ask someone,
read the manual, or check:
	/usr/include/sys/param.h
	/usr/include/limits.h		(CLK_TCK==HZ)
for this information.

REPEAT: HZ IS NOT NECESSARILY THE FREQUENCY OF YOUR AC POWER !!!!!!

There are two versions to run, one with register variables, and one
without.  Please let the benchmark run for 30,000 loops on sixteen
bit machines, and for much longer (a minute or two) on faster machines.
Please note that "time(2)" has a resolution of 1 second, and may give
variable results.  No matter how time is measured, a sanity check with
a stopwatch is prudent.  We've run into systems that lie about time,
and there is always the configuration error problem.  When it comes
to time measurement on UNIX, there is no widely adhered to standard.

For segmented architectures, it is appropriate to submit results for
all memory models, as shown below.

I dropped the CODESIZE information that I attempted to add the last
time I posted this.  People couldn't follow directions, and the
information wasn't very useful anyway.

Here's a sample submission of results:

DHRYSTONE 2.1 BENCHMARK REPORTING FORM
MANUF: IBM
MODEL: PC/AT
PROC: 80286
CLOCK: 8
OS: Venix
OVERSION: SVr2.3
COMPILER: AT&T cc
CVERSION: 11/8/84
OPTIONS: -O
NOREG: 1450
REG: 1450
NOTES: HZ=60; Small Model
DATE: 03/04/88
SUBMITTER: pcrat!rick (Rick Richardson)
MAILTO: uunet!pcrat!dry2

DHRYSTONE 2.1 BENCHMARK REPORTING FORM
MANUF: IBM
MODEL: PC/AT
PROC: 80286
CLOCK: 8
OS: Venix
OVERSION: SVr2.3
COMPILER: AT&T cc
CVERSION: 11/8/84
OPTIONS: -O -Ml
NOREG: 1043
REG: 1043
NOTES: HZ=60; Large Model
DATE: 03/04/88
SUBMITTER: pcrat!rick (Rick Richardson)
MAILTO: uunet!pcrat!dry2

The program that processes submission forms is rather dumb.  Please
do not change the order, add or removes lines in the form.  If your
NOTES are longer than the space provided, then they are too long for
the summary.  Keep it terse, please.

A form consists of all lines between:
	DHRYSTONE 2.1 BENCHMARK REPORTING FORM
and
	MAILTO: uunet!pcrat!dry2
both lines must be present for the form to be processed.  If
a field does not apply or is not known, leave it blank.  The fields
are:
	MANUF:	Computer manufacturer, e.g. AT&T, IBM
	MODEL:	Model number of computer
	PROC:	If a microprocessor CPU, the part number, e.g. 68030
	CLOCK:	Clock in Mhz, if known.  Numeric only, e.g. 16.67
	OS:	Operating system, e.g. UNIX
	OVERSION: OS version, e.g. SVR3
	COMPILER: Compiler name, e.g. cc, Microsoft, or Green Hills
	CVERSION: Compiler version, e.g. 5.10
	OPTIONS: Relevant compiler options, e.g. -O3
	NOREG:	Dhrystones/second, no register attribute
	REG:	Dhrystones/second, with register attribute
	NOTES:	Additional, terse comments on one line
	DATE:	Date of test, US format MM/DD/YY
	SUBMITTER: uucp or domain address (full name)


HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ

Please, if you are using times(2) to measure time, indicate
the HZ value you used in the NOTES field.  Something like this is OK:

	NOTES: HZ=100; no inlining...

HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ HZ

--
		Rick Richardson, President, PC Research, Inc.

(201) 542-3734 (voice, nights)   OR     (201) 389-8963 (voice, days)
uunet!pcrat!rick (UUCP)			rick%pcrat.uucp@uunet.uu.net (INTERNET)
		uunet!pcrat!dry2   (Dhrystone submission forms only)