ubix2 / src / contrib / gdtoa / test /
@reddawg reddawg authored on 26 Oct 2006
..
Q.ou0 Adding 18 years ago
Q.ou1 Adding 18 years ago
Qtest.c Adding 18 years ago
README Adding 18 years ago
d.out Adding 18 years ago
dI.out Adding 18 years ago
dIsi.out Adding 18 years ago
dItest.c Adding 18 years ago
dd.out Adding 18 years ago
ddsi.out Adding 18 years ago
ddtest.c Adding 18 years ago
dt.c Adding 18 years ago
dtest.c Adding 18 years ago
dtst.out Adding 18 years ago
f.out Adding 18 years ago
ftest.c Adding 18 years ago
getround.c Adding 18 years ago
makefile Adding 18 years ago
rtestnos Adding 18 years ago
strtoIdSI.c Adding 18 years ago
strtoIddSI.c Adding 18 years ago
strtodISI.c Adding 18 years ago
strtodt.c Adding 18 years ago
strtopddSI.c Adding 18 years ago
strtorddSI.c Adding 18 years ago
testnos Adding 18 years ago
testnos1 Adding 18 years ago
testnos3 Adding 18 years ago
x.ou0 Adding 18 years ago
x.ou1 Adding 18 years ago
xL.ou0 Adding 18 years ago
xL.ou1 Adding 18 years ago
xLtest.c Adding 18 years ago
xQtest.c Adding 18 years ago
xsum0.out Adding 18 years ago
xtest.c Adding 18 years ago
README
This directory contains source for several test programs:

dt is for conversion to/from double; it permits input of pairs of
32-bit hex integers as #hhhhhhhh hhhhhhhh (i.e., the initial '#'
indicates hex input).  No initial # ==> decimal input.
After the input number is an optional : mode ndigits
(colon, and decimal integers for parameters "mode" and "ndigits"
to gdtoa).

Qtest, ddtest, dtest, ftest, xLtest and xtest are for conversion to/from

	f	IEEE single precision
	d	IEEE double precision
	xL	IEEE extended precision, as on Motorola 680x0 chips
	x	IEEE extended precision, as on Intel 80x87 chips or
			software emulation of Motorola 680x0 chips
	Q	quad precision, as on Sun Sparc chips
	dd	double double, pairs of IEEE double numbers
		whose sum is the desired value

They're all similar, except for the precision.  They test both
directed roundings and interval input (the strtoI* routines).
Lines that begin with "r" specify or interrogate the desired rounding
direction:

	0 = toward 0
	1 = nearest (default)
	2 = toward +Infinity
	3 = toward -Infinity

These are the FPI_Round_* values in gdota.h.  The "r" value is sticky:
it stays in effect til changed.  To change the value, give a line that
starts with r followed by 0, 1, 2, or 3.  To check the value, give "r"
by itself.

Lines that begin with n followed by a number specify the ndig
argument for subsequent calls to the relevant g_*fmt routine.

Lines that start with # followed by the appropriate number of
hexadecimal strings (see the comments) give the big-endian
internal representation of the desired number.

When routines Qtest, xLtest, and xtest are used on machines whose
long double is of type "quad" (for Qtest) or "extended" (for x*test),
they try to print with %Lg as another way to show binary values.

Program ddtest also accepts (white-space separated) pairs of decimal
input numbers; it converts both with strtod and feeds the result
to g_ddfmt.

Program dItest exercises strtodI and strtoId.

Programs dItestsi and ddtestsi are for testing the sudden-underflow
logic (on double and double-double conversions).

Program strtodt tests strtod on some hard cases (in file testnos3)
posted by Fred Tydeman to comp.arch.arithmetic on 26 Feb. 1996.

These are simple test programs, not meant for exhaustive testing,
but for manually testing "interesting" cases.  Paxson's testbase
is good for more exhaustive testing, in part with random inputs.