home
***
CD-ROM
|
disk
|
FTP
|
other
***
search
/
Technotools
/
Technotools (Chestnut CD-ROM)(1993).ISO
/
lang_oth
/
mawk10
/
uconfig
< prev
next >
Wrap
Text File
|
1991-10-05
|
2KB
|
93 lines
Mawk has been compiled and tested successfully in the following
Unix or Unix like environments.
All were compiled with cc, those marked were
also compiled with gcc. The script build_mawk recognizes
the systems in the left column of the following table.
system compiler (cc and ...)
sun_os40 gcc
sun3_os40_no68881
sun_os41 gcc
sun3_os41_no68881
ultrix42_mips
ultrix41_mips gcc
ultrix41_vax
ultrix31_vax
bsd43_vax gcc
sysV gcc
sysVsco gcc
sysV_ieeefp
xenix_r2
stardentVr3
apolloSR10.3
dynix
To make mawk on one of these systems, run
build_mawk [CC=your_cc] system
V7: This port is the work of Carl Mascott. You'll need to read
his notes and use the .v7 Makefiles
NOTES:
sun_os40
sun3_os40_no68881
sun 3/[56]0 running SunOS4.0.3
sun4/110 , 4/280 OS4.0.3
sun_os41
sparc station 1
sun3_os41_no68881
never really tested, same as sun_os41 except
does not set -f68881 flag for cc
sysV
tested on ESIX System 5.3.2 Rev D, using gcc 1.39
sysVsco
tested on SCO UNIX R3.2.2 , cc and gcc(1.40)
this is the same as sysV except it caters to SCO's
ranlib strangeness
sysV_ieeefp
This should but probably won't work for any sysV with
IEEE754 floating point. i386 and i486 fall in this
category, but if SCO and ESIX are typical, somebody
forgot to rewrite libm to handle INF and NAN.
If you try it -- if it doesn't work fpe_test will catch the
problem and then use sysV.
tested and failed on SCO UNIX R3.2v2.0
tested and failed on ESIX
because the
math library is not up to speed
(Aside: if I had a 387 or 486 and libm didn't support it
I'd howl long and loud to my vendor. SunOS and
Ultrix (mips) both run on IEEE754 hardware and the
math libraries support the IEEE754 standard )
stardentVr3
tested on stardent 3000. Mips processors and traps are off
by default.
Doesn't have fpsetmask() so don't use sysV_ieeefp.
dynix
Sequent symmetry