On Tue, 16 Dec 2008 11:22:49 +0100 I. Szczesniak wrote:
> On 12/16/08, Dr. Werner Fink <[email protected]> wrote:
> > On i586/i686:
> >
> > test io.sh begins at 2008-12-15+10:37:40
> > shtests: line 91: 24720: Abort
> Please send a stack trace from the code dump to the list.
yes, please
> > test arith.sh begins at 2008-12-15+10:42:55
> > arith.sh[452]: printf '%g\n' $((-Nan)) failed -- expected -nan,
> > got nan
> > arith.sh[454]: printf '%g\n' -Nan failed -- expected -nan, got nan
> > arith.sh[456]: print -- $((-Nan)) failed -- expected -nan, got nan
> > test arith.sh failed at 2008-12-15+10:42:55 with exit code 3 [ 127 tests
> > 3 errors ]
> Did you disable IEEE754 math support? Negative nan is defined in
> IEEE754-1985 and mandatory required in IEEE 754-2008.
the sign of { 0.0 Inf NaN } is determined by the <math.h> signbit()
implementation
these architectures in our build farm have faulty/missing signbit()
HOSTTYPE signbit uname -a
-------- ------- --------
darwin.i386 faulty Darwin 9.5.0 Darwin Kernel Version 9.5.0: Wed Sep 3
11:29:43 PDT 2008; root:xnu-1228.7.58~1/RELEASE_I386 i386
ibm.risc missing AIX - 3 4 000D4CDD4C00
linux.ia64 faulty Linux - 2.4.21-sgi302r24 #1 SMP Fri Oct 22 22:43:12 PDT
2004 ia64 ia64 ia64 GNU/Linux
linux.ia64 faulty Linux - 2.6.9-11.EL #1 SMP Fri May 20 18:25:13 EDT 2005
ia64 ia64 ia64 GNU/Linux
sgi.mips3 missing IRIX64 - 6.5 07202013 IP35 mips mips3 IRIX64
sol8.sun4 missing SunOS - 5.8 Generic_108528-26 sun4u sparc
SUNW,Sun-Fire-15000
I believe the correct c99 incantations were done on the faulty architectures
because they made signbit() visible
note that ksh93 does long double arithmetic when long double is supported
-- Glenn Fowler -- AT&T Research, Florham Park NJ --
_______________________________________________
ast-developers mailing list
[email protected]
https://mailman.research.att.com/mailman/listinfo/ast-developers
