On Tue, Nov 16, 2004 at 11:49:02AM +0000 Nicholas Clark wrote:
> On Tue, Nov 16, 2004 at 07:52:09AM +0100, Tassilo von Parseval wrote:
> > Hi,
> >
> > right now I'm lost a bit in the muddle of the various perl compilation
> > flags and the sizes of some C datatypes. The task is simply to get the
> > numerical maximum of a list of values. Could someone have a look at the
> > following XSUB, especially at the #defines at the beginning? In
> > particular, I am not sure about using long doubles when 64bit integers
> > are in effect:
>
> On some platforms, sizeof(long double) is 8, so they can't accurately hold
> all the values that a 64 bit integer holds. So if you really want to be
> completely portable you're going to need something more complex than this
> which deals differently with IVs and NVs if the SV is an IV with a large
> value.
Hate!
> This will get more complex still if you need it to work pre 5.8, because
> in 5.6 and earlier the IvOK flag is set if the IV slot is valid as an
> integer, rather than if it's numerically identical to the NV, because NV maths
> takes priority. eg:
>
>
> $ ~/Reference/5.6.2/bin/perl5.6.2-32 -MDevel::Peek -e '$a = 3.2; $a & 1; Dump
> $a'
> SV = PVNV(0x8010a0) at 0x806e6c
> REFCNT = 1
> FLAGS = (IOK,NOK,pIOK,pNOK,IsUV)
> UV = 3
> NV = 3.2
> PV = 0
Hate++ !
So it looks I will have to drown my little XSUBS in even more #ifs and
#defines. At least I am now aware of two more pitfalls that I will have
to guard against.
Cheers,
Tassilo
--
$_=q#",}])!JAPH!qq(tsuJ[{@"tnirp}3..0}_$;//::niam/s~=)]3[))_$-3(rellac(=_$({
pam{rekcahbus})(rekcah{lrePbus})(lreP{rehtonabus})!JAPH!qq(rehtona{tsuJbus#;
$_=reverse,s+(?<=sub).+q#q!'"qq.\t$&."'!#+sexisexiixesixeseg;y~\n~~dddd;eval
