[Orca-dev] SE Toolkit on Solaris 10 x86 64 bit

[Alex Kiernan]

On 4/20/05, Alex Kiernan <alex.kiernan at gmail.com> wrote:
> On 4/20/05, Blair Zajac <blair at orcaware.com> wrote:
> > Alex Kiernan wrote:
> >
> > >Anyone got anywhere getting SE Toolkit to work on Solaris 10 x86 64
> > >bit (my guess is its the varargs manipulations which trip it up but
> > >I've really not looked very hard)?
> > >
> > >If not I'll see if I can russle up someone here to have a look at it.
> > >
> > >
> > >
> > What problems are you seeing?  Is this with SE 3.4?
> >
> 
> Yeah, 3.4, compiled 64 bit. Two (so far), it core dumps when it hits
> the string formatting stuff and there's something screwy with the stat
> bindings. Running it against orcallator:
> 

[...]

> But if you try avoid stat's it cores:
> 
> core 'se.632.1114013425' of 632:        /opt/RICHPse/bin/se -d
> /opt/RICHPse/examples/hostid.se
>  fffffd7fff0f2046 _ndoprnt () + 2356
>  fffffd7fff0f3d7e vfprintf () + 5e
>  000000000043cfa6 run_intf () + 2f6
>  000000000043d0f3 run_printf () + 23
>  000000000043f91d run_call () + bd
>  000000000043b4ed run_block () + ed
>  000000000043b3c1 se_run () + 2c1
>  00000000004395ce main () + f5e
>  000000000041151c ???????? ()
> 

I should have said "if you hack it enough so it at least compiles it cores".

The stat problems turn out to be trivial - assorted insertion of
defined(_LP64) in the includes and some fixing of LD_LIBRARY_PATH in
the source.

-- 
Alex Kiernan