----- Original Message -----
From: "Gilles Detillieux" <[EMAIL PROTECTED]>
To: "David J. Graff" <[EMAIL PROTECTED]>
Cc: "ht://Dig General" <[EMAIL PROTECTED]>
Sent: Friday, May 04, 2001 2:52 PM
Subject: Re: [htdig] [PATCH] A whole cornucopia of patches to allow
compilation on Solaris 2.6
> According to David J. Graff:
> > Attached is a tar.gz file of patches that I have had to make to the
source
> > tree to compile under solaris.
> >
> > I'm able to cleanly compile and link the binaries. There is an error
I'm
> > getting with a arithmetic exception when linked
> > dynamically that's odd.
>
> Did you try configuring with --disable-shared, as was suggested? This is
> a known bug from back when the shared library support was introduced in
> 3.2. It works on Linux, and possibly BSD, but fails on Solaris, HP-UX,
> and possibly others.
I was able to get it to run in static link mode. I have to contact a friend
of mine who's a wizard with
shared libraries and stuff like this. Managed to get one source tree to
compile/link on Sun, AIX, and HP.
Kinda wild!
> > Can someone please look at the change to htcommon/conf_parser.cxx. I
wasn't
> > truly sure if this was a declaration I've never seen before or something
> > else.
>
> This must be a product of yacc/bison on your system. The conf_parser.cxx
> in the CVS source tree doesn't have these strange constructs.
Ok I'll have to check into it.
> > I'll be banging on the problem I'm seeing with the arithmetic error
Monday.
> > May just be an odd variable declaration that Solaris libc doesn't like.
>
> Unless you've spotted something different than the usual cause of these
> errors, it's due to objects in shared libraries not being initialized
> properly. If you can help point to a solution for this, you'd earn the
> gratitute of many.
As I said above I'm going to contact my friend.
> I took a quick look at your patches, and most seem just fine to me.
> I think Geoff should comment on the changes to the HtHeap and HtVector
> classes, though, because I think what you're doing defeats the purpose
> of what Geoff was trying to accomplish. Unfortunately, this is beyond
> my limited understanding of C++ class hierarchies.
Ok I'll take another look at the changes and see if I can find a simple
solution that will
work all around.
--
David J. Graff
mailto:[EMAIL PROTECTED]
http://www.mindspring.com/~phlat
_______________________________________________
htdig-general mailing list <[EMAIL PROTECTED]>
To unsubscribe, send a message to <[EMAIL PROTECTED]> with a
subject of unsubscribe
FAQ: http://htdig.sourceforge.net/FAQ.html
