RE: Solaris Large File Compilation Environment and Perl

[Stefan R. <Ste...@CT...>]

What about procfs?

$ uname -srpi
SunOS 5.9 sparc SUNW,UltraSPARC-IIi-cEngine
$ grep '#error' /usr/include/sys/* | grep Cannot
/usr/include/sys/old_procfs.h:#error    "Cannot use procfs in the large =
file
compilation environment"
/usr/include/sys/procfs.h:#error        "Cannot use procfs in the large =
file
compilation environment"
/usr/include/sys/swap.h:#error  "Cannot use swapctl in the large files
compilation environment"

Stefan
P.S.: I'm not a 32/64 wizard

> -----Urspr=FCngliche Nachricht-----
> Von: Bruce Shaw [mailto:Bru...@go...]
> Gesendet: Dienstag, 25. M=E4rz 2003 00:27
> An: Net-Snmp-Coders (E-mail)
> Betreff: Solaris Large File Compilation Environment and Perl
>=20
> I believe I've found a better solution to our problems with net-snmp =
and
> Sun's version of Perl hating each other.
>=20
> The problem is, in a nutshell, the use of swap.h inside
> agent/mibgroup/host/hr_storage.c requiring the "large file =
compilation
> environment".
>=20
> If you look in http://docs.sun.com/db/doc/806-0634/6j9vo5alv?a=3Dview =
 which
> is part of the Solaris 64-bit Developers guide
> (http://docs.sun.com/doc/806-0477?q=3Dlarge+file+environment)
>=20
> It says "Applications may combine objects produced in the large file
> compilation environment with objects produced in the transitional
> compilation environment, but must be careful with respect to
> interoperability between those objects. Applications should not =
declare
> global variables of types whose sizes change between compilation
> environments."
>=20
> The problem specifically in hr_storage.c is
>=20
> #if defined(solaris2)
> #if HAVE_SYS_SWAP_H
> #include <sys/swap.h>
> #endif
> #endif
>=20
> ...
>=20
> #ifdef solaris2
> void
> sol_get_swapinfo(int *totalP, int *usedP)
> {
>     struct anoninfo ainfo;
>=20
>     if (swapctl(SC_AINFO, &ainfo) < 0) {
>         *totalP =3D *usedP =3D 0;
>         return;
>     }
>=20
>     *totalP =3D ainfo.ani_max;
>     *usedP =3D ainfo.ani_resv;
> }
> #endif                          /* solaris2 */
>=20
> If we were to put this into a separate module and compile it WITHOUT
> largefile support (which generates the OFFSET_64 problem which is =
causing
> all the fuss), and pass two simple integers instead of the anoninfo =
data
> structure it should be able to be included into the larger net-snmp =
which
> WOULD compile with the largefile (and OFFSET_64) support and play =
nice
> with
> Perl.
>=20
> How to specifically do it, especially explaining all this to =
./configure
> is
> waaaaayyy over my head.
>=20
> This communication is intended for the use of the recipient to which =
it is
> addressed, and may contain confidential, personal and or privileged
> information. Please contact us immediately if you are not the =
intended
> recipient of this communication, and do not copy, distribute, or take
> action
> relying on it. Any communication received in error, or subsequent =
reply,
> should be deleted or destroyed.
>=20
>=20
>=20
> -------------------------------------------------------
> This SF.net email is sponsored by:
> The Definitive IT and Networking Event. Be There!
> NetWorld+Interop Las Vegas 2003 -- Register today!
> http://ads.sourceforge.net/cgi-bin/redirect.pl?keyn0001en
> _______________________________________________
> Net-snmp-coders mailing list
> Net...@li...
> https://lists.sourceforge.net/lists/listinfo/net-snmp-coders