On Mon, 15 May 2000, Petro wrote:
This apparently changes things to "Deferred function
binding". Is this going to cause a problem?
Yes. RTLD_NOW is the exact opposite of RTLD_LAZY:
flag must be either RTLD_LAZY, meaning resolve undefined
symbols as code from the dynamic library is executed, or
RTLD_NOW, meaning resolve all undefined symbols before
dlopen returns, and fail if this cannot be done.
Hmmm -- this seems to imply that either RTLD_LAZY or RTLD_NOW must be
set. Maybe in freebsd simply the absence of RTLD_LAZY is sufficient to
force RTLD_NOW behavior. In that case RTLD_NOW can simply be #defined to
0.
#ifndef RTLD_NOW
#define RTLD_NOW 0
#endif
But the question is whether this is the right thing to do.
Hmmm... Ok. I put the above in soxwrap.c and restarted the compile.
That compiled just fine, but the build died when compiling
courier/libs/comstatfs.c with the error sys/statfs.h no such fie or
directory.
With OpenBSD you need to include <sys/param.h> and
<sys/mount.h> instead of <sys/statfs.h>.
I changed:
#if HAVE_STATFS
#include <sys/statfs.h>
#define DOIT
To:
#if HAVE_STATFS
#include <sys/param.h>
#include <sys/mount.h>
#define DOIT
And it (at least that portion) seems to compile OK.
Is there any way that the compiler can check to see what the
OS is so that I could do something like:
#if HAVE_STATFS
#if IS_OPENBSD
#include <sys/param.h>
#include <sys/mount.h>
#else
#include <sys/statfs.h>
#endif
#define DOIT
#endif
Or am I completely barking up the wrong tree?
4 messages in net.sourceforge.lists.courier-usersRe: [courier-users] trouble compiling...
