The meat of the log is the part relevant to the failure - in the future,
please cut and past that, so (if you get it right, natch), I don't have to
go digging... Anyway...
What you find in config.log is the series of lines telling you about each
test. If it works, you get a minimal 2 lines:
configure:9051: checking for gd.h
configure:9058: result: yes
If it fails, you get the full details of the test program and all the
messages:
configure:9072: checking for gdImageDestroy in -lgd
configure:9103: gcc -o
conftest -g -O2 -I/usr/local/include -Wshadow -Wpointer-arith -Wmissing-prot
otypes -Wmissing-declarations -Wnested-externs -fPIC -DHAVE_FILEDESCRIPTORB
UG -DFREEBSD -I/usr/local/include -I/usr/local/include -I/usr/local/include
-L/usr/local/lib -L/usr/local/lib -L/tmp/ntop/gdchart0.94c/gd-1.8.3/
conftest.c -lgd -lgdbm -lpcap -lpcap -lgdbm -lgd >&5
/usr/local/lib/libgd.a(gd.o): In function `HWB_to_RGB':
gd.o(.text+0x4c1): undefined reference to `floor'
/usr/local/lib/libgd.a(gd.o): In function `gdImageCopyResized':
gd.o(.text+0x21e8): undefined reference to `floor'
gd.o(.text+0x22b0): undefined reference to `floor'
configure:9106: $? = 1
configure: failed program was:
| #line 9080 "configure"
| /* confdefs.h. */
|
| #define PACKAGE_NAME "ntop"
| #define PACKAGE_TARNAME "ntop"
| #define PACKAGE_VERSION "2.2.96"
| #define PACKAGE_STRING "ntop 2.2.96"
| #define PACKAGE_BUGREPORT ""
| #define PACKAGE "ntop"
| #define VERSION "2.2.96"
| #define STDC_HEADERS 1
| #define HAVE_SYS_TYPES_H 1
| #define HAVE_SYS_STAT_H 1
| #define HAVE_STDLIB_H 1
| #define HAVE_STRING_H 1
| #define HAVE_MEMORY_H 1
| #define HAVE_STRINGS_H 1
| #define HAVE_INTTYPES_H 1
| #define HAVE_UNISTD_H 1
| #define CFG_LITTLE_ENDIAN 1
| #define HAVE_PCAP_H 1
| #define HAVE_LIBPCAP 1
| #define HAVE_PCAP_OPEN_DEAD 1
| #define HAVE_GDBM_H 1
| #define HAVE_LIBGDBM 1
| #define HAVE_GD_H 1
| /* end confdefs.h. */
|
| /* Override any gcc2 internal prototype to avoid an error. */
| #ifdef __cplusplus
| extern "C"
| #endif
| /* We use char because int might match the return type of a gcc2
| builtin and then its argument prototype would still apply. */
| char gdImageDestroy ();
| int
| main ()
| {
| gdImageDestroy ();
| ;
| return 0;
| }
configure:9124: result: no
So, what can you figure out from the messages??
/usr/local/lib/libgd.a(gd.o): In function `HWB_to_RGB':
gd.o(.text+0x4c1): undefined reference to `floor'
/usr/local/lib/libgd.a(gd.o): In function `gdImageCopyResized':
gd.o(.text+0x21e8): undefined reference to `floor'
gd.o(.text+0x22b0): undefined reference to `floor'
Undefined references are bad. Sounds like you're missing the library that
contains the C library mathematical functions, which is often libmath,
i.e. -lmath on the compile line.
You can always test that hypothesis by cutting the program lines out into a
file, stripping the leading | and then re-issuing the gcc line, but
with -lmath (or whatever) added. If that works, then you'll need to adjust
the configureextra/FreeBSD49 file (there are examples in the directory of
other places we've had to add libraries).
What's odd, of course, is that the packaging of the gcc math library would
change from FreeBSD 4.8 to 4.9...
-----Burton
-----Original Message-----
From: [EMAIL PROTECTED] [mailto:[EMAIL PROTECTED] Behalf Of The
Jetman
Sent: Saturday, November 15, 2003 6:48 PM
To: NTOP Mail List
Subject: [Ntop] [NTOP-CVS]CONFIGURE Fails Under FreeBSD 4.9-RELEASE w/
Attachment
Not sure if I have to submit a PR about this, but I've been trying
to build NTOP-2.2+, from the CVS repository. The attached file is the
config.log for the failed run. CONFIGURE fails in spite of the fact
that I run it w/ the proper --with-gd-lib=/usr/local/lib (which is in
the attached file.)
BTW, I invoke CONGIGURE from a one-liner shell script, which I've
successfully used at least a dozen times. In fact, I used the same
script to make NTOP *2.2* (on the same machine), which worked as
expected.
Later....Jet
PS: I submitted another note about a failed CONFIGURE from 2.2c. In
both cases, I've manually verified that the required libs are being
generated.
PPS: This one includes the attachment. Sorry.
=============== From the desk of Jethro Wright, III ================
+ Nothing causes self-delusion quite so readily as power. =
=== jetman516 at hotmail.com ====================== Liu Binyan ===
_______________________________________________
Ntop mailing list
[EMAIL PROTECTED]
http://listgateway.unipi.it/mailman/listinfo/ntop
