|
From: <don...@is...> - 2008-10-12 19:32:37
|
Sam Steingold writes: > > * Don Cohen <qba...@vf...> [2008-10-11 21:30:00 -0700]: > > > > I don't see any such thing as dbus. > > [2008-10-11 21:17:57 root@don-eve ~] > > $ find /tmp/clisp/ -name dbus > > [2008-10-11 21:19:30 root@don-eve ~] > > http://clisp.cvs.sourceforge.net/viewvc/clisp/clisp/modules/dbus/ Does this mean that cvs update from the clisp directory is not adequate? Am I supposed to do another checkout? Where should I be in order to do what checkout? > > > please reconfigure with "--enable-maintainer-mode=no". > > I guess you mean this: > > ./makemake: maintainer-mode=no: invalid feature name > > I guess I meant --disable-maintainer-mode [2008-10-12 12:08:14 root@don-eve /tmp/clisp] $ ./configure --disable-maintainer-mode=no build-dir executing /tmp/clisp/build-dir/configure --srcdir=../src --disable-option-checkin g --disable-maintainer-mode=no --cache-file=config.cache configure: error: invalid feature name: maintainer-mode=no [2008-10-12 12:08:47 root@don-eve /tmp/clisp] I guess that's not what you meant. Next try looks better. $ ./configure --disable-maintainer-mode build-dir executing /tmp/clisp/build-dir/configure --srcdir=../src --disable-option-checking --disable-maintainer-mode --cache-file=config.cache configure: loading cache config.cache checking for a BSD-compatible install... (cached) /usr/bin/install -c checking whether build environment is sane... yes ... config.status: creating po/Makefile Configure findings: FFI: no (user requested: default) Is this what you wanted? I'd have expected and preferred the opposite. readline: yes (user requested: default) libsigsegv: yes ./makemake --disable-maintainer-mode --srcdir=../src > Makefile make: `config.lisp' is up to date. To continue building CLISP, the following commands are recommended (cf. unix/INSTALL step 4 ff): cd build-dir vi config.lisp # The default stack size on your platform is insufficient # and must be increased to at least 16384. You must do either # 'ulimit -s 16384' (for Bourne shell derivatives, e.g., bash and zsh) That message is also unfamiliar to me. $ ulimit -s 10240 # or 'limit stacksize 16384' (for C shell derivarives, e.g., tcsh) make make check [2008-10-12 12:09:55 root@don-eve /tmp/clisp] $ make make: *** No targets specified and no makefile found. Stop. [2008-10-12 12:09:58 root@don-eve /tmp/clisp] $ oops $ cd build-dir/ [2008-10-12 12:14:49 root@don-eve /tmp/clisp/build-dir] $ make mkdir -p gllib ... rm -rf full CLISP_LINKKIT=. MAKE=make ./clisp-link add-module-sets base full || (rm -rf full; exit 1) lisp.run lispinit.mem modules.h modules.o makevars lisp.a libnoreadline.a width.o uniname.o localcharset.o gettext.o calls.o regexi.o regex.o regexp.dvi [2008-10-12 12:17:20 root@don-eve /tmp/clisp/build-dir] $ make check mkdir -p gllib if test ! -f gllib/Makefile; then sh config.status --file=gllib/Makefile depfiles; fi ... 640 IDs "http://clisp.podval.org/impnotes/" Bye. [2008-10-12 12:25:08 root@don-eve /tmp/clisp/build-dir] So I guess I'm now back on track. [2008-10-12 12:26:05 root@don-eve /tmp/clisp/build-dir] $ /etc/cron.daily/clisp-ap5 cvs update: Updating . cvs update: Updating benchmarks ... 212 Bye. -rwxr-xr-x 1 root root 12283170 Oct 12 12:27 /tmp/ap5-2.46++ -rwxr-xr-x 1 root root 12495666 Sep 26 06:17 /tmp/ap5-2.46+ -rwxr-xr-x 1 root root 12473843 Jul 14 06:19 /tmp/ap5-2.46 -rwxr-xr-x 1 root root 12466816 Jul 2 06:19 /tmp/ap5-2.45++ [2008-10-12 12:27:53 root@don-eve /tmp/clisp/build-dir] That looks plausible. Interesting - the latest build is a little smaller than the previous successful version: -rwxr-xr-x 1 root root 12495684 Oct 10 06:16 /tmp/ap5-2.46++ Is there some way to find out what options were passed to the original configure command? Was I supposed to have used --disable-maintainer-mode in the original configure or do you think I did something wrong in some subsequent configure (unlikely, since this nightly build was working before), or was there some change in cvs (seems more likely) that made this necessary? If I were to do a new checkout of clisp, what would be the correct configure arguments? I'd hope that ./configure build-dir would normally be the right thing, and would result in readline, FFI and libsigsegv all being "yes", at least if the required libraries exist. |