On Tue, 28 Aug 2007, Alexander K. Hansen wrote:
> Viv Kendon wrote:
>> On Tue, 28 Aug 2007, Alexander K. Hansen wrote:
>>
>>> Viv Kendon wrote:
>>>> I'm getting a really verbose failure trying to build maxima:
>>>>
>>>> Setting up fink-buildlock-maxima-5.12.0-4 (2007.08.28-13.12.56) ...
>>>> gzip -dc /sw/src/maxima-5.12.0.tar.gz | /sw/bin/tar -xf - --no-same-owner
>>>> --no-
>>>> same-permissions
>>>> patch -p1 </sw/fink/dists/unstable/crypto/finkinfo/maxima.patch
>>>> patching file interfaces/xmaxima/doc/xmaxima.texi
>>>> /var/tmp/tmp.2.sWOaKM
>>>> #!/bin/sh -ev
>>>> export LANG=C LC_ALL=C
>>>> ./configure --prefix=/sw --infodir='${prefix}/share/info'
>>>> --mandir='${prefix}/s
>>>> hare/man' --libexecdir='${prefix}/lib' --enable-recode --enable-clisp
>>>> checking for a BSD-compatible install... /sw/bin/ginstall -c
>>>> checking whether build environment is sane... yes
>>>> checking for gawk... gawk
>>>> checking whether make sets $(MAKE)... yes
>>>> checking build system type... powerpc-apple-darwin8.10.0
>>>> checking host system type... powerpc-apple-darwin8.10.0
>>>> checking for clisp... true
>>>> checking for iconv... true
>>>> checking for recode... true
>>>>
>>>> *** - handle_fault error2 ! address = 0xbf7ff760 not in
>>>> [0x64303ef0,0x66a00000)
>>>> !
>>>> SIGSEGV cannot be cured. Fault address = 0xbf7ff760.
>>>> Permanently allocated: 88480 bytes.
>>>> Currently in use: 329253756 bytes.
>>>> Free space: 23961886 bytes.
>>>> ./configure: line 3141: test: too many arguments
>>>> configure: error: clisp runtime "
>>>> *** - Invalid direction :INPUT for accessing #P/dev/fd/0
>>>> The following restarts are available:
>>>> ABORT :R1 ABORT
>>>> ABORT :R2 ABORT
>>>>
>>>> *** - UNIX error 19 (ENODEV): No such device
>>>> The following restarts are available:
>>>> ABORT :R1 ABORT
>>>> ABORT :R2 ABORT
>>>> ABORT :R3 ABORT
>>>>
>>>> *** - UNIX error 19 (ENODEV): No such device
>>>> The following restarts are available:
>>>> ABORT :R1 ABORT
>>>> ABORT :R2 ABORT
>>>> ABORT :R3 ABORT
>>>> ABORT :R4 ABORT
>>>>
>>>> [huge number of similar sequences snipped until...]
>>>>
>>>> ABORT :R1541 ABORT
>>>>
>>>> *** - " is not an executable
>>>> ### execution of /var/tmp/tmp.2.sWOaKM failed, exit code 1
>>>> Removing runtime build-lock...
>>>>
>>>> This is on PPC 10.4.10 Server, case sensitive file system and
>>>> powerpc-apple-darwin8-gcc-4.0.1 (GCC) 4.0.1 (Apple Computer, Inc. build
>>>> 5367). Up to date unstable except for the things behind maxima in the
>>>> update-all queue. I tried twice and got the same exact messages both
>>>> times (in between successfully compiling fontforge).
>>>>
>>>> I suspect it is an unhappy interaction with something specific to my
>>>> system, but I have no idea what, I've never seen anything like it before.
>>>> Help much appreciated...
>>>>
>>>> -- Viv
>>>>
>>> Hmm...very mysterious indeed. I've committed the new upstream version
>>> (5.13), so try that. Also, make sure that you don't have any non-Fink
>>> instances of the dependencies, e.g. via checking the config.log.
>>>
>>> It has worked for me on PowerPC 10.4.10 Client, with the same compiler and
>>> a case-sensitive (HFS-X) filesystem. On that note--are you on HFS
>>> case-sensitive or UFS?
>>>
>>> --
>>> Alexander K. Hansen
>>> Fink User Liaison/Documenter
>>> akh AT finkproject DOT org
>>
>> HFS-X case sensitive system...and nothing non-fink shows in
>> config.log...note, though that fink list clisp gives
>>
>> i clisp 2.41-1 ANSI Common Lisp implementation
>> clisp-maxima 2.29-1014 Older version of CLISP compatible with
>> Ma...
>>
> clisp-maxima can probably be pulled, now that Maxima seems happy with current
> CLISP versions.
>> However, that's probably not relevant, the bizzare error is more likely to
>> be from sed:
>>
>> i sed 4.1.5-11 The stream editor, GNU version
>>
>> where again most likely that is the fink version being used.
>>
>> maxima-5.13 gives the same behaviour, but this time I caught sed and
>> something called lisp.run going berserk in the process table (from "top"
>> output):
>>
>> 19326 gsed 6.0% 0:07.06 1 14 25 160K 932K 648K 29.7M
>> 19325 lisp.run 90.6% 0:42.54 2 17 9482 219M+ 2.70M 220M+
>> 249M+
>>
>> followed by an incantation of "sort" to finally produce that bizzare
>> output...which has somehow been placed in $CLISP_RUNTIME_PATH by the
>> command
>>
>> CLISP_RUNTIME_PATH=`${CLISP_NAME} -norc -q -x "(format nil \"~abase/
>> ${CLISP_RUNTIME}\" (namestring *lib-directory*))"|sed
>> 's/\"\\(.*\\)\"/\\1/'`
>>
>> (I think, if I'm reading the configure script right).
>>
>> Still baffled here...
>
> Check that your gsed is indeed Fink's via "which gsed". I built
> maxima-nox-5.13-1 this morning with fink's sed.
It is...I don't have any other gsed on this system and
/sw/bin/sed points to gsed (via /sw/etc/alternatives/sed).
I rebuilt sed but that didn't help.
OK, got it, I think...I checked for lisp.run and got
prompt> locate lisp.run
/sw/lib/clisp/base/lisp.run
/sw/lib/clisp/full/lisp.run
/sw/lib/maxima/5.12.0/binary-clisp/lisp.run
They appear to be at least similar:
4 lrwxr-xr-x 1 root admin 16 Nov 13 2006
/sw/lib/clisp/base/lisp.run -> ../full/lisp.run*
4688 -rwxr-xr-x 1 root admin 4797864 Nov 13 2006
/sw/lib/clisp/full/lisp.run*
4688 -rwxr-xr-x 1 root admin 4797864 Jul 4 22:30
/sw/lib/maxima/5.12.0/binary-clisp/lisp.run*
but it doesn't hurt to try limiting the possibilities...
So I did fink remove maxima then tried again, and it appears
to have compiled OK.
I still have the old deb of maxima-5.12 if you are curious,
but I think I'll just go with being happy it now works.
-- Viv
------------------------------------------------
Dr Viv Kendon [EMAIL PROTECTED]
Quantum Information tel: 0113 343 3858
Physics & Astronomy University of Leeds
http://quantum.leeds.ac.uk/~viv
-------------------------------------------------------------------------
This SF.net email is sponsored by: Splunk Inc.
Still grepping through log files to find problems? Stop.
Now Search log events and configuration files using AJAX and a browser.
Download your FREE copy of Splunk now >> http://get.splunk.com/
_______________________________________________
Fink-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/fink-users