| Summary: | On amd64, sci-mathematics/maxima - maxima cannot be started from wxmaxima or cantor, on ~amd, it works. | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Benjamin Schulz <schulz.benjamin> |
| Component: | New packages | Assignee: | Gentoo Science Mathematics related packages <sci-mathematics> |
| Status: | RESOLVED FIXED | ||
| Severity: | normal | ||
| Priority: | Normal | ||
| Version: | unspecified | ||
| Hardware: | All | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
|
Description
Benjamin Schulz
2013-01-27 13:02:30 UTC
i looked at cantor again. It seems that cantor could start maxima, even if it said it didn't. However, there was surely some crash, as maximas output was: cantor(19294)/kdecore (KLibrary) findLibraryInternal: plugins should not have a 'lib' prefix: "libcantorpart.so" Object::disconnect: Unexpected null parameter cantor(19294)/kdeui (kdelibs): Attempt to use QAction "" with KXMLGUIFactory! Maxima 5.29.1 http://maxima.sourceforge.net using Lisp SBCL 1.0.55 Distributed under the GNU Public License. See the file COPYING. Dedicated to the memory of William Schelter. The function bug_report() provides bug reporting information. Maxima 5.29.1 http://maxima.sourceforge.net using Lisp SBCL 1.0.55 Distributed under the GNU Public License. See the file COPYING. Dedicated to the memory of William Schelter. The function bug_report() provides bug reporting information. (%i1) (%i1) the window of cantor, however, freezes, saying it wants to initialize the session... now there was additional console output: The function bug_report() provides bug reporting information. (%i1) ; in: PROGN (SETUP-SERVER 0) ; (MAXIMA::SETUP-SERVER 0) ; ; caught STYLE-WARNING: ; undefined function: SETUP-SERVER ; ; compilation unit finished ; Undefined function: ; SETUP-SERVER ; caught 1 STYLE-WARNING condition Maxima encountered a Lisp error: The function MAXIMA::SETUP-SERVER is undefined. Automatically continuing. To enable the Lisp debugger set *debugger-hook* to nil. (%i1) Maxima 5.29.1 http://maxima.sourceforge.net using Lisp SBCL 1.0.55 Distributed under the GNU Public License. See the file COPYING. Dedicated to the memory of William Schelter. The function bug_report() provides bug reporting information. (%i1) that being said, i tried it with clisp, sbcl, and ecls. I also disabled each useflag one for one... The current stable version in portage can only compile with clisp enabled and sbcl and ecls disabled. After reverting to the stable version of maxima, i have now this error with the stable version too.... I don't know what is the cause for that... On my ~amd64 box, wxmaxima successfully starts maxima-5.29.1 with any of 5 lisps: sbcl-1.1.3, gcl-2.6.8_pre20120805, clozurecl-1.8, ecls-12.12.1, clisp-2.49-r3. On ~x86 box, also with cmucl-20c (it's not available on amd64). I cannot reproduce your problem. I don't know how cantor interacts with maxima. You'd better ask the maxima mailing list, stating the exact versions of maxima and cantor. Maybe, there are people who will be able to help with this problem. I got this behavior since an emerge sync and an "emerge -uD world" two weeks ago.... before that, wxmaxima and maxima worked well. The fact that, when I downgrade wxmaxima and maxima to stable versions, the problem persists (even if these maxima and wxmaxima versions worked before) indecates that the problem is because of some updated library that came in by "emerge -uD world"... I will now look at my lisp versions once Im at my laptop again, and see whether downgrading or upgrading them helps... by the way, the stable version of maxima refuses to compile on my system. So it maybe on the lisp end.... I compiled lisp and there we have it:
piling (DEFUN DISABLE-CALL-COUNTING ...)
; compiling (DEFUN ENABLE-CALL-COUNTING-FOR-FUNCTION ...)
; compiling (DEFUN TEST-0 ...)
; compiling (DEFUN TEST ...)
; compiling (PROVIDE (QUOTE SB-SPROF))
; SYS:CONTRIB;SB-SPROF;SB-SPROF.FASL.NEWEST written
; compilation finished in 0:00:01.210
* true
make: Leaving directory `/var/tmp/portage/dev-lisp/sbcl-1.0.55-r1/work/sbcl-1.0.55/contrib/sb-sprof'
WARNING! Some of the contrib modules did not build successfully or pass
their self-tests. Failed contribs:"
sb-bsd-sockets
* ERROR: dev-lisp/sbcl-1.0.55-r1 failed (compile phase):
* make failed
*
* Call stack:
* ebuild.sh, line 93: Called src_compile
* environment, line 2448: Called die
* The specific snippet of code:
* env - HOME="${T}" CC="$(tc-getCC)" AS="$(tc-getAS)" LD="$(tc-getLD)" CPPFLAGS="${CPPFLAGS}" CFLAGS="${CFLAGS}" ASFLAGS="${ASFLAGS}" LDFLAGS="${LDFLAGS}" GNUMAKE=make ./make.sh "sh ${bindir}/run-sbcl.sh --no-sysinit --no-userinit --disable-debugger" || die "make failed";
*
* If you need support, post the output of `emerge --info '=dev-lisp/sbcl-1.0.55-r1'`,
* the complete build log and the output of `emerge -pqv '=dev-lisp/sbcl-1.0.55-r1'`.
* The complete build log is located at '/var/tmp/portage/dev-lisp/sbcl-1.0.55-r1/temp/build.log'.
* The ebuild environment file is located at '/var/tmp/portage/dev-lisp/sbcl-1.0.55-r1/temp/environment'.
* Working directory: '/var/tmp/portage/dev-lisp/sbcl-1.0.55-r1/work/sbcl-1.0.55'
* S: '/var/tmp/portage/dev-lisp/sbcl-1.0.55-r1/work/sbcl-1.0.55'
A pity. Unfortunately, the common-lisp team in Gentoo is < 1 person, and bugs are unfixed for very long periods. I tried to push some modern stuff into the tree, in particular, sbcl-1.1.x. For me, sbcl-1.1.3 works very well - it compiles maxima and fricas, and is the fastest common lisp. But these new ebuilds break something else, and therefore are hard masked. Unless we find somebody who will join the common-lisp team, the situation will remain hopeless. This doesn't sound like a bug in maxima, and perhaps not a bug at all if you haven't recompiled related packages like cantor after changing maxima version. Could you please update the title to specify the bug correctly, or consider closing it if there is no clearly defined bug. hi, of course I have recompiled cantor and wxmaxima with different versions a dozen times or so.... They seem to call maxima through the same client-server mechanism, and it seems that they can not establish a connection to the maxima server, because of some problem with a library that must be somewhere in my system. But I do not know which library this is. I will try emerge -UD world again soon, and see whether this is fixed. Perhaps I should say that I believe this is linked to https://bugs.gentoo.org/show_bug.cgi?id=455268 always during compilation of dev-lisp/sbcl, self tests are runned by the ebuild. Build is cancelled with "Some of the contrib modules did not build successfully or pass their self-tests. Failed contribs:" sb-bsd-sockets" and exactly these sockets module is responsible for the mechanism through which wxmaxima or cantor are contacting maxima. The lisp librarys other than sbcl compile fine, however, they do not run self tests. The malfunction seems to become manifest, only if I start a probram with a client server mechanism through a socket module, like wxmaxima or cantor, I may also add that I do not have anything like selinux or something enabled, which could initiate this problem. It seems to be due to some library which lisp depends on. I did now an entire reinstall using ~amd64. Now on unstable, maxima and wxmaxima works like a charm. I still do not know thich library did prevent the communication between lisp clients and servers on my stable system. At least ~amd64 seems to work. |
