| Summary: | Missing "ansi" USE-flag in dev-lisp/gcl-2.6.6 (maxima doesn't work because of it) | ||
|---|---|---|---|
| Product: | Gentoo Linux | Reporter: | Michal Krenek <mikos> |
| Component: | [OLD] Development | Assignee: | Gentoo Lisp Project <lisp> |
| Status: | RESOLVED FIXED | ||
| Severity: | major | CC: | znmeb |
| Priority: | High | ||
| Version: | unspecified | ||
| Hardware: | x86 | ||
| OS: | Linux | ||
| Whiteboard: | |||
| Package list: | Runtime testing required: | --- | |
I have tried to add "ansi" USE-flag to gcl-2.6.6 ebuild on my own. It has compilled fine and if I run gcl, I can see: GCL (GNU Common Lisp) 2.6.6 ANSI Feb 4 2005 20:04:11 So OK, ANSI support seems to work. But when I tried to emerge maxima-5.9.1-r1 (~x86 version) again, it has stopped on same error as before re-emerge of gcl with "ansi" USE-flag. So the problem is probably somewhere else... I am using GCC 3.4. Maybe there can be problem? So this is more likely maxima-related bug. But there is still need for GCL ebuild with "ansi" USE-flag. I ran into this a few minutes ago. Technically, this may be two separate bugs: 1. The "ansi" flag was left off of the gcl-2.6.6 ebuild 2. "maxima" does not build even with an ANSI gcl-2.6.6. I have emerged all of the four Gentoo LISP processors, and they have all been recently updated. I don't really need four LISP processors. The two main applications I need LISP for are algorithmic composition and Maxima. Therefore, I rank them as follows: cmucl -- most stable, mature, most proven compiler, runs Common Music and Maxima, doesn't run on Windows clisp -- runs on Windows for portability, runs Common Music (without GUI), runs Maxima sbcl -- young spin-off of cmucl, runs Common Music with GUI but needs a patch, runs Maxima, doesn't run on Windows gcl -- won't run Common Music at all Here's what I have **working** as of a few minutes ago: clisp-2.33.2-r2 cmucl-19a-r1 maxima-5.9.1-r1 compiled by both the above clisp and cmucl I haven't tried the gcl build; once we get the gcl status straightened out in Portage, I'll attempt it. But I *did* try an SBCL build of Maxima and it failed. I've been messing around with SBCL, though, trying to get the Common Music build to work, so I probably don't have a correct SBCL environment. At the moment, I think Maxima is innocent. Michal, are you building with SBCL too? I just unmerged and emerged sbcl-0.8.19 and attempted a maxima build with *just* SBCL ... it died thusly: Summary: SBCL enabled. Executable name: "sbcl" default lisp: sbcl wish executable name: "wish" Making all in src make[1]: Entering directory `/var/tmp/portage/maxima-5.9.1-r1/work/maxima-5.9.1/src' test -d binary-sbcl || mkdir binary-sbcl test -d binary-sbcl/numerical || mkdir binary-sbcl/numerical test -d binary-sbcl/numerical/slatec || mkdir binary-sbcl/numerical/slatec sbcl --noinform --noprint --eval '(progn (load "../lisp-utils/defsystem.lisp") (funcall (intern "OPERATE-ON-SYSTEM" :mk) "maxima" :compile :verbose t) (sb-ext:quit))' && \ sbcl --noinform --noprint --eval '(progn (load "../lisp-utils/defsystem.lisp") (funcall (intern "OPERATE-ON-SYSTEM" :mk) "maxima" :load :verbose t) (sb-ext:save-lisp-and-die "binary-sbcl/maxima.core" :purify t) (sb-ext:quit))' fatal error encountered in SBCL pid 17721: can't find core file make[1]: *** [binary-sbcl/maxima.core] Error 1 make[1]: Leaving directory `/var/tmp/portage/maxima-5.9.1-r1/work/maxima-5.9.1/src' make: *** [all-recursive] Error 1 !!! ERROR: sci-mathematics/maxima-5.9.1-r1 failed. !!! Function src_compile, Line 62, Exitcode 2 !!! (no error message) !!! If you need support, post the topmost build error, NOT this status message. ------------------------------------------------------------------ It turns out that sbcl-0.8.19 is itself broken; see bug 81546. I can drop back to an earlier version of sbcl and see if it works. I just got the SBCL build of maxima to work. It turns out in bug 81546 that upgrading common-lisp-controller from 3.91 to 4.12 fixes the SBCL issue. Is that relevant here?? Added the ANSI use flag to gcl-2.6.6.ebuild. Thanks for the bug report. I have tried gcl and clisp only, because cmucl and sbcl doesn't support readline. GCL has worked for me in the past, but this new 2.6.6 version doesn't compile Maxima at all. Clisp is working, but I have read that clisp is VERY slow in comparison with cmucl or gcl. So I would be happy if it start working with gcl again... It looks like the gcl bug will get fixed. So you'll be able to do a gcl build. As to the "readline" issue, I suspect the "typical" Maxima (or Common Music, for that matter) user runs Lisp-based applications out of Emacs or XEmacs, rather than straight from a command line. I personally run Maxima from TeXmacs, which does a fantastic job of typesetting the equations. I don't know whether the Maxima ebuild includes the documentation on running from EMacs or TeXmacs; I found it on the Maxima page. Yes, 'missing "ansi" USE-flag' bug is fixed, but it doesn't help. Maxima still doesn't compile with GCL 2.6.6 even if ANSI support is enabled. But this is another issue. I have filled new bug, you can find it here: http://bugs.gentoo.org/show_bug.cgi?id=81869 I am also using TeXmacs, but I want be able to use maxima from command line. Readline support is ideal for it (you can run it from command line without readline support, but it is pain), because I personaly don't like editors like emacs (or vim). |
In my latest "emerge -uD world" portage have updated gcl to new version. But maxima (computer algebra system) dont't work anymore. I have tried to re-emerge maxima, but it gave me this error: Summary: GCL enabled. Executable name: "gcl" default lisp: gcl wish executable name: "wish" Making all in src make[1]: Entering directory `/var/tmp/portage/maxima-5.9.1-r1/work/maxima-5.9.1/src' test -d binary-gcl || mkdir binary-gcl test -d binary-gcl/numerical || mkdir binary-gcl/numerical test -d binary-gcl/numerical/slatec || mkdir binary-gcl/numerical/slatec gcl -batch -eval '(progn (load "../lisp-utils/defsystem.lisp") (funcall (intern "OPERATE-ON-SYSTEM" :mk) "maxima" :compile :verbose t))' && \ gcl -batch -eval '(progn (load "../lisp-utils/defsystem.lisp") (funcall (intern "OPERATE-ON-SYSTEM" :mk) "maxima" :load :verbose t) (when (fbo undp (quote si::sgc-on))(si::sgc-on t)) (si:save-system "binary-gcl/maxima"))' make[1]: *** [binary-gcl/maxima] Error 139 make[1]: Leaving directory `/var/tmp/portage/maxima-5.9.1-r1/work/maxima-5.9.1/src' make: *** [all-recursive] Error 1 !!! ERROR: sci-mathematics/maxima-5.9.1-r1 failed. !!! Function src_compile, Line 62, Exitcode 2 !!! (no error message) !!! If you need support, post the topmost build error, NOT this status message. On start of "emerge maxima", I found this warning: * Important - GCL must be installed with * ANSI support. Otherwise this build will * fail. To check this, start gcl and look * for a header like the following: * GCL (GNU Common Lisp) 2.6.5 ANSI * If you see CtL1 where ANSI is, then you * need to emerge gcl using something like * USE="ansi" emerge gcl But ebuild for new version of gcl don't have "ansi" USE-flag. And really, if I run gcl, I can see: GCL (GNU Common Lisp) 2.6.6 CLtL1 Feb 4 2005 17:43:46 So no ANSI support :-( And ebuilds for older versions of gcl with "ansi" USE-flags were removed... Reproducible: Always Steps to Reproduce: 1. emerge gcl 2. emerge maxima