[slime-devel] Turning off ECL Threading

[Seth Burleigh]

So, ive decided I should probably use ecl lisp. So heres da problem. Examine
this output that produced an error:

 

(:emacs-rex                                        ;;the request

 (swank:connection-info)

 "COMMON-LISP-USER" t 1)

 

;; response – read error given below!!

Debugger entered: nil

  (condition-case error (slime-net-read) (error (debug) (slime-net-close
process t) (error "net-read error: %S" error)))

  slime-net-read-or-lose(#<process SLIME Lisp>)

  slime-process-available-input(#<process SLIME Lisp>)

  slime-net-filter(#<process SLIME Lisp>

 

;;the response which produced above error

 

13(:return (:ok (:pid 5548 :style nil :lisp-implementation (:type \"ECL\"
:name \"ECL\" :version \"9.12.3\")

                                 :machine (:instance \"SE\")) 1)
;;ß---------------Threading issue???

                   TH-PC\" :type \"x86\" :version \"17\")
;;ß----Should be “SETH-PC but its broken up!!!!

 

  :features (:wsock :win32 :formatter :uint32-t :uint16-t
:relative-package-names :unicode :dffi :clos-streams :cmu-format :msvc
:ecl-pde :dlopen   

 :clos :threads :boehm-gc :ansi-cl :common-lisp :ieee-floating-point
:prefixed-api :ffi :pentium4 :common :ecl) 

 

:modules (\"SB-BSD-SOCKETS\" \"SOCKETS\")

:package (:name \"COMMON-LISP-USER\" :prompt \"CL-USER\") :version
\"2009-12-23")

 

Compare this with the pristine clisp output 

 

(:emacs-rex

 (swank:connection-info)

 "COMMON-LISP-USER" t 1)

 

(:indentation-update

 (("with-c-var" . 1)

  ("with-c-place" . 1)

  ("with-foreign-string" . 1)

  ("with-foreign-object" . 1)

  ("with-defining-c-type" . 1)

  ("with-name/options" . 1)

  ("with-window" . 0)

  ("with-output-to-printer" . 1)

  ("define-setf-method" . 2)

  ("letf*" . 1)

  ("letf" . 1)

  ("muffle-cerrors" . 0)

  ("without-global-handlers" . 0)

  ("with-http-output" . 1)

  ("with-collect" . 1)

  ("with-http-input" . 1)

  ("without-package-lock" . 1)

  ("appease-cerrors" . 0)

  ("with-restarts" . 1)

  ("fcase" . 2)

  ...))

 

(:return

 (:ok

  (:pid 5748 :style nil :lisp-implementation

                (:type "CLISP" :name "clisp" :version "2.48 (2009-07-28)
(built on stnt067 [192.168.0.1])")

                :machine

                (:instance "Seth-PC [192.168.2.3]" :type "PC/386" :version
"PC/A86")

                :features

                (:cl-fad cffi-features:x86 cffi-features:windows :cffi :x86
:windows :little-endian :asdf :readline :regexp :syscalls :i18n :loop
:compiler :clos :mop :clisp :ansi-cl :common-lisp :lisp=cl ...)

                :modules

                ("SWANK-ARGLISTS" "SWANK-C-P-C" "ASDF" "readline" "regexp"
"syscalls" "i18n")

                :package

                (:name "COMMON-LISP-USER" :prompt "CL-USER")

                :version "2009-12-23"))

 1)

 

 

 

So
 Recently ecl had made threading a requirement for windows, and maybe for
some other platforms. I don’t know if this was the start of the bug as I
have not yet gone back in ecl versions and seen which one works. The error
on startup DOES NOT occur consistently which is very puzzling. But, anyways,
I would like ecl lisp to act like a ‘single threaded lisp’ as far as slime
is concerned so that I can eliminate threading issues generated by the slime
lisp code.

 

 

-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://mailman.common-lisp.net/pipermail/slime-devel/attachments/20091226/1b466fba/attachment.html>