September 2008 - slime-devel - mailman3.common-lisp.net

[slime-devel] printing large amounts of text at once breaks with SBCL/:fd-handler
by Bart Botta 22 Sep '08

22 Sep '08

Current CVS slime breaks when a large amount of text is printed to the REPL when using :fd-handler communication style on SBCL (tested on 1.0.19 and 1.0.20). For example: (loop for i below 1000 do (format t "~a ......................................................................~%" i)) stops partway through the loop, pressing enter makes it print a bit more, entering a form causes it to evaluate that form and print a bit more of the original output. backtrace from entering (break) while REPL was stopped in middle of output: Restarts: 0: [CONTINUE] Return from BREAK. 1: [RETRY] Retry SLIME REPL evaluation request. 2: [ABORT] Abort sending output to Emacs. 3: [RETRY] Retry SLIME REPL evaluation request. 4: [ABORT] Return to SLIME's top level. 5: [ABORT] Exit debugger, returning to top level. Backtrace: 0: (BREAK "break") 1: (SB-INT:SIMPLE-EVAL-IN-LEXENV (BREAK) #<NULL-LEXENV>) 2: (SWANK::EVAL-REGION "(break)\n") 3: ((LAMBDA ())) 4: (SWANK::TRACK-PACKAGE #<CLOSURE (LAMBDA #) {B1BED2D}>) 5: (SWANK::CALL-WITH-RETRY-RESTART "Retry SLIME REPL evaluation request." #<CLOSURE (LAMBDA #) {B1BEC15.. 6: (SWANK::CALL-WITH-BUFFER-SYNTAX NIL #<CLOSURE (LAMBDA #) {B1BEBFD}>) 7: (SWANK::REPL-EVAL "(break)\n") 8: (SB-INT:SIMPLE-EVAL-IN-LEXENV (SWANK:LISTENER-EVAL "(break)\n") #<NULL-LEXENV>) 9: (SWANK::EVAL-FOR-EMACS (SWANK:LISTENER-EVAL "(break)\n") "COMMON-LISP-USER" 74) 10: (SWANK::PROCESS-REQUESTS T NIL) 11: (SB-IMPL::SUB-SUB-SERVE-EVENT NIL NIL) 12: (SB-IMPL::SUB-SERVE-EVENT NIL NIL NIL) 13: (SB-SYS:WAIT-UNTIL-FD-USABLE 6 :INPUT NIL) 14: (SB-IMPL::REFILL-INPUT-BUFFER #<SB-SYS:FD-STREAM for "a socket" {AC5A119}>) 15: (SB-IMPL::INPUT-CHAR/UTF-8 #<SB-SYS:FD-STREAM for "a socket" {AC5A119}> NIL :EOF) 16: (PEEK-CHAR NIL #<SB-SYS:FD-STREAM for "a socket" {AC5A119}> NIL NIL #<unused argument>) 17: ((LAMBDA (SWANK-BACKEND::STREAMS &OPTIONAL SWANK-BACKEND::TIMEOUT)) (#<SB-SYS:FD-STREAM for "a socke.. 18: (SWANK::WAIT-FOR-EVENT/EVENT-LOOP (:EMACS-PONG 13) NIL) 19: (SWANK::WAIT-FOR-EVENT (:EMACS-PONG 13) NIL) 20: (SWANK::SEND-USER-OUTPUT ..) 21: ((LAMBDA ())) 22: ((LAMBDA ())) 23: ((FLET #:WITHOUT-INTERRUPTS-BODY-[CALL-WITH-RECURSIVE-LOCK]508)) 24: (SB-THREAD::CALL-WITH-RECURSIVE-LOCK ..) 25: ((LAMBDA (SWANK-BACKEND::LOCK FUNCTION)) #S(SB-THREAD:MUTEX :NAME "buffer write lock" :%OWNER #<SB-T.. 26: ((SB-PCL::FAST-METHOD STREAM-FINISH-OUTPUT (SWANK-BACKEND::SLIME-OUTPUT-STREAM)) #(3 NIL 2 NIL 5 NIL.. 27: ((LAMBDA ())) 28: ((FLET #:WITHOUT-INTERRUPTS-BODY-[CALL-WITH-RECURSIVE-LOCK]508)) 29: (SB-THREAD::CALL-WITH-RECURSIVE-LOCK ..) 30: ((LAMBDA (SWANK-BACKEND::LOCK FUNCTION)) #S(SB-THREAD:MUTEX :NAME "buffer write lock" :%OWNER #<SB-T.. 31: ((SB-PCL::FAST-METHOD STREAM-WRITE-STRING (SWANK-BACKEND::SLIME-OUTPUT-STREAM T)) ..) 32: (WRITE-STRING " ......................................................................" #<SWANK-BACK.. 33: (SB-FORMAT::INTERPRET-DIRECTIVE-LIST #<SWANK-BACKEND::SLIME-OUTPUT-STREAM {A6EC099}> (" .............. 34: (SB-FORMAT::%FORMAT #<SWANK-BACKEND::SLIME-OUTPUT-STREAM {A6EC099}> "~a .............................. 35: (FORMAT T "~a ......................................................................~%")[:EXTERNAL] 36: ((LAMBDA ())) 37: (SB-INT:SIMPLE-EVAL-IN-LEXENV (LOOP FOR I BELOW 1000 DO ...) #<NULL-LEXENV>) 38: (SWANK::EVAL-REGION "(loop for i below 1000 do (format t \"~a ........................................ 39: ((LAMBDA ())) 40: (SWANK::TRACK-PACKAGE #<CLOSURE (LAMBDA #) {AFDEC95}>) 41: (SWANK::CALL-WITH-RETRY-RESTART "Retry SLIME REPL evaluation request." #<CLOSURE (LAMBDA #) {AFDEB7D.. 42: (SWANK::CALL-WITH-BUFFER-SYNTAX NIL #<CLOSURE (LAMBDA #) {AFDEB65}>) 43: (SWANK::REPL-EVAL "(loop for i below 1000 do (format t \"~a .......................................... 44: (SB-INT:SIMPLE-EVAL-IN-LEXENV (SWANK:LISTENER-EVAL "(loop for i below 1000 do (format t \"~a ......... 45: (SWANK::EVAL-FOR-EMACS (SWANK:LISTENER-EVAL "(loop for i below 1000 do (format t \"~a ................ 46: (SWANK::PROCESS-REQUESTS T NIL) 47: ((LAMBDA ())) 48: ((LAMBDA (SWANK-BACKEND::HOOK SWANK-BACKEND::FUN)) #<FUNCTION SWANK:SWANK-DEBUGGER-HOOK> #<CLOSURE (.. 49: (SWANK::CALL-WITH-REDIRECTED-IO #<SWANK::CONNECTION {AC5AD59}> #<CLOSURE (LAMBDA #) {AFCBF35}>) 50: (SWANK::CALL-WITH-CONNECTION #<SWANK::CONNECTION {AC5AD59}> #<CLOSURE (LAMBDA #) {AFCBF25}>) 51: (SWANK::HANDLE-REQUESTS #<SWANK::CONNECTION {AC5AD59}> T NIL) 52: (SB-IMPL::SUB-SUB-SERVE-EVENT NIL NIL) 53: (SB-IMPL::SUB-SERVE-EVENT NIL NIL NIL) 54: (SB-SYS:WAIT-UNTIL-FD-USABLE 0 :INPUT NIL) 55: (SB-IMPL::REFILL-INPUT-BUFFER #<SB-SYS:FD-STREAM for "standard input" {A6E57F1}>) 56: (SB-IMPL::INPUT-CHAR/UTF-8 #<SB-SYS:FD-STREAM for "standard input" {A6E57F1}> NIL #:EOF-OBJECT) 57: (READ-CHAR #<SB-SYS:FD-STREAM for "standard input" {A6E57F1}> NIL #:EOF-OBJECT #<unused argument>) 58: (READ-CHAR #<SYNONYM-STREAM :SYMBOL SB-SYS:*STDIN* {90DA579}> NIL #:EOF-OBJECT #<unused argument>) 59: (READ-PRESERVING-WHITESPACE #<SYNONYM-STREAM :SYMBOL SB-SYS:*STDIN* {90DA579}> NIL (NIL) T) 60: (READ-PRESERVING-WHITESPACE #<SYNONYM-STREAM :SYMBOL SB-SYS:*STDIN* {90DA579}> NIL (NIL) NIL) 61: (READ #<SYNONYM-STREAM :SYMBOL SB-SYS:*STDIN* {90DA579}> NIL (NIL) NIL) 62: (SB-IMPL::REPL-READ-FORM-FUN #<SYNONYM-STREAM :SYMBOL SB-SYS:*STDIN* {90DA579}> #<unavailable argume.. 63: (SB-IMPL::REPL-FUN NIL) 64: (SB-IMPL::REPL-FUN NIL)[:EXTERNAL] 65: ((LAMBDA ())) 66: ((LAMBDA ()))[:EXTERNAL] 67: (SB-IMPL::%WITH-REBOUND-IO-SYNTAX #<CLOSURE (LAMBDA #) {A6E8AD5}>) 68: (SB-IMPL::TOPLEVEL-REPL NIL) 69: (SB-IMPL::TOPLEVEL-INIT) 70: ((LABELS SB-IMPL::RESTART-LISP))

2 1

[slime-devel] Truncated error messages in the debugger
by Zach 22 Sep '08

22 Sep '08

This is a minor inconvenience since I can inspect the condition and find out what is really going on, but, just thought I would point it out. From cvs slime (a little out of date). Anyone else get this behavior? (defclass test-class () ((a :initarg :a) (b :initarg :b)) ) (defun func1 (lst) (declare (type list lst)) lst ) (defmethod print-object ((obj test-class) str) (format str "#<This is an objects special PRINT-OBJECT form ~A ~A>" (slot-value obj 'a) (slot-value obj 'b) )) ;; This causes an error (func1 expects a list) (func1 (make-array 20 :initial-contents (loop for i below 20 collect (make-instance 'test-class :a 1 :b 2) ))) In the debugger I get this (the error message here has lost the "is not of type LIST" part): The value #(#<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> ..) [Condition of type TYPE-ERROR] Restarts: 0: [ABORT] Return to SLIME's top level. 1: [TERMINATE-THREAD] Terminate this thread (#<THREAD "repl-thread" {1425AA89}>) Backtrace: 0: (FUNC1 #(#<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> #<This is an objects special PRINT-OBJECT form 1 2> ...))[:EXTERNAL] 1: (SB-INT:SIMPLE-EVAL-IN-LEXENV (FUNC1 (MAKE-ARRAY 20 :INITIAL-CONTENTS (LOOP FOR I BELOW 20 COLLECT ...))) #<NULL-LEXENV>) 2: (SWANK::EVAL-REGION "(func1 (make-array 20 :initial-contents (loop for i below 20 collect (make-instance 'test-class :a 1 :b 2)))) ")

2 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 22 Sep '08

22 Sep '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1533 slime/ChangeLog:1.1536 --- slime/ChangeLog:1.1533 Sat Sep 20 17:46:30 2008 +++ slime/ChangeLog Sun Sep 21 07:18:00 2008 @@ -1,3 +1,24 @@ +2008-09-21 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-eol-conversion-fixup): Return 0 (not nil) for + anyting other than CRLF conventions. + +2008-09-21 Helmut Eller <heller(a)common-lisp.net> + + * swank-openmcl.lisp: Try to remove the first few internal frames + from backtraces. + (guess-stack-top): New function. + (call-with-debugging-environment): Use it + + (frame-arguments): Return a list instead of a string. Don't quote + symbols. + (source-locations): Recognize (:internal FOO) functions. + +2008-09-21 Helmut Eller <heller(a)common-lisp.net> + + * swank.lisp (*backtrace-pprint-dispatch-table*): + Honor *print-escape*. + 2008-09-20 Ariel Badichi <abadichi(a)bezeqint.net> * slime.el (slime-with-output-end-mark):

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 21 Sep '08

21 Sep '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1527 slime/ChangeLog:1.1533 --- slime/ChangeLog:1.1527 Fri Sep 19 07:20:22 2008 +++ slime/ChangeLog Sat Sep 20 17:46:30 2008 @@ -1,3 +1,69 @@ +2008-09-20 Ariel Badichi <abadichi(a)bezeqint.net> + + * slime.el (slime-with-output-end-mark): + slime-repl-show-maximum-output no longer accepts any arguments. + +2008-09-20 Helmut Eller <heller(a)common-lisp.net> + + * test.sh: Parse the -S option as advertized. + + * slime.el (slime-randomize-test-order): Add the * to the + docstring. + ([def-slime-test] break2): Also CCL is expected to fail here. + +2008-09-20 Helmut Eller <heller(a)common-lisp.net> + + Fix some of the bugs that I introduced with the last commits. + + * swank-openmcl.lisp (call-with-debugging-environment): Fix typo. + (call-with-debugger-hook): Bind *break-in-sldb*. + (backtrace-context): Return nil, not tcr! + (map-backtrace): Remove the stack< test. Only test for nil. + (lisp-implementation-type-name): Return "ccl". + + (emacs-inspect (t)): Fix typo. + + (kill-thread): Use join-process. Otherwise we get strange + "process-reset" errors when disconnecting. + (thread-alive-p): Implemented with ccl::process-exhausted-p. + + (source-locations): Use labels for helper functions. + (function-source-location): Now implemented on top of + source-locations. + +2008-09-20 Helmut Eller <heller(a)common-lisp.net> + + Fix frame-source-location-for-emacs for CCL. + + * swank-openmcl.lisp (source-locations): New function. + (create-source-location): New function. + (frame-source-location-for-emacs): Use it. + +2008-09-20 Helmut Eller <heller(a)common-lisp.net> + Fix inspecting of arrays. + + * swank-openmcl.lisp (emacs-inspect :around (t)): call-next-method + may return a lazy list. Detect that case and only append to + ordinary lists. + (emacs-inspect (t)): Don't mark labels as inspectable. Just print + them. + +2008-09-20 Helmut Eller <heller(a)common-lisp.net> + + Fix BREAK and backtraces after interrupts. + + * swank-openmcl.lisp (*sldb-stack-top-hint*): New variable. + (call-with-debugging-environment, break-in-sldb) + (interrupt-thread): Use it. + (*process-to-stack-top*, record-stack-top) + (grab-stack-top): Deleted. Use *sldb-stack-top-hint* instead. + (backtrace-context): Deleted. Use %current-tcr directly. + +2008-09-20 Helmut Eller <heller(a)common-lisp.net> + + * swank-openmcl.lisp (call-with-debugging-environment): Don't set + *debugger-hook* to nil. + 2008-09-19 Helmut Eller <heller(a)common-lisp.net> * slime.el ([def-slime-test] break): Split it up in two versions

1 0

[slime-devel] slime-with-output-end-mark
by Stas Boukarev 20 Sep '08

20 Sep '08

`slime-with-output-end-mark' (used only in contrib/slime-presentations.el) has wrong argument count for the function `slime-repl-show-maximum-output', which causes an error. -- With Best Regards, Stas.

1 0

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 20 Sep '08

20 Sep '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1522 slime/ChangeLog:1.1527 --- slime/ChangeLog:1.1522 Thu Sep 18 18:35:46 2008 +++ slime/ChangeLog Fri Sep 19 07:20:22 2008 @@ -1,10 +1,27 @@ 2008-09-19 Helmut Eller <heller(a)common-lisp.net> + * slime.el ([def-slime-test] break): Split it up in two versions + to make the debugger-hook issue more explicit. + + * swank-allegro.lisp (frob-allegro-field-def): There seems to be a + new type :func handle it like :lisp. + +2008-09-19 Helmut Eller <heller(a)common-lisp.net> + * slime.el (slime-save-marker): New marcro. Use it in combination with insert-before-markers. (slime-check-buffer-contents): Use {} resp. [] to describe the position of output resp. input markers. + (slime-repl-emit-result): Update window-point. + + (slime-randomize-test-order): New variable. + (slime-shuffle-list): New function. + (slime-run-tests): Use it. + (slime-batch-test): Accept test-name and randomize arguments. + + * test.sh (usage): Accept -n and -S options. + 2008-09-18 Tobias C. Rittweiler <tcr(a)freebits.de> * swank-ecl.lisp: Forgot to update ECL's backend when introducing

1 0

[slime-devel] Slime, Clozure CL and conditions
by Stas Boukarev 19 Sep '08

19 Sep '08

I have troubles with recent Slime and CCL 1.2 with respect to conditions handling: When I have an error and end up in SLDB, and evaluate through `slime-interactive-eval' (`:' key-binding) an expression which throws another error, REPL becomes broken. -- With Best Regards, Stas.

1 0

[slime-devel] Extraneous argument passed to SLIME-REPL-SHOW-MAXIMUM-OUTPUT
by Ariel Badichi 19 Sep '08

19 Sep '08

Hello, SLIME-REPL-SHOW-MAXIMUM-OUTPUT no longer accepts any argument, but in some cases it is still passed one. The following patch eliminates these cases. Ariel diff --git a/slime.el b/slime.el index 40cdb7e..f53f1a8 100644 --- a/slime.el +++ b/slime.el @@ -2572,8 +2572,7 @@ update window-point afterwards. If point is initially not at (goto-char slime-output-end) (funcall body..)))) (when updatep.. - (slime-repl-show-maximum-output - (> (- slime-output-end slime-output-start) 1000))))) + (slime-repl-show-maximum-output)))) (defun slime-output-filter (process string) (with-current-buffer (process-buffer process)

1 0

[slime-devel] Problems with cvs slime
by Alexandre Passos 19 Sep '08

19 Sep '08

Hello, Until recently I was using slime 2.0 without any issues. This week I decided to upgrade to the cvs version. I use ubuntu Hardy Heron, sbcl 1.0.11.debian and emacs23 (ubuntu's cvs snapshot). After a small time of usage (sometimes not enough to eval a single expression in the repl, sometimes longer) slime stops responding (for example, hanging emacs when I ask for symbol completion). In the *inferior-lisp* buffer I see something like ;; Swank started at port: 60259. 60259 * fatal error encountered in SBCL pid 12153(tid 3085153968): maximum interrupt nesting depth (32) exceeded Welcome to LDB, a low-level debugger for the Lisp runtime environment. ldb> And from then all I manage to do is quit. Is there any way out, or should I just downgrade? Thanks -- - Alexandre

2 1

[slime-devel] Daily ChangeLog diff
by mbaringer＠common-lisp.net 19 Sep '08

19 Sep '08

Index: slime/ChangeLog diff -u slime/ChangeLog:1.1517 slime/ChangeLog:1.1522 --- slime/ChangeLog:1.1517 Wed Sep 17 19:14:45 2008 +++ slime/ChangeLog Thu Sep 18 18:35:46 2008 @@ -1,3 +1,50 @@ +2008-09-19 Helmut Eller <heller(a)common-lisp.net> + + * slime.el (slime-save-marker): New marcro. Use it in combination + with insert-before-markers. + (slime-check-buffer-contents): Use {} resp. [] to describe the + position of output resp. input markers. + +2008-09-18 Tobias C. Rittweiler <tcr(a)freebits.de> + + * swank-ecl.lisp: Forgot to update ECL's backend when introducing + swank-frames in commit on 2008-09-12. + +2008-09-18 Helmut Eller <heller(a)common-lisp.net> + + Some cleanups for the REPL code. + + * slime.el (slime-show-last-output) + (slime-show-last-output-function) + (slime-show-last-output-region) + (slime-maybe-display-output-buffer) + (slime-repl-last-input-start-mark): Delete unused code. + + (slime-repl-emit-result, slime-repl-insert-prompt) + (slime-repl-show-abort, slime-repl-insert-result) + (slime-insert-transcript-delimiter) + (slime-eval-with-transcript-cont): Consistently use save-excursion + and insert-before-markers. We always want to preserve the cursor + position in the input region (for type-ahead). + (slime-eval-with-transcript): Change order of arguments to + make the common case easier to use. + + (slime-batch-test): Use a timer. + (slime-check-buffer-contents): New function to test contents and + current position. Use it in various places. + + (sldb-recenter-region, def-slime-test interactive-eval) + (def-slime-test interactive-eval-output): Act slightly differently + when the test suite is executed in batch mode (without terminal). + + * swank.lisp (handle-requests): Flush output. + (interactive-eval, interactive-eval): Don't use fresh-line, as + that makes it harder to test the REPL code. + + * test.sh (usage): Add a -T switch to run slime in the current + directory without copying (and compiling) everything to a + temporary directory. + 2008-09-18 Helmut Eller <heller(a)common-lisp.net> * slime.el (slime-eval-with-transcript): Accept some more arguments

1 0