- cmucl-ticket - mailman3.common-lisp.net

[cmucl] #85: Hemlock fails to start
by cmucl 09 Jun '13

09 Jun '13

#85: Hemlock fails to start ---------------------+------------------------------------------------------ Reporter: rtoy | Owner: rtoy Type: defect | Status: new Priority: minor | Milestone: Component: Hemlock | Version: 2013-06 Keywords: | ---------------------+------------------------------------------------------ On some systems, {{{ (require :hemlock) (ed) }}} causes an X error. This is due to the fact that Hemlock wants to use Courier fonts which are not available on the system. The workaround is to install Courier fonts on the system. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/85> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 0

[cmucl] #84: motifd is a 64-bit app
by cmucl 09 Jun '13

09 Jun '13

#84: motifd is a 64-bit app --------------------+------------------------------------------------------- Reporter: rtoy | Owner: rtoy Type: defect | Status: new Priority: minor | Milestone: Component: CLM | Version: 2013-06 Keywords: | --------------------+------------------------------------------------------- {{{motifd}}} is a 64-bit application, unlike {{{lisp}}} which is a 32-bit app. This is only a problem if you install on a 32-bit OS where {{{motifd}}} won't run at all. But making {{{motifd}}} a 32-bit app may cause problems on a 64-bit OS where the necessary 32-bit X libraries are not installed. In addition, I (rtoy) have difficulties building {{{motifd}}} as a 32-bit app on linux. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/84> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 0

[cmucl] #82: Source location error on large files
by cmucl 21 May '13

21 May '13

#82: Source location error on large files --------------------+------------------------------------------------------- Reporter: rtoy | Owner: somebody Type: defect | Status: new Priority: major | Milestone: Component: Core | Version: 2013-05 Keywords: | --------------------+------------------------------------------------------- As noted in the [http://article.gmane.org/gmane.lisp.cmucl.general/6497 email from Matt Kaufmann], there is an error in computing the source location. The problem was that the file was very large with many long forms. This caused the form number to overflow the 14-bit number used to hold the form number. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/82> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 1

[cmucl] #81: Reversing a string is slow
by cmucl 19 May '13

19 May '13

#81: Reversing a string is slow --------------------+------------------------------------------------------- Reporter: rtoy | Owner: somebody Type: defect | Status: new Priority: major | Milestone: Component: Core | Version: 2013-05 Keywords: | --------------------+------------------------------------------------------- Consider this: {{{ (defparameter *s* (make-string 1000000)) (defun time-rev (s) (dotimes (k 100 nil) (declare (fixnum k)) (setf s (reverse s))) s) (compile 'time-rev) (time (prog1 t (time-rev *s*))) ; Evaluation took: ; 10.23 seconds of real time ; 10.151641 seconds of user run time ; 0.079628 seconds of system run time ; 31,310,669,055 CPU cycles ; [Run times include 0.08 seconds GC run time] ; 0 page faults and ; 200,114,224 bytes consed. }}} Since strings are basically arrays of unsigned 16-bit integers, compare the time when reversing an array: {{{ (defparameter *v* (make-array 1000000 :element-type '(unsigned-byte 16))) (time (prog1 t (time-rev *v*))) ; Evaluation took: ; 3.62 seconds of real time ; 3.581727 seconds of user run time ; 0.039302 seconds of system run time ; 11,089,005,122 CPU cycles ; [Run times include 0.05 seconds GC run time] ; 0 page faults and ; 200,116,120 bytes consed. }}} Reversing a string is 3 times slower. There is some cost since strings are utf-16 encoded, but we shouldn't be 3 times slower. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/81> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 2

[cmucl] #83: (reverse "") signals an error
by cmucl 19 May '13

19 May '13

#83: (reverse "") signals an error --------------------+------------------------------------------------------- Reporter: rtoy | Owner: somebody Type: defect | Status: new Priority: major | Milestone: Component: Core | Version: 2013-05 Keywords: | --------------------+------------------------------------------------------- As noted in [http://article.gmane.org/gmane.lisp.cmucl.general/6497 the email from Matt Kaufmann], {{{(reverse "")}}} causes an error. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/83> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 0

[cmucl-ticket] [cmucl] #80: (format t "~ve" 21 5d-324) generates an error
by cmucl 13 Apr '13

13 Apr '13

#80: (format t "~ve" 21 5d-324) generates an error --------------------+------------------------------------------------------- Reporter: rtoy | Owner: somebody Type: defect | Status: new Priority: major | Milestone: Component: Core | Version: 2013-03-a Keywords: | --------------------+------------------------------------------------------- {{{ (format t "~ve" 21 5d-324) }}} generates an error indexing into {{{*powers-of-ten*}}}. We can increase the size of {{{*powers-of-ten*}}} a bit, but there will always be cases where the table is too small. This can happen if {{{21}}} above is changed to a much larger number. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/80> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 1

[cmucl-ticket] [cmucl] #77: Add JOIN-PROCESS
by cmucl 25 Mar '13

25 Mar '13

#77: Add JOIN-PROCESS -------------------------+-------------------------------------------------- Reporter: sionescu | Owner: somebody Type: enhancement | Status: new Priority: major | Milestone: Component: Core | Version: 2013-03-a Keywords: | -------------------------+-------------------------------------------------- There should be a JOIN-PROCESS function that waits for a process to complete and returns its return value. The current implementation in Bordeaux-Threads uses (mp:process-wait "..." (lambda () (not (mp:process-alive-p process)))), but that doesn't fulfill the second requirement. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/77> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 4

[cmucl-ticket] [cmucl] #75: cl-unicode regression in 2013-03
by cmucl 24 Mar '13

24 Mar '13

#75: cl-unicode regression in 2013-03 ------------------------+--------------------------------------------------- Reporter: avodonosov | Owner: somebody Type: defect | Status: new Priority: major | Milestone: Component: Core | Version: 2013-03 Keywords: | ------------------------+--------------------------------------------------- I can load library cl-unicode in cmucl-20d, but not in 2013-03 snapshot. In the example below I use Quicklisp to load the library. To workaround ASDF upgrade issue between the recent Quicklisp and CMUCL, ASDF is loaded explicitly before Quicklisp. {{{ ./lisps/cmucl-20d/bin/lisp -noinit -nositeinit -load quicklisp/asdf.lisp -load ./quicklisp/setup.lisp (ql:quickload :cl-unicode) => OK }}} {{{ ./lisps/cmucl-2013-03/bin/lisp -noinit -nositeinit -load quicklisp/asdf.lisp -load ./quicklisp/setup.lisp (ql:quickload :cl-unicode) Error in function LISP::ASSERT-ERROR: Unknown property name "OtherGraphemeExtend". [Condition of type SIMPLE-ERROR] Restarts: 0: [CONTINUE ] Retry assertion with new value for DESIGNATOR. 1: Return NIL from load of #P"/home/testgrid/.cache/common-l isp/cmu- snapshot-2013-03__20d_unicode_-linux-x86/home/testgrid/quicklisp/dists/q uicklisp/software/cl-unicode-0.1.4/alias.sse2f". 2: [TRY-RECOMPILING] Recompile alias and try loading it again 3: [RETRY ] .Retry loading FASL for #<CL-SOURCE-FILE "cl-unicode" "alias">. 4: [ACCEPT ] Continue, treating loading FASL for #<CL-SOURCE-FILE "cl-unicode" "alias"> as having been successful. 5: [ABORT ] Give up on "cl-unicode" 6: Return to Top-Level. Debug (type H for help) (LISP::ASSERT-ERROR TEST-FUNCTION (DESIGNATOR) "Unknown property name ~S." "OtherGraphemeExtend") Source: Error finding source: Error in function DEBUG::GET-FILE-TOP-LEVEL-FORM: Source file no longer exists: target:code/macros.lisp. 0] backtrace 10 0: (LISP::ASSERT-ERROR TEST-FUNCTION (DESIGNATOR) "Unknown property name ~S." "OtherGraphemeExtend") 1: ((LABELS COLLECT-TEST-FUNCTIONS BUILD-DERIVED-TEST-FUNCTION) ("Me" "Mn" "OtherGraphemeExtend")) 2: (BUILD-DERIVED-TEST-FUNCTION ("Me" "Mn" "OtherGraphemeExtend")) 3: (BUILD-DERIVED-TEST-FUNCTIONS) 4: (BUILD-ALL-PROPERTY-TESTS) 5: (C::DO-CALL #<Code Object "Top-Level Form" {5904515F}> 16 17 7 ...) 6: (LISP::FOP-FUNCALL-FOR-EFFECT) 7: (LISP::LOAD-GROUP #<Stream for file "/home/testgrid/.cache/common-lisp/cmu-snapshot-2013 -03__20d_unicode_- linux-x86/home/testgrid/quicklisp/dists/quicklisp/software/cl- unicode-0.1.4/alias.sse2f">) 8: (LISP::FASLOAD #<Stream for file "/home/testgrid/.cache/common-lisp/cmu-snapshot-2013 -03__20d_unicode_- linux-x86/home/testgrid/quicklisp/dists/quicklisp/software/cl- unicode-0.1.4/alias.sse2f">) 9: (LISP::INTERNAL-LOAD #P"/home/testgrid/.cache/common-lisp/cmu-snapshot-2013 -03__20d_unicode_- linux-x86/home/testgrid/quicklisp/dists/quicklisp/software/cl- unicode-0.1.4/alias.sse2f" #P"/home/testgrid/.cache/common-lisp/cmu-snapshot-2013 -03__20d_unicode_- linux-x86/home/testgrid/quicklisp/dists/quicklisp/software/cl- unicode-0.1.4/alias.sse2f" :ERROR :BINARY ...) }}} -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/75> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 2

[cmucl-ticket] [cmucl] #79: File-position at start of file is wrong
by cmucl 23 Mar '13

23 Mar '13

#79: File-position at start of file is wrong --------------------+------------------------------------------------------- Reporter: rtoy | Owner: somebody Type: defect | Status: new Priority: major | Milestone: Component: Core | Version: 2013-03-a Keywords: | --------------------+------------------------------------------------------- Let {{{foo}}} be an file of latin1 characters. Then {{{ (with-open-file (s "foo" :external-format :latin1) (print (file-position s)) (read-char s) (print (file-position s))) }}} prints {{{ 512 513 }}} This is totally unexpected. I would expect 0 and 1 to be printed. If {{{foo}}} has utf8 encoding then {{{ (with-open-file (s "foo" :external-format :utf8) (print (file-position s)) (read-char s) (print (file-position s))) }}} prints something like {{{ 512 1 }}} I would expect 0 to be printed instead of 512. -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/79> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 1

[cmucl-ticket] [cmucl] #78: MP:PROCESS-WAIT should have an efficient way to wait for completion of another process
by cmucl 16 Mar '13

16 Mar '13

#78: MP:PROCESS-WAIT should have an efficient way to wait for completion of another process ----------------------+----------------------------------------------------- Reporter: sionescu | Owner: somebody Type: defect | Status: new Priority: major | Milestone: Component: Core | Version: 2013-03-a Keywords: | ----------------------+----------------------------------------------------- With the addition of a MP:PROCESS-FINISHED-P, the scheduler could avoid polling in this case: {{{ (mp:process-wait "..." #'mp:process-finished-p other-process) }}} -- Ticket URL: <http://trac.common-lisp.net/cmucl/ticket/78> cmucl <http://common-lisp.net/project/cmucl> Cmucl is a high-performance, free Common Lisp implementation.

1 0