Home My Page Projects Code Snippets Project Openings SML/NJ
Summary Activity Forums Tracker Lists Tasks Docs Surveys News SCM Files

SCM Repository

[smlnj] Diff of /sml/trunk/NOTES/HISTORY
ViewVC logotype

Diff of /sml/trunk/NOTES/HISTORY

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 594, Tue Apr 4 23:41:47 2000 UTC revision 1890, Wed Feb 22 20:08:17 2006 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15    ----------------------------------------------------------------------
16    Name: Allen Leung (leunga (at) reservoir (dot) com)
17    Date: 2006/02/22 15:07:00 EST
18    Tag: leunga-20060222-mdl-gen
19    Description:
20    
21      CM has changed.  Updated the script for rebuilding the MLRISC generated
22    files.
23    
24    ----------------------------------------------------------------------
25    Name: Matthias Blume (http://www.tti-c.org/blume/)
26    Date: 2006/02/14 17:30:00 CST
27    Tag: blume-20060214-intel-mac
28    Description:
29    
30    Hooked code for Darwin-specific Intel ABI into main compiler.
31    (This is a temporary solution which relies on the fact that the
32    compiler itself does not use NLFFI.  Eventually we need to divorce
33    intel mac from generic x86 unix code and make separate sets of binaries.)
34    
35    ----------------------------------------------------------------------
36    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
37    Date: 2006/02/13
38    Tag: jhr-20060213-intel-mac
39    Description:
40    Changed MLRISC x86 CCalls for partial support of MacOS X ABI.
41    
42    ----------------------------------------------------------------------
43    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
44    Date: 2006/02/06
45    Tag: jhr-20060206-intel-mac
46    Description:
47    Changes to support MacOS X on Intel hardware.  The C-calls support in
48    MLRISC must also be updated to support the MacOS X ABI.
49    
50    ----------------------------------------------------------------------
51    Name: Matthias Blume (http://www.tti-c.org/blume/)
52    Date: 2005/12/16 00:15:00 CST
53    Tag: blume-20051216-cm
54    Description:
55    
56    Improved error reporting and handling in CM.
57    
58    ----------------------------------------------------------------------
59    Name: Matthias Blume (http://www.tti-c.org/blume/)
60    Date: 2005/11/21 14:00:00 CST
61    Tag: blume-20051121-html-lib
62    Description:
63    
64    Enabling $/html-lib.cm in config/preloads did not work.  This is fixed
65    now.  (Since the anchor mapping for html-lib.cm is not yet in effect
66    at the time when config/preloads is processed, the library has to
67    be referred to by another name.  In 110.57 this would be
68    $SMLNJ-LIB/HTML/html-lib.cm.  I arranged for $smlnj/smlnj-lib/html-lib.cm
69    to be valid as well.)
70    
71    Thanks to Todd Wilson (Fresno) for alerting me to this issue.
72    
73    ----------------------------------------------------------------------
74    Name: Matthias Blume (http://www.tti-c.org/blume/)
75    Date: 2005/11/19 13:45:00 CST
76    Tag: Release_110_57
77    Description:
78    
79    Moved release tag for 110.57.
80    
81    ----------------------------------------------------------------------
82    Name: Matthias Blume (http://www.tti-c.org/blume/)
83    Date: 2005/11/19 23:35:00 CST
84    Tag: blume-20051119-fish
85    Description:
86    
87    Fixed a problem in config/install.sh where it tries to "fish" the
88    name of the CM metadata directory from the wrong place (because
89    the physical location of basis.cm has changed).  Also, corrected the
90    path anchor for $/html-lib.cm.  (Thanks to M. Fluet for pointing out
91    these problems.)
92    
93    ----------------------------------------------------------------------
94    Name: Matthias Blume (http://www.tti-c.org/blume/)
95    Date: 2005/11/16 14:30:00 CST
96    Tag: blume-orig-110_57
97    Description:
98    
99    New working version (110.57).  NEW BOOTFILES!
100    
101    UPDATE: The release tag has been moved to include a fix for a brown-paper-
102            bag glitch (see above).
103    
104    ----------------------------------------------------------------------
105    Name: Matthias Blume (http://www.tti-c.org/blume/)
106    Date: 2005/11/16 10:25:00 CST
107    Tag: blume-22051116-btrace
108    Description:
109    
110    Fixed problem with bogus exception message when using back-trace
111    facility.
112    
113    ----------------------------------------------------------------------
114    Name: Matthias Blume (http://www.tti-c.org/blume/)
115    Date: 2005/11/15 22:35:00 CST
116    Tag: blume-22051115-array2copy
117    Description:
118    
119    Added simple implementation of Array2.copy.  (Warning: mostly untested.)
120    
121    ----------------------------------------------------------------------
122    Name: Matthias Blume (http://www.tti-c.org/blume/)
123    Date: 2005/11/15 21:35:00 CST
124    Tag: blume-22051115-versiontool
125    Description:
126    
127    Reversed change to src/system/smlnj/internals/versiontool.cm.  This
128    file gets loaded as a tool -- by the equivalent of CM.make during the
129    run of CMB.make.  Thus, CMB's path configuration is meaningless for
130    it.  Instead, it has the status of "user code", so it should use
131    $/basis.cm to refer to the Basis library.  (At least that's true for
132    the purpose of bootstrapping the previous change.  In the future it
133    might make sense to have versiontool.cm refer to
134    $smlnj/basis/basis.cm, i.e., the version of the Basis that the
135    compiler itself uses.)
136    
137    Also patched src/system/testml to have it activate those extra anchor
138    bindings in config/extrapathconfig.
139    
140    ----------------------------------------------------------------------
141    Name: Matthias Blume (http://www.tti-c.org/blume/)
142    Date: 2005/11/15 16:35:00 CST
143    Tag: blume-22051115-libraries
144    Description:
145    
146    This change affects the way the following libraries are tied into
147    the system:
148    
149       $/basis.cm
150       $/smlnj-lib.cm
151       $/pp-lib.cm
152       $/controls-lib.cm
153       $/html-lib.cm
154       $/ml-yacc-lib.cm
155    
156    These libraries are now internally (as seen from the source code
157    of the implementation itself) known by the following names:
158    
159       $smlnj/basis/basis.cm
160       $smlnj/smlnj-lib/smlnj-lib.cm
161       $smlnj/smlnj-lib/pp-lib.cm
162       $smlnj/smlnj-lib/controls-lib.cm
163       $smlnj/smlnj-lib/html-lib.cm
164       $smlnj/ml-yacc/ml-yacc-lib.cm
165    
166    This makes it possible to work with code that requires different
167    versions of these libraries, and which refers to these libraries using
168    their "default" names (i.e., the first set of names above).  In other
169    words, one can un-define or re-define those default names without
170    compromising the proper functioning of the compiler itself.
171    
172    A similar procedure had already been performed for several of the
173    MLRISC libraries that are linked into the compiler.  I did some
174    cleanup on this code.
175    
176    A new file in the config directory (named extrapathconfig) is
177    responsible for setting up path anchors that the compiler itself does
178    not need, but that are typically required by user code.
179    
180    ----------------------------------------------------------------------
181    Name: Matthias Blume
182    Date: 2005/11/07 17:05:00 CST
183    Tag: blume-20051107-slice
184    Description:
185    
186    Fixed erroneous out-of-bounds test in the "update" function of
187    various *ArraySlice modules.  (Thanks to Vesa A. Norrman for pointing
188    out the problem.)
189    
190    Pushed some Basis changes through ML-Lex, CML, and eXene.
191    
192    ----------------------------------------------------------------------
193    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
194    Date: 2005/11/07
195    Description:
196    Fixed a Basis incompatibility: The depreciated function Substring.all
197    was removed (use Substring.full instead).
198    
199    ----------------------------------------------------------------------
200    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
201    Date: 2005/11/05
202    Description:
203    Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
204    
205    ----------------------------------------------------------------------
206    Name: Matthias Blume (http://www.tti-c.org/blume/)
207    Date: 2005/11/02 16:25:00 CST
208    Tag: blume-20051102-hex
209    Description:
210    
211    * Runtime system bootstrap code now accepts hex digits in BOOTLIST
212      in either upper case or lower case format.
213    * Pushed changes to names of Pack<N>{Big,Little} structures through
214      CML and eXene.
215    
216    ----------------------------------------------------------------------
217    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
218    Date: 2005/11/02
219    Description:
220    Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
221    be called PackWord<N>{Big,Little}.
222    
223    ----------------------------------------------------------------------
224    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
225    Date: 2005/10/28
226    Description:
227    Fixed a minor Basis incompatibility: hex digits should be upper case.
228    
229    ----------------------------------------------------------------------
230    Name: Matthias Blume (blume (at) tti - c (dot) org)
231    Date: 2005/10/25 16:00:00 CDT
232    Tag: Release_110_56
233    Description:
234    
235    New working version (110.56).  NEW BOOTFILES!
236    
237    ----------------------------------------------------------------------
238    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
239    Date: 2005/10/25
240    Description:
241    Added interval sets to utility library (signatures INTERVAL_DOMAIN
242    and INTERVAL_SET, and functor IntervalSetFn).
243    
244    ----------------------------------------------------------------------
245    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
246    Date: 2005/10/14
247    Description:
248    Add Zhong Shao's fix for datatype equality functions.
249    
250    ----------------------------------------------------------------------
251    Name: Allen Leung (leunga (at) reservoir (dot) com)
252    Date: 2005/10/14 12:30:00 EST
253    Tag: leunga-20051014-x86-spill
254    Description:
255    
256    Bug fix a bug found by Carl Hauser.
257    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
258    
259    ----------------------------------------------------------------------
260    Name: Allen Leung (leunga (at) reservoir (dot) com)
261    Date: 2005/10/14 11:15:00 EST
262    Tag: leunga-20051014-x86-asm
263    Description:
264    
265    Removed some debugging code in file x86Asm.sml.
266    The function emit_operand was printing out debugging output.
267    
268    ----------------------------------------------------------------------
269    Name: John Reppy
270    Date: 2005/07/27
271    Tag:
272    Description:
273    
274    Fixed ml-lex to recognize "\r" as representing carriage return.
275    
276    ----------------------------------------------------------------------
277    Name: John Reppy
278    Date: 2005/07/27
279    Tag:
280    Description:
281    
282    Fixed ml-yacc to work on files with non-native end-of-line
283    encodings (e.g., Windows text file on a Unix system).
284    
285    ----------------------------------------------------------------------
286    Name: John Reppy
287    Date: 2005/07/20
288    Tag:
289    Description:
290    
291    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
292    to support HPUX 11.
293    
294    ----------------------------------------------------------------------
295    Name: John Reppy
296    Date: 2005/07/06
297    Tag:
298    Description:
299    
300    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
301    
302    ----------------------------------------------------------------------
303    Name: Matthias Blume (blume (at) tti - c (dot) org)
304    Date: 2005/07/06 00:45:00 CDT
305    Tag: blume-20050706-slice-copy
306    Description:
307    
308    Fixed reversed logic for deciding whether to "copy up" or "copy down"
309    in *-array-slice.sml.
310    
311    ----------------------------------------------------------------------
312    Name: Allen Leung (leunga (at) reservoir (dot) com)
313    Date: 2005/05/31 17:00:00 EST
314    Tag: leunga-20050531-cygwin-fault-2
315    Description:
316    
317    A typo in the cygwin code fixed.
318    
319    ----------------------------------------------------------------------
320    Name: Allen Leung (leunga (at) reservoir (dot) com)
321    Date: 2005/05/31 16:47:00 EST
322    Tag: leunga-20050531-cygwin-fault
323    Description:
324    
325    Updated Cygwin's fault/signal handling to match the Windows version.
326    Updated the export list.
327    
328    ----------------------------------------------------------------------
329    Name: Matthias Blume (blume (at) tti - c (dot) org)
330    Date: 2005/05/18 13:35:00 CDT
331    Tag: Release_110_54
332    Description:
333    
334    New working version (110.54).  NEW BOOTFILES!
335    
336    ----------------------------------------------------------------------
337    Name: Matthias Blume (blume (at) tti - c (dot) org)
338    Date: 2005/05/18 11:58:00 CDT
339    Tag: blume-20050518-installer
340    Description:
341    
342    Added support scripts for Mac OS X PackageMaker and modified
343    config/install.sh so that it supports re-dumping a heap image
344    after customization.
345    
346    ----------------------------------------------------------------------
347    Name: Matthias Blume (blume (at) tti - c (dot) org)
348    Date: 2005/05/18 10:55:00 CDT
349    Tag: blume-20050518-realdiv-noovld
350    Description:
351    
352    Un-overloaded / to work around bug in overloading resolution code.
353    
354    ----------------------------------------------------------------------
355    Name: Matthias Blume (blume (at) tti - c (dot) org)
356    Date: 2005/05/16 23:50:00 CDT
357    Tag: blume-20050516-redump-heap
358    Description:
359    
360    Added mechanism for re-creating a heap file for the interactive system
361    after configuration variables have been changed.
362    
363       CM.redump_heap : string -> unit
364    
365    This is much like SMLofNJ.exportML, but starting from the resulting
366    heap does not return to the caller of CM.redump_heap but
367    restarts the interactive system from scratch.  The original call of
368    CM.redump_heap does not return but ends the interactive session.
369    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
370    
371    Internally, redump_heap winds the dynamic execution context back to
372    the point where the original heap image was created and re-executes
373    the heap image generation code in the boot code.
374    
375    ----------------------------------------------------------------------
376    Name: Matthias Blume (blume (at) tti - c (dot) org)
377    Date: 2005/05/09 21:55:00 CDT
378    Tag: blume-20050509-word64
379    Description:
380    
381    Added a hack to the existing hack known as Word64 to make fromString
382    behave correctly.  I am still not sure whether Word64.scan will work
383    as specified with respect to the interaction of radix and prefix.
384    
385    ----------------------------------------------------------------------
386    Name: Allen Leung (leunga (at) reservoir (dot) com)
387    Date: 2005/05/04 11:50:00 EST
388    Tag: leunga-20050504-checkgc
389    Description:
390    
391    Added a gc protocol checking phase.  This phase is enabled with
392    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
393    
394    ----------------------------------------------------------------------
395    Name: Matthias Blume (blume (at) tti - c (dot) org)
396    Date: 2005/05/04 10:45:00 CDT
397    Tag: blume-20050504-intinf
398    Description:
399    
400    Fixed a bug in the implementation of div and mod for IntInf.
401    Thanks to Neophytos Michael for reporting the problem.
402    
403    ----------------------------------------------------------------------
404    Name: Matthias Blume (blume (at) tti - c (dot) org)
405    Date: 2005/05/04 10:35:00 CDT
406    Tag: blume-20050504-join
407    Description:
408    
409    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
410    
411    ----------------------------------------------------------------------
412    Name: Matthias Blume (blume (at) tti - c (dot) org)
413    Date: 2005/02/28 23:40:00 CST
414    Tag: blume-20050228-mVar
415    Description:
416    
417    Fixed serious bug (brown paper bag variety) in new implementation of
418    structure Atom in CML.  (I had accidentally used a mailbox instead of
419    an mvar, leaving the door open for races.)
420    
421    ----------------------------------------------------------------------
422    Name: Matthias Blume (blume (at) tti - c (dot) org)
423    Date: 2005/02/25 15:00:00 CST
424    Tag: Release_110_53
425    Description:
426    
427    New working version (110.53).  NEW BOOTFILES!
428    
429    ----------------------------------------------------------------------
430    Name: Matthias Blume (blume (at) tti - c (dot) org)
431    Date: 2005/02/25 14:50:00 CST
432    Tag: blume-20050225-susp
433    Description:
434    
435    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
436    implemented in Core -- which means that it is the same as the one used
437    by the lazy extension.
438    
439    ----------------------------------------------------------------------
440    Name: Matthias Blume (blume (at) tti - c (dot) org)
441    Date: 2005/02/24 16:50:00 CST
442    Tag: blume-20050224-cml-atom
443    Description:
444    
445    Simpler and at the same time more general implementation of structure
446    Atom in CML.
447    
448    ----------------------------------------------------------------------
449    Name: Matthias Blume (blume (at) tti - c (dot) org)
450    Date: 2005/02/15 17:35:00 CST
451    Tag: blume-20050215-tools
452    Description:
453    
454    Created new "tools" directory under "src" and moved "TraceDebugProf"
455    there.
456    
457    ----------------------------------------------------------------------
458    Name: Matthias Blume (blume (at) tti - c (dot) org)
459    Date: 2005/02/10 17:55:00 CST
460    Tag: blume-20050210-longlong
461    Description:
462    
463    Implemented "long long" arguments and results for NLFFI.  (Only the
464    PPC/MacOS implementation is complete, the other backends still need to
465    be updated.)
466    
467    ----------------------------------------------------------------------
468    Name: Matthias Blume (blume (at) tti - c (dot) org)
469    Date: 2005/01/24 17:40:00 CST
470    Tag: blume-20050124-mlyacc
471    Description:
472    
473    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
474    problem with certain "as" patterns which previously got rendered
475    using incorrect syntax.
476    
477    ----------------------------------------------------------------------
478    Name: Matthias Blume (blume (at) tti - c (dot) org)
479    Date: 2005/01/18 12:00:00 CST
480    Tag: blume-20050118-profile
481    Description:
482    
483    Made time profiling code (interrupt handler) in runtime system aware
484    of new array representation.
485    
486    ----------------------------------------------------------------------
487    Name: Matthias Blume (blume (at) tti - c (dot) org)
488    Date: 2005/01/14 18:00:00 CST
489    Tag: blume-20050114-heap2exec
490    Description:
491    
492    Implemented new (but still experimental) heap2exec facility.  This is
493    tested under Mac OS X and should work under Linux (will test shortly).
494    It will probably also work on the Sparc (will test some time later).
495    
496      - removed old "HACKED_STANDALONE" hack from runtime
497    
498    To be able to test this, uncomment the request for "heap2asm" in
499    config/targets prior to installation.  (Notice that this is different
500    from "heap2exec" mentioned below.  Not a typo.)
501    
502    To perform an actual test, run the command
503    
504       $ bin/heap2exec heapfile execfile
505    
506    (You can put heap2exec on your shell's path.)
507    
508    For example, run
509    
510      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
511    
512    This will create a standalone executable called "mly" which you can
513    then invoke directly as a command.
514    
515    ----------------------------------------------------------------------
516    Name: Matthias Blume (blume (at) tti - c (dot) org)
517    Date: 2005/01/07 11:44:00 CST
518    Tag: blume-20050107-mlstring
519    Description:
520    
521    fixed off-by-one error in ML_STRING macro (globals.c)
522    
523    ----------------------------------------------------------------------
524    Name: Matthias Blume (blume (at) tti - c (dot) org)
525    Date: 2004/12/23 18:00:00 CST
526    Tag: blume-20041223-santa
527    Description:
528    
529    Made ml-build script "smarter" (but only very little).
530    
531    ----------------------------------------------------------------------
532    Name: Matthias Blume (blume (at) tti - c (dot) org)
533    Date: 2004/12/21 15:05:00 CST
534    Tag: blume-20041221-longlong
535    Description:
536    
537    * Implemented access to signed and unsigned long long data in NLFFI.
538    (The parameter-passing part of the picture has not complete. But
539    data structure access seems to work.)
540    
541    * Fixed CM's incorrect assumption that the PPC is little-endian.
542      (On the Mac, it is big-endian.  And that's currently our only
543      PPC platform.)
544    
545    ----------------------------------------------------------------------
546    Name: Matthias Blume (blume (at) tti - c (dot) org)
547    Date: 2004/12/21 12:50:00 CST
548    Tag: blume-20041221-memory
549    Description:
550    
551    Some cleanup in the $c/memory.cm library: separated some concerns by
552    moving allocation code and memory access code each into their own
553    files.
554    
555    ----------------------------------------------------------------------
556    Name: Allen Leung (leunga (at) reservoir (dot) com)
557    Date: 2004/12/17 16:12:00 EST
558    Tag: leunga-20041217-cygwin-smlnj-home
559    Description:
560    
561        The Unix I/O library of SML/NJ on cygwin does not understand
562    Windows style pathname, so problems arise when SMLNJ_HOME is set
563    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
564    to a POSIX pathname on cygwin.
565    
566    ----------------------------------------------------------------------
567    Name: Matthias Blume (blume (at) tti - c (dot) org)
568    Date: 2004/12/16 13:35:00 CST
569    Tag: Release_110_52
570    Description:
571    
572    Last-minute changes incorporated into 110.52.  Release tag moved.
573    
574    The changes:
575       - HashString.hashString' -> HashString.hashSubstring
576       - bug fix in UnivariateStats
577    
578    ----------------------------------------------------------------------
579    Name: Matthias Blume (blume (at) tti - c (dot) org)
580    Date: 2004/12/15 23:40:00 CST
581    Tag: blume-20041215-hashSubstring
582    Description:
583    
584    - HashString.hashString' -> HashString.hashSubstring
585    - corresponding changes in atom.sml
586    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
587      added some comments
588    
589    ----------------------------------------------------------------------
590    Name: Matthias Blume (blume (at) tti - c (dot) org)
591    Date: 2004/12/15 15:30:00 CST
592    Tag: (Release_110_52)
593    Description:
594    
595    New working version (110.52).  NEW BOOTFILES!
596    
597    ----------------------------------------------------------------------
598    Name: Matthias Blume (blume (at) tti - c (dot) org)
599    Date: 2004/12/15 12:45:00 CST
600    Tag: blume-20041215-spaces
601    Description:
602    
603    More on the space problem (this time for Win32).
604    
605    ----------------------------------------------------------------------
606    Name: Matthias Blume (blume (at) tti - c (dot) org)
607    Date: 2004/12/14 17:30:00 CST
608    Tag: blume-20041214-spaces
609    Description:
610    
611    Hacked some of the scripts (in particular: the installer) to cope with
612    spaces in filenames a bit better.  But beware: the current "solution"
613    is likely still full of bugs and inherently incomplete.  (We need to
614    do away with those shell scripts for a comprehensive solution.)
615    
616    ----------------------------------------------------------------------
617    Name: Matthias Blume (blume (at) tti - c (dot) org)
618    Date: 2004/12/13 14:45:00 CST
619    Tag: blume-20041213-ml-makedepend
620    Description:
621    
622    Fixed bug in code for ml-makedepend.
623    
624    ----------------------------------------------------------------------
625    Name: Matthias Blume (blume (at) tti - c (dot) org)
626    Date: 2004/12/09 16:30:00 CST
627    Tag: blume-20041209-statistics
628    Description:
629    
630    Added two simple but potentially useful statistics modules to SML/NJ Library.
631    (See CHANGES file there.)
632    
633    ----------------------------------------------------------------------
634    Name: Matthias Blume (blume (at) tti - c (dot) org)
635    Date: 2004/12/01 16:50:00 CST
636    Tag: blume-20041201-atom
637    Description:
638    
639    smlnj-lib:
640    
641      Added function HashString.hashString' for substrings.
642      Hand-inlined CharVector.foldl into HashString (for speed).
643      Modified implementation of structure Atom to avoid extracting
644      strings from substrings unless necessary.
645      (Also see CHANGES file for smlnj-lib.)
646    
647    ----------------------------------------------------------------------
648    Name: Matthias Blume (blume (at) tti - c (dot) org)
649    Date: 2004/11/24 22:15:00 CST
650    Tag: blume-20041124-cml
651    Description:
652    
653    Made sure CML compiles when Position = Int64.
654    
655    ----------------------------------------------------------------------
656    Name: Matthias Blume (blume (at) tti - c (dot) org)
657    Date: 2004/11/24 14:30:00 CST
658    Tag: blume-20041124-position
659    Description:
660    
661    The compiler can now be compiled in a mode that makes structure Position
662    equal to Int64.  The default, however, is unchanged (Position = Int31)
663    for the time being.
664    
665    To enable 64-bit positions, use the following procedure:
666    
667    1. Start sml
668    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
669    3. Type
670          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
671    4. Run CMB.make() as usual.
672    
673    This is barely tested.  The only test so far was a little SML program
674    counting the number of characters in an 8-gigabyte file by
675    reading it character-by-character.  That test was successful.
676    
677    In support of 64-bit positions, a number of new functions have been
678    added to the runtime system.
679    
680    ----------------------------------------------------------------------
681    Name: Matthias Blume (blume (at) tti - c (dot) org)
682    Date: 2004/11/23 14:45:00 CST
683    Tag: blume-20041123-useFile
684    Description:
685    
686    Fixed a problem with unhelpful error messages related to problems with
687    .cm- or .sml files that appear as part of the sml command line.
688    
689    ----------------------------------------------------------------------
690    Name: Matthias Blume (blume (at) tti - c (dot) org)
691    Date: 2004/11/18 15:40:00 CST
692    Tag: Release_110_51
693    Description:
694    
695    New working version (110.51).  NEW BOOTFILES!
696    
697    ----------------------------------------------------------------------
698    Name: Matthias Blume (blume (at) tti - c (dot) org)
699    Date: 2004/11/18 15:35:00 CST
700    Tag: <none>
701    Description:
702    
703    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
704    Joemann).
705    
706    ----------------------------------------------------------------------
707    Name: Allen Leung (leunga (at) reservoir (dot) com)
708    Date: 2004/11/17 16:05:21 EST 2004
709    Tag: leunga-20041117-mlrisc-live-kill
710    Description:
711    
712       Added support for MLTree constructs LIVE and KILL
713    to all the architectures.
714    
715    ----------------------------------------------------------------------
716    Name: Matthias Blume (blume (at) tti - c (dot) org)
717    Date: 2004/11/13 00:20:00 CST
718    Tag: blume-20041113-versiontool
719    Description:
720    
721    - Stripped down the versiontool: It now only handles the version number.
722      The date string is generated at bootstrap time (during makeml).
723    
724    - In a previous commit, fixed a minor issue with how polyequal is being
725      translated.  In particular, the code now "looks through" abstractions.
726      This results in slightly fewer polyEqual warnings and hopefully slightly
727      more efficient code.  Important examples for where this matters are
728      the new int64 and word64 types.
729    
730    ----------------------------------------------------------------------
731    Name: Matthias Blume (blume (at) tti - c (dot) org)
732    Date: 2004/11/12 00:30:00 CST
733    Tag: blume-20041112-int64
734    Description:
735    
736    Structure Int64 fully hooked in.  (The implementation is not very
737    efficient, though.)
738    
739    ----------------------------------------------------------------------
740    Name: Matthias Blume (blume (at) tti - c (dot) org)
741    Date: 2004/11/11 17:30:00 CST
742    Tag: blume-20041111-more64
743    Description:
744    
745    All the pieces of Word64 are now there, with the exception of the
746    conversions from and to LargeWord. (Eventually these need to be identities,
747    but for the time being they don't even make sense because LargeWord is
748    32-bit wide.)
749    
750    Also started to add similar support for Int64, but major pieces of that
751    are still missing.
752    
753    ----------------------------------------------------------------------
754    Name: Matthias Blume (blume (at) tti - c (dot) org)
755    Date: 2004/11/11 00:15:00 CST
756    Tag: blume-20041111-word64
757    Description:
758    
759    Structure Word64 is now (almost) complete, word literals and patterns
760    seem to work.  There are a few odd pieces missing.  In particular,
761    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
762    at the moment.
763    
764    Making Word64 official would mean that LargeWord becomes Word64.  But
765    this requires extreme care because most word-word conversions have to
766    go through LargeWord, so making a mistake means loss of efficiency or
767    worse.  Eventually there will be a solution similar to (but actually
768    simpler than) what I did with IntInf.
769    
770    ----------------------------------------------------------------------
771    Name: Matthias Blume (blume (at) tti - c (dot) org)
772    Date: 2004/11/10 18:12:00 CST
773    Tag: blume-20041110-64bit
774    Description:
775    
776    More 64-bit hacking (but still not even half-way there yet).
777    Also, some assorted improvements to the handling of 8-bit words.
778    
779    ----------------------------------------------------------------------
780    Name: Matthias Blume (blume (at) tti - c (dot) org)
781    Date: 2004/11/09 17:50:00 CST
782    Tag: <none>
783    Description:
784    
785    Started adding some infrastructure for supporting 64-bit int- and
786    word-types.  (Still in its very early stages.)
787    
788    ----------------------------------------------------------------------
789    Name: Matthias Blume (blume (at) tti - c (dot) org)
790    Date: 2004/10/28 10:45:00 CDT
791    Tag: Release_110_50
792    Description:
793    
794    New working version (110.50).  NEW BOOTFILES!
795    
796                 =====================
797    
798    Also:
799    
800      - Changend config/srcarchiveurl from a file just
801        containing the URL string into a file containing
802        shell script code.  The code has access to the $VERSION variable.
803      - Made corresponding changes to config/install.sh and config/unpack.
804      - Default contents of config/srcarchiveurl uses $VERSION and
805        normally does not have to be edited to reflect a version change.
806    
807      (As a result, a version change can be done by just editing
808       config/version, the rest is now automatic.)
809    
810    ----------------------------------------------------------------------
811    Name: Matthias Blume (blume (at) tti - c (dot) org)
812    Date: 2004/10/27 17:50:00 CDT
813    Tag: blume-20041027-btrace-msg
814    Description:
815    
816    BackTrace.monitor now also reports the source of the exception that
817    triggered the trace.
818    
819    ----------------------------------------------------------------------
820    Name: Matthias Blume (blume (at) tti - c (dot) org)
821    Date: 2004/10/27 17:20:00 CDT
822    Tag: blume-20041027-x86-c-calls
823    Description:
824    
825    This is the HISTORY entry for two earlier commits, both concerning
826    the x86 c-calls code in MLRISC:
827    
828      - added a missing LOAD in the code that deals with struct arguments
829      - made sure the caller does not add the wrong number of bytes to the
830        stack pointer after a call of a function returning a struct
831        (the callee already pops the implicit argument which points to
832         the space reserved for the result)
833    
834    ----------------------------------------------------------------------
835    Name: Allen Leung (leunga (at) reservoir (dot) com)
836    Date: 2004/10/24 14:00:00 EST
837    Tag: leunga-20041024-x86-gas-fucomip
838    Description:
839    
840       John discovered a bug in the syntax of fucomip.
841       The opcodes FU?COMIP? have been changed to
842    
843           fu?comip? %st(i), %st
844    
845    ----------------------------------------------------------------------
846    Name: Matthias Blume (blume (at) tti - c (dot) org)
847    Date: 2004/10/20 15:06:00 CDT
848    Tag: blume-20041020-standalone-backtrace
849    Description:
850    
851    Added a mechanism for getting back-trace information from standalone
852    programs.  Here is how it works:
853    
854       1. The part of the program from which you want to get backtrace
855          information (usually the whole program) should be wrapped with
856          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
857          main program could be modified from something like
858    
859             fun main (pgm, args) = ...
860    
861         to
862    
863            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
864    
865       2. To be able to access BackTrace.monitor, you have to add
866          library $smlnj-tdp/plugins.cm to the .cm file that contains your
867          main function.
868    
869       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
870          CM might have created in the past for your project).
871    
872       4. Build the system using this command line:
873    
874             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
875                      myprog.cm MyProg.main myprog
876    
877          instead of the usual
878    
879             ml-build myprog.cm MyProg.main myprog
880    
881    I changed a library name:
882    
883       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
884    
885    New libraries:
886    
887       $smlnj-tdp/back-trace.cm
888          -- when loaded causes the back-trace plugin to be installed
889       $smlnj-tdp/coverage.cm
890          -- when loaded causes the coverage plugin to be installed
891    
892    ----------------------------------------------------------------------
893    Name: Matthias Blume (blume (at) tti - c (dot) org)
894    Date: 2004/10/18 16:45:00 CDT
895    Tag: blume-20041018-groupowner
896    Description:
897    
898    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
899    
900    Eliminated group owner specs from .cm files throughout the source tree.
901    
902    ----------------------------------------------------------------------
903    Name: Matthias Blume (blume (at) tti - c (dot) org)
904    Date: 2004/10/15 15:45:00 CDT
905    Tag: blume-20041015-coverage
906    Description:
907    
908    * Test coverage tool added!
909    * Further reorganization of tracing-, debugging-, and profiling support:
910    
911        - moved original BTImp -- now called BackTrace -- into a separate
912          library called $/trace-debug-profile.cm
913        - eliminated all mentions of BTrace from SMLofNJ.Internals
914        - only the instrumentation mechanism is now left in the compiler proper
915        - BackTrace module is a plugin which is NOT plugged in by default
916        - Coverage module is another such plugin
917    
918    To get the benefits of any of these plugin modules, the code in
919    question must be compiled with tdp instrumentation turned on.  This
920    can be done by setting SMLofNJ.Internals.TDP.mode to true.
921    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
922    
923    Plugins are selected at link time.  (Pre-compiled instrumented code
924    can be re-loaded with different plugins in effect.)  When an
925    instrumented module is linked, whatever plugins are at that time
926    enabled will come into effect for that module.
927    
928    To enable the back-trace plugin, load library $/trace-debug-profile.cm
929    and invoke BackTrace.install() (e.g., from the interactive prompt).
930    To enable the coverage plugin, load the same library and invoke
931    Coverage.install().
932    
933    Back-traces are generated automatically on uncaught exceptions and
934    when the code in question explicitly invokes BackTrace.trigger().
935    
936    Coverage (and execution frequency-) information must be queried
937    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
938    
939    ----------------------------------------------------------------------
940    Name: Matthias Blume (blume (at) tti - c (dot) org)
941    Date: 2004/10/14 17:40:00 CDT
942    Tag: blume-20041014-tdp-core
943    Description:
944    
945    Snapshot of a significant overhaul of how the trace/debug/profile support
946    is hooked into the system (specifically: Core and SMLofNJ.Internals).
947    
948    ----------------------------------------------------------------------
949    Name: Matthias Blume (blume (at) tti - c (dot) org)
950    Date: 2004/10/13 16:34:00 CDT
951    Tag: blume-20041013-tdp
952    Description:
953    
954    Some rationalization of names:
955    
956       structure BTrace -> structure TDPInstrument
957       etc.
958    
959    This is is preparation of using the original back-trace
960    instrumentation for other purposes.  "TDP" stands for
961    Trace/Debug/Profile.
962    
963    The control flag controlling whether instrumentation is on or off is now
964    registered under a different name, so instead of running sml as
965    
966     sml -Cinstrument.btrace-mode=true
967    
968    one has to say
969    
970     sml -Ctdp.instrument=true
971    
972    ----------------------------------------------------------------------
973    Name: Matthias Blume (blume (at) tti - c (dot) org)
974    Date: 2004/10/11 16:37:00 CDT
975    Tag: blume-20041011-regions
976    Description:
977    
978    Made some minor modifications to elabcore.sml to have source regions
979    be propagated more tightly -- resulting in better (i.e., smaller)
980    regions being reported in error- and debug messages.
981    
982    ----------------------------------------------------------------------
983    Name: Matthias Blume (blume (at) tti - c (dot) org)
984    Date: 2004/10/08 22:50:00 CDT
985    Tag: blume-20041008-cmkw
986    Description:
987    
988    Fixed handling of keywords in .cm files: After seeing "is" the lexer
989    treats subsequent occurrences of "group", "library", "source", "is",
990    "*", and "-" as ordinary identifiers rather than keywords.
991    
992    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
993    is supposed to accept a tool argument called "source", but this did
994    not work because of the clash with the keyword.
995    
996    ----------------------------------------------------------------------
997    Name: Matthias Blume (blume (at) tti - c (dot) org)
998    Date: 2004/10/07 16:00:00 CDT
999    Tag: blume-20041007-cleanup
1000    Description:
1001    
1002    Assorted cleanup work:
1003    
1004      - got rid of intstrmap in favor of using the library's
1005        hash table implementation
1006      - threw out most of the pathnames stuff, as it was not used anyway
1007      - simplified tokentable implementation
1008      - fixed some minor spelling errors
1009    
1010    ----------------------------------------------------------------------
1011    Name: Matthias Blume (blume (at) tti - c (dot) org)
1012    Date: 2004/10/06 15:15:15 CDT
1013    Tag: blume-20041006-handler
1014    Description:
1015    
1016    Cleaned up the absyn to reflect the invariant that HANDLE always
1017    carries a FNexp as part of the type definition.  This eliminates some
1018    superfluous sanity checks at runtime down the road.
1019    
1020    Some minor cleanup of the btrace code.
1021    
1022    ----------------------------------------------------------------------
1023    Name: Matthias Blume (blume (at) tti - c (dot) org)
1024    Date: 2004/10/01 10:20:30 CDT
1025    Tag: blume-20041001-slave
1026    Description:
1027    
1028    Added hack to make slave mode work in the presence of the version
1029    tool.  (Still, since the master does two passes over the code for
1030    CMB.make, the release number gets bumped twice when slaves are
1031    attached.  I don't know if this is worth fixing...)
1032    
1033    ----------------------------------------------------------------------
1034    Name: Matthias Blume (blume (at) tti - c (dot) org)
1035    Date: 2004/09/30 10:55:00 CDT
1036    Tag: blume-20040930-version
1037    Description:
1038    
1039    * Moved the "version" magic into its own little library under
1040      src/system/smlnj/internal.  This avoids expensive reconstruction of
1041      a stable src/compiler/core.cm.
1042    
1043    * At the same time, structure CompilerVersion is now known as
1044      structure SMLNJVersion.
1045    
1046    * Arranged for the version tool to NOT kick in when rebuilding the system
1047      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
1048      Also, loading the versiontool does not work when rebuilding the system
1049      because CM is not properly initialized at that time.
1050    
1051    ----------------------------------------------------------------------
1052    Name: Matthias Blume (blume (at) tti - c (dot) org)
1053    Date: 2004/09/29 14:00:00 CDT
1054    Tag: blume-20040929-autoversion
1055    Description:
1056    
1057    Implemented some CM magic to have
1058        file src/compiler/TopLevel/main/version.sml
1059    generated automagically.
1060    The version is taken from two files: config/version and config/release.
1061    The first is expected to contain a two-part version number such as 110.49.
1062    The second should contain a single number, but it may be missing.
1063    
1064    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
1065    time the version tool is loaded (which is the first time you say CMB.make),
1066    then the tool will increment the value stored in config/release every
1067    time CMB.make is invoked.
1068    
1069    The binfile format is now insensitive to anything beyond the first
1070    two components of a version number, so bumping the release does not render
1071    binfiles incompatible.  Auto-bumping can be used to keep track of versions
1072    during development without invalidating existing binfiles.
1073    
1074    In any case, every CMB.make updates the date information in version.sml.
1075    (This is the date that is printed in the banner.)
1076    
1077    ----------------------------------------------------------------------
1078    Name: Matthias Blume (blume (at) tti - c (dot) org)
1079    Date: 2004/09/28 10:53:00 CDT
1080    Tag: blume-20040928-controls
1081    Description:
1082    
1083    Some cleanup of the controls code.
1084    
1085    ----------------------------------------------------------------------
1086    Name: Matthias Blume (blume (at) tti - c (dot) org)
1087    Date: 2004/09/27 22:08:00 CDT
1088    Tag: blume-20040927-controls
1089    Description:
1090    
1091    Added two pieces of functionality to the Controls interface:
1092    
1093    1.   val save'restore: 'a control -> unit -> unit
1094    
1095       grabs the current value of the control in stage 1 and restores it
1096       in stage 2.
1097    
1098    2.   val set' : 'a control * 'a -> unit -> unit
1099    
1100       stores the given value into the control in stage 2 (i.e., delayed)
1101       but does all error checking in stage 1.
1102       (This is for string controls that need to do parse their argument --
1103        something that might fail.  In some cases, notably in CM, one
1104        already knows the intended argument but wants to delay the actual
1105        assignment until a time when error recovery would be more difficult.)
1106    
1107    Changed the handling of controls in tool arguments to classes "sml" and
1108    "lazysml":
1109       - use Controls.save'restore as a more robust way of restoring the
1110         old value (in particular: without having to re-parse the string)
1111       - use controls to handle the "overload" keyword in the init group
1112         (I believe this change actually fixes a long-standing obscure bug.)
1113    
1114    ----------------------------------------------------------------------
1115    Name: Matthias Blume (blume (at) tti - c (dot) org)
1116    Date: 2004/09/27 17:00:00 CDT
1117    Tag: blume-20040927-lazysml
1118    Description:
1119    
1120    Added a new tool class called "lazysml" to CM's tool chest.  The only
1121    difference to "sml" is that compilation is done with Control.lazysml
1122    set to true.  A source of class "lazysml" is automatically recognized
1123    by a file name suffix of ".lml".
1124    
1125    In addition to the above feature, the original class "sml" now also
1126    supports a tool argument "lazy" which has the same effect.  As a
1127    result, the following three lines are equivalent:
1128    
1129        foo.sml : lazysml
1130        foo.sml : sml (lazy)
1131        foo.sml (lazy)
1132    
1133    The setting goes into effect both during parsing and during
1134    compilation.  The original setting is restored right after parsing and
1135    after compilation, respectively.
1136    
1137    In addition to all the above, there is also a general mechanism to set
1138    ANY of the "controls" that are available at the command line via
1139    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
1140    apply as well. (In fact, "lazy" is implemented as a special case of
1141    the general mechanism.)
1142    
1143    The .cm file syntax uses a new keyword tool argument called "with".
1144    There are several ways of indicating the desired settings:
1145    
1146        foo.sml (with:parser.quotations=true)
1147        foo.sml (with:(name:parser.quotations value:true))
1148        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1149        foo.sml (with:(name1=value1 name2=value2 ...))
1150        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1151    
1152        etc.
1153    
1154    Another possible abbreviation is to leave out the =v or value:v part
1155    if the name refers to a boolean control (in which case the value is
1156    taken to be true).  Thus, one could get lazy sml also by saying:
1157    
1158        foo.sml (with:parser.lazy-keyword=true)
1159        foo.sml (with:parser.lazy-keyword)
1160        foo.sml (with:(name:parser.lazy-keyword value:true))
1161        foo.sml (with:(name:parser.lazy-keyword))
1162    
1163    ----------------------------------------------------------------------
1164    Name: Matthias Blume (blume (at) tti - c (dot) org)
1165    Date: 2004/09/24 16:22:00 CDT
1166    Tag: blume-20040924-ppc-long-branch
1167    Description:
1168    
1169    Turned message about "emiting long form of branch" off by default.
1170    Added a control flag to turn it back on when desired.
1171    
1172    ----------------------------------------------------------------------
1173    Name: Matthias Blume (blume (at) tti - c (dot) org)
1174    Date: 2004/09/24 16:05:00 CDT
1175    Tag: blume-20040924-rounding
1176    Description:
1177    
1178    Applied patch for setting rounding modes under Mac OS X.  Thanks to
1179    Melissa O'Neill for providing the code!
1180    
1181    ----------------------------------------------------------------------
1182    Name: Matthias Blume (blume (at) tti - c (dot) org)
1183    Date: 2004/09/23 17:30:00 CDT
1184    Tag: blume-20040923-envvars
1185    Description:
1186    
1187    1. Changed definition of type ControlRegistry.registry_tree to
1188       include control_info (i.e., the name of the controlling environment
1189       variable).
1190    
1191    2. Added command-line flags -e and -E to print the names of environment
1192       variables that can be used to control internal settings.
1193       (This uses the new API mentioned in 1.)
1194    
1195    ----------------------------------------------------------------------
1196    Name: Matthias Blume (blume (at) tti - c (dot) org)
1197    Date: 2004/09/13 16:50:00 CDT
1198    Tag: Release_110_49
1199    Description:
1200    
1201    New working version (110.49).  NEW BOOTFILES!
1202    
1203    ----------------------------------------------------------------------
1204    Name: Matthias Blume
1205    Date: 2004/09/13 16:20:00 CDT
1206    Tag: blume-20040913-config-mlrisc
1207    Description:
1208    
1209    Put target "mlrisc" back into the default list.
1210    (There is no harm in having it, and some users have expressed their
1211    wish to have "mlrisc" included by default.)
1212    
1213    ----------------------------------------------------------------------
1214    Name: John Reppy
1215    Date: 2004/09/13
1216    Tag: jhr-20040913-signals
1217    Description:
1218    
1219    Fixed the signal masking code to properly nest mask/unmask operations
1220    on a per-signal basis.
1221    
1222    ----------------------------------------------------------------------
1223    Name: Matthias Blume (blume (at) tti - c (dot) org)
1224    Date: 2004/09/08 13:20:00 CDT
1225    Tag: blume-20040908-heap-magic
1226    Description:
1227    
1228    Bumped the heap macig number to 0x09082004 to account for the changed
1229    layout of the ML frame under MacOS X.
1230    
1231    ----------------------------------------------------------------------
1232    Name: Allen Leung (leunga (at) reservoir (dot) com)
1233    Date: 2004/09/03 11:26:00 EST
1234    Tag: leunga-20040903-cygwin-install
1235    Description:
1236    
1237       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1238    The Cygwin runtime is turned on by setting the environment
1239    variable SMLNJ_CYGWIN_RUNTIME to 1.
1240    
1241    ----------------------------------------------------------------------
1242    Name: Matthias Blume (blume (at) tti - c (dot) org)
1243    Date: 2004/08/31 17:15:00 CDT
1244    Tag: blume-20040831-core
1245    Description:
1246    
1247    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1248    
1249    ----------------------------------------------------------------------
1250    Name: Matthias Blume (blume (at) tti - c (dot) org)
1251    Date: 2004/08/30 17:55:00 CDT
1252    Tag: blume-20040830-installer
1253    Description:
1254    
1255    Upon request by Johannes Joemann:
1256    
1257    - improved ML code of installer to fall back to coping when renaming
1258      fails (i.e., when source and target are on different file systems);
1259      the code compiles but has yet to be tested in anger
1260    - removed mlrisc from list of default targets (config/targets)
1261    
1262    ----------------------------------------------------------------------
1263    Name: Matthias Blume (blume (at) tti - c (dot) org)
1264    Date: 2004/08/27 17:20:00 CDT
1265    Tag: blume-20040827-ptreql
1266    Description:
1267    
1268    Added ptreql primop to structure InlineT (upon request from Larry
1269    Paulson).
1270    
1271    ----------------------------------------------------------------------
1272    Name: Allen Leung (leunga (at) reservoir (dot) com)
1273    Date: 2004/08/15 21:21:00 EST
1274    Tag: leunga-110_48-udgraph
1275    Description:
1276    
1277    Another bug fix from Carl Hauser:
1278    
1279    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1280    > 48c48
1281    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1282    > e::L)
1283    > ---
1284    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1285    > Without this, any deletion of an edge in an undirected graph does severe
1286    > violence to the graph.
1287    
1288    ----------------------------------------------------------------------
1289    Name: Allen Leung (leunga (at) reservoir (dot) com)
1290    Date: 2004/08/10 23:35:00 EST
1291    Tag: leunga-110_48-ppc
1292    Description:
1293    
1294       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1295    
1296    ----------------------------------------------------------------------
1297    Name: Matthias Blume (blume (at) tti - c (dot) org)
1298    Date: 2004/08/10 12:00:00 CDT
1299    Tag: Release_110_48
1300    Description:
1301    
1302    New working version (110.48).  NEW BOOTFILES!
1303    
1304    ----------------------------------------------------------------------
1305    Name: Allen Leung (leunga (at) reservoir (dot) com)
1306    Date: 2004/08/09 12:21:00 EST
1307    Tag: leunga-110_47-dijsktra
1308    Description:
1309    
1310    Bug fix from Carl Hauser:
1311    
1312    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1313    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1314    
1315    The problem is that following the expression A.update(dist,s,Num.zero)
1316    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1317    
1318    ----------------------------------------------------------------------
1319    Name: Matthias Blume (blume (at) tti - c (dot) org)
1320    Date: 2004/08/06 18:10:00 CDT
1321    Tag: blume-20040806-cmdline
1322    Description:
1323    
1324    Fiddled with handling of command-line options:
1325    
1326      * sml now quits after processing the command line
1327        if -H, -S, -h<n>, or -s<n> appears as the last
1328        command-line argument
1329      * a new option -q terminates the session when encountered on
1330        the command line; subsequent arguments will be ignored
1331      * bug fixes: short (erroneous) arguments are no longer ignored
1332        completely
1333    
1334    ----------------------------------------------------------------------
1335    Name: Allen Leung (leunga (at) reservoir (dot) com)
1336    Date: 2004/08/04 18:17:00 EST
1337    Tag: leunga-110_47-ppc-ibm-asm
1338    Description:
1339    
1340      - Added minimal IBM assembly syntax support for PowerPC.
1341    
1342      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1343        been altered in the runtime.  We need an automatic way to keep the file
1344        in sync.
1345    
1346    ----------------------------------------------------------------------
1347    Name: Matthias Blume (blume (at) tti - c (dot) org)
1348    Date: 2004/08/04 14:00:00 CDT
1349    Tag: Release_110_47
1350    Description:
1351    
1352    New working version (110.47).  NEW BOOTFILES!
1353    
1354    ----------------------------------------------------------------------
1355    Name: Matthias Blume (blume (at) tti - c (dot) org)
1356    Date: 2004/08/03 14:25:00 CDT
1357    Tag: blume-20040803-callingconv
1358    Description:
1359    
1360    Added low-level support for choosing C calling conventions by
1361    twiddling the type of rawccall.  (See
1362    src/compiler/Semant/types/cproto.sml for details.)
1363    
1364    ----------------------------------------------------------------------
1365    Name: Matthias Blume (blume (at) tti - c (dot) org)
1366    Date: 2004/08/02 15:55:00 CDT
1367    Tag: blume-20040802-backout
1368    Description:
1369    
1370    Backed out of change to win32-filesys.c.  The earlier patch to
1371    get_file_time caused CM to produce files with the wrong time stamp.
1372    
1373    ----------------------------------------------------------------------
1374    Name: Matthias Blume (blume (at) tti - c (dot) org)
1375    Date: 2004/08/02 14:45:00 CDT
1376    Tag: blume-20040802-nlffi-win32
1377    Description:
1378    
1379    Added NLFFI support for Win32, adapted from a patch provided by David
1380    Hansel.  This is currently completely untested.  Also, the issue
1381    concerning stdcall vs. ccall is still unresolved.
1382    
1383    ----------------------------------------------------------------------
1384    Name: Matthias Blume (blume (at) tti - c (dot) org)
1385    Date: 2004/07/30 17:55:00 CDT
1386    Tag: blume-20040730-various
1387    Description:
1388    
1389    Gearing up towards 110.47...
1390    
1391    - various minor bugfixes to ml-nlffigen
1392    - a beginning of a manual for nlffi
1393    
1394    - eliminated 'export name=value' in config/install.sh as this does
1395      not work with certain versions of /bin/sh
1396      (Thanks to David King at Motorola for catching this.)
1397    
1398    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1399      - added a test for tm==NULL to gmtime.c and localtime.c
1400      - applied patch for incorrect GetFileTime under win32
1401      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1402    
1403    ----------------------------------------------------------------------
1404    Name: Matthias Blume (blume (at) tti - c (dot) org)
1405    Date: 2004/07/21 18:20:00 CDT
1406    Tag: blume-20040721-nlffigen
1407    Description:
1408    
1409    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1410      for a typedef to an incomplete type, but leave out the "typ" member.
1411      (This is just for consistency.)
1412    - Started to produce what is supposed to become better (i.e., comprehensive)
1413      documentation of what ml-nlffigen does and produces.
1414    
1415    ----------------------------------------------------------------------
1416    Name: Matthias Blume (blume (at) tti - c (dot) org)
1417    Date: 2004/07/14 16:25:00 CDT
1418    Tag: blume-20040714-union
1419    Description:
1420    
1421    Added C_UNION to c-calls/c-types.sml and updated the machinery
1422    (ml-nlffigen, cproto.sml) that conveys C function interface
1423    information to the code generator.
1424    
1425    However, the actual architecture-specific implementation of function
1426    arguments and results that are C unions is still not implemented.
1427    
1428    ----------------------------------------------------------------------
1429    Name: Allen Leung (leunga (at) reservoir (dot) com)
1430    Date: 2004/07/14 14:38:00 EST
1431    Tag: leunga-110_46_1-ppc-lwzu
1432    Description:
1433    
1434        Added these instructions to the PowerPC architecture:
1435        LBZU(X), LHZU(X), LWZU(X),
1436        STWU(X), STFDU, STFSU
1437    
1438        etc...
1439    
1440        Note: I haven't added their instruction encoding into the description.
1441    
1442    ----------------------------------------------------------------------
1443    Name: Allen Leung (leunga (at) reservoir (dot) com)
1444    Date: 2004/07/13 15:04:00 EST
1445    Tag: leunga-110_46_1-ppc-lwarx
1446    Description:
1447    
1448       Added the two instructions LWARX and STWCX to the PowerPC
1449    instruction set.
1450    
1451       A (untested) rewrite of loop-structure.sml.  The old version
1452    is completely broken.
1453    
1454    ----------------------------------------------------------------------
1455    Name: Matthias Blume (blume (at) tti - c (dot) org)
1456    Date: 2004/07/13 13:50:00 CDT
1457    Tag: blume-20040713-nlffi
1458    Description:
1459    
1460    - use paramAlloc to report c-calls with too many arguments
1461      (for PPC version where parameter area is pre-allocated)
1462    - added ccall_maxargspace to machspec (to implement the above)
1463    - made "make" commend in CM's "make" tool configurable
1464    - added option (default: on) for passing the name of the SML/NJ's "bin"
1465      directory to "make"; the call looks like this:
1466    
1467         make <options> SMLNJ_BINDIR=<dir> <target>
1468    
1469      This can be used by the Makefile to, e.g., pick the "right" version
1470      of ml-nlffigen.
1471    - minor code tweaks
1472    
1473    ----------------------------------------------------------------------
1474    Name: Matthias Blume (blume (at) tti - c (dot) org)
1475    Date: 2004/07/12 22:50:00 CDT
1476    Tag: blume-110_46_1-macosx-nlffi
1477    Description:
1478    
1479    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1480    though.
1481    
1482    Note:  1. You have to make a new, clean build of the runtime system.
1483           2. There are new BOOTFILES, you have to use them!
1484              (Doing the bootstrap process yourself would be *very* painful!
1485               If you absolutely have to do it, build the system under
1486               a different architecture and then cross-compile.)
1487    
1488    Version bumped to 110.46.1 to account for runtime data format changes.
1489    
1490    ----------------------------------------------------------------------
1491    Name: Matthias Blume (blume (at) tti - c (dot) org)
1492    Date: 2004/06/18 14:30:00 CDT
1493    Tag: blume-20040618-unix
1494    Description:
1495    
1496    Changed the implementation of structure Unix so that the same stream
1497    is returned every time one of the {text,bin}{In,Out}streamOf functions
1498    is invoked on the same proc.  This is not what the spec currently
1499    says -- although IMO it arguably should.  (See discussion below.)
1500    
1501    ----------------------------------------------------------------------
1502    Name: Matthias Blume (blume (at) tti - c (dot) org)
1503    Date: 2004/06/17 18:15:00 CDT
1504    Tag: Release_110_46
1505    Description:
1506    
1507    New working version (110.46).  NEW BOOTFILES!
1508    
1509    ----------------------------------------------------------------------
1510    Name: Matthias Blume (blume (at) tti - c (dot) org)
1511    Date: 2004/06/17 17:20:00 CDT
1512    Tag: blume-20040617-timer-unix
1513    Description:
1514    
1515    Changed the interface of structures Timer and Unix to match the most
1516    recent Basis spec.
1517    
1518    In the case of Unix there still seems to be an open/weird issue:
1519    
1520       The {text,bin}{In,Out}streamOf functions are supposed to create
1521       fresh streams whenever they are called -- as opposed to have them
1522       return the same stream every time.  This design is supposed to
1523       prevent space leaks caused by proc values hanging on to streams.
1524    
1525       The reap function, on the other hand, is supposed to close the
1526       streams.  This cannot be done without having a handle on the
1527       stream in proc after all...
1528    
1529    I took the liberty to implement the following stopgap solution:
1530    
1531      The proc value hangs on to the most recently created stream(s).
1532      Reap closes those.  If either or both of the two streams hadn't
1533      been created at all yet, then reap will close the corresponding
1534      file descriptors directly.
1535    
1536    PS: I don't understand the original space leak argument anymore.  If
1537    a proc hangs on to the imperative stream, then I/O operations on those
1538    will advance the state of the cached stream and avoid the space leak.
1539    
1540    ----------------------------------------------------------------------
1541    Name: Matthias Blume (blume (at) tti - c (dot) org)
1542    Date: 2004/05/28 16:45:00 CDT
1543    Tag: blume-20040528-basis
1544    Description:
1545    
1546    Added signature PACK_REAL and exported functor PrimIO.
1547    
1548    ----------------------------------------------------------------------
1549    Name: Matthias Blume (blume (at) tti - c (dot) org)
1550    Date: 2004/05/25 16:00:00 CDT
1551    Tag: blume-20040525-group-owner
1552    Description:
1553    
1554    CM now ignores (but still accepts) the "owner" information in group
1555    descriptions.  The owner of a group is its next enclosing
1556    library. Each group must have a unique owner.  (There is a virtual
1557    "toplevel" library that owns groups which are not nested within a real
1558    library.)  Previously, each group had to explicitly declare its owner,
1559    and CM would check that such a declaration is correct.  The new scheme
1560    is to have CM check that for each group there is precisely one owning
1561    library.
1562    
1563    The advantage of the new scheme is that the programmer no longer needs
1564    to maintain the somewhat annoying owner information.  The downside is
1565    that CM cannot enforce the ownership rule across multiple runs of
1566    CM.make.  Fortunately, enclosing the same group in two different
1567    libraries A and B which are not part of the same program does not
1568    cause real problems.
1569    
1570    ----------------------------------------------------------------------
1571    Name: Matthias Blume (blume (at) tti - c (dot) org)
1572    Date: 2004/05/20 16:00:00 CDT
1573    Tag: blume-20040520-win32
1574    Description:
1575    
1576    Made the win32 version work again.  (Strangely, a misplaced comma had
1577    slipped into win32-process.c which prevented the runtime from being
1578    compiled correctly.)
1579    
1580    Also, included a minor addition to ml-build.bat analogous to what was
1581    done in blume-20040519-ml-build.
1582    
1583    ----------------------------------------------------------------------
1584    Name: Matthias Blume (blume (at) tti - c (dot) org)
1585    Date: 2004/05/19 22:10:00 CDT
1586    Tag: blume-20040519-ml-build
1587    Description:
1588    
1589    Arranged for ml-build to clean up after itself a little bit better.
1590    The script generates a temporary SML source file and compiles it using
1591    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1592    gets rid of those at the end, so they don't accumulate under .cm.
1593    
1594    This required a minor change to install.sh because the name of the
1595    metadata directory (default: .cm) is actually configurable at
1596    installation time.
1597    
1598    ----------------------------------------------------------------------
1599    Name: Matthias Blume (blume (at) tti - c (dot) org)
1600    Date: 2004/05/18 15:50:00 CDT
1601    Tag: blume-20040518-mkreader
1602    Description:
1603    
1604    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1605    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1606    
1607    ----------------------------------------------------------------------
1608    Name: Matthias Blume (blume (at) tti - c (dot) org)
1609    Date: 2004/05/11 14:35:00 CDT
1610    Tag: blume-20040511-win32sock
1611    Description:
1612    
1613    Added previously missing support for many socket-related functions
1614    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1615    for the voluminous patch!
1616    
1617    (I have not tested this patch under win32 yet.)
1618    
1619    Here is David's e-mail:
1620    
1621    Hi,
1622    
1623    Attached to this email you find a diff against sml/nj 110.45
1624    that will enable socket support under Windows.
1625    
1626    To apply the patch (using unix or cygwin)
1627    1) gunzip runtime.diff.gz
1628    2) "cd" into "src/runtime" in the source tree of a fresh
1629       110.45 installation.
1630    3) patch -p 1 < [your/path/to]runtime.diff
1631    
1632    The code compiles fine but has NOT yet been extensively tested.
1633    I only ran a few tests for basic socket client functionality
1634    (which worked fine).  Especially the functions that use ioctl
1635    are not tested at all and might not work (see below).
1636    
1637    I implemented this since we want to move to a newer version of sml/nj
1638    but need socket support in order to use it.  This is the first time I
1639    even had a look at the sml/nj source,  so please review my changes
1640    before making this part of the distribution!  Here are a few issues
1641    that I think might be better for someone to solve who is more
1642    familiar with the sml/nj source (and socket programming):
1643    
1644    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1645      exception since I could not figure out what the windows equivalent
1646      of these functions is
1647    
1648    - In sockets-osdep.h there are a some #include statements that are
1649      only used in a few files that include sockets-osdep.h
1650    
1651    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1652      does not process its return value since I don't know how to
1653      report an error upwards.
1654    
1655    - It would probably be good to have a call to WSACleanup() when
1656      the library is unloaded (if there is such a possibility).
1657      Otherwise I think Windows will take care of this automatically
1658      when the process finishes.
1659    
1660    - I used ioctlsocket() as a replacement for ioctl() but I have
1661      no idea if that is actually the proper replacement on Windows.
1662    
1663    - All these issues are marked in the code by "FIXME" comments.
1664    
1665    We use sml/nj extensively in our products and are quite happy
1666    with it.  I hope this contribution will help you.
1667    
1668    Keep up the good work!
1669    
1670    David
1671    
1672    ----------------------------------------------------------------------
1673    Name: Matthias Blume (blume (at) tti - c (dot) org)
1674    Date: 2004/05/11 14:20:00 CDT
1675    Tag: blume-20040511-installml
1676    Description:
1677    
1678    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1679    the patch.)
1680    
1681    ----------------------------------------------------------------------
1682    Name: Matthias Blume (blume (at) tti - c (dot) org)
1683    Date: 2004/05/11 14:05:00 CDT
1684    Tag: blume-20040511-nlffi-netbsd
1685    Description:
1686    
1687    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1688    the patch.)
1689    
1690    ----------------------------------------------------------------------
1691    Name: Matthias Blume (blume (at) tti - c (dot) org)
1692    Date: 2004/05/11 12:05:00 CDT
1693    Tag: blume-20040511-exports
1694    Description:
1695    
1696    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1697    export lists in compiler-related .cm-files.
1698    
1699    ----------------------------------------------------------------------
1700    Name: Matthias Blume (blume (at) tti - c (dot) org)
1701    Date: 2004/05/11 11:35:00 CDT
1702    Tag: blume-20040511-allsource
1703    Description:
1704    
1705    The installer now honors the "src-smlnj" target again, although its meaning
1706    has changed from "all sources required for the compiler" to "all sources
1707    the installer knows about".  In other words, if you enable "src-smlnj"
1708    in the "targets" file, then the installer will pull in sources for
1709    everything.  (Notice that this refers to source code only.  Compiled
1710    code is still only installed for modules that were requested explicitly
1711    or which are required for other modules that were requested explicitly.)
1712    
1713    ----------------------------------------------------------------------
1714    Name: Matthias Blume (blume (at) tti - c (dot) org)
1715    Date: 2004/04/23 17:40:00 CDT
1716    Tag: blume-20040423-ieee-scan
1717    Description:
1718    
1719    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1720    in the exponent calculation we get INF or ZERO (depending on the mantissa
1721    and the sign of the exponent).
1722    
1723    ----------------------------------------------------------------------
1724    Name: Matthias Blume (blume (at) tti - c (dot) org)
1725    Date: 2004/04/23 10:40:00 CDT
1726    Tag: blume-20040423-ml-build
1727    Description:
1728    
1729    The ml-build script now terminates with a non-0 status when something
1730    goes wrong.
1731    
1732    ----------------------------------------------------------------------
1733    Name: Matthias Blume (blume (at) tti - c (dot) org)
1734    Date: 2004/04/22 16:35:00 CDT
1735    Tag: blume-20040422-Option
1736    Description:
1737    
1738    Made exception Option to be the same as exception Option.Option
1739    (as it should be).
1740    
1741    ----------------------------------------------------------------------
1742    Name: Allen Leung (leunga (at) reservoir (dot) com)
1743    Date: 2004/03/19 14:40:00 EST
1744    Tag: leunga-20040319-cygwin-nlffi
1745    Description:
1746    
1747    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1748    of SML/NJ.  The problem is that
1749    
1750       lib = dlopen(NULL, ...)
1751       f   = dlsym(lib, "malloc");
1752    
1753    does not work on Windows unless we explicitly export symbols
1754    such as 'malloc' during linking.  We fixed this by explicitly
1755    exporting the required symbols with the magic gcc incantation:
1756    
1757        -Wl,--export-all cygwin.def
1758    
1759    where cygwin.def is a file containing all the symbols that
1760    we wish to export.
1761    
1762    I suspect this is a Windows problem and we'll have to
1763    do the same (somehow with windows compilers) when
1764    we build the native win32 version with the system
1765    calls LoadLibrary/GetProcAddress.
1766    
1767    ----------------------------------------------------------------------
1768    Name: Matthias Blume (blume (at) tti - c (dot) org)
1769    Date: 2004/03/04 16:35:00 CST
1770    Tag: blume-20040304-intinf-fmt
1771    Description:
1772    
1773    Fixed problem with IntInf.fmt (sign would show up on the right instead
1774    of on the left for BIN, OCT, and HEX).
1775    
1776    ----------------------------------------------------------------------
1777    Name: Matthias Blume (blume (at) tti - c (dot) org)
1778    Date: 2004/03/04 11:25:00 CST
1779    Tag: blume-20040304-symlinks
1780    Description:
1781    
1782    Fixed problem with installer script (unix only) where bin/ml-yacc and
1783    friends pointed (via symlinks) to absolute locations instead of just
1784    .run-sml.  This was reported by Vesa A Norrman.
1785    
1786    ----------------------------------------------------------------------
1787    Name: Matthias Blume (blume (at) tti - c (dot) org)
1788    Date: 2004/02/13 14:50:00 CST
1789    Tag: Release_110_45
1790    Description:
1791    
1792    New working version (110.45). New bootfiles.
1793    
1794    ----------------------------------------------------------------------
1795    Name: Matthias Blume (blume (at) tti - c (dot) org)
1796    Date: 2004/01/26 15:15:15 CST
1797    Tag: blume-20040126-toplevel
1798    Description:
1799    
1800    Improved handling of exceptions at the interactive toplevel.
1801    
1802    ----------------------------------------------------------------------
1803    Name: Matthias Blume (blume (at) tti - c (dot) org)
1804    Date: 2004/01/26 11:25:00
1805    Tag: blume-20040126-app
1806    Description:
1807    
1808    Type of top-level "app" corrected.
1809    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1810    
1811    ----------------------------------------------------------------------
1812    Name: Matthias Blume (blume (at) tti - c (dot) org)
1813    Date: 2003/11/18 17:10 CST
1814    Tag: blume-20031118-basis-fiddle
1815    Description:
1816    
1817    - changed Timer interface to what might become the spec
1818    - POSIX_FLAGS -> BIT_FLAGS according to spec
1819    - some other minor discrepancies wrt. spec eliminated
1820    
1821    ----------------------------------------------------------------------
1822    Name: Matthias Blume (blume (at) tti - c (dot) org)
1823    Date: 2003/11/06 12:00:00 CST
1824    Tag: Release_110_44
1825    Description:
1826    
1827    New working version (110.44). New bootfiles.
1828    
1829    ----------------------------------------------------------------------
1830    Name: Matthias Blume (blume (at) tti - c (dot) org)
1831    Date: 2003/11/04 11:50:00 CST
1832    Tag: blume-20031104-move-libraries
1833    Description:
1834    
1835    Eliminated the "dont_move_libraries" directive in config/targets.
1836    (The mechanism was broken and could not be fixed easily.  Moreover,
1837    there does not seem to be any reason not to move all libraries into
1838    lib during installation.  I originally implemented this directive as a
1839    backward-compatibility feature when I first introduced the new CM.
1840    Now that things have been stable for a long time and going back to the
1841    old CM is not an option, there is no reason to keep it around.)
1842    
1843    ----------------------------------------------------------------------
1844    Name: Matthias Blume (blume (at) tti - c (dot) org)
1845    Date: 2003/11/03 16:00:00 CST
1846    Tag: blume-20031103-installdir
1847    Description:
1848    
1849    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1850    Richards for pointing out the problem and providing the solution.)
1851    
1852    ----------------------------------------------------------------------
1853    Name: Matthias Blume (blume (at) tti - c (dot) org)
1854    Date: 2003/10/01 17:05:00 CDT
1855    Tag: blume-20031001-lal-mlrisc
1856    Description:
1857    
1858    MLRISC bug fix from Lal.
1859    
1860    ----------------------------------------------------------------------
1861    Name: Matthias Blume (blume (at) tti - c (dot) org)
1862    Date: 2003/09/30 16:10:00 CDT
1863    Tag: blume-20030930-primio-bat
1864    Description:
1865    
1866    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1867    2. Improved .bat files (for Win32 port) to make things work under Win95.
1868       (thanks to Aaron S. Hawley for this one)
1869    
1870    ----------------------------------------------------------------------
1871    Name: Matthias Blume (blume (at) tti - c (dot) org)
1872    Date: 2003/09/26 16:05:00 CDT
1873    Tag: blume-20030926-wrappriv
1874    Description:
1875    
1876    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1877    
1878    ----------------------------------------------------------------------
1879    Name: Matthias Blume (blume (at) tti - c (dot) org)
1880    Date: 2003/09/26 15:00:00 CDT
1881    Tag: blume-20030926-110_43_3
1882    Description:
1883    
1884    - additional cleanup
1885    - version number bump, NEW BOOTFILES
1886    
1887    ----------------------------------------------------------------------
1888    Name: Matthias Blume (blume (at) tti - c (dot) org)
1889    Date: 2003/09/26 12:00:00 CDT
1890    Tag: blume-20030926-ppautoload
1891    Description:
1892    
1893    I modified the read-eval-print loop so that the autoloader gets
1894    invoked whenever the prettyprinter tries to look up a symbol that
1895    is not currently defined in the toplevel environment but which
1896    appears in CM's autoload registry.  As a result, we see far fewer of
1897    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1898    
1899    In addition to this I tried to clean up some pieces of the Basis
1900    implementation (e.g., Socket, Word8Array) in order to prevent other
1901    instances of these ?.Foo.Bar.xxx names from being printed.
1902    
1903    The mechanism that picks names for types still needs some work, though.
1904    (Right now it seems that if there is a type A.t which is defined to
1905    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1906    "?.B.u" although the perhaps more sensible solution would be to use
1907    "A.t" in this case.  In other words, the prettyprinter should follow
1908    a chain of DEFtycs not farther than there are corresponding toplevel
1909    names in the current environment.)
1910    
1911    ----------------------------------------------------------------------
1912    Name: Matthias Blume (blume (at) tti - c (dot) org)
1913    Date: 2003/09/24 16:31:00 CDT
1914    Tag: blume-20030924-installer
1915    Description:
1916    
1917    Another installer tweak:  All the ML code for the installer is now
1918    compiled during CMB.make and put into a little library called
1919    $smlnj/installer.cm.  The installation then simply invokes
1920    
1921       sml -m $smlnj/installer.cm
1922    
1923    and everything happens automagically.
1924    
1925    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1926    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1927            and BINDIR.
1928    
1929    The new scheme guarantees that the ML code responsible for the installation
1930    is in sync with the APIs of the main system.  Also, the installer is
1931    somewhat faster because the installer script is precompiled.
1932    
1933    ----------------------------------------------------------------------
1934    Name: Matthias Blume (blume (at) tti - c (dot) org)
1935    Date: 2003/09/24 15:35:00 CDT
1936    Tag: blume-20030924-synsock
1937    Description:
1938    
1939    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1940    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1941    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1942    implementations of non-blocking operations from CML's Socket
1943    structure.
1944    
1945    ----------------------------------------------------------------------
1946    Name: Matthias Blume (blume (at) tti - c (dot) org)
1947    Date: 2003/09/24 15:10:05 CDT
1948    Tag: blume-20030924-sockets
1949    Description:
1950    
1951    1. Fixed SOCKET API and implementation to match Basis spec.
1952       This required changing the internal representation of sockets to one
1953       that remembers (for each socket file descriptor) whether it is currently
1954       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1955       call is made only if the state actually needs to change).
1956    
1957    2. OS-specific details of sockets were moved into separate files, thus
1958       making it possible to unify the bulk of the socket implementations
1959       between Unix and Win32.
1960    
1961    3. CML's socket API changed accordingly.
1962       (Note that we need to remove non-blocking functions from this API
1963       since they are redundant in the case of CML!)
1964    
1965    4. CML's socket implementation now makes use of non-blocking functions
1966       provided by Basis, thus removing all OS-dependent code from this part
1967       of CML.
1968    
1969    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1970    
1971    ----------------------------------------------------------------------
1972    Name: Matthias Blume (blume (at) tti - c (dot) org)
1973    Date: 2003/09/22 12:10:00 CDT
1974    Tag: blume-20030922-110_43_2
1975    Description:
1976    
1977    Made a new interim version and bootfiles for developer's bootstrapping
1978    convenience.
1979    
1980       110.43.2 -- NEW BOOTFILES
1981    
1982    ----------------------------------------------------------------------
1983    Name: Matthias Blume (blume (at) tti - c (dot) org)
1984    Date: 2003/09/19 15:55:00 CDT
1985    Tag: blume-20030919-cmdir
1986    Description:
1987    
1988    1. new-install.sh -> install.sh
1989    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1990    3. tweaked installer so that another name instead of .cm can be chosen
1991       at install time (by setting the CM_DIR_ARC environment variable
1992       during installation); once installation is complete, the name is
1993       fixed
1994    
1995    ----------------------------------------------------------------------
1996    Name: Matthias Blume (blume (at) tti - c (dot) org)
1997    Date: 2003/09/18 16:00:00 CDT
1998    Tag: blume-20030918-110_43_1
1999    Description:
2000    
2001    Made a new interim version and bootfiles for developer's bootstrapping
2002    convenience.
2003    
2004       110.43.1 -- NEW BOOTFILES
2005    
2006    ----------------------------------------------------------------------
2007    Name: Matthias Blume (blume (at) tti - c (dot) org)
2008    Date: 2003/09/18 15:20:00 CDT
2009    Tag: blume-20030918-misc
2010    Description:
2011    
2012    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
2013       this seems to be controversial at the moment) and used those in
2014       Posix.ProcEnv.times.
2015    
2016    2. Added Time.{from,to}Nanoseconds to Time.
2017    
2018    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
2019       For example, fromLargeInt never needs to look at more than 3 "big
2020       digits" to get its 53 bits of precision.
2021    
2022    ----------------------------------------------------------------------
2023    Name: Matthias Blume (blume (at) tti - c (dot) org)
2024    Date: 2003/09/17 16:30:00 CDT
2025    Tag: blume-20030917-real32-slices
2026    Description:
2027    
2028    Added an entry to the primitive environment
2029    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
2030    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
2031    
2032    Removed some of the "magic" constants in real64.sml and replaced them
2033    with code that generates these values from their corresponding
2034    integer counterparts.
2035    
2036    Made all(?) the slice-related changes to the Basis and made everything
2037    compile again...
2038    
2039    ----------------------------------------------------------------------
2040    Name: Matthias Blume (blume (at) tti - c (dot) org)
2041    Date: 2003/09/15 17:45:00 CDT
2042    Tag: blume-20030915-rbase
2043    Description:
2044    
2045    Fixed bug in Real.fromLargeInt.
2046    
2047    ----------------------------------------------------------------------
2048    Name: Matthias Blume (blume (at) tti - c (dot) org)
2049    Date: 2003/09/13 18:11:00 CDT
2050    Tag: blume-20030913-libinstall
2051    Description:
2052    
2053    Minor bugfix in config/libinstall (set anchor with path to
2054    standalone tool after installing it, otherwise libraries that
2055    need ml-lex or ml-yacc won't compile the first time the installer
2056    runs).
2057    
2058    ----------------------------------------------------------------------
2059    Name: Matthias Blume (blume (at) tti - c (dot) org)
2060    Date: 2003/09/12 11:45:00 CDT
2061    Tag: blume-20030912-various
2062    Description:
2063    
2064     - fixed bug in Real.toLargeInt
2065     - fixed bug in Posix.ProcEnv.times
2066     - changed inputLine functions to return an option
2067     - minor installer improvements / bugfixes
2068     - changed default @SMLalloc parameter for x86/celeron to 64k
2069    
2070    ----------------------------------------------------------------------
2071    Name: Matthias Blume (blume (at) tti - c (dot) org)
2072    Date: 2003/09/09 22:00:00 CDT
2073    Tag: Release_110_43
2074    Description:
2075    
2076        New working release 110.43.  New bootfiles.
2077    
2078    ----------------------------------------------------------------------
2079    Name: Matthias Blume (blume (at) tti - c (dot) org)
2080    Date: 2003/09/09 19:20:00 CDT
2081    Tag: blume-20030909-installer
2082    Description:
2083    
2084    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
2085    Modified config/install.bat to take advantage of it.  Also modified
2086    config/install.sh (and called it config/new-install.sh) to take advantage
2087    of it on Unix systems.  (The SML code is (supposed to be) platform-
2088    independent.)
2089    
2090    The installer can now install everything under Win32
2091    as well as under *nix as long as it compiles.
2092    
2093    Other changes:
2094    
2095      - made CML compile again under Win32
2096      - made eXene compile under Win32 (by providing a fake structure UnixSock
2097        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
2098      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
2099        same as type int;  under Win32 it isn't
2100      - fixed some slice-related problems in the win32-specific parts of CML
2101      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
2102        (under Win32, the volume name is case-insensitive, and the
2103        OS.Path code compares volume names for equality)
2104    
2105    ----------------------------------------------------------------------
2106    Name: Matthias Blume (blume (at) tti - c (dot) org)
2107    Date: 2003/09/08 11:55:00 CDT
2108    Tag: blume-20030908-fullpath
2109    Description:
2110    
2111    Made Win32 version of OS.FileSys.fullPath return current directory
2112    when given an empty string.  This is what the spec says, and incidentally,
2113    CM depends on it.  (CM otherwise goes into an infinite loop in certain
2114    cases when presented with the name of a non-existing .cm file.)
2115    
2116    ----------------------------------------------------------------------
2117    Name: Matthias Blume (blume (at) tti - c (dot) org)
2118    Date: 2003/09/04 16:30:00 CDT
2119    Tag: blume-20030905-slices-etc
2120    Description:
2121    
2122      1. Changed interface to vectors and arrays in Basis to match
2123         (draft) Basis spec.
2124      2. Added signatures and implementations of slices according to
2125         Basis spec.
2126      3. Edited source code throughout the system to make it compile again
2127         under 1. and 2.  (In some cases code had to be added to have it
2128         match the new signatures.)
2129      4. MLRISC should be backward-compatible: the copies of the originals
2130         of files that needed to change under 3. were retained, the .cm files
2131         check the compiler version number and use old versions when
2132         appropriate.
2133      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2134         dirstream -> string option (in accordance with Basis spec).
2135      6. When generating code that counts lines, ml-lex used function
2136         CharVector.foldli, taking advantage of its old interface.
2137         This has been replaced with the corresponding code from
2138         CharVectorSlice.  (html-lex must be re-lexed!)
2139      7. BitArray in smlnj-lib/Util has been extended/modified to match the
2140         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
2141      8. Removed temporary additions (fromInternal, toInternal) from the
2142         (now obsolete) IntInf in smlnj-lib/Util.
2143      9. Cleaned up structure Byte.
2144     10. Added localOffset, scan, and fromString to Date (according to spec).
2145         Cleaned/corrected implementation of Date.
2146         (Still need to check for correctness; implement better canonicalizeDate.)
2147     11. Added "scan" to signature IEEE_REAL.
2148     12. Some improvements to IntInf [in particular: efficiency-hack for
2149         mod and rem when second operand is 2 (for parity checks).]
2150     13. Changed representation of type Time.time, using a single IntInf.int
2151         value counting microseconds.  This considerably simplified the
2152         implementation of structure Time.  We now support negative time
2153         values; scan and fromString handle signs.
2154     14. Functor PrimIO now takes two additional arguments (VectorSlice and
2155         ArraySlice).
2156    
2157    ----------------------------------------------------------------------
2158    Name: Matthias Blume (blume (at) tti - c (dot) org)
2159    Date: 2003/08/28 17:00:00 CDT
2160    Tag: blume-20030828-intinf
2161    Description:
2162    
2163    This is a major update which comes with a version number bump
2164    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2165    and an implementation of IntInf in the Basis.
2166    
2167    There are a fairly large number of related changes and updates throughout
2168    the system:
2169    
2170    Basis:
2171            - Implemented IntInf.
2172            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
2173            - Added some missing Real64 operations, most notably Real.toLargeInt.
2174            - Added FixedInt as a synonym for Int32.
2175    
2176    compiler:
2177            * Added support for a built-in intinf type.
2178              - literals
2179              - pattern matching
2180              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2181              - overloading on literals and operations
2182    
2183            This required adding a primitive type intinf, some additional
2184            primops, and implementations for several non-trivial intinf
2185            operations in Core.  (The intinf type is completely abstract
2186            to the compiler; all operations get delegated back to the Core.)
2187    
2188            * Intinf equality is handled by polyequal.  However, the compiler
2189            does not print its usual warning in this case (since polyequal
2190            is the right thing to do there).
2191    
2192            * Improved the organization of structure InlineT.
2193    
2194            * A word about conversion primops:
2195            If conversions involving intinf do not cancel out during
2196            CPS contract, then the compiler must insert calls to Core functions.
2197            Since all core access must be resolved already during the FLINT
2198            translate phase, it would be too late a the time of CPS contract
2199            to add new Core calls.  For this reason, conversion primops
2200            for intinf carry two arguments:  1. the numeric argument that
2201            they are supposed to convert, and 2. the Core function that
2202            can help with this conversion if necessary.  If CPS contract
2203            eliminates a primop, then the associated Core function becomes
2204            dead and goes away.  Intinf conversion primops that do not get
2205            eliminated by CPS contract get rewritten into calls of their
2206            core functions by a separate, new phase.
2207    
2208    interactive system:
2209            - Control.Print.intinfDepth controls max length of intinf constants
2210            being printed. (Analogous to Control.Print.stringDepth.)
2211            - Cleanup in printutil and pputil: got rid of unused stuff and
2212            duplicates; replaced some of the code with code that makes better
2213            use of library functionality.
2214    
2215    CM:
2216            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2217            are no longer silent.
2218    
2219    CKIT:
2220            Fixed mismatched uses of Int32 and LargeInt.  I always decided
2221            in favor of LargeInt -- which is now the same as IntInf.
2222            CKIT-knowledgable people should check whether this is what's
2223            intended and otherwise change things back to using Int32 or
2224            FixedInt.
2225    
2226    Throughout the code:
2227            Started using IntInf.int literals and built-in operations
2228            (e.g., comparison with 0) where this seems appropriate.
2229    
2230    
2231    ----------------------------------------------------------------------
2232    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2233    Date: 2003/08/13 11:36:00 CDT
2234    Tag: dbm-20030813-mcz-merge1
2235    Description:
2236    
2237    Merging changes from the mcz-branch development branch into trunk.
2238    These changes involve replacement of the emulated old prettyprinter
2239    interface with direct use of the SML/NJ Lib PP library, and fixing
2240    of a couple of bugs (895, 1186) relating to error messages.  A new
2241    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2242    has been added.
2243    
2244    ----------------------------------------------------------------------
2245    Name: Matthias Blume (blume (at) tti - c (dot) org)
2246    Date: 2003/08/11 15:45:00 CDT
2247    Tag: blume-20030811-windows
2248    Description:
2249    
2250    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2251    
2252        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2253    
2254    This patch restores SML/NJ's ability to run under win32.  There are a
2255    number of changes, including fixes for several bugs that had gone
2256    unnoticed until now:
2257    
2258      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2259        trying to run the win32 version from within cygwin.)
2260    
2261      - There are a number of simple .bat scripts that substitute for their
2262        corresponding Unix shell-scripts.  (See below.)
2263    
2264      - The internals of ml-build have been modified slightly.  The main
2265        difference is that instead of calling ".link-sml" (or link-sml.bat)
2266        using OS.Process.system, the ML process delegates this task back
2267        to the script.  Otherwise problems arise in mixed environments such
2268        as Cygwin where scripts look and work like Unix scripts, but
2269        where OS.Process.system cannot run them.
2270    
2271      - In CM, the srcpath pickler used native pathname syntax -- which
2272        is incorrect in the case of cross-compilation.  The new pickle format
2273        is independent of platform-specific naming conventions.
2274    
2275      - Path configuration files (such as lib/pathconfig) can now choose
2276        between native and standard syntax.  Placing a line of the form
2277    
2278            standard!
2279    
2280        into the file causes all subsequent paths to be interpreted using
2281        CM standard pathname syntax (= Unix conventions); a line
2282    
2283            native!
2284    
2285        switches back to native style.  This was needed so that
2286        path config files can be written portably, see src/system/pathconfig.
2287    
2288      - Runtime system:
2289    
2290          - win32-filesys.c:  get_file_time and set_file_time now
2291            access modification time, not creation time.
2292    
2293          - I/O code made aware of new array representation.
2294    
2295          - Bug fixes in X86.prim.masm.
2296    
2297       - src/system/makeml made aware of win32. (For use under cygwin
2298            and other Unix-environments for windows.)
2299    
2300       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2301         which caused BinIO.inputAll to fail consistently.
2302    
2303    .bat scripts:
2304    
2305       Windows .bat scripts assume that SMLNJ_HOME is defined.
2306    
2307       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2308         applications (sml, ml-yacc, ml-lex).
2309       - ml-build.bat: analogous to ml-build.
2310       - config\install.bat:  Analogous to config/install.sh.  This requires
2311         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2312         (nmake etc. must be on the path, and vcvars32 must have been run.)
2313         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2314         and the bootfile hierarchy must have been unpacked under
2315         sml.boot.x86-win32.
2316         The script is very primitive and does a poor job at error checking.
2317         It only installs the base system, ml-lex, and ml-yacc.  No other
2318         libraries are being installed (i.e., you get only those that
2319         are part of the compiler.)
2320       - link-sml.bat:  analogous to .link-sml, but not currently used
2321    
2322    Unrelated bug fixes:
2323    
2324      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2325        types.
2326      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2327        otherwise.)
2328    
2329    ----------------------------------------------------------------------
2330    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2331    Date: 2003/06/17
2332    Tag: macqueen-20030617-bug895
2333    Description:
2334    
2335    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2336    Tag will be used for new development branch (mcz-branch) for
2337    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2338    
2339    ----------------------------------------------------------------------
2340    Name: Matthias Blume (blume (at) tti - c (dot) org)
2341    Date: 2003/05/27 16:55:00 CDT
2342    Tag: blume-20030527-polyeq
2343    Description:
2344    
2345    Tried to eliminated most cases of polymorphic equality.
2346    
2347    ----------------------------------------------------------------------
2348    Name: Matthias Blume (blume (at) tti - c (dot) org)
2349    Date: 2003/05/21 17:45:00 CDT
2350    Tag: blume-20030517-complete
2351    Description:
2352    
2353    Two changes:
2354    
2355      1. Added a flag for controlling whether non-exhaustive bindings will
2356         be treated as errors (default is false).
2357      2. Cleaned up the *entire* source tree so that CMB.make goes through
2358         without a single non-exhaustive match- or bind warning.
2359    
2360    ----------------------------------------------------------------------
2361    Name: Matthias Blume (blume (at) tti - c (dot) org)
2362    Date: 2003/05/17 10:20:00 CDT
2363    Tag: blume-20030517-absyn
2364    Description:
2365    
2366    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2367    
2368       This mainly affects the quality of error messages.  However, some
2369       of the code is now more straightforward than before. (Treatment of
2370       the above four constructs in translate.sml is much simpler than
2371       the "macro-expansion" that was going on before.  Plus, the mach-
2372       compiler no longer gets invoked just to be able to compile an
2373       if-expression.)
2374    
2375    2. The ErrorMsg.Error exception is now caught and absorbed by the
2376       interactive loop.
2377    
2378    ----------------------------------------------------------------------
2379    Name: Allen Leung
2380    Date: 2003/05/16 13:05:00 CDT
2381    Tag: leunga-20030516-cygwin-runtime
2382    Description:
2383    
2384        Ported the runtime system to cygwin, which uses the unix
2385    x86-unix bin files.  Missing/buggy features:
2386    
2387       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2388         the Cygwin library.
2389       o Ctrl-C handling may be flaky.
2390       o Windows system calls and Windows I/O are not supported.
2391    
2392       A new set of binfiles is located at:
2393    
2394         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2395    
2396       This is only needed for bootstrapping the cygwin version of smlnj.
2397       Other x86 versions can use the existing binfiles.
2398    
2399    ----------------------------------------------------------------------
2400    Name: Matthias Blume
2401    Date: 2003/04/08 15:42:00 CDT
2402    Tag: blume-20030408-listpair
2403    Description:
2404    
2405    1. Added a target 'mlrisc' to installer.
2406    
2407    2. Added missing elements to structure ListPair.
2408    
2409    ----------------------------------------------------------------------
2410    Name: Allen Leung
2411    Date: 2003/01/07 10:40:00 EST
2412    Tag: leunga-20030107-int-rem
2413    Description:
2414    
2415    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2416    The arguments to the SUBL instruction were swapped.
2417    
2418    ----------------------------------------------------------------------
2419    Name: Matthias Blume
2420    Date: 2002/12/12 16:25:00 EST
2421    Tag: blume-20021212-risc-ra
2422    Description:
2423    
2424    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2425    sent the RA into an infinite loop when floating point registers get
2426    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2427    110.37 and 110.38.)
2428    There was another set of potential problems related to the handling of
2429    MLRISC annotations (but those did not yet cause real problems, apparently).
2430    
2431    ----------------------------------------------------------------------
2432    Name: Matthias Blume
2433    Date: 2002/12/06 22:40:00 EST
2434    Tag: blume-20021206-cm-fileid
2435    Description:
2436    
2437    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2438    This fixes the problem of CM getting confused by files that suddenly
2439    change their identity (e.g., by getting unlinked and recreated by some
2440    text editor such as vi).  There might be a better/cheaper/cleaner way
2441    of doing this, but for now this will have to do.
2442    
2443    ----------------------------------------------------------------------
2444    Name: Matthias Blume
2445    Date: 2002/10/28 09:50:00 EST
2446    Tag: blume-20021028-typecheck
2447    Description:
2448    
2449    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2450    
2451    ----------------------------------------------------------------------
2452    Name: Matthias Blume
2453    Date: 2002/10/17 09:10:00 EDT
2454    Tag: Release_110_42
2455    Description:
2456    
2457    In good old tradition, there has been a slight hiccup so that we have
2458    to patch 110.42 after the fact.   The old release tag has been replaced
2459    (see below).
2460    
2461    The change solves a problem with two competing approaches the
2462    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2463    each other's way.
2464    
2465    This change only affects the runtime system code and the installer script.
2466    (No new bootfiles.)
2467    
2468    ----------------------------------------------------------------------
2469    Name: Matthias Blume
2470    Date: 2002/10/16 12:00:00 EDT
2471    Tag: Release_110_42_removed
2472    Description:
2473    
2474    New working release.  New bootfiles.
2475    
2476    ----------------------------------------------------------------------
2477    Name: Matthias Blume
2478    Date: 2002/10/10 13:10:00 EDT
2479    Tag: blume-20021010-ppc-divs
2480    Description:
2481    
2482    The mltree operator DIVS must be implemented with an overflow check on
2483    the PPC because the hardware indicates divide-by-zero using "overflow" as
2484    well.
2485    
2486    ----------------------------------------------------------------------
2487    Name: Matthias Blume
2488    Date: 2002/07/23 11:45:00
2489    Tag: blume-20020723-smlnj-home
2490    Description:
2491    
2492    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2493    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2494    CM_PATHCONFIG is also set), the path configuration file is assumed
2495    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2496    the entire tree to some other place and everything will "just work".
2497    
2498    (Companion commands such as ml-build and ml-makedepend also sense this
2499    variable.)
2500    
2501    ----------------------------------------------------------------------
2502    Name: Matthias Blume
2503    Date: 2002/07/12 21:19:00 EDT
2504    Tag: blume-20020712-liveness
2505    Description:
2506    
2507    Exported two useful "step" functions from liveness module (MLRISC).
2508    
2509    ----------------------------------------------------------------------
2510    Name: Matthias Blume
2511    Date: 2002/07/05 16:00 EDT
2512    Tag: Release_110_41
2513    Description:
2514    
2515    New working release.  New bootfiles.
2516    
2517    ----------------------------------------------------------------------
2518    Name: Matthias Blume
2519    Date: 2002/07/05 10:25:00 EDT
2520    Tag: blume-20020705-btimp
2521    Description:
2522    
2523    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2524    other clients can set up backtracing support.
2525    
2526    ----------------------------------------------------------------------
2527    Name: Matthias Blume
2528    Date: 2002/06/25 17:23:00 EDT
2529    Tag: blume-20020625-fpmax
2530    Description:
2531    
2532    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2533    case.  (The sense of the isNaN test was reversed -- which made min and
2534    max always return their first argument.)
2535    
2536    ----------------------------------------------------------------------
2537    Name: Matthias Blume
2538    Date: 2002/06/11
2539    Tag: blume-20020611-unixpath
2540    Description:
2541    
2542    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2543    
2544    ----------------------------------------------------------------------
2545    Name: Matthias Blume
2546    Date: 2002/06/10 16:35:00 EDT
2547    Tag: blume-20020610-ieeereal
2548    Description:
2549    
2550    I back-ported my implementation of IEEEReal.fromString from the
2551    idlbasis-devel branch so that we can test it.
2552    
2553    Another small change is that ppDec tries to give more information
2554    than just "<sig>" in the case of functors.  However, this code is
2555    broken in some mysterious way if the functor's body's signature
2556    has not been declared by ascription but gets inferred from the
2557    implementation.  This needs fixing...
2558    
2559    ----------------------------------------------------------------------
2560    Name: Matthias Blume
2561    Date: 2002/05/31
2562    Tag: blume-20020531-btrace-mode
2563    Description:
2564    
2565    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2566    the wayside when I switched over to using Controls everywhere.)
2567    
2568    ----------------------------------------------------------------------
2569    Name: Lal George
2570    Date: 2002/05/23 12:21:40 EDT
2571    Tag: george-20020523-visual-labels
2572    Description:
2573    
2574    Labels are now displayed in the graphical output to make
2575    the fall-through and target blocks obvious.
2576    
2577    ----------------------------------------------------------------------
2578    Name: Matthias Blume
2579    Date: 2002/05/22 11:03:00 EDT
2580    Tag: blume-20020522-shrink
2581    Description:
2582    
2583    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2584    situation that comes up when allocating huge arrays.
2585    
2586    ----------------------------------------------------------------------
2587    Name: Matthias Blume
2588    Date: 2002/05/21 16:00:00 EDT
2589    Tag: Release_110_40
2590    Description:
2591    
2592    New working release (110.40).  New bootfiles.
2593    
2594    [Also: John Reppy fixed GC bug 1131.]
2595    
2596    ----------------------------------------------------------------------
2597    Name: Matthias Blume
2598    Date: 2002/05/21 12:35:00 EDT
2599    Tag: blume-20020521-cmdoc
2600    Description:
2601    
2602    CM documentation update.
2603    
2604    ----------------------------------------------------------------------
2605    Name: Matthias Blume
2606    Date: 2002/05/21 10:55:00 EDT
2607    Tag: blume-20020521-misc
2608    Description:
2609    
2610    - John tweaked runtime to be silent on heap export (except when
2611      GC messages are on).
2612    - I added a few more things (cross-compiling versions of CMB) to
2613      config/preloads (as suggestions).
2614    
2615    ----------------------------------------------------------------------
2616    Name: Matthias Blume
2617    Date: 2002/05/20 22:25:00 EDT
2618    Tag: blume-20020520-controls
2619    Description:
2620    
2621    - Added ControlUtil structure to control-lib.cm.
2622    - Use it throughout.
2623    - Used Controls facility to define MLRISC controls (as opposed to
2624      registering MLRISC control ref cells with Controls after the fact)
2625    - Fixed messed-up controls priorities.
2626    
2627    * Removed again all the stuff from config/preloads that one wouldn't
2628      be able to preload at the time the initial heap image is built.
2629      (Many libraries, e.g., CML, do not exist yet at this time.  The
2630       only libraries that can be preloaded via config/preloads are those
2631       that come bundled with the bootfiles.)
2632    
2633    ----------------------------------------------------------------------
2634    Name: Matthias Blume
2635    Date: 2002/05/20 10:59:00 EDT
2636    Tag: blume-20020520-preloads
2637    Description:
2638    
2639    Added a lot of commented-out suggestions for things to be included
2640    in config/preloads.
2641    
2642    ----------------------------------------------------------------------
2643    Name: Allen Leung
2644    Date: 2002/05/18 14:20:00 EDT
2645    Tag: leunga-20020518-mdl
2646    Description:
2647    
2648    o Made the mdl tool stuff compile and run again.
2649    o I've disabled all the stuff that depends on RTL specifications; they
2650      are all badly broken anyway.
2651    
2652    ----------------------------------------------------------------------
2653    Name: Matthias Blume
2654    Date: 2002/05/17 16:49:00 EDT
2655    Tag: blume-20020517-controls
2656    Description:
2657    
2658    0. John Reppy made several modifications to the SML/NJ library.
2659       In particular, there is a shiny new controls-lib.cm.
2660    
2661    1. Pushed new controls interface through compiler so that everything
2662       compiles again.
2663    
2664    2. Added FormatComb and FORMAT_COMB to the CML version of the
2665       SML/NJ library (so that CML compiles again).
2666    
2667    3. Modified init scripts because XXX_DEFAULT environment variables
2668       are no longer with us.  (Boot-time initialization is now done
2669       using the same environment variables that are also used for
2670       startup-time initialization of controls.)
2671    
2672    ----------------------------------------------------------------------
2673    Name: Lal George
2674    Date: 2002/05/15 09:20:10 EDT
2675    Tag: george-20020515-pseudo-op-decls
2676    Description:
2677    
2678            All pseudo-ops emitted before the first segment declaration
2679    such as TEXT, DATA, and BSS directives are assumed to be global
2680    declarations and are emitted first in the assembly file. This is
2681    useful in a number of situations where one has pseudo-ops that are not
2682    specific to any segment, and also works around the constraint that one
2683    cannot have client pseudo-ops in the TEXT segment.
2684    
2685            Because no segment is associated with these declarations it is
2686    an error to allocate any space or objects before the first segment
2687    directive and an exception will be raised. However, we cannot make
2688    this check for client pseudo-ops.
2689    
2690            These top level declarations are a field in the CFG graph_info.
2691    In theory you can continue to add to this field after the CFG has been
2692    built -- provided you know what you are doing;-)
2693    
2694    ----------------------------------------------------------------------
2695    Name: Matthias Blume
2696    Date: 2002/05/13 16:40:00 EDT
2697    Tag: blume-20020513-pp-etc
2698    Description:
2699    
2700    A few minor bugfixes:
2701    
2702      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2703        (Bogus printouts for redefined bindings still occur.  Compiler
2704        bug should no longer occur now.  We need to redo the prettyprinter
2705        from scratch.)
2706    
2707      - CM pathname printer now also adds escape sequences for ( and )
2708    
2709      - commend and docu fixes for ml-nlffi
2710    
2711    ----------------------------------------------------------------------
2712    Name: Matthias Blume
2713    Date: 2002/05/10 16:40:00 EDT
2714    Tag: blume-20020510-erg-textio
2715    Description:
2716    
2717    Applied the following bugfix provided by Emden Gansner:
2718    
2719        Output is corrupted when outputSubstr is used rather than output.
2720    
2721        The problem occurs when a substring
2722    
2723            ss = (s, dataStart, dataLen)
2724    
2725        where dataStart > 0, fills a stream buffer with avail bytes left.
2726        avail bytes of s, starting at index dataStart, are copied into the
2727        buffer, the buffer is flushed, and then the remaining dataLen-avail
2728        bytes of ss are copied into the beginning of the buffer. Instead of
2729        starting this copy at index dataStart+avail in s, the current code
2730        starts the copy at index avail.
2731    
2732        Fix:
2733        In text-io-fn.sml, change line 695 from
2734             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2735        to
2736             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2737    
2738    ----------------------------------------------------------------------
2739    Name: Matthias Blume
2740    Date: 2002/04/12 13:55:00 EDT
2741    Tag: blume-20020412-assyntax
2742    Description:
2743    
2744    1. Grabbed newer assyntax.h from the XFree86 project.
2745    2. Fiddled with how to compile X86.prim.asm without warnings.
2746    3. (Very) Minor cleanup in CM.
2747    
2748    ----------------------------------------------------------------------
2749    Name: Matthias Blume
2750    Date: 2002/04/01 (no joke!) 17:07:00 EST
2751    Tag: blume-20020401-x86div
2752    Description:
2753    
2754    Added full support for div/mod/rem/quot on the x86, using the machine
2755    instruction's two results (without clumsily recomputing the remainder)
2756    directly where appropriate.
2757    
2758    Some more extensive power-of-two support was added to the x86 instruction
2759    selector (avoiding expensive divs, mods, and muls where they can be
2760    replaced with cheaper shifts and masks).  However, this sort of thing
2761    ought to be done earlier, e.g., within the CPS optimizer so that
2762    all architectures benefit from it.
2763    
2764    The compiler compiles to a fixed point, but changes might be somewhat
2765    fragile nevertheless.  Please, report any strange things that you might
2766    see wrt. div/mod/quot/rem...
2767    
2768    ----------------------------------------------------------------------
2769    Name: Matthias Blume
2770    Date: 2002/03/29 17:22:00
2771    Tag: blume-20020329-div
2772    Description:
2773    
2774    Fixed my broken div/mod logic.  Unfortunately, this means that the
2775    inline code for div/mod now has one more comparison than before.
2776    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2777    The problem was with quotient = 0, because that alone does not tell
2778    us which way the rounding went.  One then has to look at whether
2779    remainder and divisor have the same sign...  :(
2780    
2781    Anyway, I replaced the bootfiles with fresh ones...
2782    
2783    ----------------------------------------------------------------------
2784    Name: Matthias Blume
2785    Date: 2002/03/29 14:10:00 EST
2786    Tag: blume-20020329-inlprims
2787    Description:
2788    
2789    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2790    
2791    Primops have changed. This means that the bin/boot-file formats have
2792    changed as well.
2793    
2794    To make sure that there is no confusion, I made a new version.
2795    
2796    
2797    CHANGES:
2798    
2799    * removed REMT from mltree (remainder should never overflow).
2800    
2801    * added primops to deal with divisions of all flavors to the frontend
2802    
2803    * handled these primops all the way through so they map to their respective
2804      MLRISC support
2805    
2806    * used these primops in the implementation of Int, Int32, Word, Word32
2807    
2808    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2809    
2810    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2811    
2812    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2813      floating point
2814    
2815    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2816      and Real (but Real.abs maps to a separate floating-point-only primop)
2817    
2818    
2819    TODO items:
2820    
2821    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2822      instructions because the machine instruction encoder cannot handle
2823      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2824      REM in the first place, and REM is supported by the x86 machine coder.)
2825    
2826    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2827      selection phase.  (The two can be streamlined because the hardware
2828      delivers both quotient and remainder at the same time anyway.)
2829    
2830    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2831      (Currently the behavior is inconsistent both across architectures and
2832      wrt. the draft Basis spec.)
2833    
2834    * Word8 should eventually be handled natively, too.
2835    
2836    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2837      as if there currently is no execution path that could trigger it in
2838      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2839      hold for things like multiplication and division.)
2840    
2841    ----------------------------------------------------------------------
2842    Name: Matthias Blume
2843    Date: 2002/03/27 16:27:00 EST
2844    Tag: blume-20020327-mlrisc-divisions
2845    Description:
2846    
2847    Added support for all four division operations (ML's div, mod, quot,
2848    and rem) to MLRISC.  In the course of doing so, I also rationalized
2849    the naming (no more annoying switch-around of DIV and QUOT), by
2850    parameterizing the operation by div_rounding_mode (which can be either
2851    DIV_TO_ZERO or DIV_TO_NEGINF).
2852    
2853    The generic MLTreeGen functor takes care of compiling all four
2854    operations down to only round-to-zero div.
2855    
2856    Missing pieces:
2857    
2858      * Doing something smarter than relying on MLTreeGen on architectures
2859        like, e.g., the x86 where hardware division delivers both quotient and
2860        remainder at the same time.  With this, the implementation of the
2861        round-to-neginf operations could be further streamlined.
2862    
2863      * Remove inlining support for div/mod/rem from the frontend and replace it
2864        with primops that get carried through to the backend.  Do this for all
2865        int and word types.
2866    
2867    ----------------------------------------------------------------------
2868    Name: Matthias Blume
2869    Date: 2002/03/25 17:25:00 EST
2870    Tag: blume-20020325-divmod
2871    Description:
2872    
2873    I improved (hopefully without breaking them) the implementation of Int.div,
2874    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2875    advantage of the following observations:
2876    
2877      Let  q = x quot y      r = x rem y
2878           d = x div  y      m = x mod y
2879    
2880    where "quot" is the round-to-zero version of integer division that
2881    hardware usually provides.  Then we have:
2882    
2883         r = x - q * y        where neither the * nor the - will overflow
2884         d = if q >= 0 orelse x = q * y then q else q - 1
2885                              where neither the * nor the - will overflow
2886         m = if q >= 0 orelse r = 0 then r else r + y
2887                              where the + will not overflow
2888    
2889    This results in substantial simplification of the generated code.
2890    The following table shows the number of CFG nodes and edges generated
2891    for
2892            fun f (x, y) = x OPER y
2893            (* with OPER \in div, mod, quot, rem *)
2894    
2895    
2896        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2897        --------------------------------------------------------
2898         div |         24 |         39 |         12 |         16
2899         mod |         41 |         71 |         12 |         16
2900        quot |          8 |         10 |          8 |         10
2901         rem |         10 |         14 |          8 |         10
2902    
2903    
2904    ----------------------------------------------------------------------
2905    Name: Matthias Blume
2906    Date: 2002/03/25 22:06:00 EST
2907    Tag: blume-20020325-cprotobug
2908    Description:
2909    
2910    Fixed a bug in cproto (c prototype decoder).
2911    
2912    ----------------------------------------------------------------------
2913    Name: Matthias Blume
2914    Date: 2002/03/25 16:00:00 EST
2915    Tag: blume-20020325-raw-primops
2916    Description:
2917    
2918    I did some cleanup to Allen's new primop code and
2919    replaced yesterday's bootfiles with new ones.
2920    (But they are stored in the same place.)
2921    
2922    ----------------------------------------------------------------------
2923    Name: Matthias Blume
2924    Date: 2002/03/24 22:40:00 EST
2925    Tag: blume-20020324-bootfiles
2926    Description:
2927    
2928    Made the bootfiles that Allen asked for.
2929    
2930    ----------------------------------------------------------------------
2931    Name: Allen Leung
2932    Date: 2002/03/23 15:50:00 EST
2933    Tag: leunga-20020323-flint-cps-rcc-primops
2934    Description:
2935    
2936      1. Changes to FLINT primops:
2937    
2938        (* make a call to a C-function;
2939         * The primop carries C function prototype information and specifies
2940         * which of its (ML-) arguments are floating point. C prototype
2941         * information is for use by the backend, ML information is for
2942         * use by the CPS converter. *)
2943      | RAW_CCALL of { c_proto: CTypes.c_proto,
2944                       ml_args: ccall_type list,
2945                       ml_res_opt: ccall_type option,
2946                       reentrant : bool
2947                     } option
2948       (* Allocate uninitialized storage on the heap.
2949        * The record is meant to hold short-lived C objects, i.e., they
2950        * are not ML pointers.  With the tag, the representation is
2951        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2952        *)
2953      | RAW_RECORD of {tag:bool,sz:int}
2954      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2955    
2956      2.  These CPS primops are now overloaded:
2957    
2958           rawload of {kind:numkind}
2959           rawstore of {kind:numkind}
2960    
2961          The one argument form is:
2962    
2963             rawload {kind} address
2964    
2965          The two argument form is:
2966    
2967             rawload {kind} [ml object, byte-offset]
2968    
2969      3. RAW_CCALL/RCC now takes two extra arguments:
2970    
2971         a. The first is whether the C call is reentrant, i.e., whether
2972            ML state should be saved and restored.
2973         b. The second argument is a string argument specifying the name of
2974            library and the C function.
2975    
2976         These things are currently not handled in the code generator, yet.
2977    
2978      4. In CProto,
2979    
2980         An encoding type of "bool" means "ml object" and is mapped into
2981         C prototype of PTR.  Note that "bool" is different than "string",
2982         even though "string" is also mapped into PTR, because "bool"
2983         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2984    
2985      5. Pickler/unpicker
2986    
2987         Changed to handle RAW_RECORD and newest RAW_CCALL
2988    
2989      6. MLRiscGen,
2990    
2991         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2992         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2993            in the file CodeGen/cpscompile/cps-c-calls.sml
2994    
2995      7. Added the conditional move operator
2996    
2997             condmove of branch
2998    
2999         to cps.  Generation of this is still buggy so it is currently
3000         disabled.
3001    
3002    ----------------------------------------------------------------------
3003    Name: Lal George
3004    Date: 2002/03/22 14:18:25 EST
3005    Tag: george-20020322-cps-branch-prob
3006    Description:
3007    
3008    Implemented the Ball-Larus branch prediction-heuristics, and
3009    incorporated graphical viewers for control flow graphs.
3010    
3011    Ball-Larus Heuristics:
3012    ---------------------
3013    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
3014    
3015    By design it uses the Dempster-Shafer theory for combining
3016    probabilities.  For example, in the function:
3017    
3018        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
3019    
3020    the ball-larus heuristics predicts that the n=0 is unlikely
3021    (OH-heuristic), and the 'then' branch is unlikely because of the
3022    RH-heuristic -- giving the 'then' branch an even lower combined
3023    probability using the Dempster-Shafer theory.
3024    
3025    Finally, John Reppy's loop analysis in MLRISC, further lowers the
3026    probability of the 'then' branch because of the loop in the else
3027    branch.
3028    
3029    
3030    Graphical Viewing:
3031    ------------------
3032    I merely plugged in Allen's graphical viewers into the compiler. The
3033    additional code is not much. At the top level, saying:
3034    
3035            Control.MLRISC.getFlag "cfg-graphical-view" := true;
3036    
3037    will display the graphical view of the control flow graph just before
3038    back-patching.  daVinci must be in your path for this to work. If
3039    daVinci is not available, then the default viewer can be changed
3040    using:
3041    
3042            Control.MLRISC.getString "viewer"
3043    
3044    which can be set to "dot" or "vcg" for the corresponding viewers. Of
3045    course, these viewers must be in your path.
3046    
3047    The above will display the compilation unit at the level of clusters,
3048    many of which are small, boring, and un-interesting. Also setting:
3049    
3050            Control.MLRISC.getInt "cfg-graphical-view_size"
3051    
3052    will display clusters that are larger than the value set by the above.
3053    
3054    
3055    ----------------------------------------------------------------------
3056    Name: Matthias Blume
3057    Date: 2002/03/21 22:20:00 EST
3058    Tag: blume-20020321-kmp-bugfix
3059    Description:
3060    
3061    Changed the interface to the KMP routine in PreString and fixed
3062    a minor bug in one place where it was used.
3063    
3064    ----------------------------------------------------------------------
3065    Name: Allen Leung
3066    Date: 2002/03/21 20:30:00 EST
3067    Tag: leunga-20020321-cfg
3068    Description:
3069    
3070      Fixed a potential problem in cfg edge splitting.
3071    
3072    ----------------------------------------------------------------------
3073    Name: Allen Leung
3074    Date: 2002/03/21 17:15:00 EST
3075    Tag: leunga-20020321-x86-fp-cfg
3076    Description:
3077    
3078      1. Recoded the buggy parts of x86-fp.
3079    
3080         a. All the block reordering code has been removed.
3081            We now depend on the block placement phases to do this work.
3082    
3083         b. Critical edge splitting code has been simplified and moved into the
3084            CFG modules, as where they belong.
3085    
3086         Both of these were quite buggy and complex.  The code is now much, much
3087         simpler.
3088    
3089      2. X86 backend.
3090    
3091         a. Added instructions for 64-bit support.  Instruction selection for
3092            64-bit has not been committed, however, since that
3093            requires changes to MLTREE which haven't been approved by
3094            Lal and John.
3095    
3096         b. Added support for FUCOMI and FUCOMIP when generating code for
3097            PentiumPro and above.  We only generate these instructions in
3098            the fast-fp mode.
3099    
3100         c. Added cases for JP and JNP in X86FreqProps.
3101    
3102      3. CFG
3103    
3104         CFG now has a bunch of methods for edge splitting and merging.
3105    
3106      4. Machine description.
3107    
3108         John's simplification of MLTREE_BASIS.fcond broke a few machine
3109         description things:
3110    
3111         rtl-build.{sig,sml} and hppa.mdl fixed.
3112    
3113         NOTE: the machine description stuff in the repository is still broken.
3114               Again, I can't put my fixes in because that involves
3115               changes to MLTREE.
3116    
3117    ----------------------------------------------------------------------
3118    Name: Matthias Blume
3119    Date: 2002/03/20 15:55:00 EST
3120    Tag: blume-20020320-kmp
3121    Description:
3122    
3123    Implemented Knuth-Morris-Pratt string matching in PreString and used
3124    it for String.isSubstring, Substring.isSubstring, and
3125    Substring.position.
3126    
3127    (Might need some stress-testing.  Simple examples worked fine.)
3128    
3129    ----------------------------------------------------------------------
3130    Name: Matthias Blume
3131    Date: 2002/03/19 16:37:00 EST
3132    Tag: blume-20020319-witnesses
3133    Description:
3134    
3135    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
3136    
3137    This implements a generic mechanism for changing constness qualifiers
3138    anywhere within big C types without resorting to outright "casts".
3139    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
3140    modify the constness at the outermost level.)
3141    The implementation of "convert" is based on the idea of "witness"
3142    values -- values that are not used by the operation but whose types
3143    "testify" to their applicability.  On the implementation side, "convert"
3144    is simply a projection (returning its second curried argument).  With
3145    cross-module inlining, it should not result in any machine code being
3146    generated.
3147    
3148    ----------------------------------------------------------------------
3149    Name: Matthias Blume
3150    Date: 2002/03/15 16:40:00 EST
3151    Tag: blume-20020315-basis
3152    Description:
3153    
3154    Provided (preliminary?) implementations for
3155    
3156      {String,Substring}.{concatWith,isSuffix,isSubstring}
3157    
3158    and
3159    
3160      Substring.full
3161    
3162    Those are in the Basis spec but they were missing in SML/NJ.
3163    
3164    ----------------------------------------------------------------------
3165    Name: Matthias Blume
3166    Date: 2002/03/14 21:30:00 EST
3167    Tag: blume-20020314-controls
3168    Description:
3169    
3170    Controls:
3171    ---------
3172    
3173    1. Factored out the recently-added Controls : CONTROLS stuff and put
3174       it into its own library $/controls-lib.cm.  The source tree for
3175       this is under src/smlnj-lib/Controls.
3176    
3177    2. Changed the names of types and functions in this interface, so they
3178       make a bit more "sense":
3179    
3180          module -> registry
3181          'a registry -> 'a group
3182    
3183    3. The interface now deals in ref cells only.  The getter/setter interface
3184       is (mostly) gone.
3185    
3186    4. Added a function that lets one register an already-existing ref cell.
3187    
3188    5. Made the corresponding modifications to the rest of the code so that
3189       everything compiles again.
3190    
3191    6. Changed the implementation of Controls.MLRISC back to something closer
3192       to the original.  In particular, this module (and therefore MLRISC)
3193       does not depend on Controls.  There now is some link-time code in
3194       int-sys.sml that registers the MLRISC controls with the Controls
3195       module.
3196    
3197    CM:
3198    ---
3199    
3200      * One can now specify the lambda-split aggressiveness in init.cmi.
3201    
3202    ----------------------------------------------------------------------
3203    Name: Allen Leung
3204    Date: 2002/03/13 17:30:00 EST
3205    Tag: leunga-20020313-x86-fp-unary
3206    Description:
3207    
3208    Bug fix for:
3209    
3210    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3211    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3212    > - fun f(x,(y,z)) = Real.~ y;
3213    > [autoloading]
3214    > [autoloading done]
3215    >       fchsl   (%eax), 184(%esp)
3216    > Error: MLRisc bug: X86MCEmitter.emitInstr
3217    >
3218    > uncaught exception Error
3219    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3220    
3221    The problem was that the code generator did not generate any fp registers
3222    in this case, and the ra didn't know that it needed to run the X86FP phase to
3223    translate the pseudo fp instruction.   This only happened with unary fp
3224    operators in certain situations.
3225    
3226    ----------------------------------------------------------------------
3227    Name: Matthias Blume
3228    Date: 2002/03/13 14:00:00 EST
3229    Tag: blume-20020313-overload-etc
3230    Description:
3231    
3232    1. Added _overload as a synonym for overload for backward compatibility.
3233       (Control.overloadKW must be true for either version to be accepted.)
3234    
3235    2. Fixed bug in install script that caused more things to be installed
3236       than what was requested in config/targets.
3237    
3238    3. Made CM aware of the (_)overload construct so that autoloading
3239       works.
3240    
3241    ----------------------------------------------------------------------
3242    Name: Matthias Blume
3243    Date: 2002/03/12 22:03:00 EST
3244    Tag: blume-20020312-url
3245    Description:
3246    
3247    Forgot to update BOOT and srcarchiveurl.
3248    
3249    ----------------------------------------------------------------------
3250    Name: Matthias Blume
3251    Date: 2002/03/12 17:30:00 EST
3252    Tag: blume-20020312-version110392
3253    Description:
3254    
3255    Yet another version number bump (because of small changes to the
3256    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3257    
3258    Changes:
3259    
3260      The new pid generation scheme described a few weeks ago was overly
3261      complicated.  I implemented a new mechanism that is simpler and
3262      provides a bit more "stability":  Once CM has seen a compilation
3263      unit, it keeps its identity constant (as long as you do not delete
3264      those crucial CM/GUID/* files).  This means that when you change
3265      an interface, compile, then go back to the old interface, and
3266      compile again, you arrive at the original pid.
3267    
3268      There now also is a mechanism that instructs CM to use the plain
3269      environment hash as a module's pid (effectively making its GUID
3270      the empty string).  For this, "noguid" must be specified as an
3271      option to the .sml file in question within its .cm file.
3272      This is most useful for code that is being generated by tools such
3273      as ml-nlffigen (because during development programmers tend to
3274      erase the tool's entire output directory tree including CM's cached
3275      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3276      revert to the old, broken behavior of SML/NJ, but in specific cases
3277      where there is no danger of interface confusion, its use is ok
3278      (I think).
3279    
3280      ml-nlffigen by default generates "noguid" annotations.  They can be
3281      turned off by specifying -guid in its command line.
3282    
3283    ----------------------------------------------------------------------
3284    Name: Lal George
3285    Date: 2002/03/12 12 14:42:36 EST
3286    Tag: george-20020312-frequency-computation
3287    Description:
3288    
3289    Integrated jump chaining and static block frequency into the
3290    compiler. More details and numbers later.
3291    
3292    ----------------------------------------------------------------------
3293    Name: Lal George
3294    Date: 2002/03/11 11 22:38:53 EST
3295    Tag: george-20020311-jump-chain-elim
3296    Description:
3297    
3298    Tested the jump chain elimination on all architectures (except the
3299    hppa).  This is on by default right now and is profitable for the
3300    alpha and x86, however, it may not be profitable for the sparc and ppc
3301    when compiling the compiler.
3302    
3303    The gc test will typically jump to a label at the end of the cluster,
3304    where there is another jump to an external cluster containing the actual
3305    code to invoke gc. This is to allow factoring of common gc invocation
3306    sequences. That is to say, we generate:
3307    
3308            f:
3309               testgc
3310               ja   L1      % jump if above to L1
3311    
3312            L1:
3313               jmp L2
3314    
3315    
3316    After jump chain elimination the 'ja L1' instructions is converted to
3317    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3318    up being implemented in their long form (if L2 is far away) using:
3319    
3320            jbe     L3      % jump if below or equal to L3
3321            jmp     L2
3322         L3:
3323            ...
3324    
3325    
3326    For large compilation units L2  may be far away.
3327    
3328    
3329    ----------------------------------------------------------------------
3330    Name: Matthias Blume
3331    Date: 2002/03/11 13:30:00 EST
3332    Tag: blume-20020311-mltreeeval
3333    Description:
3334    
3335    A functor parameter was missing.
3336    
3337    ----------------------------------------------------------------------
3338    Name: Allen Leung
3339    Date: 2002/03/11 10:30:00 EST
3340    Tag: leunga-20020311-runtime-string0
3341    Description:
3342    
3343       The representation of the empty string now points to a
3344    legal null terminated C string instead of unit.  It is now possible
3345    to convert an ML string into C string with InlineT.CharVector.getData.
3346    This compiles into one single machine instruction.
3347    
3348    ----------------------------------------------------------------------
3349    Name: Allen Leung
3350    Date: 2002/03/10 23:55:00 EST
3351    Tag: leunga-20020310-x86-call
3352    Description:
3353    
3354       Added machine generation for CALL instruction (relative displacement mode)
3355    
3356    ----------------------------------------------------------------------
3357    Name: Matthias Blume
3358    Date: 2002/03/08 16:05:00
3359    Tag: blume-20020308-entrypoints
3360    Description:
3361    
3362    Version number bumped to 110.39.1.  NEW BOOTFILES!
3363    
3364    Entrypoints: non-zero offset into a code object where execution should begin.
3365    
3366    - Added the notion of an entrypoint to CodeObj.
3367    - Added reading/writing of entrypoint info to Binfile.
3368    - Made runtime system bootloader aware of entrypoints.
3369    - Use the address of the label of the first function given to mlriscGen
3370      as the entrypoint.  This address is currently always 0, but it will
3371      not be 0 once we turn on block placement.
3372    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3373      with entry points) from mlriscGen.
3374    
3375    ----------------------------------------------------------------------
3376    Name: Allen Leung
3377    Date: 2002/03/07 20:45:00 EST
3378    Tag: leunga-20020307-x86-cmov
3379    Description:
3380    
3381       Bug fixes for CMOVcc on x86.
3382    
3383       1. Added machine code generation for CMOVcc
3384       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3385       3. CMOVcc cannot have an immediate operand as argument.
3386    
3387    ----------------------------------------------------------------------
3388    Name: Matthias Blume
3389    Date: 2002/03/07 16:15:00 EST
3390    Tag: blume-20020307-controls
3391    Description:
3392    
3393    This is a very large but mostly boring patch which makes (almost)
3394    every tuneable compiler knob (i.e., pretty much everything under
3395    Control.* plus a few other things) configurable via both the command
3396    line and environment variables in the style CM did its configuration
3397    until now.
3398    
3399    Try starting sml with '-h' (or, if you are brave, '-H')
3400    
3401    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3402    implements the underlying generic mechanism.
3403    
3404    The interface to some of the existing such facilities has changed somewhat.
3405    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3406    (The getFoo interface is still there for backward-compatibility, but its
3407    use is deprecated.)
3408    
3409    The ml-build script passes -Cxxx=yyy command-line arguments through so
3410    that one can now twiddle the compiler settings when using this "batch"
3411    compiler.
3412    
3413    TODO items:
3414    
3415    We should go through and throw out all controls that are no longer
3416    connected to anything.  Moreover, we should go through and provide
3417    meaningful (and correct!) documentation strings for those controls
3418    that still are connected.
3419    
3420    Currently, multiple calls to Controls.new are accepted (only the first
3421    has any effect).  Eventually we should make sure that every control
3422    is being made (via Controls.new) exactly once.  Future access can then
3423    be done using Controls.acc.
3424    
3425    Finally, it would probably be a good idea to use the getter-setter
3426    interface to controls rather than ref cells.  For the time being, both
3427    styles are provided by the Controls module, but getter-setter pairs are
3428    better if thread-safety is of any concern because they can be wrapped.
3429    
3430    *****************************************
3431    
3432    One bug fix: The function blockPlacement in three of the MLRISC
3433    backpatch files used to be hard-wired to one of two possibilities at
3434    link time (according to the value of the placementFlag).  But (I
3435    think) it should rather sense the flag every time.
3436    
3437    *****************************************
3438    
3439    Other assorted changes (by other people who did not supply a HISTORY entry):
3440    
3441    1. the cross-module inliner now works much better (Monnier)
3442    2. representation of weights, frequencies, and probabilities in MLRISC
3443       changed in preparation of using those for weighted block placement
3444       (Reppy, George)
3445    
3446    ----------------------------------------------------------------------
3447    Name: Lal George
3448    Date: 2002/03/07 14:44:24 EST 2002
3449    Tag: george-20020307-weighted-block-placement
3450    
3451    Tested the weighted block placement optimization on all architectures
3452    (except the hppa) using AMPL to generate the block and edge frequencies.
3453    Changes were required in the machine properties to correctly
3454    categorize trap instructions. There is an MLRISC flag
3455    "weighted-block-placement" that can be used to enable weighted block
3456    placement, but this will be ineffective without block/edge
3457    frequencies (coming soon).
3458    
3459    
3460    ----------------------------------------------------------------------
3461    Name: Lal George
3462    Date: 2002/03/05 17:24:48 EST
3463    Tag: george-20020305-linkage-cluster
3464    
3465    In order to support the block placement optimization, a new cluster
3466    is generated as the very first cluster (called the linkage cluster).
3467    It contains a single jump to the 'real' entry point for the compilation
3468    unit. Block placement has no effect on the linkage cluster itself, but
3469    all the other clusters  have full freedom in the manner in which they
3470    reorder blocks or functions.
3471    
3472    On the x86 the typical linkage code that is generated is:
3473       ----------------------
3474            .align 2
3475       L0:
3476            addl    $L1-L0, 72(%esp)
3477            jmp     L1
3478    
3479    
3480            .align  2
3481       L1:
3482       ----------------------
3483    
3484    72(%esp) is the memory location for the stdlink register. This
3485    must contain the address of the CPS function being called. In the
3486    above example, it contains the address of  L0; before
3487    calling L1 (the real entry point for the compilation unit), it
3488    must contain the address for L1, and hence
3489    
3490            addl $L1-L0, 72(%esp)
3491    
3492    I have tested this on all architectures except the hppa.The increase
3493    in code size is of course negligible
3494    
3495    ----------------------------------------------------------------------
3496    Name: Allen Leung
3497    Date: 2002/03/03 13:20:00 EST
3498    Tag: leunga-20020303-mlrisc-tools
3499    
3500      Added #[ ... ] expressions to mlrisc tools
3501    
3502    ----------------------------------------------------------------------
3503    Name: Matthias Blume
3504    Date: 2002/02/27 12:29:00 EST
3505    Tag: blume-20020227-cdebug
3506    Description:
3507    
3508    - made types in structure C and C_Debug to be equal
3509    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3510    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3511    
3512    ----------------------------------------------------------------------
3513    Name: Matthias Blume
3514    Date: 2002/02/26 12:00:00 EST
3515    Tag: blume-20020226-ffi
3516    Description:
3517    
3518    1. Fixed a minor bug in CM's "noweb" tool:
3519       If numbering is turned off, then truly don't number (i.e., do not
3520       supply the -L option to noweb).  The previous behavior was to supply
3521       -L'' -- which caused noweb to use the "default" line numbering scheme.
3522       Thanks to Chris Richards for pointing this out (and supplying the fix).
3523    
3524    2. Once again, I reworked some aspects of the FFI:
3525    
3526       A. The incomplete/complete type business:
3527    
3528       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3529         gone!
3530       - ML types representing an incomplete type are now *equal* to
3531         ML types representing their corresponding complete types (just like
3532         in C).  This is still safe because ml-nlffigen will not generate
3533         RTTI for incomplete types, nor will it generate functions that
3534         require access to such RTTI.   But when ML code generated from both
3535         incomplete and complete versions of the C type meet, the ML types
3536         are trivially interoperable.
3537    
3538         NOTE:  These changes restore the full generality of the translation
3539         (which was previously lost when I eliminated functorization)!
3540    
3541       B. Enum types:
3542    
3543       - Structure C now has a type constructor "enum" that is similar to
3544         how the "su" constructor works.  However, "enum" is not a phantom
3545         type because each "T enum" has values (and is isomorphic to
3546         MLRep.Signed.int).
3547       - There are generic access operations for enum objects (using
3548         MLRep.Signed.int).
3549       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3550         * The structure contains the definition of type "mlrep" (the ML-side
3551         representation type of the enum).  Normally, mlrep is the same
3552         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3553         then mlrep will be defined as a datatype -- thus facilitating
3554         pattern matching on mlrep values.
3555         ("-ec" will be suppressed if there are duplicate values in an
3556          enumeration.)
3557         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3558         will be generated for each C enum constant xxx.
3559         * Conversion functions m2i and i2m convert between mlrep and
3560         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3561         * Coversion functions c and ml convert between mlrep and "tag enum".
3562         * Access functions (get/set) fetch and store mlrep values.
3563       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3564         enumerations are merged into one single enumeration represented by
3565         structure E_'.
3566    
3567    ----------------------------------------------------------------------
3568    Name: Allen Leung
3569    Date: 2002/02/25 04:45:00 EST
3570    Tag: leunga-20020225-cps-spill
3571    
3572    This is a new implementation of the CPS spill phase.
3573    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3574    In case of problems, replace it with the old file spill.sml
3575    
3576    The current compiler runs into some serious performance problems when
3577    constructing a large record.  This can happen when we try to compile a
3578    structure with many items.  Even a very simple structure like the following
3579    makes the compiler slow down.
3580    
3581        structure Foo = struct
3582           val x_1 = 0w1 : Word32.int
3583           val x_2 = 0w2 : Word32.int
3584           val x_3 = 0w3 : Word32.int
3585           ...
3586           val x_N = 0wN : Word32.int
3587        end
3588    
3589    The following table shows the compile time, from N=1000 to N=4000,
3590    with the old compiler:
3591    
3592    N
3593    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3594           MLRISC ra                               0.06u  0.00s  0.05g
3595              (spills = 0 reloads = 0)
3596           TOTAL                                   0.63u  0.07s  0.21g
3597    
3598    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3599           MLRISC ra                               5.68u  0.59s  3.93g
3600              (spills = 0 reloads = 0)
3601           TOTAL                                   14.71u  0.99s  4.81g
3602    
3603    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3604           MLRISC ra                               5.54u  0.65s  3.91g
3605              (spills = 543 reloads = 1082)
3606           TOTAL                                   65.40u  3.13s  6.00g
3607    
3608    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3609           MLRISC ra                               0.80u  0.10s  0.55g
3610              (spills = 42 reloads = 84)
3611           TOTAL                                   129.42u  5.10s  4.13g
3612    
3613    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3614           MLRISC ra                               2.69u  0.27s  1.38g
3615              (spills = 62 reloads = 124)
3616           TOTAL                                   682.48u  19.61s  13.99g
3617    
3618    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3619           MLRISC ra                               4.96u  0.27s  2.72g
3620              (spills = 85 reloads = 170)
3621           TOTAL                                   2375.26u  57.21s  48.00g
3622    
3623    As you can see the old cps spill module suffers from some serious
3624    performance problem.  But since I cannot decipher the old code fully,
3625    instead of patching the problems up, I'm reimplementing it
3626    with a different algorithm.  The new code is more modular,
3627    smaller when compiled, and substantially faster
3628    (O(n log n) time and O(n) space).  Timing of the new spill module:
3629    
3630    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3631          MLRISC ra                               0.25u  0.02s  0.15g