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/branches/primop-branch-3/NOTES/HISTORY
ViewVC logotype

Diff of /sml/branches/primop-branch-3/NOTES/HISTORY

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

revision 643, Fri May 12 08:29:15 2000 UTC revision 1879, Fri Dec 16 06:13:13 2005 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: Matthias Blume (http://www.tti-c.org/blume/)
17    Date: 2005/12/16 00:15:00 CST
18    Tag: blume-20051216-cm
19    Description:
20    
21    Improved error reporting and handling in CM.
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume (http://www.tti-c.org/blume/)
25    Date: 2005/11/21 14:00:00 CST
26    Tag: blume-20051121-html-lib
27    Description:
28    
29    Enabling $/html-lib.cm in config/preloads did not work.  This is fixed
30    now.  (Since the anchor mapping for html-lib.cm is not yet in effect
31    at the time when config/preloads is processed, the library has to
32    be referred to by another name.  In 110.57 this would be
33    $SMLNJ-LIB/HTML/html-lib.cm.  I arranged for $smlnj/smlnj-lib/html-lib.cm
34    to be valid as well.)
35    
36    Thanks to Todd Wilson (Fresno) for alerting me to this issue.
37    
38    ----------------------------------------------------------------------
39    Name: Matthias Blume (http://www.tti-c.org/blume/)
40    Date: 2005/11/19 13:45:00 CST
41    Tag: Release_110_57
42    Description:
43    
44    Moved release tag for 110.57.
45    
46    ----------------------------------------------------------------------
47    Name: Matthias Blume (http://www.tti-c.org/blume/)
48    Date: 2005/11/19 23:35:00 CST
49    Tag: blume-20051119-fish
50    Description:
51    
52    Fixed a problem in config/install.sh where it tries to "fish" the
53    name of the CM metadata directory from the wrong place (because
54    the physical location of basis.cm has changed).  Also, corrected the
55    path anchor for $/html-lib.cm.  (Thanks to M. Fluet for pointing out
56    these problems.)
57    
58    ----------------------------------------------------------------------
59    Name: Matthias Blume (http://www.tti-c.org/blume/)
60    Date: 2005/11/16 14:30:00 CST
61    Tag: blume-orig-110_57
62    Description:
63    
64    New working version (110.57).  NEW BOOTFILES!
65    
66    UPDATE: The release tag has been moved to include a fix for a brown-paper-
67            bag glitch (see above).
68    
69    ----------------------------------------------------------------------
70    Name: Matthias Blume (http://www.tti-c.org/blume/)
71    Date: 2005/11/16 10:25:00 CST
72    Tag: blume-22051116-btrace
73    Description:
74    
75    Fixed problem with bogus exception message when using back-trace
76    facility.
77    
78    ----------------------------------------------------------------------
79    Name: Matthias Blume (http://www.tti-c.org/blume/)
80    Date: 2005/11/15 22:35:00 CST
81    Tag: blume-22051115-array2copy
82    Description:
83    
84    Added simple implementation of Array2.copy.  (Warning: mostly untested.)
85    
86    ----------------------------------------------------------------------
87    Name: Matthias Blume (http://www.tti-c.org/blume/)
88    Date: 2005/11/15 21:35:00 CST
89    Tag: blume-22051115-versiontool
90    Description:
91    
92    Reversed change to src/system/smlnj/internals/versiontool.cm.  This
93    file gets loaded as a tool -- by the equivalent of CM.make during the
94    run of CMB.make.  Thus, CMB's path configuration is meaningless for
95    it.  Instead, it has the status of "user code", so it should use
96    $/basis.cm to refer to the Basis library.  (At least that's true for
97    the purpose of bootstrapping the previous change.  In the future it
98    might make sense to have versiontool.cm refer to
99    $smlnj/basis/basis.cm, i.e., the version of the Basis that the
100    compiler itself uses.)
101    
102    Also patched src/system/testml to have it activate those extra anchor
103    bindings in config/extrapathconfig.
104    
105    ----------------------------------------------------------------------
106    Name: Matthias Blume (http://www.tti-c.org/blume/)
107    Date: 2005/11/15 16:35:00 CST
108    Tag: blume-22051115-libraries
109    Description:
110    
111    This change affects the way the following libraries are tied into
112    the system:
113    
114       $/basis.cm
115       $/smlnj-lib.cm
116       $/pp-lib.cm
117       $/controls-lib.cm
118       $/html-lib.cm
119       $/ml-yacc-lib.cm
120    
121    These libraries are now internally (as seen from the source code
122    of the implementation itself) known by the following names:
123    
124       $smlnj/basis/basis.cm
125       $smlnj/smlnj-lib/smlnj-lib.cm
126       $smlnj/smlnj-lib/pp-lib.cm
127       $smlnj/smlnj-lib/controls-lib.cm
128       $smlnj/smlnj-lib/html-lib.cm
129       $smlnj/ml-yacc/ml-yacc-lib.cm
130    
131    This makes it possible to work with code that requires different
132    versions of these libraries, and which refers to these libraries using
133    their "default" names (i.e., the first set of names above).  In other
134    words, one can un-define or re-define those default names without
135    compromising the proper functioning of the compiler itself.
136    
137    A similar procedure had already been performed for several of the
138    MLRISC libraries that are linked into the compiler.  I did some
139    cleanup on this code.
140    
141    A new file in the config directory (named extrapathconfig) is
142    responsible for setting up path anchors that the compiler itself does
143    not need, but that are typically required by user code.
144    
145    ----------------------------------------------------------------------
146    Name: Matthias Blume
147    Date: 2005/11/07 17:05:00 CST
148    Tag: blume-20051107-slice
149    Description:
150    
151    Fixed erroneous out-of-bounds test in the "update" function of
152    various *ArraySlice modules.  (Thanks to Vesa A. Norrman for pointing
153    out the problem.)
154    
155    Pushed some Basis changes through ML-Lex, CML, and eXene.
156    
157    ----------------------------------------------------------------------
158    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
159    Date: 2005/11/07
160    Description:
161    Fixed a Basis incompatibility: The depreciated function Substring.all
162    was removed (use Substring.full instead).
163    
164    ----------------------------------------------------------------------
165    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
166    Date: 2005/11/05
167    Description:
168    Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
169    
170    ----------------------------------------------------------------------
171    Name: Matthias Blume (http://www.tti-c.org/blume/)
172    Date: 2005/11/02 16:25:00 CST
173    Tag: blume-20051102-hex
174    Description:
175    
176    * Runtime system bootstrap code now accepts hex digits in BOOTLIST
177      in either upper case or lower case format.
178    * Pushed changes to names of Pack<N>{Big,Little} structures through
179      CML and eXene.
180    
181    ----------------------------------------------------------------------
182    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
183    Date: 2005/11/02
184    Description:
185    Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
186    be called PackWord<N>{Big,Little}.
187    
188    ----------------------------------------------------------------------
189    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
190    Date: 2005/10/28
191    Description:
192    Fixed a minor Basis incompatibility: hex digits should be upper case.
193    
194    ----------------------------------------------------------------------
195    Name: Matthias Blume (blume (at) tti - c (dot) org)
196    Date: 2005/10/25 16:00:00 CDT
197    Tag: Release_110_56
198    Description:
199    
200    New working version (110.56).  NEW BOOTFILES!
201    
202    ----------------------------------------------------------------------
203    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
204    Date: 2005/10/25
205    Description:
206    Added interval sets to utility library (signatures INTERVAL_DOMAIN
207    and INTERVAL_SET, and functor IntervalSetFn).
208    
209    ----------------------------------------------------------------------
210    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
211    Date: 2005/10/14
212    Description:
213    Add Zhong Shao's fix for datatype equality functions.
214    
215    ----------------------------------------------------------------------
216    Name: Allen Leung (leunga (at) reservoir (dot) com)
217    Date: 2005/10/14 12:30:00 EST
218    Tag: leunga-20051014-x86-spill
219    Description:
220    
221    Bug fix a bug found by Carl Hauser.
222    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
223    
224    ----------------------------------------------------------------------
225    Name: Allen Leung (leunga (at) reservoir (dot) com)
226    Date: 2005/10/14 11:15:00 EST
227    Tag: leunga-20051014-x86-asm
228    Description:
229    
230    Removed some debugging code in file x86Asm.sml.
231    The function emit_operand was printing out debugging output.
232    
233    ----------------------------------------------------------------------
234    Name: John Reppy
235    Date: 2005/07/27
236    Tag:
237    Description:
238    
239    Fixed ml-lex to recognize "\r" as representing carriage return.
240    
241    ----------------------------------------------------------------------
242    Name: John Reppy
243    Date: 2005/07/27
244    Tag:
245    Description:
246    
247    Fixed ml-yacc to work on files with non-native end-of-line
248    encodings (e.g., Windows text file on a Unix system).
249    
250    ----------------------------------------------------------------------
251    Name: John Reppy
252    Date: 2005/07/20
253    Tag:
254    Description:
255    
256    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
257    to support HPUX 11.
258    
259    ----------------------------------------------------------------------
260    Name: John Reppy
261    Date: 2005/07/06
262    Tag:
263    Description:
264    
265    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
266    
267    ----------------------------------------------------------------------
268    Name: Matthias Blume (blume (at) tti - c (dot) org)
269    Date: 2005/07/06 00:45:00 CDT
270    Tag: blume-20050706-slice-copy
271    Description:
272    
273    Fixed reversed logic for deciding whether to "copy up" or "copy down"
274    in *-array-slice.sml.
275    
276    ----------------------------------------------------------------------
277    Name: Allen Leung (leunga (at) reservoir (dot) com)
278    Date: 2005/05/31 17:00:00 EST
279    Tag: leunga-20050531-cygwin-fault-2
280    Description:
281    
282    A typo in the cygwin code fixed.
283    
284    ----------------------------------------------------------------------
285    Name: Allen Leung (leunga (at) reservoir (dot) com)
286    Date: 2005/05/31 16:47:00 EST
287    Tag: leunga-20050531-cygwin-fault
288    Description:
289    
290    Updated Cygwin's fault/signal handling to match the Windows version.
291    Updated the export list.
292    
293    ----------------------------------------------------------------------
294    Name: Matthias Blume (blume (at) tti - c (dot) org)
295    Date: 2005/05/18 13:35:00 CDT
296    Tag: Release_110_54
297    Description:
298    
299    New working version (110.54).  NEW BOOTFILES!
300    
301    ----------------------------------------------------------------------
302    Name: Matthias Blume (blume (at) tti - c (dot) org)
303    Date: 2005/05/18 11:58:00 CDT
304    Tag: blume-20050518-installer
305    Description:
306    
307    Added support scripts for Mac OS X PackageMaker and modified
308    config/install.sh so that it supports re-dumping a heap image
309    after customization.
310    
311    ----------------------------------------------------------------------
312    Name: Matthias Blume (blume (at) tti - c (dot) org)
313    Date: 2005/05/18 10:55:00 CDT
314    Tag: blume-20050518-realdiv-noovld
315    Description:
316    
317    Un-overloaded / to work around bug in overloading resolution code.
318    
319    ----------------------------------------------------------------------
320    Name: Matthias Blume (blume (at) tti - c (dot) org)
321    Date: 2005/05/16 23:50:00 CDT
322    Tag: blume-20050516-redump-heap
323    Description:
324    
325    Added mechanism for re-creating a heap file for the interactive system
326    after configuration variables have been changed.
327    
328       CM.redump_heap : string -> unit
329    
330    This is much like SMLofNJ.exportML, but starting from the resulting
331    heap does not return to the caller of CM.redump_heap but
332    restarts the interactive system from scratch.  The original call of
333    CM.redump_heap does not return but ends the interactive session.
334    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
335    
336    Internally, redump_heap winds the dynamic execution context back to
337    the point where the original heap image was created and re-executes
338    the heap image generation code in the boot code.
339    
340    ----------------------------------------------------------------------
341    Name: Matthias Blume (blume (at) tti - c (dot) org)
342    Date: 2005/05/09 21:55:00 CDT
343    Tag: blume-20050509-word64
344    Description:
345    
346    Added a hack to the existing hack known as Word64 to make fromString
347    behave correctly.  I am still not sure whether Word64.scan will work
348    as specified with respect to the interaction of radix and prefix.
349    
350    ----------------------------------------------------------------------
351    Name: Allen Leung (leunga (at) reservoir (dot) com)
352    Date: 2005/05/04 11:50:00 EST
353    Tag: leunga-20050504-checkgc
354    Description:
355    
356    Added a gc protocol checking phase.  This phase is enabled with
357    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
358    
359    ----------------------------------------------------------------------
360    Name: Matthias Blume (blume (at) tti - c (dot) org)
361    Date: 2005/05/04 10:45:00 CDT
362    Tag: blume-20050504-intinf
363    Description:
364    
365    Fixed a bug in the implementation of div and mod for IntInf.
366    Thanks to Neophytos Michael for reporting the problem.
367    
368    ----------------------------------------------------------------------
369    Name: Matthias Blume (blume (at) tti - c (dot) org)
370    Date: 2005/05/04 10:35:00 CDT
371    Tag: blume-20050504-join
372    Description:
373    
374    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
375    
376    ----------------------------------------------------------------------
377    Name: Matthias Blume (blume (at) tti - c (dot) org)
378    Date: 2005/02/28 23:40:00 CST
379    Tag: blume-20050228-mVar
380    Description:
381    
382    Fixed serious bug (brown paper bag variety) in new implementation of
383    structure Atom in CML.  (I had accidentally used a mailbox instead of
384    an mvar, leaving the door open for races.)
385    
386    ----------------------------------------------------------------------
387    Name: Matthias Blume (blume (at) tti - c (dot) org)
388    Date: 2005/02/25 15:00:00 CST
389    Tag: Release_110_53
390    Description:
391    
392    New working version (110.53).  NEW BOOTFILES!
393    
394    ----------------------------------------------------------------------
395    Name: Matthias Blume (blume (at) tti - c (dot) org)
396    Date: 2005/02/25 14:50:00 CST
397    Tag: blume-20050225-susp
398    Description:
399    
400    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
401    implemented in Core -- which means that it is the same as the one used
402    by the lazy extension.
403    
404    ----------------------------------------------------------------------
405    Name: Matthias Blume (blume (at) tti - c (dot) org)
406    Date: 2005/02/24 16:50:00 CST
407    Tag: blume-20050224-cml-atom
408    Description:
409    
410    Simpler and at the same time more general implementation of structure
411    Atom in CML.
412    
413    ----------------------------------------------------------------------
414    Name: Matthias Blume (blume (at) tti - c (dot) org)
415    Date: 2005/02/15 17:35:00 CST
416    Tag: blume-20050215-tools
417    Description:
418    
419    Created new "tools" directory under "src" and moved "TraceDebugProf"
420    there.
421    
422    ----------------------------------------------------------------------
423    Name: Matthias Blume (blume (at) tti - c (dot) org)
424    Date: 2005/02/10 17:55:00 CST
425    Tag: blume-20050210-longlong
426    Description:
427    
428    Implemented "long long" arguments and results for NLFFI.  (Only the
429    PPC/MacOS implementation is complete, the other backends still need to
430    be updated.)
431    
432    ----------------------------------------------------------------------
433    Name: Matthias Blume (blume (at) tti - c (dot) org)
434    Date: 2005/01/24 17:40:00 CST
435    Tag: blume-20050124-mlyacc
436    Description:
437    
438    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
439    problem with certain "as" patterns which previously got rendered
440    using incorrect syntax.
441    
442    ----------------------------------------------------------------------
443    Name: Matthias Blume (blume (at) tti - c (dot) org)
444    Date: 2005/01/18 12:00:00 CST
445    Tag: blume-20050118-profile
446    Description:
447    
448    Made time profiling code (interrupt handler) in runtime system aware
449    of new array representation.
450    
451    ----------------------------------------------------------------------
452    Name: Matthias Blume (blume (at) tti - c (dot) org)
453    Date: 2005/01/14 18:00:00 CST
454    Tag: blume-20050114-heap2exec
455    Description:
456    
457    Implemented new (but still experimental) heap2exec facility.  This is
458    tested under Mac OS X and should work under Linux (will test shortly).
459    It will probably also work on the Sparc (will test some time later).
460    
461      - removed old "HACKED_STANDALONE" hack from runtime
462    
463    To be able to test this, uncomment the request for "heap2asm" in
464    config/targets prior to installation.  (Notice that this is different
465    from "heap2exec" mentioned below.  Not a typo.)
466    
467    To perform an actual test, run the command
468    
469       $ bin/heap2exec heapfile execfile
470    
471    (You can put heap2exec on your shell's path.)
472    
473    For example, run
474    
475      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
476    
477    This will create a standalone executable called "mly" which you can
478    then invoke directly as a command.
479    
480    ----------------------------------------------------------------------
481    Name: Matthias Blume (blume (at) tti - c (dot) org)
482    Date: 2005/01/07 11:44:00 CST
483    Tag: blume-20050107-mlstring
484    Description:
485    
486    fixed off-by-one error in ML_STRING macro (globals.c)
487    
488    ----------------------------------------------------------------------
489    Name: Matthias Blume (blume (at) tti - c (dot) org)
490    Date: 2004/12/23 18:00:00 CST
491    Tag: blume-20041223-santa
492    Description:
493    
494    Made ml-build script "smarter" (but only very little).
495    
496    ----------------------------------------------------------------------
497    Name: Matthias Blume (blume (at) tti - c (dot) org)
498    Date: 2004/12/21 15:05:00 CST
499    Tag: blume-20041221-longlong
500    Description:
501    
502    * Implemented access to signed and unsigned long long data in NLFFI.
503    (The parameter-passing part of the picture has not complete. But
504    data structure access seems to work.)
505    
506    * Fixed CM's incorrect assumption that the PPC is little-endian.
507      (On the Mac, it is big-endian.  And that's currently our only
508      PPC platform.)
509    
510    ----------------------------------------------------------------------
511    Name: Matthias Blume (blume (at) tti - c (dot) org)
512    Date: 2004/12/21 12:50:00 CST
513    Tag: blume-20041221-memory
514    Description:
515    
516    Some cleanup in the $c/memory.cm library: separated some concerns by
517    moving allocation code and memory access code each into their own
518    files.
519    
520    ----------------------------------------------------------------------
521    Name: Allen Leung (leunga (at) reservoir (dot) com)
522    Date: 2004/12/17 16:12:00 EST
523    Tag: leunga-20041217-cygwin-smlnj-home
524    Description:
525    
526        The Unix I/O library of SML/NJ on cygwin does not understand
527    Windows style pathname, so problems arise when SMLNJ_HOME is set
528    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
529    to a POSIX pathname on cygwin.
530    
531    ----------------------------------------------------------------------
532    Name: Matthias Blume (blume (at) tti - c (dot) org)
533    Date: 2004/12/16 13:35:00 CST
534    Tag: Release_110_52
535    Description:
536    
537    Last-minute changes incorporated into 110.52.  Release tag moved.
538    
539    The changes:
540       - HashString.hashString' -> HashString.hashSubstring
541       - bug fix in UnivariateStats
542    
543    ----------------------------------------------------------------------
544    Name: Matthias Blume (blume (at) tti - c (dot) org)
545    Date: 2004/12/15 23:40:00 CST
546    Tag: blume-20041215-hashSubstring
547    Description:
548    
549    - HashString.hashString' -> HashString.hashSubstring
550    - corresponding changes in atom.sml
551    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
552      added some comments
553    
554    ----------------------------------------------------------------------
555    Name: Matthias Blume (blume (at) tti - c (dot) org)
556    Date: 2004/12/15 15:30:00 CST
557    Tag: (Release_110_52)
558    Description:
559    
560    New working version (110.52).  NEW BOOTFILES!
561    
562    ----------------------------------------------------------------------
563    Name: Matthias Blume (blume (at) tti - c (dot) org)
564    Date: 2004/12/15 12:45:00 CST
565    Tag: blume-20041215-spaces
566    Description:
567    
568    More on the space problem (this time for Win32).
569    
570    ----------------------------------------------------------------------
571    Name: Matthias Blume (blume (at) tti - c (dot) org)
572    Date: 2004/12/14 17:30:00 CST
573    Tag: blume-20041214-spaces
574    Description:
575    
576    Hacked some of the scripts (in particular: the installer) to cope with
577    spaces in filenames a bit better.  But beware: the current "solution"
578    is likely still full of bugs and inherently incomplete.  (We need to
579    do away with those shell scripts for a comprehensive solution.)
580    
581    ----------------------------------------------------------------------
582    Name: Matthias Blume (blume (at) tti - c (dot) org)
583    Date: 2004/12/13 14:45:00 CST
584    Tag: blume-20041213-ml-makedepend
585    Description:
586    
587    Fixed bug in code for ml-makedepend.
588    
589    ----------------------------------------------------------------------
590    Name: Matthias Blume (blume (at) tti - c (dot) org)
591    Date: 2004/12/09 16:30:00 CST
592    Tag: blume-20041209-statistics
593    Description:
594    
595    Added two simple but potentially useful statistics modules to SML/NJ Library.
596    (See CHANGES file there.)
597    
598    ----------------------------------------------------------------------
599    Name: Matthias Blume (blume (at) tti - c (dot) org)
600    Date: 2004/12/01 16:50:00 CST
601    Tag: blume-20041201-atom
602    Description:
603    
604    smlnj-lib:
605    
606      Added function HashString.hashString' for substrings.
607      Hand-inlined CharVector.foldl into HashString (for speed).
608      Modified implementation of structure Atom to avoid extracting
609      strings from substrings unless necessary.
610      (Also see CHANGES file for smlnj-lib.)
611    
612    ----------------------------------------------------------------------
613    Name: Matthias Blume (blume (at) tti - c (dot) org)
614    Date: 2004/11/24 22:15:00 CST
615    Tag: blume-20041124-cml
616    Description:
617    
618    Made sure CML compiles when Position = Int64.
619    
620    ----------------------------------------------------------------------
621    Name: Matthias Blume (blume (at) tti - c (dot) org)
622    Date: 2004/11/24 14:30:00 CST
623    Tag: blume-20041124-position
624    Description:
625    
626    The compiler can now be compiled in a mode that makes structure Position
627    equal to Int64.  The default, however, is unchanged (Position = Int31)
628    for the time being.
629    
630    To enable 64-bit positions, use the following procedure:
631    
632    1. Start sml
633    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
634    3. Type
635          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
636    4. Run CMB.make() as usual.
637    
638    This is barely tested.  The only test so far was a little SML program
639    counting the number of characters in an 8-gigabyte file by
640    reading it character-by-character.  That test was successful.
641    
642    In support of 64-bit positions, a number of new functions have been
643    added to the runtime system.
644    
645    ----------------------------------------------------------------------
646    Name: Matthias Blume (blume (at) tti - c (dot) org)
647    Date: 2004/11/23 14:45:00 CST
648    Tag: blume-20041123-useFile
649    Description:
650    
651    Fixed a problem with unhelpful error messages related to problems with
652    .cm- or .sml files that appear as part of the sml command line.
653    
654    ----------------------------------------------------------------------
655    Name: Matthias Blume (blume (at) tti - c (dot) org)
656    Date: 2004/11/18 15:40:00 CST
657    Tag: Release_110_51
658    Description:
659    
660    New working version (110.51).  NEW BOOTFILES!
661    
662    ----------------------------------------------------------------------
663    Name: Matthias Blume (blume (at) tti - c (dot) org)
664    Date: 2004/11/18 15:35:00 CST
665    Tag: <none>
666    Description:
667    
668    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
669    Joemann).
670    
671    ----------------------------------------------------------------------
672    Name: Allen Leung (leunga (at) reservoir (dot) com)
673    Date: 2004/11/17 16:05:21 EST 2004
674    Tag: leunga-20041117-mlrisc-live-kill
675    Description:
676    
677       Added support for MLTree constructs LIVE and KILL
678    to all the architectures.
679    
680    ----------------------------------------------------------------------
681    Name: Matthias Blume (blume (at) tti - c (dot) org)
682    Date: 2004/11/13 00:20:00 CST
683    Tag: blume-20041113-versiontool
684    Description:
685    
686    - Stripped down the versiontool: It now only handles the version number.
687      The date string is generated at bootstrap time (during makeml).
688    
689    - In a previous commit, fixed a minor issue with how polyequal is being
690      translated.  In particular, the code now "looks through" abstractions.
691      This results in slightly fewer polyEqual warnings and hopefully slightly
692      more efficient code.  Important examples for where this matters are
693      the new int64 and word64 types.
694    
695    ----------------------------------------------------------------------
696    Name: Matthias Blume (blume (at) tti - c (dot) org)
697    Date: 2004/11/12 00:30:00 CST
698    Tag: blume-20041112-int64
699    Description:
700    
701    Structure Int64 fully hooked in.  (The implementation is not very
702    efficient, though.)
703    
704    ----------------------------------------------------------------------
705    Name: Matthias Blume (blume (at) tti - c (dot) org)
706    Date: 2004/11/11 17:30:00 CST
707    Tag: blume-20041111-more64
708    Description:
709    
710    All the pieces of Word64 are now there, with the exception of the
711    conversions from and to LargeWord. (Eventually these need to be identities,
712    but for the time being they don't even make sense because LargeWord is
713    32-bit wide.)
714    
715    Also started to add similar support for Int64, but major pieces of that
716    are still missing.
717    
718    ----------------------------------------------------------------------
719    Name: Matthias Blume (blume (at) tti - c (dot) org)
720    Date: 2004/11/11 00:15:00 CST
721    Tag: blume-20041111-word64
722    Description:
723    
724    Structure Word64 is now (almost) complete, word literals and patterns
725    seem to work.  There are a few odd pieces missing.  In particular,
726    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
727    at the moment.
728    
729    Making Word64 official would mean that LargeWord becomes Word64.  But
730    this requires extreme care because most word-word conversions have to
731    go through LargeWord, so making a mistake means loss of efficiency or
732    worse.  Eventually there will be a solution similar to (but actually
733    simpler than) what I did with IntInf.
734    
735    ----------------------------------------------------------------------
736    Name: Matthias Blume (blume (at) tti - c (dot) org)
737    Date: 2004/11/10 18:12:00 CST
738    Tag: blume-20041110-64bit
739    Description:
740    
741    More 64-bit hacking (but still not even half-way there yet).
742    Also, some assorted improvements to the handling of 8-bit words.
743    
744    ----------------------------------------------------------------------
745    Name: Matthias Blume (blume (at) tti - c (dot) org)
746    Date: 2004/11/09 17:50:00 CST
747    Tag: <none>
748    Description:
749    
750    Started adding some infrastructure for supporting 64-bit int- and
751    word-types.  (Still in its very early stages.)
752    
753    ----------------------------------------------------------------------
754    Name: Matthias Blume (blume (at) tti - c (dot) org)
755    Date: 2004/10/28 10:45:00 CDT
756    Tag: Release_110_50
757    Description:
758    
759    New working version (110.50).  NEW BOOTFILES!
760    
761                 =====================
762    
763    Also:
764    
765      - Changend config/srcarchiveurl from a file just
766        containing the URL string into a file containing
767        shell script code.  The code has access to the $VERSION variable.
768      - Made corresponding changes to config/install.sh and config/unpack.
769      - Default contents of config/srcarchiveurl uses $VERSION and
770        normally does not have to be edited to reflect a version change.
771    
772      (As a result, a version change can be done by just editing
773       config/version, the rest is now automatic.)
774    
775    ----------------------------------------------------------------------
776    Name: Matthias Blume (blume (at) tti - c (dot) org)
777    Date: 2004/10/27 17:50:00 CDT
778    Tag: blume-20041027-btrace-msg
779    Description:
780    
781    BackTrace.monitor now also reports the source of the exception that
782    triggered the trace.
783    
784    ----------------------------------------------------------------------
785    Name: Matthias Blume (blume (at) tti - c (dot) org)
786    Date: 2004/10/27 17:20:00 CDT
787    Tag: blume-20041027-x86-c-calls
788    Description:
789    
790    This is the HISTORY entry for two earlier commits, both concerning
791    the x86 c-calls code in MLRISC:
792    
793      - added a missing LOAD in the code that deals with struct arguments
794      - made sure the caller does not add the wrong number of bytes to the
795        stack pointer after a call of a function returning a struct
796        (the callee already pops the implicit argument which points to
797         the space reserved for the result)
798    
799    ----------------------------------------------------------------------
800    Name: Allen Leung (leunga (at) reservoir (dot) com)
801    Date: 2004/10/24 14:00:00 EST
802    Tag: leunga-20041024-x86-gas-fucomip
803    Description:
804    
805       John discovered a bug in the syntax of fucomip.
806       The opcodes FU?COMIP? have been changed to
807    
808           fu?comip? %st(i), %st
809    
810    ----------------------------------------------------------------------
811    Name: Matthias Blume (blume (at) tti - c (dot) org)
812    Date: 2004/10/20 15:06:00 CDT
813    Tag: blume-20041020-standalone-backtrace
814    Description:
815    
816    Added a mechanism for getting back-trace information from standalone
817    programs.  Here is how it works:
818    
819       1. The part of the program from which you want to get backtrace
820          information (usually the whole program) should be wrapped with
821          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
822          main program could be modified from something like
823    
824             fun main (pgm, args) = ...
825    
826         to
827    
828            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
829    
830       2. To be able to access BackTrace.monitor, you have to add
831          library $smlnj-tdp/plugins.cm to the .cm file that contains your
832          main function.
833    
834       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
835          CM might have created in the past for your project).
836    
837       4. Build the system using this command line:
838    
839             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
840                      myprog.cm MyProg.main myprog
841    
842          instead of the usual
843    
844             ml-build myprog.cm MyProg.main myprog
845    
846    I changed a library name:
847    
848       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
849    
850    New libraries:
851    
852       $smlnj-tdp/back-trace.cm
853          -- when loaded causes the back-trace plugin to be installed
854       $smlnj-tdp/coverage.cm
855          -- when loaded causes the coverage plugin to be installed
856    
857    ----------------------------------------------------------------------
858    Name: Matthias Blume (blume (at) tti - c (dot) org)
859    Date: 2004/10/18 16:45:00 CDT
860    Tag: blume-20041018-groupowner
861    Description:
862    
863    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
864    
865    Eliminated group owner specs from .cm files throughout the source tree.
866    
867    ----------------------------------------------------------------------
868    Name: Matthias Blume (blume (at) tti - c (dot) org)
869    Date: 2004/10/15 15:45:00 CDT
870    Tag: blume-20041015-coverage
871    Description:
872    
873    * Test coverage tool added!
874    * Further reorganization of tracing-, debugging-, and profiling support:
875    
876        - moved original BTImp -- now called BackTrace -- into a separate
877          library called $/trace-debug-profile.cm
878        - eliminated all mentions of BTrace from SMLofNJ.Internals
879        - only the instrumentation mechanism is now left in the compiler proper
880        - BackTrace module is a plugin which is NOT plugged in by default
881        - Coverage module is another such plugin
882    
883    To get the benefits of any of these plugin modules, the code in
884    question must be compiled with tdp instrumentation turned on.  This
885    can be done by setting SMLofNJ.Internals.TDP.mode to true.
886    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
887    
888    Plugins are selected at link time.  (Pre-compiled instrumented code
889    can be re-loaded with different plugins in effect.)  When an
890    instrumented module is linked, whatever plugins are at that time
891    enabled will come into effect for that module.
892    
893    To enable the back-trace plugin, load library $/trace-debug-profile.cm
894    and invoke BackTrace.install() (e.g., from the interactive prompt).
895    To enable the coverage plugin, load the same library and invoke
896    Coverage.install().
897    
898    Back-traces are generated automatically on uncaught exceptions and
899    when the code in question explicitly invokes BackTrace.trigger().
900    
901    Coverage (and execution frequency-) information must be queried
902    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
903    
904    ----------------------------------------------------------------------
905    Name: Matthias Blume (blume (at) tti - c (dot) org)
906    Date: 2004/10/14 17:40:00 CDT
907    Tag: blume-20041014-tdp-core
908    Description:
909    
910    Snapshot of a significant overhaul of how the trace/debug/profile support
911    is hooked into the system (specifically: Core and SMLofNJ.Internals).
912    
913    ----------------------------------------------------------------------
914    Name: Matthias Blume (blume (at) tti - c (dot) org)
915    Date: 2004/10/13 16:34:00 CDT
916    Tag: blume-20041013-tdp
917    Description:
918    
919    Some rationalization of names:
920    
921       structure BTrace -> structure TDPInstrument
922       etc.
923    
924    This is is preparation of using the original back-trace
925    instrumentation for other purposes.  "TDP" stands for
926    Trace/Debug/Profile.
927    
928    The control flag controlling whether instrumentation is on or off is now
929    registered under a different name, so instead of running sml as
930    
931     sml -Cinstrument.btrace-mode=true
932    
933    one has to say
934    
935     sml -Ctdp.instrument=true
936    
937    ----------------------------------------------------------------------
938    Name: Matthias Blume (blume (at) tti - c (dot) org)
939    Date: 2004/10/11 16:37:00 CDT
940    Tag: blume-20041011-regions
941    Description:
942    
943    Made some minor modifications to elabcore.sml to have source regions
944    be propagated more tightly -- resulting in better (i.e., smaller)
945    regions being reported in error- and debug messages.
946    
947    ----------------------------------------------------------------------
948    Name: Matthias Blume (blume (at) tti - c (dot) org)
949    Date: 2004/10/08 22:50:00 CDT
950    Tag: blume-20041008-cmkw
951    Description:
952    
953    Fixed handling of keywords in .cm files: After seeing "is" the lexer
954    treats subsequent occurrences of "group", "library", "source", "is",
955    "*", and "-" as ordinary identifiers rather than keywords.
956    
957    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
958    is supposed to accept a tool argument called "source", but this did
959    not work because of the clash with the keyword.
960    
961    ----------------------------------------------------------------------
962    Name: Matthias Blume (blume (at) tti - c (dot) org)
963    Date: 2004/10/07 16:00:00 CDT
964    Tag: blume-20041007-cleanup
965    Description:
966    
967    Assorted cleanup work:
968    
969      - got rid of intstrmap in favor of using the library's
970        hash table implementation
971      - threw out most of the pathnames stuff, as it was not used anyway
972      - simplified tokentable implementation
973      - fixed some minor spelling errors
974    
975    ----------------------------------------------------------------------
976    Name: Matthias Blume (blume (at) tti - c (dot) org)
977    Date: 2004/10/06 15:15:15 CDT
978    Tag: blume-20041006-handler
979    Description:
980    
981    Cleaned up the absyn to reflect the invariant that HANDLE always
982    carries a FNexp as part of the type definition.  This eliminates some
983    superfluous sanity checks at runtime down the road.
984    
985    Some minor cleanup of the btrace code.
986    
987    ----------------------------------------------------------------------
988    Name: Matthias Blume (blume (at) tti - c (dot) org)
989    Date: 2004/10/01 10:20:30 CDT
990    Tag: blume-20041001-slave
991    Description:
992    
993    Added hack to make slave mode work in the presence of the version
994    tool.  (Still, since the master does two passes over the code for
995    CMB.make, the release number gets bumped twice when slaves are
996    attached.  I don't know if this is worth fixing...)
997    
998    ----------------------------------------------------------------------
999    Name: Matthias Blume (blume (at) tti - c (dot) org)
1000    Date: 2004/09/30 10:55:00 CDT
1001    Tag: blume-20040930-version
1002    Description:
1003    
1004    * Moved the "version" magic into its own little library under
1005      src/system/smlnj/internal.  This avoids expensive reconstruction of
1006      a stable src/compiler/core.cm.
1007    
1008    * At the same time, structure CompilerVersion is now known as
1009      structure SMLNJVersion.
1010    
1011    * Arranged for the version tool to NOT kick in when rebuilding the system
1012      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
1013      Also, loading the versiontool does not work when rebuilding the system
1014      because CM is not properly initialized at that time.
1015    
1016    ----------------------------------------------------------------------
1017    Name: Matthias Blume (blume (at) tti - c (dot) org)
1018    Date: 2004/09/29 14:00:00 CDT
1019    Tag: blume-20040929-autoversion
1020    Description:
1021    
1022    Implemented some CM magic to have
1023        file src/compiler/TopLevel/main/version.sml
1024    generated automagically.
1025    The version is taken from two files: config/version and config/release.
1026    The first is expected to contain a two-part version number such as 110.49.
1027    The second should contain a single number, but it may be missing.
1028    
1029    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
1030    time the version tool is loaded (which is the first time you say CMB.make),
1031    then the tool will increment the value stored in config/release every
1032    time CMB.make is invoked.
1033    
1034    The binfile format is now insensitive to anything beyond the first
1035    two components of a version number, so bumping the release does not render
1036    binfiles incompatible.  Auto-bumping can be used to keep track of versions
1037    during development without invalidating existing binfiles.
1038    
1039    In any case, every CMB.make updates the date information in version.sml.
1040    (This is the date that is printed in the banner.)
1041    
1042    ----------------------------------------------------------------------
1043    Name: Matthias Blume (blume (at) tti - c (dot) org)
1044    Date: 2004/09/28 10:53:00 CDT
1045    Tag: blume-20040928-controls
1046    Description:
1047    
1048    Some cleanup of the controls code.
1049    
1050    ----------------------------------------------------------------------
1051    Name: Matthias Blume (blume (at) tti - c (dot) org)
1052    Date: 2004/09/27 22:08:00 CDT
1053    Tag: blume-20040927-controls
1054    Description:
1055    
1056    Added two pieces of functionality to the Controls interface:
1057    
1058    1.   val save'restore: 'a control -> unit -> unit
1059    
1060       grabs the current value of the control in stage 1 and restores it
1061       in stage 2.
1062    
1063    2.   val set' : 'a control * 'a -> unit -> unit
1064    
1065       stores the given value into the control in stage 2 (i.e., delayed)
1066       but does all error checking in stage 1.
1067       (This is for string controls that need to do parse their argument --
1068        something that might fail.  In some cases, notably in CM, one
1069        already knows the intended argument but wants to delay the actual
1070        assignment until a time when error recovery would be more difficult.)
1071    
1072    Changed the handling of controls in tool arguments to classes "sml" and
1073    "lazysml":
1074       - use Controls.save'restore as a more robust way of restoring the
1075         old value (in particular: without having to re-parse the string)
1076       - use controls to handle the "overload" keyword in the init group
1077         (I believe this change actually fixes a long-standing obscure bug.)
1078    
1079    ----------------------------------------------------------------------
1080    Name: Matthias Blume (blume (at) tti - c (dot) org)
1081    Date: 2004/09/27 17:00:00 CDT
1082    Tag: blume-20040927-lazysml
1083    Description:
1084    
1085    Added a new tool class called "lazysml" to CM's tool chest.  The only
1086    difference to "sml" is that compilation is done with Control.lazysml
1087    set to true.  A source of class "lazysml" is automatically recognized
1088    by a file name suffix of ".lml".
1089    
1090    In addition to the above feature, the original class "sml" now also
1091    supports a tool argument "lazy" which has the same effect.  As a
1092    result, the following three lines are equivalent:
1093    
1094        foo.sml : lazysml
1095        foo.sml : sml (lazy)
1096        foo.sml (lazy)
1097    
1098    The setting goes into effect both during parsing and during
1099    compilation.  The original setting is restored right after parsing and
1100    after compilation, respectively.
1101    
1102    In addition to all the above, there is also a general mechanism to set
1103    ANY of the "controls" that are available at the command line via
1104    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
1105    apply as well. (In fact, "lazy" is implemented as a special case of
1106    the general mechanism.)
1107    
1108    The .cm file syntax uses a new keyword tool argument called "with".
1109    There are several ways of indicating the desired settings:
1110    
1111        foo.sml (with:parser.quotations=true)
1112        foo.sml (with:(name:parser.quotations value:true))
1113        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1114        foo.sml (with:(name1=value1 name2=value2 ...))
1115        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1116    
1117        etc.
1118    
1119    Another possible abbreviation is to leave out the =v or value:v part
1120    if the name refers to a boolean control (in which case the value is
1121    taken to be true).  Thus, one could get lazy sml also by saying:
1122    
1123        foo.sml (with:parser.lazy-keyword=true)
1124        foo.sml (with:parser.lazy-keyword)
1125        foo.sml (with:(name:parser.lazy-keyword value:true))
1126        foo.sml (with:(name:parser.lazy-keyword))
1127    
1128    ----------------------------------------------------------------------
1129    Name: Matthias Blume (blume (at) tti - c (dot) org)
1130    Date: 2004/09/24 16:22:00 CDT
1131    Tag: blume-20040924-ppc-long-branch
1132    Description:
1133    
1134    Turned message about "emiting long form of branch" off by default.
1135    Added a control flag to turn it back on when desired.
1136    
1137    ----------------------------------------------------------------------
1138    Name: Matthias Blume (blume (at) tti - c (dot) org)
1139    Date: 2004/09/24 16:05:00 CDT
1140    Tag: blume-20040924-rounding
1141    Description:
1142    
1143    Applied patch for setting rounding modes under Mac OS X.  Thanks to
1144    Melissa O'Neill for providing the code!
1145    
1146    ----------------------------------------------------------------------
1147    Name: Matthias Blume (blume (at) tti - c (dot) org)
1148    Date: 2004/09/23 17:30:00 CDT
1149    Tag: blume-20040923-envvars
1150    Description:
1151    
1152    1. Changed definition of type ControlRegistry.registry_tree to
1153       include control_info (i.e., the name of the controlling environment
1154       variable).
1155    
1156    2. Added command-line flags -e and -E to print the names of environment
1157       variables that can be used to control internal settings.
1158       (This uses the new API mentioned in 1.)
1159    
1160    ----------------------------------------------------------------------
1161    Name: Matthias Blume (blume (at) tti - c (dot) org)
1162    Date: 2004/09/13 16:50:00 CDT
1163    Tag: Release_110_49
1164    Description:
1165    
1166    New working version (110.49).  NEW BOOTFILES!
1167    
1168    ----------------------------------------------------------------------
1169    Name: Matthias Blume
1170    Date: 2004/09/13 16:20:00 CDT
1171    Tag: blume-20040913-config-mlrisc
1172    Description:
1173    
1174    Put target "mlrisc" back into the default list.
1175    (There is no harm in having it, and some users have expressed their
1176    wish to have "mlrisc" included by default.)
1177    
1178    ----------------------------------------------------------------------
1179    Name: John Reppy
1180    Date: 2004/09/13
1181    Tag: jhr-20040913-signals
1182    Description:
1183    
1184    Fixed the signal masking code to properly nest mask/unmask operations
1185    on a per-signal basis.
1186    
1187    ----------------------------------------------------------------------
1188    Name: Matthias Blume (blume (at) tti - c (dot) org)
1189    Date: 2004/09/08 13:20:00 CDT
1190    Tag: blume-20040908-heap-magic
1191    Description:
1192    
1193    Bumped the heap macig number to 0x09082004 to account for the changed
1194    layout of the ML frame under MacOS X.
1195    
1196    ----------------------------------------------------------------------
1197    Name: Allen Leung (leunga (at) reservoir (dot) com)
1198    Date: 2004/09/03 11:26:00 EST
1199    Tag: leunga-20040903-cygwin-install
1200    Description:
1201    
1202       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1203    The Cygwin runtime is turned on by setting the environment
1204    variable SMLNJ_CYGWIN_RUNTIME to 1.
1205    
1206    ----------------------------------------------------------------------
1207    Name: Matthias Blume (blume (at) tti - c (dot) org)
1208    Date: 2004/08/31 17:15:00 CDT
1209    Tag: blume-20040831-core
1210    Description:
1211    
1212    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1213    
1214    ----------------------------------------------------------------------
1215    Name: Matthias Blume (blume (at) tti - c (dot) org)
1216    Date: 2004/08/30 17:55:00 CDT
1217    Tag: blume-20040830-installer
1218    Description:
1219    
1220    Upon request by Johannes Joemann:
1221    
1222    - improved ML code of installer to fall back to coping when renaming
1223      fails (i.e., when source and target are on different file systems);
1224      the code compiles but has yet to be tested in anger
1225    - removed mlrisc from list of default targets (config/targets)
1226    
1227    ----------------------------------------------------------------------
1228    Name: Matthias Blume (blume (at) tti - c (dot) org)
1229    Date: 2004/08/27 17:20:00 CDT
1230    Tag: blume-20040827-ptreql
1231    Description:
1232    
1233    Added ptreql primop to structure InlineT (upon request from Larry
1234    Paulson).
1235    
1236    ----------------------------------------------------------------------
1237    Name: Allen Leung (leunga (at) reservoir (dot) com)
1238    Date: 2004/08/15 21:21:00 EST
1239    Tag: leunga-110_48-udgraph
1240    Description:
1241    
1242    Another bug fix from Carl Hauser:
1243    
1244    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1245    > 48c48
1246    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1247    > e::L)
1248    > ---
1249    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1250    > Without this, any deletion of an edge in an undirected graph does severe
1251    > violence to the graph.
1252    
1253    ----------------------------------------------------------------------
1254    Name: Allen Leung (leunga (at) reservoir (dot) com)
1255    Date: 2004/08/10 23:35:00 EST
1256    Tag: leunga-110_48-ppc
1257    Description:
1258    
1259       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1260    
1261    ----------------------------------------------------------------------
1262    Name: Matthias Blume (blume (at) tti - c (dot) org)
1263    Date: 2004/08/10 12:00:00 CDT
1264    Tag: Release_110_48
1265    Description:
1266    
1267    New working version (110.48).  NEW BOOTFILES!
1268    
1269    ----------------------------------------------------------------------
1270    Name: Allen Leung (leunga (at) reservoir (dot) com)
1271    Date: 2004/08/09 12:21:00 EST
1272    Tag: leunga-110_47-dijsktra
1273    Description:
1274    
1275    Bug fix from Carl Hauser:
1276    
1277    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1278    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1279    
1280    The problem is that following the expression A.update(dist,s,Num.zero)
1281    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1282    
1283    ----------------------------------------------------------------------
1284    Name: Matthias Blume (blume (at) tti - c (dot) org)
1285    Date: 2004/08/06 18:10:00 CDT
1286    Tag: blume-20040806-cmdline
1287    Description:
1288    
1289    Fiddled with handling of command-line options:
1290    
1291      * sml now quits after processing the command line
1292        if -H, -S, -h<n>, or -s<n> appears as the last
1293        command-line argument
1294      * a new option -q terminates the session when encountered on
1295        the command line; subsequent arguments will be ignored
1296      * bug fixes: short (erroneous) arguments are no longer ignored
1297        completely
1298    
1299    ----------------------------------------------------------------------
1300    Name: Allen Leung (leunga (at) reservoir (dot) com)
1301    Date: 2004/08/04 18:17:00 EST
1302    Tag: leunga-110_47-ppc-ibm-asm
1303    Description:
1304    
1305      - Added minimal IBM assembly syntax support for PowerPC.
1306    
1307      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1308        been altered in the runtime.  We need an automatic way to keep the file
1309        in sync.
1310    
1311    ----------------------------------------------------------------------
1312    Name: Matthias Blume (blume (at) tti - c (dot) org)
1313    Date: 2004/08/04 14:00:00 CDT
1314    Tag: Release_110_47
1315    Description:
1316    
1317    New working version (110.47).  NEW BOOTFILES!
1318    
1319    ----------------------------------------------------------------------
1320    Name: Matthias Blume (blume (at) tti - c (dot) org)
1321    Date: 2004/08/03 14:25:00 CDT
1322    Tag: blume-20040803-callingconv
1323    Description:
1324    
1325    Added low-level support for choosing C calling conventions by
1326    twiddling the type of rawccall.  (See
1327    src/compiler/Semant/types/cproto.sml for details.)
1328    
1329    ----------------------------------------------------------------------
1330    Name: Matthias Blume (blume (at) tti - c (dot) org)
1331    Date: 2004/08/02 15:55:00 CDT
1332    Tag: blume-20040802-backout
1333    Description:
1334    
1335    Backed out of change to win32-filesys.c.  The earlier patch to
1336    get_file_time caused CM to produce files with the wrong time stamp.
1337    
1338    ----------------------------------------------------------------------
1339    Name: Matthias Blume (blume (at) tti - c (dot) org)
1340    Date: 2004/08/02 14:45:00 CDT
1341    Tag: blume-20040802-nlffi-win32
1342    Description:
1343    
1344    Added NLFFI support for Win32, adapted from a patch provided by David
1345    Hansel.  This is currently completely untested.  Also, the issue
1346    concerning stdcall vs. ccall is still unresolved.
1347    
1348    ----------------------------------------------------------------------
1349    Name: Matthias Blume (blume (at) tti - c (dot) org)
1350    Date: 2004/07/30 17:55:00 CDT
1351    Tag: blume-20040730-various
1352    Description:
1353    
1354    Gearing up towards 110.47...
1355    
1356    - various minor bugfixes to ml-nlffigen
1357    - a beginning of a manual for nlffi
1358    
1359    - eliminated 'export name=value' in config/install.sh as this does
1360      not work with certain versions of /bin/sh
1361      (Thanks to David King at Motorola for catching this.)
1362    
1363    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1364      - added a test for tm==NULL to gmtime.c and localtime.c
1365      - applied patch for incorrect GetFileTime under win32
1366      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1367    
1368    ----------------------------------------------------------------------
1369    Name: Matthias Blume (blume (at) tti - c (dot) org)
1370    Date: 2004/07/21 18:20:00 CDT
1371    Tag: blume-20040721-nlffigen
1372    Description:
1373    
1374    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1375      for a typedef to an incomplete type, but leave out the "typ" member.
1376      (This is just for consistency.)
1377    - Started to produce what is supposed to become better (i.e., comprehensive)
1378      documentation of what ml-nlffigen does and produces.
1379    
1380    ----------------------------------------------------------------------
1381    Name: Matthias Blume (blume (at) tti - c (dot) org)
1382    Date: 2004/07/14 16:25:00 CDT
1383    Tag: blume-20040714-union
1384    Description:
1385    
1386    Added C_UNION to c-calls/c-types.sml and updated the machinery
1387    (ml-nlffigen, cproto.sml) that conveys C function interface
1388    information to the code generator.
1389    
1390    However, the actual architecture-specific implementation of function
1391    arguments and results that are C unions is still not implemented.
1392    
1393    ----------------------------------------------------------------------
1394    Name: Allen Leung (leunga (at) reservoir (dot) com)
1395    Date: 2004/07/14 14:38:00 EST
1396    Tag: leunga-110_46_1-ppc-lwzu
1397    Description:
1398    
1399        Added these instructions to the PowerPC architecture:
1400        LBZU(X), LHZU(X), LWZU(X),
1401        STWU(X), STFDU, STFSU
1402    
1403        etc...
1404    
1405        Note: I haven't added their instruction encoding into the description.
1406    
1407    ----------------------------------------------------------------------
1408    Name: Allen Leung (leunga (at) reservoir (dot) com)
1409    Date: 2004/07/13 15:04:00 EST
1410    Tag: leunga-110_46_1-ppc-lwarx
1411    Description:
1412    
1413       Added the two instructions LWARX and STWCX to the PowerPC
1414    instruction set.
1415    
1416       A (untested) rewrite of loop-structure.sml.  The old version
1417    is completely broken.
1418    
1419    ----------------------------------------------------------------------
1420    Name: Matthias Blume (blume (at) tti - c (dot) org)
1421    Date: 2004/07/13 13:50:00 CDT
1422    Tag: blume-20040713-nlffi
1423    Description:
1424    
1425    - use paramAlloc to report c-calls with too many arguments
1426      (for PPC version where parameter area is pre-allocated)
1427    - added ccall_maxargspace to machspec (to implement the above)
1428    - made "make" commend in CM's "make" tool configurable
1429    - added option (default: on) for passing the name of the SML/NJ's "bin"
1430      directory to "make"; the call looks like this:
1431    
1432         make <options> SMLNJ_BINDIR=<dir> <target>
1433    
1434      This can be used by the Makefile to, e.g., pick the "right" version
1435      of ml-nlffigen.
1436    - minor code tweaks
1437    
1438    ----------------------------------------------------------------------
1439    Name: Matthias Blume (blume (at) tti - c (dot) org)
1440    Date: 2004/07/12 22:50:00 CDT
1441    Tag: blume-110_46_1-macosx-nlffi
1442    Description:
1443    
1444    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1445    though.
1446    
1447    Note:  1. You have to make a new, clean build of the runtime system.
1448           2. There are new BOOTFILES, you have to use them!
1449              (Doing the bootstrap process yourself would be *very* painful!
1450               If you absolutely have to do it, build the system under
1451               a different architecture and then cross-compile.)
1452    
1453    Version bumped to 110.46.1 to account for runtime data format changes.
1454    
1455    ----------------------------------------------------------------------
1456    Name: Matthias Blume (blume (at) tti - c (dot) org)
1457    Date: 2004/06/18 14:30:00 CDT
1458    Tag: blume-20040618-unix
1459    Description:
1460    
1461    Changed the implementation of structure Unix so that the same stream
1462    is returned every time one of the {text,bin}{In,Out}streamOf functions
1463    is invoked on the same proc.  This is not what the spec currently
1464    says -- although IMO it arguably should.  (See discussion below.)
1465    
1466    ----------------------------------------------------------------------
1467    Name: Matthias Blume (blume (at) tti - c (dot) org)
1468    Date: 2004/06/17 18:15:00 CDT
1469    Tag: Release_110_46
1470    Description:
1471    
1472    New working version (110.46).  NEW BOOTFILES!
1473    
1474    ----------------------------------------------------------------------
1475    Name: Matthias Blume (blume (at) tti - c (dot) org)
1476    Date: 2004/06/17 17:20:00 CDT
1477    Tag: blume-20040617-timer-unix
1478    Description:
1479    
1480    Changed the interface of structures Timer and Unix to match the most
1481    recent Basis spec.
1482    
1483    In the case of Unix there still seems to be an open/weird issue:
1484    
1485       The {text,bin}{In,Out}streamOf functions are supposed to create
1486       fresh streams whenever they are called -- as opposed to have them
1487       return the same stream every time.  This design is supposed to
1488       prevent space leaks caused by proc values hanging on to streams.
1489    
1490       The reap function, on the other hand, is supposed to close the
1491       streams.  This cannot be done without having a handle on the
1492       stream in proc after all...
1493    
1494    I took the liberty to implement the following stopgap solution:
1495    
1496      The proc value hangs on to the most recently created stream(s).
1497      Reap closes those.  If either or both of the two streams hadn't
1498      been created at all yet, then reap will close the corresponding
1499      file descriptors directly.
1500    
1501    PS: I don't understand the original space leak argument anymore.  If
1502    a proc hangs on to the imperative stream, then I/O operations on those
1503    will advance the state of the cached stream and avoid the space leak.
1504    
1505    ----------------------------------------------------------------------
1506    Name: Matthias Blume (blume (at) tti - c (dot) org)
1507    Date: 2004/05/28 16:45:00 CDT
1508    Tag: blume-20040528-basis
1509    Description:
1510    
1511    Added signature PACK_REAL and exported functor PrimIO.
1512    
1513    ----------------------------------------------------------------------
1514    Name: Matthias Blume (blume (at) tti - c (dot) org)
1515    Date: 2004/05/25 16:00:00 CDT
1516    Tag: blume-20040525-group-owner
1517    Description:
1518    
1519    CM now ignores (but still accepts) the "owner" information in group
1520    descriptions.  The owner of a group is its next enclosing
1521    library. Each group must have a unique owner.  (There is a virtual
1522    "toplevel" library that owns groups which are not nested within a real
1523    library.)  Previously, each group had to explicitly declare its owner,
1524    and CM would check that such a declaration is correct.  The new scheme
1525    is to have CM check that for each group there is precisely one owning
1526    library.
1527    
1528    The advantage of the new scheme is that the programmer no longer needs
1529    to maintain the somewhat annoying owner information.  The downside is
1530    that CM cannot enforce the ownership rule across multiple runs of
1531    CM.make.  Fortunately, enclosing the same group in two different
1532    libraries A and B which are not part of the same program does not
1533    cause real problems.
1534    
1535    ----------------------------------------------------------------------
1536    Name: Matthias Blume (blume (at) tti - c (dot) org)
1537    Date: 2004/05/20 16:00:00 CDT
1538    Tag: blume-20040520-win32
1539    Description:
1540    
1541    Made the win32 version work again.  (Strangely, a misplaced comma had
1542    slipped into win32-process.c which prevented the runtime from being
1543    compiled correctly.)
1544    
1545    Also, included a minor addition to ml-build.bat analogous to what was
1546    done in blume-20040519-ml-build.
1547    
1548    ----------------------------------------------------------------------
1549    Name: Matthias Blume (blume (at) tti - c (dot) org)
1550    Date: 2004/05/19 22:10:00 CDT
1551    Tag: blume-20040519-ml-build
1552    Description:
1553    
1554    Arranged for ml-build to clean up after itself a little bit better.
1555    The script generates a temporary SML source file and compiles it using
1556    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1557    gets rid of those at the end, so they don't accumulate under .cm.
1558    
1559    This required a minor change to install.sh because the name of the
1560    metadata directory (default: .cm) is actually configurable at
1561    installation time.
1562    
1563    ----------------------------------------------------------------------
1564    Name: Matthias Blume (blume (at) tti - c (dot) org)
1565    Date: 2004/05/18 15:50:00 CDT
1566    Tag: blume-20040518-mkreader
1567    Description:
1568    
1569    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1570    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1571    
1572    ----------------------------------------------------------------------
1573    Name: Matthias Blume (blume (at) tti - c (dot) org)
1574    Date: 2004/05/11 14:35:00 CDT
1575    Tag: blume-20040511-win32sock
1576    Description:
1577    
1578    Added previously missing support for many socket-related functions
1579    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1580    for the voluminous patch!
1581    
1582    (I have not tested this patch under win32 yet.)
1583    
1584    Here is David's e-mail:
1585    
1586    Hi,
1587    
1588    Attached to this email you find a diff against sml/nj 110.45
1589    that will enable socket support under Windows.
1590    
1591    To apply the patch (using unix or cygwin)
1592    1) gunzip runtime.diff.gz
1593    2) "cd" into "src/runtime" in the source tree of a fresh
1594       110.45 installation.
1595    3) patch -p 1 < [your/path/to]runtime.diff
1596    
1597    The code compiles fine but has NOT yet been extensively tested.
1598    I only ran a few tests for basic socket client functionality
1599    (which worked fine).  Especially the functions that use ioctl
1600    are not tested at all and might not work (see below).
1601    
1602    I implemented this since we want to move to a newer version of sml/nj
1603    but need socket support in order to use it.  This is the first time I
1604    even had a look at the sml/nj source,  so please review my changes
1605    before making this part of the distribution!  Here are a few issues
1606    that I think might be better for someone to solve who is more
1607    familiar with the sml/nj source (and socket programming):
1608    
1609    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1610      exception since I could not figure out what the windows equivalent
1611      of these functions is
1612    
1613    - In sockets-osdep.h there are a some #include statements that are
1614      only used in a few files that include sockets-osdep.h
1615    
1616    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1617      does not process its return value since I don't know how to
1618      report an error upwards.
1619    
1620    - It would probably be good to have a call to WSACleanup() when
1621      the library is unloaded (if there is such a possibility).
1622      Otherwise I think Windows will take care of this automatically
1623      when the process finishes.
1624    
1625    - I used ioctlsocket() as a replacement for ioctl() but I have
1626      no idea if that is actually the proper replacement on Windows.
1627    
1628    - All these issues are marked in the code by "FIXME" comments.
1629    
1630    We use sml/nj extensively in our products and are quite happy
1631    with it.  I hope this contribution will help you.
1632    
1633    Keep up the good work!
1634    
1635    David
1636    
1637    ----------------------------------------------------------------------
1638    Name: Matthias Blume (blume (at) tti - c (dot) org)
1639    Date: 2004/05/11 14:20:00 CDT
1640    Tag: blume-20040511-installml
1641    Description:
1642    
1643    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1644    the patch.)
1645    
1646    ----------------------------------------------------------------------
1647    Name: Matthias Blume (blume (at) tti - c (dot) org)
1648    Date: 2004/05/11 14:05:00 CDT
1649    Tag: blume-20040511-nlffi-netbsd
1650    Description:
1651    
1652    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1653    the patch.)
1654    
1655    ----------------------------------------------------------------------
1656    Name: Matthias Blume (blume (at) tti - c (dot) org)
1657    Date: 2004/05/11 12:05:00 CDT
1658    Tag: blume-20040511-exports
1659    Description:
1660    
1661    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1662    export lists in compiler-related .cm-files.
1663    
1664    ----------------------------------------------------------------------
1665    Name: Matthias Blume (blume (at) tti - c (dot) org)
1666    Date: 2004/05/11 11:35:00 CDT
1667    Tag: blume-20040511-allsource
1668    Description:
1669    
1670    The installer now honors the "src-smlnj" target again, although its meaning
1671    has changed from "all sources required for the compiler" to "all sources
1672    the installer knows about".  In other words, if you enable "src-smlnj"
1673    in the "targets" file, then the installer will pull in sources for
1674    everything.  (Notice that this refers to source code only.  Compiled
1675    code is still only installed for modules that were requested explicitly
1676    or which are required for other modules that were requested explicitly.)
1677    
1678    ----------------------------------------------------------------------
1679    Name: Matthias Blume (blume (at) tti - c (dot) org)
1680    Date: 2004/04/23 17:40:00 CDT
1681    Tag: blume-20040423-ieee-scan
1682    Description:
1683    
1684    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1685    in the exponent calculation we get INF or ZERO (depending on the mantissa
1686    and the sign of the exponent).
1687    
1688    ----------------------------------------------------------------------
1689    Name: Matthias Blume (blume (at) tti - c (dot) org)
1690    Date: 2004/04/23 10:40:00 CDT
1691    Tag: blume-20040423-ml-build
1692    Description:
1693    
1694    The ml-build script now terminates with a non-0 status when something
1695    goes wrong.
1696    
1697    ----------------------------------------------------------------------
1698    Name: Matthias Blume (blume (at) tti - c (dot) org)
1699    Date: 2004/04/22 16:35:00 CDT
1700    Tag: blume-20040422-Option
1701    Description:
1702    
1703    Made exception Option to be the same as exception Option.Option
1704    (as it should be).
1705    
1706    ----------------------------------------------------------------------
1707    Name: Allen Leung (leunga (at) reservoir (dot) com)
1708    Date: 2004/03/19 14:40:00 EST
1709    Tag: leunga-20040319-cygwin-nlffi
1710    Description:
1711    
1712    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1713    of SML/NJ.  The problem is that
1714    
1715       lib = dlopen(NULL, ...)
1716       f   = dlsym(lib, "malloc");
1717    
1718    does not work on Windows unless we explicitly export symbols
1719    such as 'malloc' during linking.  We fixed this by explicitly
1720    exporting the required symbols with the magic gcc incantation:
1721    
1722        -Wl,--export-all cygwin.def
1723    
1724    where cygwin.def is a file containing all the symbols that
1725    we wish to export.
1726    
1727    I suspect this is a Windows problem and we'll have to
1728    do the same (somehow with windows compilers) when
1729    we build the native win32 version with the system
1730    calls LoadLibrary/GetProcAddress.
1731    
1732    ----------------------------------------------------------------------
1733    Name: Matthias Blume (blume (at) tti - c (dot) org)
1734    Date: 2004/03/04 16:35:00 CST
1735    Tag: blume-20040304-intinf-fmt
1736    Description:
1737    
1738    Fixed problem with IntInf.fmt (sign would show up on the right instead
1739    of on the left for BIN, OCT, and HEX).
1740    
1741    ----------------------------------------------------------------------
1742    Name: Matthias Blume (blume (at) tti - c (dot) org)
1743    Date: 2004/03/04 11:25:00 CST
1744    Tag: blume-20040304-symlinks
1745    Description:
1746    
1747    Fixed problem with installer script (unix only) where bin/ml-yacc and
1748    friends pointed (via symlinks) to absolute locations instead of just
1749    .run-sml.  This was reported by Vesa A Norrman.
1750    
1751    ----------------------------------------------------------------------
1752    Name: Matthias Blume (blume (at) tti - c (dot) org)
1753    Date: 2004/02/13 14:50:00 CST
1754    Tag: Release_110_45
1755    Description:
1756    
1757    New working version (110.45). New bootfiles.
1758    
1759    ----------------------------------------------------------------------
1760    Name: Matthias Blume (blume (at) tti - c (dot) org)
1761    Date: 2004/01/26 15:15:15 CST
1762    Tag: blume-20040126-toplevel
1763    Description:
1764    
1765    Improved handling of exceptions at the interactive toplevel.
1766    
1767    ----------------------------------------------------------------------
1768    Name: Matthias Blume (blume (at) tti - c (dot) org)
1769    Date: 2004/01/26 11:25:00
1770    Tag: blume-20040126-app
1771    Description:
1772    
1773    Type of top-level "app" corrected.
1774    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1775    
1776    ----------------------------------------------------------------------
1777    Name: Matthias Blume (blume (at) tti - c (dot) org)
1778    Date: 2003/11/18 17:10 CST
1779    Tag: blume-20031118-basis-fiddle
1780    Description:
1781    
1782    - changed Timer interface to what might become the spec
1783    - POSIX_FLAGS -> BIT_FLAGS according to spec
1784    - some other minor discrepancies wrt. spec eliminated
1785    
1786    ----------------------------------------------------------------------
1787    Name: Matthias Blume (blume (at) tti - c (dot) org)
1788    Date: 2003/11/06 12:00:00 CST
1789    Tag: Release_110_44
1790    Description:
1791    
1792    New working version (110.44). New bootfiles.
1793    
1794    ----------------------------------------------------------------------
1795    Name: Matthias Blume (blume (at) tti - c (dot) org)
1796    Date: 2003/11/04 11:50:00 CST
1797    Tag: blume-20031104-move-libraries
1798    Description:
1799    
1800    Eliminated the "dont_move_libraries" directive in config/targets.
1801    (The mechanism was broken and could not be fixed easily.  Moreover,
1802    there does not seem to be any reason not to move all libraries into
1803    lib during installation.  I originally implemented this directive as a
1804    backward-compatibility feature when I first introduced the new CM.
1805    Now that things have been stable for a long time and going back to the
1806    old CM is not an option, there is no reason to keep it around.)
1807    
1808    ----------------------------------------------------------------------
1809    Name: Matthias Blume (blume (at) tti - c (dot) org)
1810    Date: 2003/11/03 16:00:00 CST
1811    Tag: blume-20031103-installdir
1812    Description:
1813    
1814    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1815    Richards for pointing out the problem and providing the solution.)
1816    
1817    ----------------------------------------------------------------------
1818    Name: Matthias Blume (blume (at) tti - c (dot) org)
1819    Date: 2003/10/01 17:05:00 CDT
1820    Tag: blume-20031001-lal-mlrisc
1821    Description:
1822    
1823    MLRISC bug fix from Lal.
1824    
1825    ----------------------------------------------------------------------
1826    Name: Matthias Blume (blume (at) tti - c (dot) org)
1827    Date: 2003/09/30 16:10:00 CDT
1828    Tag: blume-20030930-primio-bat
1829    Description:
1830    
1831    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1832    2. Improved .bat files (for Win32 port) to make things work under Win95.
1833       (thanks to Aaron S. Hawley for this one)
1834    
1835    ----------------------------------------------------------------------
1836    Name: Matthias Blume (blume (at) tti - c (dot) org)
1837    Date: 2003/09/26 16:05:00 CDT
1838    Tag: blume-20030926-wrappriv
1839    Description:
1840    
1841    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1842    
1843    ----------------------------------------------------------------------
1844    Name: Matthias Blume (blume (at) tti - c (dot) org)
1845    Date: 2003/09/26 15:00:00 CDT
1846    Tag: blume-20030926-110_43_3
1847    Description:
1848    
1849    - additional cleanup
1850    - version number bump, NEW BOOTFILES
1851    
1852    ----------------------------------------------------------------------
1853    Name: Matthias Blume (blume (at) tti - c (dot) org)
1854    Date: 2003/09/26 12:00:00 CDT
1855    Tag: blume-20030926-ppautoload
1856    Description:
1857    
1858    I modified the read-eval-print loop so that the autoloader gets
1859    invoked whenever the prettyprinter tries to look up a symbol that
1860    is not currently defined in the toplevel environment but which
1861    appears in CM's autoload registry.  As a result, we see far fewer of
1862    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1863    
1864    In addition to this I tried to clean up some pieces of the Basis
1865    implementation (e.g., Socket, Word8Array) in order to prevent other
1866    instances of these ?.Foo.Bar.xxx names from being printed.
1867    
1868    The mechanism that picks names for types still needs some work, though.
1869    (Right now it seems that if there is a type A.t which is defined to
1870    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1871    "?.B.u" although the perhaps more sensible solution would be to use
1872    "A.t" in this case.  In other words, the prettyprinter should follow
1873    a chain of DEFtycs not farther than there are corresponding toplevel
1874    names in the current environment.)
1875    
1876    ----------------------------------------------------------------------
1877    Name: Matthias Blume (blume (at) tti - c (dot) org)
1878    Date: 2003/09/24 16:31:00 CDT
1879    Tag: blume-20030924-installer
1880    Description:
1881    
1882    Another installer tweak:  All the ML code for the installer is now
1883    compiled during CMB.make and put into a little library called
1884    $smlnj/installer.cm.  The installation then simply invokes
1885    
1886       sml -m $smlnj/installer.cm
1887    
1888    and everything happens automagically.
1889    
1890    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1891    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1892            and BINDIR.
1893    
1894    The new scheme guarantees that the ML code responsible for the installation
1895    is in sync with the APIs of the main system.  Also, the installer is
1896    somewhat faster because the installer script is precompiled.
1897    
1898    ----------------------------------------------------------------------
1899    Name: Matthias Blume (blume (at) tti - c (dot) org)
1900    Date: 2003/09/24 15:35:00 CDT
1901    Tag: blume-20030924-synsock
1902    Description:
1903    
1904    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1905    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1906    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1907    implementations of non-blocking operations from CML's Socket
1908    structure.
1909    
1910    ----------------------------------------------------------------------
1911    Name: Matthias Blume (blume (at) tti - c (dot) org)
1912    Date: 2003/09/24 15:10:05 CDT
1913    Tag: blume-20030924-sockets
1914    Description:
1915    
1916    1. Fixed SOCKET API and implementation to match Basis spec.
1917       This required changing the internal representation of sockets to one
1918       that remembers (for each socket file descriptor) whether it is currently
1919       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1920       call is made only if the state actually needs to change).
1921    
1922    2. OS-specific details of sockets were moved into separate files, thus
1923       making it possible to unify the bulk of the socket implementations
1924       between Unix and Win32.
1925    
1926    3. CML's socket API changed accordingly.
1927       (Note that we need to remove non-blocking functions from this API
1928       since they are redundant in the case of CML!)
1929    
1930    4. CML's socket implementation now makes use of non-blocking functions
1931       provided by Basis, thus removing all OS-dependent code from this part
1932       of CML.
1933    
1934    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1935    
1936    ----------------------------------------------------------------------
1937    Name: Matthias Blume (blume (at) tti - c (dot) org)
1938    Date: 2003/09/22 12:10:00 CDT
1939    Tag: blume-20030922-110_43_2
1940    Description:
1941    
1942    Made a new interim version and bootfiles for developer's bootstrapping
1943    convenience.
1944    
1945       110.43.2 -- NEW BOOTFILES
1946    
1947    ----------------------------------------------------------------------
1948    Name: Matthias Blume (blume (at) tti - c (dot) org)
1949    Date: 2003/09/19 15:55:00 CDT
1950    Tag: blume-20030919-cmdir
1951    Description:
1952    
1953    1. new-install.sh -> install.sh
1954    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1955    3. tweaked installer so that another name instead of .cm can be chosen
1956       at install time (by setting the CM_DIR_ARC environment variable
1957       during installation); once installation is complete, the name is
1958       fixed
1959    
1960    ----------------------------------------------------------------------
1961    Name: Matthias Blume (blume (at) tti - c (dot) org)
1962    Date: 2003/09/18 16:00:00 CDT
1963    Tag: blume-20030918-110_43_1
1964    Description:
1965    
1966    Made a new interim version and bootfiles for developer's bootstrapping
1967    convenience.
1968    
1969       110.43.1 -- NEW BOOTFILES
1970    
1971    ----------------------------------------------------------------------
1972    Name: Matthias Blume (blume (at) tti - c (dot) org)
1973    Date: 2003/09/18 15:20:00 CDT
1974    Tag: blume-20030918-misc
1975    Description:
1976    
1977    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1978       this seems to be controversial at the moment) and used those in
1979       Posix.ProcEnv.times.
1980    
1981    2. Added Time.{from,to}Nanoseconds to Time.
1982    
1983    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1984       For example, fromLargeInt never needs to look at more than 3 "big
1985       digits" to get its 53 bits of precision.
1986    
1987    ----------------------------------------------------------------------
1988    Name: Matthias Blume (blume (at) tti - c (dot) org)
1989    Date: 2003/09/17 16:30:00 CDT
1990    Tag: blume-20030917-real32-slices
1991    Description:
1992    
1993    Added an entry to the primitive environment
1994    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1995    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1996    
1997    Removed some of the "magic" constants in real64.sml and replaced them
1998    with code that generates these values from their corresponding
1999    integer counterparts.
2000    
2001    Made all(?) the slice-related changes to the Basis and made everything
2002    compile again...
2003    
2004    ----------------------------------------------------------------------
2005    Name: Matthias Blume (blume (at) tti - c (dot) org)
2006    Date: 2003/09/15 17:45:00 CDT
2007    Tag: blume-20030915-rbase
2008    Description:
2009    
2010    Fixed bug in Real.fromLargeInt.
2011    
2012    ----------------------------------------------------------------------
2013    Name: Matthias Blume (blume (at) tti - c (dot) org)
2014    Date: 2003/09/13 18:11:00 CDT
2015    Tag: blume-20030913-libinstall
2016    Description:
2017    
2018    Minor bugfix in config/libinstall (set anchor with path to
2019    standalone tool after installing it, otherwise libraries that
2020    need ml-lex or ml-yacc won't compile the first time the installer
2021    runs).
2022    
2023    ----------------------------------------------------------------------
2024    Name: Matthias Blume (blume (at) tti - c (dot) org)
2025    Date: 2003/09/12 11:45:00 CDT
2026    Tag: blume-20030912-various
2027    Description:
2028    
2029     - fixed bug in Real.toLargeInt
2030     - fixed bug in Posix.ProcEnv.times
2031     - changed inputLine functions to return an option
2032     - minor installer improvements / bugfixes
2033     - changed default @SMLalloc parameter for x86/celeron to 64k
2034    
2035    ----------------------------------------------------------------------
2036    Name: Matthias Blume (blume (at) tti - c (dot) org)
2037    Date: 2003/09/09 22:00:00 CDT
2038    Tag: Release_110_43
2039    Description:
2040    
2041        New working release 110.43.  New bootfiles.
2042    
2043    ----------------------------------------------------------------------
2044    Name: Matthias Blume (blume (at) tti - c (dot) org)
2045    Date: 2003/09/09 19:20:00 CDT
2046    Tag: blume-20030909-installer
2047    Description:
2048    
2049    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
2050    Modified config/install.bat to take advantage of it.  Also modified
2051    config/install.sh (and called it config/new-install.sh) to take advantage
2052    of it on Unix systems.  (The SML code is (supposed to be) platform-
2053    independent.)
2054    
2055    The installer can now install everything under Win32
2056    as well as under *nix as long as it compiles.
2057    
2058    Other changes:
2059    
2060      - made CML compile again under Win32
2061      - made eXene compile under Win32 (by providing a fake structure UnixSock
2062        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
2063      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
2064        same as type int;  under Win32 it isn't
2065      - fixed some slice-related problems in the win32-specific parts of CML
2066      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
2067        (under Win32, the volume name is case-insensitive, and the
2068        OS.Path code compares volume names for equality)
2069    
2070    ----------------------------------------------------------------------
2071    Name: Matthias Blume (blume (at) tti - c (dot) org)
2072    Date: 2003/09/08 11:55:00 CDT
2073    Tag: blume-20030908-fullpath
2074    Description:
2075    
2076    Made Win32 version of OS.FileSys.fullPath return current directory
2077    when given an empty string.  This is what the spec says, and incidentally,
2078    CM depends on it.  (CM otherwise goes into an infinite loop in certain
2079    cases when presented with the name of a non-existing .cm file.)
2080    
2081    ----------------------------------------------------------------------
2082    Name: Matthias Blume (blume (at) tti - c (dot) org)
2083    Date: 2003/09/04 16:30:00 CDT
2084    Tag: blume-20030905-slices-etc
2085    Description:
2086    
2087      1. Changed interface to vectors and arrays in Basis to match
2088         (draft) Basis spec.
2089      2. Added signatures and implementations of slices according to
2090         Basis spec.
2091      3. Edited source code throughout the system to make it compile again
2092         under 1. and 2.  (In some cases code had to be added to have it
2093         match the new signatures.)
2094      4. MLRISC should be backward-compatible: the copies of the originals
2095         of files that needed to change under 3. were retained, the .cm files
2096         check the compiler version number and use old versions when
2097         appropriate.
2098      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2099         dirstream -> string option (in accordance with Basis spec).
2100      6. When generating code that counts lines, ml-lex used function
2101         CharVector.foldli, taking advantage of its old interface.
2102         This has been replaced with the corresponding code from
2103         CharVectorSlice.  (html-lex must be re-lexed!)
2104      7. BitArray in smlnj-lib/Util has been extended/modified to match the
2105         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
2106      8. Removed temporary additions (fromInternal, toInternal) from the
2107         (now obsolete) IntInf in smlnj-lib/Util.
2108      9. Cleaned up structure Byte.
2109     10. Added localOffset, scan, and fromString to Date (according to spec).
2110         Cleaned/corrected implementation of Date.
2111         (Still need to check for correctness; implement better canonicalizeDate.)
2112     11. Added "scan" to signature IEEE_REAL.
2113     12. Some improvements to IntInf [in particular: efficiency-hack for
2114         mod and rem when second operand is 2 (for parity checks).]
2115     13. Changed representation of type Time.time, using a single IntInf.int
2116         value counting microseconds.  This considerably simplified the
2117         implementation of structure Time.  We now support negative time
2118         values; scan and fromString handle signs.
2119     14. Functor PrimIO now takes two additional arguments (VectorSlice and
2120         ArraySlice).
2121    
2122    ----------------------------------------------------------------------
2123    Name: Matthias Blume (blume (at) tti - c (dot) org)
2124    Date: 2003/08/28 17:00:00 CDT
2125    Tag: blume-20030828-intinf
2126    Description:
2127    
2128    This is a major update which comes with a version number bump
2129    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2130    and an implementation of IntInf in the Basis.
2131    
2132    There are a fairly large number of related changes and updates throughout
2133    the system:
2134    
2135    Basis:
2136            - Implemented IntInf.
2137            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
2138            - Added some missing Real64 operations, most notably Real.toLargeInt.
2139            - Added FixedInt as a synonym for Int32.
2140    
2141    compiler:
2142            * Added support for a built-in intinf type.
2143              - literals
2144              - pattern matching
2145              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2146              - overloading on literals and operations
2147    
2148            This required adding a primitive type intinf, some additional
2149            primops, and implementations for several non-trivial intinf
2150            operations in Core.  (The intinf type is completely abstract
2151            to the compiler; all operations get delegated back to the Core.)
2152    
2153            * Intinf equality is handled by polyequal.  However, the compiler
2154            does not print its usual warning in this case (since polyequal
2155            is the right thing to do there).
2156    
2157            * Improved the organization of structure InlineT.
2158    
2159            * A word about conversion primops:
2160            If conversions involving intinf do not cancel out during
2161            CPS contract, then the compiler must insert calls to Core functions.
2162            Since all core access must be resolved already during the FLINT
2163            translate phase, it would be too late a the time of CPS contract
2164            to add new Core calls.  For this reason, conversion primops
2165            for intinf carry two arguments:  1. the numeric argument that
2166            they are supposed to convert, and 2. the Core function that
2167            can help with this conversion if necessary.  If CPS contract
2168            eliminates a primop, then the associated Core function becomes
2169            dead and goes away.  Intinf conversion primops that do not get
2170            eliminated by CPS contract get rewritten into calls of their
2171            core functions by a separate, new phase.
2172    
2173    interactive system:
2174            - Control.Print.intinfDepth controls max length of intinf constants
2175            being printed. (Analogous to Control.Print.stringDepth.)
2176            - Cleanup in printutil and pputil: got rid of unused stuff and
2177            duplicates; replaced some of the code with code that makes better
2178            use of library functionality.
2179    
2180    CM:
2181            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2182            are no longer silent.
2183    
2184    CKIT:
2185            Fixed mismatched uses of Int32 and LargeInt.  I always decided
2186            in favor of LargeInt -- which is now the same as IntInf.
2187            CKIT-knowledgable people should check whether this is what's
2188            intended and otherwise change things back to using Int32 or
2189            FixedInt.
2190    
2191    Throughout the code:
2192            Started using IntInf.int literals and built-in operations
2193            (e.g., comparison with 0) where this seems appropriate.
2194    
2195    
2196    ----------------------------------------------------------------------
2197    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2198    Date: 2003/08/13 11:36:00 CDT
2199    Tag: dbm-20030813-mcz-merge1
2200    Description:
2201    
2202    Merging changes from the mcz-branch development branch into trunk.
2203    These changes involve replacement of the emulated old prettyprinter
2204    interface with direct use of the SML/NJ Lib PP library, and fixing
2205    of a couple of bugs (895, 1186) relating to error messages.  A new
2206    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2207    has been added.
2208    
2209    ----------------------------------------------------------------------
2210    Name: Matthias Blume (blume (at) tti - c (dot) org)
2211    Date: 2003/08/11 15:45:00 CDT
2212    Tag: blume-20030811-windows
2213    Description:
2214    
2215    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2216    
2217        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2218    
2219    This patch restores SML/NJ's ability to run under win32.  There are a
2220    number of changes, including fixes for several bugs that had gone
2221    unnoticed until now:
2222    
2223      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2224        trying to run the win32 version from within cygwin.)
2225    
2226      - There are a number of simple .bat scripts that substitute for their
2227        corresponding Unix shell-scripts.  (See below.)
2228    
2229      - The internals of ml-build have been modified slightly.  The main
2230        difference is that instead of calling ".link-sml" (or link-sml.bat)
2231        using OS.Process.system, the ML process delegates this task back
2232        to the script.  Otherwise problems arise in mixed environments such
2233        as Cygwin where scripts look and work like Unix scripts, but
2234        where OS.Process.system cannot run them.
2235    
2236      - In CM, the srcpath pickler used native pathname syntax -- which
2237        is incorrect in the case of cross-compilation.  The new pickle format
2238        is independent of platform-specific naming conventions.
2239    
2240      - Path configuration files (such as lib/pathconfig) can now choose
2241        between native and standard syntax.  Placing a line of the form
2242    
2243            standard!
2244    
2245        into the file causes all subsequent paths to be interpreted using
2246        CM standard pathname syntax (= Unix conventions); a line
2247    
2248            native!
2249    
2250        switches back to native style.  This was needed so that
2251        path config files can be written portably, see src/system/pathconfig.
2252    
2253      - Runtime system:
2254    
2255          - win32-filesys.c:  get_file_time and set_file_time now
2256            access modification time, not creation time.
2257    
2258          - I/O code made aware of new array representation.
2259    
2260          - Bug fixes in X86.prim.masm.
2261    
2262       - src/system/makeml made aware of win32. (For use under cygwin
2263            and other Unix-environments for windows.)
2264    
2265       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2266         which caused BinIO.inputAll to fail consistently.
2267    
2268    .bat scripts:
2269    
2270       Windows .bat scripts assume that SMLNJ_HOME is defined.
2271    
2272       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2273         applications (sml, ml-yacc, ml-lex).
2274       - ml-build.bat: analogous to ml-build.
2275       - config\install.bat:  Analogous to config/install.sh.  This requires
2276         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2277         (nmake etc. must be on the path, and vcvars32 must have been run.)
2278         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2279         and the bootfile hierarchy must have been unpacked under
2280         sml.boot.x86-win32.
2281         The script is very primitive and does a poor job at error checking.
2282         It only installs the base system, ml-lex, and ml-yacc.  No other
2283         libraries are being installed (i.e., you get only those that
2284         are part of the compiler.)
2285       - link-sml.bat:  analogous to .link-sml, but not currently used
2286    
2287    Unrelated bug fixes:
2288    
2289      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2290        types.
2291      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2292        otherwise.)
2293    
2294    ----------------------------------------------------------------------
2295    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2296    Date: 2003/06/17
2297    Tag: macqueen-20030617-bug895
2298    Description:
2299    
2300    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2301    Tag will be used for new development branch (mcz-branch) for
2302    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2303    
2304    ----------------------------------------------------------------------
2305    Name: Matthias Blume (blume (at) tti - c (dot) org)
2306    Date: 2003/05/27 16:55:00 CDT
2307    Tag: blume-20030527-polyeq
2308    Description:
2309    
2310    Tried to eliminated most cases of polymorphic equality.
2311    
2312    ----------------------------------------------------------------------
2313    Name: Matthias Blume (blume (at) tti - c (dot) org)
2314    Date: 2003/05/21 17:45:00 CDT
2315    Tag: blume-20030517-complete
2316    Description:
2317    
2318    Two changes:
2319    
2320      1. Added a flag for controlling whether non-exhaustive bindings will
2321         be treated as errors (default is false).
2322      2. Cleaned up the *entire* source tree so that CMB.make goes through
2323         without a single non-exhaustive match- or bind warning.
2324    
2325    ----------------------------------------------------------------------
2326    Name: Matthias Blume (blume (at) tti - c (dot) org)
2327    Date: 2003/05/17 10:20:00 CDT
2328    Tag: blume-20030517-absyn
2329    Description:
2330    
2331    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2332    
2333       This mainly affects the quality of error messages.  However, some
2334       of the code is now more straightforward than before. (Treatment of
2335       the above four constructs in translate.sml is much simpler than
2336       the "macro-expansion" that was going on before.  Plus, the mach-
2337       compiler no longer gets invoked just to be able to compile an
2338       if-expression.)
2339    
2340    2. The ErrorMsg.Error exception is now caught and absorbed by the
2341       interactive loop.
2342    
2343    ----------------------------------------------------------------------
2344    Name: Allen Leung
2345    Date: 2003/05/16 13:05:00 CDT
2346    Tag: leunga-20030516-cygwin-runtime
2347    Description:
2348    
2349        Ported the runtime system to cygwin, which uses the unix
2350    x86-unix bin files.  Missing/buggy features:
2351    
2352       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2353         the Cygwin library.
2354       o Ctrl-C handling may be flaky.
2355       o Windows system calls and Windows I/O are not supported.
2356    
2357       A new set of binfiles is located at:
2358    
2359         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2360    
2361       This is only needed for bootstrapping the cygwin version of smlnj.
2362       Other x86 versions can use the existing binfiles.
2363    
2364    ----------------------------------------------------------------------
2365    Name: Matthias Blume
2366    Date: 2003/04/08 15:42:00 CDT
2367    Tag: blume-20030408-listpair
2368    Description:
2369    
2370    1. Added a target 'mlrisc' to installer.
2371    
2372    2. Added missing elements to structure ListPair.
2373    
2374    ----------------------------------------------------------------------
2375    Name: Allen Leung
2376    Date: 2003/01/07 10:40:00 EST
2377    Tag: leunga-20030107-int-rem
2378    Description:
2379    
2380    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2381    The arguments to the SUBL instruction were swapped.
2382    
2383    ----------------------------------------------------------------------
2384    Name: Matthias Blume
2385    Date: 2002/12/12 16:25:00 EST
2386    Tag: blume-20021212-risc-ra
2387    Description:
2388    
2389    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2390    sent the RA into an infinite loop when floating point registers get
2391    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2392    110.37 and 110.38.)
2393    There was another set of potential problems related to the handling of
2394    MLRISC annotations (but those did not yet cause real problems, apparently).
2395    
2396    ----------------------------------------------------------------------
2397    Name: Matthias Blume
2398    Date: 2002/12/06 22:40:00 EST
2399    Tag: blume-20021206-cm-fileid
2400    Description:
2401    
2402    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2403    This fixes the problem of CM getting confused by files that suddenly
2404    change their identity (e.g., by getting unlinked and recreated by some
2405    text editor such as vi).  There might be a better/cheaper/cleaner way
2406    of doing this, but for now this will have to do.
2407    
2408    ----------------------------------------------------------------------
2409    Name: Matthias Blume
2410    Date: 2002/10/28 09:50:00 EST
2411    Tag: blume-20021028-typecheck
2412    Description:
2413    
2414    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2415    
2416    ----------------------------------------------------------------------
2417    Name: Matthias Blume
2418    Date: 2002/10/17 09:10:00 EDT
2419    Tag: Release_110_42
2420    Description:
2421    
2422    In good old tradition, there has been a slight hiccup so that we have
2423    to patch 110.42 after the fact.   The old release tag has been replaced
2424    (see below).
2425    
2426    The change solves a problem with two competing approaches the
2427    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2428    each other's way.
2429    
2430    This change only affects the runtime system code and the installer script.
2431    (No new bootfiles.)
2432    
2433    ----------------------------------------------------------------------
2434    Name: Matthias Blume
2435    Date: 2002/10/16 12:00:00 EDT
2436    Tag: Release_110_42_removed
2437    Description:
2438    
2439    New working release.  New bootfiles.
2440    
2441    ----------------------------------------------------------------------
2442    Name: Matthias Blume
2443    Date: 2002/10/10 13:10:00 EDT
2444    Tag: blume-20021010-ppc-divs
2445    Description:
2446    
2447    The mltree operator DIVS must be implemented with an overflow check on
2448    the PPC because the hardware indicates divide-by-zero using "overflow" as
2449    well.
2450    
2451    ----------------------------------------------------------------------
2452    Name: Matthias Blume
2453    Date: 2002/07/23 11:45:00
2454    Tag: blume-20020723-smlnj-home
2455    Description:
2456    
2457    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2458    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2459    CM_PATHCONFIG is also set), the path configuration file is assumed
2460    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2461    the entire tree to some other place and everything will "just work".
2462    
2463    (Companion commands such as ml-build and ml-makedepend also sense this
2464    variable.)
2465    
2466    ----------------------------------------------------------------------
2467    Name: Matthias Blume
2468    Date: 2002/07/12 21:19:00 EDT
2469    Tag: blume-20020712-liveness
2470    Description:
2471    
2472    Exported two useful "step" functions from liveness module (MLRISC).
2473    
2474    ----------------------------------------------------------------------
2475    Name: Matthias Blume
2476    Date: 2002/07/05 16:00 EDT
2477    Tag: Release_110_41
2478    Description:
2479    
2480    New working release.  New bootfiles.
2481    
2482    ----------------------------------------------------------------------
2483    Name: Matthias Blume
2484    Date: 2002/07/05 10:25:00 EDT
2485    Tag: blume-20020705-btimp
2486    Description:
2487    
2488    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2489    other clients can set up backtracing support.
2490    
2491    ----------------------------------------------------------------------
2492    Name: Matthias Blume
2493    Date: 2002/06/25 17:23:00 EDT
2494    Tag: blume-20020625-fpmax
2495    Description:
2496    
2497    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2498    case.  (The sense of the isNaN test was reversed -- which made min and
2499    max always return their first argument.)
2500    
2501    ----------------------------------------------------------------------
2502    Name: Matthias Blume
2503    Date: 2002/06/11
2504    Tag: blume-20020611-unixpath
2505    Description:
2506    
2507    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2508    
2509    ----------------------------------------------------------------------
2510    Name: Matthias Blume
2511    Date: 2002/06/10 16:35:00 EDT
2512    Tag: blume-20020610-ieeereal
2513    Description:
2514    
2515    I back-ported my implementation of IEEEReal.fromString from the
2516    idlbasis-devel branch so that we can test it.
2517    
2518    Another small change is that ppDec tries to give more information
2519    than just "<sig>" in the case of functors.  However, this code is
2520    broken in some mysterious way if the functor's body's signature
2521    has not been declared by ascription but gets inferred from the
2522    implementation.  This needs fixing...
2523    
2524    ----------------------------------------------------------------------
2525    Name: Matthias Blume
2526    Date: 2002/05/31
2527    Tag: blume-20020531-btrace-mode
2528    Description:
2529    
2530    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2531    the wayside when I switched over to using Controls everywhere.)
2532    
2533    ----------------------------------------------------------------------
2534    Name: Lal George
2535    Date: 2002/05/23 12:21:40 EDT
2536    Tag: george-20020523-visual-labels
2537    Description:
2538    
2539    Labels are now displayed in the graphical output to make
2540    the fall-through and target blocks obvious.
2541    
2542    ----------------------------------------------------------------------
2543    Name: Matthias Blume
2544    Date: 2002/05/22 11:03:00 EDT
2545    Tag: blume-20020522-shrink
2546    Description:
2547    
2548    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2549    situation that comes up when allocating huge arrays.
2550    
2551    ----------------------------------------------------------------------
2552    Name: Matthias Blume
2553    Date: 2002/05/21 16:00:00 EDT
2554    Tag: Release_110_40
2555    Description:
2556    
2557    New working release (110.40).  New bootfiles.
2558    
2559    [Also: John Reppy fixed GC bug 1131.]
2560    
2561    ----------------------------------------------------------------------
2562    Name: Matthias Blume
2563    Date: 2002/05/21 12:35:00 EDT
2564    Tag: blume-20020521-cmdoc
2565    Description:
2566    
2567    CM documentation update.
2568    
2569    ----------------------------------------------------------------------
2570    Name: Matthias Blume
2571    Date: 2002/05/21 10:55:00 EDT
2572    Tag: blume-20020521-misc
2573    Description:
2574    
2575    - John tweaked runtime to be silent on heap export (except when
2576      GC messages are on).
2577    - I added a few more things (cross-compiling versions of CMB) to
2578      config/preloads (as suggestions).
2579    
2580    ----------------------------------------------------------------------
2581    Name: Matthias Blume
2582    Date: 2002/05/20 22:25:00 EDT
2583    Tag: blume-20020520-controls
2584    Description:
2585    
2586    - Added ControlUtil structure to control-lib.cm.
2587    - Use it throughout.
2588    - Used Controls facility to define MLRISC controls (as opposed to
2589      registering MLRISC control ref cells with Controls after the fact)
2590    - Fixed messed-up controls priorities.
2591    
2592    * Removed again all the stuff from config/preloads that one wouldn't
2593      be able to preload at the time the initial heap image is built.
2594      (Many libraries, e.g., CML, do not exist yet at this time.  The
2595       only libraries that can be preloaded via config/preloads are those
2596       that come bundled with the bootfiles.)
2597    
2598    ----------------------------------------------------------------------
2599    Name: Matthias Blume
2600    Date: 2002/05/20 10:59:00 EDT
2601    Tag: blume-20020520-preloads
2602    Description:
2603    
2604    Added a lot of commented-out suggestions for things to be included
2605    in config/preloads.
2606    
2607    ----------------------------------------------------------------------
2608    Name: Allen Leung
2609    Date: 2002/05/18 14:20:00 EDT
2610    Tag: leunga-20020518-mdl
2611    Description:
2612    
2613    o Made the mdl tool stuff compile and run again.
2614    o I've disabled all the stuff that depends on RTL specifications; they
2615      are all badly broken anyway.
2616    
2617    ----------------------------------------------------------------------
2618    Name: Matthias Blume
2619    Date: 2002/05/17 16:49:00 EDT
2620    Tag: blume-20020517-controls
2621    Description:
2622    
2623    0. John Reppy made several modifications to the SML/NJ library.
2624       In particular, there is a shiny new controls-lib.cm.
2625    
2626    1. Pushed new controls interface through compiler so that everything
2627       compiles again.
2628    
2629    2. Added FormatComb and FORMAT_COMB to the CML version of the
2630       SML/NJ library (so that CML compiles again).
2631    
2632    3. Modified init scripts because XXX_DEFAULT environment variables
2633       are no longer with us.  (Boot-time initialization is now done
2634       using the same environment variables that are also used for
2635       startup-time initialization of controls.)
2636    
2637    ----------------------------------------------------------------------
2638    Name: Lal George
2639    Date: 2002/05/15 09:20:10 EDT
2640    Tag: george-20020515-pseudo-op-decls
2641    Description:
2642    
2643            All pseudo-ops emitted before the first segment declaration
2644    such as TEXT, DATA, and BSS directives are assumed to be global
2645    declarations and are emitted first in the assembly file. This is
2646    useful in a number of situations where one has pseudo-ops that are not
2647    specific to any segment, and also works around the constraint that one
2648    cannot have client pseudo-ops in the TEXT segment.
2649    
2650            Because no segment is associated with these declarations it is
2651    an error to allocate any space or objects before the first segment
2652    directive and an exception will be raised. However, we cannot make
2653    this check for client pseudo-ops.
2654    
2655            These top level declarations are a field in the CFG graph_info.
2656    In theory you can continue to add to this field after the CFG has been
2657    built -- provided you know what you are doing;-)
2658    
2659    ----------------------------------------------------------------------
2660    Name: Matthias Blume
2661    Date: 2002/05/13 16:40:00 EDT
2662    Tag: blume-20020513-pp-etc
2663    Description:
2664    
2665    A few minor bugfixes:
2666    
2667      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2668        (Bogus printouts for redefined bindings still occur.  Compiler
2669        bug should no longer occur now.  We need to redo the prettyprinter
2670        from scratch.)
2671    
2672      - CM pathname printer now also adds escape sequences for ( and )
2673    
2674      - commend and docu fixes for ml-nlffi
2675    
2676    ----------------------------------------------------------------------
2677    Name: Matthias Blume
2678    Date: 2002/05/10 16:40:00 EDT
2679    Tag: blume-20020510-erg-textio
2680    Description:
2681    
2682    Applied the following bugfix provided by Emden Gansner:
2683    
2684        Output is corrupted when outputSubstr is used rather than output.
2685    
2686        The problem occurs when a substring
2687    
2688            ss = (s, dataStart, dataLen)
2689    
2690        where dataStart > 0, fills a stream buffer with avail bytes left.
2691        avail bytes of s, starting at index dataStart, are copied into the
2692        buffer, the buffer is flushed, and then the remaining dataLen-avail
2693        bytes of ss are copied into the beginning of the buffer. Instead of
2694        starting this copy at index dataStart+avail in s, the current code
2695        starts the copy at index avail.
2696    
2697        Fix:
2698        In text-io-fn.sml, change line 695 from
2699             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2700        to
2701             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2702    
2703    ----------------------------------------------------------------------
2704    Name: Matthias Blume
2705    Date: 2002/04/12 13:55:00 EDT
2706    Tag: blume-20020412-assyntax
2707    Description:
2708    
2709    1. Grabbed newer assyntax.h from the XFree86 project.
2710    2. Fiddled with how to compile X86.prim.asm without warnings.
2711    3. (Very) Minor cleanup in CM.
2712    
2713    ----------------------------------------------------------------------
2714    Name: Matthias Blume
2715    Date: 2002/04/01 (no joke!) 17:07:00 EST
2716    Tag: blume-20020401-x86div
2717    Description:
2718    
2719    Added full support for div/mod/rem/quot on the x86, using the machine
2720    instruction's two results (without clumsily recomputing the remainder)
2721    directly where appropriate.
2722    
2723    Some more extensive power-of-two support was added to the x86 instruction
2724    selector (avoiding expensive divs, mods, and muls where they can be
2725    replaced with cheaper shifts and masks).  However, this sort of thing
2726    ought to be done earlier, e.g., within the CPS optimizer so that
2727    all architectures benefit from it.
2728    
2729    The compiler compiles to a fixed point, but changes might be somewhat
2730    fragile nevertheless.  Please, report any strange things that you might
2731    see wrt. div/mod/quot/rem...
2732    
2733    ----------------------------------------------------------------------
2734    Name: Matthias Blume
2735    Date: 2002/03/29 17:22:00
2736    Tag: blume-20020329-div
2737    Description:
2738    
2739    Fixed my broken div/mod logic.  Unfortunately, this means that the
2740    inline code for div/mod now has one more comparison than before.
2741    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2742    The problem was with quotient = 0, because that alone does not tell
2743    us which way the rounding went.  One then has to look at whether
2744    remainder and divisor have the same sign...  :(
2745    
2746    Anyway, I replaced the bootfiles with fresh ones...
2747    
2748    ----------------------------------------------------------------------
2749    Name: Matthias Blume
2750    Date: 2002/03/29 14:10:00 EST
2751    Tag: blume-20020329-inlprims
2752    Description:
2753    
2754    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2755    
2756    Primops have changed. This means that the bin/boot-file formats have
2757    changed as well.
2758    
2759    To make sure that there is no confusion, I made a new version.
2760    
2761    
2762    CHANGES:
2763    
2764    * removed REMT from mltree (remainder should never overflow).
2765    
2766    * added primops to deal with divisions of all flavors to the frontend
2767    
2768    * handled these primops all the way through so they map to their respective
2769      MLRISC support
2770    
2771    * used these primops in the implementation of Int, Int32, Word, Word32
2772    
2773    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2774    
2775    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2776    
2777    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2778      floating point
2779    
2780    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2781      and Real (but Real.abs maps to a separate floating-point-only primop)
2782    
2783    
2784    TODO items:
2785    
2786    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2787      instructions because the machine instruction encoder cannot handle
2788      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2789      REM in the first place, and REM is supported by the x86 machine coder.)
2790    
2791    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2792      selection phase.  (The two can be streamlined because the hardware
2793      delivers both quotient and remainder at the same time anyway.)
2794    
2795    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2796      (Currently the behavior is inconsistent both across architectures and
2797      wrt. the draft Basis spec.)
2798    
2799    * Word8 should eventually be handled natively, too.
2800    
2801    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2802      as if there currently is no execution path that could trigger it in
2803      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2804      hold for things like multiplication and division.)
2805    
2806    ----------------------------------------------------------------------
2807    Name: Matthias Blume
2808    Date: 2002/03/27 16:27:00 EST
2809    Tag: blume-20020327-mlrisc-divisions
2810    Description:
2811    
2812    Added support for all four division operations (ML's div, mod, quot,
2813    and rem) to MLRISC.  In the course of doing so, I also rationalized
2814    the naming (no more annoying switch-around of DIV and QUOT), by
2815    parameterizing the operation by div_rounding_mode (which can be either
2816    DIV_TO_ZERO or DIV_TO_NEGINF).
2817    
2818    The generic MLTreeGen functor takes care of compiling all four
2819    operations down to only round-to-zero div.
2820    
2821    Missing pieces:
2822    
2823      * Doing something smarter than relying on MLTreeGen on architectures
2824        like, e.g., the x86 where hardware division delivers both quotient and
2825        remainder at the same time.  With this, the implementation of the
2826        round-to-neginf operations could be further streamlined.
2827    
2828      * Remove inlining support for div/mod/rem from the frontend and replace it
2829        with primops that get carried through to the backend.  Do this for all
2830        int and word types.
2831    
2832    ----------------------------------------------------------------------
2833    Name: Matthias Blume
2834    Date: 2002/03/25 17:25:00 EST
2835    Tag: blume-20020325-divmod
2836    Description:
2837    
2838    I improved (hopefully without breaking them) the implementation of Int.div,
2839    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2840    advantage of the following observations:
2841    
2842      Let  q = x quot y      r = x rem y
2843           d = x div  y      m = x mod y
2844    
2845    where "quot" is the round-to-zero version of integer division that
2846    hardware usually provides.  Then we have:
2847    
2848         r = x - q * y        where neither the * nor the - will overflow
2849         d = if q >= 0 orelse x = q * y then q else q - 1
2850                              where neither the * nor the - will overflow
2851         m = if q >= 0 orelse r = 0 then r else r + y
2852                              where the + will not overflow
2853    
2854    This results in substantial simplification of the generated code.
2855    The following table shows the number of CFG nodes and edges generated
2856    for
2857            fun f (x, y) = x OPER y
2858            (* with OPER \in div, mod, quot, rem *)
2859    
2860    
2861        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2862        --------------------------------------------------------
2863         div |         24 |         39 |         12 |         16
2864         mod |         41 |         71 |         12 |         16
2865        quot |          8 |         10 |          8 |         10
2866         rem |         10 |         14 |          8 |         10
2867    
2868    
2869    ----------------------------------------------------------------------
2870    Name: Matthias Blume
2871    Date: 2002/03/25 22:06:00 EST
2872    Tag: blume-20020325-cprotobug
2873    Description:
2874    
2875    Fixed a bug in cproto (c prototype decoder).
2876    
2877    ----------------------------------------------------------------------
2878    Name: Matthias Blume
2879    Date: 2002/03/25 16:00:00 EST
2880    Tag: blume-20020325-raw-primops
2881    Description:
2882    
2883    I did some cleanup to Allen's new primop code and
2884    replaced yesterday's bootfiles with new ones.
2885    (But they are stored in the same place.)
2886    
2887    ----------------------------------------------------------------------
2888    Name: Matthias Blume
2889    Date: 2002/03/24 22:40:00 EST
2890    Tag: blume-20020324-bootfiles
2891    Description:
2892    
2893    Made the bootfiles that Allen asked for.
2894    
2895    ----------------------------------------------------------------------
2896    Name: Allen Leung
2897    Date: 2002/03/23 15:50:00 EST
2898    Tag: leunga-20020323-flint-cps-rcc-primops
2899    Description:
2900    
2901      1. Changes to FLINT primops:
2902    
2903        (* make a call to a C-function;
2904         * The primop carries C function prototype information and specifies
2905         * which of its (ML-) arguments are floating point. C prototype
2906         * information is for use by the backend, ML information is for
2907         * use by the CPS converter. *)
2908      | RAW_CCALL of { c_proto: CTypes.c_proto,
2909                       ml_args: ccall_type list,
2910                       ml_res_opt: ccall_type option,
2911                       reentrant : bool
2912                     } option
2913       (* Allocate uninitialized storage on the heap.
2914        * The record is meant to hold short-lived C objects, i.e., they
2915        * are not ML pointers.  With the tag, the representation is
2916        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2917        *)
2918      | RAW_RECORD of {tag:bool,sz:int}
2919      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2920    
2921      2.  These CPS primops are now overloaded:
2922    
2923           rawload of {kind:numkind}
2924           rawstore of {kind:numkind}
2925    
2926          The one argument form is:
2927    
2928             rawload {kind} address
2929    
2930          The two argument form is:
2931    
2932             rawload {kind} [ml object, byte-offset]
2933    
2934      3. RAW_CCALL/RCC now takes two extra arguments:
2935    
2936         a. The first is whether the C call is reentrant, i.e., whether
2937            ML state should be saved and restored.
2938         b. The second argument is a string argument specifying the name of
2939            library and the C function.
2940    
2941         These things are currently not handled in the code generator, yet.
2942    
2943      4. In CProto,
2944    
2945         An encoding type of "bool" means "ml object" and is mapped into
2946         C prototype of PTR.  Note that "bool" is different than "string",
2947         even though "string" is also mapped into PTR, because "bool"
2948         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2949    
2950      5. Pickler/unpicker
2951    
2952         Changed to handle RAW_RECORD and newest RAW_CCALL
2953    
2954      6. MLRiscGen,
2955    
2956         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2957         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2958            in the file CodeGen/cpscompile/cps-c-calls.sml
2959    
2960      7. Added the conditional move operator
2961    
2962             condmove of branch
2963    
2964         to cps.  Generation of this is still buggy so it is currently
2965         disabled.
2966    
2967    ----------------------------------------------------------------------
2968    Name: Lal George
2969    Date: 2002/03/22 14:18:25 EST
2970    Tag: george-20020322-cps-branch-prob
2971    Description:
2972    
2973    Implemented the Ball-Larus branch prediction-heuristics, and
2974    incorporated graphical viewers for control flow graphs.
2975    
2976    Ball-Larus Heuristics:
2977    ---------------------
2978    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2979    
2980    By design it uses the Dempster-Shafer theory for combining
2981    probabilities.  For example, in the function:
2982    
2983        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2984    
2985    the ball-larus heuristics predicts that the n=0 is unlikely
2986    (OH-heuristic), and the 'then' branch is unlikely because of the
2987    RH-heuristic -- giving the 'then' branch an even lower combined
2988    probability using the Dempster-Shafer theory.
2989    
2990    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2991    probability of the 'then' branch because of the loop in the else
2992    branch.
2993    
2994    
2995    Graphical Viewing:
2996    ------------------
2997    I merely plugged in Allen's graphical viewers into the compiler. The
2998    additional code is not much. At the top level, saying:
2999    
3000            Control.MLRISC.getFlag "cfg-graphical-view" := true;
3001    
3002    will display the graphical view of the control flow graph just before
3003    back-patching.  daVinci must be in your path for this to work. If
3004    daVinci is not available, then the default viewer can be changed
3005    using:
3006    
3007            Control.MLRISC.getString "viewer"
3008    
3009    which can be set to "dot" or "vcg" for the corresponding viewers. Of
3010    course, these viewers must be in your path.
3011    
3012    The above will display the compilation unit at the level of clusters,
3013    many of which are small, boring, and un-interesting. Also setting:
3014    
3015            Control.MLRISC.getInt "cfg-graphical-view_size"
3016    
3017    will display clusters that are larger than the value set by the above.
3018    
3019    
3020    ----------------------------------------------------------------------
3021    Name: Matthias Blume
3022    Date: 2002/03/21 22:20:00 EST
3023    Tag: blume-20020321-kmp-bugfix
3024    Description:
3025    
3026    Changed the interface to the KMP routine in PreString and fixed
3027    a minor bug in one place where it was used.
3028    
3029    ----------------------------------------------------------------------
3030    Name: Allen Leung
3031    Date: 2002/03/21 20:30:00 EST
3032    Tag: leunga-20020321-cfg
3033    Description:
3034    
3035      Fixed a potential problem in cfg edge splitting.
3036    
3037    ----------------------------------------------------------------------
3038    Name: Allen Leung
3039    Date: 2002/03/21 17:15:00 EST
3040    Tag: leunga-20020321-x86-fp-cfg
3041    Description:
3042    
3043      1. Recoded the buggy parts of x86-fp.
3044    
3045         a. All the block reordering code has been removed.
3046            We now depend on the block placement phases to do this work.
3047    
3048         b. Critical edge splitting code has been simplified and moved into the
3049            CFG modules, as where they belong.
3050    
3051         Both of these were quite buggy and complex.  The code is now much, much
3052         simpler.
3053    
3054      2. X86 backend.
3055    
3056         a. Added instructions for 64-bit support.  Instruction selection for
3057            64-bit has not been committed, however, since that
3058            requires changes to MLTREE which haven't been approved by
3059            Lal and John.
3060    
3061         b. Added support for FUCOMI and FUCOMIP when generating code for
3062            PentiumPro and above.  We only generate these instructions in
3063            the fast-fp mode.
3064    
3065         c. Added cases for JP and JNP in X86FreqProps.
3066    
3067      3. CFG
3068    
3069         CFG now has a bunch of methods for edge splitting and merging.
3070    
3071      4. Machine description.
3072    
3073         John's simplification of MLTREE_BASIS.fcond broke a few machine
3074         description things:
3075    
3076         rtl-build.{sig,sml} and hppa.mdl fixed.
3077    
3078         NOTE: the machine description stuff in the repository is still broken.
3079               Again, I can't put my fixes in because that involves
3080               changes to MLTREE.
3081    
3082    ----------------------------------------------------------------------
3083    Name: Matthias Blume
3084    Date: 2002/03/20 15:55:00 EST
3085    Tag: blume-20020320-kmp
3086    Description:
3087    
3088    Implemented Knuth-Morris-Pratt string matching in PreString and used
3089    it for String.isSubstring, Substring.isSubstring, and
3090    Substring.position.
3091    
3092    (Might need some stress-testing.  Simple examples worked fine.)
3093    
3094    ----------------------------------------------------------------------
3095    Name: Matthias Blume
3096    Date: 2002/03/19 16:37:00 EST
3097    Tag: blume-20020319-witnesses
3098    Description:
3099    
3100    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
3101    
3102    This implements a generic mechanism for changing constness qualifiers
3103    anywhere within big C types without resorting to outright "casts".
3104    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
3105    modify the constness at the outermost level.)
3106    The implementation of "convert" is based on the idea of "witness"
3107    values -- values that are not used by the operation but whose types
3108    "testify" to their applicability.  On the implementation side, "convert"
3109    is simply a projection (returning its second curried argument).  With
3110    cross-module inlining, it should not result in any machine code being
3111    generated.
3112    
3113    ----------------------------------------------------------------------
3114    Name: Matthias Blume
3115    Date: 2002/03/15 16:40:00 EST
3116    Tag: blume-20020315-basis
3117    Description:
3118    
3119    Provided (preliminary?) implementations for
3120    
3121      {String,Substring}.{concatWith,isSuffix,isSubstring}
3122    
3123    and
3124    
3125      Substring.full
3126    
3127    Those are in the Basis spec but they were missing in SML/NJ.
3128    
3129    ----------------------------------------------------------------------
3130    Name: Matthias Blume
3131    Date: 2002/03/14 21:30:00 EST
3132    Tag: blume-20020314-controls
3133    Description:
3134    
3135    Controls:
3136    ---------
3137    
3138    1. Factored out the recently-added Controls : CONTROLS stuff and put
3139       it into its own library $/controls-lib.cm.  The source tree for
3140       this is under src/smlnj-lib/Controls.
3141    
3142    2. Changed the names of types and functions in this interface, so they
3143       make a bit more "sense":
3144    
3145          module -> registry
3146          'a registry -> 'a group
3147    
3148    3. The interface now deals in ref cells only.  The getter/setter interface
3149       is (mostly) gone.
3150    
3151    4. Added a function that lets one register an already-existing ref cell.
3152    
3153    5. Made the corresponding modifications to the rest of the code so that
3154       everything compiles again.
3155    
3156    6. Changed the implementation of Controls.MLRISC back to something closer
3157       to the original.  In particular, this module (and therefore MLRISC)
3158       does not depend on Controls.  There now is some link-time code in
3159       int-sys.sml that registers the MLRISC controls with the Controls
3160       module.
3161    
3162    CM:
3163    ---
3164    
3165      * One can now specify the lambda-split aggressiveness in init.cmi.
3166    
3167    ----------------------------------------------------------------------
3168    Name: Allen Leung
3169    Date: 2002/03/13 17:30:00 EST
3170    Tag: leunga-20020313-x86-fp-unary
3171    Description:
3172    
3173    Bug fix for:
3174    
3175    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3176    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3177    > - fun f(x,(y,z)) = Real.~ y;
3178    > [autoloading]
3179    > [autoloading done]
3180    >       fchsl   (%eax), 184(%esp)
3181    > Error: MLRisc bug: X86MCEmitter.emitInstr
3182    >
3183    > uncaught exception Error
3184    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3185    
3186    The problem was that the code generator did not generate any fp registers
3187    in this case, and the ra didn't know that it needed to run the X86FP phase to
3188    translate the pseudo fp instruction.   This only happened with unary fp
3189    operators in certain situations.
3190    
3191    ----------------------------------------------------------------------
3192    Name: Matthias Blume
3193    Date: 2002/03/13 14:00:00 EST
3194    Tag: blume-20020313-overload-etc
3195    Description:
3196    
3197    1. Added _overload as a synonym for overload for backward compatibility.
3198       (Control.overloadKW must be true for either version to be accepted.)
3199    
3200    2. Fixed bug in install script that caused more things to be installed
3201       than what was requested in config/targets.
3202    
3203    3. Made CM aware of the (_)overload construct so that autoloading
3204       works.
3205    
3206    ----------------------------------------------------------------------
3207    Name: Matthias Blume
3208    Date: 2002/03/12 22:03:00 EST
3209    Tag: blume-20020312-url
3210    Description:
3211    
3212    Forgot to update BOOT and srcarchiveurl.
3213    
3214    ----------------------------------------------------------------------
3215    Name: Matthias Blume
3216    Date: 2002/03/12 17:30:00 EST
3217    Tag: blume-20020312-version110392
3218    Description:
3219    
3220    Yet another version number bump (because of small changes to the
3221    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3222    
3223    Changes:
3224    
3225      The new pid generation scheme described a few weeks ago was overly
3226      complicated.  I implemented a new mechanism that is simpler and
3227      provides a bit more "stability":  Once CM has seen a compilation
3228      unit, it keeps its identity constant (as long as you do not delete
3229      those crucial CM/GUID/* files).  This means that when you change
3230      an interface, compile, then go back to the old interface, and
3231      compile again, you arrive at the original pid.
3232    
3233      There now also is a mechanism that instructs CM to use the plain
3234      environment hash as a module's pid (effectively making its GUID
3235      the empty string).  For this, "noguid" must be specified as an
3236      option to the .sml file in question within its .cm file.
3237      This is most useful for code that is being generated by tools such
3238      as ml-nlffigen (because during development programmers tend to
3239      erase the tool's entire output directory tree including CM's cached
3240      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3241      revert to the old, broken behavior of SML/NJ, but in specific cases
3242      where there is no danger of interface confusion, its use is ok
3243      (I think).
3244    
3245      ml-nlffigen by default generates "noguid" annotations.  They can be
3246      turned off by specifying -guid in its command line.
3247    
3248    ----------------------------------------------------------------------
3249    Name: Lal George
3250    Date: 2002/03/12 12 14:42:36 EST
3251    Tag: george-20020312-frequency-computation
3252    Description:
3253    
3254    Integrated jump chaining and static block frequency into the
3255    compiler. More details and numbers later.
3256    
3257    ----------------------------------------------------------------------
3258    Name: Lal George
3259    Date: 2002/03/11 11 22:38:53 EST
3260    Tag: george-20020311-jump-chain-elim
3261    Description:
3262    
3263    Tested the jump chain elimination on all architectures (except the
3264    hppa).  This is on by default right now and is profitable for the
3265    alpha and x86, however, it may not be profitable for the sparc and ppc
3266    when compiling the compiler.
3267    
3268    The gc test will typically jump to a label at the end of the cluster,
3269    where there is another jump to an external cluster containing the actual
3270    code to invoke gc. This is to allow factoring of common gc invocation
3271    sequences. That is to say, we generate:
3272    
3273            f:
3274               testgc
3275               ja   L1      % jump if above to L1
3276    
3277            L1:
3278               jmp L2
3279    
3280    
3281    After jump chain elimination the 'ja L1' instructions is converted to
3282    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3283    up being implemented in their long form (if L2 is far away) using:
3284    
3285            jbe     L3      % jump if below or equal to L3
3286            jmp     L2
3287         L3:
3288            ...
3289    
3290    
3291    For large compilation units L2  may be far away.
3292    
3293    
3294    ----------------------------------------------------------------------
3295    Name: Matthias Blume
3296    Date: 2002/03/11 13:30:00 EST
3297    Tag: blume-20020311-mltreeeval
3298    Description:
3299    
3300    A functor parameter was missing.
3301    
3302    ----------------------------------------------------------------------
3303    Name: Allen Leung
3304    Date: 2002/03/11 10:30:00 EST
3305    Tag: leunga-20020311-runtime-string0
3306    Description:
3307    
3308       The representation of the empty string now points to a
3309    legal null terminated C string instead of unit.  It is now possible
3310    to convert an ML string into C string with InlineT.CharVector.getData.
3311    This compiles into one single machine instruction.
3312    
3313    ----------------------------------------------------------------------
3314    Name: Allen Leung
3315    Date: 2002/03/10 23:55:00 EST
3316    Tag: leunga-20020310-x86-call
3317    Description:
3318    
3319       Added machine generation for CALL instruction (relative displacement mode)
3320    
3321    ----------------------------------------------------------------------
3322    Name: Matthias Blume
3323    Date: 2002/03/08 16:05:00
3324    Tag: blume-20020308-entrypoints
3325    Description:
3326    
3327    Version number bumped to 110.39.1.  NEW BOOTFILES!
3328    
3329    Entrypoints: non-zero offset into a code object where execution should begin.
3330    
3331    - Added the notion of an entrypoint to CodeObj.
3332    - Added reading/writing of entrypoint info to Binfile.
3333    - Made runtime system bootloader aware of entrypoints.
3334    - Use the address of the label of the first function given to mlriscGen
3335      as the entrypoint.  This address is currently always 0, but it will
3336      not be 0 once we turn on block placement.
3337    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3338      with entry points) from mlriscGen.
3339    
3340    ----------------------------------------------------------------------
3341    Name: Allen Leung
3342    Date: 2002/03/07 20:45:00 EST
3343    Tag: leunga-20020307-x86-cmov
3344    Description:
3345    
3346       Bug fixes for CMOVcc on x86.
3347    
3348       1. Added machine code generation for CMOVcc
3349       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3350       3. CMOVcc cannot have an immediate operand as argument.
3351    
3352    ----------------------------------------------------------------------
3353    Name: Matthias Blume
3354    Date: 2002/03/07 16:15:00 EST
3355    Tag: blume-20020307-controls
3356    Description:
3357    
3358    This is a very large but mostly boring patch which makes (almost)
3359    every tuneable compiler knob (i.e., pretty much everything under
3360    Control.* plus a few other things) configurable via both the command
3361    line and environment variables in the style CM did its configuration
3362    until now.
3363    
3364    Try starting sml with '-h' (or, if you are brave, '-H')
3365    
3366    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3367    implements the underlying generic mechanism.
3368    
3369    The interface to some of the existing such facilities has changed somewhat.
3370    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3371    (The getFoo interface is still there for backward-compatibility, but its
3372    use is deprecated.)
3373    
3374    The ml-build script passes -Cxxx=yyy command-line arguments through so
3375    that one can now twiddle the compiler settings when using this "batch"
3376    compiler.
3377    
3378    TODO items:
3379    
3380    We should go through and throw out all controls that are no longer
3381    connected to anything.  Moreover, we should go through and provide
3382    meaningful (and correct!) documentation strings for those controls
3383    that still are connected.
3384    
3385    Currently, multiple calls to Controls.new are accepted (only the first
3386    has any effect).  Eventually we should make sure that every control
3387    is being made (via Controls.new) exactly once.  Future access can then
3388    be done using Controls.acc.
3389    
3390    Finally, it would probably be a good idea to use the getter-setter
3391    interface to controls rather than ref cells.  For the time being, both
3392    styles are provided by the Controls module, but getter-setter pairs are
3393    better if thread-safety is of any concern because they can be wrapped.
3394    
3395    *****************************************
3396    
3397    One bug fix: The function blockPlacement in three of the MLRISC
3398    backpatch files used to be hard-wired to one of two possibilities at
3399    link time (according to the value of the placementFlag).  But (I
3400    think) it should rather sense the flag every time.
3401    
3402    *****************************************
3403    
3404    Other assorted changes (by other people who did not supply a HISTORY entry):
3405    
3406    1. the cross-module inliner now works much better (Monnier)
3407    2. representation of weights, frequencies, and probabilities in MLRISC
3408       changed in preparation of using those for weighted block placement
3409       (Reppy, George)
3410    
3411    ----------------------------------------------------------------------
3412    Name: Lal George
3413    Date: 2002/03/07 14:44:24 EST 2002
3414    Tag: george-20020307-weighted-block-placement
3415    
3416    Tested the weighted block placement optimization on all architectures
3417    (except the hppa) using AMPL to generate the block and edge frequencies.
3418    Changes were required in the machine properties to correctly
3419    categorize trap instructions. There is an MLRISC flag
3420    "weighted-block-placement" that can be used to enable weighted block
3421    placement, but this will be ineffective without block/edge
3422    frequencies (coming soon).
3423    
3424    
3425    ----------------------------------------------------------------------
3426    Name: Lal George
3427    Date: 2002/03/05 17:24:48 EST
3428    Tag: george-20020305-linkage-cluster
3429    
3430    In order to support the block placement optimization, a new cluster
3431    is generated as the very first cluster (called the linkage cluster).
3432    It contains a single jump to the 'real' entry point for the compilation
3433    unit. Block placement has no effect on the linkage cluster itself, but
3434    all the other clusters  have full freedom in the manner in which they
3435    reorder blocks or functions.
3436    
3437    On the x86 the typical linkage code that is generated is:
3438       ----------------------
3439            .align 2
3440       L0:
3441            addl    $L1-L0, 72(%esp)
3442            jmp     L1
3443    
3444    
3445            .align  2
3446       L1:
3447       ----------------------
3448    
3449    72(%esp) is the memory location for the stdlink register. This
3450    must contain the address of the CPS function being called. In the
3451    above example, it contains the address of  L0; before
3452    calling L1 (the real entry point for the compilation unit), it
3453    must contain the address for L1, and hence
3454    
3455            addl $L1-L0, 72(%esp)
3456    
3457    I have tested this on all architectures except the hppa.The increase
3458    in code size is of course negligible
3459    
3460    ----------------------------------------------------------------------
3461    Name: Allen Leung
3462    Date: 2002/03/03 13:20:00 EST
3463    Tag: leunga-20020303-mlrisc-tools
3464    
3465      Added #[ ... ] expressions to mlrisc tools
3466    
3467    ----------------------------------------------------------------------
3468    Name: Matthias Blume
3469    Date: 2002/02/27 12:29:00 EST
3470    Tag: blume-20020227-cdebug
3471    Description:
3472    
3473    - made types in structure C and C_Debug to be equal
3474    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3475    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3476    
3477    ----------------------------------------------------------------------
3478    Name: Matthias Blume
3479    Date: 2002/02/26 12:00:00 EST
3480    Tag: blume-20020226-ffi
3481    Description:
3482    
3483    1. Fixed a minor bug in CM's "noweb" tool:
3484       If numbering is turned off, then truly don't number (i.e., do not
3485       supply the -L option to noweb).  The previous behavior was to supply
3486       -L'' -- which caused noweb to use the "default" line numbering scheme.
3487       Thanks to Chris Richards for pointing this out (and supplying the fix).
3488    
3489    2. Once again, I reworked some aspects of the FFI:
3490    
3491       A. The incomplete/complete type business:
3492    
3493       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3494         gone!
3495       - ML types representing an incomplete type are now *equal* to
3496         ML types representing their corresponding complete types (just like
3497         in C).  This is still safe because ml-nlffigen will not generate
3498         RTTI for incomplete types, nor will it generate functions that
3499         require access to such RTTI.   But when ML code generated from both
3500         incomplete and complete versions of the C type meet, the ML types
3501         are trivially interoperable.
3502    
3503         NOTE:  These changes restore the full generality of the translation
3504         (which was previously lost when I eliminated functorization)!
3505    
3506       B. Enum types:
3507    
3508       - Structure C now has a type constructor "enum" that is similar to
3509         how the "su" constructor works.  However, "enum" is not a phantom
3510         type because each "T enum" has values (and is isomorphic to
3511         MLRep.Signed.int).
3512       - There are generic access operations for enum objects (using
3513         MLRep.Signed.int).
3514       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3515         * The structure contains the definition of type "mlrep" (the ML-side
3516         representation type of the enum).  Normally, mlrep is the same
3517         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3518         then mlrep will be defined as a datatype -- thus facilitating
3519         pattern matching on mlrep values.
3520         ("-ec" will be suppressed if there are duplicate values in an
3521          enumeration.)
3522         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3523         will be generated for each C enum constant xxx.
3524         * Conversion functions m2i and i2m convert between mlrep and
3525         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3526         * Coversion functions c and ml convert between mlrep and "tag enum".
3527         * Access functions (get/set) fetch and store mlrep values.
3528       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3529         enumerations are merged into one single enumeration represented by
3530         structure E_'.
3531    
3532    ----------------------------------------------------------------------
3533    Name: Allen Leung
3534    Date: 2002/02/25 04:45:00 EST
3535    Tag: leunga-20020225-cps-spill
3536    
3537    This is a new implementation of the CPS spill phase.
3538    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3539    In case of problems, replace it with the old file spill.sml
3540    
3541    The current compiler runs into some serious performance problems when
3542    constructing a large record.  This can happen when we try to compile a
3543    structure with many items.  Even a very simple structure like the following
3544    makes the compiler slow down.
3545    
3546        structure Foo = struct
3547           val x_1 = 0w1 : Word32.int
3548           val x_2 = 0w2 : Word32.int
3549           val x_3 = 0w3 : Word32.int
3550           ...
3551           val x_N = 0wN : Word32.int
3552        end
3553    
3554    The following table shows the compile time, from N=1000 to N=4000,
3555    with the old compiler:
3556    
3557    N
3558    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3559           MLRISC ra                               0.06u  0.00s  0.05g
3560              (spills = 0 reloads = 0)
3561           TOTAL                                   0.63u  0.07s  0.21g
3562    
3563    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3564           MLRISC ra                               5.68u  0.59s  3.93g
3565              (spills = 0 reloads = 0)
3566           TOTAL                                   14.71u  0.99s  4.81g
3567    
3568    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3569           MLRISC ra                               5.54u  0.65s  3.91g
3570              (spills = 543 reloads = 1082)
3571           TOTAL                                   65.40u  3.13s  6.00g
3572    
3573    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3574           MLRISC ra                               0.80u  0.10s  0.55g
3575              (spills = 42 reloads = 84)
3576           TOTAL                                   129.42u  5.10s  4.13g
3577    
3578    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3579           MLRISC ra                               2.69u  0.27s  1.38g
3580              (spills = 62 reloads = 124)
3581           TOTAL                                   682.48u  19.61s  13.99g
3582    
3583    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3584           MLRISC ra                               4.96u  0.27s  2.72g
3585              (spills = 85 reloads = 170)
3586           TOTAL                                   2375.26u  57.21s  48.00g
3587    
3588    As you can see the old cps spill module suffers from some serious
3589    performance problem.  But since I cannot decipher the old code fully,
3590    instead of patching the problems up, I'm reimplementing it
3591    with a different algorithm.  The new code is more modular,
3592    smaller when compiled, and substantially faster
3593    (O(n log n) time and O(n) space).  Timing of the new spill module:
3594    
3595    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3596          MLRISC ra                               0.25u  0.02s  0.15g
3597             (spills=1 reloads=3)
3598          TOTAL                                   7.74u  0.34s  1.62g
3599    
3600    Implementation details:
3601    
3602    As far as I can tell, the purpose of the CPS spill module is to make sure the
3603    number of live variables at any program point (the bandwidth)
3604    does not exceed a certain limit, which is determined by the
3605    size of the spill area.
3606    
3607    When the bandwidth is too large, we decrease the register pressure by
3608    packing live variables into spill records.  How we achieve this is
3609    completely different than what we did in the old code.
3610    
3611    First, there is something about the MLRiscGen code generator
3612    that we should be aware of:
3613    
3614    o MLRiscGen performs code motion!
3615    
3616       In particular, it will move floating point computations and
3617       address computations involving only the heap pointer to
3618       their use sites (if there is only a single use).
3619       What this means is that if we have a CPS record construction
3620       statement
3621    
3622           RECORD(k,vl,w,e)
3623    
3624       we should never count the new record address w as live if w
3625       has only one use (which is often the case).
3626    
3627       We should do something similar to floating point, but the transformation
3628       there is much more complex, so I won't deal with that.
3629    
3630    Secondly, there are now two new cps primops at our disposal:
3631    
3632     1. rawrecord of record_kind option
3633        This pure operator allocates some uninitialized storage from the heap.
3634        There are two forms:
3635    
3636         rawrecord NONE [INT n]  allocates a tagless record of length n
3637         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3638                                     and initializes the tag.
3639    
3640     2. rawupdate of cty
3641          rawupdate cty (v,i,x)
3642          Assigns to x to the ith component of record v.
3643          The storelist is not updated.
3644    
3645    We use these new primops for both spilling and increment record construction.
3646    
3647     1. Spilling.
3648    
3649        This is implemented with a linear scan algorithm (but generalized
3650        to trees).  The algorithm will create a single spill record at the
3651        beginning of the cps function and use rawupdate to spill to it,
3652        and SELECT or SELp to reload from it.  So both spills and reloads
3653        are fine-grain operations.  In contrast, in the old algorithm
3654        "spills" have to be bundled together in records.
3655    
3656        Ideally, we should sink the spill record construction to where
3657        it is needed.  We can even split the spill record into multiple ones
3658        at the places where they are needed.  But CPS is not a good
3659        representation for global code motion, so I'll keep it simple and
3660        am not attempting this.
3661    
3662     2. Incremental record construction (aka record splitting).
3663    
3664        Long records with many component values which are simulatenously live
3665        (recall that single use record addresses are not considered to
3666         be live) are constructed with rawrecord and rawupdate.
3667        We allocate space on the heap with rawrecord first, then gradually
3668        fill it in with rawupdate.  This is the technique suggested to me
3669        by Matthias.
3670    
3671        Some restrictions on when this is applicable:
3672        1. It is not a VECTOR record.  The code generator currently does not handle
3673           this case. VECTOR record uses double indirection like arrays.
3674        2. All the record component values are defined in the same "basic block"
3675           as the record constructor.  This is to prevent speculative
3676           record construction.
3677    
3678    ----------------------------------------------------------------------
3679    Name: Allen Leung
3680    Date: 2002/02/22 01:02:00 EST
3681    Tag: leunga-20020222-mlrisc-tools
3682    
3683    Minor bug fixes in the parser and rewriter
3684    
3685    ----------------------------------------------------------------------
3686    Name: Allen Leung
3687    Date: 2002/02/21 20:20:00 EST
3688    Tag: leunga-20020221-peephole
3689    
3690    Regenerated the peephole files.  Some contained typos in the specification
3691    and some didn't compile because of pretty printing bugs in the old version
3692    of 'nowhere'.
3693    
3694    ----------------------------------------------------------------------
3695    Name: Allen Leung
3696    Date: 2002/02/19 20:20:00 EST
3697    Tag: leunga-20020219-mlrisc-tools
3698    Description:
3699    
3700       Minor bug fixes to the mlrisc-tools library:
3701    
3702       1.  Fixed up parsing colon suffixed keywords
3703       2.  Added the ability to shut the error messages up
3704       3.  Reimplemented the pretty printer and fixed up/improved
3705           the pretty printing of handle and -> types.
3706       4.  Fixed up generation of literal symbols in the nowhere tool.
3707       5.  Added some SML keywords to to sml.sty
3708    
3709    ----------------------------------------------------------------------
3710    Name: Matthias Blume
3711    Date: 2002/02/19 16:20:00 EST
3712    Tag: blume-20020219-cmffi
3713    Description:
3714    
3715    A wild mix of changes, some minor, some major:
3716    
3717    * All C FFI-related libraries are now anchored under $c:
3718        $/c.cm      --> $c/c.cm
3719        $/c-int.cm  --> $c/internals/c-int.cm
3720        $/memory.cm --> $c/memory/memory.cm
3721    
3722    * "make" tool (in CM) now treats its argument pathname slightly
3723      differently:
3724        1. If the native expansion is an absolute name, then before invoking
3725           the "make" command on it, CM will apply OS.Path.mkRelative
3726           (with relativeTo = OS.FileSys.getDir()) to it.
3727        2. The argument will be passed through to subsequent phases of CM
3728           processing without "going native".  In particular, if the argument
3729           was an anchored path, then "make" will not lose track of that anchor.
3730    
3731    * Compiler backends now "know" their respective C calling conventions
3732      instead of having to be told about it by ml-nlffigen.  This relieves
3733      ml-nlffigen from one of its burdens.
3734    
3735    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3736    
3737    * Export C_DEBUG and C_Debug from $c/c.cm.
3738    
3739    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3740      subtyping relationship between incomplete pointers and their complete
3741      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3742      with the convention of instantiating 'o with ('t, 'c) obj whenever
3743      the pointer target type is complete.  In the incomplete case, 'o
3744      will be instantiated with some "'c iobj" -- a type obtained by
3745      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3746    
3747      Operations that work on both incomplete and complete pointer types are
3748      typed as taking an 'o ptr while operations that require the target to
3749      be known are typed as taking some ('t, 'c) obj ptr.
3750    
3751      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3752      where void is an eqtype without any values.  This makes it possible
3753      to work on voidptr values using functions meant to operate on light
3754      incomplete pointers.
3755    
3756    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3757      been vastly simplified.
3758    
3759    ----------------------------------------------------------------------
3760    Name: Matthias Blume
3761    Date: 2002/02/19 10:48:00 EST
3762    Tag: blume-20020219-pqfix
3763    Description:
3764    
3765    Applied Chris Okasaki's bug fix for priority queues.
3766    
3767    ----------------------------------------------------------------------
3768    Name: Matthias Blume
3769    Date: 2002/02/15 17:05:00
3770    Tag: Release_110_39
3771    Description:
3772    
3773    Last-minute retagging is becoming a tradition... :-(
3774    
3775    This is the working release 110.39.
3776    
3777    ----------------------------------------------------------------------
3778    Name: Matthias Blume
3779    Date: 2002/02/15 16:00:00 EST
3780    Tag: Release_110_39-orig
3781    Description:
3782    
3783    Working release 110.39.  New bootfiles.
3784    
3785    (Update: There was a small bug in the installer so it wouldn't work
3786    with all shells.  So I retagged. -Matthias)
3787    
3788    ----------------------------------------------------------------------
3789    Name: Matthias Blume
3790    Date: 2002/02/15 14:17:00 EST
3791    Tag: blume-20020215-showbindings
3792    Description:
3793    
3794    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3795    the latter can be useful for exploring what bindings are available at
3796    the interactive prompt.  (The first function returns only the list
3797    of symbols that are really bound, the second prints those but also the
3798    ones that CM's autoloading mechanism knows about.)
3799    
3800    ----------------------------------------------------------------------
3801    Name: Matthias Blume
3802    Date: 2002/02/15 12:08:00 EST
3803    Tag: blume-20020215-iptrs
3804    Description:
3805    
3806    Two improvements to ml-nlffigen:
3807    
3808      1. Write files only if they do not exist or if their current contents
3809         do not coincide with what's being written.  (That is, avoid messing
3810         with the time stamps unless absolutely necessary.)
3811    
3812      2. Implement a "repository" mechanism for generated files related
3813         to "incomplete pointer types".   See the README file for details.
3814    
3815    ----------------------------------------------------------------------
3816    Name: Matthias Blume
3817    Date: 2002/02/14 11:50:00 EST
3818    Tag: blume-20020214-quote
3819    Description:
3820    
3821    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3822    because of the new and improved tag generation scheme.  (Thanks to Allen
3823    Leung for pointing it out.)
3824    
3825    ----------------------------------------------------------------------
3826    Name: Lal George
3827    Date: 2002/02/14 09:55:27 EST 2002
3828    Tag: george-20020214-isabelle-bug
3829    Description:
3830    
3831    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3832    of Isabelle on the x86.
3833    
3834    From Allen:
3835    -----------
3836     I've found the problem:
3837    
3838         in ra-core.sml, I use the counter "blocked" to keep track of the
3839         true number of elements in the freeze queue.  When the counter goes
3840         to zero, I skip examining the queue.  But I've messed up the
3841         bookkeeping in combine():
3842    
3843             else ();
3844             case !ucol of
3845               PSEUDO => (if !cntv > 0 then
3846                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3847                                        ^^^^^^^^^^^^^^^^^^^^^^^
3848                      moveu := mergeMoveList(!movev, !moveu)
3849                     )
3850                  else ();
3851    
3852         combine() is called to coalesce two nodes u and v.
3853         I think I was thinking that if the move counts of u and v are both
3854         greater than zero then after they are coalesced then one node is
3855         removed from the freeze queue.  Apparently I was thinking that
3856         both u and v are of low degree, but that's clearly not necessarily true.
3857    
3858    
3859    02/12/2002:
3860        Here's the patch.  HOL now compiles.
3861    
3862        I don't know how this impact on performance (compile
3863        time or runtime).  This bug caused the RA (especially on the x86)
3864        to go thru the potential spill phase when there are still nodes on the
3865        freeze queue.
3866    
3867    
3868    
3869    
3870    ----------------------------------------------------------------------
3871    Name: Matthias Blume
3872    Date: 2002/02/13 22:40:00 EST
3873    Tag: blume-20020213-fptr-rtti
3874    Description:
3875    
3876    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3877    updates.
3878    
3879    ----------------------------------------------------------------------
3880    Name: Matthias Blume
3881    Date: 2002/02/13 16:41:00 EST
3882    Tag: blume-20020213-cmlpq
3883    Description:
3884    
3885    Added new priority queue export symbols (which have just been added to
3886    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3887    not compile and the installer would choke.)
3888    
3889    ----------------------------------------------------------------------
3890    Name: Matthias Blume
3891    Date: 2002/02/13 16:15:00 EST
3892    Tag: blume-20020213-various
3893    Description:
3894    
3895    1. More tweaks to ml-nlffigen:
3896    
3897       - better internal datastructures (resulting in slight speedup)
3898       - "-match" option requires exact match
3899       - "localized" gensym counters (untagged structs/unions nested within
3900         other structs/unions or within typedefs get a fresh counter; their
3901         tag will be prefixed by a concatenation of their parents' tags)
3902       - bug fixes (related to calculation of transitive closure of types
3903         to be included in the output)
3904    
3905    2. Minor Basis updates:
3906    
3907       - added implementations for List.collate and Option.app
3908    
3909    ----------------------------------------------------------------------
3910    Name: Matthias Blume
3911    Date: 2002/02/11 15:55:00 EST
3912    Tag: blume-20020211-gensym
3913    Description:
3914    
3915    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3916    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3917    names (ML structure names that correspond to unnamed C structs, unions,
3918    and enums), so that separate runs of ml-nlffigen do not clash.
3919    
3920    ----------------------------------------------------------------------
3921    Name: Matthias Blume
3922    Date: 2002/02/11 12:05:00 EST
3923    Tag: blume-20020211-gensml
3924    Description:
3925    
3926    A quick fix for a problem with GenSML (in the pgraph-util library):
3927    Make generation of toplevel "local" optional.  (Strictly speaking,
3928    signature definitions within "local" are not legal SML.)
3929    
3930    Other than that: updates to INSTALL and cm/TODO.
3931    
3932    ----------------------------------------------------------------------
3933    Name: Matthias Blume
3934    Date: 2002/02/08 15:00:00 EST
3935    Tag: blume-20020208-uniquepid
3936    Description:
3937    
3938    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3939    
3940    1. The installer (config/install.sh) has gotten smarter:
3941    
3942         - Configuration options are a bit easier to specify now
3943           (in config/targets).
3944         - Bug in recognizing .tar.bz2 files fixed.
3945         - Installer automatically resolves dependencies between
3946           configuration options (e.g., if you ask for eXene, you will
3947           also get cml -- regardless whether you asked for it or not).
3948         - Installer can run in "quieter mode" by setting the environment
3949           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3950           "completely silent", though.
3951         - Build HashCons library as part of smlnj-lib.
3952    
3953    2. A new scheme for assigning persistent identifiers to compilation
3954       units (and, by extension, to types etc.) has been put into place.
3955       This fixes a long-standing bug where types and even dynamic values
3956       can get internally confused, thereby compromising type safety
3957       (abstraction) and dynamic correctness.  See
3958    
3959         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3960    
3961       for an example of how things could go wrong until now.
3962    
3963       The downside of the new scheme is that pids are not quite as
3964       persistent as they used to be: CM will generate a fresh pid
3965       for every compilation unit that it thinks it sees for the first
3966       time.  That means that if you compile starting from a clean, fresh
3967       source tree at two different times, you end up with different
3968       binaries.
3969    
3970       Cutoff recompilation, however, has not been compromised because
3971       CM keeps pid information in special caches between runs.
3972    
3973    ----------------------------------------------------------------------
3974    Name: Lal George
3975    Date: 2002/02/07 15:34:13 EST 2002
3976    Tag: <none>
3977    Description:
3978    
3979    Compilers that generate assembly code may produce  global labels
3980    whose value is resolved at link time. The various peephole optimization
3981    modules did not take this in account.
3982    
3983    TODO. The Labels.addrOf function should really return an option
3984    type so that clients are forced to deal with this issue, rather
3985    than an exception being raised.
3986    
3987    ----------------------------------------------------------------------
3988    Name: Lal George
3989    Date: 2002/02/06 13:55:02 EST
3990    Tag: george-20020206-ra-breakup
3991    Description:
3992    
3993    1. A bug fix from Allen.
3994    
3995        A typo causes extra fstp %st(0)'s to be generated at compensation
3996        edges, which might cause stack underflow traps at runtime.  This
3997        occurs in fft where there are extraneous fstps right before the 'into'
3998        trap instruction (in this case they are harmless since none of the
3999        integers overflow.)
4000    
4001    2. Pulled out various utility modules that were embedded in the modules
4002       of the register allocator. I need these modules for other purposes, but
4003       they are not complete enough to put into a library (just yet).
4004    ----------------------------------------------------------------------
4005    Name: Matthias Blume
4006    Date: 2002/01/31 16:05:00 EST
4007    Tag: blume-20020131-sparc-ccalls
4008    Description:
4009    
4010    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
4011       of extra stack space by mistake.  Fixed.
4012    
4013    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
4014    
4015    ----------------------------------------------------------------------
4016    Name: Allen Leung
4017    Date: 2002/01/30
4018    Tag: leunga-20020130-nowhere-bug-fix
4019    Description:
4020    
4021       MLRISC bug fixes:
4022       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
4023       2. MachineInt.fromString was negating its value.
4024    
4025    ----------------------------------------------------------------------
4026    Name: Matthias Blume
4027    Date: 2002/01/29
4028    Tag: blume-20020129-INSTALL
4029    Description:
4030    
4031    - Added somewhat detailed installation instructions (file INSTALL).
4032    - Fixed curl-detection bug in config/install.sh.
4033    - It is now possible to select the URL getter using the URLGETTER
4034      environment variable:
4035    
4036          not set / "unknown"      --> automatic detection (script tries wget,
4037                                       curl, and lynx)
4038          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
4039                                       how to properly invoke them)
4040          other                    --> use $URLGETTER directly, it must take
4041                                       precisely two command-line arguments
4042                                       (source URL and destination file name)
4043    
4044    ----------------------------------------------------------------------
4045    Name: Matthias Blume
4046    Date: 2002/01/28
4047    Tag: blume-20020128-sparc-ccalls
4048    Description:
4049    
4050    - Fixed problem with calculation of "used" registers in sparc-c-calls.
4051    - Make use of the allocParam argument in sparc-c-calls.
4052    
4053    ----------------------------------------------------------------------
4054    Name: Matthias Blume
4055    Date: 2002/01/28
4056    Tag: blume-20020128-allocParam
4057    Description:
4058    
4059    John Reppy:  Changes c-calls API to accept client-callback for
4060    allocating extra stack space.
4061    me: Corresponding changes to mlriscGen (using a dummy argument that
4062        does not change the current behavior).
4063    
4064    ----------------------------------------------------------------------
4065    Name: Matthias Blume
4066    Date: 2002/01/28 12:00:00
4067    Tag: Release_110_38
4068    Description:
4069    
4070    This time for real!!!
4071    
4072    ----------------------------------------------------------------------
4073    Name: Matthias Blume
4074    Date: 2002/01/28 10:56:00 EST
4075    Tag: blume-20020128-retraction
4076    Description:
4077    
4078    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
4079       with blume-Release_110_38-retracted.)
4080    
4081    1. Fixed a problem with incorrect rounding modes in real64.sml.
4082       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
4083    
4084    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
4085       and enums fixed.  The naming of corresponding ML identifiers should
4086       now be consistent again.
4087    
4088    ----------------------------------------------------------------------
4089    Name: Allen Leung
4090    Date: 2002/01/27
4091    Tag: leunga-20020127-nowhere
4092    Description:
4093    
4094       Added a target called nowhere in the configuration scripts.
4095       Enabling this will build the MLRISC 'nowhere' tool (for translating
4096       programs with where-clauses into legal SML code) during installation.
4097    
4098    --------------------------------------------------------