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 890, Thu Jul 19 20:38:56 2001 UTC revision 1886, Tue Feb 14 23:31:16 2006 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Matthias Blume (http://www.tti-c.org/blume/)
17    Date: 2006/02/14 17:30:00 CST
18    Tag: blume-20060214-intel-mac
19    Description:
20    
21    Hooked code for Darwin-specific Intel ABI into main compiler.
22    (This is a temporary solution which relies on the fact that the
23    compiler itself does not use NLFFI.  Eventually we need to divorce
24    intel mac from generic x86 unix code and make separate sets of binaries.)
25    
26    ----------------------------------------------------------------------
27    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
28    Date: 2006/02/13
29    Tag: jhr-20060213-intel-mac
30    Description:
31    Changed MLRISC x86 CCalls for partial support of MacOS X ABI.
32    
33    ----------------------------------------------------------------------
34    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
35    Date: 2006/02/06
36    Tag: jhr-20060206-intel-mac
37    Description:
38    Changes to support MacOS X on Intel hardware.  The C-calls support in
39    MLRISC must also be updated to support the MacOS X ABI.
40    
41    ----------------------------------------------------------------------
42    Name: Matthias Blume (http://www.tti-c.org/blume/)
43    Date: 2005/12/16 00:15:00 CST
44    Tag: blume-20051216-cm
45    Description:
46    
47    Improved error reporting and handling in CM.
48    
49    ----------------------------------------------------------------------
50    Name: Matthias Blume (http://www.tti-c.org/blume/)
51    Date: 2005/11/21 14:00:00 CST
52    Tag: blume-20051121-html-lib
53    Description:
54    
55    Enabling $/html-lib.cm in config/preloads did not work.  This is fixed
56    now.  (Since the anchor mapping for html-lib.cm is not yet in effect
57    at the time when config/preloads is processed, the library has to
58    be referred to by another name.  In 110.57 this would be
59    $SMLNJ-LIB/HTML/html-lib.cm.  I arranged for $smlnj/smlnj-lib/html-lib.cm
60    to be valid as well.)
61    
62    Thanks to Todd Wilson (Fresno) for alerting me to this issue.
63    
64    ----------------------------------------------------------------------
65    Name: Matthias Blume (http://www.tti-c.org/blume/)
66    Date: 2005/11/19 13:45:00 CST
67    Tag: Release_110_57
68    Description:
69    
70    Moved release tag for 110.57.
71    
72    ----------------------------------------------------------------------
73    Name: Matthias Blume (http://www.tti-c.org/blume/)
74    Date: 2005/11/19 23:35:00 CST
75    Tag: blume-20051119-fish
76    Description:
77    
78    Fixed a problem in config/install.sh where it tries to "fish" the
79    name of the CM metadata directory from the wrong place (because
80    the physical location of basis.cm has changed).  Also, corrected the
81    path anchor for $/html-lib.cm.  (Thanks to M. Fluet for pointing out
82    these problems.)
83    
84    ----------------------------------------------------------------------
85    Name: Matthias Blume (http://www.tti-c.org/blume/)
86    Date: 2005/11/16 14:30:00 CST
87    Tag: blume-orig-110_57
88    Description:
89    
90    New working version (110.57).  NEW BOOTFILES!
91    
92    UPDATE: The release tag has been moved to include a fix for a brown-paper-
93            bag glitch (see above).
94    
95    ----------------------------------------------------------------------
96    Name: Matthias Blume (http://www.tti-c.org/blume/)
97    Date: 2005/11/16 10:25:00 CST
98    Tag: blume-22051116-btrace
99    Description:
100    
101    Fixed problem with bogus exception message when using back-trace
102    facility.
103    
104    ----------------------------------------------------------------------
105    Name: Matthias Blume (http://www.tti-c.org/blume/)
106    Date: 2005/11/15 22:35:00 CST
107    Tag: blume-22051115-array2copy
108    Description:
109    
110    Added simple implementation of Array2.copy.  (Warning: mostly untested.)
111    
112    ----------------------------------------------------------------------
113    Name: Matthias Blume (http://www.tti-c.org/blume/)
114    Date: 2005/11/15 21:35:00 CST
115    Tag: blume-22051115-versiontool
116    Description:
117    
118    Reversed change to src/system/smlnj/internals/versiontool.cm.  This
119    file gets loaded as a tool -- by the equivalent of CM.make during the
120    run of CMB.make.  Thus, CMB's path configuration is meaningless for
121    it.  Instead, it has the status of "user code", so it should use
122    $/basis.cm to refer to the Basis library.  (At least that's true for
123    the purpose of bootstrapping the previous change.  In the future it
124    might make sense to have versiontool.cm refer to
125    $smlnj/basis/basis.cm, i.e., the version of the Basis that the
126    compiler itself uses.)
127    
128    Also patched src/system/testml to have it activate those extra anchor
129    bindings in config/extrapathconfig.
130    
131    ----------------------------------------------------------------------
132    Name: Matthias Blume (http://www.tti-c.org/blume/)
133    Date: 2005/11/15 16:35:00 CST
134    Tag: blume-22051115-libraries
135    Description:
136    
137    This change affects the way the following libraries are tied into
138    the system:
139    
140       $/basis.cm
141       $/smlnj-lib.cm
142       $/pp-lib.cm
143       $/controls-lib.cm
144       $/html-lib.cm
145       $/ml-yacc-lib.cm
146    
147    These libraries are now internally (as seen from the source code
148    of the implementation itself) known by the following names:
149    
150       $smlnj/basis/basis.cm
151       $smlnj/smlnj-lib/smlnj-lib.cm
152       $smlnj/smlnj-lib/pp-lib.cm
153       $smlnj/smlnj-lib/controls-lib.cm
154       $smlnj/smlnj-lib/html-lib.cm
155       $smlnj/ml-yacc/ml-yacc-lib.cm
156    
157    This makes it possible to work with code that requires different
158    versions of these libraries, and which refers to these libraries using
159    their "default" names (i.e., the first set of names above).  In other
160    words, one can un-define or re-define those default names without
161    compromising the proper functioning of the compiler itself.
162    
163    A similar procedure had already been performed for several of the
164    MLRISC libraries that are linked into the compiler.  I did some
165    cleanup on this code.
166    
167    A new file in the config directory (named extrapathconfig) is
168    responsible for setting up path anchors that the compiler itself does
169    not need, but that are typically required by user code.
170    
171    ----------------------------------------------------------------------
172    Name: Matthias Blume
173    Date: 2005/11/07 17:05:00 CST
174    Tag: blume-20051107-slice
175    Description:
176    
177    Fixed erroneous out-of-bounds test in the "update" function of
178    various *ArraySlice modules.  (Thanks to Vesa A. Norrman for pointing
179    out the problem.)
180    
181    Pushed some Basis changes through ML-Lex, CML, and eXene.
182    
183    ----------------------------------------------------------------------
184    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
185    Date: 2005/11/07
186    Description:
187    Fixed a Basis incompatibility: The depreciated function Substring.all
188    was removed (use Substring.full instead).
189    
190    ----------------------------------------------------------------------
191    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
192    Date: 2005/11/05
193    Description:
194    Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
195    
196    ----------------------------------------------------------------------
197    Name: Matthias Blume (http://www.tti-c.org/blume/)
198    Date: 2005/11/02 16:25:00 CST
199    Tag: blume-20051102-hex
200    Description:
201    
202    * Runtime system bootstrap code now accepts hex digits in BOOTLIST
203      in either upper case or lower case format.
204    * Pushed changes to names of Pack<N>{Big,Little} structures through
205      CML and eXene.
206    
207    ----------------------------------------------------------------------
208    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
209    Date: 2005/11/02
210    Description:
211    Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
212    be called PackWord<N>{Big,Little}.
213    
214    ----------------------------------------------------------------------
215    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
216    Date: 2005/10/28
217    Description:
218    Fixed a minor Basis incompatibility: hex digits should be upper case.
219    
220    ----------------------------------------------------------------------
221    Name: Matthias Blume (blume (at) tti - c (dot) org)
222    Date: 2005/10/25 16:00:00 CDT
223    Tag: Release_110_56
224    Description:
225    
226    New working version (110.56).  NEW BOOTFILES!
227    
228    ----------------------------------------------------------------------
229    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
230    Date: 2005/10/25
231    Description:
232    Added interval sets to utility library (signatures INTERVAL_DOMAIN
233    and INTERVAL_SET, and functor IntervalSetFn).
234    
235    ----------------------------------------------------------------------
236    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
237    Date: 2005/10/14
238    Description:
239    Add Zhong Shao's fix for datatype equality functions.
240    
241    ----------------------------------------------------------------------
242    Name: Allen Leung (leunga (at) reservoir (dot) com)
243    Date: 2005/10/14 12:30:00 EST
244    Tag: leunga-20051014-x86-spill
245    Description:
246    
247    Bug fix a bug found by Carl Hauser.
248    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
249    
250    ----------------------------------------------------------------------
251    Name: Allen Leung (leunga (at) reservoir (dot) com)
252    Date: 2005/10/14 11:15:00 EST
253    Tag: leunga-20051014-x86-asm
254    Description:
255    
256    Removed some debugging code in file x86Asm.sml.
257    The function emit_operand was printing out debugging output.
258    
259    ----------------------------------------------------------------------
260    Name: John Reppy
261    Date: 2005/07/27
262    Tag:
263    Description:
264    
265    Fixed ml-lex to recognize "\r" as representing carriage return.
266    
267    ----------------------------------------------------------------------
268    Name: John Reppy
269    Date: 2005/07/27
270    Tag:
271    Description:
272    
273    Fixed ml-yacc to work on files with non-native end-of-line
274    encodings (e.g., Windows text file on a Unix system).
275    
276    ----------------------------------------------------------------------
277    Name: John Reppy
278    Date: 2005/07/20
279    Tag:
280    Description:
281    
282    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
283    to support HPUX 11.
284    
285    ----------------------------------------------------------------------
286    Name: John Reppy
287    Date: 2005/07/06
288    Tag:
289    Description:
290    
291    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
292    
293    ----------------------------------------------------------------------
294    Name: Matthias Blume (blume (at) tti - c (dot) org)
295    Date: 2005/07/06 00:45:00 CDT
296    Tag: blume-20050706-slice-copy
297    Description:
298    
299    Fixed reversed logic for deciding whether to "copy up" or "copy down"
300    in *-array-slice.sml.
301    
302    ----------------------------------------------------------------------
303    Name: Allen Leung (leunga (at) reservoir (dot) com)
304    Date: 2005/05/31 17:00:00 EST
305    Tag: leunga-20050531-cygwin-fault-2
306    Description:
307    
308    A typo in the cygwin code fixed.
309    
310    ----------------------------------------------------------------------
311    Name: Allen Leung (leunga (at) reservoir (dot) com)
312    Date: 2005/05/31 16:47:00 EST
313    Tag: leunga-20050531-cygwin-fault
314    Description:
315    
316    Updated Cygwin's fault/signal handling to match the Windows version.
317    Updated the export list.
318    
319    ----------------------------------------------------------------------
320    Name: Matthias Blume (blume (at) tti - c (dot) org)
321    Date: 2005/05/18 13:35:00 CDT
322    Tag: Release_110_54
323    Description:
324    
325    New working version (110.54).  NEW BOOTFILES!
326    
327    ----------------------------------------------------------------------
328    Name: Matthias Blume (blume (at) tti - c (dot) org)
329    Date: 2005/05/18 11:58:00 CDT
330    Tag: blume-20050518-installer
331    Description:
332    
333    Added support scripts for Mac OS X PackageMaker and modified
334    config/install.sh so that it supports re-dumping a heap image
335    after customization.
336    
337    ----------------------------------------------------------------------
338    Name: Matthias Blume (blume (at) tti - c (dot) org)
339    Date: 2005/05/18 10:55:00 CDT
340    Tag: blume-20050518-realdiv-noovld
341    Description:
342    
343    Un-overloaded / to work around bug in overloading resolution code.
344    
345    ----------------------------------------------------------------------
346    Name: Matthias Blume (blume (at) tti - c (dot) org)
347    Date: 2005/05/16 23:50:00 CDT
348    Tag: blume-20050516-redump-heap
349    Description:
350    
351    Added mechanism for re-creating a heap file for the interactive system
352    after configuration variables have been changed.
353    
354       CM.redump_heap : string -> unit
355    
356    This is much like SMLofNJ.exportML, but starting from the resulting
357    heap does not return to the caller of CM.redump_heap but
358    restarts the interactive system from scratch.  The original call of
359    CM.redump_heap does not return but ends the interactive session.
360    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
361    
362    Internally, redump_heap winds the dynamic execution context back to
363    the point where the original heap image was created and re-executes
364    the heap image generation code in the boot code.
365    
366    ----------------------------------------------------------------------
367    Name: Matthias Blume (blume (at) tti - c (dot) org)
368    Date: 2005/05/09 21:55:00 CDT
369    Tag: blume-20050509-word64
370    Description:
371    
372    Added a hack to the existing hack known as Word64 to make fromString
373    behave correctly.  I am still not sure whether Word64.scan will work
374    as specified with respect to the interaction of radix and prefix.
375    
376    ----------------------------------------------------------------------
377    Name: Allen Leung (leunga (at) reservoir (dot) com)
378    Date: 2005/05/04 11:50:00 EST
379    Tag: leunga-20050504-checkgc
380    Description:
381    
382    Added a gc protocol checking phase.  This phase is enabled with
383    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
384    
385    ----------------------------------------------------------------------
386    Name: Matthias Blume (blume (at) tti - c (dot) org)
387    Date: 2005/05/04 10:45:00 CDT
388    Tag: blume-20050504-intinf
389    Description:
390    
391    Fixed a bug in the implementation of div and mod for IntInf.
392    Thanks to Neophytos Michael for reporting the problem.
393    
394    ----------------------------------------------------------------------
395    Name: Matthias Blume (blume (at) tti - c (dot) org)
396    Date: 2005/05/04 10:35:00 CDT
397    Tag: blume-20050504-join
398    Description:
399    
400    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
401    
402    ----------------------------------------------------------------------
403    Name: Matthias Blume (blume (at) tti - c (dot) org)
404    Date: 2005/02/28 23:40:00 CST
405    Tag: blume-20050228-mVar
406    Description:
407    
408    Fixed serious bug (brown paper bag variety) in new implementation of
409    structure Atom in CML.  (I had accidentally used a mailbox instead of
410    an mvar, leaving the door open for races.)
411    
412    ----------------------------------------------------------------------
413    Name: Matthias Blume (blume (at) tti - c (dot) org)
414    Date: 2005/02/25 15:00:00 CST
415    Tag: Release_110_53
416    Description:
417    
418    New working version (110.53).  NEW BOOTFILES!
419    
420    ----------------------------------------------------------------------
421    Name: Matthias Blume (blume (at) tti - c (dot) org)
422    Date: 2005/02/25 14:50:00 CST
423    Tag: blume-20050225-susp
424    Description:
425    
426    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
427    implemented in Core -- which means that it is the same as the one used
428    by the lazy extension.
429    
430    ----------------------------------------------------------------------
431    Name: Matthias Blume (blume (at) tti - c (dot) org)
432    Date: 2005/02/24 16:50:00 CST
433    Tag: blume-20050224-cml-atom
434    Description:
435    
436    Simpler and at the same time more general implementation of structure
437    Atom in CML.
438    
439    ----------------------------------------------------------------------
440    Name: Matthias Blume (blume (at) tti - c (dot) org)
441    Date: 2005/02/15 17:35:00 CST
442    Tag: blume-20050215-tools
443    Description:
444    
445    Created new "tools" directory under "src" and moved "TraceDebugProf"
446    there.
447    
448    ----------------------------------------------------------------------
449    Name: Matthias Blume (blume (at) tti - c (dot) org)
450    Date: 2005/02/10 17:55:00 CST
451    Tag: blume-20050210-longlong
452    Description:
453    
454    Implemented "long long" arguments and results for NLFFI.  (Only the
455    PPC/MacOS implementation is complete, the other backends still need to
456    be updated.)
457    
458    ----------------------------------------------------------------------
459    Name: Matthias Blume (blume (at) tti - c (dot) org)
460    Date: 2005/01/24 17:40:00 CST
461    Tag: blume-20050124-mlyacc
462    Description:
463    
464    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
465    problem with certain "as" patterns which previously got rendered
466    using incorrect syntax.
467    
468    ----------------------------------------------------------------------
469    Name: Matthias Blume (blume (at) tti - c (dot) org)
470    Date: 2005/01/18 12:00:00 CST
471    Tag: blume-20050118-profile
472    Description:
473    
474    Made time profiling code (interrupt handler) in runtime system aware
475    of new array representation.
476    
477    ----------------------------------------------------------------------
478    Name: Matthias Blume (blume (at) tti - c (dot) org)
479    Date: 2005/01/14 18:00:00 CST
480    Tag: blume-20050114-heap2exec
481    Description:
482    
483    Implemented new (but still experimental) heap2exec facility.  This is
484    tested under Mac OS X and should work under Linux (will test shortly).
485    It will probably also work on the Sparc (will test some time later).
486    
487      - removed old "HACKED_STANDALONE" hack from runtime
488    
489    To be able to test this, uncomment the request for "heap2asm" in
490    config/targets prior to installation.  (Notice that this is different
491    from "heap2exec" mentioned below.  Not a typo.)
492    
493    To perform an actual test, run the command
494    
495       $ bin/heap2exec heapfile execfile
496    
497    (You can put heap2exec on your shell's path.)
498    
499    For example, run
500    
501      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
502    
503    This will create a standalone executable called "mly" which you can
504    then invoke directly as a command.
505    
506    ----------------------------------------------------------------------
507    Name: Matthias Blume (blume (at) tti - c (dot) org)
508    Date: 2005/01/07 11:44:00 CST
509    Tag: blume-20050107-mlstring
510    Description:
511    
512    fixed off-by-one error in ML_STRING macro (globals.c)
513    
514    ----------------------------------------------------------------------
515    Name: Matthias Blume (blume (at) tti - c (dot) org)
516    Date: 2004/12/23 18:00:00 CST
517    Tag: blume-20041223-santa
518    Description:
519    
520    Made ml-build script "smarter" (but only very little).
521    
522    ----------------------------------------------------------------------
523    Name: Matthias Blume (blume (at) tti - c (dot) org)
524    Date: 2004/12/21 15:05:00 CST
525    Tag: blume-20041221-longlong
526    Description:
527    
528    * Implemented access to signed and unsigned long long data in NLFFI.
529    (The parameter-passing part of the picture has not complete. But
530    data structure access seems to work.)
531    
532    * Fixed CM's incorrect assumption that the PPC is little-endian.
533      (On the Mac, it is big-endian.  And that's currently our only
534      PPC platform.)
535    
536    ----------------------------------------------------------------------
537    Name: Matthias Blume (blume (at) tti - c (dot) org)
538    Date: 2004/12/21 12:50:00 CST
539    Tag: blume-20041221-memory
540    Description:
541    
542    Some cleanup in the $c/memory.cm library: separated some concerns by
543    moving allocation code and memory access code each into their own
544    files.
545    
546    ----------------------------------------------------------------------
547    Name: Allen Leung (leunga (at) reservoir (dot) com)
548    Date: 2004/12/17 16:12:00 EST
549    Tag: leunga-20041217-cygwin-smlnj-home
550    Description:
551    
552        The Unix I/O library of SML/NJ on cygwin does not understand
553    Windows style pathname, so problems arise when SMLNJ_HOME is set
554    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
555    to a POSIX pathname on cygwin.
556    
557    ----------------------------------------------------------------------
558    Name: Matthias Blume (blume (at) tti - c (dot) org)
559    Date: 2004/12/16 13:35:00 CST
560    Tag: Release_110_52
561    Description:
562    
563    Last-minute changes incorporated into 110.52.  Release tag moved.
564    
565    The changes:
566       - HashString.hashString' -> HashString.hashSubstring
567       - bug fix in UnivariateStats
568    
569    ----------------------------------------------------------------------
570    Name: Matthias Blume (blume (at) tti - c (dot) org)
571    Date: 2004/12/15 23:40:00 CST
572    Tag: blume-20041215-hashSubstring
573    Description:
574    
575    - HashString.hashString' -> HashString.hashSubstring
576    - corresponding changes in atom.sml
577    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
578      added some comments
579    
580    ----------------------------------------------------------------------
581    Name: Matthias Blume (blume (at) tti - c (dot) org)
582    Date: 2004/12/15 15:30:00 CST
583    Tag: (Release_110_52)
584    Description:
585    
586    New working version (110.52).  NEW BOOTFILES!
587    
588    ----------------------------------------------------------------------
589    Name: Matthias Blume (blume (at) tti - c (dot) org)
590    Date: 2004/12/15 12:45:00 CST
591    Tag: blume-20041215-spaces
592    Description:
593    
594    More on the space problem (this time for Win32).
595    
596    ----------------------------------------------------------------------
597    Name: Matthias Blume (blume (at) tti - c (dot) org)
598    Date: 2004/12/14 17:30:00 CST
599    Tag: blume-20041214-spaces
600    Description:
601    
602    Hacked some of the scripts (in particular: the installer) to cope with
603    spaces in filenames a bit better.  But beware: the current "solution"
604    is likely still full of bugs and inherently incomplete.  (We need to
605    do away with those shell scripts for a comprehensive solution.)
606    
607    ----------------------------------------------------------------------
608    Name: Matthias Blume (blume (at) tti - c (dot) org)
609    Date: 2004/12/13 14:45:00 CST
610    Tag: blume-20041213-ml-makedepend
611    Description:
612    
613    Fixed bug in code for ml-makedepend.
614    
615    ----------------------------------------------------------------------
616    Name: Matthias Blume (blume (at) tti - c (dot) org)
617    Date: 2004/12/09 16:30:00 CST
618    Tag: blume-20041209-statistics
619    Description:
620    
621    Added two simple but potentially useful statistics modules to SML/NJ Library.
622    (See CHANGES file there.)
623    
624    ----------------------------------------------------------------------
625    Name: Matthias Blume (blume (at) tti - c (dot) org)
626    Date: 2004/12/01 16:50:00 CST
627    Tag: blume-20041201-atom
628    Description:
629    
630    smlnj-lib:
631    
632      Added function HashString.hashString' for substrings.
633      Hand-inlined CharVector.foldl into HashString (for speed).
634      Modified implementation of structure Atom to avoid extracting
635      strings from substrings unless necessary.
636      (Also see CHANGES file for smlnj-lib.)
637    
638    ----------------------------------------------------------------------
639    Name: Matthias Blume (blume (at) tti - c (dot) org)
640    Date: 2004/11/24 22:15:00 CST
641    Tag: blume-20041124-cml
642    Description:
643    
644    Made sure CML compiles when Position = Int64.
645    
646    ----------------------------------------------------------------------
647    Name: Matthias Blume (blume (at) tti - c (dot) org)
648    Date: 2004/11/24 14:30:00 CST
649    Tag: blume-20041124-position
650    Description:
651    
652    The compiler can now be compiled in a mode that makes structure Position
653    equal to Int64.  The default, however, is unchanged (Position = Int31)
654    for the time being.
655    
656    To enable 64-bit positions, use the following procedure:
657    
658    1. Start sml
659    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
660    3. Type
661          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
662    4. Run CMB.make() as usual.
663    
664    This is barely tested.  The only test so far was a little SML program
665    counting the number of characters in an 8-gigabyte file by
666    reading it character-by-character.  That test was successful.
667    
668    In support of 64-bit positions, a number of new functions have been
669    added to the runtime system.
670    
671    ----------------------------------------------------------------------
672    Name: Matthias Blume (blume (at) tti - c (dot) org)
673    Date: 2004/11/23 14:45:00 CST
674    Tag: blume-20041123-useFile
675    Description:
676    
677    Fixed a problem with unhelpful error messages related to problems with
678    .cm- or .sml files that appear as part of the sml command line.
679    
680    ----------------------------------------------------------------------
681    Name: Matthias Blume (blume (at) tti - c (dot) org)
682    Date: 2004/11/18 15:40:00 CST
683    Tag: Release_110_51
684    Description:
685    
686    New working version (110.51).  NEW BOOTFILES!
687    
688    ----------------------------------------------------------------------
689    Name: Matthias Blume (blume (at) tti - c (dot) org)
690    Date: 2004/11/18 15:35:00 CST
691    Tag: <none>
692    Description:
693    
694    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
695    Joemann).
696    
697    ----------------------------------------------------------------------
698    Name: Allen Leung (leunga (at) reservoir (dot) com)
699    Date: 2004/11/17 16:05:21 EST 2004
700    Tag: leunga-20041117-mlrisc-live-kill
701    Description:
702    
703       Added support for MLTree constructs LIVE and KILL
704    to all the architectures.
705    
706    ----------------------------------------------------------------------
707    Name: Matthias Blume (blume (at) tti - c (dot) org)
708    Date: 2004/11/13 00:20:00 CST
709    Tag: blume-20041113-versiontool
710    Description:
711    
712    - Stripped down the versiontool: It now only handles the version number.
713      The date string is generated at bootstrap time (during makeml).
714    
715    - In a previous commit, fixed a minor issue with how polyequal is being
716      translated.  In particular, the code now "looks through" abstractions.
717      This results in slightly fewer polyEqual warnings and hopefully slightly
718      more efficient code.  Important examples for where this matters are
719      the new int64 and word64 types.
720    
721    ----------------------------------------------------------------------
722    Name: Matthias Blume (blume (at) tti - c (dot) org)
723    Date: 2004/11/12 00:30:00 CST
724    Tag: blume-20041112-int64
725    Description:
726    
727    Structure Int64 fully hooked in.  (The implementation is not very
728    efficient, though.)
729    
730    ----------------------------------------------------------------------
731    Name: Matthias Blume (blume (at) tti - c (dot) org)
732    Date: 2004/11/11 17:30:00 CST
733    Tag: blume-20041111-more64
734    Description:
735    
736    All the pieces of Word64 are now there, with the exception of the
737    conversions from and to LargeWord. (Eventually these need to be identities,
738    but for the time being they don't even make sense because LargeWord is
739    32-bit wide.)
740    
741    Also started to add similar support for Int64, but major pieces of that
742    are still missing.
743    
744    ----------------------------------------------------------------------
745    Name: Matthias Blume (blume (at) tti - c (dot) org)
746    Date: 2004/11/11 00:15:00 CST
747    Tag: blume-20041111-word64
748    Description:
749    
750    Structure Word64 is now (almost) complete, word literals and patterns
751    seem to work.  There are a few odd pieces missing.  In particular,
752    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
753    at the moment.
754    
755    Making Word64 official would mean that LargeWord becomes Word64.  But
756    this requires extreme care because most word-word conversions have to
757    go through LargeWord, so making a mistake means loss of efficiency or
758    worse.  Eventually there will be a solution similar to (but actually
759    simpler than) what I did with IntInf.
760    
761    ----------------------------------------------------------------------
762    Name: Matthias Blume (blume (at) tti - c (dot) org)
763    Date: 2004/11/10 18:12:00 CST
764    Tag: blume-20041110-64bit
765    Description:
766    
767    More 64-bit hacking (but still not even half-way there yet).
768    Also, some assorted improvements to the handling of 8-bit words.
769    
770    ----------------------------------------------------------------------
771    Name: Matthias Blume (blume (at) tti - c (dot) org)
772    Date: 2004/11/09 17:50:00 CST
773    Tag: <none>
774    Description:
775    
776    Started adding some infrastructure for supporting 64-bit int- and
777    word-types.  (Still in its very early stages.)
778    
779    ----------------------------------------------------------------------
780    Name: Matthias Blume (blume (at) tti - c (dot) org)
781    Date: 2004/10/28 10:45:00 CDT
782    Tag: Release_110_50
783    Description:
784    
785    New working version (110.50).  NEW BOOTFILES!
786    
787                 =====================
788    
789    Also:
790    
791      - Changend config/srcarchiveurl from a file just
792        containing the URL string into a file containing
793        shell script code.  The code has access to the $VERSION variable.
794      - Made corresponding changes to config/install.sh and config/unpack.
795      - Default contents of config/srcarchiveurl uses $VERSION and
796        normally does not have to be edited to reflect a version change.
797    
798      (As a result, a version change can be done by just editing
799       config/version, the rest is now automatic.)
800    
801    ----------------------------------------------------------------------
802    Name: Matthias Blume (blume (at) tti - c (dot) org)
803    Date: 2004/10/27 17:50:00 CDT
804    Tag: blume-20041027-btrace-msg
805    Description:
806    
807    BackTrace.monitor now also reports the source of the exception that
808    triggered the trace.
809    
810    ----------------------------------------------------------------------
811    Name: Matthias Blume (blume (at) tti - c (dot) org)
812    Date: 2004/10/27 17:20:00 CDT
813    Tag: blume-20041027-x86-c-calls
814    Description:
815    
816    This is the HISTORY entry for two earlier commits, both concerning
817    the x86 c-calls code in MLRISC:
818    
819      - added a missing LOAD in the code that deals with struct arguments
820      - made sure the caller does not add the wrong number of bytes to the
821        stack pointer after a call of a function returning a struct
822        (the callee already pops the implicit argument which points to
823         the space reserved for the result)
824    
825    ----------------------------------------------------------------------
826    Name: Allen Leung (leunga (at) reservoir (dot) com)
827    Date: 2004/10/24 14:00:00 EST
828    Tag: leunga-20041024-x86-gas-fucomip
829    Description:
830    
831       John discovered a bug in the syntax of fucomip.
832       The opcodes FU?COMIP? have been changed to
833    
834           fu?comip? %st(i), %st
835    
836    ----------------------------------------------------------------------
837    Name: Matthias Blume (blume (at) tti - c (dot) org)
838    Date: 2004/10/20 15:06:00 CDT
839    Tag: blume-20041020-standalone-backtrace
840    Description:
841    
842    Added a mechanism for getting back-trace information from standalone
843    programs.  Here is how it works:
844    
845       1. The part of the program from which you want to get backtrace
846          information (usually the whole program) should be wrapped with
847          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
848          main program could be modified from something like
849    
850             fun main (pgm, args) = ...
851    
852         to
853    
854            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
855    
856       2. To be able to access BackTrace.monitor, you have to add
857          library $smlnj-tdp/plugins.cm to the .cm file that contains your
858          main function.
859    
860       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
861          CM might have created in the past for your project).
862    
863       4. Build the system using this command line:
864    
865             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
866                      myprog.cm MyProg.main myprog
867    
868          instead of the usual
869    
870             ml-build myprog.cm MyProg.main myprog
871    
872    I changed a library name:
873    
874       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
875    
876    New libraries:
877    
878       $smlnj-tdp/back-trace.cm
879          -- when loaded causes the back-trace plugin to be installed
880       $smlnj-tdp/coverage.cm
881          -- when loaded causes the coverage plugin to be installed
882    
883    ----------------------------------------------------------------------
884    Name: Matthias Blume (blume (at) tti - c (dot) org)
885    Date: 2004/10/18 16:45:00 CDT
886    Tag: blume-20041018-groupowner
887    Description:
888    
889    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
890    
891    Eliminated group owner specs from .cm files throughout the source tree.
892    
893    ----------------------------------------------------------------------
894    Name: Matthias Blume (blume (at) tti - c (dot) org)
895    Date: 2004/10/15 15:45:00 CDT
896    Tag: blume-20041015-coverage
897    Description:
898    
899    * Test coverage tool added!
900    * Further reorganization of tracing-, debugging-, and profiling support:
901    
902        - moved original BTImp -- now called BackTrace -- into a separate
903          library called $/trace-debug-profile.cm
904        - eliminated all mentions of BTrace from SMLofNJ.Internals
905        - only the instrumentation mechanism is now left in the compiler proper
906        - BackTrace module is a plugin which is NOT plugged in by default
907        - Coverage module is another such plugin
908    
909    To get the benefits of any of these plugin modules, the code in
910    question must be compiled with tdp instrumentation turned on.  This
911    can be done by setting SMLofNJ.Internals.TDP.mode to true.
912    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
913    
914    Plugins are selected at link time.  (Pre-compiled instrumented code
915    can be re-loaded with different plugins in effect.)  When an
916    instrumented module is linked, whatever plugins are at that time
917    enabled will come into effect for that module.
918    
919    To enable the back-trace plugin, load library $/trace-debug-profile.cm
920    and invoke BackTrace.install() (e.g., from the interactive prompt).
921    To enable the coverage plugin, load the same library and invoke
922    Coverage.install().
923    
924    Back-traces are generated automatically on uncaught exceptions and
925    when the code in question explicitly invokes BackTrace.trigger().
926    
927    Coverage (and execution frequency-) information must be queried
928    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
929    
930    ----------------------------------------------------------------------
931    Name: Matthias Blume (blume (at) tti - c (dot) org)
932    Date: 2004/10/14 17:40:00 CDT
933    Tag: blume-20041014-tdp-core
934    Description:
935    
936    Snapshot of a significant overhaul of how the trace/debug/profile support
937    is hooked into the system (specifically: Core and SMLofNJ.Internals).
938    
939    ----------------------------------------------------------------------
940    Name: Matthias Blume (blume (at) tti - c (dot) org)
941    Date: 2004/10/13 16:34:00 CDT
942    Tag: blume-20041013-tdp
943    Description:
944    
945    Some rationalization of names:
946    
947       structure BTrace -> structure TDPInstrument
948       etc.
949    
950    This is is preparation of using the original back-trace
951    instrumentation for other purposes.  "TDP" stands for
952    Trace/Debug/Profile.
953    
954    The control flag controlling whether instrumentation is on or off is now
955    registered under a different name, so instead of running sml as
956    
957     sml -Cinstrument.btrace-mode=true
958    
959    one has to say
960    
961     sml -Ctdp.instrument=true
962    
963    ----------------------------------------------------------------------
964    Name: Matthias Blume (blume (at) tti - c (dot) org)
965    Date: 2004/10/11 16:37:00 CDT
966    Tag: blume-20041011-regions
967    Description:
968    
969    Made some minor modifications to elabcore.sml to have source regions
970    be propagated more tightly -- resulting in better (i.e., smaller)
971    regions being reported in error- and debug messages.
972    
973    ----------------------------------------------------------------------
974    Name: Matthias Blume (blume (at) tti - c (dot) org)
975    Date: 2004/10/08 22:50:00 CDT
976    Tag: blume-20041008-cmkw
977    Description:
978    
979    Fixed handling of keywords in .cm files: After seeing "is" the lexer
980    treats subsequent occurrences of "group", "library", "source", "is",
981    "*", and "-" as ordinary identifiers rather than keywords.
982    
983    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
984    is supposed to accept a tool argument called "source", but this did
985    not work because of the clash with the keyword.
986    
987    ----------------------------------------------------------------------
988    Name: Matthias Blume (blume (at) tti - c (dot) org)
989    Date: 2004/10/07 16:00:00 CDT
990    Tag: blume-20041007-cleanup
991    Description:
992    
993    Assorted cleanup work:
994    
995      - got rid of intstrmap in favor of using the library's
996        hash table implementation
997      - threw out most of the pathnames stuff, as it was not used anyway
998      - simplified tokentable implementation
999      - fixed some minor spelling errors
1000    
1001    ----------------------------------------------------------------------
1002    Name: Matthias Blume (blume (at) tti - c (dot) org)
1003    Date: 2004/10/06 15:15:15 CDT
1004    Tag: blume-20041006-handler
1005    Description:
1006    
1007    Cleaned up the absyn to reflect the invariant that HANDLE always
1008    carries a FNexp as part of the type definition.  This eliminates some
1009    superfluous sanity checks at runtime down the road.
1010    
1011    Some minor cleanup of the btrace code.
1012    
1013    ----------------------------------------------------------------------
1014    Name: Matthias Blume (blume (at) tti - c (dot) org)
1015    Date: 2004/10/01 10:20:30 CDT
1016    Tag: blume-20041001-slave
1017    Description:
1018    
1019    Added hack to make slave mode work in the presence of the version
1020    tool.  (Still, since the master does two passes over the code for
1021    CMB.make, the release number gets bumped twice when slaves are
1022    attached.  I don't know if this is worth fixing...)
1023    
1024    ----------------------------------------------------------------------
1025    Name: Matthias Blume (blume (at) tti - c (dot) org)
1026    Date: 2004/09/30 10:55:00 CDT
1027    Tag: blume-20040930-version
1028    Description:
1029    
1030    * Moved the "version" magic into its own little library under
1031      src/system/smlnj/internal.  This avoids expensive reconstruction of
1032      a stable src/compiler/core.cm.
1033    
1034    * At the same time, structure CompilerVersion is now known as
1035      structure SMLNJVersion.
1036    
1037    * Arranged for the version tool to NOT kick in when rebuilding the system
1038      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
1039      Also, loading the versiontool does not work when rebuilding the system
1040      because CM is not properly initialized at that time.
1041    
1042    ----------------------------------------------------------------------
1043    Name: Matthias Blume (blume (at) tti - c (dot) org)
1044    Date: 2004/09/29 14:00:00 CDT
1045    Tag: blume-20040929-autoversion
1046    Description:
1047    
1048    Implemented some CM magic to have
1049        file src/compiler/TopLevel/main/version.sml
1050    generated automagically.
1051    The version is taken from two files: config/version and config/release.
1052    The first is expected to contain a two-part version number such as 110.49.
1053    The second should contain a single number, but it may be missing.
1054    
1055    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
1056    time the version tool is loaded (which is the first time you say CMB.make),
1057    then the tool will increment the value stored in config/release every
1058    time CMB.make is invoked.
1059    
1060    The binfile format is now insensitive to anything beyond the first
1061    two components of a version number, so bumping the release does not render
1062    binfiles incompatible.  Auto-bumping can be used to keep track of versions
1063    during development without invalidating existing binfiles.
1064    
1065    In any case, every CMB.make updates the date information in version.sml.
1066    (This is the date that is printed in the banner.)
1067    
1068    ----------------------------------------------------------------------
1069    Name: Matthias Blume (blume (at) tti - c (dot) org)
1070    Date: 2004/09/28 10:53:00 CDT
1071    Tag: blume-20040928-controls
1072    Description:
1073    
1074    Some cleanup of the controls code.
1075    
1076    ----------------------------------------------------------------------
1077    Name: Matthias Blume (blume (at) tti - c (dot) org)
1078    Date: 2004/09/27 22:08:00 CDT
1079    Tag: blume-20040927-controls
1080    Description:
1081    
1082    Added two pieces of functionality to the Controls interface:
1083    
1084    1.   val save'restore: 'a control -> unit -> unit
1085    
1086       grabs the current value of the control in stage 1 and restores it
1087       in stage 2.
1088    
1089    2.   val set' : 'a control * 'a -> unit -> unit
1090    
1091       stores the given value into the control in stage 2 (i.e., delayed)
1092       but does all error checking in stage 1.
1093       (This is for string controls that need to do parse their argument --
1094        something that might fail.  In some cases, notably in CM, one
1095        already knows the intended argument but wants to delay the actual
1096        assignment until a time when error recovery would be more difficult.)
1097    
1098    Changed the handling of controls in tool arguments to classes "sml" and
1099    "lazysml":
1100       - use Controls.save'restore as a more robust way of restoring the
1101         old value (in particular: without having to re-parse the string)
1102       - use controls to handle the "overload" keyword in the init group
1103         (I believe this change actually fixes a long-standing obscure bug.)
1104    
1105    ----------------------------------------------------------------------
1106    Name: Matthias Blume (blume (at) tti - c (dot) org)
1107    Date: 2004/09/27 17:00:00 CDT
1108    Tag: blume-20040927-lazysml
1109    Description:
1110    
1111    Added a new tool class called "lazysml" to CM's tool chest.  The only
1112    difference to "sml" is that compilation is done with Control.lazysml
1113    set to true.  A source of class "lazysml" is automatically recognized
1114    by a file name suffix of ".lml".
1115    
1116    In addition to the above feature, the original class "sml" now also
1117    supports a tool argument "lazy" which has the same effect.  As a
1118    result, the following three lines are equivalent:
1119    
1120        foo.sml : lazysml
1121        foo.sml : sml (lazy)
1122        foo.sml (lazy)
1123    
1124    The setting goes into effect both during parsing and during
1125    compilation.  The original setting is restored right after parsing and
1126    after compilation, respectively.
1127    
1128    In addition to all the above, there is also a general mechanism to set
1129    ANY of the "controls" that are available at the command line via
1130    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
1131    apply as well. (In fact, "lazy" is implemented as a special case of
1132    the general mechanism.)
1133    
1134    The .cm file syntax uses a new keyword tool argument called "with".
1135    There are several ways of indicating the desired settings:
1136    
1137        foo.sml (with:parser.quotations=true)
1138        foo.sml (with:(name:parser.quotations value:true))
1139        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1140        foo.sml (with:(name1=value1 name2=value2 ...))
1141        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1142    
1143        etc.
1144    
1145    Another possible abbreviation is to leave out the =v or value:v part
1146    if the name refers to a boolean control (in which case the value is
1147    taken to be true).  Thus, one could get lazy sml also by saying:
1148    
1149        foo.sml (with:parser.lazy-keyword=true)
1150        foo.sml (with:parser.lazy-keyword)
1151        foo.sml (with:(name:parser.lazy-keyword value:true))
1152        foo.sml (with:(name:parser.lazy-keyword))
1153    
1154    ----------------------------------------------------------------------
1155    Name: Matthias Blume (blume (at) tti - c (dot) org)
1156    Date: 2004/09/24 16:22:00 CDT
1157    Tag: blume-20040924-ppc-long-branch
1158    Description:
1159    
1160    Turned message about "emiting long form of branch" off by default.
1161    Added a control flag to turn it back on when desired.
1162    
1163    ----------------------------------------------------------------------
1164    Name: Matthias Blume (blume (at) tti - c (dot) org)
1165    Date: 2004/09/24 16:05:00 CDT
1166    Tag: blume-20040924-rounding
1167    Description:
1168    
1169    Applied patch for setting rounding modes under Mac OS X.  Thanks to
1170    Melissa O'Neill for providing the code!
1171    
1172    ----------------------------------------------------------------------
1173    Name: Matthias Blume (blume (at) tti - c (dot) org)
1174    Date: 2004/09/23 17:30:00 CDT
1175    Tag: blume-20040923-envvars
1176    Description:
1177    
1178    1. Changed definition of type ControlRegistry.registry_tree to
1179       include control_info (i.e., the name of the controlling environment
1180       variable).
1181    
1182    2. Added command-line flags -e and -E to print the names of environment
1183       variables that can be used to control internal settings.
1184       (This uses the new API mentioned in 1.)
1185    
1186    ----------------------------------------------------------------------
1187    Name: Matthias Blume (blume (at) tti - c (dot) org)
1188    Date: 2004/09/13 16:50:00 CDT
1189    Tag: Release_110_49
1190    Description:
1191    
1192    New working version (110.49).  NEW BOOTFILES!
1193    
1194    ----------------------------------------------------------------------
1195    Name: Matthias Blume
1196    Date: 2004/09/13 16:20:00 CDT
1197    Tag: blume-20040913-config-mlrisc
1198    Description:
1199    
1200    Put target "mlrisc" back into the default list.
1201    (There is no harm in having it, and some users have expressed their
1202    wish to have "mlrisc" included by default.)
1203    
1204    ----------------------------------------------------------------------
1205    Name: John Reppy
1206    Date: 2004/09/13
1207    Tag: jhr-20040913-signals
1208    Description:
1209    
1210    Fixed the signal masking code to properly nest mask/unmask operations
1211    on a per-signal basis.
1212    
1213    ----------------------------------------------------------------------
1214    Name: Matthias Blume (blume (at) tti - c (dot) org)
1215    Date: 2004/09/08 13:20:00 CDT
1216    Tag: blume-20040908-heap-magic
1217    Description:
1218    
1219    Bumped the heap macig number to 0x09082004 to account for the changed
1220    layout of the ML frame under MacOS X.
1221    
1222    ----------------------------------------------------------------------
1223    Name: Allen Leung (leunga (at) reservoir (dot) com)
1224    Date: 2004/09/03 11:26:00 EST
1225    Tag: leunga-20040903-cygwin-install
1226    Description:
1227    
1228       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1229    The Cygwin runtime is turned on by setting the environment
1230    variable SMLNJ_CYGWIN_RUNTIME to 1.
1231    
1232    ----------------------------------------------------------------------
1233    Name: Matthias Blume (blume (at) tti - c (dot) org)
1234    Date: 2004/08/31 17:15:00 CDT
1235    Tag: blume-20040831-core
1236    Description:
1237    
1238    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1239    
1240    ----------------------------------------------------------------------
1241    Name: Matthias Blume (blume (at) tti - c (dot) org)
1242    Date: 2004/08/30 17:55:00 CDT
1243    Tag: blume-20040830-installer
1244    Description:
1245    
1246    Upon request by Johannes Joemann:
1247    
1248    - improved ML code of installer to fall back to coping when renaming
1249      fails (i.e., when source and target are on different file systems);
1250      the code compiles but has yet to be tested in anger
1251    - removed mlrisc from list of default targets (config/targets)
1252    
1253    ----------------------------------------------------------------------
1254    Name: Matthias Blume (blume (at) tti - c (dot) org)
1255    Date: 2004/08/27 17:20:00 CDT
1256    Tag: blume-20040827-ptreql
1257    Description:
1258    
1259    Added ptreql primop to structure InlineT (upon request from Larry
1260    Paulson).
1261    
1262    ----------------------------------------------------------------------
1263    Name: Allen Leung (leunga (at) reservoir (dot) com)
1264    Date: 2004/08/15 21:21:00 EST
1265    Tag: leunga-110_48-udgraph
1266    Description:
1267    
1268    Another bug fix from Carl Hauser:
1269    
1270    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1271    > 48c48
1272    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1273    > e::L)
1274    > ---
1275    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1276    > Without this, any deletion of an edge in an undirected graph does severe
1277    > violence to the graph.
1278    
1279    ----------------------------------------------------------------------
1280    Name: Allen Leung (leunga (at) reservoir (dot) com)
1281    Date: 2004/08/10 23:35:00 EST
1282    Tag: leunga-110_48-ppc
1283    Description:
1284    
1285       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1286    
1287    ----------------------------------------------------------------------
1288    Name: Matthias Blume (blume (at) tti - c (dot) org)
1289    Date: 2004/08/10 12:00:00 CDT
1290    Tag: Release_110_48
1291    Description:
1292    
1293    New working version (110.48).  NEW BOOTFILES!
1294    
1295    ----------------------------------------------------------------------
1296    Name: Allen Leung (leunga (at) reservoir (dot) com)
1297    Date: 2004/08/09 12:21:00 EST
1298    Tag: leunga-110_47-dijsktra
1299    Description:
1300    
1301    Bug fix from Carl Hauser:
1302    
1303    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1304    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1305    
1306    The problem is that following the expression A.update(dist,s,Num.zero)
1307    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1308    
1309    ----------------------------------------------------------------------
1310    Name: Matthias Blume (blume (at) tti - c (dot) org)
1311    Date: 2004/08/06 18:10:00 CDT
1312    Tag: blume-20040806-cmdline
1313    Description:
1314    
1315    Fiddled with handling of command-line options:
1316    
1317      * sml now quits after processing the command line
1318        if -H, -S, -h<n>, or -s<n> appears as the last
1319        command-line argument
1320      * a new option -q terminates the session when encountered on
1321        the command line; subsequent arguments will be ignored
1322      * bug fixes: short (erroneous) arguments are no longer ignored
1323        completely
1324    
1325    ----------------------------------------------------------------------
1326    Name: Allen Leung (leunga (at) reservoir (dot) com)
1327    Date: 2004/08/04 18:17:00 EST
1328    Tag: leunga-110_47-ppc-ibm-asm
1329    Description:
1330    
1331      - Added minimal IBM assembly syntax support for PowerPC.
1332    
1333      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1334        been altered in the runtime.  We need an automatic way to keep the file
1335        in sync.
1336    
1337    ----------------------------------------------------------------------
1338    Name: Matthias Blume (blume (at) tti - c (dot) org)
1339    Date: 2004/08/04 14:00:00 CDT
1340    Tag: Release_110_47
1341    Description:
1342    
1343    New working version (110.47).  NEW BOOTFILES!
1344    
1345    ----------------------------------------------------------------------
1346    Name: Matthias Blume (blume (at) tti - c (dot) org)
1347    Date: 2004/08/03 14:25:00 CDT
1348    Tag: blume-20040803-callingconv
1349    Description:
1350    
1351    Added low-level support for choosing C calling conventions by
1352    twiddling the type of rawccall.  (See
1353    src/compiler/Semant/types/cproto.sml for details.)
1354    
1355    ----------------------------------------------------------------------
1356    Name: Matthias Blume (blume (at) tti - c (dot) org)
1357    Date: 2004/08/02 15:55:00 CDT
1358    Tag: blume-20040802-backout
1359    Description:
1360    
1361    Backed out of change to win32-filesys.c.  The earlier patch to
1362    get_file_time caused CM to produce files with the wrong time stamp.
1363    
1364    ----------------------------------------------------------------------
1365    Name: Matthias Blume (blume (at) tti - c (dot) org)
1366    Date: 2004/08/02 14:45:00 CDT
1367    Tag: blume-20040802-nlffi-win32
1368    Description:
1369    
1370    Added NLFFI support for Win32, adapted from a patch provided by David
1371    Hansel.  This is currently completely untested.  Also, the issue
1372    concerning stdcall vs. ccall is still unresolved.
1373    
1374    ----------------------------------------------------------------------
1375    Name: Matthias Blume (blume (at) tti - c (dot) org)
1376    Date: 2004/07/30 17:55:00 CDT
1377    Tag: blume-20040730-various
1378    Description:
1379    
1380    Gearing up towards 110.47...
1381    
1382    - various minor bugfixes to ml-nlffigen
1383    - a beginning of a manual for nlffi
1384    
1385    - eliminated 'export name=value' in config/install.sh as this does
1386      not work with certain versions of /bin/sh
1387      (Thanks to David King at Motorola for catching this.)
1388    
1389    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1390      - added a test for tm==NULL to gmtime.c and localtime.c
1391      - applied patch for incorrect GetFileTime under win32
1392      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1393    
1394    ----------------------------------------------------------------------
1395    Name: Matthias Blume (blume (at) tti - c (dot) org)
1396    Date: 2004/07/21 18:20:00 CDT
1397    Tag: blume-20040721-nlffigen
1398    Description:
1399    
1400    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1401      for a typedef to an incomplete type, but leave out the "typ" member.
1402      (This is just for consistency.)
1403    - Started to produce what is supposed to become better (i.e., comprehensive)
1404      documentation of what ml-nlffigen does and produces.
1405    
1406    ----------------------------------------------------------------------
1407    Name: Matthias Blume (blume (at) tti - c (dot) org)
1408    Date: 2004/07/14 16:25:00 CDT
1409    Tag: blume-20040714-union
1410    Description:
1411    
1412    Added C_UNION to c-calls/c-types.sml and updated the machinery
1413    (ml-nlffigen, cproto.sml) that conveys C function interface
1414    information to the code generator.
1415    
1416    However, the actual architecture-specific implementation of function
1417    arguments and results that are C unions is still not implemented.
1418    
1419    ----------------------------------------------------------------------
1420    Name: Allen Leung (leunga (at) reservoir (dot) com)
1421    Date: 2004/07/14 14:38:00 EST
1422    Tag: leunga-110_46_1-ppc-lwzu
1423    Description:
1424    
1425        Added these instructions to the PowerPC architecture:
1426        LBZU(X), LHZU(X), LWZU(X),
1427        STWU(X), STFDU, STFSU
1428    
1429        etc...
1430    
1431        Note: I haven't added their instruction encoding into the description.
1432    
1433    ----------------------------------------------------------------------
1434    Name: Allen Leung (leunga (at) reservoir (dot) com)
1435    Date: 2004/07/13 15:04:00 EST
1436    Tag: leunga-110_46_1-ppc-lwarx
1437    Description:
1438    
1439       Added the two instructions LWARX and STWCX to the PowerPC
1440    instruction set.
1441    
1442       A (untested) rewrite of loop-structure.sml.  The old version
1443    is completely broken.
1444    
1445    ----------------------------------------------------------------------
1446    Name: Matthias Blume (blume (at) tti - c (dot) org)
1447    Date: 2004/07/13 13:50:00 CDT
1448    Tag: blume-20040713-nlffi
1449    Description:
1450    
1451    - use paramAlloc to report c-calls with too many arguments
1452      (for PPC version where parameter area is pre-allocated)
1453    - added ccall_maxargspace to machspec (to implement the above)
1454    - made "make" commend in CM's "make" tool configurable
1455    - added option (default: on) for passing the name of the SML/NJ's "bin"
1456      directory to "make"; the call looks like this:
1457    
1458         make <options> SMLNJ_BINDIR=<dir> <target>
1459    
1460      This can be used by the Makefile to, e.g., pick the "right" version
1461      of ml-nlffigen.
1462    - minor code tweaks
1463    
1464    ----------------------------------------------------------------------
1465    Name: Matthias Blume (blume (at) tti - c (dot) org)
1466    Date: 2004/07/12 22:50:00 CDT
1467    Tag: blume-110_46_1-macosx-nlffi
1468    Description:
1469    
1470    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1471    though.
1472    
1473    Note:  1. You have to make a new, clean build of the runtime system.
1474           2. There are new BOOTFILES, you have to use them!
1475              (Doing the bootstrap process yourself would be *very* painful!
1476               If you absolutely have to do it, build the system under
1477               a different architecture and then cross-compile.)
1478    
1479    Version bumped to 110.46.1 to account for runtime data format changes.
1480    
1481    ----------------------------------------------------------------------
1482    Name: Matthias Blume (blume (at) tti - c (dot) org)
1483    Date: 2004/06/18 14:30:00 CDT
1484    Tag: blume-20040618-unix
1485    Description:
1486    
1487    Changed the implementation of structure Unix so that the same stream
1488    is returned every time one of the {text,bin}{In,Out}streamOf functions
1489    is invoked on the same proc.  This is not what the spec currently
1490    says -- although IMO it arguably should.  (See discussion below.)
1491    
1492    ----------------------------------------------------------------------
1493    Name: Matthias Blume (blume (at) tti - c (dot) org)
1494    Date: 2004/06/17 18:15:00 CDT
1495    Tag: Release_110_46
1496    Description:
1497    
1498    New working version (110.46).  NEW BOOTFILES!
1499    
1500    ----------------------------------------------------------------------
1501    Name: Matthias Blume (blume (at) tti - c (dot) org)
1502    Date: 2004/06/17 17:20:00 CDT
1503    Tag: blume-20040617-timer-unix
1504    Description:
1505    
1506    Changed the interface of structures Timer and Unix to match the most
1507    recent Basis spec.
1508    
1509    In the case of Unix there still seems to be an open/weird issue:
1510    
1511       The {text,bin}{In,Out}streamOf functions are supposed to create
1512       fresh streams whenever they are called -- as opposed to have them
1513       return the same stream every time.  This design is supposed to
1514       prevent space leaks caused by proc values hanging on to streams.
1515    
1516       The reap function, on the other hand, is supposed to close the
1517       streams.  This cannot be done without having a handle on the
1518       stream in proc after all...
1519    
1520    I took the liberty to implement the following stopgap solution:
1521    
1522      The proc value hangs on to the most recently created stream(s).
1523      Reap closes those.  If either or both of the two streams hadn't
1524      been created at all yet, then reap will close the corresponding
1525      file descriptors directly.
1526    
1527    PS: I don't understand the original space leak argument anymore.  If
1528    a proc hangs on to the imperative stream, then I/O operations on those
1529    will advance the state of the cached stream and avoid the space leak.
1530    
1531    ----------------------------------------------------------------------
1532    Name: Matthias Blume (blume (at) tti - c (dot) org)
1533    Date: 2004/05/28 16:45:00 CDT
1534    Tag: blume-20040528-basis
1535    Description:
1536    
1537    Added signature PACK_REAL and exported functor PrimIO.
1538    
1539    ----------------------------------------------------------------------
1540    Name: Matthias Blume (blume (at) tti - c (dot) org)
1541    Date: 2004/05/25 16:00:00 CDT
1542    Tag: blume-20040525-group-owner
1543    Description:
1544    
1545    CM now ignores (but still accepts) the "owner" information in group
1546    descriptions.  The owner of a group is its next enclosing
1547    library. Each group must have a unique owner.  (There is a virtual
1548    "toplevel" library that owns groups which are not nested within a real
1549    library.)  Previously, each group had to explicitly declare its owner,
1550    and CM would check that such a declaration is correct.  The new scheme
1551    is to have CM check that for each group there is precisely one owning
1552    library.
1553    
1554    The advantage of the new scheme is that the programmer no longer needs
1555    to maintain the somewhat annoying owner information.  The downside is
1556    that CM cannot enforce the ownership rule across multiple runs of
1557    CM.make.  Fortunately, enclosing the same group in two different
1558    libraries A and B which are not part of the same program does not
1559    cause real problems.
1560    
1561    ----------------------------------------------------------------------
1562    Name: Matthias Blume (blume (at) tti - c (dot) org)
1563    Date: 2004/05/20 16:00:00 CDT
1564    Tag: blume-20040520-win32
1565    Description:
1566    
1567    Made the win32 version work again.  (Strangely, a misplaced comma had
1568    slipped into win32-process.c which prevented the runtime from being
1569    compiled correctly.)
1570    
1571    Also, included a minor addition to ml-build.bat analogous to what was
1572    done in blume-20040519-ml-build.
1573    
1574    ----------------------------------------------------------------------
1575    Name: Matthias Blume (blume (at) tti - c (dot) org)
1576    Date: 2004/05/19 22:10:00 CDT
1577    Tag: blume-20040519-ml-build
1578    Description:
1579    
1580    Arranged for ml-build to clean up after itself a little bit better.
1581    The script generates a temporary SML source file and compiles it using
1582    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1583    gets rid of those at the end, so they don't accumulate under .cm.
1584    
1585    This required a minor change to install.sh because the name of the
1586    metadata directory (default: .cm) is actually configurable at
1587    installation time.
1588    
1589    ----------------------------------------------------------------------
1590    Name: Matthias Blume (blume (at) tti - c (dot) org)
1591    Date: 2004/05/18 15:50:00 CDT
1592    Tag: blume-20040518-mkreader
1593    Description:
1594    
1595    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1596    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1597    
1598    ----------------------------------------------------------------------
1599    Name: Matthias Blume (blume (at) tti - c (dot) org)
1600    Date: 2004/05/11 14:35:00 CDT
1601    Tag: blume-20040511-win32sock
1602    Description:
1603    
1604    Added previously missing support for many socket-related functions
1605    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1606    for the voluminous patch!
1607    
1608    (I have not tested this patch under win32 yet.)
1609    
1610    Here is David's e-mail:
1611    
1612    Hi,
1613    
1614    Attached to this email you find a diff against sml/nj 110.45
1615    that will enable socket support under Windows.
1616    
1617    To apply the patch (using unix or cygwin)
1618    1) gunzip runtime.diff.gz
1619    2) "cd" into "src/runtime" in the source tree of a fresh
1620       110.45 installation.
1621    3) patch -p 1 < [your/path/to]runtime.diff
1622    
1623    The code compiles fine but has NOT yet been extensively tested.
1624    I only ran a few tests for basic socket client functionality
1625    (which worked fine).  Especially the functions that use ioctl
1626    are not tested at all and might not work (see below).
1627    
1628    I implemented this since we want to move to a newer version of sml/nj
1629    but need socket support in order to use it.  This is the first time I
1630    even had a look at the sml/nj source,  so please review my changes
1631    before making this part of the distribution!  Here are a few issues
1632    that I think might be better for someone to solve who is more
1633    familiar with the sml/nj source (and socket programming):
1634    
1635    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1636      exception since I could not figure out what the windows equivalent
1637      of these functions is
1638    
1639    - In sockets-osdep.h there are a some #include statements that are
1640      only used in a few files that include sockets-osdep.h
1641    
1642    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1643      does not process its return value since I don't know how to
1644      report an error upwards.
1645    
1646    - It would probably be good to have a call to WSACleanup() when
1647      the library is unloaded (if there is such a possibility).
1648      Otherwise I think Windows will take care of this automatically
1649      when the process finishes.
1650    
1651    - I used ioctlsocket() as a replacement for ioctl() but I have
1652      no idea if that is actually the proper replacement on Windows.
1653    
1654    - All these issues are marked in the code by "FIXME" comments.
1655    
1656    We use sml/nj extensively in our products and are quite happy
1657    with it.  I hope this contribution will help you.
1658    
1659    Keep up the good work!
1660    
1661    David
1662    
1663    ----------------------------------------------------------------------
1664    Name: Matthias Blume (blume (at) tti - c (dot) org)
1665    Date: 2004/05/11 14:20:00 CDT
1666    Tag: blume-20040511-installml
1667    Description:
1668    
1669    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1670    the patch.)
1671    
1672    ----------------------------------------------------------------------
1673    Name: Matthias Blume (blume (at) tti - c (dot) org)
1674    Date: 2004/05/11 14:05:00 CDT
1675    Tag: blume-20040511-nlffi-netbsd
1676    Description:
1677    
1678    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1679    the patch.)
1680    
1681    ----------------------------------------------------------------------
1682    Name: Matthias Blume (blume (at) tti - c (dot) org)
1683    Date: 2004/05/11 12:05:00 CDT
1684    Tag: blume-20040511-exports
1685    Description:
1686    
1687    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1688    export lists in compiler-related .cm-files.
1689    
1690    ----------------------------------------------------------------------
1691    Name: Matthias Blume (blume (at) tti - c (dot) org)
1692    Date: 2004/05/11 11:35:00 CDT
1693    Tag: blume-20040511-allsource
1694    Description:
1695    
1696    The installer now honors the "src-smlnj" target again, although its meaning
1697    has changed from "all sources required for the compiler" to "all sources
1698    the installer knows about".  In other words, if you enable "src-smlnj"
1699    in the "targets" file, then the installer will pull in sources for
1700    everything.  (Notice that this refers to source code only.  Compiled
1701    code is still only installed for modules that were requested explicitly
1702    or which are required for other modules that were requested explicitly.)
1703    
1704    ----------------------------------------------------------------------
1705    Name: Matthias Blume (blume (at) tti - c (dot) org)
1706    Date: 2004/04/23 17:40:00 CDT
1707    Tag: blume-20040423-ieee-scan
1708    Description:
1709    
1710    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1711    in the exponent calculation we get INF or ZERO (depending on the mantissa
1712    and the sign of the exponent).
1713    
1714    ----------------------------------------------------------------------
1715    Name: Matthias Blume (blume (at) tti - c (dot) org)
1716    Date: 2004/04/23 10:40:00 CDT
1717    Tag: blume-20040423-ml-build
1718    Description:
1719    
1720    The ml-build script now terminates with a non-0 status when something
1721    goes wrong.
1722    
1723    ----------------------------------------------------------------------
1724    Name: Matthias Blume (blume (at) tti - c (dot) org)
1725    Date: 2004/04/22 16:35:00 CDT
1726    Tag: blume-20040422-Option
1727    Description:
1728    
1729    Made exception Option to be the same as exception Option.Option
1730    (as it should be).
1731    
1732    ----------------------------------------------------------------------
1733    Name: Allen Leung (leunga (at) reservoir (dot) com)
1734    Date: 2004/03/19 14:40:00 EST
1735    Tag: leunga-20040319-cygwin-nlffi
1736    Description:
1737    
1738    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1739    of SML/NJ.  The problem is that
1740    
1741       lib = dlopen(NULL, ...)
1742       f   = dlsym(lib, "malloc");
1743    
1744    does not work on Windows unless we explicitly export symbols
1745    such as 'malloc' during linking.  We fixed this by explicitly
1746    exporting the required symbols with the magic gcc incantation:
1747    
1748        -Wl,--export-all cygwin.def
1749    
1750    where cygwin.def is a file containing all the symbols that
1751    we wish to export.
1752    
1753    I suspect this is a Windows problem and we'll have to
1754    do the same (somehow with windows compilers) when
1755    we build the native win32 version with the system
1756    calls LoadLibrary/GetProcAddress.
1757    
1758    ----------------------------------------------------------------------
1759    Name: Matthias Blume (blume (at) tti - c (dot) org)
1760    Date: 2004/03/04 16:35:00 CST
1761    Tag: blume-20040304-intinf-fmt
1762    Description:
1763    
1764    Fixed problem with IntInf.fmt (sign would show up on the right instead
1765    of on the left for BIN, OCT, and HEX).
1766    
1767    ----------------------------------------------------------------------
1768    Name: Matthias Blume (blume (at) tti - c (dot) org)
1769    Date: 2004/03/04 11:25:00 CST
1770    Tag: blume-20040304-symlinks
1771    Description:
1772    
1773    Fixed problem with installer script (unix only) where bin/ml-yacc and
1774    friends pointed (via symlinks) to absolute locations instead of just
1775    .run-sml.  This was reported by Vesa A Norrman.
1776    
1777    ----------------------------------------------------------------------
1778    Name: Matthias Blume (blume (at) tti - c (dot) org)
1779    Date: 2004/02/13 14:50:00 CST
1780    Tag: Release_110_45
1781    Description:
1782    
1783    New working version (110.45). New bootfiles.
1784    
1785    ----------------------------------------------------------------------
1786    Name: Matthias Blume (blume (at) tti - c (dot) org)
1787    Date: 2004/01/26 15:15:15 CST
1788    Tag: blume-20040126-toplevel
1789    Description:
1790    
1791    Improved handling of exceptions at the interactive toplevel.
1792    
1793    ----------------------------------------------------------------------
1794    Name: Matthias Blume (blume (at) tti - c (dot) org)
1795    Date: 2004/01/26 11:25:00
1796    Tag: blume-20040126-app
1797    Description:
1798    
1799    Type of top-level "app" corrected.
1800    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1801    
1802    ----------------------------------------------------------------------
1803    Name: Matthias Blume (blume (at) tti - c (dot) org)
1804    Date: 2003/11/18 17:10 CST
1805    Tag: blume-20031118-basis-fiddle
1806    Description:
1807    
1808    - changed Timer interface to what might become the spec
1809    - POSIX_FLAGS -> BIT_FLAGS according to spec
1810    - some other minor discrepancies wrt. spec eliminated
1811    
1812    ----------------------------------------------------------------------
1813    Name: Matthias Blume (blume (at) tti - c (dot) org)
1814    Date: 2003/11/06 12:00:00 CST
1815    Tag: Release_110_44
1816    Description:
1817    
1818    New working version (110.44). New bootfiles.
1819    
1820    ----------------------------------------------------------------------
1821    Name: Matthias Blume (blume (at) tti - c (dot) org)
1822    Date: 2003/11/04 11:50:00 CST
1823    Tag: blume-20031104-move-libraries
1824    Description:
1825    
1826    Eliminated the "dont_move_libraries" directive in config/targets.
1827    (The mechanism was broken and could not be fixed easily.  Moreover,
1828    there does not seem to be any reason not to move all libraries into
1829    lib during installation.  I originally implemented this directive as a
1830    backward-compatibility feature when I first introduced the new CM.
1831    Now that things have been stable for a long time and going back to the
1832    old CM is not an option, there is no reason to keep it around.)
1833    
1834    ----------------------------------------------------------------------
1835    Name: Matthias Blume (blume (at) tti - c (dot) org)
1836    Date: 2003/11/03 16:00:00 CST
1837    Tag: blume-20031103-installdir
1838    Description:
1839    
1840    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1841    Richards for pointing out the problem and providing the solution.)
1842    
1843    ----------------------------------------------------------------------
1844    Name: Matthias Blume (blume (at) tti - c (dot) org)
1845    Date: 2003/10/01 17:05:00 CDT
1846    Tag: blume-20031001-lal-mlrisc
1847    Description:
1848    
1849    MLRISC bug fix from Lal.
1850    
1851    ----------------------------------------------------------------------
1852    Name: Matthias Blume (blume (at) tti - c (dot) org)
1853    Date: 2003/09/30 16:10:00 CDT
1854    Tag: blume-20030930-primio-bat
1855    Description:
1856    
1857    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1858    2. Improved .bat files (for Win32 port) to make things work under Win95.
1859       (thanks to Aaron S. Hawley for this one)
1860    
1861    ----------------------------------------------------------------------
1862    Name: Matthias Blume (blume (at) tti - c (dot) org)
1863    Date: 2003/09/26 16:05:00 CDT
1864    Tag: blume-20030926-wrappriv
1865    Description:
1866    
1867    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1868    
1869    ----------------------------------------------------------------------
1870    Name: Matthias Blume (blume (at) tti - c (dot) org)
1871    Date: 2003/09/26 15:00:00 CDT
1872    Tag: blume-20030926-110_43_3
1873    Description:
1874    
1875    - additional cleanup
1876    - version number bump, NEW BOOTFILES
1877    
1878    ----------------------------------------------------------------------
1879    Name: Matthias Blume (blume (at) tti - c (dot) org)
1880    Date: 2003/09/26 12:00:00 CDT
1881    Tag: blume-20030926-ppautoload
1882    Description:
1883    
1884    I modified the read-eval-print loop so that the autoloader gets
1885    invoked whenever the prettyprinter tries to look up a symbol that
1886    is not currently defined in the toplevel environment but which
1887    appears in CM's autoload registry.  As a result, we see far fewer of
1888    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1889    
1890    In addition to this I tried to clean up some pieces of the Basis
1891    implementation (e.g., Socket, Word8Array) in order to prevent other
1892    instances of these ?.Foo.Bar.xxx names from being printed.
1893    
1894    The mechanism that picks names for types still needs some work, though.
1895    (Right now it seems that if there is a type A.t which is defined to
1896    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1897    "?.B.u" although the perhaps more sensible solution would be to use
1898    "A.t" in this case.  In other words, the prettyprinter should follow
1899    a chain of DEFtycs not farther than there are corresponding toplevel
1900    names in the current environment.)
1901    
1902    ----------------------------------------------------------------------
1903    Name: Matthias Blume (blume (at) tti - c (dot) org)
1904    Date: 2003/09/24 16:31:00 CDT
1905    Tag: blume-20030924-installer
1906    Description:
1907    
1908    Another installer tweak:  All the ML code for the installer is now
1909    compiled during CMB.make and put into a little library called
1910    $smlnj/installer.cm.  The installation then simply invokes
1911    
1912       sml -m $smlnj/installer.cm
1913    
1914    and everything happens automagically.
1915    
1916    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1917    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1918            and BINDIR.
1919    
1920    The new scheme guarantees that the ML code responsible for the installation
1921    is in sync with the APIs of the main system.  Also, the installer is
1922    somewhat faster because the installer script is precompiled.
1923    
1924    ----------------------------------------------------------------------
1925    Name: Matthias Blume (blume (at) tti - c (dot) org)
1926    Date: 2003/09/24 15:35:00 CDT
1927    Tag: blume-20030924-synsock
1928    Description:
1929    
1930    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1931    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1932    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1933    implementations of non-blocking operations from CML's Socket
1934    structure.
1935    
1936    ----------------------------------------------------------------------
1937    Name: Matthias Blume (blume (at) tti - c (dot) org)
1938    Date: 2003/09/24 15:10:05 CDT
1939    Tag: blume-20030924-sockets
1940    Description:
1941    
1942    1. Fixed SOCKET API and implementation to match Basis spec.
1943       This required changing the internal representation of sockets to one
1944       that remembers (for each socket file descriptor) whether it is currently
1945       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1946       call is made only if the state actually needs to change).
1947    
1948    2. OS-specific details of sockets were moved into separate files, thus
1949       making it possible to unify the bulk of the socket implementations
1950       between Unix and Win32.
1951    
1952    3. CML's socket API changed accordingly.
1953       (Note that we need to remove non-blocking functions from this API
1954       since they are redundant in the case of CML!)
1955    
1956    4. CML's socket implementation now makes use of non-blocking functions
1957       provided by Basis, thus removing all OS-dependent code from this part
1958       of CML.
1959    
1960    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1961    
1962    ----------------------------------------------------------------------
1963    Name: Matthias Blume (blume (at) tti - c (dot) org)
1964    Date: 2003/09/22 12:10:00 CDT
1965    Tag: blume-20030922-110_43_2
1966    Description:
1967    
1968    Made a new interim version and bootfiles for developer's bootstrapping
1969    convenience.
1970    
1971       110.43.2 -- NEW BOOTFILES
1972    
1973    ----------------------------------------------------------------------
1974    Name: Matthias Blume (blume (at) tti - c (dot) org)
1975    Date: 2003/09/19 15:55:00 CDT
1976    Tag: blume-20030919-cmdir
1977    Description:
1978    
1979    1. new-install.sh -> install.sh
1980    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1981    3. tweaked installer so that another name instead of .cm can be chosen
1982       at install time (by setting the CM_DIR_ARC environment variable
1983       during installation); once installation is complete, the name is
1984       fixed
1985    
1986    ----------------------------------------------------------------------
1987    Name: Matthias Blume (blume (at) tti - c (dot) org)
1988    Date: 2003/09/18 16:00:00 CDT
1989    Tag: blume-20030918-110_43_1
1990    Description:
1991    
1992    Made a new interim version and bootfiles for developer's bootstrapping
1993    convenience.
1994    
1995       110.43.1 -- NEW BOOTFILES
1996    
1997    ----------------------------------------------------------------------
1998    Name: Matthias Blume (blume (at) tti - c (dot) org)
1999    Date: 2003/09/18 15:20:00 CDT
2000    Tag: blume-20030918-misc
2001    Description:
2002    
2003    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
2004       this seems to be controversial at the moment) and used those in
2005       Posix.ProcEnv.times.
2006    
2007    2. Added Time.{from,to}Nanoseconds to Time.
2008    
2009    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
2010       For example, fromLargeInt never needs to look at more than 3 "big
2011       digits" to get its 53 bits of precision.
2012    
2013    ----------------------------------------------------------------------
2014    Name: Matthias Blume (blume (at) tti - c (dot) org)
2015    Date: 2003/09/17 16:30:00 CDT
2016    Tag: blume-20030917-real32-slices
2017    Description:
2018    
2019    Added an entry to the primitive environment
2020    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
2021    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
2022    
2023    Removed some of the "magic" constants in real64.sml and replaced them
2024    with code that generates these values from their corresponding
2025    integer counterparts.
2026    
2027    Made all(?) the slice-related changes to the Basis and made everything
2028    compile again...
2029    
2030    ----------------------------------------------------------------------
2031    Name: Matthias Blume (blume (at) tti - c (dot) org)
2032    Date: 2003/09/15 17:45:00 CDT
2033    Tag: blume-20030915-rbase
2034    Description:
2035    
2036    Fixed bug in Real.fromLargeInt.
2037    
2038    ----------------------------------------------------------------------
2039    Name: Matthias Blume (blume (at) tti - c (dot) org)
2040    Date: 2003/09/13 18:11:00 CDT
2041    Tag: blume-20030913-libinstall
2042    Description:
2043    
2044    Minor bugfix in config/libinstall (set anchor with path to
2045    standalone tool after installing it, otherwise libraries that
2046    need ml-lex or ml-yacc won't compile the first time the installer
2047    runs).
2048    
2049    ----------------------------------------------------------------------
2050    Name: Matthias Blume (blume (at) tti - c (dot) org)
2051    Date: 2003/09/12 11:45:00 CDT
2052    Tag: blume-20030912-various
2053    Description:
2054    
2055     - fixed bug in Real.toLargeInt
2056     - fixed bug in Posix.ProcEnv.times
2057     - changed inputLine functions to return an option
2058     - minor installer improvements / bugfixes
2059     - changed default @SMLalloc parameter for x86/celeron to 64k
2060    
2061    ----------------------------------------------------------------------
2062    Name: Matthias Blume (blume (at) tti - c (dot) org)
2063    Date: 2003/09/09 22:00:00 CDT
2064    Tag: Release_110_43
2065    Description:
2066    
2067        New working release 110.43.  New bootfiles.
2068    
2069    ----------------------------------------------------------------------
2070    Name: Matthias Blume (blume (at) tti - c (dot) org)
2071    Date: 2003/09/09 19:20:00 CDT
2072    Tag: blume-20030909-installer
2073    Description:
2074    
2075    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
2076    Modified config/install.bat to take advantage of it.  Also modified
2077    config/install.sh (and called it config/new-install.sh) to take advantage
2078    of it on Unix systems.  (The SML code is (supposed to be) platform-
2079    independent.)
2080    
2081    The installer can now install everything under Win32
2082    as well as under *nix as long as it compiles.
2083    
2084    Other changes:
2085    
2086      - made CML compile again under Win32
2087      - made eXene compile under Win32 (by providing a fake structure UnixSock
2088        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
2089      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
2090        same as type int;  under Win32 it isn't
2091      - fixed some slice-related problems in the win32-specific parts of CML
2092      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
2093        (under Win32, the volume name is case-insensitive, and the
2094        OS.Path code compares volume names for equality)
2095    
2096    ----------------------------------------------------------------------
2097    Name: Matthias Blume (blume (at) tti - c (dot) org)
2098    Date: 2003/09/08 11:55:00 CDT
2099    Tag: blume-20030908-fullpath
2100    Description:
2101    
2102    Made Win32 version of OS.FileSys.fullPath return current directory
2103    when given an empty string.  This is what the spec says, and incidentally,
2104    CM depends on it.  (CM otherwise goes into an infinite loop in certain
2105    cases when presented with the name of a non-existing .cm file.)
2106    
2107    ----------------------------------------------------------------------
2108    Name: Matthias Blume (blume (at) tti - c (dot) org)
2109    Date: 2003/09/04 16:30:00 CDT
2110    Tag: blume-20030905-slices-etc
2111    Description:
2112    
2113      1. Changed interface to vectors and arrays in Basis to match
2114         (draft) Basis spec.
2115      2. Added signatures and implementations of slices according to
2116         Basis spec.
2117      3. Edited source code throughout the system to make it compile again
2118         under 1. and 2.  (In some cases code had to be added to have it
2119         match the new signatures.)
2120      4. MLRISC should be backward-compatible: the copies of the originals
2121         of files that needed to change under 3. were retained, the .cm files
2122         check the compiler version number and use old versions when
2123         appropriate.
2124      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2125         dirstream -> string option (in accordance with Basis spec).
2126      6. When generating code that counts lines, ml-lex used function
2127         CharVector.foldli, taking advantage of its old interface.
2128         This has been replaced with the corresponding code from
2129         CharVectorSlice.  (html-lex must be re-lexed!)
2130      7. BitArray in smlnj-lib/Util has been extended/modified to match the
2131         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
2132      8. Removed temporary additions (fromInternal, toInternal) from the
2133         (now obsolete) IntInf in smlnj-lib/Util.
2134      9. Cleaned up structure Byte.
2135     10. Added localOffset, scan, and fromString to Date (according to spec).
2136         Cleaned/corrected implementation of Date.
2137         (Still need to check for correctness; implement better canonicalizeDate.)
2138     11. Added "scan" to signature IEEE_REAL.
2139     12. Some improvements to IntInf [in particular: efficiency-hack for
2140         mod and rem when second operand is 2 (for parity checks).]
2141     13. Changed representation of type Time.time, using a single IntInf.int
2142         value counting microseconds.  This considerably simplified the
2143         implementation of structure Time.  We now support negative time
2144         values; scan and fromString handle signs.
2145     14. Functor PrimIO now takes two additional arguments (VectorSlice and
2146         ArraySlice).
2147    
2148    ----------------------------------------------------------------------
2149    Name: Matthias Blume (blume (at) tti - c (dot) org)
2150    Date: 2003/08/28 17:00:00 CDT
2151    Tag: blume-20030828-intinf
2152    Description:
2153    
2154    This is a major update which comes with a version number bump
2155    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2156    and an implementation of IntInf in the Basis.
2157    
2158    There are a fairly large number of related changes and updates throughout
2159    the system:
2160    
2161    Basis:
2162            - Implemented IntInf.
2163            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
2164            - Added some missing Real64 operations, most notably Real.toLargeInt.
2165            - Added FixedInt as a synonym for Int32.
2166    
2167    compiler:
2168            * Added support for a built-in intinf type.
2169              - literals
2170              - pattern matching
2171              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2172              - overloading on literals and operations
2173    
2174            This required adding a primitive type intinf, some additional
2175            primops, and implementations for several non-trivial intinf
2176            operations in Core.  (The intinf type is completely abstract
2177            to the compiler; all operations get delegated back to the Core.)
2178    
2179            * Intinf equality is handled by polyequal.  However, the compiler
2180            does not print its usual warning in this case (since polyequal
2181            is the right thing to do there).
2182    
2183            * Improved the organization of structure InlineT.
2184    
2185            * A word about conversion primops:
2186            If conversions involving intinf do not cancel out during
2187            CPS contract, then the compiler must insert calls to Core functions.
2188            Since all core access must be resolved already during the FLINT
2189            translate phase, it would be too late a the time of CPS contract
2190            to add new Core calls.  For this reason, conversion primops
2191            for intinf carry two arguments:  1. the numeric argument that
2192            they are supposed to convert, and 2. the Core function that
2193            can help with this conversion if necessary.  If CPS contract
2194            eliminates a primop, then the associated Core function becomes
2195            dead and goes away.  Intinf conversion primops that do not get
2196            eliminated by CPS contract get rewritten into calls of their
2197            core functions by a separate, new phase.
2198    
2199    interactive system:
2200            - Control.Print.intinfDepth controls max length of intinf constants
2201            being printed. (Analogous to Control.Print.stringDepth.)
2202            - Cleanup in printutil and pputil: got rid of unused stuff and
2203            duplicates; replaced some of the code with code that makes better
2204            use of library functionality.
2205    
2206    CM:
2207            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2208            are no longer silent.
2209    
2210    CKIT:
2211            Fixed mismatched uses of Int32 and LargeInt.  I always decided
2212            in favor of LargeInt -- which is now the same as IntInf.
2213            CKIT-knowledgable people should check whether this is what's
2214            intended and otherwise change things back to using Int32 or
2215            FixedInt.
2216    
2217    Throughout the code:
2218            Started using IntInf.int literals and built-in operations
2219            (e.g., comparison with 0) where this seems appropriate.
2220    
2221    
2222    ----------------------------------------------------------------------
2223    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2224    Date: 2003/08/13 11:36:00 CDT
2225    Tag: dbm-20030813-mcz-merge1
2226    Description:
2227    
2228    Merging changes from the mcz-branch development branch into trunk.
2229    These changes involve replacement of the emulated old prettyprinter
2230    interface with direct use of the SML/NJ Lib PP library, and fixing
2231    of a couple of bugs (895, 1186) relating to error messages.  A new
2232    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2233    has been added.
2234    
2235    ----------------------------------------------------------------------
2236    Name: Matthias Blume (blume (at) tti - c (dot) org)
2237    Date: 2003/08/11 15:45:00 CDT
2238    Tag: blume-20030811-windows
2239    Description:
2240    
2241    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2242    
2243        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2244    
2245    This patch restores SML/NJ's ability to run under win32.  There are a
2246    number of changes, including fixes for several bugs that had gone
2247    unnoticed until now:
2248    
2249      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2250        trying to run the win32 version from within cygwin.)
2251    
2252      - There are a number of simple .bat scripts that substitute for their
2253        corresponding Unix shell-scripts.  (See below.)
2254    
2255      - The internals of ml-build have been modified slightly.  The main
2256        difference is that instead of calling ".link-sml" (or link-sml.bat)
2257        using OS.Process.system, the ML process delegates this task back
2258        to the script.  Otherwise problems arise in mixed environments such
2259        as Cygwin where scripts look and work like Unix scripts, but
2260        where OS.Process.system cannot run them.
2261    
2262      - In CM, the srcpath pickler used native pathname syntax -- which
2263        is incorrect in the case of cross-compilation.  The new pickle format
2264        is independent of platform-specific naming conventions.
2265    
2266      - Path configuration files (such as lib/pathconfig) can now choose
2267        between native and standard syntax.  Placing a line of the form
2268    
2269            standard!
2270    
2271        into the file causes all subsequent paths to be interpreted using
2272        CM standard pathname syntax (= Unix conventions); a line
2273    
2274            native!
2275    
2276        switches back to native style.  This was needed so that
2277        path config files can be written portably, see src/system/pathconfig.
2278    
2279      - Runtime system:
2280    
2281          - win32-filesys.c:  get_file_time and set_file_time now
2282            access modification time, not creation time.
2283    
2284          - I/O code made aware of new array representation.
2285    
2286          - Bug fixes in X86.prim.masm.
2287    
2288       - src/system/makeml made aware of win32. (For use under cygwin
2289            and other Unix-environments for windows.)
2290    
2291       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2292         which caused BinIO.inputAll to fail consistently.
2293    
2294    .bat scripts:
2295    
2296       Windows .bat scripts assume that SMLNJ_HOME is defined.
2297    
2298       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2299         applications (sml, ml-yacc, ml-lex).
2300       - ml-build.bat: analogous to ml-build.
2301       - config\install.bat:  Analogous to config/install.sh.  This requires
2302         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2303         (nmake etc. must be on the path, and vcvars32 must have been run.)
2304         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2305         and the bootfile hierarchy must have been unpacked under
2306         sml.boot.x86-win32.
2307         The script is very primitive and does a poor job at error checking.
2308         It only installs the base system, ml-lex, and ml-yacc.  No other
2309         libraries are being installed (i.e., you get only those that
2310         are part of the compiler.)
2311       - link-sml.bat:  analogous to .link-sml, but not currently used
2312    
2313    Unrelated bug fixes:
2314    
2315      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2316        types.
2317      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2318        otherwise.)
2319    
2320    ----------------------------------------------------------------------
2321    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2322    Date: 2003/06/17
2323    Tag: macqueen-20030617-bug895
2324    Description:
2325    
2326    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2327    Tag will be used for new development branch (mcz-branch) for
2328    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2329    
2330    ----------------------------------------------------------------------
2331    Name: Matthias Blume (blume (at) tti - c (dot) org)
2332    Date: 2003/05/27 16:55:00 CDT
2333    Tag: blume-20030527-polyeq
2334    Description:
2335    
2336    Tried to eliminated most cases of polymorphic equality.
2337    
2338    ----------------------------------------------------------------------
2339    Name: Matthias Blume (blume (at) tti - c (dot) org)
2340    Date: 2003/05/21 17:45:00 CDT
2341    Tag: blume-20030517-complete
2342    Description:
2343    
2344    Two changes:
2345    
2346      1. Added a flag for controlling whether non-exhaustive bindings will
2347         be treated as errors (default is false).
2348      2. Cleaned up the *entire* source tree so that CMB.make goes through
2349         without a single non-exhaustive match- or bind warning.
2350    
2351    ----------------------------------------------------------------------
2352    Name: Matthias Blume (blume (at) tti - c (dot) org)
2353    Date: 2003/05/17 10:20:00 CDT
2354    Tag: blume-20030517-absyn
2355    Description:
2356    
2357    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2358    
2359       This mainly affects the quality of error messages.  However, some
2360       of the code is now more straightforward than before. (Treatment of
2361       the above four constructs in translate.sml is much simpler than
2362       the "macro-expansion" that was going on before.  Plus, the mach-
2363       compiler no longer gets invoked just to be able to compile an
2364       if-expression.)
2365    
2366    2. The ErrorMsg.Error exception is now caught and absorbed by the
2367       interactive loop.
2368    
2369    ----------------------------------------------------------------------
2370    Name: Allen Leung
2371    Date: 2003/05/16 13:05:00 CDT
2372    Tag: leunga-20030516-cygwin-runtime
2373    Description:
2374    
2375        Ported the runtime system to cygwin, which uses the unix
2376    x86-unix bin files.  Missing/buggy features:
2377    
2378       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2379         the Cygwin library.
2380       o Ctrl-C handling may be flaky.
2381       o Windows system calls and Windows I/O are not supported.
2382    
2383       A new set of binfiles is located at:
2384    
2385         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2386    
2387       This is only needed for bootstrapping the cygwin version of smlnj.
2388       Other x86 versions can use the existing binfiles.
2389    
2390    ----------------------------------------------------------------------
2391    Name: Matthias Blume
2392    Date: 2003/04/08 15:42:00 CDT
2393    Tag: blume-20030408-listpair
2394    Description:
2395    
2396    1. Added a target 'mlrisc' to installer.
2397    
2398    2. Added missing elements to structure ListPair.
2399    
2400    ----------------------------------------------------------------------
2401    Name: Allen Leung
2402    Date: 2003/01/07 10:40:00 EST
2403    Tag: leunga-20030107-int-rem
2404    Description:
2405    
2406    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2407    The arguments to the SUBL instruction were swapped.
2408    
2409    ----------------------------------------------------------------------
2410    Name: Matthias Blume
2411    Date: 2002/12/12 16:25:00 EST
2412    Tag: blume-20021212-risc-ra
2413    Description:
2414    
2415    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2416    sent the RA into an infinite loop when floating point registers get
2417    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2418    110.37 and 110.38.)
2419    There was another set of potential problems related to the handling of
2420    MLRISC annotations (but those did not yet cause real problems, apparently).
2421    
2422    ----------------------------------------------------------------------
2423    Name: Matthias Blume
2424    Date: 2002/12/06 22:40:00 EST
2425    Tag: blume-20021206-cm-fileid
2426    Description:
2427    
2428    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2429    This fixes the problem of CM getting confused by files that suddenly
2430    change their identity (e.g., by getting unlinked and recreated by some
2431    text editor such as vi).  There might be a better/cheaper/cleaner way
2432    of doing this, but for now this will have to do.
2433    
2434    ----------------------------------------------------------------------
2435    Name: Matthias Blume
2436    Date: 2002/10/28 09:50:00 EST
2437    Tag: blume-20021028-typecheck
2438    Description:
2439    
2440    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2441    
2442    ----------------------------------------------------------------------
2443    Name: Matthias Blume
2444    Date: 2002/10/17 09:10:00 EDT
2445    Tag: Release_110_42
2446    Description:
2447    
2448    In good old tradition, there has been a slight hiccup so that we have
2449    to patch 110.42 after the fact.   The old release tag has been replaced
2450    (see below).
2451    
2452    The change solves a problem with two competing approaches the
2453    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2454    each other's way.
2455    
2456    This change only affects the runtime system code and the installer script.
2457    (No new bootfiles.)
2458    
2459    ----------------------------------------------------------------------
2460    Name: Matthias Blume
2461    Date: 2002/10/16 12:00:00 EDT
2462    Tag: Release_110_42_removed
2463    Description:
2464    
2465    New working release.  New bootfiles.
2466    
2467    ----------------------------------------------------------------------
2468    Name: Matthias Blume
2469    Date: 2002/10/10 13:10:00 EDT
2470    Tag: blume-20021010-ppc-divs
2471    Description:
2472    
2473    The mltree operator DIVS must be implemented with an overflow check on
2474    the PPC because the hardware indicates divide-by-zero using "overflow" as
2475    well.
2476    
2477    ----------------------------------------------------------------------
2478    Name: Matthias Blume
2479    Date: 2002/07/23 11:45:00
2480    Tag: blume-20020723-smlnj-home
2481    Description:
2482    
2483    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2484    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2485    CM_PATHCONFIG is also set), the path configuration file is assumed
2486    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2487    the entire tree to some other place and everything will "just work".
2488    
2489    (Companion commands such as ml-build and ml-makedepend also sense this
2490    variable.)
2491    
2492    ----------------------------------------------------------------------
2493    Name: Matthias Blume
2494    Date: 2002/07/12 21:19:00 EDT
2495    Tag: blume-20020712-liveness
2496    Description:
2497    
2498    Exported two useful "step" functions from liveness module (MLRISC).
2499    
2500    ----------------------------------------------------------------------
2501    Name: Matthias Blume
2502    Date: 2002/07/05 16:00 EDT
2503    Tag: Release_110_41
2504    Description:
2505    
2506    New working release.  New bootfiles.
2507    
2508    ----------------------------------------------------------------------
2509    Name: Matthias Blume
2510    Date: 2002/07/05 10:25:00 EDT
2511    Tag: blume-20020705-btimp
2512    Description:
2513    
2514    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2515    other clients can set up backtracing support.
2516    
2517    ----------------------------------------------------------------------
2518    Name: Matthias Blume
2519    Date: 2002/06/25 17:23:00 EDT
2520    Tag: blume-20020625-fpmax
2521    Description:
2522    
2523    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2524    case.  (The sense of the isNaN test was reversed -- which made min and
2525    max always return their first argument.)
2526    
2527    ----------------------------------------------------------------------
2528    Name: Matthias Blume
2529    Date: 2002/06/11
2530    Tag: blume-20020611-unixpath
2531    Description:
2532    
2533    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2534    
2535    ----------------------------------------------------------------------
2536    Name: Matthias Blume
2537    Date: 2002/06/10 16:35:00 EDT
2538    Tag: blume-20020610-ieeereal
2539    Description:
2540    
2541    I back-ported my implementation of IEEEReal.fromString from the
2542    idlbasis-devel branch so that we can test it.
2543    
2544    Another small change is that ppDec tries to give more information
2545    than just "<sig>" in the case of functors.  However, this code is
2546    broken in some mysterious way if the functor's body's signature
2547    has not been declared by ascription but gets inferred from the
2548    implementation.  This needs fixing...
2549    
2550    ----------------------------------------------------------------------
2551    Name: Matthias Blume
2552    Date: 2002/05/31
2553    Tag: blume-20020531-btrace-mode
2554    Description:
2555    
2556    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2557    the wayside when I switched over to using Controls everywhere.)
2558    
2559    ----------------------------------------------------------------------
2560    Name: Lal George
2561    Date: 2002/05/23 12:21:40 EDT
2562    Tag: george-20020523-visual-labels
2563    Description:
2564    
2565    Labels are now displayed in the graphical output to make
2566    the fall-through and target blocks obvious.
2567    
2568    ----------------------------------------------------------------------
2569    Name: Matthias Blume
2570    Date: 2002/05/22 11:03:00 EDT
2571    Tag: blume-20020522-shrink
2572    Description:
2573    
2574    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2575    situation that comes up when allocating huge arrays.
2576    
2577    ----------------------------------------------------------------------
2578    Name: Matthias Blume
2579    Date: 2002/05/21 16:00:00 EDT
2580    Tag: Release_110_40
2581    Description:
2582    
2583    New working release (110.40).  New bootfiles.
2584    
2585    [Also: John Reppy fixed GC bug 1131.]
2586    
2587    ----------------------------------------------------------------------
2588    Name: Matthias Blume
2589    Date: 2002/05/21 12:35:00 EDT
2590    Tag: blume-20020521-cmdoc
2591    Description:
2592    
2593    CM documentation update.
2594    
2595    ----------------------------------------------------------------------
2596    Name: Matthias Blume
2597    Date: 2002/05/21 10:55:00 EDT
2598    Tag: blume-20020521-misc
2599    Description:
2600    
2601    - John tweaked runtime to be silent on heap export (except when
2602      GC messages are on).
2603    - I added a few more things (cross-compiling versions of CMB) to
2604      config/preloads (as suggestions).
2605    
2606    ----------------------------------------------------------------------
2607    Name: Matthias Blume
2608    Date: 2002/05/20 22:25:00 EDT
2609    Tag: blume-20020520-controls
2610    Description:
2611    
2612    - Added ControlUtil structure to control-lib.cm.
2613    - Use it throughout.
2614    - Used Controls facility to define MLRISC controls (as opposed to
2615      registering MLRISC control ref cells with Controls after the fact)
2616    - Fixed messed-up controls priorities.
2617    
2618    * Removed again all the stuff from config/preloads that one wouldn't
2619      be able to preload at the time the initial heap image is built.
2620      (Many libraries, e.g., CML, do not exist yet at this time.  The
2621       only libraries that can be preloaded via config/preloads are those
2622       that come bundled with the bootfiles.)
2623    
2624    ----------------------------------------------------------------------
2625    Name: Matthias Blume
2626    Date: 2002/05/20 10:59:00 EDT
2627    Tag: blume-20020520-preloads
2628    Description:
2629    
2630    Added a lot of commented-out suggestions for things to be included
2631    in config/preloads.
2632    
2633    ----------------------------------------------------------------------
2634    Name: Allen Leung
2635    Date: 2002/05/18 14:20:00 EDT
2636    Tag: leunga-20020518-mdl
2637    Description:
2638    
2639    o Made the mdl tool stuff compile and run again.
2640    o I've disabled all the stuff that depends on RTL specifications; they
2641      are all badly broken anyway.
2642    
2643    ----------------------------------------------------------------------
2644    Name: Matthias Blume
2645    Date: 2002/05/17 16:49:00 EDT
2646    Tag: blume-20020517-controls
2647    Description:
2648    
2649    0. John Reppy made several modifications to the SML/NJ library.
2650       In particular, there is a shiny new controls-lib.cm.
2651    
2652    1. Pushed new controls interface through compiler so that everything
2653       compiles again.
2654    
2655    2. Added FormatComb and FORMAT_COMB to the CML version of the
2656       SML/NJ library (so that CML compiles again).
2657    
2658    3. Modified init scripts because XXX_DEFAULT environment variables
2659       are no longer with us.  (Boot-time initialization is now done
2660       using the same environment variables that are also used for
2661       startup-time initialization of controls.)
2662    
2663    ----------------------------------------------------------------------
2664    Name: Lal George
2665    Date: 2002/05/15 09:20:10 EDT
2666    Tag: george-20020515-pseudo-op-decls
2667    Description:
2668    
2669            All pseudo-ops emitted before the first segment declaration
2670    such as TEXT, DATA, and BSS directives are assumed to be global
2671    declarations and are emitted first in the assembly file. This is
2672    useful in a number of situations where one has pseudo-ops that are not
2673    specific to any segment, and also works around the constraint that one
2674    cannot have client pseudo-ops in the TEXT segment.
2675    
2676            Because no segment is associated with these declarations it is
2677    an error to allocate any space or objects before the first segment
2678    directive and an exception will be raised. However, we cannot make
2679    this check for client pseudo-ops.
2680    
2681            These top level declarations are a field in the CFG graph_info.
2682    In theory you can continue to add to this field after the CFG has been
2683    built -- provided you know what you are doing;-)
2684    
2685    ----------------------------------------------------------------------
2686    Name: Matthias Blume
2687    Date: 2002/05/13 16:40:00 EDT
2688    Tag: blume-20020513-pp-etc
2689    Description:
2690    
2691    A few minor bugfixes:
2692    
2693      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2694        (Bogus printouts for redefined bindings still occur.  Compiler
2695        bug should no longer occur now.  We need to redo the prettyprinter
2696        from scratch.)
2697    
2698      - CM pathname printer now also adds escape sequences for ( and )
2699    
2700      - commend and docu fixes for ml-nlffi
2701    
2702    ----------------------------------------------------------------------
2703    Name: Matthias Blume
2704    Date: 2002/05/10 16:40:00 EDT
2705    Tag: blume-20020510-erg-textio
2706    Description:
2707    
2708    Applied the following bugfix provided by Emden Gansner:
2709    
2710        Output is corrupted when outputSubstr is used rather than output.
2711    
2712        The problem occurs when a substring
2713    
2714            ss = (s, dataStart, dataLen)
2715    
2716        where dataStart > 0, fills a stream buffer with avail bytes left.
2717        avail bytes of s, starting at index dataStart, are copied into the
2718        buffer, the buffer is flushed, and then the remaining dataLen-avail
2719        bytes of ss are copied into the beginning of the buffer. Instead of
2720        starting this copy at index dataStart+avail in s, the current code
2721        starts the copy at index avail.
2722    
2723        Fix:
2724        In text-io-fn.sml, change line 695 from
2725             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2726        to
2727             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2728    
2729    ----------------------------------------------------------------------
2730    Name: Matthias Blume
2731    Date: 2002/04/12 13:55:00 EDT
2732    Tag: blume-20020412-assyntax
2733    Description:
2734    
2735    1. Grabbed newer assyntax.h from the XFree86 project.
2736    2. Fiddled with how to compile X86.prim.asm without warnings.
2737    3. (Very) Minor cleanup in CM.
2738    
2739    ----------------------------------------------------------------------
2740    Name: Matthias Blume
2741    Date: 2002/04/01 (no joke!) 17:07:00 EST
2742    Tag: blume-20020401-x86div
2743    Description:
2744    
2745    Added full support for div/mod/rem/quot on the x86, using the machine
2746    instruction's two results (without clumsily recomputing the remainder)
2747    directly where appropriate.
2748    
2749    Some more extensive power-of-two support was added to the x86 instruction
2750    selector (avoiding expensive divs, mods, and muls where they can be
2751    replaced with cheaper shifts and masks).  However, this sort of thing
2752    ought to be done earlier, e.g., within the CPS optimizer so that
2753    all architectures benefit from it.
2754    
2755    The compiler compiles to a fixed point, but changes might be somewhat
2756    fragile nevertheless.  Please, report any strange things that you might
2757    see wrt. div/mod/quot/rem...
2758    
2759    ----------------------------------------------------------------------
2760    Name: Matthias Blume
2761    Date: 2002/03/29 17:22:00
2762    Tag: blume-20020329-div
2763    Description:
2764    
2765    Fixed my broken div/mod logic.  Unfortunately, this means that the
2766    inline code for div/mod now has one more comparison than before.
2767    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2768    The problem was with quotient = 0, because that alone does not tell
2769    us which way the rounding went.  One then has to look at whether
2770    remainder and divisor have the same sign...  :(
2771    
2772    Anyway, I replaced the bootfiles with fresh ones...
2773    
2774    ----------------------------------------------------------------------
2775    Name: Matthias Blume
2776    Date: 2002/03/29 14:10:00 EST
2777    Tag: blume-20020329-inlprims
2778    Description:
2779    
2780    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2781    
2782    Primops have changed. This means that the bin/boot-file formats have
2783    changed as well.
2784    
2785    To make sure that there is no confusion, I made a new version.
2786    
2787    
2788    CHANGES:
2789    
2790    * removed REMT from mltree (remainder should never overflow).
2791    
2792    * added primops to deal with divisions of all flavors to the frontend
2793    
2794    * handled these primops all the way through so they map to their respective
2795      MLRISC support
2796    
2797    * used these primops in the implementation of Int, Int32, Word, Word32
2798    
2799    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2800    
2801    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2802    
2803    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2804      floating point
2805    
2806    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2807      and Real (but Real.abs maps to a separate floating-point-only primop)
2808    
2809    
2810    TODO items:
2811    
2812    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2813      instructions because the machine instruction encoder cannot handle
2814      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2815      REM in the first place, and REM is supported by the x86 machine coder.)
2816    
2817    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2818      selection phase.  (The two can be streamlined because the hardware
2819      delivers both quotient and remainder at the same time anyway.)
2820    
2821    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2822      (Currently the behavior is inconsistent both across architectures and
2823      wrt. the draft Basis spec.)
2824    
2825    * Word8 should eventually be handled natively, too.
2826    
2827    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2828      as if there currently is no execution path that could trigger it in
2829      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2830      hold for things like multiplication and division.)
2831    
2832    ----------------------------------------------------------------------
2833    Name: Matthias Blume
2834    Date: 2002/03/27 16:27:00 EST
2835    Tag: blume-20020327-mlrisc-divisions
2836    Description:
2837    
2838    Added support for all four division operations (ML's div, mod, quot,
2839    and rem) to MLRISC.  In the course of doing so, I also rationalized
2840    the naming (no more annoying switch-around of DIV and QUOT), by
2841    parameterizing the operation by div_rounding_mode (which can be either
2842    DIV_TO_ZERO or DIV_TO_NEGINF).
2843    
2844    The generic MLTreeGen functor takes care of compiling all four
2845    operations down to only round-to-zero div.
2846    
2847    Missing pieces:
2848    
2849      * Doing something smarter than relying on MLTreeGen on architectures
2850        like, e.g., the x86 where hardware division delivers both quotient and
2851        remainder at the same time.  With this, the implementation of the
2852        round-to-neginf operations could be further streamlined.
2853    
2854      * Remove inlining support for div/mod/rem from the frontend and replace it
2855        with primops that get carried through to the backend.  Do this for all
2856        int and word types.
2857    
2858    ----------------------------------------------------------------------
2859    Name: Matthias Blume
2860    Date: 2002/03/25 17:25:00 EST
2861    Tag: blume-20020325-divmod
2862    Description:
2863    
2864    I improved (hopefully without breaking them) the implementation of Int.div,
2865    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2866    advantage of the following observations:
2867    
2868      Let  q = x quot y      r = x rem y
2869           d = x div  y      m = x mod y
2870    
2871    where "quot" is the round-to-zero version of integer division that
2872    hardware usually provides.  Then we have:
2873    
2874         r = x - q * y        where neither the * nor the - will overflow
2875         d = if q >= 0 orelse x = q * y then q else q - 1
2876                              where neither the * nor the - will overflow
2877         m = if q >= 0 orelse r = 0 then r else r + y
2878                              where the + will not overflow
2879    
2880    This results in substantial simplification of the generated code.
2881    The following table shows the number of CFG nodes and edges generated
2882    for
2883            fun f (x, y) = x OPER y
2884            (* with OPER \in div, mod, quot, rem *)
2885    
2886    
2887        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2888        --------------------------------------------------------
2889         div |         24 |         39 |         12 |         16
2890         mod |         41 |         71 |         12 |         16
2891        quot |          8 |         10 |          8 |         10
2892         rem |         10 |         14 |          8 |         10
2893    
2894    
2895    ----------------------------------------------------------------------
2896    Name: Matthias Blume
2897    Date: 2002/03/25 22:06:00 EST
2898    Tag: blume-20020325-cprotobug
2899    Description:
2900    
2901    Fixed a bug in cproto (c prototype decoder).
2902    
2903    ----------------------------------------------------------------------
2904    Name: Matthias Blume
2905    Date: 2002/03/25 16:00:00 EST
2906    Tag: blume-20020325-raw-primops
2907    Description:
2908    
2909    I did some cleanup to Allen's new primop code and
2910    replaced yesterday's bootfiles with new ones.
2911    (But they are stored in the same place.)
2912    
2913    ----------------------------------------------------------------------
2914    Name: Matthias Blume
2915    Date: 2002/03/24 22:40:00 EST
2916    Tag: blume-20020324-bootfiles
2917    Description:
2918    
2919    Made the bootfiles that Allen asked for.
2920    
2921    ----------------------------------------------------------------------
2922    Name: Allen Leung
2923    Date: 2002/03/23 15:50:00 EST
2924    Tag: leunga-20020323-flint-cps-rcc-primops
2925    Description:
2926    
2927      1. Changes to FLINT primops:
2928    
2929        (* make a call to a C-function;
2930         * The primop carries C function prototype information and specifies
2931         * which of its (ML-) arguments are floating point. C prototype
2932         * information is for use by the backend, ML information is for
2933         * use by the CPS converter. *)
2934      | RAW_CCALL of { c_proto: CTypes.c_proto,
2935                       ml_args: ccall_type list,
2936                       ml_res_opt: ccall_type option,
2937                       reentrant : bool
2938                     } option
2939       (* Allocate uninitialized storage on the heap.
2940        * The record is meant to hold short-lived C objects, i.e., they
2941        * are not ML pointers.  With the tag, the representation is
2942        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2943        *)
2944      | RAW_RECORD of {tag:bool,sz:int}
2945      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2946    
2947      2.  These CPS primops are now overloaded:
2948    
2949           rawload of {kind:numkind}
2950           rawstore of {kind:numkind}
2951    
2952          The one argument form is:
2953    
2954             rawload {kind} address
2955    
2956          The two argument form is:
2957    
2958             rawload {kind} [ml object, byte-offset]
2959    
2960      3. RAW_CCALL/RCC now takes two extra arguments:
2961    
2962         a. The first is whether the C call is reentrant, i.e., whether
2963            ML state should be saved and restored.
2964         b. The second argument is a string argument specifying the name of
2965            library and the C function.
2966    
2967         These things are currently not handled in the code generator, yet.
2968    
2969      4. In CProto,
2970    
2971         An encoding type of "bool" means "ml object" and is mapped into
2972         C prototype of PTR.  Note that "bool" is different than "string",
2973         even though "string" is also mapped into PTR, because "bool"
2974         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2975    
2976      5. Pickler/unpicker
2977    
2978         Changed to handle RAW_RECORD and newest RAW_CCALL
2979    
2980      6. MLRiscGen,
2981    
2982         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2983         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2984            in the file CodeGen/cpscompile/cps-c-calls.sml
2985    
2986      7. Added the conditional move operator
2987    
2988             condmove of branch
2989    
2990         to cps.  Generation of this is still buggy so it is currently
2991         disabled.
2992    
2993    ----------------------------------------------------------------------
2994    Name: Lal George
2995    Date: 2002/03/22 14:18:25 EST
2996    Tag: george-20020322-cps-branch-prob
2997    Description:
2998    
2999    Implemented the Ball-Larus branch prediction-heuristics, and
3000    incorporated graphical viewers for control flow graphs.
3001    
3002    Ball-Larus Heuristics:
3003    ---------------------
3004    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
3005    
3006    By design it uses the Dempster-Shafer theory for combining
3007    probabilities.  For example, in the function:
3008    
3009        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
3010    
3011    the ball-larus heuristics predicts that the n=0 is unlikely
3012    (OH-heuristic), and the 'then' branch is unlikely because of the
3013    RH-heuristic -- giving the 'then' branch an even lower combined
3014    probability using the Dempster-Shafer theory.
3015    
3016    Finally, John Reppy's loop analysis in MLRISC, further lowers the
3017    probability of the 'then' branch because of the loop in the else
3018    branch.
3019    
3020    
3021    Graphical Viewing:
3022    ------------------
3023    I merely plugged in Allen's graphical viewers into the compiler. The
3024    additional code is not much. At the top level, saying:
3025    
3026            Control.MLRISC.getFlag "cfg-graphical-view" := true;
3027    
3028    will display the graphical view of the control flow graph just before
3029    back-patching.  daVinci must be in your path for this to work. If
3030    daVinci is not available, then the default viewer can be changed
3031    using:
3032    
3033            Control.MLRISC.getString "viewer"
3034    
3035    which can be set to "dot" or "vcg" for the corresponding viewers. Of
3036    course, these viewers must be in your path.
3037    
3038    The above will display the compilation unit at the level of clusters,
3039    many of which are small, boring, and un-interesting. Also setting:
3040    
3041            Control.MLRISC.getInt "cfg-graphical-view_size"
3042    
3043    will display clusters that are larger than the value set by the above.
3044    
3045    
3046    ----------------------------------------------------------------------
3047    Name: Matthias Blume
3048    Date: 2002/03/21 22:20:00 EST
3049    Tag: blume-20020321-kmp-bugfix
3050    Description:
3051    
3052    Changed the interface to the KMP routine in PreString and fixed
3053    a minor bug in one place where it was used.
3054    
3055    ----------------------------------------------------------------------
3056    Name: Allen Leung
3057    Date: 2002/03/21 20:30:00 EST
3058    Tag: leunga-20020321-cfg
3059    Description:
3060    
3061      Fixed a potential problem in cfg edge splitting.
3062    
3063    ----------------------------------------------------------------------
3064    Name: Allen Leung
3065    Date: 2002/03/21 17:15:00 EST
3066    Tag: leunga-20020321-x86-fp-cfg
3067    Description:
3068    
3069      1. Recoded the buggy parts of x86-fp.
3070    
3071         a. All the block reordering code has been removed.
3072            We now depend on the block placement phases to do this work.
3073    
3074         b. Critical edge splitting code has been simplified and moved into the
3075            CFG modules, as where they belong.
3076    
3077         Both of these were quite buggy and complex.  The code is now much, much
3078         simpler.
3079    
3080      2. X86 backend.
3081    
3082         a. Added instructions for 64-bit support.  Instruction selection for
3083            64-bit has not been committed, however, since that
3084            requires changes to MLTREE which haven't been approved by
3085            Lal and John.
3086    
3087         b. Added support for FUCOMI and FUCOMIP when generating code for
3088            PentiumPro and above.  We only generate these instructions in
3089            the fast-fp mode.
3090    
3091         c. Added cases for JP and JNP in X86FreqProps.
3092    
3093      3. CFG
3094    
3095         CFG now has a bunch of methods for edge splitting and merging.
3096    
3097      4. Machine description.
3098    
3099         John's simplification of MLTREE_BASIS.fcond broke a few machine
3100         description things:
3101    
3102         rtl-build.{sig,sml} and hppa.mdl fixed.
3103    
3104         NOTE: the machine description stuff in the repository is still broken.
3105               Again, I can't put my fixes in because that involves
3106               changes to MLTREE.
3107    
3108    ----------------------------------------------------------------------
3109    Name: Matthias Blume
3110    Date: 2002/03/20 15:55:00 EST
3111    Tag: blume-20020320-kmp
3112    Description:
3113    
3114    Implemented Knuth-Morris-Pratt string matching in PreString and used
3115    it for String.isSubstring, Substring.isSubstring, and
3116    Substring.position.
3117    
3118    (Might need some stress-testing.  Simple examples worked fine.)
3119    
3120    ----------------------------------------------------------------------
3121    Name: Matthias Blume
3122    Date: 2002/03/19 16:37:00 EST
3123    Tag: blume-20020319-witnesses
3124    Description:
3125    
3126    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
3127    
3128    This implements a generic mechanism for changing constness qualifiers
3129    anywhere within big C types without resorting to outright "casts".
3130    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
3131    modify the constness at the outermost level.)
3132    The implementation of "convert" is based on the idea of "witness"
3133    values -- values that are not used by the operation but whose types
3134    "testify" to their applicability.  On the implementation side, "convert"
3135    is simply a projection (returning its second curried argument).  With
3136    cross-module inlining, it should not result in any machine code being
3137    generated.
3138    
3139    ----------------------------------------------------------------------
3140    Name: Matthias Blume
3141    Date: 2002/03/15 16:40:00 EST
3142    Tag: blume-20020315-basis
3143    Description:
3144    
3145    Provided (preliminary?) implementations for
3146    
3147      {String,Substring}.{concatWith,isSuffix,isSubstring}
3148    
3149    and
3150    
3151      Substring.full
3152    
3153    Those are in the Basis spec but they were missing in SML/NJ.
3154    
3155    ----------------------------------------------------------------------
3156    Name: Matthias Blume
3157    Date: 2002/03/14 21:30:00 EST
3158    Tag: blume-20020314-controls
3159    Description:
3160    
3161    Controls:
3162    ---------
3163    
3164    1. Factored out the recently-added Controls : CONTROLS stuff and put
3165       it into its own library $/controls-lib.cm.  The source tree for
3166       this is under src/smlnj-lib/Controls.
3167    
3168    2. Changed the names of types and functions in this interface, so they
3169       make a bit more "sense":
3170    
3171          module -> registry
3172          'a registry -> 'a group
3173    
3174    3. The interface now deals in ref cells only.  The getter/setter interface
3175       is (mostly) gone.
3176    
3177    4. Added a function that lets one register an already-existing ref cell.
3178    
3179    5. Made the corresponding modifications to the rest of the code so that
3180       everything compiles again.
3181    
3182    6. Changed the implementation of Controls.MLRISC back to something closer
3183       to the original.  In particular, this module (and therefore MLRISC)
3184       does not depend on Controls.  There now is some link-time code in
3185       int-sys.sml that registers the MLRISC controls with the Controls
3186       module.
3187    
3188    CM:
3189    ---
3190    
3191      * One can now specify the lambda-split aggressiveness in init.cmi.
3192    
3193    ----------------------------------------------------------------------
3194    Name: Allen Leung
3195    Date: 2002/03/13 17:30:00 EST
3196    Tag: leunga-20020313-x86-fp-unary
3197    Description:
3198    
3199    Bug fix for:
3200    
3201    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3202    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3203    > - fun f(x,(y,z)) = Real.~ y;
3204    > [autoloading]
3205    > [autoloading done]
3206    >       fchsl   (%eax), 184(%esp)
3207    > Error: MLRisc bug: X86MCEmitter.emitInstr
3208    >
3209    > uncaught exception Error
3210    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3211    
3212    The problem was that the code generator did not generate any fp registers
3213    in this case, and the ra didn't know that it needed to run the X86FP phase to
3214    translate the pseudo fp instruction.   This only happened with unary fp
3215    operators in certain situations.
3216    
3217    ----------------------------------------------------------------------
3218    Name: Matthias Blume
3219    Date: 2002/03/13 14:00:00 EST
3220    Tag: blume-20020313-overload-etc
3221    Description:
3222    
3223    1. Added _overload as a synonym for overload for backward compatibility.
3224       (Control.overloadKW must be true for either version to be accepted.)
3225    
3226    2. Fixed bug in install script that caused more things to be installed
3227       than what was requested in config/targets.
3228    
3229    3. Made CM aware of the (_)overload construct so that autoloading
3230       works.
3231    
3232    ----------------------------------------------------------------------
3233    Name: Matthias Blume
3234    Date: 2002/03/12 22:03:00 EST
3235    Tag: blume-20020312-url
3236    Description:
3237    
3238    Forgot to update BOOT and srcarchiveurl.
3239    
3240    ----------------------------------------------------------------------
3241    Name: Matthias Blume
3242    Date: 2002/03/12 17:30:00 EST
3243    Tag: blume-20020312-version110392
3244    Description:
3245    
3246    Yet another version number bump (because of small changes to the
3247    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3248    
3249    Changes:
3250    
3251      The new pid generation scheme described a few weeks ago was overly
3252      complicated.  I implemented a new mechanism that is simpler and
3253      provides a bit more "stability":  Once CM has seen a compilation
3254      unit, it keeps its identity constant (as long as you do not delete
3255      those crucial CM/GUID/* files).  This means that when you change
3256      an interface, compile, then go back to the old interface, and
3257      compile again, you arrive at the original pid.
3258    
3259      There now also is a mechanism that instructs CM to use the plain
3260      environment hash as a module's pid (effectively making its GUID
3261      the empty string).  For this, "noguid" must be specified as an
3262      option to the .sml file in question within its .cm file.
3263      This is most useful for code that is being generated by tools such
3264      as ml-nlffigen (because during development programmers tend to
3265      erase the tool's entire output directory tree including CM's cached
3266      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3267      revert to the old, broken behavior of SML/NJ, but in specific cases
3268      where there is no danger of interface confusion, its use is ok
3269      (I think).
3270    
3271      ml-nlffigen by default generates "noguid" annotations.  They can be
3272      turned off by specifying -guid in its command line.
3273    
3274    ----------------------------------------------------------------------
3275    Name: Lal George
3276    Date: 2002/03/12 12 14:42:36 EST
3277    Tag: george-20020312-frequency-computation
3278    Description:
3279    
3280    Integrated jump chaining and static block frequency into the
3281    compiler. More details and numbers later.
3282    
3283    ----------------------------------------------------------------------
3284    Name: Lal George
3285    Date: 2002/03/11 11 22:38:53 EST
3286    Tag: george-20020311-jump-chain-elim
3287    Description:
3288    
3289    Tested the jump chain elimination on all architectures (except the
3290    hppa).  This is on by default right now and is profitable for the
3291    alpha and x86, however, it may not be profitable for the sparc and ppc
3292    when compiling the compiler.
3293    
3294    The gc test will typically jump to a label at the end of the cluster,
3295    where there is another jump to an external cluster containing the actual
3296    code to invoke gc. This is to allow factoring of common gc invocation
3297    sequences. That is to say, we generate:
3298    
3299            f:
3300               testgc
3301               ja   L1      % jump if above to L1
3302    
3303            L1:
3304               jmp L2
3305    
3306    
3307    After jump chain elimination the 'ja L1' instructions is converted to
3308    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3309    up being implemented in their long form (if L2 is far away) using:
3310    
3311            jbe     L3      % jump if below or equal to L3
3312            jmp     L2
3313         L3:
3314            ...
3315    
3316    
3317    For large compilation units L2  may be far away.
3318    
3319    
3320    ----------------------------------------------------------------------
3321    Name: Matthias Blume
3322    Date: 2002/03/11 13:30:00 EST
3323    Tag: blume-20020311-mltreeeval
3324    Description:
3325    
3326    A functor parameter was missing.
3327    
3328    ----------------------------------------------------------------------
3329    Name: Allen Leung
3330    Date: 2002/03/11 10:30:00 EST
3331    Tag: leunga-20020311-runtime-string0
3332    Description:
3333    
3334       The representation of the empty string now points to a
3335    legal null terminated C string instead of unit.  It is now possible
3336    to convert an ML string into C string with InlineT.CharVector.getData.
3337    This compiles into one single machine instruction.
3338    
3339    ----------------------------------------------------------------------
3340    Name: Allen Leung
3341    Date: 2002/03/10 23:55:00 EST
3342    Tag: leunga-20020310-x86-call
3343    Description:
3344    
3345       Added machine generation for CALL instruction (relative displacement mode)
3346    
3347    ----------------------------------------------------------------------
3348    Name: Matthias Blume
3349    Date: 2002/03/08 16:05:00
3350    Tag: blume-20020308-entrypoints
3351    Description:
3352    
3353    Version number bumped to 110.39.1.  NEW BOOTFILES!
3354    
3355    Entrypoints: non-zero offset into a code object where execution should begin.
3356    
3357    - Added the notion of an entrypoint to CodeObj.
3358    - Added reading/writing of entrypoint info to Binfile.
3359    - Made runtime system bootloader aware of entrypoints.
3360    - Use the address of the label of the first function given to mlriscGen
3361      as the entrypoint.  This address is currently always 0, but it will
3362      not be 0 once we turn on block placement.
3363    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3364      with entry points) from mlriscGen.
3365    
3366    ----------------------------------------------------------------------
3367    Name: Allen Leung
3368    Date: 2002/03/07 20:45:00 EST
3369    Tag: leunga-20020307-x86-cmov
3370    Description:
3371    
3372       Bug fixes for CMOVcc on x86.
3373    
3374       1. Added machine code generation for CMOVcc
3375       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3376       3. CMOVcc cannot have an immediate operand as argument.
3377    
3378    ----------------------------------------------------------------------
3379    Name: Matthias Blume
3380    Date: 2002/03/07 16:15:00 EST
3381    Tag: blume-20020307-controls
3382    Description:
3383    
3384    This is a very large but mostly boring patch which makes (almost)
3385    every tuneable compiler knob (i.e., pretty much everything under
3386    Control.* plus a few other things) configurable via both the command
3387    line and environment variables in the style CM did its configuration
3388    until now.
3389    
3390    Try starting sml with '-h' (or, if you are brave, '-H')
3391    
3392    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3393    implements the underlying generic mechanism.
3394    
3395    The interface to some of the existing such facilities has changed somewhat.
3396    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3397    (The getFoo interface is still there for backward-compatibility, but its
3398    use is deprecated.)
3399    
3400    The ml-build script passes -Cxxx=yyy command-line arguments through so
3401    that one can now twiddle the compiler settings when using this "batch"
3402    compiler.
3403    
3404    TODO items:
3405    
3406    We should go through and throw out all controls that are no longer
3407    connected to anything.  Moreover, we should go through and provide
3408    meaningful (and correct!) documentation strings for those controls
3409    that still are connected.
3410    
3411    Currently, multiple calls to Controls.new are accepted (only the first
3412    has any effect).  Eventually we should make sure that every control
3413    is being made (via Controls.new) exactly once.  Future access can then
3414    be done using Controls.acc.
3415    
3416    Finally, it would probably be a good idea to use the getter-setter
3417    interface to controls rather than ref cells.  For the time being, both
3418    styles are provided by the Controls module, but getter-setter pairs are
3419    better if thread-safety is of any concern because they can be wrapped.
3420    
3421    *****************************************
3422    
3423    One bug fix: The function blockPlacement in three of the MLRISC
3424    backpatch files used to be hard-wired to one of two possibilities at
3425    link time (according to the value of the placementFlag).  But (I
3426    think) it should rather sense the flag every time.
3427    
3428    *****************************************
3429    
3430    Other assorted changes (by other people who did not supply a HISTORY entry):
3431    
3432    1. the cross-module inliner now works much better (Monnier)
3433    2. representation of weights, frequencies, and probabilities in MLRISC
3434       changed in preparation of using those for weighted block placement
3435       (Reppy, George)
3436    
3437    ----------------------------------------------------------------------
3438    Name: Lal George
3439    Date: 2002/03/07 14:44:24 EST 2002
3440    Tag: george-20020307-weighted-block-placement
3441    
3442    Tested the weighted block placement optimization on all architectures
3443    (except the hppa) using AMPL to generate the block and edge frequencies.
3444    Changes were required in the machine properties to correctly
3445    categorize trap instructions. There is an MLRISC flag
3446    "weighted-block-placement" that can be used to enable weighted block
3447    placement, but this will be ineffective without block/edge
3448    frequencies (coming soon).
3449    
3450    
3451    ----------------------------------------------------------------------
3452    Name: Lal George
3453    Date: 2002/03/05 17:24:48 EST
3454    Tag: george-20020305-linkage-cluster
3455    
3456    In order to support the block placement optimization, a new cluster
3457    is generated as the very first cluster (called the linkage cluster).
3458    It contains a single jump to the 'real' entry point for the compilation
3459    unit. Block placement has no effect on the linkage cluster itself, but
3460    all the other clusters  have full freedom in the manner in which they
3461    reorder blocks or functions.
3462    
3463    On the x86 the typical linkage code that is generated is:
3464       ----------------------
3465            .align 2
3466       L0:
3467            addl    $L1-L0, 72(%esp)
3468            jmp     L1
3469    
3470    
3471            .align  2
3472       L1:
3473       ----------------------
3474    
3475    72(%esp) is the memory location for the stdlink register. This
3476    must contain the address of the CPS function being called. In the
3477    above example, it contains the address of  L0; before
3478    calling L1 (the real entry point for the compilation unit), it
3479    must contain the address for L1, and hence
3480    
3481            addl $L1-L0, 72(%esp)
3482    
3483    I have tested this on all architectures except the hppa.The increase
3484    in code size is of course negligible
3485    
3486    ----------------------------------------------------------------------
3487    Name: Allen Leung
3488    Date: 2002/03/03 13:20:00 EST
3489    Tag: leunga-20020303-mlrisc-tools
3490    
3491      Added #[ ... ] expressions to mlrisc tools
3492    
3493    ----------------------------------------------------------------------
3494    Name: Matthias Blume
3495    Date: 2002/02/27 12:29:00 EST
3496    Tag: blume-20020227-cdebug
3497    Description:
3498    
3499    - made types in structure C and C_Debug to be equal
3500    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3501    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3502    
3503    ----------------------------------------------------------------------
3504    Name: Matthias Blume
3505    Date: 2002/02/26 12:00:00 EST
3506    Tag: blume-20020226-ffi
3507    Description:
3508    
3509    1. Fixed a minor bug in CM's "noweb" tool:
3510       If numbering is turned off, then truly don't number (i.e., do not
3511       supply the -L option to noweb).  The previous behavior was to supply
3512       -L'' -- which caused noweb to use the "default" line numbering scheme.
3513       Thanks to Chris Richards for pointing this out (and supplying the fix).
3514    
3515    2. Once again, I reworked some aspects of the FFI:
3516    
3517       A. The incomplete/complete type business:
3518    
3519       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3520         gone!
3521       - ML types representing an incomplete type are now *equal* to
3522         ML types representing their corresponding complete types (just like
3523         in C).  This is still safe because ml-nlffigen will not generate
3524         RTTI for incomplete types, nor will it generate functions that
3525         require access to such RTTI.   But when ML code generated from both
3526         incomplete and complete versions of the C type meet, the ML types
3527         are trivially interoperable.
3528    
3529         NOTE:  These changes restore the full generality of the translation
3530         (which was previously lost when I eliminated functorization)!
3531    
3532       B. Enum types:
3533    
3534       - Structure C now has a type constructor "enum" that is similar to
3535         how the "su" constructor works.  However, "enum" is not a phantom
3536         type because each "T enum" has values (and is isomorphic to
3537         MLRep.Signed.int).
3538       - There are generic access operations for enum objects (using
3539         MLRep.Signed.int).
3540       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3541         * The structure contains the definition of type "mlrep" (the ML-side
3542         representation type of the enum).  Normally, mlrep is the same
3543         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3544         then mlrep will be defined as a datatype -- thus facilitating
3545         pattern matching on mlrep values.
3546         ("-ec" will be suppressed if there are duplicate values in an
3547          enumeration.)
3548         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3549         will be generated for each C enum constant xxx.
3550         * Conversion functions m2i and i2m convert between mlrep and
3551         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3552         * Coversion functions c and ml convert between mlrep and "tag enum".
3553         * Access functions (get/set) fetch and store mlrep values.
3554       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3555         enumerations are merged into one single enumeration represented by
3556         structure E_'.
3557    
3558    ----------------------------------------------------------------------
3559    Name: Allen Leung
3560    Date: 2002/02/25 04:45:00 EST
3561    Tag: leunga-20020225-cps-spill
3562    
3563    This is a new implementation of the CPS spill phase.
3564    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3565    In case of problems, replace it with the old file spill.sml
3566    
3567    The current compiler runs into some serious performance problems when
3568    constructing a large record.  This can happen when we try to compile a
3569    structure with many items.  Even a very simple structure like the following
3570    makes the compiler slow down.
3571    
3572        structure Foo = struct
3573           val x_1 = 0w1 : Word32.int
3574           val x_2 = 0w2 : Word32.int
3575           val x_3 = 0w3 : Word32.int
3576           ...
3577           val x_N = 0wN : Word32.int
3578        end
3579    
3580    The following table shows the compile time, from N=1000 to N=4000,
3581    with the old compiler:
3582    
3583    N
3584    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3585           MLRISC ra                               0.06u  0.00s  0.05g
3586              (spills = 0 reloads = 0)
3587           TOTAL                                   0.63u  0.07s  0.21g
3588    
3589    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3590           MLRISC ra                               5.68u  0.59s  3.93g
3591              (spills = 0 reloads = 0)
3592           TOTAL                                   14.71u  0.99s  4.81g
3593    
3594    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3595           MLRISC ra                               5.54u  0.65s  3.91g
3596              (spills = 543 reloads = 1082)
3597           TOTAL                                   65.40u  3.13s  6.00g
3598    
3599    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3600           MLRISC ra                               0.80u  0.10s  0.55g
3601              (spills = 42 reloads = 84)
3602           TOTAL                                   129.42u  5.10s  4.13g
3603    
3604    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3605           MLRISC ra                               2.69u  0.27s  1.38g
3606              (spills = 62 reloads = 124)
3607           TOTAL                                   682.48u  19.61s  13.99g
3608    
3609    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3610           MLRISC ra                               4.96u  0.27s  2.72g
3611              (spills = 85 reloads = 170)
3612           TOTAL                                   2375.26u  57.21s  48.00g
3613    
3614    As you can see the old cps spill module suffers from some serious
3615    performance problem.  But since I cannot decipher the old code fully,
3616    instead of patching the problems up, I'm reimplementing it
3617    with a different algorithm.  The new code is more modular,
3618    smaller when compiled, and substantially faster
3619    (O(n log n) time and O(n) space).  Timing of the new spill module:
3620    
3621    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3622          MLRISC ra                               0.25u  0.02s  0.15g
3623             (spills=1 reloads=3)
3624          TOTAL                                   7.74u  0.34s  1.62g
3625    
3626    Implementation details:
3627    
3628    As far as I can tell, the purpose of the CPS spill module is to make sure the
3629    number of live variables at any program point (the bandwidth)
3630    does not exceed a certain limit, which is determined by the
3631    size of the spill area.
3632    
3633    When the bandwidth is too large, we decrease the register pressure by
3634    packing live variables into spill records.  How we achieve this is
3635    completely different than what we did in the old code.
3636    
3637    First, there is something about the MLRiscGen code generator
3638    that we should be aware of:
3639    
3640    o MLRiscGen performs code motion!
3641    
3642       In particular, it will move floating point computations and
3643       address computations involving only the heap pointer to
3644       their use sites (if there is only a single use).
3645       What this means is that if we have a CPS record construction
3646       statement
3647    
3648           RECORD(k,vl,w,e)
3649    
3650       we should never count the new record address w as live if w
3651       has only one use (which is often the case).
3652    
3653       We should do something similar to floating point, but the transformation
3654       there is much more complex, so I won't deal with that.
3655    
3656    Secondly, there are now two new cps primops at our disposal:
3657    
3658     1. rawrecord of record_kind option
3659        This pure operator allocates some uninitialized storage from the heap.
3660        There are two forms:
3661    
3662         rawrecord NONE [INT n]  allocates a tagless record of length n
3663         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3664                                     and initializes the tag.
3665    
3666     2. rawupdate of cty
3667          rawupdate cty (v,i,x)
3668          Assigns to x to the ith component of record v.
3669          The storelist is not updated.
3670    
3671    We use these new primops for both spilling and increment record construction.
3672    
3673     1. Spilling.
3674    
3675        This is implemented with a linear scan algorithm (but generalized
3676        to trees).  The algorithm will create a single spill record at the
3677        beginning of the cps function and use rawupdate to spill to it,
3678        and SELECT or SELp to reload from it.  So both spills and reloads
3679        are fine-grain operations.  In contrast, in the old algorithm
3680        "spills" have to be bundled together in records.
3681    
3682        Ideally, we should sink the spill record construction to where
3683        it is needed.  We can even split the spill record into multiple ones
3684        at the places where they are needed.  But CPS is not a good
3685        representation for global code motion, so I'll keep it simple and
3686        am not attempting this.
3687    
3688     2. Incremental record construction (aka record splitting).
3689    
3690        Long records with many component values which are simulatenously live
3691        (recall that single use record addresses are not considered to
3692         be live) are constructed with rawrecord and rawupdate.
3693        We allocate space on the heap with rawrecord first, then gradually
3694        fill it in with rawupdate.  This is the technique suggested to me
3695        by Matthias.
3696    
3697        Some restrictions on when this is applicable:
3698        1. It is not a VECTOR record.  The code generator currently does not handle
3699           this case. VECTOR record uses double indirection like arrays.
3700        2. All the record component values are defined in the same "basic block"
3701           as the record constructor.  This is to prevent speculative
3702           record construction.
3703    
3704    ----------------------------------------------------------------------
3705    Name: Allen Leung
3706    Date: 2002/02/22 01:02:00 EST
3707    Tag: leunga-20020222-mlrisc-tools
3708    
3709    Minor bug fixes in the parser and rewriter
3710    
3711    ----------------------------------------------------------------------
3712    Name: Allen Leung
3713    Date: 2002/02/21 20:20:00 EST
3714    Tag: leunga-20020221-peephole
3715    
3716    Regenerated the peephole files.  Some contained typos in the specification
3717    and some didn't compile because of pretty printing bugs in the old version
3718    of 'nowhere'.
3719    
3720    ----------------------------------------------------------------------
3721    Name: Allen Leung
3722    Date: 2002/02/19 20:20:00 EST
3723    Tag: leunga-20020219-mlrisc-tools
3724    Description:
3725    
3726       Minor bug fixes to the mlrisc-tools library:
3727    
3728       1.  Fixed up parsing colon suffixed keywords
3729       2.  Added the ability to shut the error messages up
3730       3.  Reimplemented the pretty printer and fixed up/improved
3731           the pretty printing of handle and -> types.
3732       4.  Fixed up generation of literal symbols in the nowhere tool.
3733       5.  Added some SML keywords to to sml.sty
3734    
3735    ----------------------------------------------------------------------
3736    Name: Matthias Blume
3737    Date: 2002/02/19 16:20:00 EST
3738    Tag: blume-20020219-cmffi
3739    Description:
3740    
3741    A wild mix of changes, some minor, some major:
3742    
3743    * All C FFI-related libraries are now anchored under $c:
3744        $/c.cm      --> $c/c.cm
3745        $/c-int.cm  --> $c/internals/c-int.cm
3746        $/memory.cm --> $c/memory/memory.cm
3747    
3748    * "make" tool (in CM) now treats its argument pathname slightly
3749      differently:
3750        1. If the native expansion is an absolute name, then before invoking
3751           the "make" command on it, CM will apply OS.Path.mkRelative
3752           (with relativeTo = OS.FileSys.getDir()) to it.
3753        2. The argument will be passed through to subsequent phases of CM
3754           processing without "going native".  In particular, if the argument
3755           was an anchored path, then "make" will not lose track of that anchor.
3756    
3757    * Compiler backends now "know" their respective C calling conventions
3758      instead of having to be told about it by ml-nlffigen.  This relieves
3759      ml-nlffigen from one of its burdens.
3760    
3761    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3762    
3763    * Export C_DEBUG and C_Debug from $c/c.cm.
3764    
3765    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3766      subtyping relationship between incomplete pointers and their complete
3767      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3768      with the convention of instantiating 'o with ('t, 'c) obj whenever
3769      the pointer target type is complete.  In the incomplete case, 'o
3770      will be instantiated with some "'c iobj" -- a type obtained by
3771      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3772    
3773      Operations that work on both incomplete and complete pointer types are
3774      typed as taking an 'o ptr while operations that require the target to
3775      be known are typed as taking some ('t, 'c) obj ptr.
3776    
3777      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3778      where void is an eqtype without any values.  This makes it possible
3779      to work on voidptr values using functions meant to operate on light
3780      incomplete pointers.
3781    
3782    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3783      been vastly simplified.
3784    
3785    ----------------------------------------------------------------------
3786    Name: Matthias Blume
3787    Date: 2002/02/19 10:48:00 EST
3788    Tag: blume-20020219-pqfix
3789    Description:
3790    
3791    Applied Chris Okasaki's bug fix for priority queues.
3792    
3793    ----------------------------------------------------------------------
3794    Name: Matthias Blume
3795    Date: 2002/02/15 17:05:00
3796    Tag: Release_110_39
3797    Description:
3798    
3799    Last-minute retagging is becoming a tradition... :-(
3800    
3801    This is the working release 110.39.
3802    
3803    ----------------------------------------------------------------------
3804    Name: Matthias Blume
3805    Date: 2002/02/15 16:00:00 EST
3806    Tag: Release_110_39-orig
3807    Description:
3808    
3809    Working release 110.39.  New bootfiles.
3810    
3811    (Update: There was a small bug in the installer so it wouldn't work
3812    with all shells.  So I retagged. -Matthias)
3813    
3814    ----------------------------------------------------------------------
3815    Name: Matthias Blume
3816    Date: 2002/02/15 14:17:00 EST
3817    Tag: blume-20020215-showbindings
3818    Description:
3819    
3820    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3821    the latter can be useful for exploring what bindings are available at
3822    the interactive prompt.  (The first function returns only the list
3823    of symbols that are really bound, the second prints those but also the
3824    ones that CM's autoloading mechanism knows about.)
3825    
3826    ----------------------------------------------------------------------
3827    Name: Matthias Blume
3828    Date: 2002/02/15 12:08:00 EST
3829    Tag: blume-20020215-iptrs
3830    Description:
3831    
3832    Two improvements to ml-nlffigen:
3833    
3834      1. Write files only if they do not exist or if their current contents
3835         do not coincide with what's being written.  (That is, avoid messing
3836         with the time stamps unless absolutely necessary.)
3837    
3838      2. Implement a "repository" mechanism for generated files related
3839         to "incomplete pointer types".   See the README file for details.
3840    
3841    ----------------------------------------------------------------------
3842    Name: Matthias Blume
3843    Date: 2002/02/14 11:50:00 EST
3844    Tag: blume-20020214-quote
3845    Description:
3846    
3847    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3848    because of the new and improved tag generation scheme.  (Thanks to Allen
3849    Leung for pointing it out.)
3850    
3851    ----------------------------------------------------------------------
3852    Name: Lal George
3853    Date: 2002/02/14 09:55:27 EST 2002
3854    Tag: george-20020214-isabelle-bug
3855    Description:
3856    
3857    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3858    of Isabelle on the x86.
3859    
3860    From Allen:
3861    -----------
3862     I've found the problem:
3863    
3864         in ra-core.sml, I use the counter "blocked" to keep track of the
3865         true number of elements in the freeze queue.  When the counter goes
3866         to zero, I skip examining the queue.  But I've messed up the
3867         bookkeeping in combine():
3868    
3869             else ();
3870             case !ucol of
3871               PSEUDO => (if !cntv > 0 then
3872                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3873                                        ^^^^^^^^^^^^^^^^^^^^^^^
3874                      moveu := mergeMoveList(!movev, !moveu)
3875                     )
3876                  else ();
3877    
3878         combine() is called to coalesce two nodes u and v.
3879         I think I was thinking that if the move counts of u and v are both
3880         greater than zero then after they are coalesced then one node is
3881         removed from the freeze queue.  Apparently I was thinking that
3882         both u and v are of low degree, but that's clearly not necessarily true.
3883    
3884    
3885    02/12/2002:
3886        Here's the patch.  HOL now compiles.
3887    
3888        I don't know how this impact on performance (compile
3889        time or runtime).  This bug caused the RA (especially on the x86)
3890        to go thru the potential spill phase when there are still nodes on the
3891        freeze queue.
3892    
3893    
3894    
3895    
3896    ----------------------------------------------------------------------
3897    Name: Matthias Blume
3898    Date: 2002/02/13 22:40:00 EST
3899    Tag: blume-20020213-fptr-rtti
3900    Description:
3901    
3902    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3903    updates.
3904    
3905    ----------------------------------------------------------------------
3906    Name: Matthias Blume
3907    Date: 2002/02/13 16:41:00 EST
3908    Tag: blume-20020213-cmlpq
3909    Description:
3910    
3911    Added new priority queue export symbols (which have just been added to
3912    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3913    not compile and the installer would choke.)
3914    
3915    ----------------------------------------------------------------------
3916    Name: Matthias Blume
3917    Date: 2002/02/13 16:15:00 EST
3918    Tag: blume-20020213-various
3919    Description:
3920    
3921    1. More tweaks to ml-nlffigen:
3922    
3923       - better internal datastructures (resulting in slight speedup)
3924       - "-match" option requires exact match
3925       - "localized" gensym counters (untagged structs/unions nested within
3926         other structs/unions or within typedefs get a fresh counter; their
3927         tag will be prefixed by a concatenation of their parents' tags)
3928       - bug fixes (related to calculation of transitive closure of types
3929         to be included in the output)
3930    
3931    2. Minor Basis updates:
3932    
3933       - added implementations for List.collate and Option.app
3934    
3935    ----------------------------------------------------------------------
3936    Name: Matthias Blume
3937    Date: 2002/02/11 15:55:00 EST
3938    Tag: blume-20020211-gensym
3939    Description:
3940    
3941    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3942    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3943    names (ML structure names that correspond to unnamed C structs, unions,
3944    and enums), so that separate runs of ml-nlffigen do not clash.
3945    
3946    ----------------------------------------------------------------------
3947    Name: Matthias Blume
3948    Date: 2002/02/11 12:05:00 EST
3949    Tag: blume-20020211-gensml
3950    Description:
3951    
3952    A quick fix for a problem with GenSML (in the pgraph-util library):
3953    Make generation of toplevel "local" optional.  (Strictly speaking,
3954    signature definitions within "local" are not legal SML.)
3955    
3956    Other than that: updates to INSTALL and cm/TODO.
3957    
3958    ----------------------------------------------------------------------
3959    Name: Matthias Blume
3960    Date: 2002/02/08 15:00:00 EST
3961    Tag: blume-20020208-uniquepid
3962    Description:
3963    
3964    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3965    
3966    1. The installer (config/install.sh) has gotten smarter:
3967    
3968         - Configuration options are a bit easier to specify now
3969           (in config/targets).
3970         - Bug in recognizing .tar.bz2 files fixed.
3971         - Installer automatically resolves dependencies between
3972           configuration options (e.g., if you ask for eXene, you will
3973           also get cml -- regardless whether you asked for it or not).
3974         - Installer can run in "quieter mode" by setting the environment
3975           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3976           "completely silent", though.
3977         - Build HashCons library as part of smlnj-lib.
3978    
3979    2. A new scheme for assigning persistent identifiers to compilation
3980       units (and, by extension, to types etc.) has been put into place.
3981       This fixes a long-standing bug where types and even dynamic values
3982       can get internally confused, thereby compromising type safety
3983       (abstraction) and dynamic correctness.  See
3984    
3985         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3986    
3987       for an example of how things could go wrong until now.
3988    
3989       The downside of the new scheme is that pids are not quite as
3990       persistent as they used to be: CM will generate a fresh pid
3991       for every compilation unit that it thinks it sees for the first
3992       time.  That means that if you compile starting from a clean, fresh
3993       source tree at two different times, you end up with different
3994       binaries.
3995    
3996       Cutoff recompilation, however, has not been compromised because
3997       CM keeps pid information in special caches between runs.
3998    
3999    ----------------------------------------------------------------------
4000    Name: Lal George
4001    Date: 2002/02/07 15:34:13 EST 2002
4002    Tag: <none>
4003    Description:
4004    
4005    Compilers that generate assembly code may produce  global labels
4006    whose value is resolved at link time. The various peephole optimization
4007    modules did not take this in account.
4008    
4009    TODO. The Labels.addrOf function should really return an option
4010    type so that clients are forced to deal with this issue, rather
4011    than an exception being raised.
4012    
4013    ----------------------------------------------------------------------
4014    Name: Lal George
4015    Date: 2002/02/06 13:55:02 EST
4016    Tag: george-20020206-ra-breakup
4017    Description:
4018    
4019    1. A bug fix from Allen.
4020    
4021        A typo causes extra fstp %st(0)'s to be generated at compensation
4022        edges, which might cause stack underflow traps at runtime.  This
4023        occurs in fft where there are extraneous fstps right before the 'into'
4024        trap instruction (in this case they are harmless since none of the
4025        integers overflow.)
4026    
4027    2. Pulled out various utility modules that were embedded in the modules
4028       of the register allocator. I need these modules for other purposes, but
4029       they are not complete enough to put into a library (just yet).
4030    ----------------------------------------------------------------------
4031    Name: Matthias Blume
4032    Date: 2002/01/31 16:05:00 EST
4033    Tag: blume-20020131-sparc-ccalls
4034    Description:
4035    
4036    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
4037       of extra stack space by mistake.  Fixed.
4038    
4039    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
4040    
4041    ----------------------------------------------------------------------
4042    Name: Allen Leung
4043    Date: 2002/01/30
4044    Tag: leunga-20020130-nowhere-bug-fix
4045    Description:
4046    
4047       MLRISC bug fixes:
4048       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
4049       2. MachineInt.fromString was negating its value.
4050    
4051    ----------------------------------------------------------------------
4052    Name: Matthias Blume
4053    Date: 2002/01/29
4054    Tag: blume-20020129-INSTALL
4055    Description:
4056    
4057    - Added somewhat detailed installation instructions (file INSTALL).
4058    - Fixed curl-detection bug in config/install.sh.
4059    - It is now possible to select the URL getter using the URLGETTER
4060      environment variable:
4061    
4062          not set / "unknown"      --> automatic detection (script tries wget,
4063                                       curl, and lynx)
4064          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
4065                                       how to properly invoke them)
4066          other                    --> use $URLGETTER directly, it must take
4067                                       precisely two command-line arguments
4068                                       (source URL and destination file name)
4069    
4070    ----------------------------------------------------------------------
4071    Name: Matthias Blume
4072    Date: 2002/01/28
4073    Tag: blume-20020128-sparc-ccalls
4074    Description:
4075    
4076    - Fixed problem with calculation of "used" registers in sparc-c-calls.
4077    - Make use of the allocParam argument in sparc-c-calls.
4078    
4079    ----------------------------------------------------------------------
4080    Name: Matthias Blume
4081    Date: 2002/01/28
4082    Tag: blume-20020128-allocParam
4083    Description:
4084    
4085    John Reppy:  Changes c-calls API to accept client-callback for
4086    allocating extra stack space.
4087    me: Corresponding changes to mlriscGen (using a dummy argument that
4088        does not change the current behavior).
4089    
4090    ----------------------------------------------------------------------
4091    Name: Matthias Blume
4092    Date: 2002/01/28 12:00:00
4093    Tag: Release_110_38
4094    Description:
4095    
4096    This time for real!!!
4097