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

SCM Repository

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

Diff of /sml/trunk/HISTORY

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

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