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