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