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 993, Fri Nov 23 02:37:10 2001 UTC revision 1806, Tue May 31 20:47:40 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: Allen Leung (leunga (at) reservoir (dot) com)
17    Date: 2005/05/31 16:47:00 EST
18    Tag: leunga-20050531-cygwin-fault
19    Description:
20    
21    Updated Cygwin's fault/signal handling to match the Windows version.
22    Updated the export list.
23    
24    ----------------------------------------------------------------------
25    Name: Matthias Blume (blume (at) tti - c (dot) org)
26    Date: 2005/05/18 13:35:00 CDT
27    Tag: Release_110_54
28    Description:
29    
30    New working version (110.54).  NEW BOOTFILES!
31    
32    ----------------------------------------------------------------------
33    Name: Matthias Blume (blume (at) tti - c (dot) org)
34    Date: 2005/05/18 11:58:00 CDT
35    Tag: blume-20050518-installer
36    Description:
37    
38    Added support scripts for Mac OS X PackageMaker and modified
39    config/install.sh so that it supports re-dumping a heap image
40    after customization.
41    
42    ----------------------------------------------------------------------
43    Name: Matthias Blume (blume (at) tti - c (dot) org)
44    Date: 2005/05/18 10:55:00 CDT
45    Tag: blume-20050518-realdiv-noovld
46    Description:
47    
48    Un-overloaded / to work around bug in overloading resolution code.
49    
50    ----------------------------------------------------------------------
51    Name: Matthias Blume (blume (at) tti - c (dot) org)
52    Date: 2005/05/16 23:50:00 CDT
53    Tag: blume-20050516-redump-heap
54    Description:
55    
56    Added mechanism for re-creating a heap file for the interactive system
57    after configuration variables have been changed.
58    
59       CM.redump_heap : string -> unit
60    
61    This is much like SMLofNJ.exportML, but starting from the resulting
62    heap does not return to the caller of CM.redump_heap but
63    restarts the interactive system from scratch.  The original call of
64    CM.redump_heap does not return but ends the interactive session.
65    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
66    
67    Internally, redump_heap winds the dynamic execution context back to
68    the point where the original heap image was created and re-executes
69    the heap image generation code in the boot code.
70    
71    ----------------------------------------------------------------------
72    Name: Matthias Blume (blume (at) tti - c (dot) org)
73    Date: 2005/05/09 21:55:00 CDT
74    Tag: blume-20050509-word64
75    Description:
76    
77    Added a hack to the existing hack known as Word64 to make fromString
78    behave correctly.  I am still not sure whether Word64.scan will work
79    as specified with respect to the interaction of radix and prefix.
80    
81    ----------------------------------------------------------------------
82    Name: Allen Leung (leunga (at) reservoir (dot) com)
83    Date: 2005/05/04 11:50:00 EST
84    Tag: leunga-20050504-checkgc
85    Description:
86    
87    Added a gc protocol checking phase.  This phase is enabled with
88    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
89    
90    ----------------------------------------------------------------------
91    Name: Matthias Blume (blume (at) tti - c (dot) org)
92    Date: 2005/05/04 10:45:00 CDT
93    Tag: blume-20050504-intinf
94    Description:
95    
96    Fixed a bug in the implementation of div and mod for IntInf.
97    Thanks to Neophytos Michael for reporting the problem.
98    
99    ----------------------------------------------------------------------
100    Name: Matthias Blume (blume (at) tti - c (dot) org)
101    Date: 2005/05/04 10:35:00 CDT
102    Tag: blume-20050504-join
103    Description:
104    
105    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
106    
107    ----------------------------------------------------------------------
108    Name: Matthias Blume (blume (at) tti - c (dot) org)
109    Date: 2005/02/28 23:40:00 CST
110    Tag: blume-20050228-mVar
111    Description:
112    
113    Fixed serious bug (brown paper bag variety) in new implementation of
114    structure Atom in CML.  (I had accidentally used a mailbox instead of
115    an mvar, leaving the door open for races.)
116    
117    ----------------------------------------------------------------------
118    Name: Matthias Blume (blume (at) tti - c (dot) org)
119    Date: 2005/02/25 15:00:00 CST
120    Tag: Release_110_53
121    Description:
122    
123    New working version (110.53).  NEW BOOTFILES!
124    
125    ----------------------------------------------------------------------
126    Name: Matthias Blume (blume (at) tti - c (dot) org)
127    Date: 2005/02/25 14:50:00 CST
128    Tag: blume-20050225-susp
129    Description:
130    
131    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
132    implemented in Core -- which means that it is the same as the one used
133    by the lazy extension.
134    
135    ----------------------------------------------------------------------
136    Name: Matthias Blume (blume (at) tti - c (dot) org)
137    Date: 2005/02/24 16:50:00 CST
138    Tag: blume-20050224-cml-atom
139    Description:
140    
141    Simpler and at the same time more general implementation of structure
142    Atom in CML.
143    
144    ----------------------------------------------------------------------
145    Name: Matthias Blume (blume (at) tti - c (dot) org)
146    Date: 2005/02/15 17:35:00 CST
147    Tag: blume-20050215-tools
148    Description:
149    
150    Created new "tools" directory under "src" and moved "TraceDebugProf"
151    there.
152    
153    ----------------------------------------------------------------------
154    Name: Matthias Blume (blume (at) tti - c (dot) org)
155    Date: 2005/02/10 17:55:00 CST
156    Tag: blume-20050210-longlong
157    Description:
158    
159    Implemented "long long" arguments and results for NLFFI.  (Only the
160    PPC/MacOS implementation is complete, the other backends still need to
161    be updated.)
162    
163    ----------------------------------------------------------------------
164    Name: Matthias Blume (blume (at) tti - c (dot) org)
165    Date: 2005/01/24 17:40:00 CST
166    Tag: blume-20050124-mlyacc
167    Description:
168    
169    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
170    problem with certain "as" patterns which previously got rendered
171    using incorrect syntax.
172    
173    ----------------------------------------------------------------------
174    Name: Matthias Blume (blume (at) tti - c (dot) org)
175    Date: 2005/01/18 12:00:00 CST
176    Tag: blume-20050118-profile
177    Description:
178    
179    Made time profiling code (interrupt handler) in runtime system aware
180    of new array representation.
181    
182    ----------------------------------------------------------------------
183    Name: Matthias Blume (blume (at) tti - c (dot) org)
184    Date: 2005/01/14 18:00:00 CST
185    Tag: blume-20050114-heap2exec
186    Description:
187    
188    Implemented new (but still experimental) heap2exec facility.  This is
189    tested under Mac OS X and should work under Linux (will test shortly).
190    It will probably also work on the Sparc (will test some time later).
191    
192      - removed old "HACKED_STANDALONE" hack from runtime
193    
194    To be able to test this, uncomment the request for "heap2asm" in
195    config/targets prior to installation.  (Notice that this is different
196    from "heap2exec" mentioned below.  Not a typo.)
197    
198    To perform an actual test, run the command
199    
200       $ bin/heap2exec heapfile execfile
201    
202    (You can put heap2exec on your shell's path.)
203    
204    For example, run
205    
206      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
207    
208    This will create a standalone executable called "mly" which you can
209    then invoke directly as a command.
210    
211    ----------------------------------------------------------------------
212    Name: Matthias Blume (blume (at) tti - c (dot) org)
213    Date: 2005/01/07 11:44:00 CST
214    Tag: blume-20050107-mlstring
215    Description:
216    
217    fixed off-by-one error in ML_STRING macro (globals.c)
218    
219    ----------------------------------------------------------------------
220    Name: Matthias Blume (blume (at) tti - c (dot) org)
221    Date: 2004/12/23 18:00:00 CST
222    Tag: blume-20041223-santa
223    Description:
224    
225    Made ml-build script "smarter" (but only very little).
226    
227    ----------------------------------------------------------------------
228    Name: Matthias Blume (blume (at) tti - c (dot) org)
229    Date: 2004/12/21 15:05:00 CST
230    Tag: blume-20041221-longlong
231    Description:
232    
233    * Implemented access to signed and unsigned long long data in NLFFI.
234    (The parameter-passing part of the picture has not complete. But
235    data structure access seems to work.)
236    
237    * Fixed CM's incorrect assumption that the PPC is little-endian.
238      (On the Mac, it is big-endian.  And that's currently our only
239      PPC platform.)
240    
241    ----------------------------------------------------------------------
242    Name: Matthias Blume (blume (at) tti - c (dot) org)
243    Date: 2004/12/21 12:50:00 CST
244    Tag: blume-20041221-memory
245    Description:
246    
247    Some cleanup in the $c/memory.cm library: separated some concerns by
248    moving allocation code and memory access code each into their own
249    files.
250    
251    ----------------------------------------------------------------------
252    Name: Allen Leung (leunga (at) reservoir (dot) com)
253    Date: 2004/12/17 16:12:00 EST
254    Tag: leunga-20041217-cygwin-smlnj-home
255    Description:
256    
257        The Unix I/O library of SML/NJ on cygwin does not understand
258    Windows style pathname, so problems arise when SMLNJ_HOME is set
259    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
260    to a POSIX pathname on cygwin.
261    
262    ----------------------------------------------------------------------
263    Name: Matthias Blume (blume (at) tti - c (dot) org)
264    Date: 2004/12/16 13:35:00 CST
265    Tag: Release_110_52
266    Description:
267    
268    Last-minute changes incorporated into 110.52.  Release tag moved.
269    
270    The changes:
271       - HashString.hashString' -> HashString.hashSubstring
272       - bug fix in UnivariateStats
273    
274    ----------------------------------------------------------------------
275    Name: Matthias Blume (blume (at) tti - c (dot) org)
276    Date: 2004/12/15 23:40:00 CST
277    Tag: blume-20041215-hashSubstring
278    Description:
279    
280    - HashString.hashString' -> HashString.hashSubstring
281    - corresponding changes in atom.sml
282    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
283      added some comments
284    
285    ----------------------------------------------------------------------
286    Name: Matthias Blume (blume (at) tti - c (dot) org)
287    Date: 2004/12/15 15:30:00 CST
288    Tag: (Release_110_52)
289    Description:
290    
291    New working version (110.52).  NEW BOOTFILES!
292    
293    ----------------------------------------------------------------------
294    Name: Matthias Blume (blume (at) tti - c (dot) org)
295    Date: 2004/12/15 12:45:00 CST
296    Tag: blume-20041215-spaces
297    Description:
298    
299    More on the space problem (this time for Win32).
300    
301    ----------------------------------------------------------------------
302    Name: Matthias Blume (blume (at) tti - c (dot) org)
303    Date: 2004/12/14 17:30:00 CST
304    Tag: blume-20041214-spaces
305    Description:
306    
307    Hacked some of the scripts (in particular: the installer) to cope with
308    spaces in filenames a bit better.  But beware: the current "solution"
309    is likely still full of bugs and inherently incomplete.  (We need to
310    do away with those shell scripts for a comprehensive solution.)
311    
312    ----------------------------------------------------------------------
313    Name: Matthias Blume (blume (at) tti - c (dot) org)
314    Date: 2004/12/13 14:45:00 CST
315    Tag: blume-20041213-ml-makedepend
316    Description:
317    
318    Fixed bug in code for ml-makedepend.
319    
320    ----------------------------------------------------------------------
321    Name: Matthias Blume (blume (at) tti - c (dot) org)
322    Date: 2004/12/09 16:30:00 CST
323    Tag: blume-20041209-statistics
324    Description:
325    
326    Added two simple but potentially useful statistics modules to SML/NJ Library.
327    (See CHANGES file there.)
328    
329    ----------------------------------------------------------------------
330    Name: Matthias Blume (blume (at) tti - c (dot) org)
331    Date: 2004/12/01 16:50:00 CST
332    Tag: blume-20041201-atom
333    Description:
334    
335    smlnj-lib:
336    
337      Added function HashString.hashString' for substrings.
338      Hand-inlined CharVector.foldl into HashString (for speed).
339      Modified implementation of structure Atom to avoid extracting
340      strings from substrings unless necessary.
341      (Also see CHANGES file for smlnj-lib.)
342    
343    ----------------------------------------------------------------------
344    Name: Matthias Blume (blume (at) tti - c (dot) org)
345    Date: 2004/11/24 22:15:00 CST
346    Tag: blume-20041124-cml
347    Description:
348    
349    Made sure CML compiles when Position = Int64.
350    
351    ----------------------------------------------------------------------
352    Name: Matthias Blume (blume (at) tti - c (dot) org)
353    Date: 2004/11/24 14:30:00 CST
354    Tag: blume-20041124-position
355    Description:
356    
357    The compiler can now be compiled in a mode that makes structure Position
358    equal to Int64.  The default, however, is unchanged (Position = Int31)
359    for the time being.
360    
361    To enable 64-bit positions, use the following procedure:
362    
363    1. Start sml
364    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
365    3. Type
366          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
367    4. Run CMB.make() as usual.
368    
369    This is barely tested.  The only test so far was a little SML program
370    counting the number of characters in an 8-gigabyte file by
371    reading it character-by-character.  That test was successful.
372    
373    In support of 64-bit positions, a number of new functions have been
374    added to the runtime system.
375    
376    ----------------------------------------------------------------------
377    Name: Matthias Blume (blume (at) tti - c (dot) org)
378    Date: 2004/11/23 14:45:00 CST
379    Tag: blume-20041123-useFile
380    Description:
381    
382    Fixed a problem with unhelpful error messages related to problems with
383    .cm- or .sml files that appear as part of the sml command line.
384    
385    ----------------------------------------------------------------------
386    Name: Matthias Blume (blume (at) tti - c (dot) org)
387    Date: 2004/11/18 15:40:00 CST
388    Tag: Release_110_51
389    Description:
390    
391    New working version (110.51).  NEW BOOTFILES!
392    
393    ----------------------------------------------------------------------
394    Name: Matthias Blume (blume (at) tti - c (dot) org)
395    Date: 2004/11/18 15:35:00 CST
396    Tag: <none>
397    Description:
398    
399    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
400    Joemann).
401    
402    ----------------------------------------------------------------------
403    Name: Allen Leung (leunga (at) reservoir (dot) com)
404    Date: 2004/11/17 16:05:21 EST 2004
405    Tag: leunga-20041117-mlrisc-live-kill
406    Description:
407    
408       Added support for MLTree constructs LIVE and KILL
409    to all the architectures.
410    
411    ----------------------------------------------------------------------
412    Name: Matthias Blume (blume (at) tti - c (dot) org)
413    Date: 2004/11/13 00:20:00 CST
414    Tag: blume-20041113-versiontool
415    Description:
416    
417    - Stripped down the versiontool: It now only handles the version number.
418      The date string is generated at bootstrap time (during makeml).
419    
420    - In a previous commit, fixed a minor issue with how polyequal is being
421      translated.  In particular, the code now "looks through" abstractions.
422      This results in slightly fewer polyEqual warnings and hopefully slightly
423      more efficient code.  Important examples for where this matters are
424      the new int64 and word64 types.
425    
426    ----------------------------------------------------------------------
427    Name: Matthias Blume (blume (at) tti - c (dot) org)
428    Date: 2004/11/12 00:30:00 CST
429    Tag: blume-20041112-int64
430    Description:
431    
432    Structure Int64 fully hooked in.  (The implementation is not very
433    efficient, though.)
434    
435    ----------------------------------------------------------------------
436    Name: Matthias Blume (blume (at) tti - c (dot) org)
437    Date: 2004/11/11 17:30:00 CST
438    Tag: blume-20041111-more64
439    Description:
440    
441    All the pieces of Word64 are now there, with the exception of the
442    conversions from and to LargeWord. (Eventually these need to be identities,
443    but for the time being they don't even make sense because LargeWord is
444    32-bit wide.)
445    
446    Also started to add similar support for Int64, but major pieces of that
447    are still missing.
448    
449    ----------------------------------------------------------------------
450    Name: Matthias Blume (blume (at) tti - c (dot) org)
451    Date: 2004/11/11 00:15:00 CST
452    Tag: blume-20041111-word64
453    Description:
454    
455    Structure Word64 is now (almost) complete, word literals and patterns
456    seem to work.  There are a few odd pieces missing.  In particular,
457    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
458    at the moment.
459    
460    Making Word64 official would mean that LargeWord becomes Word64.  But
461    this requires extreme care because most word-word conversions have to
462    go through LargeWord, so making a mistake means loss of efficiency or
463    worse.  Eventually there will be a solution similar to (but actually
464    simpler than) what I did with IntInf.
465    
466    ----------------------------------------------------------------------
467    Name: Matthias Blume (blume (at) tti - c (dot) org)
468    Date: 2004/11/10 18:12:00 CST
469    Tag: blume-20041110-64bit
470    Description:
471    
472    More 64-bit hacking (but still not even half-way there yet).
473    Also, some assorted improvements to the handling of 8-bit words.
474    
475    ----------------------------------------------------------------------
476    Name: Matthias Blume (blume (at) tti - c (dot) org)
477    Date: 2004/11/09 17:50:00 CST
478    Tag: <none>
479    Description:
480    
481    Started adding some infrastructure for supporting 64-bit int- and
482    word-types.  (Still in its very early stages.)
483    
484    ----------------------------------------------------------------------
485    Name: Matthias Blume (blume (at) tti - c (dot) org)
486    Date: 2004/10/28 10:45:00 CDT
487    Tag: Release_110_50
488    Description:
489    
490    New working version (110.50).  NEW BOOTFILES!
491    
492                 =====================
493    
494    Also:
495    
496      - Changend config/srcarchiveurl from a file just
497        containing the URL string into a file containing
498        shell script code.  The code has access to the $VERSION variable.
499      - Made corresponding changes to config/install.sh and config/unpack.
500      - Default contents of config/srcarchiveurl uses $VERSION and
501        normally does not have to be edited to reflect a version change.
502    
503      (As a result, a version change can be done by just editing
504       config/version, the rest is now automatic.)
505    
506    ----------------------------------------------------------------------
507    Name: Matthias Blume (blume (at) tti - c (dot) org)
508    Date: 2004/10/27 17:50:00 CDT
509    Tag: blume-20041027-btrace-msg
510    Description:
511    
512    BackTrace.monitor now also reports the source of the exception that
513    triggered the trace.
514    
515    ----------------------------------------------------------------------
516    Name: Matthias Blume (blume (at) tti - c (dot) org)
517    Date: 2004/10/27 17:20:00 CDT
518    Tag: blume-20041027-x86-c-calls
519    Description:
520    
521    This is the HISTORY entry for two earlier commits, both concerning
522    the x86 c-calls code in MLRISC:
523    
524      - added a missing LOAD in the code that deals with struct arguments
525      - made sure the caller does not add the wrong number of bytes to the
526        stack pointer after a call of a function returning a struct
527        (the callee already pops the implicit argument which points to
528         the space reserved for the result)
529    
530    ----------------------------------------------------------------------
531    Name: Allen Leung (leunga (at) reservoir (dot) com)
532    Date: 2004/10/24 14:00:00 EST
533    Tag: leunga-20041024-x86-gas-fucomip
534    Description:
535    
536       John discovered a bug in the syntax of fucomip.
537       The opcodes FU?COMIP? have been changed to
538    
539           fu?comip? %st(i), %st
540    
541    ----------------------------------------------------------------------
542    Name: Matthias Blume (blume (at) tti - c (dot) org)
543    Date: 2004/10/20 15:06:00 CDT
544    Tag: blume-20041020-standalone-backtrace
545    Description:
546    
547    Added a mechanism for getting back-trace information from standalone
548    programs.  Here is how it works:
549    
550       1. The part of the program from which you want to get backtrace
551          information (usually the whole program) should be wrapped with
552          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
553          main program could be modified from something like
554    
555             fun main (pgm, args) = ...
556    
557         to
558    
559            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
560    
561       2. To be able to access BackTrace.monitor, you have to add
562          library $smlnj-tdp/plugins.cm to the .cm file that contains your
563          main function.
564    
565       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
566          CM might have created in the past for your project).
567    
568       4. Build the system using this command line:
569    
570             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
571                      myprog.cm MyProg.main myprog
572    
573          instead of the usual
574    
575             ml-build myprog.cm MyProg.main myprog
576    
577    I changed a library name:
578    
579       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
580    
581    New libraries:
582    
583       $smlnj-tdp/back-trace.cm
584          -- when loaded causes the back-trace plugin to be installed
585       $smlnj-tdp/coverage.cm
586          -- when loaded causes the coverage plugin to be installed
587    
588    ----------------------------------------------------------------------
589    Name: Matthias Blume (blume (at) tti - c (dot) org)
590    Date: 2004/10/18 16:45:00 CDT
591    Tag: blume-20041018-groupowner
592    Description:
593    
594    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
595    
596    Eliminated group owner specs from .cm files throughout the source tree.
597    
598    ----------------------------------------------------------------------
599    Name: Matthias Blume (blume (at) tti - c (dot) org)
600    Date: 2004/10/15 15:45:00 CDT
601    Tag: blume-20041015-coverage
602    Description:
603    
604    * Test coverage tool added!
605    * Further reorganization of tracing-, debugging-, and profiling support:
606    
607        - moved original BTImp -- now called BackTrace -- into a separate
608          library called $/trace-debug-profile.cm
609        - eliminated all mentions of BTrace from SMLofNJ.Internals
610        - only the instrumentation mechanism is now left in the compiler proper
611        - BackTrace module is a plugin which is NOT plugged in by default
612        - Coverage module is another such plugin
613    
614    To get the benefits of any of these plugin modules, the code in
615    question must be compiled with tdp instrumentation turned on.  This
616    can be done by setting SMLofNJ.Internals.TDP.mode to true.
617    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
618    
619    Plugins are selected at link time.  (Pre-compiled instrumented code
620    can be re-loaded with different plugins in effect.)  When an
621    instrumented module is linked, whatever plugins are at that time
622    enabled will come into effect for that module.
623    
624    To enable the back-trace plugin, load library $/trace-debug-profile.cm
625    and invoke BackTrace.install() (e.g., from the interactive prompt).
626    To enable the coverage plugin, load the same library and invoke
627    Coverage.install().
628    
629    Back-traces are generated automatically on uncaught exceptions and
630    when the code in question explicitly invokes BackTrace.trigger().
631    
632    Coverage (and execution frequency-) information must be queried
633    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
634    
635    ----------------------------------------------------------------------
636    Name: Matthias Blume (blume (at) tti - c (dot) org)
637    Date: 2004/10/14 17:40:00 CDT
638    Tag: blume-20041014-tdp-core
639    Description:
640    
641    Snapshot of a significant overhaul of how the trace/debug/profile support
642    is hooked into the system (specifically: Core and SMLofNJ.Internals).
643    
644    ----------------------------------------------------------------------
645    Name: Matthias Blume (blume (at) tti - c (dot) org)
646    Date: 2004/10/13 16:34:00 CDT
647    Tag: blume-20041013-tdp
648    Description:
649    
650    Some rationalization of names:
651    
652       structure BTrace -> structure TDPInstrument
653       etc.
654    
655    This is is preparation of using the original back-trace
656    instrumentation for other purposes.  "TDP" stands for
657    Trace/Debug/Profile.
658    
659    The control flag controlling whether instrumentation is on or off is now
660    registered under a different name, so instead of running sml as
661    
662     sml -Cinstrument.btrace-mode=true
663    
664    one has to say
665    
666     sml -Ctdp.instrument=true
667    
668    ----------------------------------------------------------------------
669    Name: Matthias Blume (blume (at) tti - c (dot) org)
670    Date: 2004/10/11 16:37:00 CDT
671    Tag: blume-20041011-regions
672    Description:
673    
674    Made some minor modifications to elabcore.sml to have source regions
675    be propagated more tightly -- resulting in better (i.e., smaller)
676    regions being reported in error- and debug messages.
677    
678    ----------------------------------------------------------------------
679    Name: Matthias Blume (blume (at) tti - c (dot) org)
680    Date: 2004/10/08 22:50:00 CDT
681    Tag: blume-20041008-cmkw
682    Description:
683    
684    Fixed handling of keywords in .cm files: After seeing "is" the lexer
685    treats subsequent occurrences of "group", "library", "source", "is",
686    "*", and "-" as ordinary identifiers rather than keywords.
687    
688    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
689    is supposed to accept a tool argument called "source", but this did
690    not work because of the clash with the keyword.
691    
692    ----------------------------------------------------------------------
693    Name: Matthias Blume (blume (at) tti - c (dot) org)
694    Date: 2004/10/07 16:00:00 CDT
695    Tag: blume-20041007-cleanup
696    Description:
697    
698    Assorted cleanup work:
699    
700      - got rid of intstrmap in favor of using the library's
701        hash table implementation
702      - threw out most of the pathnames stuff, as it was not used anyway
703      - simplified tokentable implementation
704      - fixed some minor spelling errors
705    
706    ----------------------------------------------------------------------
707    Name: Matthias Blume (blume (at) tti - c (dot) org)
708    Date: 2004/10/06 15:15:15 CDT
709    Tag: blume-20041006-handler
710    Description:
711    
712    Cleaned up the absyn to reflect the invariant that HANDLE always
713    carries a FNexp as part of the type definition.  This eliminates some
714    superfluous sanity checks at runtime down the road.
715    
716    Some minor cleanup of the btrace code.
717    
718    ----------------------------------------------------------------------
719    Name: Matthias Blume (blume (at) tti - c (dot) org)
720    Date: 2004/10/01 10:20:30 CDT
721    Tag: blume-20041001-slave
722    Description:
723    
724    Added hack to make slave mode work in the presence of the version
725    tool.  (Still, since the master does two passes over the code for
726    CMB.make, the release number gets bumped twice when slaves are
727    attached.  I don't know if this is worth fixing...)
728    
729    ----------------------------------------------------------------------
730    Name: Matthias Blume (blume (at) tti - c (dot) org)
731    Date: 2004/09/30 10:55:00 CDT
732    Tag: blume-20040930-version
733    Description:
734    
735    * Moved the "version" magic into its own little library under
736      src/system/smlnj/internal.  This avoids expensive reconstruction of
737      a stable src/compiler/core.cm.
738    
739    * At the same time, structure CompilerVersion is now known as
740      structure SMLNJVersion.
741    
742    * Arranged for the version tool to NOT kick in when rebuilding the system
743      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
744      Also, loading the versiontool does not work when rebuilding the system
745      because CM is not properly initialized at that time.
746    
747    ----------------------------------------------------------------------
748    Name: Matthias Blume (blume (at) tti - c (dot) org)
749    Date: 2004/09/29 14:00:00 CDT
750    Tag: blume-20040929-autoversion
751    Description:
752    
753    Implemented some CM magic to have
754        file src/compiler/TopLevel/main/version.sml
755    generated automagically.
756    The version is taken from two files: config/version and config/release.
757    The first is expected to contain a two-part version number such as 110.49.
758    The second should contain a single number, but it may be missing.
759    
760    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
761    time the version tool is loaded (which is the first time you say CMB.make),
762    then the tool will increment the value stored in config/release every
763    time CMB.make is invoked.
764    
765    The binfile format is now insensitive to anything beyond the first
766    two components of a version number, so bumping the release does not render
767    binfiles incompatible.  Auto-bumping can be used to keep track of versions
768    during development without invalidating existing binfiles.
769    
770    In any case, every CMB.make updates the date information in version.sml.
771    (This is the date that is printed in the banner.)
772    
773    ----------------------------------------------------------------------
774    Name: Matthias Blume (blume (at) tti - c (dot) org)
775    Date: 2004/09/28 10:53:00 CDT
776    Tag: blume-20040928-controls
777    Description:
778    
779    Some cleanup of the controls code.
780    
781    ----------------------------------------------------------------------
782    Name: Matthias Blume (blume (at) tti - c (dot) org)
783    Date: 2004/09/27 22:08:00 CDT
784    Tag: blume-20040927-controls
785    Description:
786    
787    Added two pieces of functionality to the Controls interface:
788    
789    1.   val save'restore: 'a control -> unit -> unit
790    
791       grabs the current value of the control in stage 1 and restores it
792       in stage 2.
793    
794    2.   val set' : 'a control * 'a -> unit -> unit
795    
796       stores the given value into the control in stage 2 (i.e., delayed)
797       but does all error checking in stage 1.
798       (This is for string controls that need to do parse their argument --
799        something that might fail.  In some cases, notably in CM, one
800        already knows the intended argument but wants to delay the actual
801        assignment until a time when error recovery would be more difficult.)
802    
803    Changed the handling of controls in tool arguments to classes "sml" and
804    "lazysml":
805       - use Controls.save'restore as a more robust way of restoring the
806         old value (in particular: without having to re-parse the string)
807       - use controls to handle the "overload" keyword in the init group
808         (I believe this change actually fixes a long-standing obscure bug.)
809    
810    ----------------------------------------------------------------------
811    Name: Matthias Blume (blume (at) tti - c (dot) org)
812    Date: 2004/09/27 17:00:00 CDT
813    Tag: blume-20040927-lazysml
814    Description:
815    
816    Added a new tool class called "lazysml" to CM's tool chest.  The only
817    difference to "sml" is that compilation is done with Control.lazysml
818    set to true.  A source of class "lazysml" is automatically recognized
819    by a file name suffix of ".lml".
820    
821    In addition to the above feature, the original class "sml" now also
822    supports a tool argument "lazy" which has the same effect.  As a
823    result, the following three lines are equivalent:
824    
825        foo.sml : lazysml
826        foo.sml : sml (lazy)
827        foo.sml (lazy)
828    
829    The setting goes into effect both during parsing and during
830    compilation.  The original setting is restored right after parsing and
831    after compilation, respectively.
832    
833    In addition to all the above, there is also a general mechanism to set
834    ANY of the "controls" that are available at the command line via
835    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
836    apply as well. (In fact, "lazy" is implemented as a special case of
837    the general mechanism.)
838    
839    The .cm file syntax uses a new keyword tool argument called "with".
840    There are several ways of indicating the desired settings:
841    
842        foo.sml (with:parser.quotations=true)
843        foo.sml (with:(name:parser.quotations value:true))
844        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
845        foo.sml (with:(name1=value1 name2=value2 ...))
846        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
847    
848        etc.
849    
850    Another possible abbreviation is to leave out the =v or value:v part
851    if the name refers to a boolean control (in which case the value is
852    taken to be true).  Thus, one could get lazy sml also by saying:
853    
854        foo.sml (with:parser.lazy-keyword=true)
855        foo.sml (with:parser.lazy-keyword)
856        foo.sml (with:(name:parser.lazy-keyword value:true))
857        foo.sml (with:(name:parser.lazy-keyword))
858    
859    ----------------------------------------------------------------------
860    Name: Matthias Blume (blume (at) tti - c (dot) org)
861    Date: 2004/09/24 16:22:00 CDT
862    Tag: blume-20040924-ppc-long-branch
863    Description:
864    
865    Turned message about "emiting long form of branch" off by default.
866    Added a control flag to turn it back on when desired.
867    
868    ----------------------------------------------------------------------
869    Name: Matthias Blume (blume (at) tti - c (dot) org)
870    Date: 2004/09/24 16:05:00 CDT
871    Tag: blume-20040924-rounding
872    Description:
873    
874    Applied patch for setting rounding modes under Mac OS X.  Thanks to
875    Melissa O'Neill for providing the code!
876    
877    ----------------------------------------------------------------------
878    Name: Matthias Blume (blume (at) tti - c (dot) org)
879    Date: 2004/09/23 17:30:00 CDT
880    Tag: blume-20040923-envvars
881    Description:
882    
883    1. Changed definition of type ControlRegistry.registry_tree to
884       include control_info (i.e., the name of the controlling environment
885       variable).
886    
887    2. Added command-line flags -e and -E to print the names of environment
888       variables that can be used to control internal settings.
889       (This uses the new API mentioned in 1.)
890    
891    ----------------------------------------------------------------------
892    Name: Matthias Blume (blume (at) tti - c (dot) org)
893    Date: 2004/09/13 16:50:00 CDT
894    Tag: Release_110_49
895    Description:
896    
897    New working version (110.49).  NEW BOOTFILES!
898    
899    ----------------------------------------------------------------------
900    Name: Matthias Blume
901    Date: 2004/09/13 16:20:00 CDT
902    Tag: blume-20040913-config-mlrisc
903    Description:
904    
905    Put target "mlrisc" back into the default list.
906    (There is no harm in having it, and some users have expressed their
907    wish to have "mlrisc" included by default.)
908    
909    ----------------------------------------------------------------------
910    Name: John Reppy
911    Date: 2004/09/13
912    Tag: jhr-20040913-signals
913    Description:
914    
915    Fixed the signal masking code to properly nest mask/unmask operations
916    on a per-signal basis.
917    
918    ----------------------------------------------------------------------
919    Name: Matthias Blume (blume (at) tti - c (dot) org)
920    Date: 2004/09/08 13:20:00 CDT
921    Tag: blume-20040908-heap-magic
922    Description:
923    
924    Bumped the heap macig number to 0x09082004 to account for the changed
925    layout of the ML frame under MacOS X.
926    
927    ----------------------------------------------------------------------
928    Name: Allen Leung (leunga (at) reservoir (dot) com)
929    Date: 2004/09/03 11:26:00 EST
930    Tag: leunga-20040903-cygwin-install
931    Description:
932    
933       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
934    The Cygwin runtime is turned on by setting the environment
935    variable SMLNJ_CYGWIN_RUNTIME to 1.
936    
937    ----------------------------------------------------------------------
938    Name: Matthias Blume (blume (at) tti - c (dot) org)
939    Date: 2004/08/31 17:15:00 CDT
940    Tag: blume-20040831-core
941    Description:
942    
943    Added some exports to src/compiler/core.cm upon request by J. Joemann.
944    
945    ----------------------------------------------------------------------
946    Name: Matthias Blume (blume (at) tti - c (dot) org)
947    Date: 2004/08/30 17:55:00 CDT
948    Tag: blume-20040830-installer
949    Description:
950    
951    Upon request by Johannes Joemann:
952    
953    - improved ML code of installer to fall back to coping when renaming
954      fails (i.e., when source and target are on different file systems);
955      the code compiles but has yet to be tested in anger
956    - removed mlrisc from list of default targets (config/targets)
957    
958    ----------------------------------------------------------------------
959    Name: Matthias Blume (blume (at) tti - c (dot) org)
960    Date: 2004/08/27 17:20:00 CDT
961    Tag: blume-20040827-ptreql
962    Description:
963    
964    Added ptreql primop to structure InlineT (upon request from Larry
965    Paulson).
966    
967    ----------------------------------------------------------------------
968    Name: Allen Leung (leunga (at) reservoir (dot) com)
969    Date: 2004/08/15 21:21:00 EST
970    Tag: leunga-110_48-udgraph
971    Description:
972    
973    Another bug fix from Carl Hauser:
974    
975    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
976    > 48c48
977    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
978    > e::L)
979    > ---
980    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
981    > Without this, any deletion of an edge in an undirected graph does severe
982    > violence to the graph.
983    
984    ----------------------------------------------------------------------
985    Name: Allen Leung (leunga (at) reservoir (dot) com)
986    Date: 2004/08/10 23:35:00 EST
987    Tag: leunga-110_48-ppc
988    Description:
989    
990       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
991    
992    ----------------------------------------------------------------------
993    Name: Matthias Blume (blume (at) tti - c (dot) org)
994    Date: 2004/08/10 12:00:00 CDT
995    Tag: Release_110_48
996    Description:
997    
998    New working version (110.48).  NEW BOOTFILES!
999    
1000    ----------------------------------------------------------------------
1001    Name: Allen Leung (leunga (at) reservoir (dot) com)
1002    Date: 2004/08/09 12:21:00 EST
1003    Tag: leunga-110_47-dijsktra
1004    Description:
1005    
1006    Bug fix from Carl Hauser:
1007    
1008    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1009    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1010    
1011    The problem is that following the expression A.update(dist,s,Num.zero)
1012    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1013    
1014    ----------------------------------------------------------------------
1015    Name: Matthias Blume (blume (at) tti - c (dot) org)
1016    Date: 2004/08/06 18:10:00 CDT
1017    Tag: blume-20040806-cmdline
1018    Description:
1019    
1020    Fiddled with handling of command-line options:
1021    
1022      * sml now quits after processing the command line
1023        if -H, -S, -h<n>, or -s<n> appears as the last
1024        command-line argument
1025      * a new option -q terminates the session when encountered on
1026        the command line; subsequent arguments will be ignored
1027      * bug fixes: short (erroneous) arguments are no longer ignored
1028        completely
1029    
1030    ----------------------------------------------------------------------
1031    Name: Allen Leung (leunga (at) reservoir (dot) com)
1032    Date: 2004/08/04 18:17:00 EST
1033    Tag: leunga-110_47-ppc-ibm-asm
1034    Description:
1035    
1036      - Added minimal IBM assembly syntax support for PowerPC.
1037    
1038      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1039        been altered in the runtime.  We need an automatic way to keep the file
1040        in sync.
1041    
1042    ----------------------------------------------------------------------
1043    Name: Matthias Blume (blume (at) tti - c (dot) org)
1044    Date: 2004/08/04 14:00:00 CDT
1045    Tag: Release_110_47
1046    Description:
1047    
1048    New working version (110.47).  NEW BOOTFILES!
1049    
1050    ----------------------------------------------------------------------
1051    Name: Matthias Blume (blume (at) tti - c (dot) org)
1052    Date: 2004/08/03 14:25:00 CDT
1053    Tag: blume-20040803-callingconv
1054    Description:
1055    
1056    Added low-level support for choosing C calling conventions by
1057    twiddling the type of rawccall.  (See
1058    src/compiler/Semant/types/cproto.sml for details.)
1059    
1060    ----------------------------------------------------------------------
1061    Name: Matthias Blume (blume (at) tti - c (dot) org)
1062    Date: 2004/08/02 15:55:00 CDT
1063    Tag: blume-20040802-backout
1064    Description:
1065    
1066    Backed out of change to win32-filesys.c.  The earlier patch to
1067    get_file_time caused CM to produce files with the wrong time stamp.
1068    
1069    ----------------------------------------------------------------------
1070    Name: Matthias Blume (blume (at) tti - c (dot) org)
1071    Date: 2004/08/02 14:45:00 CDT
1072    Tag: blume-20040802-nlffi-win32
1073    Description:
1074    
1075    Added NLFFI support for Win32, adapted from a patch provided by David
1076    Hansel.  This is currently completely untested.  Also, the issue
1077    concerning stdcall vs. ccall is still unresolved.
1078    
1079    ----------------------------------------------------------------------
1080    Name: Matthias Blume (blume (at) tti - c (dot) org)
1081    Date: 2004/07/30 17:55:00 CDT
1082    Tag: blume-20040730-various
1083    Description:
1084    
1085    Gearing up towards 110.47...
1086    
1087    - various minor bugfixes to ml-nlffigen
1088    - a beginning of a manual for nlffi
1089    
1090    - eliminated 'export name=value' in config/install.sh as this does
1091      not work with certain versions of /bin/sh
1092      (Thanks to David King at Motorola for catching this.)
1093    
1094    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1095      - added a test for tm==NULL to gmtime.c and localtime.c
1096      - applied patch for incorrect GetFileTime under win32
1097      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1098    
1099    ----------------------------------------------------------------------
1100    Name: Matthias Blume (blume (at) tti - c (dot) org)
1101    Date: 2004/07/21 18:20:00 CDT
1102    Tag: blume-20040721-nlffigen
1103    Description:
1104    
1105    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1106      for a typedef to an incomplete type, but leave out the "typ" member.
1107      (This is just for consistency.)
1108    - Started to produce what is supposed to become better (i.e., comprehensive)
1109      documentation of what ml-nlffigen does and produces.
1110    
1111    ----------------------------------------------------------------------
1112    Name: Matthias Blume (blume (at) tti - c (dot) org)
1113    Date: 2004/07/14 16:25:00 CDT
1114    Tag: blume-20040714-union
1115    Description:
1116    
1117    Added C_UNION to c-calls/c-types.sml and updated the machinery
1118    (ml-nlffigen, cproto.sml) that conveys C function interface
1119    information to the code generator.
1120    
1121    However, the actual architecture-specific implementation of function
1122    arguments and results that are C unions is still not implemented.
1123    
1124    ----------------------------------------------------------------------
1125    Name: Allen Leung (leunga (at) reservoir (dot) com)
1126    Date: 2004/07/14 14:38:00 EST
1127    Tag: leunga-110_46_1-ppc-lwzu
1128    Description:
1129    
1130        Added these instructions to the PowerPC architecture:
1131        LBZU(X), LHZU(X), LWZU(X),
1132        STWU(X), STFDU, STFSU
1133    
1134        etc...
1135    
1136        Note: I haven't added their instruction encoding into the description.
1137    
1138    ----------------------------------------------------------------------
1139    Name: Allen Leung (leunga (at) reservoir (dot) com)
1140    Date: 2004/07/13 15:04:00 EST
1141    Tag: leunga-110_46_1-ppc-lwarx
1142    Description:
1143    
1144       Added the two instructions LWARX and STWCX to the PowerPC
1145    instruction set.
1146    
1147       A (untested) rewrite of loop-structure.sml.  The old version
1148    is completely broken.
1149    
1150    ----------------------------------------------------------------------
1151    Name: Matthias Blume (blume (at) tti - c (dot) org)
1152    Date: 2004/07/13 13:50:00 CDT
1153    Tag: blume-20040713-nlffi
1154    Description:
1155    
1156    - use paramAlloc to report c-calls with too many arguments
1157      (for PPC version where parameter area is pre-allocated)
1158    - added ccall_maxargspace to machspec (to implement the above)
1159    - made "make" commend in CM's "make" tool configurable
1160    - added option (default: on) for passing the name of the SML/NJ's "bin"
1161      directory to "make"; the call looks like this:
1162    
1163         make <options> SMLNJ_BINDIR=<dir> <target>
1164    
1165      This can be used by the Makefile to, e.g., pick the "right" version
1166      of ml-nlffigen.
1167    - minor code tweaks
1168    
1169    ----------------------------------------------------------------------
1170    Name: Matthias Blume (blume (at) tti - c (dot) org)
1171    Date: 2004/07/12 22:50:00 CDT
1172    Tag: blume-110_46_1-macosx-nlffi
1173    Description:
1174    
1175    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1176    though.
1177    
1178    Note:  1. You have to make a new, clean build of the runtime system.
1179           2. There are new BOOTFILES, you have to use them!
1180              (Doing the bootstrap process yourself would be *very* painful!
1181               If you absolutely have to do it, build the system under
1182               a different architecture and then cross-compile.)
1183    
1184    Version bumped to 110.46.1 to account for runtime data format changes.
1185    
1186    ----------------------------------------------------------------------
1187    Name: Matthias Blume (blume (at) tti - c (dot) org)
1188    Date: 2004/06/18 14:30:00 CDT
1189    Tag: blume-20040618-unix
1190    Description:
1191    
1192    Changed the implementation of structure Unix so that the same stream
1193    is returned every time one of the {text,bin}{In,Out}streamOf functions
1194    is invoked on the same proc.  This is not what the spec currently
1195    says -- although IMO it arguably should.  (See discussion below.)
1196    
1197    ----------------------------------------------------------------------
1198    Name: Matthias Blume (blume (at) tti - c (dot) org)
1199    Date: 2004/06/17 18:15:00 CDT
1200    Tag: Release_110_46
1201    Description:
1202    
1203    New working version (110.46).  NEW BOOTFILES!
1204    
1205    ----------------------------------------------------------------------
1206    Name: Matthias Blume (blume (at) tti - c (dot) org)
1207    Date: 2004/06/17 17:20:00 CDT
1208    Tag: blume-20040617-timer-unix
1209    Description:
1210    
1211    Changed the interface of structures Timer and Unix to match the most
1212    recent Basis spec.
1213    
1214    In the case of Unix there still seems to be an open/weird issue:
1215    
1216       The {text,bin}{In,Out}streamOf functions are supposed to create
1217       fresh streams whenever they are called -- as opposed to have them
1218       return the same stream every time.  This design is supposed to
1219       prevent space leaks caused by proc values hanging on to streams.
1220    
1221       The reap function, on the other hand, is supposed to close the
1222       streams.  This cannot be done without having a handle on the
1223       stream in proc after all...
1224    
1225    I took the liberty to implement the following stopgap solution:
1226    
1227      The proc value hangs on to the most recently created stream(s).
1228      Reap closes those.  If either or both of the two streams hadn't
1229      been created at all yet, then reap will close the corresponding
1230      file descriptors directly.
1231    
1232    PS: I don't understand the original space leak argument anymore.  If
1233    a proc hangs on to the imperative stream, then I/O operations on those
1234    will advance the state of the cached stream and avoid the space leak.
1235    
1236    ----------------------------------------------------------------------
1237    Name: Matthias Blume (blume (at) tti - c (dot) org)
1238    Date: 2004/05/28 16:45:00 CDT
1239    Tag: blume-20040528-basis
1240    Description:
1241    
1242    Added signature PACK_REAL and exported functor PrimIO.
1243    
1244    ----------------------------------------------------------------------
1245    Name: Matthias Blume (blume (at) tti - c (dot) org)
1246    Date: 2004/05/25 16:00:00 CDT
1247    Tag: blume-20040525-group-owner
1248    Description:
1249    
1250    CM now ignores (but still accepts) the "owner" information in group
1251    descriptions.  The owner of a group is its next enclosing
1252    library. Each group must have a unique owner.  (There is a virtual
1253    "toplevel" library that owns groups which are not nested within a real
1254    library.)  Previously, each group had to explicitly declare its owner,
1255    and CM would check that such a declaration is correct.  The new scheme
1256    is to have CM check that for each group there is precisely one owning
1257    library.
1258    
1259    The advantage of the new scheme is that the programmer no longer needs
1260    to maintain the somewhat annoying owner information.  The downside is
1261    that CM cannot enforce the ownership rule across multiple runs of
1262    CM.make.  Fortunately, enclosing the same group in two different
1263    libraries A and B which are not part of the same program does not
1264    cause real problems.
1265    
1266    ----------------------------------------------------------------------
1267    Name: Matthias Blume (blume (at) tti - c (dot) org)
1268    Date: 2004/05/20 16:00:00 CDT
1269    Tag: blume-20040520-win32
1270    Description:
1271    
1272    Made the win32 version work again.  (Strangely, a misplaced comma had
1273    slipped into win32-process.c which prevented the runtime from being
1274    compiled correctly.)
1275    
1276    Also, included a minor addition to ml-build.bat analogous to what was
1277    done in blume-20040519-ml-build.
1278    
1279    ----------------------------------------------------------------------
1280    Name: Matthias Blume (blume (at) tti - c (dot) org)
1281    Date: 2004/05/19 22:10:00 CDT
1282    Tag: blume-20040519-ml-build
1283    Description:
1284    
1285    Arranged for ml-build to clean up after itself a little bit better.
1286    The script generates a temporary SML source file and compiles it using
1287    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1288    gets rid of those at the end, so they don't accumulate under .cm.
1289    
1290    This required a minor change to install.sh because the name of the
1291    metadata directory (default: .cm) is actually configurable at
1292    installation time.
1293    
1294    ----------------------------------------------------------------------
1295    Name: Matthias Blume (blume (at) tti - c (dot) org)
1296    Date: 2004/05/18 15:50:00 CDT
1297    Tag: blume-20040518-mkreader
1298    Description:
1299    
1300    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1301    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1302    
1303    ----------------------------------------------------------------------
1304    Name: Matthias Blume (blume (at) tti - c (dot) org)
1305    Date: 2004/05/11 14:35:00 CDT
1306    Tag: blume-20040511-win32sock
1307    Description:
1308    
1309    Added previously missing support for many socket-related functions
1310    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1311    for the voluminous patch!
1312    
1313    (I have not tested this patch under win32 yet.)
1314    
1315    Here is David's e-mail:
1316    
1317    Hi,
1318    
1319    Attached to this email you find a diff against sml/nj 110.45
1320    that will enable socket support under Windows.
1321    
1322    To apply the patch (using unix or cygwin)
1323    1) gunzip runtime.diff.gz
1324    2) "cd" into "src/runtime" in the source tree of a fresh
1325       110.45 installation.
1326    3) patch -p 1 < [your/path/to]runtime.diff
1327    
1328    The code compiles fine but has NOT yet been extensively tested.
1329    I only ran a few tests for basic socket client functionality
1330    (which worked fine).  Especially the functions that use ioctl
1331    are not tested at all and might not work (see below).
1332    
1333    I implemented this since we want to move to a newer version of sml/nj
1334    but need socket support in order to use it.  This is the first time I
1335    even had a look at the sml/nj source,  so please review my changes
1336    before making this part of the distribution!  Here are a few issues
1337    that I think might be better for someone to solve who is more
1338    familiar with the sml/nj source (and socket programming):
1339    
1340    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1341      exception since I could not figure out what the windows equivalent
1342      of these functions is
1343    
1344    - In sockets-osdep.h there are a some #include statements that are
1345      only used in a few files that include sockets-osdep.h
1346    
1347    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1348      does not process its return value since I don't know how to
1349      report an error upwards.
1350    
1351    - It would probably be good to have a call to WSACleanup() when
1352      the library is unloaded (if there is such a possibility).
1353      Otherwise I think Windows will take care of this automatically
1354      when the process finishes.
1355    
1356    - I used ioctlsocket() as a replacement for ioctl() but I have
1357      no idea if that is actually the proper replacement on Windows.
1358    
1359    - All these issues are marked in the code by "FIXME" comments.
1360    
1361    We use sml/nj extensively in our products and are quite happy
1362    with it.  I hope this contribution will help you.
1363    
1364    Keep up the good work!
1365    
1366    David
1367    
1368    ----------------------------------------------------------------------
1369    Name: Matthias Blume (blume (at) tti - c (dot) org)
1370    Date: 2004/05/11 14:20:00 CDT
1371    Tag: blume-20040511-installml
1372    Description:
1373    
1374    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1375    the patch.)
1376    
1377    ----------------------------------------------------------------------
1378    Name: Matthias Blume (blume (at) tti - c (dot) org)
1379    Date: 2004/05/11 14:05:00 CDT
1380    Tag: blume-20040511-nlffi-netbsd
1381    Description:
1382    
1383    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1384    the patch.)
1385    
1386    ----------------------------------------------------------------------
1387    Name: Matthias Blume (blume (at) tti - c (dot) org)
1388    Date: 2004/05/11 12:05:00 CDT
1389    Tag: blume-20040511-exports
1390    Description:
1391    
1392    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1393    export lists in compiler-related .cm-files.
1394    
1395    ----------------------------------------------------------------------
1396    Name: Matthias Blume (blume (at) tti - c (dot) org)
1397    Date: 2004/05/11 11:35:00 CDT
1398    Tag: blume-20040511-allsource
1399    Description:
1400    
1401    The installer now honors the "src-smlnj" target again, although its meaning
1402    has changed from "all sources required for the compiler" to "all sources
1403    the installer knows about".  In other words, if you enable "src-smlnj"
1404    in the "targets" file, then the installer will pull in sources for
1405    everything.  (Notice that this refers to source code only.  Compiled
1406    code is still only installed for modules that were requested explicitly
1407    or which are required for other modules that were requested explicitly.)
1408    
1409    ----------------------------------------------------------------------
1410    Name: Matthias Blume (blume (at) tti - c (dot) org)
1411    Date: 2004/04/23 17:40:00 CDT
1412    Tag: blume-20040423-ieee-scan
1413    Description:
1414    
1415    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1416    in the exponent calculation we get INF or ZERO (depending on the mantissa
1417    and the sign of the exponent).
1418    
1419    ----------------------------------------------------------------------
1420    Name: Matthias Blume (blume (at) tti - c (dot) org)
1421    Date: 2004/04/23 10:40:00 CDT
1422    Tag: blume-20040423-ml-build
1423    Description:
1424    
1425    The ml-build script now terminates with a non-0 status when something
1426    goes wrong.
1427    
1428    ----------------------------------------------------------------------
1429    Name: Matthias Blume (blume (at) tti - c (dot) org)
1430    Date: 2004/04/22 16:35:00 CDT
1431    Tag: blume-20040422-Option
1432    Description:
1433    
1434    Made exception Option to be the same as exception Option.Option
1435    (as it should be).
1436    
1437    ----------------------------------------------------------------------
1438    Name: Allen Leung (leunga (at) reservoir (dot) com)
1439    Date: 2004/03/19 14:40:00 EST
1440    Tag: leunga-20040319-cygwin-nlffi
1441    Description:
1442    
1443    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1444    of SML/NJ.  The problem is that
1445    
1446       lib = dlopen(NULL, ...)
1447       f   = dlsym(lib, "malloc");
1448    
1449    does not work on Windows unless we explicitly export symbols
1450    such as 'malloc' during linking.  We fixed this by explicitly
1451    exporting the required symbols with the magic gcc incantation:
1452    
1453        -Wl,--export-all cygwin.def
1454    
1455    where cygwin.def is a file containing all the symbols that
1456    we wish to export.
1457    
1458    I suspect this is a Windows problem and we'll have to
1459    do the same (somehow with windows compilers) when
1460    we build the native win32 version with the system
1461    calls LoadLibrary/GetProcAddress.
1462    
1463    ----------------------------------------------------------------------
1464    Name: Matthias Blume (blume (at) tti - c (dot) org)
1465    Date: 2004/03/04 16:35:00 CST
1466    Tag: blume-20040304-intinf-fmt
1467    Description:
1468    
1469    Fixed problem with IntInf.fmt (sign would show up on the right instead
1470    of on the left for BIN, OCT, and HEX).
1471    
1472    ----------------------------------------------------------------------
1473    Name: Matthias Blume (blume (at) tti - c (dot) org)
1474    Date: 2004/03/04 11:25:00 CST
1475    Tag: blume-20040304-symlinks
1476    Description:
1477    
1478    Fixed problem with installer script (unix only) where bin/ml-yacc and
1479    friends pointed (via symlinks) to absolute locations instead of just
1480    .run-sml.  This was reported by Vesa A Norrman.
1481    
1482    ----------------------------------------------------------------------
1483    Name: Matthias Blume (blume (at) tti - c (dot) org)
1484    Date: 2004/02/13 14:50:00 CST
1485    Tag: Release_110_45
1486    Description:
1487    
1488    New working version (110.45). New bootfiles.
1489    
1490    ----------------------------------------------------------------------
1491    Name: Matthias Blume (blume (at) tti - c (dot) org)
1492    Date: 2004/01/26 15:15:15 CST
1493    Tag: blume-20040126-toplevel
1494    Description:
1495    
1496    Improved handling of exceptions at the interactive toplevel.
1497    
1498    ----------------------------------------------------------------------
1499    Name: Matthias Blume (blume (at) tti - c (dot) org)
1500    Date: 2004/01/26 11:25:00
1501    Tag: blume-20040126-app
1502    Description:
1503    
1504    Type of top-level "app" corrected.
1505    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1506    
1507    ----------------------------------------------------------------------
1508    Name: Matthias Blume (blume (at) tti - c (dot) org)
1509    Date: 2003/11/18 17:10 CST
1510    Tag: blume-20031118-basis-fiddle
1511    Description:
1512    
1513    - changed Timer interface to what might become the spec
1514    - POSIX_FLAGS -> BIT_FLAGS according to spec
1515    - some other minor discrepancies wrt. spec eliminated
1516    
1517    ----------------------------------------------------------------------
1518    Name: Matthias Blume (blume (at) tti - c (dot) org)
1519    Date: 2003/11/06 12:00:00 CST
1520    Tag: Release_110_44
1521    Description:
1522    
1523    New working version (110.44). New bootfiles.
1524    
1525    ----------------------------------------------------------------------
1526    Name: Matthias Blume (blume (at) tti - c (dot) org)
1527    Date: 2003/11/04 11:50:00 CST
1528    Tag: blume-20031104-move-libraries
1529    Description:
1530    
1531    Eliminated the "dont_move_libraries" directive in config/targets.
1532    (The mechanism was broken and could not be fixed easily.  Moreover,
1533    there does not seem to be any reason not to move all libraries into
1534    lib during installation.  I originally implemented this directive as a
1535    backward-compatibility feature when I first introduced the new CM.
1536    Now that things have been stable for a long time and going back to the
1537    old CM is not an option, there is no reason to keep it around.)
1538    
1539    ----------------------------------------------------------------------
1540    Name: Matthias Blume (blume (at) tti - c (dot) org)
1541    Date: 2003/11/03 16:00:00 CST
1542    Tag: blume-20031103-installdir
1543    Description:
1544    
1545    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1546    Richards for pointing out the problem and providing the solution.)
1547    
1548    ----------------------------------------------------------------------
1549    Name: Matthias Blume (blume (at) tti - c (dot) org)
1550    Date: 2003/10/01 17:05:00 CDT
1551    Tag: blume-20031001-lal-mlrisc
1552    Description:
1553    
1554    MLRISC bug fix from Lal.
1555    
1556    ----------------------------------------------------------------------
1557    Name: Matthias Blume (blume (at) tti - c (dot) org)
1558    Date: 2003/09/30 16:10:00 CDT
1559    Tag: blume-20030930-primio-bat
1560    Description:
1561    
1562    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1563    2. Improved .bat files (for Win32 port) to make things work under Win95.
1564       (thanks to Aaron S. Hawley for this one)
1565    
1566    ----------------------------------------------------------------------
1567    Name: Matthias Blume (blume (at) tti - c (dot) org)
1568    Date: 2003/09/26 16:05:00 CDT
1569    Tag: blume-20030926-wrappriv
1570    Description:
1571    
1572    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1573    
1574    ----------------------------------------------------------------------
1575    Name: Matthias Blume (blume (at) tti - c (dot) org)
1576    Date: 2003/09/26 15:00:00 CDT
1577    Tag: blume-20030926-110_43_3
1578    Description:
1579    
1580    - additional cleanup
1581    - version number bump, NEW BOOTFILES
1582    
1583    ----------------------------------------------------------------------
1584    Name: Matthias Blume (blume (at) tti - c (dot) org)
1585    Date: 2003/09/26 12:00:00 CDT
1586    Tag: blume-20030926-ppautoload
1587    Description:
1588    
1589    I modified the read-eval-print loop so that the autoloader gets
1590    invoked whenever the prettyprinter tries to look up a symbol that
1591    is not currently defined in the toplevel environment but which
1592    appears in CM's autoload registry.  As a result, we see far fewer of
1593    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1594    
1595    In addition to this I tried to clean up some pieces of the Basis
1596    implementation (e.g., Socket, Word8Array) in order to prevent other
1597    instances of these ?.Foo.Bar.xxx names from being printed.
1598    
1599    The mechanism that picks names for types still needs some work, though.
1600    (Right now it seems that if there is a type A.t which is defined to
1601    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1602    "?.B.u" although the perhaps more sensible solution would be to use
1603    "A.t" in this case.  In other words, the prettyprinter should follow
1604    a chain of DEFtycs not farther than there are corresponding toplevel
1605    names in the current environment.)
1606    
1607    ----------------------------------------------------------------------
1608    Name: Matthias Blume (blume (at) tti - c (dot) org)
1609    Date: 2003/09/24 16:31:00 CDT
1610    Tag: blume-20030924-installer
1611    Description:
1612    
1613    Another installer tweak:  All the ML code for the installer is now
1614    compiled during CMB.make and put into a little library called
1615    $smlnj/installer.cm.  The installation then simply invokes
1616    
1617       sml -m $smlnj/installer.cm
1618    
1619    and everything happens automagically.
1620    
1621    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1622    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1623            and BINDIR.
1624    
1625    The new scheme guarantees that the ML code responsible for the installation
1626    is in sync with the APIs of the main system.  Also, the installer is
1627    somewhat faster because the installer script is precompiled.
1628    
1629    ----------------------------------------------------------------------
1630    Name: Matthias Blume (blume (at) tti - c (dot) org)
1631    Date: 2003/09/24 15:35:00 CDT
1632    Tag: blume-20030924-synsock
1633    Description:
1634    
1635    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1636    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1637    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1638    implementations of non-blocking operations from CML's Socket
1639    structure.
1640    
1641    ----------------------------------------------------------------------
1642    Name: Matthias Blume (blume (at) tti - c (dot) org)
1643    Date: 2003/09/24 15:10:05 CDT
1644    Tag: blume-20030924-sockets
1645    Description:
1646    
1647    1. Fixed SOCKET API and implementation to match Basis spec.
1648       This required changing the internal representation of sockets to one
1649       that remembers (for each socket file descriptor) whether it is currently
1650       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1651       call is made only if the state actually needs to change).
1652    
1653    2. OS-specific details of sockets were moved into separate files, thus
1654       making it possible to unify the bulk of the socket implementations
1655       between Unix and Win32.
1656    
1657    3. CML's socket API changed accordingly.
1658       (Note that we need to remove non-blocking functions from this API
1659       since they are redundant in the case of CML!)
1660    
1661    4. CML's socket implementation now makes use of non-blocking functions
1662       provided by Basis, thus removing all OS-dependent code from this part
1663       of CML.
1664    
1665    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1666    
1667    ----------------------------------------------------------------------
1668    Name: Matthias Blume (blume (at) tti - c (dot) org)
1669    Date: 2003/09/22 12:10:00 CDT
1670    Tag: blume-20030922-110_43_2
1671    Description:
1672    
1673    Made a new interim version and bootfiles for developer's bootstrapping
1674    convenience.
1675    
1676       110.43.2 -- NEW BOOTFILES
1677    
1678    ----------------------------------------------------------------------
1679    Name: Matthias Blume (blume (at) tti - c (dot) org)
1680    Date: 2003/09/19 15:55:00 CDT
1681    Tag: blume-20030919-cmdir
1682    Description:
1683    
1684    1. new-install.sh -> install.sh
1685    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1686    3. tweaked installer so that another name instead of .cm can be chosen
1687       at install time (by setting the CM_DIR_ARC environment variable
1688       during installation); once installation is complete, the name is
1689       fixed
1690    
1691    ----------------------------------------------------------------------
1692    Name: Matthias Blume (blume (at) tti - c (dot) org)
1693    Date: 2003/09/18 16:00:00 CDT
1694    Tag: blume-20030918-110_43_1
1695    Description:
1696    
1697    Made a new interim version and bootfiles for developer's bootstrapping
1698    convenience.
1699    
1700       110.43.1 -- NEW BOOTFILES
1701    
1702    ----------------------------------------------------------------------
1703    Name: Matthias Blume (blume (at) tti - c (dot) org)
1704    Date: 2003/09/18 15:20:00 CDT
1705    Tag: blume-20030918-misc
1706    Description:
1707    
1708    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1709       this seems to be controversial at the moment) and used those in
1710       Posix.ProcEnv.times.
1711    
1712    2. Added Time.{from,to}Nanoseconds to Time.
1713    
1714    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1715       For example, fromLargeInt never needs to look at more than 3 "big
1716       digits" to get its 53 bits of precision.
1717    
1718    ----------------------------------------------------------------------
1719    Name: Matthias Blume (blume (at) tti - c (dot) org)
1720    Date: 2003/09/17 16:30:00 CDT
1721    Tag: blume-20030917-real32-slices
1722    Description:
1723    
1724    Added an entry to the primitive environment
1725    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1726    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1727    
1728    Removed some of the "magic" constants in real64.sml and replaced them
1729    with code that generates these values from their corresponding
1730    integer counterparts.
1731    
1732    Made all(?) the slice-related changes to the Basis and made everything
1733    compile again...
1734    
1735    ----------------------------------------------------------------------
1736    Name: Matthias Blume (blume (at) tti - c (dot) org)
1737    Date: 2003/09/15 17:45:00 CDT
1738    Tag: blume-20030915-rbase
1739    Description:
1740    
1741    Fixed bug in Real.fromLargeInt.
1742    
1743    ----------------------------------------------------------------------
1744    Name: Matthias Blume (blume (at) tti - c (dot) org)
1745    Date: 2003/09/13 18:11:00 CDT
1746    Tag: blume-20030913-libinstall
1747    Description:
1748    
1749    Minor bugfix in config/libinstall (set anchor with path to
1750    standalone tool after installing it, otherwise libraries that
1751    need ml-lex or ml-yacc won't compile the first time the installer
1752    runs).
1753    
1754    ----------------------------------------------------------------------
1755    Name: Matthias Blume (blume (at) tti - c (dot) org)
1756    Date: 2003/09/12 11:45:00 CDT
1757    Tag: blume-20030912-various
1758    Description:
1759    
1760     - fixed bug in Real.toLargeInt
1761     - fixed bug in Posix.ProcEnv.times
1762     - changed inputLine functions to return an option
1763     - minor installer improvements / bugfixes
1764     - changed default @SMLalloc parameter for x86/celeron to 64k
1765    
1766    ----------------------------------------------------------------------
1767    Name: Matthias Blume (blume (at) tti - c (dot) org)
1768    Date: 2003/09/09 22:00:00 CDT
1769    Tag: Release_110_43
1770    Description:
1771    
1772        New working release 110.43.  New bootfiles.
1773    
1774    ----------------------------------------------------------------------
1775    Name: Matthias Blume (blume (at) tti - c (dot) org)
1776    Date: 2003/09/09 19:20:00 CDT
1777    Tag: blume-20030909-installer
1778    Description:
1779    
1780    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1781    Modified config/install.bat to take advantage of it.  Also modified
1782    config/install.sh (and called it config/new-install.sh) to take advantage
1783    of it on Unix systems.  (The SML code is (supposed to be) platform-
1784    independent.)
1785    
1786    The installer can now install everything under Win32
1787    as well as under *nix as long as it compiles.
1788    
1789    Other changes:
1790    
1791      - made CML compile again under Win32
1792      - made eXene compile under Win32 (by providing a fake structure UnixSock
1793        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1794      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1795        same as type int;  under Win32 it isn't
1796      - fixed some slice-related problems in the win32-specific parts of CML
1797      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1798        (under Win32, the volume name is case-insensitive, and the
1799        OS.Path code compares volume names for equality)
1800    
1801    ----------------------------------------------------------------------
1802    Name: Matthias Blume (blume (at) tti - c (dot) org)
1803    Date: 2003/09/08 11:55:00 CDT
1804    Tag: blume-20030908-fullpath
1805    Description:
1806    
1807    Made Win32 version of OS.FileSys.fullPath return current directory
1808    when given an empty string.  This is what the spec says, and incidentally,
1809    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1810    cases when presented with the name of a non-existing .cm file.)
1811    
1812    ----------------------------------------------------------------------
1813    Name: Matthias Blume (blume (at) tti - c (dot) org)
1814    Date: 2003/09/04 16:30:00 CDT
1815    Tag: blume-20030905-slices-etc
1816    Description:
1817    
1818      1. Changed interface to vectors and arrays in Basis to match
1819         (draft) Basis spec.
1820      2. Added signatures and implementations of slices according to
1821         Basis spec.
1822      3. Edited source code throughout the system to make it compile again
1823         under 1. and 2.  (In some cases code had to be added to have it
1824         match the new signatures.)
1825      4. MLRISC should be backward-compatible: the copies of the originals
1826         of files that needed to change under 3. were retained, the .cm files
1827         check the compiler version number and use old versions when
1828         appropriate.
1829      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1830         dirstream -> string option (in accordance with Basis spec).
1831      6. When generating code that counts lines, ml-lex used function
1832         CharVector.foldli, taking advantage of its old interface.
1833         This has been replaced with the corresponding code from
1834         CharVectorSlice.  (html-lex must be re-lexed!)
1835      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1836         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1837      8. Removed temporary additions (fromInternal, toInternal) from the
1838         (now obsolete) IntInf in smlnj-lib/Util.
1839      9. Cleaned up structure Byte.
1840     10. Added localOffset, scan, and fromString to Date (according to spec).
1841         Cleaned/corrected implementation of Date.
1842         (Still need to check for correctness; implement better canonicalizeDate.)
1843     11. Added "scan" to signature IEEE_REAL.
1844     12. Some improvements to IntInf [in particular: efficiency-hack for
1845         mod and rem when second operand is 2 (for parity checks).]
1846     13. Changed representation of type Time.time, using a single IntInf.int
1847         value counting microseconds.  This considerably simplified the
1848         implementation of structure Time.  We now support negative time
1849         values; scan and fromString handle signs.
1850     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1851         ArraySlice).
1852    
1853    ----------------------------------------------------------------------
1854    Name: Matthias Blume (blume (at) tti - c (dot) org)
1855    Date: 2003/08/28 17:00:00 CDT
1856    Tag: blume-20030828-intinf
1857    Description:
1858    
1859    This is a major update which comes with a version number bump
1860    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1861    and an implementation of IntInf in the Basis.
1862    
1863    There are a fairly large number of related changes and updates throughout
1864    the system:
1865    
1866    Basis:
1867            - Implemented IntInf.
1868            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1869            - Added some missing Real64 operations, most notably Real.toLargeInt.
1870            - Added FixedInt as a synonym for Int32.
1871    
1872    compiler:
1873            * Added support for a built-in intinf type.
1874              - literals
1875              - pattern matching
1876              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1877              - overloading on literals and operations
1878    
1879            This required adding a primitive type intinf, some additional
1880            primops, and implementations for several non-trivial intinf
1881            operations in Core.  (The intinf type is completely abstract
1882            to the compiler; all operations get delegated back to the Core.)
1883    
1884            * Intinf equality is handled by polyequal.  However, the compiler
1885            does not print its usual warning in this case (since polyequal
1886            is the right thing to do there).
1887    
1888            * Improved the organization of structure InlineT.
1889    
1890            * A word about conversion primops:
1891            If conversions involving intinf do not cancel out during
1892            CPS contract, then the compiler must insert calls to Core functions.
1893            Since all core access must be resolved already during the FLINT
1894            translate phase, it would be too late a the time of CPS contract
1895            to add new Core calls.  For this reason, conversion primops
1896            for intinf carry two arguments:  1. the numeric argument that
1897            they are supposed to convert, and 2. the Core function that
1898            can help with this conversion if necessary.  If CPS contract
1899            eliminates a primop, then the associated Core function becomes
1900            dead and goes away.  Intinf conversion primops that do not get
1901            eliminated by CPS contract get rewritten into calls of their
1902            core functions by a separate, new phase.
1903    
1904    interactive system:
1905            - Control.Print.intinfDepth controls max length of intinf constants
1906            being printed. (Analogous to Control.Print.stringDepth.)
1907            - Cleanup in printutil and pputil: got rid of unused stuff and
1908            duplicates; replaced some of the code with code that makes better
1909            use of library functionality.
1910    
1911    CM:
1912            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1913            are no longer silent.
1914    
1915    CKIT:
1916            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1917            in favor of LargeInt -- which is now the same as IntInf.
1918            CKIT-knowledgable people should check whether this is what's
1919            intended and otherwise change things back to using Int32 or
1920            FixedInt.
1921    
1922    Throughout the code:
1923            Started using IntInf.int literals and built-in operations
1924            (e.g., comparison with 0) where this seems appropriate.
1925    
1926    
1927    ----------------------------------------------------------------------
1928    Name: Dave MacQueen (dbm@cs.uchicago.edu)
1929    Date: 2003/08/13 11:36:00 CDT
1930    Tag: dbm-20030813-mcz-merge1
1931    Description:
1932    
1933    Merging changes from the mcz-branch development branch into trunk.
1934    These changes involve replacement of the emulated old prettyprinter
1935    interface with direct use of the SML/NJ Lib PP library, and fixing
1936    of a couple of bugs (895, 1186) relating to error messages.  A new
1937    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1938    has been added.
1939    
1940    ----------------------------------------------------------------------
1941    Name: Matthias Blume (blume (at) tti - c (dot) org)
1942    Date: 2003/08/11 15:45:00 CDT
1943    Tag: blume-20030811-windows
1944    Description:
1945    
1946    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
1947    
1948        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
1949    
1950    This patch restores SML/NJ's ability to run under win32.  There are a
1951    number of changes, including fixes for several bugs that had gone
1952    unnoticed until now:
1953    
1954      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
1955        trying to run the win32 version from within cygwin.)
1956    
1957      - There are a number of simple .bat scripts that substitute for their
1958        corresponding Unix shell-scripts.  (See below.)
1959    
1960      - The internals of ml-build have been modified slightly.  The main
1961        difference is that instead of calling ".link-sml" (or link-sml.bat)
1962        using OS.Process.system, the ML process delegates this task back
1963        to the script.  Otherwise problems arise in mixed environments such
1964        as Cygwin where scripts look and work like Unix scripts, but
1965        where OS.Process.system cannot run them.
1966    
1967      - In CM, the srcpath pickler used native pathname syntax -- which
1968        is incorrect in the case of cross-compilation.  The new pickle format
1969        is independent of platform-specific naming conventions.
1970    
1971      - Path configuration files (such as lib/pathconfig) can now choose
1972        between native and standard syntax.  Placing a line of the form
1973    
1974            standard!
1975    
1976        into the file causes all subsequent paths to be interpreted using
1977        CM standard pathname syntax (= Unix conventions); a line
1978    
1979            native!
1980    
1981        switches back to native style.  This was needed so that
1982        path config files can be written portably, see src/system/pathconfig.
1983    
1984      - Runtime system:
1985    
1986          - win32-filesys.c:  get_file_time and set_file_time now
1987            access modification time, not creation time.
1988    
1989          - I/O code made aware of new array representation.
1990    
1991          - Bug fixes in X86.prim.masm.
1992    
1993       - src/system/makeml made aware of win32. (For use under cygwin
1994            and other Unix-environments for windows.)
1995    
1996       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
1997         which caused BinIO.inputAll to fail consistently.
1998    
1999    .bat scripts:
2000    
2001       Windows .bat scripts assume that SMLNJ_HOME is defined.
2002    
2003       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2004         applications (sml, ml-yacc, ml-lex).
2005       - ml-build.bat: analogous to ml-build.
2006       - config\install.bat:  Analogous to config/install.sh.  This requires
2007         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2008         (nmake etc. must be on the path, and vcvars32 must have been run.)
2009         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2010         and the bootfile hierarchy must have been unpacked under
2011         sml.boot.x86-win32.
2012         The script is very primitive and does a poor job at error checking.
2013         It only installs the base system, ml-lex, and ml-yacc.  No other
2014         libraries are being installed (i.e., you get only those that
2015         are part of the compiler.)
2016       - link-sml.bat:  analogous to .link-sml, but not currently used
2017    
2018    Unrelated bug fixes:
2019    
2020      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2021        types.
2022      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2023        otherwise.)
2024    
2025    ----------------------------------------------------------------------
2026    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2027    Date: 2003/06/17
2028    Tag: macqueen-20030617-bug895
2029    Description:
2030    
2031    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2032    Tag will be used for new development branch (mcz-branch) for
2033    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2034    
2035    ----------------------------------------------------------------------
2036    Name: Matthias Blume (blume (at) tti - c (dot) org)
2037    Date: 2003/05/27 16:55:00 CDT
2038    Tag: blume-20030527-polyeq
2039    Description:
2040    
2041    Tried to eliminated most cases of polymorphic equality.
2042    
2043    ----------------------------------------------------------------------
2044    Name: Matthias Blume (blume (at) tti - c (dot) org)
2045    Date: 2003/05/21 17:45:00 CDT
2046    Tag: blume-20030517-complete
2047    Description:
2048    
2049    Two changes:
2050    
2051      1. Added a flag for controlling whether non-exhaustive bindings will
2052         be treated as errors (default is false).
2053      2. Cleaned up the *entire* source tree so that CMB.make goes through
2054         without a single non-exhaustive match- or bind warning.
2055    
2056    ----------------------------------------------------------------------
2057    Name: Matthias Blume (blume (at) tti - c (dot) org)
2058    Date: 2003/05/17 10:20:00 CDT
2059    Tag: blume-20030517-absyn
2060    Description:
2061    
2062    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2063    
2064       This mainly affects the quality of error messages.  However, some
2065       of the code is now more straightforward than before. (Treatment of
2066       the above four constructs in translate.sml is much simpler than
2067       the "macro-expansion" that was going on before.  Plus, the mach-
2068       compiler no longer gets invoked just to be able to compile an
2069       if-expression.)
2070    
2071    2. The ErrorMsg.Error exception is now caught and absorbed by the
2072       interactive loop.
2073    
2074    ----------------------------------------------------------------------
2075    Name: Allen Leung
2076    Date: 2003/05/16 13:05:00 CDT
2077    Tag: leunga-20030516-cygwin-runtime
2078    Description:
2079    
2080        Ported the runtime system to cygwin, which uses the unix
2081    x86-unix bin files.  Missing/buggy features:
2082    
2083       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2084         the Cygwin library.
2085       o Ctrl-C handling may be flaky.
2086       o Windows system calls and Windows I/O are not supported.
2087    
2088       A new set of binfiles is located at:
2089    
2090         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2091    
2092       This is only needed for bootstrapping the cygwin version of smlnj.
2093       Other x86 versions can use the existing binfiles.
2094    
2095    ----------------------------------------------------------------------
2096    Name: Matthias Blume
2097    Date: 2003/04/08 15:42:00 CDT
2098    Tag: blume-20030408-listpair
2099    Description:
2100    
2101    1. Added a target 'mlrisc' to installer.
2102    
2103    2. Added missing elements to structure ListPair.
2104    
2105    ----------------------------------------------------------------------
2106    Name: Allen Leung
2107    Date: 2003/01/07 10:40:00 EST
2108    Tag: leunga-20030107-int-rem
2109    Description:
2110    
2111    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2112    The arguments to the SUBL instruction were swapped.
2113    
2114    ----------------------------------------------------------------------
2115    Name: Matthias Blume
2116    Date: 2002/12/12 16:25:00 EST
2117    Tag: blume-20021212-risc-ra
2118    Description:
2119    
2120    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2121    sent the RA into an infinite loop when floating point registers get
2122    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2123    110.37 and 110.38.)
2124    There was another set of potential problems related to the handling of
2125    MLRISC annotations (but those did not yet cause real problems, apparently).
2126    
2127    ----------------------------------------------------------------------
2128    Name: Matthias Blume
2129    Date: 2002/12/06 22:40:00 EST
2130    Tag: blume-20021206-cm-fileid
2131    Description:
2132    
2133    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2134    This fixes the problem of CM getting confused by files that suddenly
2135    change their identity (e.g., by getting unlinked and recreated by some
2136    text editor such as vi).  There might be a better/cheaper/cleaner way
2137    of doing this, but for now this will have to do.
2138    
2139    ----------------------------------------------------------------------
2140    Name: Matthias Blume
2141    Date: 2002/10/28 09:50:00 EST
2142    Tag: blume-20021028-typecheck
2143    Description:
2144    
2145    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2146    
2147    ----------------------------------------------------------------------
2148    Name: Matthias Blume
2149    Date: 2002/10/17 09:10:00 EDT
2150    Tag: Release_110_42
2151    Description:
2152    
2153    In good old tradition, there has been a slight hiccup so that we have
2154    to patch 110.42 after the fact.   The old release tag has been replaced
2155    (see below).
2156    
2157    The change solves a problem with two competing approaches the
2158    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2159    each other's way.
2160    
2161    This change only affects the runtime system code and the installer script.
2162    (No new bootfiles.)
2163    
2164    ----------------------------------------------------------------------
2165    Name: Matthias Blume
2166    Date: 2002/10/16 12:00:00 EDT
2167    Tag: Release_110_42_removed
2168    Description:
2169    
2170    New working release.  New bootfiles.
2171    
2172    ----------------------------------------------------------------------
2173    Name: Matthias Blume
2174    Date: 2002/10/10 13:10:00 EDT
2175    Tag: blume-20021010-ppc-divs
2176    Description:
2177    
2178    The mltree operator DIVS must be implemented with an overflow check on
2179    the PPC because the hardware indicates divide-by-zero using "overflow" as
2180    well.
2181    
2182    ----------------------------------------------------------------------
2183    Name: Matthias Blume
2184    Date: 2002/07/23 11:45:00
2185    Tag: blume-20020723-smlnj-home
2186    Description:
2187    
2188    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2189    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2190    CM_PATHCONFIG is also set), the path configuration file is assumed
2191    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2192    the entire tree to some other place and everything will "just work".
2193    
2194    (Companion commands such as ml-build and ml-makedepend also sense this
2195    variable.)
2196    
2197    ----------------------------------------------------------------------
2198    Name: Matthias Blume
2199    Date: 2002/07/12 21:19:00 EDT
2200    Tag: blume-20020712-liveness
2201    Description:
2202    
2203    Exported two useful "step" functions from liveness module (MLRISC).
2204    
2205    ----------------------------------------------------------------------
2206    Name: Matthias Blume
2207    Date: 2002/07/05 16:00 EDT
2208    Tag: Release_110_41
2209    Description:
2210    
2211    New working release.  New bootfiles.
2212    
2213    ----------------------------------------------------------------------
2214    Name: Matthias Blume
2215    Date: 2002/07/05 10:25:00 EDT
2216    Tag: blume-20020705-btimp
2217    Description:
2218    
2219    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2220    other clients can set up backtracing support.
2221    
2222    ----------------------------------------------------------------------
2223    Name: Matthias Blume
2224    Date: 2002/06/25 17:23:00 EDT
2225    Tag: blume-20020625-fpmax
2226    Description:
2227    
2228    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2229    case.  (The sense of the isNaN test was reversed -- which made min and
2230    max always return their first argument.)
2231    
2232    ----------------------------------------------------------------------
2233    Name: Matthias Blume
2234    Date: 2002/06/11
2235    Tag: blume-20020611-unixpath
2236    Description:
2237    
2238    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2239    
2240    ----------------------------------------------------------------------
2241    Name: Matthias Blume
2242    Date: 2002/06/10 16:35:00 EDT
2243    Tag: blume-20020610-ieeereal
2244    Description:
2245    
2246    I back-ported my implementation of IEEEReal.fromString from the
2247    idlbasis-devel branch so that we can test it.
2248    
2249    Another small change is that ppDec tries to give more information
2250    than just "<sig>" in the case of functors.  However, this code is
2251    broken in some mysterious way if the functor's body's signature
2252    has not been declared by ascription but gets inferred from the
2253    implementation.  This needs fixing...
2254    
2255    ----------------------------------------------------------------------
2256    Name: Matthias Blume
2257    Date: 2002/05/31
2258    Tag: blume-20020531-btrace-mode
2259    Description:
2260    
2261    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2262    the wayside when I switched over to using Controls everywhere.)
2263    
2264    ----------------------------------------------------------------------
2265    Name: Lal George
2266    Date: 2002/05/23 12:21:40 EDT
2267    Tag: george-20020523-visual-labels
2268    Description:
2269    
2270    Labels are now displayed in the graphical output to make
2271    the fall-through and target blocks obvious.
2272    
2273    ----------------------------------------------------------------------
2274    Name: Matthias Blume
2275    Date: 2002/05/22 11:03:00 EDT
2276    Tag: blume-20020522-shrink
2277    Description:
2278    
2279    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2280    situation that comes up when allocating huge arrays.
2281    
2282    ----------------------------------------------------------------------
2283    Name: Matthias Blume
2284    Date: 2002/05/21 16:00:00 EDT
2285    Tag: Release_110_40
2286    Description:
2287    
2288    New working release (110.40).  New bootfiles.
2289    
2290    [Also: John Reppy fixed GC bug 1131.]
2291    
2292    ----------------------------------------------------------------------
2293    Name: Matthias Blume
2294    Date: 2002/05/21 12:35:00 EDT
2295    Tag: blume-20020521-cmdoc
2296    Description:
2297    
2298    CM documentation update.
2299    
2300    ----------------------------------------------------------------------
2301    Name: Matthias Blume
2302    Date: 2002/05/21 10:55:00 EDT
2303    Tag: blume-20020521-misc
2304    Description:
2305    
2306    - John tweaked runtime to be silent on heap export (except when
2307      GC messages are on).
2308    - I added a few more things (cross-compiling versions of CMB) to
2309      config/preloads (as suggestions).
2310    
2311    ----------------------------------------------------------------------
2312    Name: Matthias Blume
2313    Date: 2002/05/20 22:25:00 EDT
2314    Tag: blume-20020520-controls
2315    Description:
2316    
2317    - Added ControlUtil structure to control-lib.cm.
2318    - Use it throughout.
2319    - Used Controls facility to define MLRISC controls (as opposed to
2320      registering MLRISC control ref cells with Controls after the fact)
2321    - Fixed messed-up controls priorities.
2322    
2323    * Removed again all the stuff from config/preloads that one wouldn't
2324      be able to preload at the time the initial heap image is built.
2325      (Many libraries, e.g., CML, do not exist yet at this time.  The
2326       only libraries that can be preloaded via config/preloads are those
2327       that come bundled with the bootfiles.)
2328    
2329    ----------------------------------------------------------------------
2330    Name: Matthias Blume
2331    Date: 2002/05/20 10:59:00 EDT
2332    Tag: blume-20020520-preloads
2333    Description:
2334    
2335    Added a lot of commented-out suggestions for things to be included
2336    in config/preloads.
2337    
2338    ----------------------------------------------------------------------
2339    Name: Allen Leung
2340    Date: 2002/05/18 14:20:00 EDT
2341    Tag: leunga-20020518-mdl
2342    Description:
2343    
2344    o Made the mdl tool stuff compile and run again.
2345    o I've disabled all the stuff that depends on RTL specifications; they
2346      are all badly broken anyway.
2347    
2348    ----------------------------------------------------------------------
2349    Name: Matthias Blume
2350    Date: 2002/05/17 16:49:00 EDT
2351    Tag: blume-20020517-controls
2352    Description:
2353    
2354    0. John Reppy made several modifications to the SML/NJ library.
2355       In particular, there is a shiny new controls-lib.cm.
2356    
2357    1. Pushed new controls interface through compiler so that everything
2358       compiles again.
2359    
2360    2. Added FormatComb and FORMAT_COMB to the CML version of the
2361       SML/NJ library (so that CML compiles again).
2362    
2363    3. Modified init scripts because XXX_DEFAULT environment variables
2364       are no longer with us.  (Boot-time initialization is now done
2365       using the same environment variables that are also used for
2366       startup-time initialization of controls.)
2367    
2368    ----------------------------------------------------------------------
2369    Name: Lal George
2370    Date: 2002/05/15 09:20:10 EDT
2371    Tag: george-20020515-pseudo-op-decls
2372    Description:
2373    
2374            All pseudo-ops emitted before the first segment declaration
2375    such as TEXT, DATA, and BSS directives are assumed to be global
2376    declarations and are emitted first in the assembly file. This is
2377    useful in a number of situations where one has pseudo-ops that are not
2378    specific to any segment, and also works around the constraint that one
2379    cannot have client pseudo-ops in the TEXT segment.
2380    
2381            Because no segment is associated with these declarations it is
2382    an error to allocate any space or objects before the first segment
2383    directive and an exception will be raised. However, we cannot make
2384    this check for client pseudo-ops.
2385    
2386            These top level declarations are a field in the CFG graph_info.
2387    In theory you can continue to add to this field after the CFG has been
2388    built -- provided you know what you are doing;-)
2389    
2390    ----------------------------------------------------------------------
2391    Name: Matthias Blume
2392    Date: 2002/05/13 16:40:00 EDT
2393    Tag: blume-20020513-pp-etc
2394    Description:
2395    
2396    A few minor bugfixes:
2397    
2398      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2399        (Bogus printouts for redefined bindings still occur.  Compiler
2400        bug should no longer occur now.  We need to redo the prettyprinter
2401        from scratch.)
2402    
2403      - CM pathname printer now also adds escape sequences for ( and )
2404    
2405      - commend and docu fixes for ml-nlffi
2406    
2407    ----------------------------------------------------------------------
2408    Name: Matthias Blume
2409    Date: 2002/05/10 16:40:00 EDT
2410    Tag: blume-20020510-erg-textio
2411    Description:
2412    
2413    Applied the following bugfix provided by Emden Gansner:
2414    
2415        Output is corrupted when outputSubstr is used rather than output.
2416    
2417        The problem occurs when a substring
2418    
2419            ss = (s, dataStart, dataLen)
2420    
2421        where dataStart > 0, fills a stream buffer with avail bytes left.
2422        avail bytes of s, starting at index dataStart, are copied into the
2423        buffer, the buffer is flushed, and then the remaining dataLen-avail
2424        bytes of ss are copied into the beginning of the buffer. Instead of
2425        starting this copy at index dataStart+avail in s, the current code
2426        starts the copy at index avail.
2427    
2428        Fix:
2429        In text-io-fn.sml, change line 695 from
2430             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2431        to
2432             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2433    
2434    ----------------------------------------------------------------------
2435    Name: Matthias Blume
2436    Date: 2002/04/12 13:55:00 EDT
2437    Tag: blume-20020412-assyntax
2438    Description:
2439    
2440    1. Grabbed newer assyntax.h from the XFree86 project.
2441    2. Fiddled with how to compile X86.prim.asm without warnings.
2442    3. (Very) Minor cleanup in CM.
2443    
2444    ----------------------------------------------------------------------
2445    Name: Matthias Blume
2446    Date: 2002/04/01 (no joke!) 17:07:00 EST
2447    Tag: blume-20020401-x86div
2448    Description:
2449    
2450    Added full support for div/mod/rem/quot on the x86, using the machine
2451    instruction's two results (without clumsily recomputing the remainder)
2452    directly where appropriate.
2453    
2454    Some more extensive power-of-two support was added to the x86 instruction
2455    selector (avoiding expensive divs, mods, and muls where they can be
2456    replaced with cheaper shifts and masks).  However, this sort of thing
2457    ought to be done earlier, e.g., within the CPS optimizer so that
2458    all architectures benefit from it.
2459    
2460    The compiler compiles to a fixed point, but changes might be somewhat
2461    fragile nevertheless.  Please, report any strange things that you might
2462    see wrt. div/mod/quot/rem...
2463    
2464    ----------------------------------------------------------------------
2465    Name: Matthias Blume
2466    Date: 2002/03/29 17:22:00
2467    Tag: blume-20020329-div
2468    Description:
2469    
2470    Fixed my broken div/mod logic.  Unfortunately, this means that the
2471    inline code for div/mod now has one more comparison than before.
2472    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2473    The problem was with quotient = 0, because that alone does not tell
2474    us which way the rounding went.  One then has to look at whether
2475    remainder and divisor have the same sign...  :(
2476    
2477    Anyway, I replaced the bootfiles with fresh ones...
2478    
2479    ----------------------------------------------------------------------
2480    Name: Matthias Blume
2481    Date: 2002/03/29 14:10:00 EST
2482    Tag: blume-20020329-inlprims
2483    Description:
2484    
2485    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2486    
2487    Primops have changed. This means that the bin/boot-file formats have
2488    changed as well.
2489    
2490    To make sure that there is no confusion, I made a new version.
2491    
2492    
2493    CHANGES:
2494    
2495    * removed REMT from mltree (remainder should never overflow).
2496    
2497    * added primops to deal with divisions of all flavors to the frontend
2498    
2499    * handled these primops all the way through so they map to their respective
2500      MLRISC support
2501    
2502    * used these primops in the implementation of Int, Int32, Word, Word32
2503    
2504    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2505    
2506    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2507    
2508    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2509      floating point
2510    
2511    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2512      and Real (but Real.abs maps to a separate floating-point-only primop)
2513    
2514    
2515    TODO items:
2516    
2517    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2518      instructions because the machine instruction encoder cannot handle
2519      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2520      REM in the first place, and REM is supported by the x86 machine coder.)
2521    
2522    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2523      selection phase.  (The two can be streamlined because the hardware
2524      delivers both quotient and remainder at the same time anyway.)
2525    
2526    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2527      (Currently the behavior is inconsistent both across architectures and
2528      wrt. the draft Basis spec.)
2529    
2530    * Word8 should eventually be handled natively, too.
2531    
2532    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2533      as if there currently is no execution path that could trigger it in
2534      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2535      hold for things like multiplication and division.)
2536    
2537    ----------------------------------------------------------------------
2538    Name: Matthias Blume
2539    Date: 2002/03/27 16:27:00 EST
2540    Tag: blume-20020327-mlrisc-divisions
2541    Description:
2542    
2543    Added support for all four division operations (ML's div, mod, quot,
2544    and rem) to MLRISC.  In the course of doing so, I also rationalized
2545    the naming (no more annoying switch-around of DIV and QUOT), by
2546    parameterizing the operation by div_rounding_mode (which can be either
2547    DIV_TO_ZERO or DIV_TO_NEGINF).
2548    
2549    The generic MLTreeGen functor takes care of compiling all four
2550    operations down to only round-to-zero div.
2551    
2552    Missing pieces:
2553    
2554      * Doing something smarter than relying on MLTreeGen on architectures
2555        like, e.g., the x86 where hardware division delivers both quotient and
2556        remainder at the same time.  With this, the implementation of the
2557        round-to-neginf operations could be further streamlined.
2558    
2559      * Remove inlining support for div/mod/rem from the frontend and replace it
2560        with primops that get carried through to the backend.  Do this for all
2561        int and word types.
2562    
2563    ----------------------------------------------------------------------
2564    Name: Matthias Blume
2565    Date: 2002/03/25 17:25:00 EST
2566    Tag: blume-20020325-divmod
2567    Description:
2568    
2569    I improved (hopefully without breaking them) the implementation of Int.div,
2570    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2571    advantage of the following observations:
2572    
2573      Let  q = x quot y      r = x rem y
2574           d = x div  y      m = x mod y
2575    
2576    where "quot" is the round-to-zero version of integer division that
2577    hardware usually provides.  Then we have:
2578    
2579         r = x - q * y        where neither the * nor the - will overflow
2580         d = if q >= 0 orelse x = q * y then q else q - 1
2581                              where neither the * nor the - will overflow
2582         m = if q >= 0 orelse r = 0 then r else r + y
2583                              where the + will not overflow
2584    
2585    This results in substantial simplification of the generated code.
2586    The following table shows the number of CFG nodes and edges generated
2587    for
2588            fun f (x, y) = x OPER y
2589            (* with OPER \in div, mod, quot, rem *)
2590    
2591    
2592        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2593        --------------------------------------------------------
2594         div |         24 |         39 |         12 |         16
2595         mod |         41 |         71 |         12 |         16
2596        quot |          8 |         10 |          8 |         10
2597         rem |         10 |         14 |          8 |         10
2598    
2599    
2600    ----------------------------------------------------------------------
2601    Name: Matthias Blume
2602    Date: 2002/03/25 22:06:00 EST
2603    Tag: blume-20020325-cprotobug
2604    Description:
2605    
2606    Fixed a bug in cproto (c prototype decoder).
2607    
2608    ----------------------------------------------------------------------
2609    Name: Matthias Blume
2610    Date: 2002/03/25 16:00:00 EST
2611    Tag: blume-20020325-raw-primops
2612    Description:
2613    
2614    I did some cleanup to Allen's new primop code and
2615    replaced yesterday's bootfiles with new ones.
2616    (But they are stored in the same place.)
2617    
2618    ----------------------------------------------------------------------
2619    Name: Matthias Blume
2620    Date: 2002/03/24 22:40:00 EST
2621    Tag: blume-20020324-bootfiles
2622    Description:
2623    
2624    Made the bootfiles that Allen asked for.
2625    
2626    ----------------------------------------------------------------------
2627    Name: Allen Leung
2628    Date: 2002/03/23 15:50:00 EST
2629    Tag: leunga-20020323-flint-cps-rcc-primops
2630    Description:
2631    
2632      1. Changes to FLINT primops:
2633    
2634        (* make a call to a C-function;
2635         * The primop carries C function prototype information and specifies
2636         * which of its (ML-) arguments are floating point. C prototype
2637         * information is for use by the backend, ML information is for
2638         * use by the CPS converter. *)
2639      | RAW_CCALL of { c_proto: CTypes.c_proto,
2640                       ml_args: ccall_type list,
2641                       ml_res_opt: ccall_type option,
2642                       reentrant : bool
2643                     } option
2644       (* Allocate uninitialized storage on the heap.
2645        * The record is meant to hold short-lived C objects, i.e., they
2646        * are not ML pointers.  With the tag, the representation is
2647        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2648        *)
2649      | RAW_RECORD of {tag:bool,sz:int}
2650      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2651    
2652      2.  These CPS primops are now overloaded:
2653    
2654           rawload of {kind:numkind}
2655           rawstore of {kind:numkind}
2656    
2657          The one argument form is:
2658    
2659             rawload {kind} address
2660    
2661          The two argument form is:
2662    
2663             rawload {kind} [ml object, byte-offset]
2664    
2665      3. RAW_CCALL/RCC now takes two extra arguments:
2666    
2667         a. The first is whether the C call is reentrant, i.e., whether
2668            ML state should be saved and restored.
2669         b. The second argument is a string argument specifying the name of
2670            library and the C function.
2671    
2672         These things are currently not handled in the code generator, yet.
2673    
2674      4. In CProto,
2675    
2676         An encoding type of "bool" means "ml object" and is mapped into
2677         C prototype of PTR.  Note that "bool" is different than "string",
2678         even though "string" is also mapped into PTR, because "bool"
2679         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2680    
2681      5. Pickler/unpicker
2682    
2683         Changed to handle RAW_RECORD and newest RAW_CCALL
2684    
2685      6. MLRiscGen,
2686    
2687         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2688         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2689            in the file CodeGen/cpscompile/cps-c-calls.sml
2690    
2691      7. Added the conditional move operator
2692    
2693             condmove of branch
2694    
2695         to cps.  Generation of this is still buggy so it is currently
2696         disabled.
2697    
2698    ----------------------------------------------------------------------
2699    Name: Lal George
2700    Date: 2002/03/22 14:18:25 EST
2701    Tag: george-20020322-cps-branch-prob
2702    Description:
2703    
2704    Implemented the Ball-Larus branch prediction-heuristics, and
2705    incorporated graphical viewers for control flow graphs.
2706    
2707    Ball-Larus Heuristics:
2708    ---------------------
2709    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2710    
2711    By design it uses the Dempster-Shafer theory for combining
2712    probabilities.  For example, in the function:
2713    
2714        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2715    
2716    the ball-larus heuristics predicts that the n=0 is unlikely
2717    (OH-heuristic), and the 'then' branch is unlikely because of the
2718    RH-heuristic -- giving the 'then' branch an even lower combined
2719    probability using the Dempster-Shafer theory.
2720    
2721    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2722    probability of the 'then' branch because of the loop in the else
2723    branch.
2724    
2725    
2726    Graphical Viewing:
2727    ------------------
2728    I merely plugged in Allen's graphical viewers into the compiler. The
2729    additional code is not much. At the top level, saying:
2730    
2731            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2732    
2733    will display the graphical view of the control flow graph just before
2734    back-patching.  daVinci must be in your path for this to work. If
2735    daVinci is not available, then the default viewer can be changed
2736    using:
2737    
2738            Control.MLRISC.getString "viewer"
2739    
2740    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2741    course, these viewers must be in your path.
2742    
2743    The above will display the compilation unit at the level of clusters,
2744    many of which are small, boring, and un-interesting. Also setting:
2745    
2746            Control.MLRISC.getInt "cfg-graphical-view_size"
2747    
2748    will display clusters that are larger than the value set by the above.
2749    
2750    
2751    ----------------------------------------------------------------------
2752    Name: Matthias Blume
2753    Date: 2002/03/21 22:20:00 EST
2754    Tag: blume-20020321-kmp-bugfix
2755    Description:
2756    
2757    Changed the interface to the KMP routine in PreString and fixed
2758    a minor bug in one place where it was used.
2759    
2760    ----------------------------------------------------------------------
2761    Name: Allen Leung
2762    Date: 2002/03/21 20:30:00 EST
2763    Tag: leunga-20020321-cfg
2764    Description:
2765    
2766      Fixed a potential problem in cfg edge splitting.
2767    
2768    ----------------------------------------------------------------------
2769    Name: Allen Leung
2770    Date: 2002/03/21 17:15:00 EST
2771    Tag: leunga-20020321-x86-fp-cfg
2772    Description:
2773    
2774      1. Recoded the buggy parts of x86-fp.
2775    
2776         a. All the block reordering code has been removed.
2777            We now depend on the block placement phases to do this work.
2778    
2779         b. Critical edge splitting code has been simplified and moved into the
2780            CFG modules, as where they belong.
2781    
2782         Both of these were quite buggy and complex.  The code is now much, much
2783         simpler.
2784    
2785      2. X86 backend.
2786    
2787         a. Added instructions for 64-bit support.  Instruction selection for
2788            64-bit has not been committed, however, since that
2789            requires changes to MLTREE which haven't been approved by
2790            Lal and John.
2791    
2792         b. Added support for FUCOMI and FUCOMIP when generating code for
2793            PentiumPro and above.  We only generate these instructions in
2794            the fast-fp mode.
2795    
2796         c. Added cases for JP and JNP in X86FreqProps.
2797    
2798      3. CFG
2799    
2800         CFG now has a bunch of methods for edge splitting and merging.
2801    
2802      4. Machine description.
2803    
2804         John's simplification of MLTREE_BASIS.fcond broke a few machine
2805         description things:
2806    
2807         rtl-build.{sig,sml} and hppa.mdl fixed.
2808    
2809         NOTE: the machine description stuff in the repository is still broken.
2810               Again, I can't put my fixes in because that involves
2811               changes to MLTREE.
2812    
2813    ----------------------------------------------------------------------
2814    Name: Matthias Blume
2815    Date: 2002/03/20 15:55:00 EST
2816    Tag: blume-20020320-kmp
2817    Description:
2818    
2819    Implemented Knuth-Morris-Pratt string matching in PreString and used
2820    it for String.isSubstring, Substring.isSubstring, and
2821    Substring.position.
2822    
2823    (Might need some stress-testing.  Simple examples worked fine.)
2824    
2825    ----------------------------------------------------------------------
2826    Name: Matthias Blume
2827    Date: 2002/03/19 16:37:00 EST
2828    Tag: blume-20020319-witnesses
2829    Description:
2830    
2831    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2832    
2833    This implements a generic mechanism for changing constness qualifiers
2834    anywhere within big C types without resorting to outright "casts".
2835    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2836    modify the constness at the outermost level.)
2837    The implementation of "convert" is based on the idea of "witness"
2838    values -- values that are not used by the operation but whose types
2839    "testify" to their applicability.  On the implementation side, "convert"
2840    is simply a projection (returning its second curried argument).  With
2841    cross-module inlining, it should not result in any machine code being
2842    generated.
2843    
2844    ----------------------------------------------------------------------
2845    Name: Matthias Blume
2846    Date: 2002/03/15 16:40:00 EST
2847    Tag: blume-20020315-basis
2848    Description:
2849    
2850    Provided (preliminary?) implementations for
2851    
2852      {String,Substring}.{concatWith,isSuffix,isSubstring}
2853    
2854    and
2855    
2856      Substring.full
2857    
2858    Those are in the Basis spec but they were missing in SML/NJ.
2859    
2860    ----------------------------------------------------------------------
2861    Name: Matthias Blume
2862    Date: 2002/03/14 21:30:00 EST
2863    Tag: blume-20020314-controls
2864    Description:
2865    
2866    Controls:
2867    ---------
2868    
2869    1. Factored out the recently-added Controls : CONTROLS stuff and put
2870       it into its own library $/controls-lib.cm.  The source tree for
2871       this is under src/smlnj-lib/Controls.
2872    
2873    2. Changed the names of types and functions in this interface, so they
2874       make a bit more "sense":
2875    
2876          module -> registry
2877          'a registry -> 'a group
2878    
2879    3. The interface now deals in ref cells only.  The getter/setter interface
2880       is (mostly) gone.
2881    
2882    4. Added a function that lets one register an already-existing ref cell.
2883    
2884    5. Made the corresponding modifications to the rest of the code so that
2885       everything compiles again.
2886    
2887    6. Changed the implementation of Controls.MLRISC back to something closer
2888       to the original.  In particular, this module (and therefore MLRISC)
2889       does not depend on Controls.  There now is some link-time code in
2890       int-sys.sml that registers the MLRISC controls with the Controls
2891       module.
2892    
2893    CM:
2894    ---
2895    
2896      * One can now specify the lambda-split aggressiveness in init.cmi.
2897    
2898    ----------------------------------------------------------------------
2899    Name: Allen Leung
2900    Date: 2002/03/13 17:30:00 EST
2901    Tag: leunga-20020313-x86-fp-unary
2902    Description:
2903    
2904    Bug fix for:
2905    
2906    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2907    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2908    > - fun f(x,(y,z)) = Real.~ y;
2909    > [autoloading]
2910    > [autoloading done]
2911    >       fchsl   (%eax), 184(%esp)
2912    > Error: MLRisc bug: X86MCEmitter.emitInstr
2913    >
2914    > uncaught exception Error
2915    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2916    
2917    The problem was that the code generator did not generate any fp registers
2918    in this case, and the ra didn't know that it needed to run the X86FP phase to
2919    translate the pseudo fp instruction.   This only happened with unary fp
2920    operators in certain situations.
2921    
2922    ----------------------------------------------------------------------
2923    Name: Matthias Blume
2924    Date: 2002/03/13 14:00:00 EST
2925    Tag: blume-20020313-overload-etc
2926    Description:
2927    
2928    1. Added _overload as a synonym for overload for backward compatibility.
2929       (Control.overloadKW must be true for either version to be accepted.)
2930    
2931    2. Fixed bug in install script that caused more things to be installed
2932       than what was requested in config/targets.
2933    
2934    3. Made CM aware of the (_)overload construct so that autoloading
2935       works.
2936    
2937    ----------------------------------------------------------------------
2938    Name: Matthias Blume
2939    Date: 2002/03/12 22:03:00 EST
2940    Tag: blume-20020312-url
2941    Description:
2942    
2943    Forgot to update BOOT and srcarchiveurl.
2944    
2945    ----------------------------------------------------------------------
2946    Name: Matthias Blume
2947    Date: 2002/03/12 17:30:00 EST
2948    Tag: blume-20020312-version110392
2949    Description:
2950    
2951    Yet another version number bump (because of small changes to the
2952    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
2953    
2954    Changes:
2955    
2956      The new pid generation scheme described a few weeks ago was overly
2957      complicated.  I implemented a new mechanism that is simpler and
2958      provides a bit more "stability":  Once CM has seen a compilation
2959      unit, it keeps its identity constant (as long as you do not delete
2960      those crucial CM/GUID/* files).  This means that when you change
2961      an interface, compile, then go back to the old interface, and
2962      compile again, you arrive at the original pid.
2963    
2964      There now also is a mechanism that instructs CM to use the plain
2965      environment hash as a module's pid (effectively making its GUID
2966      the empty string).  For this, "noguid" must be specified as an
2967      option to the .sml file in question within its .cm file.
2968      This is most useful for code that is being generated by tools such
2969      as ml-nlffigen (because during development programmers tend to
2970      erase the tool's entire output directory tree including CM's cached
2971      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
2972      revert to the old, broken behavior of SML/NJ, but in specific cases
2973      where there is no danger of interface confusion, its use is ok
2974      (I think).
2975    
2976      ml-nlffigen by default generates "noguid" annotations.  They can be
2977      turned off by specifying -guid in its command line.
2978    
2979    ----------------------------------------------------------------------
2980    Name: Lal George
2981    Date: 2002/03/12 12 14:42:36 EST
2982    Tag: george-20020312-frequency-computation
2983    Description:
2984    
2985    Integrated jump chaining and static block frequency into the
2986    compiler. More details and numbers later.
2987    
2988    ----------------------------------------------------------------------
2989    Name: Lal George
2990    Date: 2002/03/11 11 22:38:53 EST
2991    Tag: george-20020311-jump-chain-elim
2992    Description:
2993    
2994    Tested the jump chain elimination on all architectures (except the
2995    hppa).  This is on by default right now and is profitable for the
2996    alpha and x86, however, it may not be profitable for the sparc and ppc
2997    when compiling the compiler.
2998    
2999    The gc test will typically jump to a label at the end of the cluster,
3000    where there is another jump to an external cluster containing the actual
3001    code to invoke gc. This is to allow factoring of common gc invocation
3002    sequences. That is to say, we generate:
3003    
3004            f:
3005               testgc
3006               ja   L1      % jump if above to L1
3007    
3008            L1:
3009               jmp L2
3010    
3011    
3012    After jump chain elimination the 'ja L1' instructions is converted to
3013    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3014    up being implemented in their long form (if L2 is far away) using:
3015    
3016            jbe     L3      % jump if below or equal to L3
3017            jmp     L2
3018         L3:
3019            ...
3020    
3021    
3022    For large compilation units L2  may be far away.
3023    
3024    
3025    ----------------------------------------------------------------------
3026    Name: Matthias Blume
3027    Date: 2002/03/11 13:30:00 EST
3028    Tag: blume-20020311-mltreeeval
3029    Description:
3030    
3031    A functor parameter was missing.
3032    
3033    ----------------------------------------------------------------------
3034    Name: Allen Leung
3035    Date: 2002/03/11 10:30:00 EST
3036    Tag: leunga-20020311-runtime-string0
3037    Description:
3038    
3039       The representation of the empty string now points to a
3040    legal null terminated C string instead of unit.  It is now possible
3041    to convert an ML string into C string with InlineT.CharVector.getData.
3042    This compiles into one single machine instruction.
3043    
3044    ----------------------------------------------------------------------
3045    Name: Allen Leung
3046    Date: 2002/03/10 23:55:00 EST
3047    Tag: leunga-20020310-x86-call
3048    Description:
3049    
3050       Added machine generation for CALL instruction (relative displacement mode)
3051    
3052    ----------------------------------------------------------------------
3053    Name: Matthias Blume
3054    Date: 2002/03/08 16:05:00
3055    Tag: blume-20020308-entrypoints
3056    Description:
3057    
3058    Version number bumped to 110.39.1.  NEW BOOTFILES!
3059    
3060    Entrypoints: non-zero offset into a code object where execution should begin.
3061    
3062    - Added the notion of an entrypoint to CodeObj.
3063    - Added reading/writing of entrypoint info to Binfile.
3064    - Made runtime system bootloader aware of entrypoints.
3065    - Use the address of the label of the first function given to mlriscGen
3066      as the entrypoint.  This address is currently always 0, but it will
3067      not be 0 once we turn on block placement.
3068    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3069      with entry points) from mlriscGen.
3070    
3071    ----------------------------------------------------------------------
3072    Name: Allen Leung
3073    Date: 2002/03/07 20:45:00 EST
3074    Tag: leunga-20020307-x86-cmov
3075    Description:
3076    
3077       Bug fixes for CMOVcc on x86.
3078    
3079       1. Added machine code generation for CMOVcc
3080       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3081       3. CMOVcc cannot have an immediate operand as argument.
3082    
3083    ----------------------------------------------------------------------
3084    Name: Matthias Blume
3085    Date: 2002/03/07 16:15:00 EST
3086    Tag: blume-20020307-controls
3087    Description:
3088    
3089    This is a very large but mostly boring patch which makes (almost)
3090    every tuneable compiler knob (i.e., pretty much everything under
3091    Control.* plus a few other things) configurable via both the command
3092    line and environment variables in the style CM did its configuration
3093    until now.
3094    
3095    Try starting sml with '-h' (or, if you are brave, '-H')
3096    
3097    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3098    implements the underlying generic mechanism.
3099    
3100    The interface to some of the existing such facilities has changed somewhat.
3101    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3102    (The getFoo interface is still there for backward-compatibility, but its
3103    use is deprecated.)
3104    
3105    The ml-build script passes -Cxxx=yyy command-line arguments through so
3106    that one can now twiddle the compiler settings when using this "batch"
3107    compiler.
3108    
3109    TODO items:
3110    
3111    We should go through and throw out all controls that are no longer
3112    connected to anything.  Moreover, we should go through and provide
3113    meaningful (and correct!) documentation strings for those controls
3114    that still are connected.
3115    
3116    Currently, multiple calls to Controls.new are accepted (only the first
3117    has any effect).  Eventually we should make sure that every control
3118    is being made (via Controls.new) exactly once.  Future access can then
3119    be done using Controls.acc.
3120    
3121    Finally, it would probably be a good idea to use the getter-setter
3122    interface to controls rather than ref cells.  For the time being, both
3123    styles are provided by the Controls module, but getter-setter pairs are
3124    better if thread-safety is of any concern because they can be wrapped.
3125    
3126    *****************************************
3127    
3128    One bug fix: The function blockPlacement in three of the MLRISC
3129    backpatch files used to be hard-wired to one of two possibilities at
3130    link time (according to the value of the placementFlag).  But (I
3131    think) it should rather sense the flag every time.
3132    
3133    *****************************************
3134    
3135    Other assorted changes (by other people who did not supply a HISTORY entry):
3136    
3137    1. the cross-module inliner now works much better (Monnier)
3138    2. representation of weights, frequencies, and probabilities in MLRISC
3139       changed in preparation of using those for weighted block placement
3140       (Reppy, George)
3141    
3142    ----------------------------------------------------------------------
3143    Name: Lal George
3144    Date: 2002/03/07 14:44:24 EST 2002
3145    Tag: george-20020307-weighted-block-placement
3146    
3147    Tested the weighted block placement optimization on all architectures
3148    (except the hppa) using AMPL to generate the block and edge frequencies.
3149    Changes were required in the machine properties to correctly
3150    categorize trap instructions. There is an MLRISC flag
3151    "weighted-block-placement" that can be used to enable weighted block
3152    placement, but this will be ineffective without block/edge
3153    frequencies (coming soon).
3154    
3155    
3156    ----------------------------------------------------------------------
3157    Name: Lal George
3158    Date: 2002/03/05 17:24:48 EST
3159    Tag: george-20020305-linkage-cluster
3160    
3161    In order to support the block placement optimization, a new cluster
3162    is generated as the very first cluster (called the linkage cluster).
3163    It contains a single jump to the 'real' entry point for the compilation
3164    unit. Block placement has no effect on the linkage cluster itself, but
3165    all the other clusters  have full freedom in the manner in which they
3166    reorder blocks or functions.
3167    
3168    On the x86 the typical linkage code that is generated is:
3169       ----------------------
3170            .align 2
3171       L0:
3172            addl    $L1-L0, 72(%esp)
3173            jmp     L1
3174    
3175    
3176            .align  2
3177       L1:
3178       ----------------------
3179    
3180    72(%esp) is the memory location for the stdlink register. This
3181    must contain the address of the CPS function being called. In the
3182    above example, it contains the address of  L0; before
3183    calling L1 (the real entry point for the compilation unit), it
3184    must contain the address for L1, and hence
3185    
3186            addl $L1-L0, 72(%esp)
3187    
3188    I have tested this on all architectures except the hppa.The increase
3189    in code size is of course negligible
3190    
3191    ----------------------------------------------------------------------
3192    Name: Allen Leung
3193    Date: 2002/03/03 13:20:00 EST
3194    Tag: leunga-20020303-mlrisc-tools
3195    
3196      Added #[ ... ] expressions to mlrisc tools
3197    
3198    ----------------------------------------------------------------------
3199    Name: Matthias Blume
3200    Date: 2002/02/27 12:29:00 EST
3201    Tag: blume-20020227-cdebug
3202    Description:
3203    
3204    - made types in structure C and C_Debug to be equal
3205    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3206    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3207    
3208    ----------------------------------------------------------------------
3209    Name: Matthias Blume
3210    Date: 2002/02/26 12:00:00 EST
3211    Tag: blume-20020226-ffi
3212    Description:
3213    
3214    1. Fixed a minor bug in CM's "noweb" tool:
3215       If numbering is turned off, then truly don't number (i.e., do not
3216       supply the -L option to noweb).  The previous behavior was to supply
3217       -L'' -- which caused noweb to use the "default" line numbering scheme.
3218       Thanks to Chris Richards for pointing this out (and supplying the fix).
3219    
3220    2. Once again, I reworked some aspects of the FFI:
3221    
3222       A. The incomplete/complete type business:
3223    
3224       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3225         gone!
3226       - ML types representing an incomplete type are now *equal* to
3227         ML types representing their corresponding complete types (just like
3228         in C).  This is still safe because ml-nlffigen will not generate
3229         RTTI for incomplete types, nor will it generate functions that
3230         require access to such RTTI.   But when ML code generated from both
3231         incomplete and complete versions of the C type meet, the ML types
3232         are trivially interoperable.
3233    
3234         NOTE:  These changes restore the full generality of the translation
3235         (which was previously lost when I eliminated functorization)!
3236    
3237       B. Enum types:
3238    
3239       - Structure C now has a type constructor "enum" that is similar to
3240         how the "su" constructor works.  However, "enum" is not a phantom
3241         type because each "T enum" has values (and is isomorphic to
3242         MLRep.Signed.int).
3243       - There are generic access operations for enum objects (using
3244         MLRep.Signed.int).
3245       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3246         * The structure contains the definition of type "mlrep" (the ML-side
3247         representation type of the enum).  Normally, mlrep is the same
3248         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3249         then mlrep will be defined as a datatype -- thus facilitating
3250         pattern matching on mlrep values.
3251         ("-ec" will be suppressed if there are duplicate values in an
3252          enumeration.)
3253         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3254         will be generated for each C enum constant xxx.
3255         * Conversion functions m2i and i2m convert between mlrep and
3256         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3257         * Coversion functions c and ml convert between mlrep and "tag enum".
3258         * Access functions (get/set) fetch and store mlrep values.
3259       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3260         enumerations are merged into one single enumeration represented by
3261         structure E_'.
3262    
3263    ----------------------------------------------------------------------
3264    Name: Allen Leung
3265    Date: 2002/02/25 04:45:00 EST
3266    Tag: leunga-20020225-cps-spill
3267    
3268    This is a new implementation of the CPS spill phase.
3269    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3270    In case of problems, replace it with the old file spill.sml
3271    
3272    The current compiler runs into some serious performance problems when
3273    constructing a large record.  This can happen when we try to compile a
3274    structure with many items.  Even a very simple structure like the following
3275    makes the compiler slow down.
3276    
3277        structure Foo = struct
3278           val x_1 = 0w1 : Word32.int
3279           val x_2 = 0w2 : Word32.int
3280           val x_3 = 0w3 : Word32.int
3281           ...
3282           val x_N = 0wN : Word32.int
3283        end
3284    
3285    The following table shows the compile time, from N=1000 to N=4000,
3286    with the old compiler:
3287    
3288    N
3289    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3290           MLRISC ra                               0.06u  0.00s  0.05g
3291              (spills = 0 reloads = 0)
3292           TOTAL                                   0.63u  0.07s  0.21g
3293    
3294    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3295           MLRISC ra                               5.68u  0.59s  3.93g
3296              (spills = 0 reloads = 0)
3297           TOTAL                                   14.71u  0.99s  4.81g
3298    
3299    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3300           MLRISC ra                               5.54u  0.65s  3.91g
3301              (spills = 543 reloads = 1082)
3302           TOTAL                                   65.40u  3.13s  6.00g
3303    
3304    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3305           MLRISC ra                               0.80u  0.10s  0.55g
3306              (spills = 42 reloads = 84)
3307           TOTAL                                   129.42u  5.10s  4.13g
3308    
3309    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3310           MLRISC ra                               2.69u  0.27s  1.38g
3311              (spills = 62 reloads = 124)
3312           TOTAL                                   682.48u  19.61s  13.99g
3313    
3314    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3315           MLRISC ra                               4.96u  0.27s  2.72g
3316              (spills = 85 reloads = 170)
3317           TOTAL                                   2375.26u  57.21s  48.00g
3318    
3319    As you can see the old cps spill module suffers from some serious
3320    performance problem.  But since I cannot decipher the old code fully,
3321    instead of patching the problems up, I'm reimplementing it
3322    with a different algorithm.  The new code is more modular,
3323    smaller when compiled, and substantially faster
3324    (O(n log n) time and O(n) space).  Timing of the new spill module:
3325    
3326    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3327          MLRISC ra                               0.25u  0.02s  0.15g
3328             (spills=1 reloads=3)
3329          TOTAL                                   7.74u  0.34s  1.62g
3330    
3331    Implementation details:
3332    
3333    As far as I can tell, the purpose of the CPS spill module is to make sure the
3334    number of live variables at any program point (the bandwidth)
3335    does not exceed a certain limit, which is determined by the
3336    size of the spill area.
3337    
3338    When the bandwidth is too large, we decrease the register pressure by
3339    packing live variables into spill records.  How we achieve this is
3340    completely different than what we did in the old code.
3341    
3342    First, there is something about the MLRiscGen code generator
3343    that we should be aware of:
3344    
3345    o MLRiscGen performs code motion!
3346    
3347       In particular, it will move floating point computations and
3348       address computations involving only the heap pointer to
3349       their use sites (if there is only a single use).
3350       What this means is that if we have a CPS record construction
3351       statement
3352    
3353           RECORD(k,vl,w,e)
3354    
3355       we should never count the new record address w as live if w
3356       has only one use (which is often the case).
3357    
3358       We should do something similar to floating point, but the transformation
3359       there is much more complex, so I won't deal with that.
3360    
3361    Secondly, there are now two new cps primops at our disposal:
3362    
3363     1. rawrecord of record_kind option
3364        This pure operator allocates some uninitialized storage from the heap.
3365        There are two forms:
3366    
3367         rawrecord NONE [INT n]  allocates a tagless record of length n
3368         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3369                                     and initializes the tag.
3370    
3371     2. rawupdate of cty
3372          rawupdate cty (v,i,x)
3373          Assigns to x to the ith component of record v.
3374          The storelist is not updated.
3375    
3376    We use these new primops for both spilling and increment record construction.
3377    
3378     1. Spilling.
3379    
3380        This is implemented with a linear scan algorithm (but generalized
3381        to trees).  The algorithm will create a single spill record at the
3382        beginning of the cps function and use rawupdate to spill to it,
3383        and SELECT or SELp to reload from it.  So both spills and reloads
3384        are fine-grain operations.  In contrast, in the old algorithm
3385        "spills" have to be bundled together in records.
3386    
3387        Ideally, we should sink the spill record construction to where
3388        it is needed.  We can even split the spill record into multiple ones
3389        at the places where they are needed.  But CPS is not a good
3390        representation for global code motion, so I'll keep it simple and
3391        am not attempting this.
3392    
3393     2. Incremental record construction (aka record splitting).
3394    
3395        Long records with many component values which are simulatenously live
3396        (recall that single use record addresses are not considered to
3397         be live) are constructed with rawrecord and rawupdate.
3398        We allocate space on the heap with rawrecord first, then gradually
3399        fill it in with rawupdate.  This is the technique suggested to me
3400        by Matthias.
3401    
3402        Some restrictions on when this is applicable:
3403        1. It is not a VECTOR record.  The code generator currently does not handle
3404           this case. VECTOR record uses double indirection like arrays.
3405        2. All the record component values are defined in the same "basic block"
3406           as the record constructor.  This is to prevent speculative
3407           record construction.
3408    
3409    ----------------------------------------------------------------------
3410    Name: Allen Leung
3411    Date: 2002/02/22 01:02:00 EST
3412    Tag: leunga-20020222-mlrisc-tools
3413    
3414    Minor bug fixes in the parser and rewriter
3415    
3416    ----------------------------------------------------------------------
3417    Name: Allen Leung
3418    Date: 2002/02/21 20:20:00 EST
3419    Tag: leunga-20020221-peephole
3420    
3421    Regenerated the peephole files.  Some contained typos in the specification
3422    and some didn't compile because of pretty printing bugs in the old version
3423    of 'nowhere'.
3424    
3425    ----------------------------------------------------------------------
3426    Name: Allen Leung
3427    Date: 2002/02/19 20:20:00 EST
3428    Tag: leunga-20020219-mlrisc-tools
3429    Description:
3430    
3431       Minor bug fixes to the mlrisc-tools library:
3432    
3433       1.  Fixed up parsing colon suffixed keywords
3434       2.  Added the ability to shut the error messages up
3435       3.  Reimplemented the pretty printer and fixed up/improved
3436           the pretty printing of handle and -> types.
3437       4.  Fixed up generation of literal symbols in the nowhere tool.
3438       5.  Added some SML keywords to to sml.sty
3439    
3440    ----------------------------------------------------------------------
3441    Name: Matthias Blume
3442    Date: 2002/02/19 16:20:00 EST
3443    Tag: blume-20020219-cmffi
3444    Description:
3445    
3446    A wild mix of changes, some minor, some major:
3447    
3448    * All C FFI-related libraries are now anchored under $c:
3449        $/c.cm      --> $c/c.cm
3450        $/c-int.cm  --> $c/internals/c-int.cm
3451        $/memory.cm --> $c/memory/memory.cm
3452    
3453    * "make" tool (in CM) now treats its argument pathname slightly
3454      differently:
3455        1. If the native expansion is an absolute name, then before invoking
3456           the "make" command on it, CM will apply OS.Path.mkRelative
3457           (with relativeTo = OS.FileSys.getDir()) to it.
3458        2. The argument will be passed through to subsequent phases of CM
3459           processing without "going native".  In particular, if the argument
3460           was an anchored path, then "make" will not lose track of that anchor.
3461    
3462    * Compiler backends now "know" their respective C calling conventions
3463      instead of having to be told about it by ml-nlffigen.  This relieves
3464      ml-nlffigen from one of its burdens.
3465    
3466    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3467    
3468    * Export C_DEBUG and C_Debug from $c/c.cm.
3469    
3470    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3471      subtyping relationship between incomplete pointers and their complete
3472      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3473      with the convention of instantiating 'o with ('t, 'c) obj whenever
3474      the pointer target type is complete.  In the incomplete case, 'o
3475      will be instantiated with some "'c iobj" -- a type obtained by
3476      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3477    
3478      Operations that work on both incomplete and complete pointer types are
3479      typed as taking an 'o ptr while operations that require the target to
3480      be known are typed as taking some ('t, 'c) obj ptr.
3481    
3482      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3483      where void is an eqtype without any values.  This makes it possible
3484      to work on voidptr values using functions meant to operate on light
3485      incomplete pointers.
3486    
3487    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3488      been vastly simplified.
3489    
3490    ----------------------------------------------------------------------
3491    Name: Matthias Blume
3492    Date: 2002/02/19 10:48:00 EST
3493    Tag: blume-20020219-pqfix
3494    Description:
3495    
3496    Applied Chris Okasaki's bug fix for priority queues.
3497    
3498    ----------------------------------------------------------------------
3499    Name: Matthias Blume
3500    Date: 2002/02/15 17:05:00
3501    Tag: Release_110_39
3502    Description:
3503    
3504    Last-minute retagging is becoming a tradition... :-(
3505    
3506    This is the working release 110.39.
3507    
3508    ----------------------------------------------------------------------
3509    Name: Matthias Blume
3510    Date: 2002/02/15 16:00:00 EST
3511    Tag: Release_110_39-orig
3512    Description:
3513    
3514    Working release 110.39.  New bootfiles.
3515    
3516    (Update: There was a small bug in the installer so it wouldn't work
3517    with all shells.  So I retagged. -Matthias)
3518    
3519    ----------------------------------------------------------------------
3520    Name: Matthias Blume
3521    Date: 2002/02/15 14:17:00 EST
3522    Tag: blume-20020215-showbindings
3523    Description:
3524    
3525    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3526    the latter can be useful for exploring what bindings are available at
3527    the interactive prompt.  (The first function returns only the list
3528    of symbols that are really bound, the second prints those but also the
3529    ones that CM's autoloading mechanism knows about.)
3530    
3531    ----------------------------------------------------------------------
3532    Name: Matthias Blume
3533    Date: 2002/02/15 12:08:00 EST
3534    Tag: blume-20020215-iptrs
3535    Description:
3536    
3537    Two improvements to ml-nlffigen:
3538    
3539      1. Write files only if they do not exist or if their current contents
3540         do not coincide with what's being written.  (That is, avoid messing
3541         with the time stamps unless absolutely necessary.)
3542    
3543      2. Implement a "repository" mechanism for generated files related
3544         to "incomplete pointer types".   See the README file for details.
3545    
3546    ----------------------------------------------------------------------
3547    Name: Matthias Blume
3548    Date: 2002/02/14 11:50:00 EST
3549    Tag: blume-20020214-quote
3550    Description:
3551    
3552    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3553    because of the new and improved tag generation scheme.  (Thanks to Allen
3554    Leung for pointing it out.)
3555    
3556    ----------------------------------------------------------------------
3557    Name: Lal George
3558    Date: 2002/02/14 09:55:27 EST 2002
3559    Tag: george-20020214-isabelle-bug
3560    Description:
3561    
3562    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3563    of Isabelle on the x86.
3564    
3565    From Allen:
3566    -----------
3567     I've found the problem:
3568    
3569         in ra-core.sml, I use the counter "blocked" to keep track of the
3570         true number of elements in the freeze queue.  When the counter goes
3571         to zero, I skip examining the queue.  But I've messed up the
3572         bookkeeping in combine():
3573    
3574             else ();
3575             case !ucol of
3576               PSEUDO => (if !cntv > 0 then
3577                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3578                                        ^^^^^^^^^^^^^^^^^^^^^^^
3579                      moveu := mergeMoveList(!movev, !moveu)
3580                     )
3581                  else ();
3582    
3583         combine() is called to coalesce two nodes u and v.
3584         I think I was thinking that if the move counts of u and v are both
3585         greater than zero then after they are coalesced then one node is
3586         removed from the freeze queue.  Apparently I was thinking that
3587         both u and v are of low degree, but that's clearly not necessarily true.
3588    
3589    
3590    02/12/2002:
3591        Here's the patch.  HOL now compiles.
3592    
3593        I don't know how this impact on performance (compile
3594        time or runtime).  This bug caused the RA (especially on the x86)
3595        to go thru the potential spill phase when there are still nodes on the
3596        freeze queue.
3597    
3598    
3599    
3600    
3601    ----------------------------------------------------------------------
3602    Name: Matthias Blume
3603    Date: 2002/02/13 22:40:00 EST
3604    Tag: blume-20020213-fptr-rtti
3605    Description:
3606    
3607    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3608    updates.
3609    
3610    ----------------------------------------------------------------------
3611    Name: Matthias Blume
3612    Date: 2002/02/13 16:41:00 EST
3613    Tag: blume-20020213-cmlpq
3614    Description:
3615    
3616    Added new priority queue export symbols (which have just been added to
3617    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3618    not compile and the installer would choke.)
3619    
3620    ----------------------------------------------------------------------
3621    Name: Matthias Blume
3622    Date: 2002/02/13 16:15:00 EST
3623    Tag: blume-20020213-various
3624    Description:
3625    
3626    1. More tweaks to ml-nlffigen:
3627    
3628       - better internal datastructures (resulting in slight speedup)
3629       - "-match" option requires exact match
3630       - "localized" gensym counters (untagged structs/unions nested within
3631         other structs/unions or within typedefs get a fresh counter; their
3632         tag will be prefixed by a concatenation of their parents' tags)
3633       - bug fixes (related to calculation of transitive closure of types
3634         to be included in the output)
3635    
3636    2. Minor Basis updates:
3637    
3638       - added implementations for List.collate and Option.app
3639    
3640    ----------------------------------------------------------------------
3641    Name: Matthias Blume
3642    Date: 2002/02/11 15:55:00 EST
3643    Tag: blume-20020211-gensym
3644    Description:
3645    
3646    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3647    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3648    names (ML structure names that correspond to unnamed C structs, unions,
3649    and enums), so that separate runs of ml-nlffigen do not clash.
3650    
3651    ----------------------------------------------------------------------
3652    Name: Matthias Blume
3653    Date: 2002/02/11 12:05:00 EST
3654    Tag: blume-20020211-gensml
3655    Description:
3656    
3657    A quick fix for a problem with GenSML (in the pgraph-util library):
3658    Make generation of toplevel "local" optional.  (Strictly speaking,
3659    signature definitions within "local" are not legal SML.)
3660    
3661    Other than that: updates to INSTALL and cm/TODO.
3662    
3663    ----------------------------------------------------------------------
3664    Name: Matthias Blume
3665    Date: 2002/02/08 15:00:00 EST
3666    Tag: blume-20020208-uniquepid
3667    Description:
3668    
3669    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3670    
3671    1. The installer (config/install.sh) has gotten smarter:
3672    
3673         - Configuration options are a bit easier to specify now
3674           (in config/targets).
3675         - Bug in recognizing .tar.bz2 files fixed.
3676         - Installer automatically resolves dependencies between
3677           configuration options (e.g., if you ask for eXene, you will
3678           also get cml -- regardless whether you asked for it or not).
3679         - Installer can run in "quieter mode" by setting the environment
3680           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3681           "completely silent", though.
3682         - Build HashCons library as part of smlnj-lib.
3683    
3684    2. A new scheme for assigning persistent identifiers to compilation
3685       units (and, by extension, to types etc.) has been put into place.
3686       This fixes a long-standing bug where types and even dynamic values
3687       can get internally confused, thereby compromising type safety
3688       (abstraction) and dynamic correctness.  See
3689    
3690         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3691    
3692       for an example of how things could go wrong until now.
3693    
3694       The downside of the new scheme is that pids are not quite as
3695       persistent as they used to be: CM will generate a fresh pid
3696       for every compilation unit that it thinks it sees for the first
3697       time.  That means that if you compile starting from a clean, fresh
3698       source tree at two different times, you end up with different
3699       binaries.
3700    
3701       Cutoff recompilation, however, has not been compromised because
3702       CM keeps pid information in special caches between runs.
3703    
3704    ----------------------------------------------------------------------
3705    Name: Lal George
3706    Date: 2002/02/07 15:34:13 EST 2002
3707    Tag: <none>
3708    Description:
3709    
3710    Compilers that generate assembly code may produce  global labels
3711    whose value is resolved at link time. The various peephole optimization
3712    modules did not take this in account.
3713    
3714    TODO. The Labels.addrOf function should really return an option
3715    type so that clients are forced to deal with this issue, rather
3716    than an exception being raised.
3717    
3718    ----------------------------------------------------------------------
3719    Name: Lal George
3720    Date: 2002/02/06 13:55:02 EST
3721    Tag: george-20020206-ra-breakup
3722    Description:
3723    
3724    1. A bug fix from Allen.
3725    
3726        A typo causes extra fstp %st(0)'s to be generated at compensation
3727        edges, which might cause stack underflow traps at runtime.  This
3728        occurs in fft where there are extraneous fstps right before the 'into'
3729        trap instruction (in this case they are harmless since none of the
3730        integers overflow.)
3731    
3732    2. Pulled out various utility modules that were embedded in the modules
3733       of the register allocator. I need these modules for other purposes, but
3734       they are not complete enough to put into a library (just yet).
3735    ----------------------------------------------------------------------
3736    Name: Matthias Blume
3737    Date: 2002/01/31 16:05:00 EST
3738    Tag: blume-20020131-sparc-ccalls
3739    Description:
3740    
3741    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3742       of extra stack space by mistake.  Fixed.
3743    
3744    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3745    
3746    ----------------------------------------------------------------------
3747    Name: Allen Leung
3748    Date: 2002/01/30
3749    Tag: leunga-20020130-nowhere-bug-fix
3750    Description:
3751    
3752       MLRISC bug fixes:
3753       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3754       2. MachineInt.fromString was negating its value.
3755    
3756    ----------------------------------------------------------------------
3757    Name: Matthias Blume
3758    Date: 2002/01/29
3759    Tag: blume-20020129-INSTALL
3760    Description:
3761    
3762    - Added somewhat detailed installation instructions (file INSTALL).
3763    - Fixed curl-detection bug in config/install.sh.
3764    - It is now possible to select the URL getter using the URLGETTER
3765      environment variable:
3766    
3767          not set / "unknown"      --> automatic detection (script tries wget,
3768                                       curl, and lynx)
3769          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3770                                       how to properly invoke them)
3771          other                    --> use $URLGETTER directly, it must take
3772                                       precisely two command-line arguments
3773                                       (source URL and destination file name)
3774    
3775    ----------------------------------------------------------------------
3776    Name: Matthias Blume
3777    Date: 2002/01/28
3778    Tag: blume-20020128-sparc-ccalls
3779    Description:
3780    
3781    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3782    - Make use of the allocParam argument in sparc-c-calls.
3783    
3784    ----------------------------------------------------------------------
3785    Name: Matthias Blume
3786    Date: 2002/01/28
3787    Tag: blume-20020128-allocParam
3788    Description:
3789    
3790    John Reppy:  Changes c-calls API to accept client-callback for
3791    allocating extra stack space.
3792    me: Corresponding changes to mlriscGen (using a dummy argument that
3793        does not change the current behavior).
3794    
3795    ----------------------------------------------------------------------
3796    Name: Matthias Blume
3797    Date: 2002/01/28 12:00:00
3798    Tag: Release_110_38
3799    Description:
3800    
3801    This time for real!!!
3802    
3803    ----------------------------------------------------------------------
3804    Name: Matthias Blume
3805    Date: 2002/01/28 10:56:00 EST
3806    Tag: blume-20020128-retraction
3807    Description:
3808    
3809    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
3810       with blume-Release_110_38-retracted.)
3811    
3812    1. Fixed a problem with incorrect rounding modes in real64.sml.
3813       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
3814    
3815    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
3816       and enums fixed.  The naming of corresponding ML identifiers should
3817       now be consistent again.
3818    
3819    ----------------------------------------------------------------------
3820    Name: Allen Leung
3821    Date: 2002/01/27
3822    Tag: leunga-20020127-nowhere
3823    Description:
3824    
3825       Added a target called nowhere in the configuration scripts.
3826       Enabling this will build the MLRISC 'nowhere' tool (for translating
3827       programs with where-clauses into legal SML code) during installation.
3828    
3829    ----------------------------------------------------------------------
3830    Name: Matthias Blume
3831    Date: 2002/01/25 21:27:00 EST
3832    Tag: blume-Release_110_38-retracted
3833    Description:
3834    
3835    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
3836    
3837    README will be added later.
3838    
3839    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
3840    (Release_110_38) removed.  Reason: Last-minute bug fixes.
3841    
3842    ----------------------------------------------------------------------
3843    Name: Matthias Blume
3844    Date: 2002/01/25
3845    Tag: blume-20020125-ffi
3846    Description:
3847    
3848    A large number of tweaks and improvements to ml-nlffi-lib and
3849    ml-nlffigen:
3850    
3851       - ML represenation types have been streamlined
3852       - getter and setter functions work with concrete values, not abstract
3853         ones where possible
3854       - ml-nlffigen command line