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 972, Sun Oct 28 00:37:02 2001 UTC revision 1795, Wed May 18 18:42:26 2005 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Matthias Blume (blume (at) tti - c (dot) org)
17    Date: 2005/05/18 13:35:00 CDT
18    Tag: Release_110_54
19    Description:
20    
21    New working version (110.54).  NEW BOOTFILES!
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume (blume (at) tti - c (dot) org)
25    Date: 2005/05/18 11:58:00 CDT
26    Tag: blume-20050518-installer
27    Description:
28    
29    Added support scripts for Mac OS X PackageMaker and modified
30    config/install.sh so that it supports re-dumping a heap image
31    after customization.
32    
33    ----------------------------------------------------------------------
34    Name: Matthias Blume (blume (at) tti - c (dot) org)
35    Date: 2005/05/18 10:55:00 CDT
36    Tag: blume-20050518-realdiv-noovld
37    Description:
38    
39    Un-overloaded / to work around bug in overloading resolution code.
40    
41    ----------------------------------------------------------------------
42    Name: Matthias Blume (blume (at) tti - c (dot) org)
43    Date: 2005/05/16 23:50:00 CDT
44    Tag: blume-20050516-redump-heap
45    Description:
46    
47    Added mechanism for re-creating a heap file for the interactive system
48    after configuration variables have been changed.
49    
50       CM.redump_heap : string -> unit
51    
52    This is much like SMLofNJ.exportML, but starting from the resulting
53    heap does not return to the caller of CM.redump_heap but
54    restarts the interactive system from scratch.  The original call of
55    CM.redump_heap does not return but ends the interactive session.
56    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
57    
58    Internally, redump_heap winds the dynamic execution context back to
59    the point where the original heap image was created and re-executes
60    the heap image generation code in the boot code.
61    
62    ----------------------------------------------------------------------
63    Name: Matthias Blume (blume (at) tti - c (dot) org)
64    Date: 2005/05/09 21:55:00 CDT
65    Tag: blume-20050509-word64
66    Description:
67    
68    Added a hack to the existing hack known as Word64 to make fromString
69    behave correctly.  I am still not sure whether Word64.scan will work
70    as specified with respect to the interaction of radix and prefix.
71    
72    ----------------------------------------------------------------------
73    Name: Allen Leung (leunga (at) reservoir (dot) com)
74    Date: 2005/05/04 11:50:00 EST
75    Tag: leunga-20050504-checkgc
76    Description:
77    
78    Added a gc protocol checking phase.  This phase is enabled with
79    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
80    
81    ----------------------------------------------------------------------
82    Name: Matthias Blume (blume (at) tti - c (dot) org)
83    Date: 2005/05/04 10:45:00 CDT
84    Tag: blume-20050504-intinf
85    Description:
86    
87    Fixed a bug in the implementation of div and mod for IntInf.
88    Thanks to Neophytos Michael for reporting the problem.
89    
90    ----------------------------------------------------------------------
91    Name: Matthias Blume (blume (at) tti - c (dot) org)
92    Date: 2005/05/04 10:35:00 CDT
93    Tag: blume-20050504-join
94    Description:
95    
96    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
97    
98    ----------------------------------------------------------------------
99    Name: Matthias Blume (blume (at) tti - c (dot) org)
100    Date: 2005/02/28 23:40:00 CST
101    Tag: blume-20050228-mVar
102    Description:
103    
104    Fixed serious bug (brown paper bag variety) in new implementation of
105    structure Atom in CML.  (I had accidentally used a mailbox instead of
106    an mvar, leaving the door open for races.)
107    
108    ----------------------------------------------------------------------
109    Name: Matthias Blume (blume (at) tti - c (dot) org)
110    Date: 2005/02/25 15:00:00 CST
111    Tag: Release_110_53
112    Description:
113    
114    New working version (110.53).  NEW BOOTFILES!
115    
116    ----------------------------------------------------------------------
117    Name: Matthias Blume (blume (at) tti - c (dot) org)
118    Date: 2005/02/25 14:50:00 CST
119    Tag: blume-20050225-susp
120    Description:
121    
122    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
123    implemented in Core -- which means that it is the same as the one used
124    by the lazy extension.
125    
126    ----------------------------------------------------------------------
127    Name: Matthias Blume (blume (at) tti - c (dot) org)
128    Date: 2005/02/24 16:50:00 CST
129    Tag: blume-20050224-cml-atom
130    Description:
131    
132    Simpler and at the same time more general implementation of structure
133    Atom in CML.
134    
135    ----------------------------------------------------------------------
136    Name: Matthias Blume (blume (at) tti - c (dot) org)
137    Date: 2005/02/15 17:35:00 CST
138    Tag: blume-20050215-tools
139    Description:
140    
141    Created new "tools" directory under "src" and moved "TraceDebugProf"
142    there.
143    
144    ----------------------------------------------------------------------
145    Name: Matthias Blume (blume (at) tti - c (dot) org)
146    Date: 2005/02/10 17:55:00 CST
147    Tag: blume-20050210-longlong
148    Description:
149    
150    Implemented "long long" arguments and results for NLFFI.  (Only the
151    PPC/MacOS implementation is complete, the other backends still need to
152    be updated.)
153    
154    ----------------------------------------------------------------------
155    Name: Matthias Blume (blume (at) tti - c (dot) org)
156    Date: 2005/01/24 17:40:00 CST
157    Tag: blume-20050124-mlyacc
158    Description:
159    
160    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
161    problem with certain "as" patterns which previously got rendered
162    using incorrect syntax.
163    
164    ----------------------------------------------------------------------
165    Name: Matthias Blume (blume (at) tti - c (dot) org)
166    Date: 2005/01/18 12:00:00 CST
167    Tag: blume-20050118-profile
168    Description:
169    
170    Made time profiling code (interrupt handler) in runtime system aware
171    of new array representation.
172    
173    ----------------------------------------------------------------------
174    Name: Matthias Blume (blume (at) tti - c (dot) org)
175    Date: 2005/01/14 18:00:00 CST
176    Tag: blume-20050114-heap2exec
177    Description:
178    
179    Implemented new (but still experimental) heap2exec facility.  This is
180    tested under Mac OS X and should work under Linux (will test shortly).
181    It will probably also work on the Sparc (will test some time later).
182    
183      - removed old "HACKED_STANDALONE" hack from runtime
184    
185    To be able to test this, uncomment the request for "heap2asm" in
186    config/targets prior to installation.  (Notice that this is different
187    from "heap2exec" mentioned below.  Not a typo.)
188    
189    To perform an actual test, run the command
190    
191       $ bin/heap2exec heapfile execfile
192    
193    (You can put heap2exec on your shell's path.)
194    
195    For example, run
196    
197      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
198    
199    This will create a standalone executable called "mly" which you can
200    then invoke directly as a command.
201    
202    ----------------------------------------------------------------------
203    Name: Matthias Blume (blume (at) tti - c (dot) org)
204    Date: 2005/01/07 11:44:00 CST
205    Tag: blume-20050107-mlstring
206    Description:
207    
208    fixed off-by-one error in ML_STRING macro (globals.c)
209    
210    ----------------------------------------------------------------------
211    Name: Matthias Blume (blume (at) tti - c (dot) org)
212    Date: 2004/12/23 18:00:00 CST
213    Tag: blume-20041223-santa
214    Description:
215    
216    Made ml-build script "smarter" (but only very little).
217    
218    ----------------------------------------------------------------------
219    Name: Matthias Blume (blume (at) tti - c (dot) org)
220    Date: 2004/12/21 15:05:00 CST
221    Tag: blume-20041221-longlong
222    Description:
223    
224    * Implemented access to signed and unsigned long long data in NLFFI.
225    (The parameter-passing part of the picture has not complete. But
226    data structure access seems to work.)
227    
228    * Fixed CM's incorrect assumption that the PPC is little-endian.
229      (On the Mac, it is big-endian.  And that's currently our only
230      PPC platform.)
231    
232    ----------------------------------------------------------------------
233    Name: Matthias Blume (blume (at) tti - c (dot) org)
234    Date: 2004/12/21 12:50:00 CST
235    Tag: blume-20041221-memory
236    Description:
237    
238    Some cleanup in the $c/memory.cm library: separated some concerns by
239    moving allocation code and memory access code each into their own
240    files.
241    
242    ----------------------------------------------------------------------
243    Name: Allen Leung (leunga (at) reservoir (dot) com)
244    Date: 2004/12/17 16:12:00 EST
245    Tag: leunga-20041217-cygwin-smlnj-home
246    Description:
247    
248        The Unix I/O library of SML/NJ on cygwin does not understand
249    Windows style pathname, so problems arise when SMLNJ_HOME is set
250    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
251    to a POSIX pathname on cygwin.
252    
253    ----------------------------------------------------------------------
254    Name: Matthias Blume (blume (at) tti - c (dot) org)
255    Date: 2004/12/16 13:35:00 CST
256    Tag: Release_110_52
257    Description:
258    
259    Last-minute changes incorporated into 110.52.  Release tag moved.
260    
261    The changes:
262       - HashString.hashString' -> HashString.hashSubstring
263       - bug fix in UnivariateStats
264    
265    ----------------------------------------------------------------------
266    Name: Matthias Blume (blume (at) tti - c (dot) org)
267    Date: 2004/12/15 23:40:00 CST
268    Tag: blume-20041215-hashSubstring
269    Description:
270    
271    - HashString.hashString' -> HashString.hashSubstring
272    - corresponding changes in atom.sml
273    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
274      added some comments
275    
276    ----------------------------------------------------------------------
277    Name: Matthias Blume (blume (at) tti - c (dot) org)
278    Date: 2004/12/15 15:30:00 CST
279    Tag: (Release_110_52)
280    Description:
281    
282    New working version (110.52).  NEW BOOTFILES!
283    
284    ----------------------------------------------------------------------
285    Name: Matthias Blume (blume (at) tti - c (dot) org)
286    Date: 2004/12/15 12:45:00 CST
287    Tag: blume-20041215-spaces
288    Description:
289    
290    More on the space problem (this time for Win32).
291    
292    ----------------------------------------------------------------------
293    Name: Matthias Blume (blume (at) tti - c (dot) org)
294    Date: 2004/12/14 17:30:00 CST
295    Tag: blume-20041214-spaces
296    Description:
297    
298    Hacked some of the scripts (in particular: the installer) to cope with
299    spaces in filenames a bit better.  But beware: the current "solution"
300    is likely still full of bugs and inherently incomplete.  (We need to
301    do away with those shell scripts for a comprehensive solution.)
302    
303    ----------------------------------------------------------------------
304    Name: Matthias Blume (blume (at) tti - c (dot) org)
305    Date: 2004/12/13 14:45:00 CST
306    Tag: blume-20041213-ml-makedepend
307    Description:
308    
309    Fixed bug in code for ml-makedepend.
310    
311    ----------------------------------------------------------------------
312    Name: Matthias Blume (blume (at) tti - c (dot) org)
313    Date: 2004/12/09 16:30:00 CST
314    Tag: blume-20041209-statistics
315    Description:
316    
317    Added two simple but potentially useful statistics modules to SML/NJ Library.
318    (See CHANGES file there.)
319    
320    ----------------------------------------------------------------------
321    Name: Matthias Blume (blume (at) tti - c (dot) org)
322    Date: 2004/12/01 16:50:00 CST
323    Tag: blume-20041201-atom
324    Description:
325    
326    smlnj-lib:
327    
328      Added function HashString.hashString' for substrings.
329      Hand-inlined CharVector.foldl into HashString (for speed).
330      Modified implementation of structure Atom to avoid extracting
331      strings from substrings unless necessary.
332      (Also see CHANGES file for smlnj-lib.)
333    
334    ----------------------------------------------------------------------
335    Name: Matthias Blume (blume (at) tti - c (dot) org)
336    Date: 2004/11/24 22:15:00 CST
337    Tag: blume-20041124-cml
338    Description:
339    
340    Made sure CML compiles when Position = Int64.
341    
342    ----------------------------------------------------------------------
343    Name: Matthias Blume (blume (at) tti - c (dot) org)
344    Date: 2004/11/24 14:30:00 CST
345    Tag: blume-20041124-position
346    Description:
347    
348    The compiler can now be compiled in a mode that makes structure Position
349    equal to Int64.  The default, however, is unchanged (Position = Int31)
350    for the time being.
351    
352    To enable 64-bit positions, use the following procedure:
353    
354    1. Start sml
355    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
356    3. Type
357          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
358    4. Run CMB.make() as usual.
359    
360    This is barely tested.  The only test so far was a little SML program
361    counting the number of characters in an 8-gigabyte file by
362    reading it character-by-character.  That test was successful.
363    
364    In support of 64-bit positions, a number of new functions have been
365    added to the runtime system.
366    
367    ----------------------------------------------------------------------
368    Name: Matthias Blume (blume (at) tti - c (dot) org)
369    Date: 2004/11/23 14:45:00 CST
370    Tag: blume-20041123-useFile
371    Description:
372    
373    Fixed a problem with unhelpful error messages related to problems with
374    .cm- or .sml files that appear as part of the sml command line.
375    
376    ----------------------------------------------------------------------
377    Name: Matthias Blume (blume (at) tti - c (dot) org)
378    Date: 2004/11/18 15:40:00 CST
379    Tag: Release_110_51
380    Description:
381    
382    New working version (110.51).  NEW BOOTFILES!
383    
384    ----------------------------------------------------------------------
385    Name: Matthias Blume (blume (at) tti - c (dot) org)
386    Date: 2004/11/18 15:35:00 CST
387    Tag: <none>
388    Description:
389    
390    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
391    Joemann).
392    
393    ----------------------------------------------------------------------
394    Name: Allen Leung (leunga (at) reservoir (dot) com)
395    Date: 2004/11/17 16:05:21 EST 2004
396    Tag: leunga-20041117-mlrisc-live-kill
397    Description:
398    
399       Added support for MLTree constructs LIVE and KILL
400    to all the architectures.
401    
402    ----------------------------------------------------------------------
403    Name: Matthias Blume (blume (at) tti - c (dot) org)
404    Date: 2004/11/13 00:20:00 CST
405    Tag: blume-20041113-versiontool
406    Description:
407    
408    - Stripped down the versiontool: It now only handles the version number.
409      The date string is generated at bootstrap time (during makeml).
410    
411    - In a previous commit, fixed a minor issue with how polyequal is being
412      translated.  In particular, the code now "looks through" abstractions.
413      This results in slightly fewer polyEqual warnings and hopefully slightly
414      more efficient code.  Important examples for where this matters are
415      the new int64 and word64 types.
416    
417    ----------------------------------------------------------------------
418    Name: Matthias Blume (blume (at) tti - c (dot) org)
419    Date: 2004/11/12 00:30:00 CST
420    Tag: blume-20041112-int64
421    Description:
422    
423    Structure Int64 fully hooked in.  (The implementation is not very
424    efficient, though.)
425    
426    ----------------------------------------------------------------------
427    Name: Matthias Blume (blume (at) tti - c (dot) org)
428    Date: 2004/11/11 17:30:00 CST
429    Tag: blume-20041111-more64
430    Description:
431    
432    All the pieces of Word64 are now there, with the exception of the
433    conversions from and to LargeWord. (Eventually these need to be identities,
434    but for the time being they don't even make sense because LargeWord is
435    32-bit wide.)
436    
437    Also started to add similar support for Int64, but major pieces of that
438    are still missing.
439    
440    ----------------------------------------------------------------------
441    Name: Matthias Blume (blume (at) tti - c (dot) org)
442    Date: 2004/11/11 00:15:00 CST
443    Tag: blume-20041111-word64
444    Description:
445    
446    Structure Word64 is now (almost) complete, word literals and patterns
447    seem to work.  There are a few odd pieces missing.  In particular,
448    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
449    at the moment.
450    
451    Making Word64 official would mean that LargeWord becomes Word64.  But
452    this requires extreme care because most word-word conversions have to
453    go through LargeWord, so making a mistake means loss of efficiency or
454    worse.  Eventually there will be a solution similar to (but actually
455    simpler than) what I did with IntInf.
456    
457    ----------------------------------------------------------------------
458    Name: Matthias Blume (blume (at) tti - c (dot) org)
459    Date: 2004/11/10 18:12:00 CST
460    Tag: blume-20041110-64bit
461    Description:
462    
463    More 64-bit hacking (but still not even half-way there yet).
464    Also, some assorted improvements to the handling of 8-bit words.
465    
466    ----------------------------------------------------------------------
467    Name: Matthias Blume (blume (at) tti - c (dot) org)
468    Date: 2004/11/09 17:50:00 CST
469    Tag: <none>
470    Description:
471    
472    Started adding some infrastructure for supporting 64-bit int- and
473    word-types.  (Still in its very early stages.)
474    
475    ----------------------------------------------------------------------
476    Name: Matthias Blume (blume (at) tti - c (dot) org)
477    Date: 2004/10/28 10:45:00 CDT
478    Tag: Release_110_50
479    Description:
480    
481    New working version (110.50).  NEW BOOTFILES!
482    
483                 =====================
484    
485    Also:
486    
487      - Changend config/srcarchiveurl from a file just
488        containing the URL string into a file containing
489        shell script code.  The code has access to the $VERSION variable.
490      - Made corresponding changes to config/install.sh and config/unpack.
491      - Default contents of config/srcarchiveurl uses $VERSION and
492        normally does not have to be edited to reflect a version change.
493    
494      (As a result, a version change can be done by just editing
495       config/version, the rest is now automatic.)
496    
497    ----------------------------------------------------------------------
498    Name: Matthias Blume (blume (at) tti - c (dot) org)
499    Date: 2004/10/27 17:50:00 CDT
500    Tag: blume-20041027-btrace-msg
501    Description:
502    
503    BackTrace.monitor now also reports the source of the exception that
504    triggered the trace.
505    
506    ----------------------------------------------------------------------
507    Name: Matthias Blume (blume (at) tti - c (dot) org)
508    Date: 2004/10/27 17:20:00 CDT
509    Tag: blume-20041027-x86-c-calls
510    Description:
511    
512    This is the HISTORY entry for two earlier commits, both concerning
513    the x86 c-calls code in MLRISC:
514    
515      - added a missing LOAD in the code that deals with struct arguments
516      - made sure the caller does not add the wrong number of bytes to the
517        stack pointer after a call of a function returning a struct
518        (the callee already pops the implicit argument which points to
519         the space reserved for the result)
520    
521    ----------------------------------------------------------------------
522    Name: Allen Leung (leunga (at) reservoir (dot) com)
523    Date: 2004/10/24 14:00:00 EST
524    Tag: leunga-20041024-x86-gas-fucomip
525    Description:
526    
527       John discovered a bug in the syntax of fucomip.
528       The opcodes FU?COMIP? have been changed to
529    
530           fu?comip? %st(i), %st
531    
532    ----------------------------------------------------------------------
533    Name: Matthias Blume (blume (at) tti - c (dot) org)
534    Date: 2004/10/20 15:06:00 CDT
535    Tag: blume-20041020-standalone-backtrace
536    Description:
537    
538    Added a mechanism for getting back-trace information from standalone
539    programs.  Here is how it works:
540    
541       1. The part of the program from which you want to get backtrace
542          information (usually the whole program) should be wrapped with
543          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
544          main program could be modified from something like
545    
546             fun main (pgm, args) = ...
547    
548         to
549    
550            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
551    
552       2. To be able to access BackTrace.monitor, you have to add
553          library $smlnj-tdp/plugins.cm to the .cm file that contains your
554          main function.
555    
556       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
557          CM might have created in the past for your project).
558    
559       4. Build the system using this command line:
560    
561             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
562                      myprog.cm MyProg.main myprog
563    
564          instead of the usual
565    
566             ml-build myprog.cm MyProg.main myprog
567    
568    I changed a library name:
569    
570       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
571    
572    New libraries:
573    
574       $smlnj-tdp/back-trace.cm
575          -- when loaded causes the back-trace plugin to be installed
576       $smlnj-tdp/coverage.cm
577          -- when loaded causes the coverage plugin to be installed
578    
579    ----------------------------------------------------------------------
580    Name: Matthias Blume (blume (at) tti - c (dot) org)
581    Date: 2004/10/18 16:45:00 CDT
582    Tag: blume-20041018-groupowner
583    Description:
584    
585    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
586    
587    Eliminated group owner specs from .cm files throughout the source tree.
588    
589    ----------------------------------------------------------------------
590    Name: Matthias Blume (blume (at) tti - c (dot) org)
591    Date: 2004/10/15 15:45:00 CDT
592    Tag: blume-20041015-coverage
593    Description:
594    
595    * Test coverage tool added!
596    * Further reorganization of tracing-, debugging-, and profiling support:
597    
598        - moved original BTImp -- now called BackTrace -- into a separate
599          library called $/trace-debug-profile.cm
600        - eliminated all mentions of BTrace from SMLofNJ.Internals
601        - only the instrumentation mechanism is now left in the compiler proper
602        - BackTrace module is a plugin which is NOT plugged in by default
603        - Coverage module is another such plugin
604    
605    To get the benefits of any of these plugin modules, the code in
606    question must be compiled with tdp instrumentation turned on.  This
607    can be done by setting SMLofNJ.Internals.TDP.mode to true.
608    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
609    
610    Plugins are selected at link time.  (Pre-compiled instrumented code
611    can be re-loaded with different plugins in effect.)  When an
612    instrumented module is linked, whatever plugins are at that time
613    enabled will come into effect for that module.
614    
615    To enable the back-trace plugin, load library $/trace-debug-profile.cm
616    and invoke BackTrace.install() (e.g., from the interactive prompt).
617    To enable the coverage plugin, load the same library and invoke
618    Coverage.install().
619    
620    Back-traces are generated automatically on uncaught exceptions and
621    when the code in question explicitly invokes BackTrace.trigger().
622    
623    Coverage (and execution frequency-) information must be queried
624    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
625    
626    ----------------------------------------------------------------------
627    Name: Matthias Blume (blume (at) tti - c (dot) org)
628    Date: 2004/10/14 17:40:00 CDT
629    Tag: blume-20041014-tdp-core
630    Description:
631    
632    Snapshot of a significant overhaul of how the trace/debug/profile support
633    is hooked into the system (specifically: Core and SMLofNJ.Internals).
634    
635    ----------------------------------------------------------------------
636    Name: Matthias Blume (blume (at) tti - c (dot) org)
637    Date: 2004/10/13 16:34:00 CDT
638    Tag: blume-20041013-tdp
639    Description:
640    
641    Some rationalization of names:
642    
643       structure BTrace -> structure TDPInstrument
644       etc.
645    
646    This is is preparation of using the original back-trace
647    instrumentation for other purposes.  "TDP" stands for
648    Trace/Debug/Profile.
649    
650    The control flag controlling whether instrumentation is on or off is now
651    registered under a different name, so instead of running sml as
652    
653     sml -Cinstrument.btrace-mode=true
654    
655    one has to say
656    
657     sml -Ctdp.instrument=true
658    
659    ----------------------------------------------------------------------
660    Name: Matthias Blume (blume (at) tti - c (dot) org)
661    Date: 2004/10/11 16:37:00 CDT
662    Tag: blume-20041011-regions
663    Description:
664    
665    Made some minor modifications to elabcore.sml to have source regions
666    be propagated more tightly -- resulting in better (i.e., smaller)
667    regions being reported in error- and debug messages.
668    
669    ----------------------------------------------------------------------
670    Name: Matthias Blume (blume (at) tti - c (dot) org)
671    Date: 2004/10/08 22:50:00 CDT
672    Tag: blume-20041008-cmkw
673    Description:
674    
675    Fixed handling of keywords in .cm files: After seeing "is" the lexer
676    treats subsequent occurrences of "group", "library", "source", "is",
677    "*", and "-" as ordinary identifiers rather than keywords.
678    
679    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
680    is supposed to accept a tool argument called "source", but this did
681    not work because of the clash with the keyword.
682    
683    ----------------------------------------------------------------------
684    Name: Matthias Blume (blume (at) tti - c (dot) org)
685    Date: 2004/10/07 16:00:00 CDT
686    Tag: blume-20041007-cleanup
687    Description:
688    
689    Assorted cleanup work:
690    
691      - got rid of intstrmap in favor of using the library's
692        hash table implementation
693      - threw out most of the pathnames stuff, as it was not used anyway
694      - simplified tokentable implementation
695      - fixed some minor spelling errors
696    
697    ----------------------------------------------------------------------
698    Name: Matthias Blume (blume (at) tti - c (dot) org)
699    Date: 2004/10/06 15:15:15 CDT
700    Tag: blume-20041006-handler
701    Description:
702    
703    Cleaned up the absyn to reflect the invariant that HANDLE always
704    carries a FNexp as part of the type definition.  This eliminates some
705    superfluous sanity checks at runtime down the road.
706    
707    Some minor cleanup of the btrace code.
708    
709    ----------------------------------------------------------------------
710    Name: Matthias Blume (blume (at) tti - c (dot) org)
711    Date: 2004/10/01 10:20:30 CDT
712    Tag: blume-20041001-slave
713    Description:
714    
715    Added hack to make slave mode work in the presence of the version
716    tool.  (Still, since the master does two passes over the code for
717    CMB.make, the release number gets bumped twice when slaves are
718    attached.  I don't know if this is worth fixing...)
719    
720    ----------------------------------------------------------------------
721    Name: Matthias Blume (blume (at) tti - c (dot) org)
722    Date: 2004/09/30 10:55:00 CDT
723    Tag: blume-20040930-version
724    Description:
725    
726    * Moved the "version" magic into its own little library under
727      src/system/smlnj/internal.  This avoids expensive reconstruction of
728      a stable src/compiler/core.cm.
729    
730    * At the same time, structure CompilerVersion is now known as
731      structure SMLNJVersion.
732    
733    * Arranged for the version tool to NOT kick in when rebuilding the system
734      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
735      Also, loading the versiontool does not work when rebuilding the system
736      because CM is not properly initialized at that time.
737    
738    ----------------------------------------------------------------------
739    Name: Matthias Blume (blume (at) tti - c (dot) org)
740    Date: 2004/09/29 14:00:00 CDT
741    Tag: blume-20040929-autoversion
742    Description:
743    
744    Implemented some CM magic to have
745        file src/compiler/TopLevel/main/version.sml
746    generated automagically.
747    The version is taken from two files: config/version and config/release.
748    The first is expected to contain a two-part version number such as 110.49.
749    The second should contain a single number, but it may be missing.
750    
751    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
752    time the version tool is loaded (which is the first time you say CMB.make),
753    then the tool will increment the value stored in config/release every
754    time CMB.make is invoked.
755    
756    The binfile format is now insensitive to anything beyond the first
757    two components of a version number, so bumping the release does not render
758    binfiles incompatible.  Auto-bumping can be used to keep track of versions
759    during development without invalidating existing binfiles.
760    
761    In any case, every CMB.make updates the date information in version.sml.
762    (This is the date that is printed in the banner.)
763    
764    ----------------------------------------------------------------------
765    Name: Matthias Blume (blume (at) tti - c (dot) org)
766    Date: 2004/09/28 10:53:00 CDT
767    Tag: blume-20040928-controls
768    Description:
769    
770    Some cleanup of the controls code.
771    
772    ----------------------------------------------------------------------
773    Name: Matthias Blume (blume (at) tti - c (dot) org)
774    Date: 2004/09/27 22:08:00 CDT
775    Tag: blume-20040927-controls
776    Description:
777    
778    Added two pieces of functionality to the Controls interface:
779    
780    1.   val save'restore: 'a control -> unit -> unit
781    
782       grabs the current value of the control in stage 1 and restores it
783       in stage 2.
784    
785    2.   val set' : 'a control * 'a -> unit -> unit
786    
787       stores the given value into the control in stage 2 (i.e., delayed)
788       but does all error checking in stage 1.
789       (This is for string controls that need to do parse their argument --
790        something that might fail.  In some cases, notably in CM, one
791        already knows the intended argument but wants to delay the actual
792        assignment until a time when error recovery would be more difficult.)
793    
794    Changed the handling of controls in tool arguments to classes "sml" and
795    "lazysml":
796       - use Controls.save'restore as a more robust way of restoring the
797         old value (in particular: without having to re-parse the string)
798       - use controls to handle the "overload" keyword in the init group
799         (I believe this change actually fixes a long-standing obscure bug.)
800    
801    ----------------------------------------------------------------------
802    Name: Matthias Blume (blume (at) tti - c (dot) org)
803    Date: 2004/09/27 17:00:00 CDT
804    Tag: blume-20040927-lazysml
805    Description:
806    
807    Added a new tool class called "lazysml" to CM's tool chest.  The only
808    difference to "sml" is that compilation is done with Control.lazysml
809    set to true.  A source of class "lazysml" is automatically recognized
810    by a file name suffix of ".lml".
811    
812    In addition to the above feature, the original class "sml" now also
813    supports a tool argument "lazy" which has the same effect.  As a
814    result, the following three lines are equivalent:
815    
816        foo.sml : lazysml
817        foo.sml : sml (lazy)
818        foo.sml (lazy)
819    
820    The setting goes into effect both during parsing and during
821    compilation.  The original setting is restored right after parsing and
822    after compilation, respectively.
823    
824    In addition to all the above, there is also a general mechanism to set
825    ANY of the "controls" that are available at the command line via
826    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
827    apply as well. (In fact, "lazy" is implemented as a special case of
828    the general mechanism.)
829    
830    The .cm file syntax uses a new keyword tool argument called "with".
831    There are several ways of indicating the desired settings:
832    
833        foo.sml (with:parser.quotations=true)
834        foo.sml (with:(name:parser.quotations value:true))
835        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
836        foo.sml (with:(name1=value1 name2=value2 ...))
837        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
838    
839        etc.
840    
841    Another possible abbreviation is to leave out the =v or value:v part
842    if the name refers to a boolean control (in which case the value is
843    taken to be true).  Thus, one could get lazy sml also by saying:
844    
845        foo.sml (with:parser.lazy-keyword=true)
846        foo.sml (with:parser.lazy-keyword)
847        foo.sml (with:(name:parser.lazy-keyword value:true))
848        foo.sml (with:(name:parser.lazy-keyword))
849    
850    ----------------------------------------------------------------------
851    Name: Matthias Blume (blume (at) tti - c (dot) org)
852    Date: 2004/09/24 16:22:00 CDT
853    Tag: blume-20040924-ppc-long-branch
854    Description:
855    
856    Turned message about "emiting long form of branch" off by default.
857    Added a control flag to turn it back on when desired.
858    
859    ----------------------------------------------------------------------
860    Name: Matthias Blume (blume (at) tti - c (dot) org)
861    Date: 2004/09/24 16:05:00 CDT
862    Tag: blume-20040924-rounding
863    Description:
864    
865    Applied patch for setting rounding modes under Mac OS X.  Thanks to
866    Melissa O'Neill for providing the code!
867    
868    ----------------------------------------------------------------------
869    Name: Matthias Blume (blume (at) tti - c (dot) org)
870    Date: 2004/09/23 17:30:00 CDT
871    Tag: blume-20040923-envvars
872    Description:
873    
874    1. Changed definition of type ControlRegistry.registry_tree to
875       include control_info (i.e., the name of the controlling environment
876       variable).
877    
878    2. Added command-line flags -e and -E to print the names of environment
879       variables that can be used to control internal settings.
880       (This uses the new API mentioned in 1.)
881    
882    ----------------------------------------------------------------------
883    Name: Matthias Blume (blume (at) tti - c (dot) org)
884    Date: 2004/09/13 16:50:00 CDT
885    Tag: Release_110_49
886    Description:
887    
888    New working version (110.49).  NEW BOOTFILES!
889    
890    ----------------------------------------------------------------------
891    Name: Matthias Blume
892    Date: 2004/09/13 16:20:00 CDT
893    Tag: blume-20040913-config-mlrisc
894    Description:
895    
896    Put target "mlrisc" back into the default list.
897    (There is no harm in having it, and some users have expressed their
898    wish to have "mlrisc" included by default.)
899    
900    ----------------------------------------------------------------------
901    Name: John Reppy
902    Date: 2004/09/13
903    Tag: jhr-20040913-signals
904    Description:
905    
906    Fixed the signal masking code to properly nest mask/unmask operations
907    on a per-signal basis.
908    
909    ----------------------------------------------------------------------
910    Name: Matthias Blume (blume (at) tti - c (dot) org)
911    Date: 2004/09/08 13:20:00 CDT
912    Tag: blume-20040908-heap-magic
913    Description:
914    
915    Bumped the heap macig number to 0x09082004 to account for the changed
916    layout of the ML frame under MacOS X.
917    
918    ----------------------------------------------------------------------
919    Name: Allen Leung (leunga (at) reservoir (dot) com)
920    Date: 2004/09/03 11:26:00 EST
921    Tag: leunga-20040903-cygwin-install
922    Description:
923    
924       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
925    The Cygwin runtime is turned on by setting the environment
926    variable SMLNJ_CYGWIN_RUNTIME to 1.
927    
928    ----------------------------------------------------------------------
929    Name: Matthias Blume (blume (at) tti - c (dot) org)
930    Date: 2004/08/31 17:15:00 CDT
931    Tag: blume-20040831-core
932    Description:
933    
934    Added some exports to src/compiler/core.cm upon request by J. Joemann.
935    
936    ----------------------------------------------------------------------
937    Name: Matthias Blume (blume (at) tti - c (dot) org)
938    Date: 2004/08/30 17:55:00 CDT
939    Tag: blume-20040830-installer
940    Description:
941    
942    Upon request by Johannes Joemann:
943    
944    - improved ML code of installer to fall back to coping when renaming
945      fails (i.e., when source and target are on different file systems);
946      the code compiles but has yet to be tested in anger
947    - removed mlrisc from list of default targets (config/targets)
948    
949    ----------------------------------------------------------------------
950    Name: Matthias Blume (blume (at) tti - c (dot) org)
951    Date: 2004/08/27 17:20:00 CDT
952    Tag: blume-20040827-ptreql
953    Description:
954    
955    Added ptreql primop to structure InlineT (upon request from Larry
956    Paulson).
957    
958    ----------------------------------------------------------------------
959    Name: Allen Leung (leunga (at) reservoir (dot) com)
960    Date: 2004/08/15 21:21:00 EST
961    Tag: leunga-110_48-udgraph
962    Description:
963    
964    Another bug fix from Carl Hauser:
965    
966    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
967    > 48c48
968    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
969    > e::L)
970    > ---
971    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
972    > Without this, any deletion of an edge in an undirected graph does severe
973    > violence to the graph.
974    
975    ----------------------------------------------------------------------
976    Name: Allen Leung (leunga (at) reservoir (dot) com)
977    Date: 2004/08/10 23:35:00 EST
978    Tag: leunga-110_48-ppc
979    Description:
980    
981       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
982    
983    ----------------------------------------------------------------------
984    Name: Matthias Blume (blume (at) tti - c (dot) org)
985    Date: 2004/08/10 12:00:00 CDT
986    Tag: Release_110_48
987    Description:
988    
989    New working version (110.48).  NEW BOOTFILES!
990    
991    ----------------------------------------------------------------------
992    Name: Allen Leung (leunga (at) reservoir (dot) com)
993    Date: 2004/08/09 12:21:00 EST
994    Tag: leunga-110_47-dijsktra
995    Description:
996    
997    Bug fix from Carl Hauser:
998    
999    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1000    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1001    
1002    The problem is that following the expression A.update(dist,s,Num.zero)
1003    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1004    
1005    ----------------------------------------------------------------------
1006    Name: Matthias Blume (blume (at) tti - c (dot) org)
1007    Date: 2004/08/06 18:10:00 CDT
1008    Tag: blume-20040806-cmdline
1009    Description:
1010    
1011    Fiddled with handling of command-line options:
1012    
1013      * sml now quits after processing the command line
1014        if -H, -S, -h<n>, or -s<n> appears as the last
1015        command-line argument
1016      * a new option -q terminates the session when encountered on
1017        the command line; subsequent arguments will be ignored
1018      * bug fixes: short (erroneous) arguments are no longer ignored
1019        completely
1020    
1021    ----------------------------------------------------------------------
1022    Name: Allen Leung (leunga (at) reservoir (dot) com)
1023    Date: 2004/08/04 18:17:00 EST
1024    Tag: leunga-110_47-ppc-ibm-asm
1025    Description:
1026    
1027      - Added minimal IBM assembly syntax support for PowerPC.
1028    
1029      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1030        been altered in the runtime.  We need an automatic way to keep the file
1031        in sync.
1032    
1033    ----------------------------------------------------------------------
1034    Name: Matthias Blume (blume (at) tti - c (dot) org)
1035    Date: 2004/08/04 14:00:00 CDT
1036    Tag: Release_110_47
1037    Description:
1038    
1039    New working version (110.47).  NEW BOOTFILES!
1040    
1041    ----------------------------------------------------------------------
1042    Name: Matthias Blume (blume (at) tti - c (dot) org)
1043    Date: 2004/08/03 14:25:00 CDT
1044    Tag: blume-20040803-callingconv
1045    Description:
1046    
1047    Added low-level support for choosing C calling conventions by
1048    twiddling the type of rawccall.  (See
1049    src/compiler/Semant/types/cproto.sml for details.)
1050    
1051    ----------------------------------------------------------------------
1052    Name: Matthias Blume (blume (at) tti - c (dot) org)
1053    Date: 2004/08/02 15:55:00 CDT
1054    Tag: blume-20040802-backout
1055    Description:
1056    
1057    Backed out of change to win32-filesys.c.  The earlier patch to
1058    get_file_time caused CM to produce files with the wrong time stamp.
1059    
1060    ----------------------------------------------------------------------
1061    Name: Matthias Blume (blume (at) tti - c (dot) org)
1062    Date: 2004/08/02 14:45:00 CDT
1063    Tag: blume-20040802-nlffi-win32
1064    Description:
1065    
1066    Added NLFFI support for Win32, adapted from a patch provided by David
1067    Hansel.  This is currently completely untested.  Also, the issue
1068    concerning stdcall vs. ccall is still unresolved.
1069    
1070    ----------------------------------------------------------------------
1071    Name: Matthias Blume (blume (at) tti - c (dot) org)
1072    Date: 2004/07/30 17:55:00 CDT
1073    Tag: blume-20040730-various
1074    Description:
1075    
1076    Gearing up towards 110.47...
1077    
1078    - various minor bugfixes to ml-nlffigen
1079    - a beginning of a manual for nlffi
1080    
1081    - eliminated 'export name=value' in config/install.sh as this does
1082      not work with certain versions of /bin/sh
1083      (Thanks to David King at Motorola for catching this.)
1084    
1085    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1086      - added a test for tm==NULL to gmtime.c and localtime.c
1087      - applied patch for incorrect GetFileTime under win32
1088      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1089    
1090    ----------------------------------------------------------------------
1091    Name: Matthias Blume (blume (at) tti - c (dot) org)
1092    Date: 2004/07/21 18:20:00 CDT
1093    Tag: blume-20040721-nlffigen
1094    Description:
1095    
1096    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1097      for a typedef to an incomplete type, but leave out the "typ" member.
1098      (This is just for consistency.)
1099    - Started to produce what is supposed to become better (i.e., comprehensive)
1100      documentation of what ml-nlffigen does and produces.
1101    
1102    ----------------------------------------------------------------------
1103    Name: Matthias Blume (blume (at) tti - c (dot) org)
1104    Date: 2004/07/14 16:25:00 CDT
1105    Tag: blume-20040714-union
1106    Description:
1107    
1108    Added C_UNION to c-calls/c-types.sml and updated the machinery
1109    (ml-nlffigen, cproto.sml) that conveys C function interface
1110    information to the code generator.
1111    
1112    However, the actual architecture-specific implementation of function
1113    arguments and results that are C unions is still not implemented.
1114    
1115    ----------------------------------------------------------------------
1116    Name: Allen Leung (leunga (at) reservoir (dot) com)
1117    Date: 2004/07/14 14:38:00 EST
1118    Tag: leunga-110_46_1-ppc-lwzu
1119    Description:
1120    
1121        Added these instructions to the PowerPC architecture:
1122        LBZU(X), LHZU(X), LWZU(X),
1123        STWU(X), STFDU, STFSU
1124    
1125        etc...
1126    
1127        Note: I haven't added their instruction encoding into the description.
1128    
1129    ----------------------------------------------------------------------
1130    Name: Allen Leung (leunga (at) reservoir (dot) com)
1131    Date: 2004/07/13 15:04:00 EST
1132    Tag: leunga-110_46_1-ppc-lwarx
1133    Description:
1134    
1135       Added the two instructions LWARX and STWCX to the PowerPC
1136    instruction set.
1137    
1138       A (untested) rewrite of loop-structure.sml.  The old version
1139    is completely broken.
1140    
1141    ----------------------------------------------------------------------
1142    Name: Matthias Blume (blume (at) tti - c (dot) org)
1143    Date: 2004/07/13 13:50:00 CDT
1144    Tag: blume-20040713-nlffi
1145    Description:
1146    
1147    - use paramAlloc to report c-calls with too many arguments
1148      (for PPC version where parameter area is pre-allocated)
1149    - added ccall_maxargspace to machspec (to implement the above)
1150    - made "make" commend in CM's "make" tool configurable
1151    - added option (default: on) for passing the name of the SML/NJ's "bin"
1152      directory to "make"; the call looks like this:
1153    
1154         make <options> SMLNJ_BINDIR=<dir> <target>
1155    
1156      This can be used by the Makefile to, e.g., pick the "right" version
1157      of ml-nlffigen.
1158    - minor code tweaks
1159    
1160    ----------------------------------------------------------------------
1161    Name: Matthias Blume (blume (at) tti - c (dot) org)
1162    Date: 2004/07/12 22:50:00 CDT
1163    Tag: blume-110_46_1-macosx-nlffi
1164    Description:
1165    
1166    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1167    though.
1168    
1169    Note:  1. You have to make a new, clean build of the runtime system.
1170           2. There are new BOOTFILES, you have to use them!
1171              (Doing the bootstrap process yourself would be *very* painful!
1172               If you absolutely have to do it, build the system under
1173               a different architecture and then cross-compile.)
1174    
1175    Version bumped to 110.46.1 to account for runtime data format changes.
1176    
1177    ----------------------------------------------------------------------
1178    Name: Matthias Blume (blume (at) tti - c (dot) org)
1179    Date: 2004/06/18 14:30:00 CDT
1180    Tag: blume-20040618-unix
1181    Description:
1182    
1183    Changed the implementation of structure Unix so that the same stream
1184    is returned every time one of the {text,bin}{In,Out}streamOf functions
1185    is invoked on the same proc.  This is not what the spec currently
1186    says -- although IMO it arguably should.  (See discussion below.)
1187    
1188    ----------------------------------------------------------------------
1189    Name: Matthias Blume (blume (at) tti - c (dot) org)
1190    Date: 2004/06/17 18:15:00 CDT
1191    Tag: Release_110_46
1192    Description:
1193    
1194    New working version (110.46).  NEW BOOTFILES!
1195    
1196    ----------------------------------------------------------------------
1197    Name: Matthias Blume (blume (at) tti - c (dot) org)
1198    Date: 2004/06/17 17:20:00 CDT
1199    Tag: blume-20040617-timer-unix
1200    Description:
1201    
1202    Changed the interface of structures Timer and Unix to match the most
1203    recent Basis spec.
1204    
1205    In the case of Unix there still seems to be an open/weird issue:
1206    
1207       The {text,bin}{In,Out}streamOf functions are supposed to create
1208       fresh streams whenever they are called -- as opposed to have them
1209       return the same stream every time.  This design is supposed to
1210       prevent space leaks caused by proc values hanging on to streams.
1211    
1212       The reap function, on the other hand, is supposed to close the
1213       streams.  This cannot be done without having a handle on the
1214       stream in proc after all...
1215    
1216    I took the liberty to implement the following stopgap solution:
1217    
1218      The proc value hangs on to the most recently created stream(s).
1219      Reap closes those.  If either or both of the two streams hadn't
1220      been created at all yet, then reap will close the corresponding
1221      file descriptors directly.
1222    
1223    PS: I don't understand the original space leak argument anymore.  If
1224    a proc hangs on to the imperative stream, then I/O operations on those
1225    will advance the state of the cached stream and avoid the space leak.
1226    
1227    ----------------------------------------------------------------------
1228    Name: Matthias Blume (blume (at) tti - c (dot) org)
1229    Date: 2004/05/28 16:45:00 CDT
1230    Tag: blume-20040528-basis
1231    Description:
1232    
1233    Added signature PACK_REAL and exported functor PrimIO.
1234    
1235    ----------------------------------------------------------------------
1236    Name: Matthias Blume (blume (at) tti - c (dot) org)
1237    Date: 2004/05/25 16:00:00 CDT
1238    Tag: blume-20040525-group-owner
1239    Description:
1240    
1241    CM now ignores (but still accepts) the "owner" information in group
1242    descriptions.  The owner of a group is its next enclosing
1243    library. Each group must have a unique owner.  (There is a virtual
1244    "toplevel" library that owns groups which are not nested within a real
1245    library.)  Previously, each group had to explicitly declare its owner,
1246    and CM would check that such a declaration is correct.  The new scheme
1247    is to have CM check that for each group there is precisely one owning
1248    library.
1249    
1250    The advantage of the new scheme is that the programmer no longer needs
1251    to maintain the somewhat annoying owner information.  The downside is
1252    that CM cannot enforce the ownership rule across multiple runs of
1253    CM.make.  Fortunately, enclosing the same group in two different
1254    libraries A and B which are not part of the same program does not
1255    cause real problems.
1256    
1257    ----------------------------------------------------------------------
1258    Name: Matthias Blume (blume (at) tti - c (dot) org)
1259    Date: 2004/05/20 16:00:00 CDT
1260    Tag: blume-20040520-win32
1261    Description:
1262    
1263    Made the win32 version work again.  (Strangely, a misplaced comma had
1264    slipped into win32-process.c which prevented the runtime from being
1265    compiled correctly.)
1266    
1267    Also, included a minor addition to ml-build.bat analogous to what was
1268    done in blume-20040519-ml-build.
1269    
1270    ----------------------------------------------------------------------
1271    Name: Matthias Blume (blume (at) tti - c (dot) org)
1272    Date: 2004/05/19 22:10:00 CDT
1273    Tag: blume-20040519-ml-build
1274    Description:
1275    
1276    Arranged for ml-build to clean up after itself a little bit better.
1277    The script generates a temporary SML source file and compiles it using
1278    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1279    gets rid of those at the end, so they don't accumulate under .cm.
1280    
1281    This required a minor change to install.sh because the name of the
1282    metadata directory (default: .cm) is actually configurable at
1283    installation time.
1284    
1285    ----------------------------------------------------------------------
1286    Name: Matthias Blume (blume (at) tti - c (dot) org)
1287    Date: 2004/05/18 15:50:00 CDT
1288    Tag: blume-20040518-mkreader
1289    Description:
1290    
1291    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1292    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1293    
1294    ----------------------------------------------------------------------
1295    Name: Matthias Blume (blume (at) tti - c (dot) org)
1296    Date: 2004/05/11 14:35:00 CDT
1297    Tag: blume-20040511-win32sock
1298    Description:
1299    
1300    Added previously missing support for many socket-related functions
1301    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1302    for the voluminous patch!
1303    
1304    (I have not tested this patch under win32 yet.)
1305    
1306    Here is David's e-mail:
1307    
1308    Hi,
1309    
1310    Attached to this email you find a diff against sml/nj 110.45
1311    that will enable socket support under Windows.
1312    
1313    To apply the patch (using unix or cygwin)
1314    1) gunzip runtime.diff.gz
1315    2) "cd" into "src/runtime" in the source tree of a fresh
1316       110.45 installation.
1317    3) patch -p 1 < [your/path/to]runtime.diff
1318    
1319    The code compiles fine but has NOT yet been extensively tested.
1320    I only ran a few tests for basic socket client functionality
1321    (which worked fine).  Especially the functions that use ioctl
1322    are not tested at all and might not work (see below).
1323    
1324    I implemented this since we want to move to a newer version of sml/nj
1325    but need socket support in order to use it.  This is the first time I
1326    even had a look at the sml/nj source,  so please review my changes
1327    before making this part of the distribution!  Here are a few issues
1328    that I think might be better for someone to solve who is more
1329    familiar with the sml/nj source (and socket programming):
1330    
1331    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1332      exception since I could not figure out what the windows equivalent
1333      of these functions is
1334    
1335    - In sockets-osdep.h there are a some #include statements that are
1336      only used in a few files that include sockets-osdep.h
1337    
1338    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1339      does not process its return value since I don't know how to
1340      report an error upwards.
1341    
1342    - It would probably be good to have a call to WSACleanup() when
1343      the library is unloaded (if there is such a possibility).
1344      Otherwise I think Windows will take care of this automatically
1345      when the process finishes.
1346    
1347    - I used ioctlsocket() as a replacement for ioctl() but I have
1348      no idea if that is actually the proper replacement on Windows.
1349    
1350    - All these issues are marked in the code by "FIXME" comments.
1351    
1352    We use sml/nj extensively in our products and are quite happy
1353    with it.  I hope this contribution will help you.
1354    
1355    Keep up the good work!
1356    
1357    David
1358    
1359    ----------------------------------------------------------------------
1360    Name: Matthias Blume (blume (at) tti - c (dot) org)
1361    Date: 2004/05/11 14:20:00 CDT
1362    Tag: blume-20040511-installml
1363    Description:
1364    
1365    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1366    the patch.)
1367    
1368    ----------------------------------------------------------------------
1369    Name: Matthias Blume (blume (at) tti - c (dot) org)
1370    Date: 2004/05/11 14:05:00 CDT
1371    Tag: blume-20040511-nlffi-netbsd
1372    Description:
1373    
1374    Added support for nlffi under netbsd.  (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 12:05:00 CDT
1380    Tag: blume-20040511-exports
1381    Description:
1382    
1383    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1384    export lists in compiler-related .cm-files.
1385    
1386    ----------------------------------------------------------------------
1387    Name: Matthias Blume (blume (at) tti - c (dot) org)
1388    Date: 2004/05/11 11:35:00 CDT
1389    Tag: blume-20040511-allsource
1390    Description:
1391    
1392    The installer now honors the "src-smlnj" target again, although its meaning
1393    has changed from "all sources required for the compiler" to "all sources
1394    the installer knows about".  In other words, if you enable "src-smlnj"
1395    in the "targets" file, then the installer will pull in sources for
1396    everything.  (Notice that this refers to source code only.  Compiled
1397    code is still only installed for modules that were requested explicitly
1398    or which are required for other modules that were requested explicitly.)
1399    
1400    ----------------------------------------------------------------------
1401    Name: Matthias Blume (blume (at) tti - c (dot) org)
1402    Date: 2004/04/23 17:40:00 CDT
1403    Tag: blume-20040423-ieee-scan
1404    Description:
1405    
1406    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1407    in the exponent calculation we get INF or ZERO (depending on the mantissa
1408    and the sign of the exponent).
1409    
1410    ----------------------------------------------------------------------
1411    Name: Matthias Blume (blume (at) tti - c (dot) org)
1412    Date: 2004/04/23 10:40:00 CDT
1413    Tag: blume-20040423-ml-build
1414    Description:
1415    
1416    The ml-build script now terminates with a non-0 status when something
1417    goes wrong.
1418    
1419    ----------------------------------------------------------------------
1420    Name: Matthias Blume (blume (at) tti - c (dot) org)
1421    Date: 2004/04/22 16:35:00 CDT
1422    Tag: blume-20040422-Option
1423    Description:
1424    
1425    Made exception Option to be the same as exception Option.Option
1426    (as it should be).
1427    
1428    ----------------------------------------------------------------------
1429    Name: Allen Leung (leunga (at) reservoir (dot) com)
1430    Date: 2004/03/19 14:40:00 EST
1431    Tag: leunga-20040319-cygwin-nlffi
1432    Description:
1433    
1434    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1435    of SML/NJ.  The problem is that
1436    
1437       lib = dlopen(NULL, ...)
1438       f   = dlsym(lib, "malloc");
1439    
1440    does not work on Windows unless we explicitly export symbols
1441    such as 'malloc' during linking.  We fixed this by explicitly
1442    exporting the required symbols with the magic gcc incantation:
1443    
1444        -Wl,--export-all cygwin.def
1445    
1446    where cygwin.def is a file containing all the symbols that
1447    we wish to export.
1448    
1449    I suspect this is a Windows problem and we'll have to
1450    do the same (somehow with windows compilers) when
1451    we build the native win32 version with the system
1452    calls LoadLibrary/GetProcAddress.
1453    
1454    ----------------------------------------------------------------------
1455    Name: Matthias Blume (blume (at) tti - c (dot) org)
1456    Date: 2004/03/04 16:35:00 CST
1457    Tag: blume-20040304-intinf-fmt
1458    Description:
1459    
1460    Fixed problem with IntInf.fmt (sign would show up on the right instead
1461    of on the left for BIN, OCT, and HEX).
1462    
1463    ----------------------------------------------------------------------
1464    Name: Matthias Blume (blume (at) tti - c (dot) org)
1465    Date: 2004/03/04 11:25:00 CST
1466    Tag: blume-20040304-symlinks
1467    Description:
1468    
1469    Fixed problem with installer script (unix only) where bin/ml-yacc and
1470    friends pointed (via symlinks) to absolute locations instead of just
1471    .run-sml.  This was reported by Vesa A Norrman.
1472    
1473    ----------------------------------------------------------------------
1474    Name: Matthias Blume (blume (at) tti - c (dot) org)
1475    Date: 2004/02/13 14:50:00 CST
1476    Tag: Release_110_45
1477    Description:
1478    
1479    New working version (110.45). New bootfiles.
1480    
1481    ----------------------------------------------------------------------
1482    Name: Matthias Blume (blume (at) tti - c (dot) org)
1483    Date: 2004/01/26 15:15:15 CST
1484    Tag: blume-20040126-toplevel
1485    Description:
1486    
1487    Improved handling of exceptions at the interactive toplevel.
1488    
1489    ----------------------------------------------------------------------
1490    Name: Matthias Blume (blume (at) tti - c (dot) org)
1491    Date: 2004/01/26 11:25:00
1492    Tag: blume-20040126-app
1493    Description:
1494    
1495    Type of top-level "app" corrected.
1496    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1497    
1498    ----------------------------------------------------------------------
1499    Name: Matthias Blume (blume (at) tti - c (dot) org)
1500    Date: 2003/11/18 17:10 CST
1501    Tag: blume-20031118-basis-fiddle
1502    Description:
1503    
1504    - changed Timer interface to what might become the spec
1505    - POSIX_FLAGS -> BIT_FLAGS according to spec
1506    - some other minor discrepancies wrt. spec eliminated
1507    
1508    ----------------------------------------------------------------------
1509    Name: Matthias Blume (blume (at) tti - c (dot) org)
1510    Date: 2003/11/06 12:00:00 CST
1511    Tag: Release_110_44
1512    Description:
1513    
1514    New working version (110.44). New bootfiles.
1515    
1516    ----------------------------------------------------------------------
1517    Name: Matthias Blume (blume (at) tti - c (dot) org)
1518    Date: 2003/11/04 11:50:00 CST
1519    Tag: blume-20031104-move-libraries
1520    Description:
1521    
1522    Eliminated the "dont_move_libraries" directive in config/targets.
1523    (The mechanism was broken and could not be fixed easily.  Moreover,
1524    there does not seem to be any reason not to move all libraries into
1525    lib during installation.  I originally implemented this directive as a
1526    backward-compatibility feature when I first introduced the new CM.
1527    Now that things have been stable for a long time and going back to the
1528    old CM is not an option, there is no reason to keep it around.)
1529    
1530    ----------------------------------------------------------------------
1531    Name: Matthias Blume (blume (at) tti - c (dot) org)
1532    Date: 2003/11/03 16:00:00 CST
1533    Tag: blume-20031103-installdir
1534    Description:
1535    
1536    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1537    Richards for pointing out the problem and providing the solution.)
1538    
1539    ----------------------------------------------------------------------
1540    Name: Matthias Blume (blume (at) tti - c (dot) org)
1541    Date: 2003/10/01 17:05:00 CDT
1542    Tag: blume-20031001-lal-mlrisc
1543    Description:
1544    
1545    MLRISC bug fix from Lal.
1546    
1547    ----------------------------------------------------------------------
1548    Name: Matthias Blume (blume (at) tti - c (dot) org)
1549    Date: 2003/09/30 16:10:00 CDT
1550    Tag: blume-20030930-primio-bat
1551    Description:
1552    
1553    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1554    2. Improved .bat files (for Win32 port) to make things work under Win95.
1555       (thanks to Aaron S. Hawley for this one)
1556    
1557    ----------------------------------------------------------------------
1558    Name: Matthias Blume (blume (at) tti - c (dot) org)
1559    Date: 2003/09/26 16:05:00 CDT
1560    Tag: blume-20030926-wrappriv
1561    Description:
1562    
1563    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1564    
1565    ----------------------------------------------------------------------
1566    Name: Matthias Blume (blume (at) tti - c (dot) org)
1567    Date: 2003/09/26 15:00:00 CDT
1568    Tag: blume-20030926-110_43_3
1569    Description:
1570    
1571    - additional cleanup
1572    - version number bump, NEW BOOTFILES
1573    
1574    ----------------------------------------------------------------------
1575    Name: Matthias Blume (blume (at) tti - c (dot) org)
1576    Date: 2003/09/26 12:00:00 CDT
1577    Tag: blume-20030926-ppautoload
1578    Description:
1579    
1580    I modified the read-eval-print loop so that the autoloader gets
1581    invoked whenever the prettyprinter tries to look up a symbol that
1582    is not currently defined in the toplevel environment but which
1583    appears in CM's autoload registry.  As a result, we see far fewer of
1584    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1585    
1586    In addition to this I tried to clean up some pieces of the Basis
1587    implementation (e.g., Socket, Word8Array) in order to prevent other
1588    instances of these ?.Foo.Bar.xxx names from being printed.
1589    
1590    The mechanism that picks names for types still needs some work, though.
1591    (Right now it seems that if there is a type A.t which is defined to
1592    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1593    "?.B.u" although the perhaps more sensible solution would be to use
1594    "A.t" in this case.  In other words, the prettyprinter should follow
1595    a chain of DEFtycs not farther than there are corresponding toplevel
1596    names in the current environment.)
1597    
1598    ----------------------------------------------------------------------
1599    Name: Matthias Blume (blume (at) tti - c (dot) org)
1600    Date: 2003/09/24 16:31:00 CDT
1601    Tag: blume-20030924-installer
1602    Description:
1603    
1604    Another installer tweak:  All the ML code for the installer is now
1605    compiled during CMB.make and put into a little library called
1606    $smlnj/installer.cm.  The installation then simply invokes
1607    
1608       sml -m $smlnj/installer.cm
1609    
1610    and everything happens automagically.
1611    
1612    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1613    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1614            and BINDIR.
1615    
1616    The new scheme guarantees that the ML code responsible for the installation
1617    is in sync with the APIs of the main system.  Also, the installer is
1618    somewhat faster because the installer script is precompiled.
1619    
1620    ----------------------------------------------------------------------
1621    Name: Matthias Blume (blume (at) tti - c (dot) org)
1622    Date: 2003/09/24 15:35:00 CDT
1623    Tag: blume-20030924-synsock
1624    Description:
1625    
1626    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1627    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1628    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1629    implementations of non-blocking operations from CML's Socket
1630    structure.
1631    
1632    ----------------------------------------------------------------------
1633    Name: Matthias Blume (blume (at) tti - c (dot) org)
1634    Date: 2003/09/24 15:10:05 CDT
1635    Tag: blume-20030924-sockets
1636    Description:
1637    
1638    1. Fixed SOCKET API and implementation to match Basis spec.
1639       This required changing the internal representation of sockets to one
1640       that remembers (for each socket file descriptor) whether it is currently
1641       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1642       call is made only if the state actually needs to change).
1643    
1644    2. OS-specific details of sockets were moved into separate files, thus
1645       making it possible to unify the bulk of the socket implementations
1646       between Unix and Win32.
1647    
1648    3. CML's socket API changed accordingly.
1649       (Note that we need to remove non-blocking functions from this API
1650       since they are redundant in the case of CML!)
1651    
1652    4. CML's socket implementation now makes use of non-blocking functions
1653       provided by Basis, thus removing all OS-dependent code from this part
1654       of CML.
1655    
1656    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1657    
1658    ----------------------------------------------------------------------
1659    Name: Matthias Blume (blume (at) tti - c (dot) org)
1660    Date: 2003/09/22 12:10:00 CDT
1661    Tag: blume-20030922-110_43_2
1662    Description:
1663    
1664    Made a new interim version and bootfiles for developer's bootstrapping
1665    convenience.
1666    
1667       110.43.2 -- NEW BOOTFILES
1668    
1669    ----------------------------------------------------------------------
1670    Name: Matthias Blume (blume (at) tti - c (dot) org)
1671    Date: 2003/09/19 15:55:00 CDT
1672    Tag: blume-20030919-cmdir
1673    Description:
1674    
1675    1. new-install.sh -> install.sh
1676    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1677    3. tweaked installer so that another name instead of .cm can be chosen
1678       at install time (by setting the CM_DIR_ARC environment variable
1679       during installation); once installation is complete, the name is
1680       fixed
1681    
1682    ----------------------------------------------------------------------
1683    Name: Matthias Blume (blume (at) tti - c (dot) org)
1684    Date: 2003/09/18 16:00:00 CDT
1685    Tag: blume-20030918-110_43_1
1686    Description:
1687    
1688    Made a new interim version and bootfiles for developer's bootstrapping
1689    convenience.
1690    
1691       110.43.1 -- NEW BOOTFILES
1692    
1693    ----------------------------------------------------------------------
1694    Name: Matthias Blume (blume (at) tti - c (dot) org)
1695    Date: 2003/09/18 15:20:00 CDT
1696    Tag: blume-20030918-misc
1697    Description:
1698    
1699    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1700       this seems to be controversial at the moment) and used those in
1701       Posix.ProcEnv.times.
1702    
1703    2. Added Time.{from,to}Nanoseconds to Time.
1704    
1705    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1706       For example, fromLargeInt never needs to look at more than 3 "big
1707       digits" to get its 53 bits of precision.
1708    
1709    ----------------------------------------------------------------------
1710    Name: Matthias Blume (blume (at) tti - c (dot) org)
1711    Date: 2003/09/17 16:30:00 CDT
1712    Tag: blume-20030917-real32-slices
1713    Description:
1714    
1715    Added an entry to the primitive environment
1716    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1717    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1718    
1719    Removed some of the "magic" constants in real64.sml and replaced them
1720    with code that generates these values from their corresponding
1721    integer counterparts.
1722    
1723    Made all(?) the slice-related changes to the Basis and made everything
1724    compile again...
1725    
1726    ----------------------------------------------------------------------
1727    Name: Matthias Blume (blume (at) tti - c (dot) org)
1728    Date: 2003/09/15 17:45:00 CDT
1729    Tag: blume-20030915-rbase
1730    Description:
1731    
1732    Fixed bug in Real.fromLargeInt.
1733    
1734    ----------------------------------------------------------------------
1735    Name: Matthias Blume (blume (at) tti - c (dot) org)
1736    Date: 2003/09/13 18:11:00 CDT
1737    Tag: blume-20030913-libinstall
1738    Description:
1739    
1740    Minor bugfix in config/libinstall (set anchor with path to
1741    standalone tool after installing it, otherwise libraries that
1742    need ml-lex or ml-yacc won't compile the first time the installer
1743    runs).
1744    
1745    ----------------------------------------------------------------------
1746    Name: Matthias Blume (blume (at) tti - c (dot) org)
1747    Date: 2003/09/12 11:45:00 CDT
1748    Tag: blume-20030912-various
1749    Description:
1750    
1751     - fixed bug in Real.toLargeInt
1752     - fixed bug in Posix.ProcEnv.times
1753     - changed inputLine functions to return an option
1754     - minor installer improvements / bugfixes
1755     - changed default @SMLalloc parameter for x86/celeron to 64k
1756    
1757    ----------------------------------------------------------------------
1758    Name: Matthias Blume (blume (at) tti - c (dot) org)
1759    Date: 2003/09/09 22:00:00 CDT
1760    Tag: Release_110_43
1761    Description:
1762    
1763        New working release 110.43.  New bootfiles.
1764    
1765    ----------------------------------------------------------------------
1766    Name: Matthias Blume (blume (at) tti - c (dot) org)
1767    Date: 2003/09/09 19:20:00 CDT
1768    Tag: blume-20030909-installer
1769    Description:
1770    
1771    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1772    Modified config/install.bat to take advantage of it.  Also modified
1773    config/install.sh (and called it config/new-install.sh) to take advantage
1774    of it on Unix systems.  (The SML code is (supposed to be) platform-
1775    independent.)
1776    
1777    The installer can now install everything under Win32
1778    as well as under *nix as long as it compiles.
1779    
1780    Other changes:
1781    
1782      - made CML compile again under Win32
1783      - made eXene compile under Win32 (by providing a fake structure UnixSock
1784        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1785      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1786        same as type int;  under Win32 it isn't
1787      - fixed some slice-related problems in the win32-specific parts of CML
1788      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1789        (under Win32, the volume name is case-insensitive, and the
1790        OS.Path code compares volume names for equality)
1791    
1792    ----------------------------------------------------------------------
1793    Name: Matthias Blume (blume (at) tti - c (dot) org)
1794    Date: 2003/09/08 11:55:00 CDT
1795    Tag: blume-20030908-fullpath
1796    Description:
1797    
1798    Made Win32 version of OS.FileSys.fullPath return current directory
1799    when given an empty string.  This is what the spec says, and incidentally,
1800    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1801    cases when presented with the name of a non-existing .cm file.)
1802    
1803    ----------------------------------------------------------------------
1804    Name: Matthias Blume (blume (at) tti - c (dot) org)
1805    Date: 2003/09/04 16:30:00 CDT
1806    Tag: blume-20030905-slices-etc
1807    Description:
1808    
1809      1. Changed interface to vectors and arrays in Basis to match
1810         (draft) Basis spec.
1811      2. Added signatures and implementations of slices according to
1812         Basis spec.
1813      3. Edited source code throughout the system to make it compile again
1814         under 1. and 2.  (In some cases code had to be added to have it
1815         match the new signatures.)
1816      4. MLRISC should be backward-compatible: the copies of the originals
1817         of files that needed to change under 3. were retained, the .cm files
1818         check the compiler version number and use old versions when
1819         appropriate.
1820      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1821         dirstream -> string option (in accordance with Basis spec).
1822      6. When generating code that counts lines, ml-lex used function
1823         CharVector.foldli, taking advantage of its old interface.
1824         This has been replaced with the corresponding code from
1825         CharVectorSlice.  (html-lex must be re-lexed!)
1826      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1827         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1828      8. Removed temporary additions (fromInternal, toInternal) from the
1829         (now obsolete) IntInf in smlnj-lib/Util.
1830      9. Cleaned up structure Byte.
1831     10. Added localOffset, scan, and fromString to Date (according to spec).
1832         Cleaned/corrected implementation of Date.
1833         (Still need to check for correctness; implement better canonicalizeDate.)
1834     11. Added "scan" to signature IEEE_REAL.
1835     12. Some improvements to IntInf [in particular: efficiency-hack for
1836         mod and rem when second operand is 2 (for parity checks).]
1837     13. Changed representation of type Time.time, using a single IntInf.int
1838         value counting microseconds.  This considerably simplified the
1839         implementation of structure Time.  We now support negative time
1840         values; scan and fromString handle signs.
1841     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1842         ArraySlice).
1843    
1844    ----------------------------------------------------------------------
1845    Name: Matthias Blume (blume (at) tti - c (dot) org)
1846    Date: 2003/08/28 17:00:00 CDT
1847    Tag: blume-20030828-intinf
1848    Description:
1849    
1850    This is a major update which comes with a version number bump
1851    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1852    and an implementation of IntInf in the Basis.
1853    
1854    There are a fairly large number of related changes and updates throughout
1855    the system:
1856    
1857    Basis:
1858            - Implemented IntInf.
1859            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1860            - Added some missing Real64 operations, most notably Real.toLargeInt.
1861            - Added FixedInt as a synonym for Int32.
1862    
1863    compiler:
1864            * Added support for a built-in intinf type.
1865              - literals
1866              - pattern matching
1867              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1868              - overloading on literals and operations
1869    
1870            This required adding a primitive type intinf, some additional
1871            primops, and implementations for several non-trivial intinf
1872            operations in Core.  (The intinf type is completely abstract
1873            to the compiler; all operations get delegated back to the Core.)
1874    
1875            * Intinf equality is handled by polyequal.  However, the compiler
1876            does not print its usual warning in this case (since polyequal
1877            is the right thing to do there).
1878    
1879            * Improved the organization of structure InlineT.
1880    
1881            * A word about conversion primops:
1882            If conversions involving intinf do not cancel out during
1883            CPS contract, then the compiler must insert calls to Core functions.
1884            Since all core access must be resolved already during the FLINT
1885            translate phase, it would be too late a the time of CPS contract
1886            to add new Core calls.  For this reason, conversion primops
1887            for intinf carry two arguments:  1. the numeric argument that
1888            they are supposed to convert, and 2. the Core function that
1889            can help with this conversion if necessary.  If CPS contract
1890            eliminates a primop, then the associated Core function becomes
1891            dead and goes away.  Intinf conversion primops that do not get
1892            eliminated by CPS contract get rewritten into calls of their
1893            core functions by a separate, new phase.
1894    
1895    interactive system:
1896            - Control.Print.intinfDepth controls max length of intinf constants
1897            being printed. (Analogous to Control.Print.stringDepth.)
1898            - Cleanup in printutil and pputil: got rid of unused stuff and
1899            duplicates; replaced some of the code with code that makes better
1900            use of library functionality.
1901    
1902    CM:
1903            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1904            are no longer silent.
1905    
1906    CKIT:
1907            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1908            in favor of LargeInt -- which is now the same as IntInf.
1909            CKIT-knowledgable people should check whether this is what's
1910            intended and otherwise change things back to using Int32 or
1911            FixedInt.
1912    
1913    Throughout the code:
1914            Started using IntInf.int literals and built-in operations
1915            (e.g., comparison with 0) where this seems appropriate.
1916    
1917    
1918    ----------------------------------------------------------------------
1919    Name: Dave MacQueen (dbm@cs.uchicago.edu)
1920    Date: 2003/08/13 11:36:00 CDT
1921    Tag: dbm-20030813-mcz-merge1
1922    Description:
1923    
1924    Merging changes from the mcz-branch development branch into trunk.
1925    These changes involve replacement of the emulated old prettyprinter
1926    interface with direct use of the SML/NJ Lib PP library, and fixing
1927    of a couple of bugs (895, 1186) relating to error messages.  A new
1928    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1929    has been added.
1930    
1931    ----------------------------------------------------------------------
1932    Name: Matthias Blume (blume (at) tti - c (dot) org)
1933    Date: 2003/08/11 15:45:00 CDT
1934    Tag: blume-20030811-windows
1935    Description:
1936    
1937    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
1938    
1939        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
1940    
1941    This patch restores SML/NJ's ability to run under win32.  There are a
1942    number of changes, including fixes for several bugs that had gone
1943    unnoticed until now:
1944    
1945      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
1946        trying to run the win32 version from within cygwin.)
1947    
1948      - There are a number of simple .bat scripts that substitute for their
1949        corresponding Unix shell-scripts.  (See below.)
1950    
1951      - The internals of ml-build have been modified slightly.  The main
1952        difference is that instead of calling ".link-sml" (or link-sml.bat)
1953        using OS.Process.system, the ML process delegates this task back
1954        to the script.  Otherwise problems arise in mixed environments such
1955        as Cygwin where scripts look and work like Unix scripts, but
1956        where OS.Process.system cannot run them.
1957    
1958      - In CM, the srcpath pickler used native pathname syntax -- which
1959        is incorrect in the case of cross-compilation.  The new pickle format
1960        is independent of platform-specific naming conventions.
1961    
1962      - Path configuration files (such as lib/pathconfig) can now choose
1963        between native and standard syntax.  Placing a line of the form
1964    
1965            standard!
1966    
1967        into the file causes all subsequent paths to be interpreted using
1968        CM standard pathname syntax (= Unix conventions); a line
1969    
1970            native!
1971    
1972        switches back to native style.  This was needed so that
1973        path config files can be written portably, see src/system/pathconfig.
1974    
1975      - Runtime system:
1976    
1977          - win32-filesys.c:  get_file_time and set_file_time now
1978            access modification time, not creation time.
1979    
1980          - I/O code made aware of new array representation.
1981    
1982          - Bug fixes in X86.prim.masm.
1983    
1984       - src/system/makeml made aware of win32. (For use under cygwin
1985            and other Unix-environments for windows.)
1986    
1987       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
1988         which caused BinIO.inputAll to fail consistently.
1989    
1990    .bat scripts:
1991    
1992       Windows .bat scripts assume that SMLNJ_HOME is defined.
1993    
1994       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
1995         applications (sml, ml-yacc, ml-lex).
1996       - ml-build.bat: analogous to ml-build.
1997       - config\install.bat:  Analogous to config/install.sh.  This requires
1998         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
1999         (nmake etc. must be on the path, and vcvars32 must have been run.)
2000         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2001         and the bootfile hierarchy must have been unpacked under
2002         sml.boot.x86-win32.
2003         The script is very primitive and does a poor job at error checking.
2004         It only installs the base system, ml-lex, and ml-yacc.  No other
2005         libraries are being installed (i.e., you get only those that
2006         are part of the compiler.)
2007       - link-sml.bat:  analogous to .link-sml, but not currently used
2008    
2009    Unrelated bug fixes:
2010    
2011      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2012        types.
2013      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2014        otherwise.)
2015    
2016    ----------------------------------------------------------------------
2017    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2018    Date: 2003/06/17
2019    Tag: macqueen-20030617-bug895
2020    Description:
2021    
2022    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2023    Tag will be used for new development branch (mcz-branch) for
2024    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2025    
2026    ----------------------------------------------------------------------
2027    Name: Matthias Blume (blume (at) tti - c (dot) org)
2028    Date: 2003/05/27 16:55:00 CDT
2029    Tag: blume-20030527-polyeq
2030    Description:
2031    
2032    Tried to eliminated most cases of polymorphic equality.
2033    
2034    ----------------------------------------------------------------------
2035    Name: Matthias Blume (blume (at) tti - c (dot) org)
2036    Date: 2003/05/21 17:45:00 CDT
2037    Tag: blume-20030517-complete
2038    Description:
2039    
2040    Two changes:
2041    
2042      1. Added a flag for controlling whether non-exhaustive bindings will
2043         be treated as errors (default is false).
2044      2. Cleaned up the *entire* source tree so that CMB.make goes through
2045         without a single non-exhaustive match- or bind warning.
2046    
2047    ----------------------------------------------------------------------
2048    Name: Matthias Blume (blume (at) tti - c (dot) org)
2049    Date: 2003/05/17 10:20:00 CDT
2050    Tag: blume-20030517-absyn
2051    Description:
2052    
2053    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2054    
2055       This mainly affects the quality of error messages.  However, some
2056       of the code is now more straightforward than before. (Treatment of
2057       the above four constructs in translate.sml is much simpler than
2058       the "macro-expansion" that was going on before.  Plus, the mach-
2059       compiler no longer gets invoked just to be able to compile an
2060       if-expression.)
2061    
2062    2. The ErrorMsg.Error exception is now caught and absorbed by the
2063       interactive loop.
2064    
2065    ----------------------------------------------------------------------
2066    Name: Allen Leung
2067    Date: 2003/05/16 13:05:00 CDT
2068    Tag: leunga-20030516-cygwin-runtime
2069    Description:
2070    
2071        Ported the runtime system to cygwin, which uses the unix
2072    x86-unix bin files.  Missing/buggy features:
2073    
2074       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2075         the Cygwin library.
2076       o Ctrl-C handling may be flaky.
2077       o Windows system calls and Windows I/O are not supported.
2078    
2079       A new set of binfiles is located at:
2080    
2081         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2082    
2083       This is only needed for bootstrapping the cygwin version of smlnj.
2084       Other x86 versions can use the existing binfiles.
2085    
2086    ----------------------------------------------------------------------
2087    Name: Matthias Blume
2088    Date: 2003/04/08 15:42:00 CDT
2089    Tag: blume-20030408-listpair
2090    Description:
2091    
2092    1. Added a target 'mlrisc' to installer.
2093    
2094    2. Added missing elements to structure ListPair.
2095    
2096    ----------------------------------------------------------------------
2097    Name: Allen Leung
2098    Date: 2003/01/07 10:40:00 EST
2099    Tag: leunga-20030107-int-rem
2100    Description:
2101    
2102    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2103    The arguments to the SUBL instruction were swapped.
2104    
2105    ----------------------------------------------------------------------
2106    Name: Matthias Blume
2107    Date: 2002/12/12 16:25:00 EST
2108    Tag: blume-20021212-risc-ra
2109    Description:
2110    
2111    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2112    sent the RA into an infinite loop when floating point registers get
2113    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2114    110.37 and 110.38.)
2115    There was another set of potential problems related to the handling of
2116    MLRISC annotations (but those did not yet cause real problems, apparently).
2117    
2118    ----------------------------------------------------------------------
2119    Name: Matthias Blume
2120    Date: 2002/12/06 22:40:00 EST
2121    Tag: blume-20021206-cm-fileid
2122    Description:
2123    
2124    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2125    This fixes the problem of CM getting confused by files that suddenly
2126    change their identity (e.g., by getting unlinked and recreated by some
2127    text editor such as vi).  There might be a better/cheaper/cleaner way
2128    of doing this, but for now this will have to do.
2129    
2130    ----------------------------------------------------------------------
2131    Name: Matthias Blume
2132    Date: 2002/10/28 09:50:00 EST
2133    Tag: blume-20021028-typecheck
2134    Description:
2135    
2136    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2137    
2138    ----------------------------------------------------------------------
2139    Name: Matthias Blume
2140    Date: 2002/10/17 09:10:00 EDT
2141    Tag: Release_110_42
2142    Description:
2143    
2144    In good old tradition, there has been a slight hiccup so that we have
2145    to patch 110.42 after the fact.   The old release tag has been replaced
2146    (see below).
2147    
2148    The change solves a problem with two competing approaches the
2149    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2150    each other's way.
2151    
2152    This change only affects the runtime system code and the installer script.
2153    (No new bootfiles.)
2154    
2155    ----------------------------------------------------------------------
2156    Name: Matthias Blume
2157    Date: 2002/10/16 12:00:00 EDT
2158    Tag: Release_110_42_removed
2159    Description:
2160    
2161    New working release.  New bootfiles.
2162    
2163    ----------------------------------------------------------------------
2164    Name: Matthias Blume
2165    Date: 2002/10/10 13:10:00 EDT
2166    Tag: blume-20021010-ppc-divs
2167    Description:
2168    
2169    The mltree operator DIVS must be implemented with an overflow check on
2170    the PPC because the hardware indicates divide-by-zero using "overflow" as
2171    well.
2172    
2173    ----------------------------------------------------------------------
2174    Name: Matthias Blume
2175    Date: 2002/07/23 11:45:00
2176    Tag: blume-20020723-smlnj-home
2177    Description:
2178    
2179    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2180    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2181    CM_PATHCONFIG is also set), the path configuration file is assumed
2182    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2183    the entire tree to some other place and everything will "just work".
2184    
2185    (Companion commands such as ml-build and ml-makedepend also sense this
2186    variable.)
2187    
2188    ----------------------------------------------------------------------
2189    Name: Matthias Blume
2190    Date: 2002/07/12 21:19:00 EDT
2191    Tag: blume-20020712-liveness
2192    Description:
2193    
2194    Exported two useful "step" functions from liveness module (MLRISC).
2195    
2196    ----------------------------------------------------------------------
2197    Name: Matthias Blume
2198    Date: 2002/07/05 16:00 EDT
2199    Tag: Release_110_41
2200    Description:
2201    
2202    New working release.  New bootfiles.
2203    
2204    ----------------------------------------------------------------------
2205    Name: Matthias Blume
2206    Date: 2002/07/05 10:25:00 EDT
2207    Tag: blume-20020705-btimp
2208    Description:
2209    
2210    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2211    other clients can set up backtracing support.
2212    
2213    ----------------------------------------------------------------------
2214    Name: Matthias Blume
2215    Date: 2002/06/25 17:23:00 EDT
2216    Tag: blume-20020625-fpmax
2217    Description:
2218    
2219    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2220    case.  (The sense of the isNaN test was reversed -- which made min and
2221    max always return their first argument.)
2222    
2223    ----------------------------------------------------------------------
2224    Name: Matthias Blume
2225    Date: 2002/06/11
2226    Tag: blume-20020611-unixpath
2227    Description:
2228    
2229    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2230    
2231    ----------------------------------------------------------------------
2232    Name: Matthias Blume
2233    Date: 2002/06/10 16:35:00 EDT
2234    Tag: blume-20020610-ieeereal
2235    Description:
2236    
2237    I back-ported my implementation of IEEEReal.fromString from the
2238    idlbasis-devel branch so that we can test it.
2239    
2240    Another small change is that ppDec tries to give more information
2241    than just "<sig>" in the case of functors.  However, this code is
2242    broken in some mysterious way if the functor's body's signature
2243    has not been declared by ascription but gets inferred from the
2244    implementation.  This needs fixing...
2245    
2246    ----------------------------------------------------------------------
2247    Name: Matthias Blume
2248    Date: 2002/05/31
2249    Tag: blume-20020531-btrace-mode
2250    Description:
2251    
2252    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2253    the wayside when I switched over to using Controls everywhere.)
2254    
2255    ----------------------------------------------------------------------
2256    Name: Lal George
2257    Date: 2002/05/23 12:21:40 EDT
2258    Tag: george-20020523-visual-labels
2259    Description:
2260    
2261    Labels are now displayed in the graphical output to make
2262    the fall-through and target blocks obvious.
2263    
2264    ----------------------------------------------------------------------
2265    Name: Matthias Blume
2266    Date: 2002/05/22 11:03:00 EDT
2267    Tag: blume-20020522-shrink
2268    Description:
2269    
2270    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2271    situation that comes up when allocating huge arrays.
2272    
2273    ----------------------------------------------------------------------
2274    Name: Matthias Blume
2275    Date: 2002/05/21 16:00:00 EDT
2276    Tag: Release_110_40
2277    Description:
2278    
2279    New working release (110.40).  New bootfiles.
2280    
2281    [Also: John Reppy fixed GC bug 1131.]
2282    
2283    ----------------------------------------------------------------------
2284    Name: Matthias Blume
2285    Date: 2002/05/21 12:35:00 EDT
2286    Tag: blume-20020521-cmdoc
2287    Description:
2288    
2289    CM documentation update.
2290    
2291    ----------------------------------------------------------------------
2292    Name: Matthias Blume
2293    Date: 2002/05/21 10:55:00 EDT
2294    Tag: blume-20020521-misc
2295    Description:
2296    
2297    - John tweaked runtime to be silent on heap export (except when
2298      GC messages are on).
2299    - I added a few more things (cross-compiling versions of CMB) to
2300      config/preloads (as suggestions).
2301    
2302    ----------------------------------------------------------------------
2303    Name: Matthias Blume
2304    Date: 2002/05/20 22:25:00 EDT
2305    Tag: blume-20020520-controls
2306    Description:
2307    
2308    - Added ControlUtil structure to control-lib.cm.
2309    - Use it throughout.
2310    - Used Controls facility to define MLRISC controls (as opposed to
2311      registering MLRISC control ref cells with Controls after the fact)
2312    - Fixed messed-up controls priorities.
2313    
2314    * Removed again all the stuff from config/preloads that one wouldn't
2315      be able to preload at the time the initial heap image is built.
2316      (Many libraries, e.g., CML, do not exist yet at this time.  The
2317       only libraries that can be preloaded via config/preloads are those
2318       that come bundled with the bootfiles.)
2319    
2320    ----------------------------------------------------------------------
2321    Name: Matthias Blume
2322    Date: 2002/05/20 10:59:00 EDT
2323    Tag: blume-20020520-preloads
2324    Description:
2325    
2326    Added a lot of commented-out suggestions for things to be included
2327    in config/preloads.
2328    
2329    ----------------------------------------------------------------------
2330    Name: Allen Leung
2331    Date: 2002/05/18 14:20:00 EDT
2332    Tag: leunga-20020518-mdl
2333    Description:
2334    
2335    o Made the mdl tool stuff compile and run again.
2336    o I've disabled all the stuff that depends on RTL specifications; they
2337      are all badly broken anyway.
2338    
2339    ----------------------------------------------------------------------
2340    Name: Matthias Blume
2341    Date: 2002/05/17 16:49:00 EDT
2342    Tag: blume-20020517-controls
2343    Description:
2344    
2345    0. John Reppy made several modifications to the SML/NJ library.
2346       In particular, there is a shiny new controls-lib.cm.
2347    
2348    1. Pushed new controls interface through compiler so that everything
2349       compiles again.
2350    
2351    2. Added FormatComb and FORMAT_COMB to the CML version of the
2352       SML/NJ library (so that CML compiles again).
2353    
2354    3. Modified init scripts because XXX_DEFAULT environment variables
2355       are no longer with us.  (Boot-time initialization is now done
2356       using the same environment variables that are also used for
2357       startup-time initialization of controls.)
2358    
2359    ----------------------------------------------------------------------
2360    Name: Lal George
2361    Date: 2002/05/15 09:20:10 EDT
2362    Tag: george-20020515-pseudo-op-decls
2363    Description:
2364    
2365            All pseudo-ops emitted before the first segment declaration
2366    such as TEXT, DATA, and BSS directives are assumed to be global
2367    declarations and are emitted first in the assembly file. This is
2368    useful in a number of situations where one has pseudo-ops that are not
2369    specific to any segment, and also works around the constraint that one
2370    cannot have client pseudo-ops in the TEXT segment.
2371    
2372            Because no segment is associated with these declarations it is
2373    an error to allocate any space or objects before the first segment
2374    directive and an exception will be raised. However, we cannot make
2375    this check for client pseudo-ops.
2376    
2377            These top level declarations are a field in the CFG graph_info.
2378    In theory you can continue to add to this field after the CFG has been
2379    built -- provided you know what you are doing;-)
2380    
2381    ----------------------------------------------------------------------
2382    Name: Matthias Blume
2383    Date: 2002/05/13 16:40:00 EDT
2384    Tag: blume-20020513-pp-etc
2385    Description:
2386    
2387    A few minor bugfixes:
2388    
2389      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2390        (Bogus printouts for redefined bindings still occur.  Compiler
2391        bug should no longer occur now.  We need to redo the prettyprinter
2392        from scratch.)
2393    
2394      - CM pathname printer now also adds escape sequences for ( and )
2395    
2396      - commend and docu fixes for ml-nlffi
2397    
2398    ----------------------------------------------------------------------
2399    Name: Matthias Blume
2400    Date: 2002/05/10 16:40:00 EDT
2401    Tag: blume-20020510-erg-textio
2402    Description:
2403    
2404    Applied the following bugfix provided by Emden Gansner:
2405    
2406        Output is corrupted when outputSubstr is used rather than output.
2407    
2408        The problem occurs when a substring
2409    
2410            ss = (s, dataStart, dataLen)
2411    
2412        where dataStart > 0, fills a stream buffer with avail bytes left.
2413        avail bytes of s, starting at index dataStart, are copied into the
2414        buffer, the buffer is flushed, and then the remaining dataLen-avail
2415        bytes of ss are copied into the beginning of the buffer. Instead of
2416        starting this copy at index dataStart+avail in s, the current code
2417        starts the copy at index avail.
2418    
2419        Fix:
2420        In text-io-fn.sml, change line 695 from
2421             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2422        to
2423             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2424    
2425    ----------------------------------------------------------------------
2426    Name: Matthias Blume
2427    Date: 2002/04/12 13:55:00 EDT
2428    Tag: blume-20020412-assyntax
2429    Description:
2430    
2431    1. Grabbed newer assyntax.h from the XFree86 project.
2432    2. Fiddled with how to compile X86.prim.asm without warnings.
2433    3. (Very) Minor cleanup in CM.
2434    
2435    ----------------------------------------------------------------------
2436    Name: Matthias Blume
2437    Date: 2002/04/01 (no joke!) 17:07:00 EST
2438    Tag: blume-20020401-x86div
2439    Description:
2440    
2441    Added full support for div/mod/rem/quot on the x86, using the machine
2442    instruction's two results (without clumsily recomputing the remainder)
2443    directly where appropriate.
2444    
2445    Some more extensive power-of-two support was added to the x86 instruction
2446    selector (avoiding expensive divs, mods, and muls where they can be
2447    replaced with cheaper shifts and masks).  However, this sort of thing
2448    ought to be done earlier, e.g., within the CPS optimizer so that
2449    all architectures benefit from it.
2450    
2451    The compiler compiles to a fixed point, but changes might be somewhat
2452    fragile nevertheless.  Please, report any strange things that you might
2453    see wrt. div/mod/quot/rem...
2454    
2455    ----------------------------------------------------------------------
2456    Name: Matthias Blume
2457    Date: 2002/03/29 17:22:00
2458    Tag: blume-20020329-div
2459    Description:
2460    
2461    Fixed my broken div/mod logic.  Unfortunately, this means that the
2462    inline code for div/mod now has one more comparison than before.
2463    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2464    The problem was with quotient = 0, because that alone does not tell
2465    us which way the rounding went.  One then has to look at whether
2466    remainder and divisor have the same sign...  :(
2467    
2468    Anyway, I replaced the bootfiles with fresh ones...
2469    
2470    ----------------------------------------------------------------------
2471    Name: Matthias Blume
2472    Date: 2002/03/29 14:10:00 EST
2473    Tag: blume-20020329-inlprims
2474    Description:
2475    
2476    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2477    
2478    Primops have changed. This means that the bin/boot-file formats have
2479    changed as well.
2480    
2481    To make sure that there is no confusion, I made a new version.
2482    
2483    
2484    CHANGES:
2485    
2486    * removed REMT from mltree (remainder should never overflow).
2487    
2488    * added primops to deal with divisions of all flavors to the frontend
2489    
2490    * handled these primops all the way through so they map to their respective
2491      MLRISC support
2492    
2493    * used these primops in the implementation of Int, Int32, Word, Word32
2494    
2495    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2496    
2497    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2498    
2499    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2500      floating point
2501    
2502    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2503      and Real (but Real.abs maps to a separate floating-point-only primop)
2504    
2505    
2506    TODO items:
2507    
2508    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2509      instructions because the machine instruction encoder cannot handle
2510      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2511      REM in the first place, and REM is supported by the x86 machine coder.)
2512    
2513    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2514      selection phase.  (The two can be streamlined because the hardware
2515      delivers both quotient and remainder at the same time anyway.)
2516    
2517    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2518      (Currently the behavior is inconsistent both across architectures and
2519      wrt. the draft Basis spec.)
2520    
2521    * Word8 should eventually be handled natively, too.
2522    
2523    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2524      as if there currently is no execution path that could trigger it in
2525      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2526      hold for things like multiplication and division.)
2527    
2528    ----------------------------------------------------------------------
2529    Name: Matthias Blume
2530    Date: 2002/03/27 16:27:00 EST
2531    Tag: blume-20020327-mlrisc-divisions
2532    Description:
2533    
2534    Added support for all four division operations (ML's div, mod, quot,
2535    and rem) to MLRISC.  In the course of doing so, I also rationalized
2536    the naming (no more annoying switch-around of DIV and QUOT), by
2537    parameterizing the operation by div_rounding_mode (which can be either
2538    DIV_TO_ZERO or DIV_TO_NEGINF).
2539    
2540    The generic MLTreeGen functor takes care of compiling all four
2541    operations down to only round-to-zero div.
2542    
2543    Missing pieces:
2544    
2545      * Doing something smarter than relying on MLTreeGen on architectures
2546        like, e.g., the x86 where hardware division delivers both quotient and
2547        remainder at the same time.  With this, the implementation of the
2548        round-to-neginf operations could be further streamlined.
2549    
2550      * Remove inlining support for div/mod/rem from the frontend and replace it
2551        with primops that get carried through to the backend.  Do this for all
2552        int and word types.
2553    
2554    ----------------------------------------------------------------------
2555    Name: Matthias Blume
2556    Date: 2002/03/25 17:25:00 EST
2557    Tag: blume-20020325-divmod
2558    Description:
2559    
2560    I improved (hopefully without breaking them) the implementation of Int.div,
2561    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2562    advantage of the following observations:
2563    
2564      Let  q = x quot y      r = x rem y
2565           d = x div  y      m = x mod y
2566    
2567    where "quot" is the round-to-zero version of integer division that
2568    hardware usually provides.  Then we have:
2569    
2570         r = x - q * y        where neither the * nor the - will overflow
2571         d = if q >= 0 orelse x = q * y then q else q - 1
2572                              where neither the * nor the - will overflow
2573         m = if q >= 0 orelse r = 0 then r else r + y
2574                              where the + will not overflow
2575    
2576    This results in substantial simplification of the generated code.
2577    The following table shows the number of CFG nodes and edges generated
2578    for
2579            fun f (x, y) = x OPER y
2580            (* with OPER \in div, mod, quot, rem *)
2581    
2582    
2583        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2584        --------------------------------------------------------
2585         div |         24 |         39 |         12 |         16
2586         mod |         41 |         71 |         12 |         16
2587        quot |          8 |         10 |          8 |         10
2588         rem |         10 |         14 |          8 |         10
2589    
2590    
2591    ----------------------------------------------------------------------
2592    Name: Matthias Blume
2593    Date: 2002/03/25 22:06:00 EST
2594    Tag: blume-20020325-cprotobug
2595    Description:
2596    
2597    Fixed a bug in cproto (c prototype decoder).
2598    
2599    ----------------------------------------------------------------------
2600    Name: Matthias Blume
2601    Date: 2002/03/25 16:00:00 EST
2602    Tag: blume-20020325-raw-primops
2603    Description:
2604    
2605    I did some cleanup to Allen's new primop code and
2606    replaced yesterday's bootfiles with new ones.
2607    (But they are stored in the same place.)
2608    
2609    ----------------------------------------------------------------------
2610    Name: Matthias Blume
2611    Date: 2002/03/24 22:40:00 EST
2612    Tag: blume-20020324-bootfiles
2613    Description:
2614    
2615    Made the bootfiles that Allen asked for.
2616    
2617    ----------------------------------------------------------------------
2618    Name: Allen Leung
2619    Date: 2002/03/23 15:50:00 EST
2620    Tag: leunga-20020323-flint-cps-rcc-primops
2621    Description:
2622    
2623      1. Changes to FLINT primops:
2624    
2625        (* make a call to a C-function;
2626         * The primop carries C function prototype information and specifies
2627         * which of its (ML-) arguments are floating point. C prototype
2628         * information is for use by the backend, ML information is for
2629         * use by the CPS converter. *)
2630      | RAW_CCALL of { c_proto: CTypes.c_proto,
2631                       ml_args: ccall_type list,
2632                       ml_res_opt: ccall_type option,
2633                       reentrant : bool
2634                     } option
2635       (* Allocate uninitialized storage on the heap.
2636        * The record is meant to hold short-lived C objects, i.e., they
2637        * are not ML pointers.  With the tag, the representation is
2638        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2639        *)
2640      | RAW_RECORD of {tag:bool,sz:int}
2641      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2642    
2643      2.  These CPS primops are now overloaded:
2644    
2645           rawload of {kind:numkind}
2646           rawstore of {kind:numkind}
2647    
2648          The one argument form is:
2649    
2650             rawload {kind} address
2651    
2652          The two argument form is:
2653    
2654             rawload {kind} [ml object, byte-offset]
2655    
2656      3. RAW_CCALL/RCC now takes two extra arguments:
2657    
2658         a. The first is whether the C call is reentrant, i.e., whether
2659            ML state should be saved and restored.
2660         b. The second argument is a string argument specifying the name of
2661            library and the C function.
2662    
2663         These things are currently not handled in the code generator, yet.
2664    
2665      4. In CProto,
2666    
2667         An encoding type of "bool" means "ml object" and is mapped into
2668         C prototype of PTR.  Note that "bool" is different than "string",
2669         even though "string" is also mapped into PTR, because "bool"
2670         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2671    
2672      5. Pickler/unpicker
2673    
2674         Changed to handle RAW_RECORD and newest RAW_CCALL
2675    
2676      6. MLRiscGen,
2677    
2678         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2679         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2680            in the file CodeGen/cpscompile/cps-c-calls.sml
2681    
2682      7. Added the conditional move operator
2683    
2684             condmove of branch
2685    
2686         to cps.  Generation of this is still buggy so it is currently
2687         disabled.
2688    
2689    ----------------------------------------------------------------------
2690    Name: Lal George
2691    Date: 2002/03/22 14:18:25 EST
2692    Tag: george-20020322-cps-branch-prob
2693    Description:
2694    
2695    Implemented the Ball-Larus branch prediction-heuristics, and
2696    incorporated graphical viewers for control flow graphs.
2697    
2698    Ball-Larus Heuristics:
2699    ---------------------
2700    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2701    
2702    By design it uses the Dempster-Shafer theory for combining
2703    probabilities.  For example, in the function:
2704    
2705        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2706    
2707    the ball-larus heuristics predicts that the n=0 is unlikely
2708    (OH-heuristic), and the 'then' branch is unlikely because of the
2709    RH-heuristic -- giving the 'then' branch an even lower combined
2710    probability using the Dempster-Shafer theory.
2711    
2712    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2713    probability of the 'then' branch because of the loop in the else
2714    branch.
2715    
2716    
2717    Graphical Viewing:
2718    ------------------
2719    I merely plugged in Allen's graphical viewers into the compiler. The
2720    additional code is not much. At the top level, saying:
2721    
2722            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2723    
2724    will display the graphical view of the control flow graph just before
2725    back-patching.  daVinci must be in your path for this to work. If
2726    daVinci is not available, then the default viewer can be changed
2727    using:
2728    
2729            Control.MLRISC.getString "viewer"
2730    
2731    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2732    course, these viewers must be in your path.
2733    
2734    The above will display the compilation unit at the level of clusters,
2735    many of which are small, boring, and un-interesting. Also setting:
2736    
2737            Control.MLRISC.getInt "cfg-graphical-view_size"
2738    
2739    will display clusters that are larger than the value set by the above.
2740    
2741    
2742    ----------------------------------------------------------------------
2743    Name: Matthias Blume
2744    Date: 2002/03/21 22:20:00 EST
2745    Tag: blume-20020321-kmp-bugfix
2746    Description:
2747    
2748    Changed the interface to the KMP routine in PreString and fixed
2749    a minor bug in one place where it was used.
2750    
2751    ----------------------------------------------------------------------
2752    Name: Allen Leung
2753    Date: 2002/03/21 20:30:00 EST
2754    Tag: leunga-20020321-cfg
2755    Description:
2756    
2757      Fixed a potential problem in cfg edge splitting.
2758    
2759    ----------------------------------------------------------------------
2760    Name: Allen Leung
2761    Date: 2002/03/21 17:15:00 EST
2762    Tag: leunga-20020321-x86-fp-cfg
2763    Description:
2764    
2765      1. Recoded the buggy parts of x86-fp.
2766    
2767         a. All the block reordering code has been removed.
2768            We now depend on the block placement phases to do this work.
2769    
2770         b. Critical edge splitting code has been simplified and moved into the
2771            CFG modules, as where they belong.
2772    
2773         Both of these were quite buggy and complex.  The code is now much, much
2774         simpler.
2775    
2776      2. X86 backend.
2777    
2778         a. Added instructions for 64-bit support.  Instruction selection for
2779            64-bit has not been committed, however, since that
2780            requires changes to MLTREE which haven't been approved by
2781            Lal and John.
2782    
2783         b. Added support for FUCOMI and FUCOMIP when generating code for
2784            PentiumPro and above.  We only generate these instructions in
2785            the fast-fp mode.
2786    
2787         c. Added cases for JP and JNP in X86FreqProps.
2788    
2789      3. CFG
2790    
2791         CFG now has a bunch of methods for edge splitting and merging.
2792    
2793      4. Machine description.
2794    
2795         John's simplification of MLTREE_BASIS.fcond broke a few machine
2796         description things:
2797    
2798         rtl-build.{sig,sml} and hppa.mdl fixed.
2799    
2800         NOTE: the machine description stuff in the repository is still broken.
2801               Again, I can't put my fixes in because that involves
2802               changes to MLTREE.
2803    
2804    ----------------------------------------------------------------------
2805    Name: Matthias Blume
2806    Date: 2002/03/20 15:55:00 EST
2807    Tag: blume-20020320-kmp
2808    Description:
2809    
2810    Implemented Knuth-Morris-Pratt string matching in PreString and used
2811    it for String.isSubstring, Substring.isSubstring, and
2812    Substring.position.
2813    
2814    (Might need some stress-testing.  Simple examples worked fine.)
2815    
2816    ----------------------------------------------------------------------
2817    Name: Matthias Blume
2818    Date: 2002/03/19 16:37:00 EST
2819    Tag: blume-20020319-witnesses
2820    Description:
2821    
2822    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2823    
2824    This implements a generic mechanism for changing constness qualifiers
2825    anywhere within big C types without resorting to outright "casts".
2826    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2827    modify the constness at the outermost level.)
2828    The implementation of "convert" is based on the idea of "witness"
2829    values -- values that are not used by the operation but whose types
2830    "testify" to their applicability.  On the implementation side, "convert"
2831    is simply a projection (returning its second curried argument).  With
2832    cross-module inlining, it should not result in any machine code being
2833    generated.
2834    
2835    ----------------------------------------------------------------------
2836    Name: Matthias Blume
2837    Date: 2002/03/15 16:40:00 EST
2838    Tag: blume-20020315-basis
2839    Description:
2840    
2841    Provided (preliminary?) implementations for
2842    
2843      {String,Substring}.{concatWith,isSuffix,isSubstring}
2844    
2845    and
2846    
2847      Substring.full
2848    
2849    Those are in the Basis spec but they were missing in SML/NJ.
2850    
2851    ----------------------------------------------------------------------
2852    Name: Matthias Blume
2853    Date: 2002/03/14 21:30:00 EST
2854    Tag: blume-20020314-controls
2855    Description:
2856    
2857    Controls:
2858    ---------
2859    
2860    1. Factored out the recently-added Controls : CONTROLS stuff and put
2861       it into its own library $/controls-lib.cm.  The source tree for
2862       this is under src/smlnj-lib/Controls.
2863    
2864    2. Changed the names of types and functions in this interface, so they
2865       make a bit more "sense":
2866    
2867          module -> registry
2868          'a registry -> 'a group
2869    
2870    3. The interface now deals in ref cells only.  The getter/setter interface
2871       is (mostly) gone.
2872    
2873    4. Added a function that lets one register an already-existing ref cell.
2874    
2875    5. Made the corresponding modifications to the rest of the code so that
2876       everything compiles again.
2877    
2878    6. Changed the implementation of Controls.MLRISC back to something closer
2879       to the original.  In particular, this module (and therefore MLRISC)
2880       does not depend on Controls.  There now is some link-time code in
2881       int-sys.sml that registers the MLRISC controls with the Controls
2882       module.
2883    
2884    CM:
2885    ---
2886    
2887      * One can now specify the lambda-split aggressiveness in init.cmi.
2888    
2889    ----------------------------------------------------------------------
2890    Name: Allen Leung
2891    Date: 2002/03/13 17:30:00 EST
2892    Tag: leunga-20020313-x86-fp-unary
2893    Description:
2894    
2895    Bug fix for:
2896    
2897    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2898    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2899    > - fun f(x,(y,z)) = Real.~ y;
2900    > [autoloading]
2901    > [autoloading done]
2902    >       fchsl   (%eax), 184(%esp)
2903    > Error: MLRisc bug: X86MCEmitter.emitInstr
2904    >
2905    > uncaught exception Error
2906    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2907    
2908    The problem was that the code generator did not generate any fp registers
2909    in this case, and the ra didn't know that it needed to run the X86FP phase to
2910    translate the pseudo fp instruction.   This only happened with unary fp
2911    operators in certain situations.
2912    
2913    ----------------------------------------------------------------------
2914    Name: Matthias Blume
2915    Date: 2002/03/13 14:00:00 EST
2916    Tag: blume-20020313-overload-etc
2917    Description:
2918    
2919    1. Added _overload as a synonym for overload for backward compatibility.
2920       (Control.overloadKW must be true for either version to be accepted.)
2921    
2922    2. Fixed bug in install script that caused more things to be installed
2923       than what was requested in config/targets.
2924    
2925    3. Made CM aware of the (_)overload construct so that autoloading
2926       works.
2927    
2928    ----------------------------------------------------------------------
2929    Name: Matthias Blume
2930    Date: 2002/03/12 22:03:00 EST
2931    Tag: blume-20020312-url
2932    Description:
2933    
2934    Forgot to update BOOT and srcarchiveurl.
2935    
2936    ----------------------------------------------------------------------
2937    Name: Matthias Blume
2938    Date: 2002/03/12 17:30:00 EST
2939    Tag: blume-20020312-version110392
2940    Description:
2941    
2942    Yet another version number bump (because of small changes to the
2943    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
2944    
2945    Changes:
2946    
2947      The new pid generation scheme described a few weeks ago was overly
2948      complicated.  I implemented a new mechanism that is simpler and
2949      provides a bit more "stability":  Once CM has seen a compilation
2950      unit, it keeps its identity constant (as long as you do not delete
2951      those crucial CM/GUID/* files).  This means that when you change
2952      an interface, compile, then go back to the old interface, and
2953      compile again, you arrive at the original pid.
2954    
2955      There now also is a mechanism that instructs CM to use the plain
2956      environment hash as a module's pid (effectively making its GUID
2957      the empty string).  For this, "noguid" must be specified as an
2958      option to the .sml file in question within its .cm file.
2959      This is most useful for code that is being generated by tools such
2960      as ml-nlffigen (because during development programmers tend to
2961      erase the tool's entire output directory tree including CM's cached
2962      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
2963      revert to the old, broken behavior of SML/NJ, but in specific cases
2964      where there is no danger of interface confusion, its use is ok
2965      (I think).
2966    
2967      ml-nlffigen by default generates "noguid" annotations.  They can be
2968      turned off by specifying -guid in its command line.
2969    
2970    ----------------------------------------------------------------------
2971    Name: Lal George
2972    Date: 2002/03/12 12 14:42:36 EST
2973    Tag: george-20020312-frequency-computation
2974    Description:
2975    
2976    Integrated jump chaining and static block frequency into the
2977    compiler. More details and numbers later.
2978    
2979    ----------------------------------------------------------------------
2980    Name: Lal George
2981    Date: 2002/03/11 11 22:38:53 EST
2982    Tag: george-20020311-jump-chain-elim
2983    Description:
2984    
2985    Tested the jump chain elimination on all architectures (except the
2986    hppa).  This is on by default right now and is profitable for the
2987    alpha and x86, however, it may not be profitable for the sparc and ppc
2988    when compiling the compiler.
2989    
2990    The gc test will typically jump to a label at the end of the cluster,
2991    where there is another jump to an external cluster containing the actual
2992    code to invoke gc. This is to allow factoring of common gc invocation
2993    sequences. That is to say, we generate:
2994    
2995            f:
2996               testgc
2997               ja   L1      % jump if above to L1
2998    
2999            L1:
3000               jmp L2
3001    
3002    
3003    After jump chain elimination the 'ja L1' instructions is converted to
3004    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3005    up being implemented in their long form (if L2 is far away) using:
3006    
3007            jbe     L3      % jump if below or equal to L3
3008            jmp     L2
3009         L3:
3010            ...
3011    
3012    
3013    For large compilation units L2  may be far away.
3014    
3015    
3016    ----------------------------------------------------------------------
3017    Name: Matthias Blume
3018    Date: 2002/03/11 13:30:00 EST
3019    Tag: blume-20020311-mltreeeval
3020    Description:
3021    
3022    A functor parameter was missing.
3023    
3024    ----------------------------------------------------------------------
3025    Name: Allen Leung
3026    Date: 2002/03/11 10:30:00 EST
3027    Tag: leunga-20020311-runtime-string0
3028    Description:
3029    
3030       The representation of the empty string now points to a
3031    legal null terminated C string instead of unit.  It is now possible
3032    to convert an ML string into C string with InlineT.CharVector.getData.
3033    This compiles into one single machine instruction.
3034    
3035    ----------------------------------------------------------------------
3036    Name: Allen Leung
3037    Date: 2002/03/10 23:55:00 EST
3038    Tag: leunga-20020310-x86-call
3039    Description:
3040    
3041       Added machine generation for CALL instruction (relative displacement mode)
3042    
3043    ----------------------------------------------------------------------
3044    Name: Matthias Blume
3045    Date: 2002/03/08 16:05:00
3046    Tag: blume-20020308-entrypoints
3047    Description:
3048    
3049    Version number bumped to 110.39.1.  NEW BOOTFILES!
3050    
3051    Entrypoints: non-zero offset into a code object where execution should begin.
3052    
3053    - Added the notion of an entrypoint to CodeObj.
3054    - Added reading/writing of entrypoint info to Binfile.
3055    - Made runtime system bootloader aware of entrypoints.
3056    - Use the address of the label of the first function given to mlriscGen
3057      as the entrypoint.  This address is currently always 0, but it will
3058      not be 0 once we turn on block placement.
3059    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3060      with entry points) from mlriscGen.
3061    
3062    ----------------------------------------------------------------------
3063    Name: Allen Leung
3064    Date: 2002/03/07 20:45:00 EST
3065    Tag: leunga-20020307-x86-cmov
3066    Description:
3067    
3068       Bug fixes for CMOVcc on x86.
3069    
3070       1. Added machine code generation for CMOVcc
3071       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3072       3. CMOVcc cannot have an immediate operand as argument.
3073    
3074    ----------------------------------------------------------------------
3075    Name: Matthias Blume
3076    Date: 2002/03/07 16:15:00 EST
3077    Tag: blume-20020307-controls
3078    Description:
3079    
3080    This is a very large but mostly boring patch which makes (almost)
3081    every tuneable compiler knob (i.e., pretty much everything under
3082    Control.* plus a few other things) configurable via both the command
3083    line and environment variables in the style CM did its configuration
3084    until now.
3085    
3086    Try starting sml with '-h' (or, if you are brave, '-H')
3087    
3088    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3089    implements the underlying generic mechanism.
3090    
3091    The interface to some of the existing such facilities has changed somewhat.
3092    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3093    (The getFoo interface is still there for backward-compatibility, but its
3094    use is deprecated.)
3095    
3096    The ml-build script passes -Cxxx=yyy command-line arguments through so
3097    that one can now twiddle the compiler settings when using this "batch"
3098    compiler.
3099    
3100    TODO items:
3101    
3102    We should go through and throw out all controls that are no longer
3103    connected to anything.  Moreover, we should go through and provide
3104    meaningful (and correct!) documentation strings for those controls
3105    that still are connected.
3106    
3107    Currently, multiple calls to Controls.new are accepted (only the first
3108    has any effect).  Eventually we should make sure that every control
3109    is being made (via Controls.new) exactly once.  Future access can then
3110    be done using Controls.acc.
3111    
3112    Finally, it would probably be a good idea to use the getter-setter
3113    interface to controls rather than ref cells.  For the time being, both
3114    styles are provided by the Controls module, but getter-setter pairs are
3115    better if thread-safety is of any concern because they can be wrapped.
3116    
3117    *****************************************
3118    
3119    One bug fix: The function blockPlacement in three of the MLRISC
3120    backpatch files used to be hard-wired to one of two possibilities at
3121    link time (according to the value of the placementFlag).  But (I
3122    think) it should rather sense the flag every time.
3123    
3124    *****************************************
3125    
3126    Other assorted changes (by other people who did not supply a HISTORY entry):
3127    
3128    1. the cross-module inliner now works much better (Monnier)
3129    2. representation of weights, frequencies, and probabilities in MLRISC
3130       changed in preparation of using those for weighted block placement
3131       (Reppy, George)
3132    
3133    ----------------------------------------------------------------------
3134    Name: Lal George
3135    Date: 2002/03/07 14:44:24 EST 2002
3136    Tag: george-20020307-weighted-block-placement
3137    
3138    Tested the weighted block placement optimization on all architectures
3139    (except the hppa) using AMPL to generate the block and edge frequencies.
3140    Changes were required in the machine properties to correctly
3141    categorize trap instructions. There is an MLRISC flag
3142    "weighted-block-placement" that can be used to enable weighted block
3143    placement, but this will be ineffective without block/edge
3144    frequencies (coming soon).
3145    
3146    
3147    ----------------------------------------------------------------------
3148    Name: Lal George
3149    Date: 2002/03/05 17:24:48 EST
3150    Tag: george-20020305-linkage-cluster
3151    
3152    In order to support the block placement optimization, a new cluster
3153    is generated as the very first cluster (called the linkage cluster).
3154    It contains a single jump to the 'real' entry point for the compilation
3155    unit. Block placement has no effect on the linkage cluster itself, but
3156    all the other clusters  have full freedom in the manner in which they
3157    reorder blocks or functions.
3158    
3159    On the x86 the typical linkage code that is generated is:
3160       ----------------------
3161            .align 2
3162       L0:
3163            addl    $L1-L0, 72(%esp)
3164            jmp     L1
3165    
3166    
3167            .align  2
3168       L1:
3169       ----------------------
3170    
3171    72(%esp) is the memory location for the stdlink register. This
3172    must contain the address of the CPS function being called. In the
3173    above example, it contains the address of  L0; before
3174    calling L1 (the real entry point for the compilation unit), it
3175    must contain the address for L1, and hence
3176    
3177            addl $L1-L0, 72(%esp)
3178    
3179    I have tested this on all architectures except the hppa.The increase
3180    in code size is of course negligible
3181    
3182    ----------------------------------------------------------------------
3183    Name: Allen Leung
3184    Date: 2002/03/03 13:20:00 EST
3185    Tag: leunga-20020303-mlrisc-tools
3186    
3187      Added #[ ... ] expressions to mlrisc tools
3188    
3189    ----------------------------------------------------------------------
3190    Name: Matthias Blume
3191    Date: 2002/02/27 12:29:00 EST
3192    Tag: blume-20020227-cdebug
3193    Description:
3194    
3195    - made types in structure C and C_Debug to be equal
3196    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3197    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3198    
3199    ----------------------------------------------------------------------
3200    Name: Matthias Blume
3201    Date: 2002/02/26 12:00:00 EST
3202    Tag: blume-20020226-ffi
3203    Description:
3204    
3205    1. Fixed a minor bug in CM's "noweb" tool:
3206       If numbering is turned off, then truly don't number (i.e., do not
3207       supply the -L option to noweb).  The previous behavior was to supply
3208       -L'' -- which caused noweb to use the "default" line numbering scheme.
3209       Thanks to Chris Richards for pointing this out (and supplying the fix).
3210    
3211    2. Once again, I reworked some aspects of the FFI:
3212    
3213       A. The incomplete/complete type business:
3214    
3215       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3216         gone!
3217       - ML types representing an incomplete type are now *equal* to
3218         ML types representing their corresponding complete types (just like
3219         in C).  This is still safe because ml-nlffigen will not generate
3220         RTTI for incomplete types, nor will it generate functions that
3221         require access to such RTTI.   But when ML code generated from both
3222         incomplete and complete versions of the C type meet, the ML types
3223         are trivially interoperable.
3224    
3225         NOTE:  These changes restore the full generality of the translation
3226         (which was previously lost when I eliminated functorization)!
3227    
3228       B. Enum types:
3229    
3230       - Structure C now has a type constructor "enum" that is similar to
3231         how the "su" constructor works.  However, "enum" is not a phantom
3232         type because each "T enum" has values (and is isomorphic to
3233         MLRep.Signed.int).
3234       - There are generic access operations for enum objects (using
3235         MLRep.Signed.int).
3236       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3237         * The structure contains the definition of type "mlrep" (the ML-side
3238         representation type of the enum).  Normally, mlrep is the same
3239         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3240         then mlrep will be defined as a datatype -- thus facilitating
3241         pattern matching on mlrep values.
3242         ("-ec" will be suppressed if there are duplicate values in an
3243          enumeration.)
3244         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3245         will be generated for each C enum constant xxx.
3246         * Conversion functions m2i and i2m convert between mlrep and
3247         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3248         * Coversion functions c and ml convert between mlrep and "tag enum".
3249         * Access functions (get/set) fetch and store mlrep values.
3250       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3251         enumerations are merged into one single enumeration represented by
3252         structure E_'.
3253    
3254    ----------------------------------------------------------------------
3255    Name: Allen Leung
3256    Date: 2002/02/25 04:45:00 EST
3257    Tag: leunga-20020225-cps-spill
3258    
3259    This is a new implementation of the CPS spill phase.
3260    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3261    In case of problems, replace it with the old file spill.sml
3262    
3263    The current compiler runs into some serious performance problems when
3264    constructing a large record.  This can happen when we try to compile a
3265    structure with many items.  Even a very simple structure like the following
3266    makes the compiler slow down.
3267    
3268        structure Foo = struct
3269           val x_1 = 0w1 : Word32.int
3270           val x_2 = 0w2 : Word32.int
3271           val x_3 = 0w3 : Word32.int
3272           ...
3273           val x_N = 0wN : Word32.int
3274        end
3275    
3276    The following table shows the compile time, from N=1000 to N=4000,
3277    with the old compiler:
3278    
3279    N
3280    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3281           MLRISC ra                               0.06u  0.00s  0.05g
3282              (spills = 0 reloads = 0)
3283           TOTAL                                   0.63u  0.07s  0.21g
3284    
3285    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3286           MLRISC ra                               5.68u  0.59s  3.93g
3287              (spills = 0 reloads = 0)
3288           TOTAL                                   14.71u  0.99s  4.81g
3289    
3290    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3291           MLRISC ra                               5.54u  0.65s  3.91g
3292              (spills = 543 reloads = 1082)
3293           TOTAL                                   65.40u  3.13s  6.00g
3294    
3295    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3296           MLRISC ra                               0.80u  0.10s  0.55g
3297              (spills = 42 reloads = 84)
3298           TOTAL                                   129.42u  5.10s  4.13g
3299    
3300    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3301           MLRISC ra                               2.69u  0.27s  1.38g
3302              (spills = 62 reloads = 124)
3303           TOTAL                                   682.48u  19.61s  13.99g
3304    
3305    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3306           MLRISC ra                               4.96u  0.27s  2.72g
3307              (spills = 85 reloads = 170)
3308           TOTAL                                   2375.26u  57.21s  48.00g
3309    
3310    As you can see the old cps spill module suffers from some serious
3311    performance problem.  But since I cannot decipher the old code fully,
3312    instead of patching the problems up, I'm reimplementing it
3313    with a different algorithm.  The new code is more modular,
3314    smaller when compiled, and substantially faster
3315    (O(n log n) time and O(n) space).  Timing of the new spill module:
3316    
3317    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3318          MLRISC ra                               0.25u  0.02s  0.15g
3319             (spills=1 reloads=3)
3320          TOTAL                                   7.74u  0.34s  1.62g
3321    
3322    Implementation details:
3323    
3324    As far as I can tell, the purpose of the CPS spill module is to make sure the
3325    number of live variables at any program point (the bandwidth)
3326    does not exceed a certain limit, which is determined by the
3327    size of the spill area.
3328    
3329    When the bandwidth is too large, we decrease the register pressure by
3330    packing live variables into spill records.  How we achieve this is
3331    completely different than what we did in the old code.
3332    
3333    First, there is something about the MLRiscGen code generator
3334    that we should be aware of:
3335    
3336    o MLRiscGen performs code motion!
3337    
3338       In particular, it will move floating point computations and
3339       address computations involving only the heap pointer to
3340       their use sites (if there is only a single use).
3341       What this means is that if we have a CPS record construction
3342       statement
3343    
3344           RECORD(k,vl,w,e)
3345    
3346       we should never count the new record address w as live if w
3347       has only one use (which is often the case).
3348    
3349       We should do something similar to floating point, but the transformation
3350       there is much more complex, so I won't deal with that.
3351    
3352    Secondly, there are now two new cps primops at our disposal:
3353    
3354     1. rawrecord of record_kind option
3355        This pure operator allocates some uninitialized storage from the heap.
3356        There are two forms:
3357    
3358         rawrecord NONE [INT n]  allocates a tagless record of length n
3359         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3360                                     and initializes the tag.
3361    
3362     2. rawupdate of cty
3363          rawupdate cty (v,i,x)
3364          Assigns to x to the ith component of record v.
3365          The storelist is not updated.
3366    
3367    We use these new primops for both spilling and increment record construction.
3368    
3369     1. Spilling.
3370    
3371        This is implemented with a linear scan algorithm (but generalized
3372        to trees).  The algorithm will create a single spill record at the
3373        beginning of the cps function and use rawupdate to spill to it,
3374        and SELECT or SELp to reload from it.  So both spills and reloads
3375        are fine-grain operations.  In contrast, in the old algorithm
3376        "spills" have to be bundled together in records.
3377    
3378        Ideally, we should sink the spill record construction to where
3379        it is needed.  We can even split the spill record into multiple ones
3380        at the places where they are needed.  But CPS is not a good
3381        representation for global code motion, so I'll keep it simple and
3382        am not attempting this.
3383    
3384     2. Incremental record construction (aka record splitting).
3385    
3386        Long records with many component values which are simulatenously live
3387        (recall that single use record addresses are not considered to
3388         be live) are constructed with rawrecord and rawupdate.
3389        We allocate space on the heap with rawrecord first, then gradually
3390        fill it in with rawupdate.  This is the technique suggested to me
3391        by Matthias.
3392    
3393        Some restrictions on when this is applicable:
3394        1. It is not a VECTOR record.  The code generator currently does not handle
3395           this case. VECTOR record uses double indirection like arrays.
3396        2. All the record component values are defined in the same "basic block"
3397           as the record constructor.  This is to prevent speculative
3398           record construction.
3399    
3400    ----------------------------------------------------------------------
3401    Name: Allen Leung
3402    Date: 2002/02/22 01:02:00 EST
3403    Tag: leunga-20020222-mlrisc-tools
3404    
3405    Minor bug fixes in the parser and rewriter
3406    
3407    ----------------------------------------------------------------------
3408    Name: Allen Leung
3409    Date: 2002/02/21 20:20:00 EST
3410    Tag: leunga-20020221-peephole
3411    
3412    Regenerated the peephole files.  Some contained typos in the specification
3413    and some didn't compile because of pretty printing bugs in the old version
3414    of 'nowhere'.
3415    
3416    ----------------------------------------------------------------------
3417    Name: Allen Leung
3418    Date: 2002/02/19 20:20:00 EST
3419    Tag: leunga-20020219-mlrisc-tools
3420    Description:
3421    
3422       Minor bug fixes to the mlrisc-tools library:
3423    
3424       1.  Fixed up parsing colon suffixed keywords
3425       2.  Added the ability to shut the error messages up
3426       3.  Reimplemented the pretty printer and fixed up/improved
3427           the pretty printing of handle and -> types.
3428       4.  Fixed up generation of literal symbols in the nowhere tool.
3429       5.  Added some SML keywords to to sml.sty
3430    
3431    ----------------------------------------------------------------------
3432    Name: Matthias Blume
3433    Date: 2002/02/19 16:20:00 EST
3434    Tag: blume-20020219-cmffi
3435    Description:
3436    
3437    A wild mix of changes, some minor, some major:
3438    
3439    * All C FFI-related libraries are now anchored under $c:
3440        $/c.cm      --> $c/c.cm
3441        $/c-int.cm  --> $c/internals/c-int.cm
3442        $/memory.cm --> $c/memory/memory.cm
3443    
3444    * "make" tool (in CM) now treats its argument pathname slightly
3445      differently:
3446        1. If the native expansion is an absolute name, then before invoking
3447           the "make" command on it, CM will apply OS.Path.mkRelative
3448           (with relativeTo = OS.FileSys.getDir()) to it.
3449        2. The argument will be passed through to subsequent phases of CM
3450           processing without "going native".  In particular, if the argument
3451           was an anchored path, then "make" will not lose track of that anchor.
3452    
3453    * Compiler backends now "know" their respective C calling conventions
3454      instead of having to be told about it by ml-nlffigen.  This relieves
3455      ml-nlffigen from one of its burdens.
3456    
3457    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3458    
3459    * Export C_DEBUG and C_Debug from $c/c.cm.
3460    
3461    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3462      subtyping relationship between incomplete pointers and their complete
3463      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3464      with the convention of instantiating 'o with ('t, 'c) obj whenever
3465      the pointer target type is complete.  In the incomplete case, 'o
3466      will be instantiated with some "'c iobj" -- a type obtained by
3467      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3468    
3469      Operations that work on both incomplete and complete pointer types are
3470      typed as taking an 'o ptr while operations that require the target to
3471      be known are typed as taking some ('t, 'c) obj ptr.
3472    
3473      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3474      where void is an eqtype without any values.  This makes it possible
3475      to work on voidptr values using functions meant to operate on light
3476      incomplete pointers.
3477    
3478    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3479      been vastly simplified.
3480    
3481    ----------------------------------------------------------------------
3482    Name: Matthias Blume
3483    Date: 2002/02/19 10:48:00 EST
3484    Tag: blume-20020219-pqfix
3485    Description:
3486    
3487    Applied Chris Okasaki's bug fix for priority queues.
3488    
3489    ----------------------------------------------------------------------
3490    Name: Matthias Blume
3491    Date: 2002/02/15 17:05:00
3492    Tag: Release_110_39
3493    Description:
3494    
3495    Last-minute retagging is becoming a tradition... :-(
3496    
3497    This is the working release 110.39.
3498    
3499    ----------------------------------------------------------------------
3500    Name: Matthias Blume
3501    Date: 2002/02/15 16:00:00 EST
3502    Tag: Release_110_39-orig
3503    Description:
3504    
3505    Working release 110.39.  New bootfiles.
3506    
3507    (Update: There was a small bug in the installer so it wouldn't work
3508    with all shells.  So I retagged. -Matthias)
3509    
3510    ----------------------------------------------------------------------
3511    Name: Matthias Blume
3512    Date: 2002/02/15 14:17:00 EST
3513    Tag: blume-20020215-showbindings
3514    Description:
3515    
3516    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3517    the latter can be useful for exploring what bindings are available at
3518    the interactive prompt.  (The first function returns only the list
3519    of symbols that are really bound, the second prints those but also the
3520    ones that CM's autoloading mechanism knows about.)
3521    
3522    ----------------------------------------------------------------------
3523    Name: Matthias Blume
3524    Date: 2002/02/15 12:08:00 EST
3525    Tag: blume-20020215-iptrs
3526    Description:
3527    
3528    Two improvements to ml-nlffigen:
3529    
3530      1. Write files only if they do not exist or if their current contents
3531         do not coincide with what's being written.  (That is, avoid messing
3532         with the time stamps unless absolutely necessary.)
3533    
3534      2. Implement a "repository" mechanism for generated files related
3535         to "incomplete pointer types".   See the README file for details.
3536    
3537    ----------------------------------------------------------------------
3538    Name: Matthias Blume
3539    Date: 2002/02/14 11:50:00 EST
3540    Tag: blume-20020214-quote
3541    Description:
3542    
3543    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3544    because of the new and improved tag generation scheme.  (Thanks to Allen
3545    Leung for pointing it out.)
3546    
3547    ----------------------------------------------------------------------
3548    Name: Lal George
3549    Date: 2002/02/14 09:55:27 EST 2002
3550    Tag: george-20020214-isabelle-bug
3551    Description:
3552    
3553    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3554    of Isabelle on the x86.
3555    
3556    From Allen:
3557    -----------
3558     I've found the problem:
3559    
3560         in ra-core.sml, I use the counter "blocked" to keep track of the
3561         true number of elements in the freeze queue.  When the counter goes
3562         to zero, I skip examining the queue.  But I've messed up the
3563         bookkeeping in combine():
3564    
3565             else ();
3566             case !ucol of
3567               PSEUDO => (if !cntv > 0 then
3568                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3569                                        ^^^^^^^^^^^^^^^^^^^^^^^
3570                      moveu := mergeMoveList(!movev, !moveu)
3571                     )
3572                  else ();
3573    
3574         combine() is called to coalesce two nodes u and v.
3575         I think I was thinking that if the move counts of u and v are both
3576         greater than zero then after they are coalesced then one node is
3577         removed from the freeze queue.  Apparently I was thinking that
3578         both u and v are of low degree, but that's clearly not necessarily true.
3579    
3580    
3581    02/12/2002:
3582        Here's the patch.  HOL now compiles.
3583    
3584        I don't know how this impact on performance (compile
3585        time or runtime).  This bug caused the RA (especially on the x86)
3586        to go thru the potential spill phase when there are still nodes on the
3587        freeze queue.
3588    
3589    
3590    
3591    
3592    ----------------------------------------------------------------------
3593    Name: Matthias Blume
3594    Date: 2002/02/13 22:40:00 EST
3595    Tag: blume-20020213-fptr-rtti
3596    Description:
3597    
3598    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3599    updates.
3600    
3601    ----------------------------------------------------------------------
3602    Name: Matthias Blume
3603    Date: 2002/02/13 16:41:00 EST
3604    Tag: blume-20020213-cmlpq
3605    Description:
3606    
3607    Added new priority queue export symbols (which have just been added to
3608    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3609    not compile and the installer would choke.)
3610    
3611    ----------------------------------------------------------------------
3612    Name: Matthias Blume
3613    Date: 2002/02/13 16:15:00 EST
3614    Tag: blume-20020213-various
3615    Description:
3616    
3617    1. More tweaks to ml-nlffigen:
3618    
3619       - better internal datastructures (resulting in slight speedup)
3620       - "-match" option requires exact match
3621       - "localized" gensym counters (untagged structs/unions nested within
3622         other structs/unions or within typedefs get a fresh counter; their
3623         tag will be prefixed by a concatenation of their parents' tags)
3624       - bug fixes (related to calculation of transitive closure of types
3625         to be included in the output)
3626    
3627    2. Minor Basis updates:
3628    
3629       - added implementations for List.collate and Option.app
3630    
3631    ----------------------------------------------------------------------
3632    Name: Matthias Blume
3633    Date: 2002/02/11 15:55:00 EST
3634    Tag: blume-20020211-gensym
3635    Description:
3636    
3637    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3638    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3639    names (ML structure names that correspond to unnamed C structs, unions,
3640    and enums), so that separate runs of ml-nlffigen do not clash.
3641    
3642    ----------------------------------------------------------------------
3643    Name: Matthias Blume
3644    Date: 2002/02/11 12:05:00 EST
3645    Tag: blume-20020211-gensml
3646    Description:
3647    
3648    A quick fix for a problem with GenSML (in the pgraph-util library):
3649    Make generation of toplevel "local" optional.  (Strictly speaking,
3650    signature definitions within "local" are not legal SML.)
3651    
3652    Other than that: updates to INSTALL and cm/TODO.
3653    
3654    ----------------------------------------------------------------------
3655    Name: Matthias Blume
3656    Date: 2002/02/08 15:00:00 EST
3657    Tag: blume-20020208-uniquepid
3658    Description:
3659    
3660    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3661    
3662    1. The installer (config/install.sh) has gotten smarter:
3663    
3664         - Configuration options are a bit easier to specify now
3665           (in config/targets).
3666         - Bug in recognizing .tar.bz2 files fixed.
3667         - Installer automatically resolves dependencies between
3668           configuration options (e.g., if you ask for eXene, you will
3669           also get cml -- regardless whether you asked for it or not).
3670         - Installer can run in "quieter mode" by setting the environment
3671           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3672           "completely silent", though.
3673         - Build HashCons library as part of smlnj-lib.
3674    
3675    2. A new scheme for assigning persistent identifiers to compilation
3676       units (and, by extension, to types etc.) has been put into place.
3677       This fixes a long-standing bug where types and even dynamic values
3678       can get internally confused, thereby compromising type safety
3679       (abstraction) and dynamic correctness.  See
3680    
3681         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3682    
3683       for an example of how things could go wrong until now.
3684    
3685       The downside of the new scheme is that pids are not quite as
3686       persistent as they used to be: CM will generate a fresh pid
3687       for every compilation unit that it thinks it sees for the first
3688       time.  That means that if you compile starting from a clean, fresh
3689       source tree at two different times, you end up with different
3690       binaries.
3691    
3692       Cutoff recompilation, however, has not been compromised because
3693       CM keeps pid information in special caches between runs.
3694    
3695    ----------------------------------------------------------------------
3696    Name: Lal George
3697    Date: 2002/02/07 15:34:13 EST 2002
3698    Tag: <none>
3699    Description:
3700    
3701    Compilers that generate assembly code may produce  global labels
3702    whose value is resolved at link time. The various peephole optimization
3703    modules did not take this in account.
3704    
3705    TODO. The Labels.addrOf function should really return an option
3706    type so that clients are forced to deal with this issue, rather
3707    than an exception being raised.
3708    
3709    ----------------------------------------------------------------------
3710    Name: Lal George
3711    Date: 2002/02/06 13:55:02 EST
3712    Tag: george-20020206-ra-breakup
3713    Description:
3714    
3715    1. A bug fix from Allen.
3716    
3717        A typo causes extra fstp %st(0)'s to be generated at compensation
3718        edges, which might cause stack underflow traps at runtime.  This
3719        occurs in fft where there are extraneous fstps right before the 'into'
3720        trap instruction (in this case they are harmless since none of the
3721        integers overflow.)
3722    
3723    2. Pulled out various utility modules that were embedded in the modules
3724       of the register allocator. I need these modules for other purposes, but
3725       they are not complete enough to put into a library (just yet).
3726    ----------------------------------------------------------------------
3727    Name: Matthias Blume
3728    Date: 2002/01/31 16:05:00 EST
3729    Tag: blume-20020131-sparc-ccalls
3730    Description:
3731    
3732    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3733       of extra stack space by mistake.  Fixed.
3734    
3735    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3736    
3737    ----------------------------------------------------------------------
3738    Name: Allen Leung
3739    Date: 2002/01/30
3740    Tag: leunga-20020130-nowhere-bug-fix
3741    Description:
3742    
3743       MLRISC bug fixes:
3744       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3745       2. MachineInt.fromString was negating its value.
3746    
3747    ----------------------------------------------------------------------
3748    Name: Matthias Blume
3749    Date: 2002/01/29
3750    Tag: blume-20020129-INSTALL
3751    Description:
3752    
3753    - Added somewhat detailed installation instructions (file INSTALL).
3754    - Fixed curl-detection bug in config/install.sh.
3755    - It is now possible to select the URL getter using the URLGETTER
3756      environment variable:
3757    
3758          not set / "unknown"      --> automatic detection (script tries wget,
3759                                       curl, and lynx)
3760          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3761                                       how to properly invoke them)
3762          other                    --> use $URLGETTER directly, it must take
3763                                       precisely two command-line arguments
3764                                       (source URL and destination file name)
3765    
3766    ----------------------------------------------------------------------
3767    Name: Matthias Blume
3768    Date: 2002/01/28
3769    Tag: blume-20020128-sparc-ccalls
3770    Description:
3771    
3772    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3773    - Make use of the allocParam argument in sparc-c-calls.
3774    
3775    ----------------------------------------------------------------------
3776    Name: Matthias Blume
3777    Date: 2002/01/28
3778    Tag: blume-20020128-allocParam
3779    Description:
3780    
3781    John Reppy:  Changes c-calls API to accept client-callback for
3782    allocating extra stack space.
3783    me: Corresponding changes to mlriscGen (using a dummy argument that
3784        does not change the current behavior).
3785    
3786    ----------------------------------------------------------------------
3787    Name: Matthias Blume
3788    Date: 2002/01/28 12:00:00
3789    Tag: Release_110_38
3790    Description:
3791    
3792    This time for real!!!
3793    
3794    ----------------------------------------------------------------------
3795    Name: Matthias Blume
3796    Date: 2002/01/28 10:56:00 EST
3797    Tag: blume-20020128-retraction
3798    Description:
3799    
3800    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
3801       with blume-Release_110_38-retracted.)
3802    
3803    1. Fixed a problem with incorrect rounding modes in real64.sml.
3804       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
3805    
3806    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
3807       and enums fixed.  The naming of corresponding ML identifiers should
3808       now be consistent again.
3809    
3810    ----------------------------------------------------------------------
3811    Name: Allen Leung
3812    Date: 2002/01/27
3813    Tag: leunga-20020127-nowhere
3814    Description:
3815    
3816       Added a target called nowhere in the configuration scripts.
3817       Enabling this will build the MLRISC 'nowhere' tool (for translating
3818       programs with where-clauses into legal SML code) during installation.
3819    
3820    ----------------------------------------------------------------------
3821    Name: Matthias Blume
3822    Date: 2002/01/25 21:27:00 EST
3823    Tag: blume-Release_110_38-retracted
3824    Description:
3825    
3826    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
3827    
3828    README will be added later.
3829    
3830    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
3831    (Release_110_38) removed.  Reason: Last-minute bug fixes.
3832    
3833    ----------------------------------------------------------------------
3834    Name: Matthias Blume
3835    Date: 2002/01/25
3836    Tag: blume-20020125-ffi
3837    Description:
3838    
3839    A large number of tweaks and improvements to ml-nlffi-lib and
3840    ml-nlffigen:
3841    
3842       - ML represenation types have been streamlined
3843       - getter and setter functions work with concrete values, not abstract
3844         ones where possible
3845       - ml-nlffigen command line more flexible (see README file there)
3846       - some bugs have been fixed (hopefully)
3847    
3848    ----------------------------------------------------------------------
3849    Name: Lal George
3850    Date: 2002/01/24
3851    Tag: george-20020124-risc-ra-interface
3852    Description:
3853    
3854       There is a dramatic simplification in the