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 657, Fri Jun 9 05:20:54 2000 UTC revision 1793, Wed May 18 15:54:17 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 10:55:00 CDT
18    Tag: blume-20050518-realdiv-noovld
19    Description:
20    
21    Un-overloaded / to work around bug in overloading resolution code.
22    
23    ----------------------------------------------------------------------
24    Name: Matthias Blume (blume (at) tti - c (dot) org)
25    Date: 2005/05/16 23:50:00 CDT
26    Tag: blume-20050516-redump-heap
27    Description:
28    
29    Added mechanism for re-creating a heap file for the interactive system
30    after configuration variables have been changed.
31    
32       CM.redump_heap : string -> unit
33    
34    This is much like SMLofNJ.exportML, but starting from the resulting
35    heap does not return to the caller of CM.redump_heap but
36    restarts the interactive system from scratch.  The original call of
37    CM.redump_heap does not return but ends the interactive session.
38    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
39    
40    Internally, redump_heap winds the dynamic execution context back to
41    the point where the original heap image was created and re-executes
42    the heap image generation code in the boot code.
43    
44    ----------------------------------------------------------------------
45    Name: Matthias Blume (blume (at) tti - c (dot) org)
46    Date: 2005/05/09 21:55:00 CDT
47    Tag: blume-20050509-word64
48    Description:
49    
50    Added a hack to the existing hack known as Word64 to make fromString
51    behave correctly.  I am still not sure whether Word64.scan will work
52    as specified with respect to the interaction of radix and prefix.
53    
54    ----------------------------------------------------------------------
55    Name: Allen Leung (leunga (at) reservoir (dot) com)
56    Date: 2005/05/04 11:50:00 EST
57    Tag: leunga-20050504-checkgc
58    Description:
59    
60    Added a gc protocol checking phase.  This phase is enabled with
61    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
62    
63    ----------------------------------------------------------------------
64    Name: Matthias Blume (blume (at) tti - c (dot) org)
65    Date: 2005/05/04 10:45:00 CDT
66    Tag: blume-20050504-intinf
67    Description:
68    
69    Fixed a bug in the implementation of div and mod for IntInf.
70    Thanks to Neophytos Michael for reporting the problem.
71    
72    ----------------------------------------------------------------------
73    Name: Matthias Blume (blume (at) tti - c (dot) org)
74    Date: 2005/05/04 10:35:00 CDT
75    Tag: blume-20050504-join
76    Description:
77    
78    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
79    
80    ----------------------------------------------------------------------
81    Name: Matthias Blume (blume (at) tti - c (dot) org)
82    Date: 2005/02/28 23:40:00 CST
83    Tag: blume-20050228-mVar
84    Description:
85    
86    Fixed serious bug (brown paper bag variety) in new implementation of
87    structure Atom in CML.  (I had accidentally used a mailbox instead of
88    an mvar, leaving the door open for races.)
89    
90    ----------------------------------------------------------------------
91    Name: Matthias Blume (blume (at) tti - c (dot) org)
92    Date: 2005/02/25 15:00:00 CST
93    Tag: Release_110_53
94    Description:
95    
96    New working version (110.53).  NEW BOOTFILES!
97    
98    ----------------------------------------------------------------------
99    Name: Matthias Blume (blume (at) tti - c (dot) org)
100    Date: 2005/02/25 14:50:00 CST
101    Tag: blume-20050225-susp
102    Description:
103    
104    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
105    implemented in Core -- which means that it is the same as the one used
106    by the lazy extension.
107    
108    ----------------------------------------------------------------------
109    Name: Matthias Blume (blume (at) tti - c (dot) org)
110    Date: 2005/02/24 16:50:00 CST
111    Tag: blume-20050224-cml-atom
112    Description:
113    
114    Simpler and at the same time more general implementation of structure
115    Atom in CML.
116    
117    ----------------------------------------------------------------------
118    Name: Matthias Blume (blume (at) tti - c (dot) org)
119    Date: 2005/02/15 17:35:00 CST
120    Tag: blume-20050215-tools
121    Description:
122    
123    Created new "tools" directory under "src" and moved "TraceDebugProf"
124    there.
125    
126    ----------------------------------------------------------------------
127    Name: Matthias Blume (blume (at) tti - c (dot) org)
128    Date: 2005/02/10 17:55:00 CST
129    Tag: blume-20050210-longlong
130    Description:
131    
132    Implemented "long long" arguments and results for NLFFI.  (Only the
133    PPC/MacOS implementation is complete, the other backends still need to
134    be updated.)
135    
136    ----------------------------------------------------------------------
137    Name: Matthias Blume (blume (at) tti - c (dot) org)
138    Date: 2005/01/24 17:40:00 CST
139    Tag: blume-20050124-mlyacc
140    Description:
141    
142    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
143    problem with certain "as" patterns which previously got rendered
144    using incorrect syntax.
145    
146    ----------------------------------------------------------------------
147    Name: Matthias Blume (blume (at) tti - c (dot) org)
148    Date: 2005/01/18 12:00:00 CST
149    Tag: blume-20050118-profile
150    Description:
151    
152    Made time profiling code (interrupt handler) in runtime system aware
153    of new array representation.
154    
155    ----------------------------------------------------------------------
156    Name: Matthias Blume (blume (at) tti - c (dot) org)
157    Date: 2005/01/14 18:00:00 CST
158    Tag: blume-20050114-heap2exec
159    Description:
160    
161    Implemented new (but still experimental) heap2exec facility.  This is
162    tested under Mac OS X and should work under Linux (will test shortly).
163    It will probably also work on the Sparc (will test some time later).
164    
165      - removed old "HACKED_STANDALONE" hack from runtime
166    
167    To be able to test this, uncomment the request for "heap2asm" in
168    config/targets prior to installation.  (Notice that this is different
169    from "heap2exec" mentioned below.  Not a typo.)
170    
171    To perform an actual test, run the command
172    
173       $ bin/heap2exec heapfile execfile
174    
175    (You can put heap2exec on your shell's path.)
176    
177    For example, run
178    
179      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
180    
181    This will create a standalone executable called "mly" which you can
182    then invoke directly as a command.
183    
184    ----------------------------------------------------------------------
185    Name: Matthias Blume (blume (at) tti - c (dot) org)
186    Date: 2005/01/07 11:44:00 CST
187    Tag: blume-20050107-mlstring
188    Description:
189    
190    fixed off-by-one error in ML_STRING macro (globals.c)
191    
192    ----------------------------------------------------------------------
193    Name: Matthias Blume (blume (at) tti - c (dot) org)
194    Date: 2004/12/23 18:00:00 CST
195    Tag: blume-20041223-santa
196    Description:
197    
198    Made ml-build script "smarter" (but only very little).
199    
200    ----------------------------------------------------------------------
201    Name: Matthias Blume (blume (at) tti - c (dot) org)
202    Date: 2004/12/21 15:05:00 CST
203    Tag: blume-20041221-longlong
204    Description:
205    
206    * Implemented access to signed and unsigned long long data in NLFFI.
207    (The parameter-passing part of the picture has not complete. But
208    data structure access seems to work.)
209    
210    * Fixed CM's incorrect assumption that the PPC is little-endian.
211      (On the Mac, it is big-endian.  And that's currently our only
212      PPC platform.)
213    
214    ----------------------------------------------------------------------
215    Name: Matthias Blume (blume (at) tti - c (dot) org)
216    Date: 2004/12/21 12:50:00 CST
217    Tag: blume-20041221-memory
218    Description:
219    
220    Some cleanup in the $c/memory.cm library: separated some concerns by
221    moving allocation code and memory access code each into their own
222    files.
223    
224    ----------------------------------------------------------------------
225    Name: Allen Leung (leunga (at) reservoir (dot) com)
226    Date: 2004/12/17 16:12:00 EST
227    Tag: leunga-20041217-cygwin-smlnj-home
228    Description:
229    
230        The Unix I/O library of SML/NJ on cygwin does not understand
231    Windows style pathname, so problems arise when SMLNJ_HOME is set
232    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
233    to a POSIX pathname on cygwin.
234    
235    ----------------------------------------------------------------------
236    Name: Matthias Blume (blume (at) tti - c (dot) org)
237    Date: 2004/12/16 13:35:00 CST
238    Tag: Release_110_52
239    Description:
240    
241    Last-minute changes incorporated into 110.52.  Release tag moved.
242    
243    The changes:
244       - HashString.hashString' -> HashString.hashSubstring
245       - bug fix in UnivariateStats
246    
247    ----------------------------------------------------------------------
248    Name: Matthias Blume (blume (at) tti - c (dot) org)
249    Date: 2004/12/15 23:40:00 CST
250    Tag: blume-20041215-hashSubstring
251    Description:
252    
253    - HashString.hashString' -> HashString.hashSubstring
254    - corresponding changes in atom.sml
255    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
256      added some comments
257    
258    ----------------------------------------------------------------------
259    Name: Matthias Blume (blume (at) tti - c (dot) org)
260    Date: 2004/12/15 15:30:00 CST
261    Tag: (Release_110_52)
262    Description:
263    
264    New working version (110.52).  NEW BOOTFILES!
265    
266    ----------------------------------------------------------------------
267    Name: Matthias Blume (blume (at) tti - c (dot) org)
268    Date: 2004/12/15 12:45:00 CST
269    Tag: blume-20041215-spaces
270    Description:
271    
272    More on the space problem (this time for Win32).
273    
274    ----------------------------------------------------------------------
275    Name: Matthias Blume (blume (at) tti - c (dot) org)
276    Date: 2004/12/14 17:30:00 CST
277    Tag: blume-20041214-spaces
278    Description:
279    
280    Hacked some of the scripts (in particular: the installer) to cope with
281    spaces in filenames a bit better.  But beware: the current "solution"
282    is likely still full of bugs and inherently incomplete.  (We need to
283    do away with those shell scripts for a comprehensive solution.)
284    
285    ----------------------------------------------------------------------
286    Name: Matthias Blume (blume (at) tti - c (dot) org)
287    Date: 2004/12/13 14:45:00 CST
288    Tag: blume-20041213-ml-makedepend
289    Description:
290    
291    Fixed bug in code for ml-makedepend.
292    
293    ----------------------------------------------------------------------
294    Name: Matthias Blume (blume (at) tti - c (dot) org)
295    Date: 2004/12/09 16:30:00 CST
296    Tag: blume-20041209-statistics
297    Description:
298    
299    Added two simple but potentially useful statistics modules to SML/NJ Library.
300    (See CHANGES file there.)
301    
302    ----------------------------------------------------------------------
303    Name: Matthias Blume (blume (at) tti - c (dot) org)
304    Date: 2004/12/01 16:50:00 CST
305    Tag: blume-20041201-atom
306    Description:
307    
308    smlnj-lib:
309    
310      Added function HashString.hashString' for substrings.
311      Hand-inlined CharVector.foldl into HashString (for speed).
312      Modified implementation of structure Atom to avoid extracting
313      strings from substrings unless necessary.
314      (Also see CHANGES file for smlnj-lib.)
315    
316    ----------------------------------------------------------------------
317    Name: Matthias Blume (blume (at) tti - c (dot) org)
318    Date: 2004/11/24 22:15:00 CST
319    Tag: blume-20041124-cml
320    Description:
321    
322    Made sure CML compiles when Position = Int64.
323    
324    ----------------------------------------------------------------------
325    Name: Matthias Blume (blume (at) tti - c (dot) org)
326    Date: 2004/11/24 14:30:00 CST
327    Tag: blume-20041124-position
328    Description:
329    
330    The compiler can now be compiled in a mode that makes structure Position
331    equal to Int64.  The default, however, is unchanged (Position = Int31)
332    for the time being.
333    
334    To enable 64-bit positions, use the following procedure:
335    
336    1. Start sml
337    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
338    3. Type
339          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
340    4. Run CMB.make() as usual.
341    
342    This is barely tested.  The only test so far was a little SML program
343    counting the number of characters in an 8-gigabyte file by
344    reading it character-by-character.  That test was successful.
345    
346    In support of 64-bit positions, a number of new functions have been
347    added to the runtime system.
348    
349    ----------------------------------------------------------------------
350    Name: Matthias Blume (blume (at) tti - c (dot) org)
351    Date: 2004/11/23 14:45:00 CST
352    Tag: blume-20041123-useFile
353    Description:
354    
355    Fixed a problem with unhelpful error messages related to problems with
356    .cm- or .sml files that appear as part of the sml command line.
357    
358    ----------------------------------------------------------------------
359    Name: Matthias Blume (blume (at) tti - c (dot) org)
360    Date: 2004/11/18 15:40:00 CST
361    Tag: Release_110_51
362    Description:
363    
364    New working version (110.51).  NEW BOOTFILES!
365    
366    ----------------------------------------------------------------------
367    Name: Matthias Blume (blume (at) tti - c (dot) org)
368    Date: 2004/11/18 15:35:00 CST
369    Tag: <none>
370    Description:
371    
372    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
373    Joemann).
374    
375    ----------------------------------------------------------------------
376    Name: Allen Leung (leunga (at) reservoir (dot) com)
377    Date: 2004/11/17 16:05:21 EST 2004
378    Tag: leunga-20041117-mlrisc-live-kill
379    Description:
380    
381       Added support for MLTree constructs LIVE and KILL
382    to all the architectures.
383    
384    ----------------------------------------------------------------------
385    Name: Matthias Blume (blume (at) tti - c (dot) org)
386    Date: 2004/11/13 00:20:00 CST
387    Tag: blume-20041113-versiontool
388    Description:
389    
390    - Stripped down the versiontool: It now only handles the version number.
391      The date string is generated at bootstrap time (during makeml).
392    
393    - In a previous commit, fixed a minor issue with how polyequal is being
394      translated.  In particular, the code now "looks through" abstractions.
395      This results in slightly fewer polyEqual warnings and hopefully slightly
396      more efficient code.  Important examples for where this matters are
397      the new int64 and word64 types.
398    
399    ----------------------------------------------------------------------
400    Name: Matthias Blume (blume (at) tti - c (dot) org)
401    Date: 2004/11/12 00:30:00 CST
402    Tag: blume-20041112-int64
403    Description:
404    
405    Structure Int64 fully hooked in.  (The implementation is not very
406    efficient, though.)
407    
408    ----------------------------------------------------------------------
409    Name: Matthias Blume (blume (at) tti - c (dot) org)
410    Date: 2004/11/11 17:30:00 CST
411    Tag: blume-20041111-more64
412    Description:
413    
414    All the pieces of Word64 are now there, with the exception of the
415    conversions from and to LargeWord. (Eventually these need to be identities,
416    but for the time being they don't even make sense because LargeWord is
417    32-bit wide.)
418    
419    Also started to add similar support for Int64, but major pieces of that
420    are still missing.
421    
422    ----------------------------------------------------------------------
423    Name: Matthias Blume (blume (at) tti - c (dot) org)
424    Date: 2004/11/11 00:15:00 CST
425    Tag: blume-20041111-word64
426    Description:
427    
428    Structure Word64 is now (almost) complete, word literals and patterns
429    seem to work.  There are a few odd pieces missing.  In particular,
430    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
431    at the moment.
432    
433    Making Word64 official would mean that LargeWord becomes Word64.  But
434    this requires extreme care because most word-word conversions have to
435    go through LargeWord, so making a mistake means loss of efficiency or
436    worse.  Eventually there will be a solution similar to (but actually
437    simpler than) what I did with IntInf.
438    
439    ----------------------------------------------------------------------
440    Name: Matthias Blume (blume (at) tti - c (dot) org)
441    Date: 2004/11/10 18:12:00 CST
442    Tag: blume-20041110-64bit
443    Description:
444    
445    More 64-bit hacking (but still not even half-way there yet).
446    Also, some assorted improvements to the handling of 8-bit words.
447    
448    ----------------------------------------------------------------------
449    Name: Matthias Blume (blume (at) tti - c (dot) org)
450    Date: 2004/11/09 17:50:00 CST
451    Tag: <none>
452    Description:
453    
454    Started adding some infrastructure for supporting 64-bit int- and
455    word-types.  (Still in its very early stages.)
456    
457    ----------------------------------------------------------------------
458    Name: Matthias Blume (blume (at) tti - c (dot) org)
459    Date: 2004/10/28 10:45:00 CDT
460    Tag: Release_110_50
461    Description:
462    
463    New working version (110.50).  NEW BOOTFILES!
464    
465                 =====================
466    
467    Also:
468    
469      - Changend config/srcarchiveurl from a file just
470        containing the URL string into a file containing
471        shell script code.  The code has access to the $VERSION variable.
472      - Made corresponding changes to config/install.sh and config/unpack.
473      - Default contents of config/srcarchiveurl uses $VERSION and
474        normally does not have to be edited to reflect a version change.
475    
476      (As a result, a version change can be done by just editing
477       config/version, the rest is now automatic.)
478    
479    ----------------------------------------------------------------------
480    Name: Matthias Blume (blume (at) tti - c (dot) org)
481    Date: 2004/10/27 17:50:00 CDT
482    Tag: blume-20041027-btrace-msg
483    Description:
484    
485    BackTrace.monitor now also reports the source of the exception that
486    triggered the trace.
487    
488    ----------------------------------------------------------------------
489    Name: Matthias Blume (blume (at) tti - c (dot) org)
490    Date: 2004/10/27 17:20:00 CDT
491    Tag: blume-20041027-x86-c-calls
492    Description:
493    
494    This is the HISTORY entry for two earlier commits, both concerning
495    the x86 c-calls code in MLRISC:
496    
497      - added a missing LOAD in the code that deals with struct arguments
498      - made sure the caller does not add the wrong number of bytes to the
499        stack pointer after a call of a function returning a struct
500        (the callee already pops the implicit argument which points to
501         the space reserved for the result)
502    
503    ----------------------------------------------------------------------
504    Name: Allen Leung (leunga (at) reservoir (dot) com)
505    Date: 2004/10/24 14:00:00 EST
506    Tag: leunga-20041024-x86-gas-fucomip
507    Description:
508    
509       John discovered a bug in the syntax of fucomip.
510       The opcodes FU?COMIP? have been changed to
511    
512           fu?comip? %st(i), %st
513    
514    ----------------------------------------------------------------------
515    Name: Matthias Blume (blume (at) tti - c (dot) org)
516    Date: 2004/10/20 15:06:00 CDT
517    Tag: blume-20041020-standalone-backtrace
518    Description:
519    
520    Added a mechanism for getting back-trace information from standalone
521    programs.  Here is how it works:
522    
523       1. The part of the program from which you want to get backtrace
524          information (usually the whole program) should be wrapped with
525          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
526          main program could be modified from something like
527    
528             fun main (pgm, args) = ...
529    
530         to
531    
532            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
533    
534       2. To be able to access BackTrace.monitor, you have to add
535          library $smlnj-tdp/plugins.cm to the .cm file that contains your
536          main function.
537    
538       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
539          CM might have created in the past for your project).
540    
541       4. Build the system using this command line:
542    
543             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
544                      myprog.cm MyProg.main myprog
545    
546          instead of the usual
547    
548             ml-build myprog.cm MyProg.main myprog
549    
550    I changed a library name:
551    
552       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
553    
554    New libraries:
555    
556       $smlnj-tdp/back-trace.cm
557          -- when loaded causes the back-trace plugin to be installed
558       $smlnj-tdp/coverage.cm
559          -- when loaded causes the coverage plugin to be installed
560    
561    ----------------------------------------------------------------------
562    Name: Matthias Blume (blume (at) tti - c (dot) org)
563    Date: 2004/10/18 16:45:00 CDT
564    Tag: blume-20041018-groupowner
565    Description:
566    
567    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
568    
569    Eliminated group owner specs from .cm files throughout the source tree.
570    
571    ----------------------------------------------------------------------
572    Name: Matthias Blume (blume (at) tti - c (dot) org)
573    Date: 2004/10/15 15:45:00 CDT
574    Tag: blume-20041015-coverage
575    Description:
576    
577    * Test coverage tool added!
578    * Further reorganization of tracing-, debugging-, and profiling support:
579    
580        - moved original BTImp -- now called BackTrace -- into a separate
581          library called $/trace-debug-profile.cm
582        - eliminated all mentions of BTrace from SMLofNJ.Internals
583        - only the instrumentation mechanism is now left in the compiler proper
584        - BackTrace module is a plugin which is NOT plugged in by default
585        - Coverage module is another such plugin
586    
587    To get the benefits of any of these plugin modules, the code in
588    question must be compiled with tdp instrumentation turned on.  This
589    can be done by setting SMLofNJ.Internals.TDP.mode to true.
590    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
591    
592    Plugins are selected at link time.  (Pre-compiled instrumented code
593    can be re-loaded with different plugins in effect.)  When an
594    instrumented module is linked, whatever plugins are at that time
595    enabled will come into effect for that module.
596    
597    To enable the back-trace plugin, load library $/trace-debug-profile.cm
598    and invoke BackTrace.install() (e.g., from the interactive prompt).
599    To enable the coverage plugin, load the same library and invoke
600    Coverage.install().
601    
602    Back-traces are generated automatically on uncaught exceptions and
603    when the code in question explicitly invokes BackTrace.trigger().
604    
605    Coverage (and execution frequency-) information must be queried
606    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
607    
608    ----------------------------------------------------------------------
609    Name: Matthias Blume (blume (at) tti - c (dot) org)
610    Date: 2004/10/14 17:40:00 CDT
611    Tag: blume-20041014-tdp-core
612    Description:
613    
614    Snapshot of a significant overhaul of how the trace/debug/profile support
615    is hooked into the system (specifically: Core and SMLofNJ.Internals).
616    
617    ----------------------------------------------------------------------
618    Name: Matthias Blume (blume (at) tti - c (dot) org)
619    Date: 2004/10/13 16:34:00 CDT
620    Tag: blume-20041013-tdp
621    Description:
622    
623    Some rationalization of names:
624    
625       structure BTrace -> structure TDPInstrument
626       etc.
627    
628    This is is preparation of using the original back-trace
629    instrumentation for other purposes.  "TDP" stands for
630    Trace/Debug/Profile.
631    
632    The control flag controlling whether instrumentation is on or off is now
633    registered under a different name, so instead of running sml as
634    
635     sml -Cinstrument.btrace-mode=true
636    
637    one has to say
638    
639     sml -Ctdp.instrument=true
640    
641    ----------------------------------------------------------------------
642    Name: Matthias Blume (blume (at) tti - c (dot) org)
643    Date: 2004/10/11 16:37:00 CDT
644    Tag: blume-20041011-regions
645    Description:
646    
647    Made some minor modifications to elabcore.sml to have source regions
648    be propagated more tightly -- resulting in better (i.e., smaller)
649    regions being reported in error- and debug messages.
650    
651    ----------------------------------------------------------------------
652    Name: Matthias Blume (blume (at) tti - c (dot) org)
653    Date: 2004/10/08 22:50:00 CDT
654    Tag: blume-20041008-cmkw
655    Description:
656    
657    Fixed handling of keywords in .cm files: After seeing "is" the lexer
658    treats subsequent occurrences of "group", "library", "source", "is",
659    "*", and "-" as ordinary identifiers rather than keywords.
660    
661    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
662    is supposed to accept a tool argument called "source", but this did
663    not work because of the clash with the keyword.
664    
665    ----------------------------------------------------------------------
666    Name: Matthias Blume (blume (at) tti - c (dot) org)
667    Date: 2004/10/07 16:00:00 CDT
668    Tag: blume-20041007-cleanup
669    Description:
670    
671    Assorted cleanup work:
672    
673      - got rid of intstrmap in favor of using the library's
674        hash table implementation
675      - threw out most of the pathnames stuff, as it was not used anyway
676      - simplified tokentable implementation
677      - fixed some minor spelling errors
678    
679    ----------------------------------------------------------------------
680    Name: Matthias Blume (blume (at) tti - c (dot) org)
681    Date: 2004/10/06 15:15:15 CDT
682    Tag: blume-20041006-handler
683    Description:
684    
685    Cleaned up the absyn to reflect the invariant that HANDLE always
686    carries a FNexp as part of the type definition.  This eliminates some
687    superfluous sanity checks at runtime down the road.
688    
689    Some minor cleanup of the btrace code.
690    
691    ----------------------------------------------------------------------
692    Name: Matthias Blume (blume (at) tti - c (dot) org)
693    Date: 2004/10/01 10:20:30 CDT
694    Tag: blume-20041001-slave
695    Description:
696    
697    Added hack to make slave mode work in the presence of the version
698    tool.  (Still, since the master does two passes over the code for
699    CMB.make, the release number gets bumped twice when slaves are
700    attached.  I don't know if this is worth fixing...)
701    
702    ----------------------------------------------------------------------
703    Name: Matthias Blume (blume (at) tti - c (dot) org)
704    Date: 2004/09/30 10:55:00 CDT
705    Tag: blume-20040930-version
706    Description:
707    
708    * Moved the "version" magic into its own little library under
709      src/system/smlnj/internal.  This avoids expensive reconstruction of
710      a stable src/compiler/core.cm.
711    
712    * At the same time, structure CompilerVersion is now known as
713      structure SMLNJVersion.
714    
715    * Arranged for the version tool to NOT kick in when rebuilding the system
716      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
717      Also, loading the versiontool does not work when rebuilding the system
718      because CM is not properly initialized at that time.
719    
720    ----------------------------------------------------------------------
721    Name: Matthias Blume (blume (at) tti - c (dot) org)
722    Date: 2004/09/29 14:00:00 CDT
723    Tag: blume-20040929-autoversion
724    Description:
725    
726    Implemented some CM magic to have
727        file src/compiler/TopLevel/main/version.sml
728    generated automagically.
729    The version is taken from two files: config/version and config/release.
730    The first is expected to contain a two-part version number such as 110.49.
731    The second should contain a single number, but it may be missing.
732    
733    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
734    time the version tool is loaded (which is the first time you say CMB.make),
735    then the tool will increment the value stored in config/release every
736    time CMB.make is invoked.
737    
738    The binfile format is now insensitive to anything beyond the first
739    two components of a version number, so bumping the release does not render
740    binfiles incompatible.  Auto-bumping can be used to keep track of versions
741    during development without invalidating existing binfiles.
742    
743    In any case, every CMB.make updates the date information in version.sml.
744    (This is the date that is printed in the banner.)
745    
746    ----------------------------------------------------------------------
747    Name: Matthias Blume (blume (at) tti - c (dot) org)
748    Date: 2004/09/28 10:53:00 CDT
749    Tag: blume-20040928-controls
750    Description:
751    
752    Some cleanup of the controls code.
753    
754    ----------------------------------------------------------------------
755    Name: Matthias Blume (blume (at) tti - c (dot) org)
756    Date: 2004/09/27 22:08:00 CDT
757    Tag: blume-20040927-controls
758    Description:
759    
760    Added two pieces of functionality to the Controls interface:
761    
762    1.   val save'restore: 'a control -> unit -> unit
763    
764       grabs the current value of the control in stage 1 and restores it
765       in stage 2.
766    
767    2.   val set' : 'a control * 'a -> unit -> unit
768    
769       stores the given value into the control in stage 2 (i.e., delayed)
770       but does all error checking in stage 1.
771       (This is for string controls that need to do parse their argument --
772        something that might fail.  In some cases, notably in CM, one
773        already knows the intended argument but wants to delay the actual
774        assignment until a time when error recovery would be more difficult.)
775    
776    Changed the handling of controls in tool arguments to classes "sml" and
777    "lazysml":
778       - use Controls.save'restore as a more robust way of restoring the
779         old value (in particular: without having to re-parse the string)
780       - use controls to handle the "overload" keyword in the init group
781         (I believe this change actually fixes a long-standing obscure bug.)
782    
783    ----------------------------------------------------------------------
784    Name: Matthias Blume (blume (at) tti - c (dot) org)
785    Date: 2004/09/27 17:00:00 CDT
786    Tag: blume-20040927-lazysml
787    Description:
788    
789    Added a new tool class called "lazysml" to CM's tool chest.  The only
790    difference to "sml" is that compilation is done with Control.lazysml
791    set to true.  A source of class "lazysml" is automatically recognized
792    by a file name suffix of ".lml".
793    
794    In addition to the above feature, the original class "sml" now also
795    supports a tool argument "lazy" which has the same effect.  As a
796    result, the following three lines are equivalent:
797    
798        foo.sml : lazysml
799        foo.sml : sml (lazy)
800        foo.sml (lazy)
801    
802    The setting goes into effect both during parsing and during
803    compilation.  The original setting is restored right after parsing and
804    after compilation, respectively.
805    
806    In addition to all the above, there is also a general mechanism to set
807    ANY of the "controls" that are available at the command line via
808    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
809    apply as well. (In fact, "lazy" is implemented as a special case of
810    the general mechanism.)
811    
812    The .cm file syntax uses a new keyword tool argument called "with".
813    There are several ways of indicating the desired settings:
814    
815        foo.sml (with:parser.quotations=true)
816        foo.sml (with:(name:parser.quotations value:true))
817        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
818        foo.sml (with:(name1=value1 name2=value2 ...))
819        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
820    
821        etc.
822    
823    Another possible abbreviation is to leave out the =v or value:v part
824    if the name refers to a boolean control (in which case the value is
825    taken to be true).  Thus, one could get lazy sml also by saying:
826    
827        foo.sml (with:parser.lazy-keyword=true)
828        foo.sml (with:parser.lazy-keyword)
829        foo.sml (with:(name:parser.lazy-keyword value:true))
830        foo.sml (with:(name:parser.lazy-keyword))
831    
832    ----------------------------------------------------------------------
833    Name: Matthias Blume (blume (at) tti - c (dot) org)
834    Date: 2004/09/24 16:22:00 CDT
835    Tag: blume-20040924-ppc-long-branch
836    Description:
837    
838    Turned message about "emiting long form of branch" off by default.
839    Added a control flag to turn it back on when desired.
840    
841    ----------------------------------------------------------------------
842    Name: Matthias Blume (blume (at) tti - c (dot) org)
843    Date: 2004/09/24 16:05:00 CDT
844    Tag: blume-20040924-rounding
845    Description:
846    
847    Applied patch for setting rounding modes under Mac OS X.  Thanks to
848    Melissa O'Neill for providing the code!
849    
850    ----------------------------------------------------------------------
851    Name: Matthias Blume (blume (at) tti - c (dot) org)
852    Date: 2004/09/23 17:30:00 CDT
853    Tag: blume-20040923-envvars
854    Description:
855    
856    1. Changed definition of type ControlRegistry.registry_tree to
857       include control_info (i.e., the name of the controlling environment
858       variable).
859    
860    2. Added command-line flags -e and -E to print the names of environment
861       variables that can be used to control internal settings.
862       (This uses the new API mentioned in 1.)
863    
864    ----------------------------------------------------------------------
865    Name: Matthias Blume (blume (at) tti - c (dot) org)
866    Date: 2004/09/13 16:50:00 CDT
867    Tag: Release_110_49
868    Description:
869    
870    New working version (110.49).  NEW BOOTFILES!
871    
872    ----------------------------------------------------------------------
873    Name: Matthias Blume
874    Date: 2004/09/13 16:20:00 CDT
875    Tag: blume-20040913-config-mlrisc
876    Description:
877    
878    Put target "mlrisc" back into the default list.
879    (There is no harm in having it, and some users have expressed their
880    wish to have "mlrisc" included by default.)
881    
882    ----------------------------------------------------------------------
883    Name: John Reppy
884    Date: 2004/09/13
885    Tag: jhr-20040913-signals
886    Description:
887    
888    Fixed the signal masking code to properly nest mask/unmask operations
889    on a per-signal basis.
890    
891    ----------------------------------------------------------------------
892    Name: Matthias Blume (blume (at) tti - c (dot) org)
893    Date: 2004/09/08 13:20:00 CDT
894    Tag: blume-20040908-heap-magic
895    Description:
896    
897    Bumped the heap macig number to 0x09082004 to account for the changed
898    layout of the ML frame under MacOS X.
899    
900    ----------------------------------------------------------------------
901    Name: Allen Leung (leunga (at) reservoir (dot) com)
902    Date: 2004/09/03 11:26:00 EST
903    Tag: leunga-20040903-cygwin-install
904    Description:
905    
906       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
907    The Cygwin runtime is turned on by setting the environment
908    variable SMLNJ_CYGWIN_RUNTIME to 1.
909    
910    ----------------------------------------------------------------------
911    Name: Matthias Blume (blume (at) tti - c (dot) org)
912    Date: 2004/08/31 17:15:00 CDT
913    Tag: blume-20040831-core
914    Description:
915    
916    Added some exports to src/compiler/core.cm upon request by J. Joemann.
917    
918    ----------------------------------------------------------------------
919    Name: Matthias Blume (blume (at) tti - c (dot) org)
920    Date: 2004/08/30 17:55:00 CDT
921    Tag: blume-20040830-installer
922    Description:
923    
924    Upon request by Johannes Joemann:
925    
926    - improved ML code of installer to fall back to coping when renaming
927      fails (i.e., when source and target are on different file systems);
928      the code compiles but has yet to be tested in anger
929    - removed mlrisc from list of default targets (config/targets)
930    
931    ----------------------------------------------------------------------
932    Name: Matthias Blume (blume (at) tti - c (dot) org)
933    Date: 2004/08/27 17:20:00 CDT
934    Tag: blume-20040827-ptreql
935    Description:
936    
937    Added ptreql primop to structure InlineT (upon request from Larry
938    Paulson).
939    
940    ----------------------------------------------------------------------
941    Name: Allen Leung (leunga (at) reservoir (dot) com)
942    Date: 2004/08/15 21:21:00 EST
943    Tag: leunga-110_48-udgraph
944    Description:
945    
946    Another bug fix from Carl Hauser:
947    
948    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
949    > 48c48
950    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
951    > e::L)
952    > ---
953    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
954    > Without this, any deletion of an edge in an undirected graph does severe
955    > violence to the graph.
956    
957    ----------------------------------------------------------------------
958    Name: Allen Leung (leunga (at) reservoir (dot) com)
959    Date: 2004/08/10 23:35:00 EST
960    Tag: leunga-110_48-ppc
961    Description:
962    
963       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
964    
965    ----------------------------------------------------------------------
966    Name: Matthias Blume (blume (at) tti - c (dot) org)
967    Date: 2004/08/10 12:00:00 CDT
968    Tag: Release_110_48
969    Description:
970    
971    New working version (110.48).  NEW BOOTFILES!
972    
973    ----------------------------------------------------------------------
974    Name: Allen Leung (leunga (at) reservoir (dot) com)
975    Date: 2004/08/09 12:21:00 EST
976    Tag: leunga-110_47-dijsktra
977    Description:
978    
979    Bug fix from Carl Hauser:
980    
981    single_source_shortest_paths in dijkstra.sml was observed to get wrong
982    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
983    
984    The problem is that following the expression A.update(dist,s,Num.zero)
985    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
986    
987    ----------------------------------------------------------------------
988    Name: Matthias Blume (blume (at) tti - c (dot) org)
989    Date: 2004/08/06 18:10:00 CDT
990    Tag: blume-20040806-cmdline
991    Description:
992    
993    Fiddled with handling of command-line options:
994    
995      * sml now quits after processing the command line
996        if -H, -S, -h<n>, or -s<n> appears as the last
997        command-line argument
998      * a new option -q terminates the session when encountered on
999        the command line; subsequent arguments will be ignored
1000      * bug fixes: short (erroneous) arguments are no longer ignored
1001        completely
1002    
1003    ----------------------------------------------------------------------
1004    Name: Allen Leung (leunga (at) reservoir (dot) com)
1005    Date: 2004/08/04 18:17:00 EST
1006    Tag: leunga-110_47-ppc-ibm-asm
1007    Description:
1008    
1009      - Added minimal IBM assembly syntax support for PowerPC.
1010    
1011      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1012        been altered in the runtime.  We need an automatic way to keep the file
1013        in sync.
1014    
1015    ----------------------------------------------------------------------
1016    Name: Matthias Blume (blume (at) tti - c (dot) org)
1017    Date: 2004/08/04 14:00:00 CDT
1018    Tag: Release_110_47
1019    Description:
1020    
1021    New working version (110.47).  NEW BOOTFILES!
1022    
1023    ----------------------------------------------------------------------
1024    Name: Matthias Blume (blume (at) tti - c (dot) org)
1025    Date: 2004/08/03 14:25:00 CDT
1026    Tag: blume-20040803-callingconv
1027    Description:
1028    
1029    Added low-level support for choosing C calling conventions by
1030    twiddling the type of rawccall.  (See
1031    src/compiler/Semant/types/cproto.sml for details.)
1032    
1033    ----------------------------------------------------------------------
1034    Name: Matthias Blume (blume (at) tti - c (dot) org)
1035    Date: 2004/08/02 15:55:00 CDT
1036    Tag: blume-20040802-backout
1037    Description:
1038    
1039    Backed out of change to win32-filesys.c.  The earlier patch to
1040    get_file_time caused CM to produce files with the wrong time stamp.
1041    
1042    ----------------------------------------------------------------------
1043    Name: Matthias Blume (blume (at) tti - c (dot) org)
1044    Date: 2004/08/02 14:45:00 CDT
1045    Tag: blume-20040802-nlffi-win32
1046    Description:
1047    
1048    Added NLFFI support for Win32, adapted from a patch provided by David
1049    Hansel.  This is currently completely untested.  Also, the issue
1050    concerning stdcall vs. ccall is still unresolved.
1051    
1052    ----------------------------------------------------------------------
1053    Name: Matthias Blume (blume (at) tti - c (dot) org)
1054    Date: 2004/07/30 17:55:00 CDT
1055    Tag: blume-20040730-various
1056    Description:
1057    
1058    Gearing up towards 110.47...
1059    
1060    - various minor bugfixes to ml-nlffigen
1061    - a beginning of a manual for nlffi
1062    
1063    - eliminated 'export name=value' in config/install.sh as this does
1064      not work with certain versions of /bin/sh
1065      (Thanks to David King at Motorola for catching this.)
1066    
1067    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1068      - added a test for tm==NULL to gmtime.c and localtime.c
1069      - applied patch for incorrect GetFileTime under win32
1070      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1071    
1072    ----------------------------------------------------------------------
1073    Name: Matthias Blume (blume (at) tti - c (dot) org)
1074    Date: 2004/07/21 18:20:00 CDT
1075    Tag: blume-20040721-nlffigen
1076    Description:
1077    
1078    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1079      for a typedef to an incomplete type, but leave out the "typ" member.
1080      (This is just for consistency.)
1081    - Started to produce what is supposed to become better (i.e., comprehensive)
1082      documentation of what ml-nlffigen does and produces.
1083    
1084    ----------------------------------------------------------------------
1085    Name: Matthias Blume (blume (at) tti - c (dot) org)
1086    Date: 2004/07/14 16:25:00 CDT
1087    Tag: blume-20040714-union
1088    Description:
1089    
1090    Added C_UNION to c-calls/c-types.sml and updated the machinery
1091    (ml-nlffigen, cproto.sml) that conveys C function interface
1092    information to the code generator.
1093    
1094    However, the actual architecture-specific implementation of function
1095    arguments and results that are C unions is still not implemented.
1096    
1097    ----------------------------------------------------------------------
1098    Name: Allen Leung (leunga (at) reservoir (dot) com)
1099    Date: 2004/07/14 14:38:00 EST
1100    Tag: leunga-110_46_1-ppc-lwzu
1101    Description:
1102    
1103        Added these instructions to the PowerPC architecture:
1104        LBZU(X), LHZU(X), LWZU(X),
1105        STWU(X), STFDU, STFSU
1106    
1107        etc...
1108    
1109        Note: I haven't added their instruction encoding into the description.
1110    
1111    ----------------------------------------------------------------------
1112    Name: Allen Leung (leunga (at) reservoir (dot) com)
1113    Date: 2004/07/13 15:04:00 EST
1114    Tag: leunga-110_46_1-ppc-lwarx
1115    Description:
1116    
1117       Added the two instructions LWARX and STWCX to the PowerPC
1118    instruction set.
1119    
1120       A (untested) rewrite of loop-structure.sml.  The old version
1121    is completely broken.
1122    
1123    ----------------------------------------------------------------------
1124    Name: Matthias Blume (blume (at) tti - c (dot) org)
1125    Date: 2004/07/13 13:50:00 CDT
1126    Tag: blume-20040713-nlffi
1127    Description:
1128    
1129    - use paramAlloc to report c-calls with too many arguments
1130      (for PPC version where parameter area is pre-allocated)
1131    - added ccall_maxargspace to machspec (to implement the above)
1132    - made "make" commend in CM's "make" tool configurable
1133    - added option (default: on) for passing the name of the SML/NJ's "bin"
1134      directory to "make"; the call looks like this:
1135    
1136         make <options> SMLNJ_BINDIR=<dir> <target>
1137    
1138      This can be used by the Makefile to, e.g., pick the "right" version
1139      of ml-nlffigen.
1140    - minor code tweaks
1141    
1142    ----------------------------------------------------------------------
1143    Name: Matthias Blume (blume (at) tti - c (dot) org)
1144    Date: 2004/07/12 22:50:00 CDT
1145    Tag: blume-110_46_1-macosx-nlffi
1146    Description:
1147    
1148    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1149    though.
1150    
1151    Note:  1. You have to make a new, clean build of the runtime system.
1152           2. There are new BOOTFILES, you have to use them!
1153              (Doing the bootstrap process yourself would be *very* painful!
1154               If you absolutely have to do it, build the system under
1155               a different architecture and then cross-compile.)
1156    
1157    Version bumped to 110.46.1 to account for runtime data format changes.
1158    
1159    ----------------------------------------------------------------------
1160    Name: Matthias Blume (blume (at) tti - c (dot) org)
1161    Date: 2004/06/18 14:30:00 CDT
1162    Tag: blume-20040618-unix
1163    Description:
1164    
1165    Changed the implementation of structure Unix so that the same stream
1166    is returned every time one of the {text,bin}{In,Out}streamOf functions
1167    is invoked on the same proc.  This is not what the spec currently
1168    says -- although IMO it arguably should.  (See discussion below.)
1169    
1170    ----------------------------------------------------------------------
1171    Name: Matthias Blume (blume (at) tti - c (dot) org)
1172    Date: 2004/06/17 18:15:00 CDT
1173    Tag: Release_110_46
1174    Description:
1175    
1176    New working version (110.46).  NEW BOOTFILES!
1177    
1178    ----------------------------------------------------------------------
1179    Name: Matthias Blume (blume (at) tti - c (dot) org)
1180    Date: 2004/06/17 17:20:00 CDT
1181    Tag: blume-20040617-timer-unix
1182    Description:
1183    
1184    Changed the interface of structures Timer and Unix to match the most
1185    recent Basis spec.
1186    
1187    In the case of Unix there still seems to be an open/weird issue:
1188    
1189       The {text,bin}{In,Out}streamOf functions are supposed to create
1190       fresh streams whenever they are called -- as opposed to have them
1191       return the same stream every time.  This design is supposed to
1192       prevent space leaks caused by proc values hanging on to streams.
1193    
1194       The reap function, on the other hand, is supposed to close the
1195       streams.  This cannot be done without having a handle on the
1196       stream in proc after all...
1197    
1198    I took the liberty to implement the following stopgap solution:
1199    
1200      The proc value hangs on to the most recently created stream(s).
1201      Reap closes those.  If either or both of the two streams hadn't
1202      been created at all yet, then reap will close the corresponding
1203      file descriptors directly.
1204    
1205    PS: I don't understand the original space leak argument anymore.  If
1206    a proc hangs on to the imperative stream, then I/O operations on those
1207    will advance the state of the cached stream and avoid the space leak.
1208    
1209    ----------------------------------------------------------------------
1210    Name: Matthias Blume (blume (at) tti - c (dot) org)
1211    Date: 2004/05/28 16:45:00 CDT
1212    Tag: blume-20040528-basis
1213    Description:
1214    
1215    Added signature PACK_REAL and exported functor PrimIO.
1216    
1217    ----------------------------------------------------------------------
1218    Name: Matthias Blume (blume (at) tti - c (dot) org)
1219    Date: 2004/05/25 16:00:00 CDT
1220    Tag: blume-20040525-group-owner
1221    Description:
1222    
1223    CM now ignores (but still accepts) the "owner" information in group
1224    descriptions.  The owner of a group is its next enclosing
1225    library. Each group must have a unique owner.  (There is a virtual
1226    "toplevel" library that owns groups which are not nested within a real
1227    library.)  Previously, each group had to explicitly declare its owner,
1228    and CM would check that such a declaration is correct.  The new scheme
1229    is to have CM check that for each group there is precisely one owning
1230    library.
1231    
1232    The advantage of the new scheme is that the programmer no longer needs
1233    to maintain the somewhat annoying owner information.  The downside is
1234    that CM cannot enforce the ownership rule across multiple runs of
1235    CM.make.  Fortunately, enclosing the same group in two different
1236    libraries A and B which are not part of the same program does not
1237    cause real problems.
1238    
1239    ----------------------------------------------------------------------
1240    Name: Matthias Blume (blume (at) tti - c (dot) org)
1241    Date: 2004/05/20 16:00:00 CDT
1242    Tag: blume-20040520-win32
1243    Description:
1244    
1245    Made the win32 version work again.  (Strangely, a misplaced comma had
1246    slipped into win32-process.c which prevented the runtime from being
1247    compiled correctly.)
1248    
1249    Also, included a minor addition to ml-build.bat analogous to what was
1250    done in blume-20040519-ml-build.
1251    
1252    ----------------------------------------------------------------------
1253    Name: Matthias Blume (blume (at) tti - c (dot) org)
1254    Date: 2004/05/19 22:10:00 CDT
1255    Tag: blume-20040519-ml-build
1256    Description:
1257    
1258    Arranged for ml-build to clean up after itself a little bit better.
1259    The script generates a temporary SML source file and compiles it using
1260    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1261    gets rid of those at the end, so they don't accumulate under .cm.
1262    
1263    This required a minor change to install.sh because the name of the
1264    metadata directory (default: .cm) is actually configurable at
1265    installation time.
1266    
1267    ----------------------------------------------------------------------
1268    Name: Matthias Blume (blume (at) tti - c (dot) org)
1269    Date: 2004/05/18 15:50:00 CDT
1270    Tag: blume-20040518-mkreader
1271    Description:
1272    
1273    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1274    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1275    
1276    ----------------------------------------------------------------------
1277    Name: Matthias Blume (blume (at) tti - c (dot) org)
1278    Date: 2004/05/11 14:35:00 CDT
1279    Tag: blume-20040511-win32sock
1280    Description:
1281    
1282    Added previously missing support for many socket-related functions
1283    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1284    for the voluminous patch!
1285    
1286    (I have not tested this patch under win32 yet.)
1287    
1288    Here is David's e-mail:
1289    
1290    Hi,
1291    
1292    Attached to this email you find a diff against sml/nj 110.45
1293    that will enable socket support under Windows.
1294    
1295    To apply the patch (using unix or cygwin)
1296    1) gunzip runtime.diff.gz
1297    2) "cd" into "src/runtime" in the source tree of a fresh
1298       110.45 installation.
1299    3) patch -p 1 < [your/path/to]runtime.diff
1300    
1301    The code compiles fine but has NOT yet been extensively tested.
1302    I only ran a few tests for basic socket client functionality
1303    (which worked fine).  Especially the functions that use ioctl
1304    are not tested at all and might not work (see below).
1305    
1306    I implemented this since we want to move to a newer version of sml/nj
1307    but need socket support in order to use it.  This is the first time I
1308    even had a look at the sml/nj source,  so please review my changes
1309    before making this part of the distribution!  Here are a few issues
1310    that I think might be better for someone to solve who is more
1311    familiar with the sml/nj source (and socket programming):
1312    
1313    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1314      exception since I could not figure out what the windows equivalent
1315      of these functions is
1316    
1317    - In sockets-osdep.h there are a some #include statements that are
1318      only used in a few files that include sockets-osdep.h
1319    
1320    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1321      does not process its return value since I don't know how to
1322      report an error upwards.
1323    
1324    - It would probably be good to have a call to WSACleanup() when
1325      the library is unloaded (if there is such a possibility).
1326      Otherwise I think Windows will take care of this automatically
1327      when the process finishes.
1328    
1329    - I used ioctlsocket() as a replacement for ioctl() but I have
1330      no idea if that is actually the proper replacement on Windows.
1331    
1332    - All these issues are marked in the code by "FIXME" comments.
1333    
1334    We use sml/nj extensively in our products and are quite happy
1335    with it.  I hope this contribution will help you.
1336    
1337    Keep up the good work!
1338    
1339    David
1340    
1341    ----------------------------------------------------------------------
1342    Name: Matthias Blume (blume (at) tti - c (dot) org)
1343    Date: 2004/05/11 14:20:00 CDT
1344    Tag: blume-20040511-installml
1345    Description:
1346    
1347    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1348    the patch.)
1349    
1350    ----------------------------------------------------------------------
1351    Name: Matthias Blume (blume (at) tti - c (dot) org)
1352    Date: 2004/05/11 14:05:00 CDT
1353    Tag: blume-20040511-nlffi-netbsd
1354    Description:
1355    
1356    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1357    the patch.)
1358    
1359    ----------------------------------------------------------------------
1360    Name: Matthias Blume (blume (at) tti - c (dot) org)
1361    Date: 2004/05/11 12:05:00 CDT
1362    Tag: blume-20040511-exports
1363    Description:
1364    
1365    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1366    export lists in compiler-related .cm-files.
1367    
1368    ----------------------------------------------------------------------
1369    Name: Matthias Blume (blume (at) tti - c (dot) org)
1370    Date: 2004/05/11 11:35:00 CDT
1371    Tag: blume-20040511-allsource
1372    Description:
1373    
1374    The installer now honors the "src-smlnj" target again, although its meaning
1375    has changed from "all sources required for the compiler" to "all sources
1376    the installer knows about".  In other words, if you enable "src-smlnj"
1377    in the "targets" file, then the installer will pull in sources for
1378    everything.  (Notice that this refers to source code only.  Compiled
1379    code is still only installed for modules that were requested explicitly
1380    or which are required for other modules that were requested explicitly.)
1381    
1382    ----------------------------------------------------------------------
1383    Name: Matthias Blume (blume (at) tti - c (dot) org)
1384    Date: 2004/04/23 17:40:00 CDT
1385    Tag: blume-20040423-ieee-scan
1386    Description:
1387    
1388    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1389    in the exponent calculation we get INF or ZERO (depending on the mantissa
1390    and the sign of the exponent).
1391    
1392    ----------------------------------------------------------------------
1393    Name: Matthias Blume (blume (at) tti - c (dot) org)
1394    Date: 2004/04/23 10:40:00 CDT
1395    Tag: blume-20040423-ml-build
1396    Description:
1397    
1398    The ml-build script now terminates with a non-0 status when something
1399    goes wrong.
1400    
1401    ----------------------------------------------------------------------
1402    Name: Matthias Blume (blume (at) tti - c (dot) org)
1403    Date: 2004/04/22 16:35:00 CDT
1404    Tag: blume-20040422-Option
1405    Description:
1406    
1407    Made exception Option to be the same as exception Option.Option
1408    (as it should be).
1409    
1410    ----------------------------------------------------------------------
1411    Name: Allen Leung (leunga (at) reservoir (dot) com)
1412    Date: 2004/03/19 14:40:00 EST
1413    Tag: leunga-20040319-cygwin-nlffi
1414    Description:
1415    
1416    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1417    of SML/NJ.  The problem is that
1418    
1419       lib = dlopen(NULL, ...)
1420       f   = dlsym(lib, "malloc");
1421    
1422    does not work on Windows unless we explicitly export symbols
1423    such as 'malloc' during linking.  We fixed this by explicitly
1424    exporting the required symbols with the magic gcc incantation:
1425    
1426        -Wl,--export-all cygwin.def
1427    
1428    where cygwin.def is a file containing all the symbols that
1429    we wish to export.
1430    
1431    I suspect this is a Windows problem and we'll have to
1432    do the same (somehow with windows compilers) when
1433    we build the native win32 version with the system
1434    calls LoadLibrary/GetProcAddress.
1435    
1436    ----------------------------------------------------------------------
1437    Name: Matthias Blume (blume (at) tti - c (dot) org)
1438    Date: 2004/03/04 16:35:00 CST
1439    Tag: blume-20040304-intinf-fmt
1440    Description:
1441    
1442    Fixed problem with IntInf.fmt (sign would show up on the right instead
1443    of on the left for BIN, OCT, and HEX).
1444    
1445    ----------------------------------------------------------------------
1446    Name: Matthias Blume (blume (at) tti - c (dot) org)
1447    Date: 2004/03/04 11:25:00 CST
1448    Tag: blume-20040304-symlinks
1449    Description:
1450    
1451    Fixed problem with installer script (unix only) where bin/ml-yacc and
1452    friends pointed (via symlinks) to absolute locations instead of just
1453    .run-sml.  This was reported by Vesa A Norrman.
1454    
1455    ----------------------------------------------------------------------
1456    Name: Matthias Blume (blume (at) tti - c (dot) org)
1457    Date: 2004/02/13 14:50:00 CST
1458    Tag: Release_110_45
1459    Description:
1460    
1461    New working version (110.45). New bootfiles.
1462    
1463    ----------------------------------------------------------------------
1464    Name: Matthias Blume (blume (at) tti - c (dot) org)
1465    Date: 2004/01/26 15:15:15 CST
1466    Tag: blume-20040126-toplevel
1467    Description:
1468    
1469    Improved handling of exceptions at the interactive toplevel.
1470    
1471    ----------------------------------------------------------------------
1472    Name: Matthias Blume (blume (at) tti - c (dot) org)
1473    Date: 2004/01/26 11:25:00
1474    Tag: blume-20040126-app
1475    Description:
1476    
1477    Type of top-level "app" corrected.
1478    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1479    
1480    ----------------------------------------------------------------------
1481    Name: Matthias Blume (blume (at) tti - c (dot) org)
1482    Date: 2003/11/18 17:10 CST
1483    Tag: blume-20031118-basis-fiddle
1484    Description:
1485    
1486    - changed Timer interface to what might become the spec
1487    - POSIX_FLAGS -> BIT_FLAGS according to spec
1488    - some other minor discrepancies wrt. spec eliminated
1489    
1490    ----------------------------------------------------------------------
1491    Name: Matthias Blume (blume (at) tti - c (dot) org)
1492    Date: 2003/11/06 12:00:00 CST
1493    Tag: Release_110_44
1494    Description:
1495    
1496    New working version (110.44). New bootfiles.
1497    
1498    ----------------------------------------------------------------------
1499    Name: Matthias Blume (blume (at) tti - c (dot) org)
1500    Date: 2003/11/04 11:50:00 CST
1501    Tag: blume-20031104-move-libraries
1502    Description:
1503    
1504    Eliminated the "dont_move_libraries" directive in config/targets.
1505    (The mechanism was broken and could not be fixed easily.  Moreover,
1506    there does not seem to be any reason not to move all libraries into
1507    lib during installation.  I originally implemented this directive as a
1508    backward-compatibility feature when I first introduced the new CM.
1509    Now that things have been stable for a long time and going back to the
1510    old CM is not an option, there is no reason to keep it around.)
1511    
1512    ----------------------------------------------------------------------
1513    Name: Matthias Blume (blume (at) tti - c (dot) org)
1514    Date: 2003/11/03 16:00:00 CST
1515    Tag: blume-20031103-installdir
1516    Description:
1517    
1518    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1519    Richards for pointing out the problem and providing the solution.)
1520    
1521    ----------------------------------------------------------------------
1522    Name: Matthias Blume (blume (at) tti - c (dot) org)
1523    Date: 2003/10/01 17:05:00 CDT
1524    Tag: blume-20031001-lal-mlrisc
1525    Description:
1526    
1527    MLRISC bug fix from Lal.
1528    
1529    ----------------------------------------------------------------------
1530    Name: Matthias Blume (blume (at) tti - c (dot) org)
1531    Date: 2003/09/30 16:10:00 CDT
1532    Tag: blume-20030930-primio-bat
1533    Description:
1534    
1535    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1536    2. Improved .bat files (for Win32 port) to make things work under Win95.
1537       (thanks to Aaron S. Hawley for this one)
1538    
1539    ----------------------------------------------------------------------
1540    Name: Matthias Blume (blume (at) tti - c (dot) org)
1541    Date: 2003/09/26 16:05:00 CDT
1542    Tag: blume-20030926-wrappriv
1543    Description:
1544    
1545    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1546    
1547    ----------------------------------------------------------------------
1548    Name: Matthias Blume (blume (at) tti - c (dot) org)
1549    Date: 2003/09/26 15:00:00 CDT
1550    Tag: blume-20030926-110_43_3
1551    Description:
1552    
1553    - additional cleanup
1554    - version number bump, NEW BOOTFILES
1555    
1556    ----------------------------------------------------------------------
1557    Name: Matthias Blume (blume (at) tti - c (dot) org)
1558    Date: 2003/09/26 12:00:00 CDT
1559    Tag: blume-20030926-ppautoload
1560    Description:
1561    
1562    I modified the read-eval-print loop so that the autoloader gets
1563    invoked whenever the prettyprinter tries to look up a symbol that
1564    is not currently defined in the toplevel environment but which
1565    appears in CM's autoload registry.  As a result, we see far fewer of
1566    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1567    
1568    In addition to this I tried to clean up some pieces of the Basis
1569    implementation (e.g., Socket, Word8Array) in order to prevent other
1570    instances of these ?.Foo.Bar.xxx names from being printed.
1571    
1572    The mechanism that picks names for types still needs some work, though.
1573    (Right now it seems that if there is a type A.t which is defined to
1574    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1575    "?.B.u" although the perhaps more sensible solution would be to use
1576    "A.t" in this case.  In other words, the prettyprinter should follow
1577    a chain of DEFtycs not farther than there are corresponding toplevel
1578    names in the current environment.)
1579    
1580    ----------------------------------------------------------------------
1581    Name: Matthias Blume (blume (at) tti - c (dot) org)
1582    Date: 2003/09/24 16:31:00 CDT
1583    Tag: blume-20030924-installer
1584    Description:
1585    
1586    Another installer tweak:  All the ML code for the installer is now
1587    compiled during CMB.make and put into a little library called
1588    $smlnj/installer.cm.  The installation then simply invokes
1589    
1590       sml -m $smlnj/installer.cm
1591    
1592    and everything happens automagically.
1593    
1594    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1595    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1596            and BINDIR.
1597    
1598    The new scheme guarantees that the ML code responsible for the installation
1599    is in sync with the APIs of the main system.  Also, the installer is
1600    somewhat faster because the installer script is precompiled.
1601    
1602    ----------------------------------------------------------------------
1603    Name: Matthias Blume (blume (at) tti - c (dot) org)
1604    Date: 2003/09/24 15:35:00 CDT
1605    Tag: blume-20030924-synsock
1606    Description:
1607    
1608    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1609    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1610    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1611    implementations of non-blocking operations from CML's Socket
1612    structure.
1613    
1614    ----------------------------------------------------------------------
1615    Name: Matthias Blume (blume (at) tti - c (dot) org)
1616    Date: 2003/09/24 15:10:05 CDT
1617    Tag: blume-20030924-sockets
1618    Description:
1619    
1620    1. Fixed SOCKET API and implementation to match Basis spec.
1621       This required changing the internal representation of sockets to one
1622       that remembers (for each socket file descriptor) whether it is currently
1623       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1624       call is made only if the state actually needs to change).
1625    
1626    2. OS-specific details of sockets were moved into separate files, thus
1627       making it possible to unify the bulk of the socket implementations
1628       between Unix and Win32.
1629    
1630    3. CML's socket API changed accordingly.
1631       (Note that we need to remove non-blocking functions from this API
1632       since they are redundant in the case of CML!)
1633    
1634    4. CML's socket implementation now makes use of non-blocking functions
1635       provided by Basis, thus removing all OS-dependent code from this part
1636       of CML.
1637    
1638    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1639    
1640    ----------------------------------------------------------------------
1641    Name: Matthias Blume (blume (at) tti - c (dot) org)
1642    Date: 2003/09/22 12:10:00 CDT
1643    Tag: blume-20030922-110_43_2
1644    Description:
1645    
1646    Made a new interim version and bootfiles for developer's bootstrapping
1647    convenience.
1648    
1649       110.43.2 -- NEW BOOTFILES
1650    
1651    ----------------------------------------------------------------------
1652    Name: Matthias Blume (blume (at) tti - c (dot) org)
1653    Date: 2003/09/19 15:55:00 CDT
1654    Tag: blume-20030919-cmdir
1655    Description:
1656    
1657    1. new-install.sh -> install.sh
1658    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1659    3. tweaked installer so that another name instead of .cm can be chosen
1660       at install time (by setting the CM_DIR_ARC environment variable
1661       during installation); once installation is complete, the name is
1662       fixed
1663    
1664    ----------------------------------------------------------------------
1665    Name: Matthias Blume (blume (at) tti - c (dot) org)
1666    Date: 2003/09/18 16:00:00 CDT
1667    Tag: blume-20030918-110_43_1
1668    Description:
1669    
1670    Made a new interim version and bootfiles for developer's bootstrapping
1671    convenience.
1672    
1673       110.43.1 -- NEW BOOTFILES
1674    
1675    ----------------------------------------------------------------------
1676    Name: Matthias Blume (blume (at) tti - c (dot) org)
1677    Date: 2003/09/18 15:20:00 CDT
1678    Tag: blume-20030918-misc
1679    Description:
1680    
1681    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1682       this seems to be controversial at the moment) and used those in
1683       Posix.ProcEnv.times.
1684    
1685    2. Added Time.{from,to}Nanoseconds to Time.
1686    
1687    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1688       For example, fromLargeInt never needs to look at more than 3 "big
1689       digits" to get its 53 bits of precision.
1690    
1691    ----------------------------------------------------------------------
1692    Name: Matthias Blume (blume (at) tti - c (dot) org)
1693    Date: 2003/09/17 16:30:00 CDT
1694    Tag: blume-20030917-real32-slices
1695    Description:
1696    
1697    Added an entry to the primitive environment
1698    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1699    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1700    
1701    Removed some of the "magic" constants in real64.sml and replaced them
1702    with code that generates these values from their corresponding
1703    integer counterparts.
1704    
1705    Made all(?) the slice-related changes to the Basis and made everything
1706    compile again...
1707    
1708    ----------------------------------------------------------------------
1709    Name: Matthias Blume (blume (at) tti - c (dot) org)
1710    Date: 2003/09/15 17:45:00 CDT
1711    Tag: blume-20030915-rbase
1712    Description:
1713    
1714    Fixed bug in Real.fromLargeInt.
1715    
1716    ----------------------------------------------------------------------
1717    Name: Matthias Blume (blume (at) tti - c (dot) org)
1718    Date: 2003/09/13 18:11:00 CDT
1719    Tag: blume-20030913-libinstall
1720    Description:
1721    
1722    Minor bugfix in config/libinstall (set anchor with path to
1723    standalone tool after installing it, otherwise libraries that
1724    need ml-lex or ml-yacc won't compile the first time the installer
1725    runs).
1726    
1727    ----------------------------------------------------------------------
1728    Name: Matthias Blume (blume (at) tti - c (dot) org)
1729    Date: 2003/09/12 11:45:00 CDT
1730    Tag: blume-20030912-various
1731    Description:
1732    
1733     - fixed bug in Real.toLargeInt
1734     - fixed bug in Posix.ProcEnv.times
1735     - changed inputLine functions to return an option
1736     - minor installer improvements / bugfixes
1737     - changed default @SMLalloc parameter for x86/celeron to 64k
1738    
1739    ----------------------------------------------------------------------
1740    Name: Matthias Blume (blume (at) tti - c (dot) org)
1741    Date: 2003/09/09 22:00:00 CDT
1742    Tag: Release_110_43
1743    Description:
1744    
1745        New working release 110.43.  New bootfiles.
1746    
1747    ----------------------------------------------------------------------
1748    Name: Matthias Blume (blume (at) tti - c (dot) org)
1749    Date: 2003/09/09 19:20:00 CDT
1750    Tag: blume-20030909-installer
1751    Description:
1752    
1753    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1754    Modified config/install.bat to take advantage of it.  Also modified
1755    config/install.sh (and called it config/new-install.sh) to take advantage
1756    of it on Unix systems.  (The SML code is (supposed to be) platform-
1757    independent.)
1758    
1759    The installer can now install everything under Win32
1760    as well as under *nix as long as it compiles.
1761    
1762    Other changes:
1763    
1764      - made CML compile again under Win32
1765      - made eXene compile under Win32 (by providing a fake structure UnixSock
1766        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1767      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1768        same as type int;  under Win32 it isn't
1769      - fixed some slice-related problems in the win32-specific parts of CML
1770      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1771        (under Win32, the volume name is case-insensitive, and the
1772        OS.Path code compares volume names for equality)
1773    
1774    ----------------------------------------------------------------------
1775    Name: Matthias Blume (blume (at) tti - c (dot) org)
1776    Date: 2003/09/08 11:55:00 CDT
1777    Tag: blume-20030908-fullpath
1778    Description:
1779    
1780    Made Win32 version of OS.FileSys.fullPath return current directory
1781    when given an empty string.  This is what the spec says, and incidentally,
1782    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1783    cases when presented with the name of a non-existing .cm file.)
1784    
1785    ----------------------------------------------------------------------
1786    Name: Matthias Blume (blume (at) tti - c (dot) org)
1787    Date: 2003/09/04 16:30:00 CDT
1788    Tag: blume-20030905-slices-etc
1789    Description:
1790    
1791      1. Changed interface to vectors and arrays in Basis to match
1792         (draft) Basis spec.
1793      2. Added signatures and implementations of slices according to
1794         Basis spec.
1795      3. Edited source code throughout the system to make it compile again
1796         under 1. and 2.  (In some cases code had to be added to have it
1797         match the new signatures.)
1798      4. MLRISC should be backward-compatible: the copies of the originals
1799         of files that needed to change under 3. were retained, the .cm files
1800         check the compiler version number and use old versions when
1801         appropriate.
1802      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1803         dirstream -> string option (in accordance with Basis spec).
1804      6. When generating code that counts lines, ml-lex used function
1805         CharVector.foldli, taking advantage of its old interface.
1806         This has been replaced with the corresponding code from
1807         CharVectorSlice.  (html-lex must be re-lexed!)
1808      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1809         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1810      8. Removed temporary additions (fromInternal, toInternal) from the
1811         (now obsolete) IntInf in smlnj-lib/Util.
1812      9. Cleaned up structure Byte.
1813     10. Added localOffset, scan, and fromString to Date (according to spec).
1814         Cleaned/corrected implementation of Date.
1815         (Still need to check for correctness; implement better canonicalizeDate.)
1816     11. Added "scan" to signature IEEE_REAL.
1817     12. Some improvements to IntInf [in particular: efficiency-hack for
1818         mod and rem when second operand is 2 (for parity checks).]
1819     13. Changed representation of type Time.time, using a single IntInf.int
1820         value counting microseconds.  This considerably simplified the
1821         implementation of structure Time.  We now support negative time
1822         values; scan and fromString handle signs.
1823     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1824         ArraySlice).
1825    
1826    ----------------------------------------------------------------------
1827    Name: Matthias Blume (blume (at) tti - c (dot) org)
1828    Date: 2003/08/28 17:00:00 CDT
1829    Tag: blume-20030828-intinf
1830    Description:
1831    
1832    This is a major update which comes with a version number bump
1833    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1834    and an implementation of IntInf in the Basis.
1835    
1836    There are a fairly large number of related changes and updates throughout
1837    the system:
1838    
1839    Basis:
1840            - Implemented IntInf.
1841            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1842            - Added some missing Real64 operations, most notably Real.toLargeInt.
1843            - Added FixedInt as a synonym for Int32.
1844    
1845    compiler:
1846            * Added support for a built-in intinf type.
1847              - literals
1848              - pattern matching
1849              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1850              - overloading on literals and operations
1851    
1852            This required adding a primitive type intinf, some additional
1853            primops, and implementations for several non-trivial intinf
1854            operations in Core.  (The intinf type is completely abstract
1855            to the compiler; all operations get delegated back to the Core.)
1856    
1857            * Intinf equality is handled by polyequal.  However, the compiler
1858            does not print its usual warning in this case (since polyequal
1859            is the right thing to do there).
1860    
1861            * Improved the organization of structure InlineT.
1862    
1863            * A word about conversion primops:
1864            If conversions involving intinf do not cancel out during
1865            CPS contract, then the compiler must insert calls to Core functions.
1866            Since all core access must be resolved already during the FLINT
1867            translate phase, it would be too late a the time of CPS contract
1868            to add new Core calls.  For this reason, conversion primops
1869            for intinf carry two arguments:  1. the numeric argument that
1870            they are supposed to convert, and 2. the Core function that
1871            can help with this conversion if necessary.  If CPS contract
1872            eliminates a primop, then the associated Core function becomes
1873            dead and goes away.  Intinf conversion primops that do not get
1874            eliminated by CPS contract get rewritten into calls of their
1875            core functions by a separate, new phase.
1876    
1877    interactive system:
1878            - Control.Print.intinfDepth controls max length of intinf constants
1879            being printed. (Analogous to Control.Print.stringDepth.)
1880            - Cleanup in printutil and pputil: got rid of unused stuff and
1881            duplicates; replaced some of the code with code that makes better
1882            use of library functionality.
1883    
1884    CM:
1885            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1886            are no longer silent.
1887    
1888    CKIT:
1889            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1890            in favor of LargeInt -- which is now the same as IntInf.
1891            CKIT-knowledgable people should check whether this is what's
1892            intended and otherwise change things back to using Int32 or
1893            FixedInt.
1894    
1895    Throughout the code:
1896            Started using IntInf.int literals and built-in operations
1897            (e.g., comparison with 0) where this seems appropriate.
1898    
1899    
1900    ----------------------------------------------------------------------
1901    Name: Dave MacQueen (dbm@cs.uchicago.edu)
1902    Date: 2003/08/13 11:36:00 CDT
1903    Tag: dbm-20030813-mcz-merge1
1904    Description:
1905    
1906    Merging changes from the mcz-branch development branch into trunk.
1907    These changes involve replacement of the emulated old prettyprinter
1908    interface with direct use of the SML/NJ Lib PP library, and fixing
1909    of a couple of bugs (895, 1186) relating to error messages.  A new
1910    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1911    has been added.
1912    
1913    ----------------------------------------------------------------------
1914    Name: Matthias Blume (blume (at) tti - c (dot) org)
1915    Date: 2003/08/11 15:45:00 CDT
1916    Tag: blume-20030811-windows
1917    Description:
1918    
1919    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
1920    
1921        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
1922    
1923    This patch restores SML/NJ's ability to run under win32.  There are a
1924    number of changes, including fixes for several bugs that had gone
1925    unnoticed until now:
1926    
1927      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
1928        trying to run the win32 version from within cygwin.)
1929    
1930      - There are a number of simple .bat scripts that substitute for their
1931        corresponding Unix shell-scripts.  (See below.)
1932    
1933      - The internals of ml-build have been modified slightly.  The main
1934        difference is that instead of calling ".link-sml" (or link-sml.bat)
1935        using OS.Process.system, the ML process delegates this task back
1936        to the script.  Otherwise problems arise in mixed environments such
1937        as Cygwin where scripts look and work like Unix scripts, but
1938        where OS.Process.system cannot run them.
1939    
1940      - In CM, the srcpath pickler used native pathname syntax -- which
1941        is incorrect in the case of cross-compilation.  The new pickle format
1942        is independent of platform-specific naming conventions.
1943    
1944      - Path configuration files (such as lib/pathconfig) can now choose
1945        between native and standard syntax.  Placing a line of the form
1946    
1947            standard!
1948    
1949        into the file causes all subsequent paths to be interpreted using
1950        CM standard pathname syntax (= Unix conventions); a line
1951    
1952            native!
1953    
1954        switches back to native style.  This was needed so that
1955        path config files can be written portably, see src/system/pathconfig.
1956    
1957      - Runtime system:
1958    
1959          - win32-filesys.c:  get_file_time and set_file_time now
1960            access modification time, not creation time.
1961    
1962          - I/O code made aware of new array representation.
1963    
1964          - Bug fixes in X86.prim.masm.
1965    
1966       - src/system/makeml made aware of win32. (For use under cygwin
1967            and other Unix-environments for windows.)
1968    
1969       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
1970         which caused BinIO.inputAll to fail consistently.
1971    
1972    .bat scripts:
1973    
1974       Windows .bat scripts assume that SMLNJ_HOME is defined.
1975    
1976       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
1977         applications (sml, ml-yacc, ml-lex).
1978       - ml-build.bat: analogous to ml-build.
1979       - config\install.bat:  Analogous to config/install.sh.  This requires
1980         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
1981         (nmake etc. must be on the path, and vcvars32 must have been run.)
1982         Moreover, sources for ml-lex and ml-yacc need to exist under src,
1983         and the bootfile hierarchy must have been unpacked under
1984         sml.boot.x86-win32.
1985         The script is very primitive and does a poor job at error checking.
1986         It only installs the base system, ml-lex, and ml-yacc.  No other
1987         libraries are being installed (i.e., you get only those that
1988         are part of the compiler.)
1989       - link-sml.bat:  analogous to .link-sml, but not currently used
1990    
1991    Unrelated bug fixes:
1992    
1993      - ml-nlffigen now exports structures ST_* corresponding to incomplete
1994        types.
1995      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
1996        otherwise.)
1997    
1998    ----------------------------------------------------------------------
1999    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2000    Date: 2003/06/17
2001    Tag: macqueen-20030617-bug895
2002    Description:
2003    
2004    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2005    Tag will be used for new development branch (mcz-branch) for
2006    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2007    
2008    ----------------------------------------------------------------------
2009    Name: Matthias Blume (blume (at) tti - c (dot) org)
2010    Date: 2003/05/27 16:55:00 CDT
2011    Tag: blume-20030527-polyeq
2012    Description:
2013    
2014    Tried to eliminated most cases of polymorphic equality.
2015    
2016    ----------------------------------------------------------------------
2017    Name: Matthias Blume (blume (at) tti - c (dot) org)
2018    Date: 2003/05/21 17:45:00 CDT
2019    Tag: blume-20030517-complete
2020    Description:
2021    
2022    Two changes:
2023    
2024      1. Added a flag for controlling whether non-exhaustive bindings will
2025         be treated as errors (default is false).
2026      2. Cleaned up the *entire* source tree so that CMB.make goes through
2027         without a single non-exhaustive match- or bind warning.
2028    
2029    ----------------------------------------------------------------------
2030    Name: Matthias Blume (blume (at) tti - c (dot) org)
2031    Date: 2003/05/17 10:20:00 CDT
2032    Tag: blume-20030517-absyn
2033    Description:
2034    
2035    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2036    
2037       This mainly affects the quality of error messages.  However, some
2038       of the code is now more straightforward than before. (Treatment of
2039       the above four constructs in translate.sml is much simpler than
2040       the "macro-expansion" that was going on before.  Plus, the mach-
2041       compiler no longer gets invoked just to be able to compile an
2042       if-expression.)
2043    
2044    2. The ErrorMsg.Error exception is now caught and absorbed by the
2045       interactive loop.
2046    
2047    ----------------------------------------------------------------------
2048    Name: Allen Leung
2049    Date: 2003/05/16 13:05:00 CDT
2050    Tag: leunga-20030516-cygwin-runtime
2051    Description:
2052    
2053        Ported the runtime system to cygwin, which uses the unix
2054    x86-unix bin files.  Missing/buggy features:
2055    
2056       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2057         the Cygwin library.
2058       o Ctrl-C handling may be flaky.
2059       o Windows system calls and Windows I/O are not supported.
2060    
2061       A new set of binfiles is located at:
2062    
2063         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2064    
2065       This is only needed for bootstrapping the cygwin version of smlnj.
2066       Other x86 versions can use the existing binfiles.
2067    
2068    ----------------------------------------------------------------------
2069    Name: Matthias Blume
2070    Date: 2003/04/08 15:42:00 CDT
2071    Tag: blume-20030408-listpair
2072    Description:
2073    
2074    1. Added a target 'mlrisc' to installer.
2075    
2076    2. Added missing elements to structure ListPair.
2077    
2078    ----------------------------------------------------------------------
2079    Name: Allen Leung
2080    Date: 2003/01/07 10:40:00 EST
2081    Tag: leunga-20030107-int-rem
2082    Description:
2083    
2084    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2085    The arguments to the SUBL instruction were swapped.
2086    
2087    ----------------------------------------------------------------------
2088    Name: Matthias Blume
2089    Date: 2002/12/12 16:25:00 EST
2090    Tag: blume-20021212-risc-ra
2091    Description:
2092    
2093    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2094    sent the RA into an infinite loop when floating point registers get
2095    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2096    110.37 and 110.38.)
2097    There was another set of potential problems related to the handling of
2098    MLRISC annotations (but those did not yet cause real problems, apparently).
2099    
2100    ----------------------------------------------------------------------
2101    Name: Matthias Blume
2102    Date: 2002/12/06 22:40:00 EST
2103    Tag: blume-20021206-cm-fileid
2104    Description:
2105    
2106    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2107    This fixes the problem of CM getting confused by files that suddenly
2108    change their identity (e.g., by getting unlinked and recreated by some
2109    text editor such as vi).  There might be a better/cheaper/cleaner way
2110    of doing this, but for now this will have to do.
2111    
2112    ----------------------------------------------------------------------
2113    Name: Matthias Blume
2114    Date: 2002/10/28 09:50:00 EST
2115    Tag: blume-20021028-typecheck
2116    Description:
2117    
2118    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2119    
2120    ----------------------------------------------------------------------
2121    Name: Matthias Blume
2122    Date: 2002/10/17 09:10:00 EDT
2123    Tag: Release_110_42
2124    Description:
2125    
2126    In good old tradition, there has been a slight hiccup so that we have
2127    to patch 110.42 after the fact.   The old release tag has been replaced
2128    (see below).
2129    
2130    The change solves a problem with two competing approaches the
2131    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2132    each other's way.
2133    
2134    This change only affects the runtime system code and the installer script.
2135    (No new bootfiles.)
2136    
2137    ----------------------------------------------------------------------
2138    Name: Matthias Blume
2139    Date: 2002/10/16 12:00:00 EDT
2140    Tag: Release_110_42_removed
2141    Description:
2142    
2143    New working release.  New bootfiles.
2144    
2145    ----------------------------------------------------------------------
2146    Name: Matthias Blume
2147    Date: 2002/10/10 13:10:00 EDT
2148    Tag: blume-20021010-ppc-divs
2149    Description:
2150    
2151    The mltree operator DIVS must be implemented with an overflow check on
2152    the PPC because the hardware indicates divide-by-zero using "overflow" as
2153    well.
2154    
2155    ----------------------------------------------------------------------
2156    Name: Matthias Blume
2157    Date: 2002/07/23 11:45:00
2158    Tag: blume-20020723-smlnj-home
2159    Description:
2160    
2161    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2162    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2163    CM_PATHCONFIG is also set), the path configuration file is assumed
2164    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2165    the entire tree to some other place and everything will "just work".
2166    
2167    (Companion commands such as ml-build and ml-makedepend also sense this
2168    variable.)
2169    
2170    ----------------------------------------------------------------------
2171    Name: Matthias Blume
2172    Date: 2002/07/12 21:19:00 EDT
2173    Tag: blume-20020712-liveness
2174    Description:
2175    
2176    Exported two useful "step" functions from liveness module (MLRISC).
2177    
2178    ----------------------------------------------------------------------
2179    Name: Matthias Blume
2180    Date: 2002/07/05 16:00 EDT
2181    Tag: Release_110_41
2182    Description:
2183    
2184    New working release.  New bootfiles.
2185    
2186    ----------------------------------------------------------------------
2187    Name: Matthias Blume
2188    Date: 2002/07/05 10:25:00 EDT
2189    Tag: blume-20020705-btimp
2190    Description:
2191    
2192    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2193    other clients can set up backtracing support.
2194    
2195    ----------------------------------------------------------------------
2196    Name: Matthias Blume
2197    Date: 2002/06/25 17:23:00 EDT
2198    Tag: blume-20020625-fpmax
2199    Description:
2200    
2201    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2202    case.  (The sense of the isNaN test was reversed -- which made min and
2203    max always return their first argument.)
2204    
2205    ----------------------------------------------------------------------
2206    Name: Matthias Blume
2207    Date: 2002/06/11
2208    Tag: blume-20020611-unixpath
2209    Description:
2210    
2211    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2212    
2213    ----------------------------------------------------------------------
2214    Name: Matthias Blume
2215    Date: 2002/06/10 16:35:00 EDT
2216    Tag: blume-20020610-ieeereal
2217    Description:
2218    
2219    I back-ported my implementation of IEEEReal.fromString from the
2220    idlbasis-devel branch so that we can test it.
2221    
2222    Another small change is that ppDec tries to give more information
2223    than just "<sig>" in the case of functors.  However, this code is
2224    broken in some mysterious way if the functor's body's signature
2225    has not been declared by ascription but gets inferred from the
2226    implementation.  This needs fixing...
2227    
2228    ----------------------------------------------------------------------
2229    Name: Matthias Blume
2230    Date: 2002/05/31
2231    Tag: blume-20020531-btrace-mode
2232    Description:
2233    
2234    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2235    the wayside when I switched over to using Controls everywhere.)
2236    
2237    ----------------------------------------------------------------------
2238    Name: Lal George
2239    Date: 2002/05/23 12:21:40 EDT
2240    Tag: george-20020523-visual-labels
2241    Description:
2242    
2243    Labels are now displayed in the graphical output to make
2244    the fall-through and target blocks obvious.
2245    
2246    ----------------------------------------------------------------------
2247    Name: Matthias Blume
2248    Date: 2002/05/22 11:03:00 EDT
2249    Tag: blume-20020522-shrink
2250    Description:
2251    
2252    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2253    situation that comes up when allocating huge arrays.
2254    
2255    ----------------------------------------------------------------------
2256    Name: Matthias Blume
2257    Date: 2002/05/21 16:00:00 EDT
2258    Tag: Release_110_40
2259    Description:
2260    
2261    New working release (110.40).  New bootfiles.
2262    
2263    [Also: John Reppy fixed GC bug 1131.]
2264    
2265    ----------------------------------------------------------------------
2266    Name: Matthias Blume
2267    Date: 2002/05/21 12:35:00 EDT
2268    Tag: blume-20020521-cmdoc
2269    Description:
2270    
2271    CM documentation update.
2272    
2273    ----------------------------------------------------------------------
2274    Name: Matthias Blume
2275    Date: 2002/05/21 10:55:00 EDT
2276    Tag: blume-20020521-misc
2277    Description:
2278    
2279    - John tweaked runtime to be silent on heap export (except when
2280      GC messages are on).
2281    - I added a few more things (cross-compiling versions of CMB) to
2282      config/preloads (as suggestions).
2283    
2284    ----------------------------------------------------------------------
2285    Name: Matthias Blume
2286    Date: 2002/05/20 22:25:00 EDT
2287    Tag: blume-20020520-controls
2288    Description:
2289    
2290    - Added ControlUtil structure to control-lib.cm.
2291    - Use it throughout.
2292    - Used Controls facility to define MLRISC controls (as opposed to
2293      registering MLRISC control ref cells with Controls after the fact)
2294    - Fixed messed-up controls priorities.
2295    
2296    * Removed again all the stuff from config/preloads that one wouldn't
2297      be able to preload at the time the initial heap image is built.
2298      (Many libraries, e.g., CML, do not exist yet at this time.  The
2299       only libraries that can be preloaded via config/preloads are those
2300       that come bundled with the bootfiles.)
2301    
2302    ----------------------------------------------------------------------
2303    Name: Matthias Blume
2304    Date: 2002/05/20 10:59:00 EDT
2305    Tag: blume-20020520-preloads
2306    Description:
2307    
2308    Added a lot of commented-out suggestions for things to be included
2309    in config/preloads.
2310    
2311    ----------------------------------------------------------------------
2312    Name: Allen Leung
2313    Date: 2002/05/18 14:20:00 EDT
2314    Tag: leunga-20020518-mdl
2315    Description:
2316    
2317    o Made the mdl tool stuff compile and run again.
2318    o I've disabled all the stuff that depends on RTL specifications; they
2319      are all badly broken anyway.
2320    
2321    ----------------------------------------------------------------------
2322    Name: Matthias Blume
2323    Date: 2002/05/17 16:49:00 EDT
2324    Tag: blume-20020517-controls
2325    Description:
2326    
2327    0. John Reppy made several modifications to the SML/NJ library.
2328       In particular, there is a shiny new controls-lib.cm.
2329    
2330    1. Pushed new controls interface through compiler so that everything
2331       compiles again.
2332    
2333    2. Added FormatComb and FORMAT_COMB to the CML version of the
2334       SML/NJ library (so that CML compiles again).
2335    
2336    3. Modified init scripts because XXX_DEFAULT environment variables
2337       are no longer with us.  (Boot-time initialization is now done
2338       using the same environment variables that are also used for
2339       startup-time initialization of controls.)
2340    
2341    ----------------------------------------------------------------------
2342    Name: Lal George
2343    Date: 2002/05/15 09:20:10 EDT
2344    Tag: george-20020515-pseudo-op-decls
2345    Description:
2346    
2347            All pseudo-ops emitted before the first segment declaration
2348    such as TEXT, DATA, and BSS directives are assumed to be global
2349    declarations and are emitted first in the assembly file. This is
2350    useful in a number of situations where one has pseudo-ops that are not
2351    specific to any segment, and also works around the constraint that one
2352    cannot have client pseudo-ops in the TEXT segment.
2353    
2354            Because no segment is associated with these declarations it is
2355    an error to allocate any space or objects before the first segment
2356    directive and an exception will be raised. However, we cannot make
2357    this check for client pseudo-ops.
2358    
2359            These top level declarations are a field in the CFG graph_info.
2360    In theory you can continue to add to this field after the CFG has been
2361    built -- provided you know what you are doing;-)
2362    
2363    ----------------------------------------------------------------------
2364    Name: Matthias Blume
2365    Date: 2002/05/13 16:40:00 EDT
2366    Tag: blume-20020513-pp-etc
2367    Description:
2368    
2369    A few minor bugfixes:
2370    
2371      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2372        (Bogus printouts for redefined bindings still occur.  Compiler
2373        bug should no longer occur now.  We need to redo the prettyprinter
2374        from scratch.)
2375    
2376      - CM pathname printer now also adds escape sequences for ( and )
2377    
2378      - commend and docu fixes for ml-nlffi
2379    
2380    ----------------------------------------------------------------------
2381    Name: Matthias Blume
2382    Date: 2002/05/10 16:40:00 EDT
2383    Tag: blume-20020510-erg-textio
2384    Description:
2385    
2386    Applied the following bugfix provided by Emden Gansner:
2387    
2388        Output is corrupted when outputSubstr is used rather than output.
2389    
2390        The problem occurs when a substring
2391    
2392            ss = (s, dataStart, dataLen)
2393    
2394        where dataStart > 0, fills a stream buffer with avail bytes left.
2395        avail bytes of s, starting at index dataStart, are copied into the
2396        buffer, the buffer is flushed, and then the remaining dataLen-avail
2397        bytes of ss are copied into the beginning of the buffer. Instead of
2398        starting this copy at index dataStart+avail in s, the current code
2399        starts the copy at index avail.
2400    
2401        Fix:
2402        In text-io-fn.sml, change line 695 from
2403             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2404        to
2405             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2406    
2407    ----------------------------------------------------------------------
2408    Name: Matthias Blume
2409    Date: 2002/04/12 13:55:00 EDT
2410    Tag: blume-20020412-assyntax
2411    Description:
2412    
2413    1. Grabbed newer assyntax.h from the XFree86 project.
2414    2. Fiddled with how to compile X86.prim.asm without warnings.
2415    3. (Very) Minor cleanup in CM.
2416    
2417    ----------------------------------------------------------------------
2418    Name: Matthias Blume
2419    Date: 2002/04/01 (no joke!) 17:07:00 EST
2420    Tag: blume-20020401-x86div
2421    Description:
2422    
2423    Added full support for div/mod/rem/quot on the x86, using the machine
2424    instruction's two results (without clumsily recomputing the remainder)
2425    directly where appropriate.
2426    
2427    Some more extensive power-of-two support was added to the x86 instruction
2428    selector (avoiding expensive divs, mods, and muls where they can be
2429    replaced with cheaper shifts and masks).  However, this sort of thing
2430    ought to be done earlier, e.g., within the CPS optimizer so that
2431    all architectures benefit from it.
2432    
2433    The compiler compiles to a fixed point, but changes might be somewhat
2434    fragile nevertheless.  Please, report any strange things that you might
2435    see wrt. div/mod/quot/rem...
2436    
2437    ----------------------------------------------------------------------
2438    Name: Matthias Blume
2439    Date: 2002/03/29 17:22:00
2440    Tag: blume-20020329-div
2441    Description:
2442    
2443    Fixed my broken div/mod logic.  Unfortunately, this means that the
2444    inline code for div/mod now has one more comparison than before.
2445    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2446    The problem was with quotient = 0, because that alone does not tell
2447    us which way the rounding went.  One then has to look at whether
2448    remainder and divisor have the same sign...  :(
2449    
2450    Anyway, I replaced the bootfiles with fresh ones...
2451    
2452    ----------------------------------------------------------------------
2453    Name: Matthias Blume
2454    Date: 2002/03/29 14:10:00 EST
2455    Tag: blume-20020329-inlprims
2456    Description:
2457    
2458    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2459    
2460    Primops have changed. This means that the bin/boot-file formats have
2461    changed as well.
2462    
2463    To make sure that there is no confusion, I made a new version.
2464    
2465    
2466    CHANGES:
2467    
2468    * removed REMT from mltree (remainder should never overflow).
2469    
2470    * added primops to deal with divisions of all flavors to the frontend
2471    
2472    * handled these primops all the way through so they map to their respective
2473      MLRISC support
2474    
2475    * used these primops in the implementation of Int, Int32, Word, Word32
2476    
2477    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2478    
2479    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2480    
2481    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2482      floating point
2483    
2484    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2485      and Real (but Real.abs maps to a separate floating-point-only primop)
2486    
2487    
2488    TODO items:
2489    
2490    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2491      instructions because the machine instruction encoder cannot handle
2492      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2493      REM in the first place, and REM is supported by the x86 machine coder.)
2494    
2495    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2496      selection phase.  (The two can be streamlined because the hardware
2497      delivers both quotient and remainder at the same time anyway.)
2498    
2499    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2500      (Currently the behavior is inconsistent both across architectures and
2501      wrt. the draft Basis spec.)
2502    
2503    * Word8 should eventually be handled natively, too.
2504    
2505    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2506      as if there currently is no execution path that could trigger it in
2507      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2508      hold for things like multiplication and division.)
2509    
2510    ----------------------------------------------------------------------
2511    Name: Matthias Blume
2512    Date: 2002/03/27 16:27:00 EST
2513    Tag: blume-20020327-mlrisc-divisions
2514    Description:
2515    
2516    Added support for all four division operations (ML's div, mod, quot,
2517    and rem) to MLRISC.  In the course of doing so, I also rationalized
2518    the naming (no more annoying switch-around of DIV and QUOT), by
2519    parameterizing the operation by div_rounding_mode (which can be either
2520    DIV_TO_ZERO or DIV_TO_NEGINF).
2521    
2522    The generic MLTreeGen functor takes care of compiling all four
2523    operations down to only round-to-zero div.
2524    
2525    Missing pieces:
2526    
2527      * Doing something smarter than relying on MLTreeGen on architectures
2528        like, e.g., the x86 where hardware division delivers both quotient and
2529        remainder at the same time.  With this, the implementation of the
2530        round-to-neginf operations could be further streamlined.
2531    
2532      * Remove inlining support for div/mod/rem from the frontend and replace it
2533        with primops that get carried through to the backend.  Do this for all
2534        int and word types.
2535    
2536    ----------------------------------------------------------------------
2537    Name: Matthias Blume
2538    Date: 2002/03/25 17:25:00 EST
2539    Tag: blume-20020325-divmod
2540    Description:
2541    
2542    I improved (hopefully without breaking them) the implementation of Int.div,
2543    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2544    advantage of the following observations:
2545    
2546      Let  q = x quot y      r = x rem y
2547           d = x div  y      m = x mod y
2548    
2549    where "quot" is the round-to-zero version of integer division that
2550    hardware usually provides.  Then we have:
2551    
2552         r = x - q * y        where neither the * nor the - will overflow
2553         d = if q >= 0 orelse x = q * y then q else q - 1
2554                              where neither the * nor the - will overflow
2555         m = if q >= 0 orelse r = 0 then r else r + y
2556                              where the + will not overflow
2557    
2558    This results in substantial simplification of the generated code.
2559    The following table shows the number of CFG nodes and edges generated
2560    for
2561            fun f (x, y) = x OPER y
2562            (* with OPER \in div, mod, quot, rem *)
2563    
2564    
2565        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2566        --------------------------------------------------------
2567         div |         24 |         39 |         12 |         16
2568         mod |         41 |         71 |         12 |         16
2569        quot |          8 |         10 |          8 |         10
2570         rem |         10 |         14 |          8 |         10
2571    
2572    
2573    ----------------------------------------------------------------------
2574    Name: Matthias Blume
2575    Date: 2002/03/25 22:06:00 EST
2576    Tag: blume-20020325-cprotobug
2577    Description:
2578    
2579    Fixed a bug in cproto (c prototype decoder).
2580    
2581    ----------------------------------------------------------------------
2582    Name: Matthias Blume
2583    Date: 2002/03/25 16:00:00 EST
2584    Tag: blume-20020325-raw-primops
2585    Description:
2586    
2587    I did some cleanup to Allen's new primop code and
2588    replaced yesterday's bootfiles with new ones.
2589    (But they are stored in the same place.)
2590    
2591    ----------------------------------------------------------------------
2592    Name: Matthias Blume
2593    Date: 2002/03/24 22:40:00 EST
2594    Tag: blume-20020324-bootfiles
2595    Description:
2596    
2597    Made the bootfiles that Allen asked for.
2598    
2599    ----------------------------------------------------------------------
2600    Name: Allen Leung
2601    Date: 2002/03/23 15:50:00 EST
2602    Tag: leunga-20020323-flint-cps-rcc-primops
2603    Description:
2604    
2605      1. Changes to FLINT primops:
2606    
2607        (* make a call to a C-function;
2608         * The primop carries C function prototype information and specifies
2609         * which of its (ML-) arguments are floating point. C prototype
2610         * information is for use by the backend, ML information is for
2611         * use by the CPS converter. *)
2612      | RAW_CCALL of { c_proto: CTypes.c_proto,
2613                       ml_args: ccall_type list,
2614                       ml_res_opt: ccall_type option,
2615                       reentrant : bool
2616                     } option
2617       (* Allocate uninitialized storage on the heap.
2618        * The record is meant to hold short-lived C objects, i.e., they
2619        * are not ML pointers.  With the tag, the representation is
2620        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2621        *)
2622      | RAW_RECORD of {tag:bool,sz:int}
2623      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2624    
2625      2.  These CPS primops are now overloaded:
2626    
2627           rawload of {kind:numkind}
2628           rawstore of {kind:numkind}
2629    
2630          The one argument form is:
2631    
2632             rawload {kind} address
2633    
2634          The two argument form is:
2635    
2636             rawload {kind} [ml object, byte-offset]
2637    
2638      3. RAW_CCALL/RCC now takes two extra arguments:
2639    
2640         a. The first is whether the C call is reentrant, i.e., whether
2641            ML state should be saved and restored.
2642         b. The second argument is a string argument specifying the name of
2643            library and the C function.
2644    
2645         These things are currently not handled in the code generator, yet.
2646    
2647      4. In CProto,
2648    
2649         An encoding type of "bool" means "ml object" and is mapped into
2650         C prototype of PTR.  Note that "bool" is different than "string",
2651         even though "string" is also mapped into PTR, because "bool"
2652         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2653    
2654      5. Pickler/unpicker
2655    
2656         Changed to handle RAW_RECORD and newest RAW_CCALL
2657    
2658      6. MLRiscGen,
2659    
2660         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2661         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2662            in the file CodeGen/cpscompile/cps-c-calls.sml
2663    
2664      7. Added the conditional move operator
2665    
2666             condmove of branch
2667    
2668         to cps.  Generation of this is still buggy so it is currently
2669         disabled.
2670    
2671    ----------------------------------------------------------------------
2672    Name: Lal George
2673    Date: 2002/03/22 14:18:25 EST
2674    Tag: george-20020322-cps-branch-prob
2675    Description:
2676    
2677    Implemented the Ball-Larus branch prediction-heuristics, and
2678    incorporated graphical viewers for control flow graphs.
2679    
2680    Ball-Larus Heuristics:
2681    ---------------------
2682    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2683    
2684    By design it uses the Dempster-Shafer theory for combining
2685    probabilities.  For example, in the function:
2686    
2687        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2688    
2689    the ball-larus heuristics predicts that the n=0 is unlikely
2690    (OH-heuristic), and the 'then' branch is unlikely because of the
2691    RH-heuristic -- giving the 'then' branch an even lower combined
2692    probability using the Dempster-Shafer theory.
2693    
2694    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2695    probability of the 'then' branch because of the loop in the else
2696    branch.
2697    
2698    
2699    Graphical Viewing:
2700    ------------------
2701    I merely plugged in Allen's graphical viewers into the compiler. The
2702    additional code is not much. At the top level, saying:
2703    
2704            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2705    
2706    will display the graphical view of the control flow graph just before
2707    back-patching.  daVinci must be in your path for this to work. If
2708    daVinci is not available, then the default viewer can be changed
2709    using:
2710    
2711            Control.MLRISC.getString "viewer"
2712    
2713    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2714    course, these viewers must be in your path.
2715    
2716    The above will display the compilation unit at the level of clusters,
2717    many of which are small, boring, and un-interesting. Also setting:
2718    
2719            Control.MLRISC.getInt "cfg-graphical-view_size"
2720    
2721    will display clusters that are larger than the value set by the above.
2722    
2723    
2724    ----------------------------------------------------------------------
2725    Name: Matthias Blume
2726    Date: 2002/03/21 22:20:00 EST
2727    Tag: blume-20020321-kmp-bugfix
2728    Description:
2729    
2730    Changed the interface to the KMP routine in PreString and fixed
2731    a minor bug in one place where it was used.
2732    
2733    ----------------------------------------------------------------------
2734    Name: Allen Leung
2735    Date: 2002/03/21 20:30:00 EST
2736    Tag: leunga-20020321-cfg
2737    Description:
2738    
2739      Fixed a potential problem in cfg edge splitting.
2740    
2741    ----------------------------------------------------------------------
2742    Name: Allen Leung
2743    Date: 2002/03/21 17:15:00 EST
2744    Tag: leunga-20020321-x86-fp-cfg
2745    Description:
2746    
2747      1. Recoded the buggy parts of x86-fp.
2748    
2749         a. All the block reordering code has been removed.
2750            We now depend on the block placement phases to do this work.
2751    
2752         b. Critical edge splitting code has been simplified and moved into the
2753            CFG modules, as where they belong.
2754    
2755         Both of these were quite buggy and complex.  The code is now much, much
2756         simpler.
2757    
2758      2. X86 backend.
2759    
2760         a. Added instructions for 64-bit support.  Instruction selection for
2761            64-bit has not been committed, however, since that
2762            requires changes to MLTREE which haven't been approved by
2763            Lal and John.
2764    
2765         b. Added support for FUCOMI and FUCOMIP when generating code for
2766            PentiumPro and above.  We only generate these instructions in
2767            the fast-fp mode.
2768    
2769         c. Added cases for JP and JNP in X86FreqProps.
2770    
2771      3. CFG
2772    
2773         CFG now has a bunch of methods for edge splitting and merging.
2774    
2775      4. Machine description.
2776    
2777         John's simplification of MLTREE_BASIS.fcond broke a few machine
2778         description things:
2779    
2780         rtl-build.{sig,sml} and hppa.mdl fixed.
2781    
2782         NOTE: the machine description stuff in the repository is still broken.
2783               Again, I can't put my fixes in because that involves
2784               changes to MLTREE.
2785    
2786    ----------------------------------------------------------------------
2787    Name: Matthias Blume
2788    Date: 2002/03/20 15:55:00 EST
2789    Tag: blume-20020320-kmp
2790    Description:
2791    
2792    Implemented Knuth-Morris-Pratt string matching in PreString and used
2793    it for String.isSubstring, Substring.isSubstring, and
2794    Substring.position.
2795    
2796    (Might need some stress-testing.  Simple examples worked fine.)
2797    
2798    ----------------------------------------------------------------------
2799    Name: Matthias Blume
2800    Date: 2002/03/19 16:37:00 EST
2801    Tag: blume-20020319-witnesses
2802    Description:
2803    
2804    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2805    
2806    This implements a generic mechanism for changing constness qualifiers
2807    anywhere within big C types without resorting to outright "casts".
2808    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2809    modify the constness at the outermost level.)
2810    The implementation of "convert" is based on the idea of "witness"
2811    values -- values that are not used by the operation but whose types
2812    "testify" to their applicability.  On the implementation side, "convert"
2813    is simply a projection (returning its second curried argument).  With
2814    cross-module inlining, it should not result in any machine code being
2815    generated.
2816    
2817    ----------------------------------------------------------------------
2818    Name: Matthias Blume
2819    Date: 2002/03/15 16:40:00 EST
2820    Tag: blume-20020315-basis
2821    Description:
2822    
2823    Provided (preliminary?) implementations for
2824    
2825      {String,Substring}.{concatWith,isSuffix,isSubstring}
2826    
2827    and
2828    
2829      Substring.full
2830    
2831    Those are in the Basis spec but they were missing in SML/NJ.
2832    
2833    ----------------------------------------------------------------------
2834    Name: Matthias Blume
2835    Date: 2002/03/14 21:30:00 EST
2836    Tag: blume-20020314-controls
2837    Description:
2838    
2839    Controls:
2840    ---------
2841    
2842    1. Factored out the recently-added Controls : CONTROLS stuff and put
2843       it into its own library $/controls-lib.cm.  The source tree for
2844       this is under src/smlnj-lib/Controls.
2845    
2846    2. Changed the names of types and functions in this interface, so they
2847       make a bit more "sense":
2848    
2849          module -> registry
2850          'a registry -> 'a group
2851    
2852    3. The interface now deals in ref cells only.  The getter/setter interface
2853       is (mostly) gone.
2854    
2855    4. Added a function that lets one register an already-existing ref cell.
2856    
2857    5. Made the corresponding modifications to the rest of the code so that
2858       everything compiles again.
2859    
2860    6. Changed the implementation of Controls.MLRISC back to something closer
2861       to the original.  In particular, this module (and therefore MLRISC)
2862       does not depend on Controls.  There now is some link-time code in
2863       int-sys.sml that registers the MLRISC controls with the Controls
2864       module.
2865    
2866    CM:
2867    ---
2868    
2869      * One can now specify the lambda-split aggressiveness in init.cmi.
2870    
2871    ----------------------------------------------------------------------
2872    Name: Allen Leung
2873    Date: 2002/03/13 17:30:00 EST
2874    Tag: leunga-20020313-x86-fp-unary
2875    Description:
2876    
2877    Bug fix for:
2878    
2879    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2880    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2881    > - fun f(x,(y,z)) = Real.~ y;
2882    > [autoloading]
2883    > [autoloading done]
2884    >       fchsl   (%eax), 184(%esp)
2885    > Error: MLRisc bug: X86MCEmitter.emitInstr
2886    >
2887    > uncaught exception Error
2888    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2889    
2890    The problem was that the code generator did not generate any fp registers
2891    in this case, and the ra didn't know that it needed to run the X86FP phase to
2892    translate the pseudo fp instruction.   This only happened with unary fp
2893    operators in certain situations.
2894    
2895    ----------------------------------------------------------------------
2896    Name: Matthias Blume
2897    Date: 2002/03/13 14:00:00 EST
2898    Tag: blume-20020313-overload-etc
2899    Description:
2900    
2901    1. Added _overload as a synonym for overload for backward compatibility.
2902       (Control.overloadKW must be true for either version to be accepted.)
2903    
2904    2. Fixed bug in install script that caused more things to be installed
2905       than what was requested in config/targets.
2906    
2907    3. Made CM aware of the (_)overload construct so that autoloading
2908       works.
2909    
2910    ----------------------------------------------------------------------
2911    Name: Matthias Blume
2912    Date: 2002/03/12 22:03:00 EST
2913    Tag: blume-20020312-url
2914    Description:
2915    
2916    Forgot to update BOOT and srcarchiveurl.
2917    
2918    ----------------------------------------------------------------------
2919    Name: Matthias Blume
2920    Date: 2002/03/12 17:30:00 EST
2921    Tag: blume-20020312-version110392
2922    Description:
2923    
2924    Yet another version number bump (because of small changes to the
2925    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
2926    
2927    Changes:
2928    
2929      The new pid generation scheme described a few weeks ago was overly
2930      complicated.  I implemented a new mechanism that is simpler and
2931      provides a bit more "stability":  Once CM has seen a compilation
2932      unit, it keeps its identity constant (as long as you do not delete
2933      those crucial CM/GUID/* files).  This means that when you change
2934      an interface, compile, then go back to the old interface, and
2935      compile again, you arrive at the original pid.
2936    
2937      There now also is a mechanism that instructs CM to use the plain
2938      environment hash as a module's pid (effectively making its GUID
2939      the empty string).  For this, "noguid" must be specified as an
2940      option to the .sml file in question within its .cm file.
2941      This is most useful for code that is being generated by tools such
2942      as ml-nlffigen (because during development programmers tend to
2943      erase the tool's entire output directory tree including CM's cached
2944      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
2945      revert to the old, broken behavior of SML/NJ, but in specific cases
2946      where there is no danger of interface confusion, its use is ok
2947      (I think).
2948    
2949      ml-nlffigen by default generates "noguid" annotations.  They can be
2950      turned off by specifying -guid in its command line.
2951    
2952    ----------------------------------------------------------------------
2953    Name: Lal George
2954    Date: 2002/03/12 12 14:42:36 EST
2955    Tag: george-20020312-frequency-computation
2956    Description:
2957    
2958    Integrated jump chaining and static block frequency into the
2959    compiler. More details and numbers later.
2960    
2961    ----------------------------------------------------------------------
2962    Name: Lal George
2963    Date: 2002/03/11 11 22:38:53 EST
2964    Tag: george-20020311-jump-chain-elim
2965    Description:
2966    
2967    Tested the jump chain elimination on all architectures (except the
2968    hppa).  This is on by default right now and is profitable for the
2969    alpha and x86, however, it may not be profitable for the sparc and ppc
2970    when compiling the compiler.
2971    
2972    The gc test will typically jump to a label at the end of the cluster,
2973    where there is another jump to an external cluster containing the actual
2974    code to invoke gc. This is to allow factoring of common gc invocation
2975    sequences. That is to say, we generate:
2976    
2977            f:
2978               testgc
2979               ja   L1      % jump if above to L1
2980    
2981            L1:
2982               jmp L2
2983    
2984    
2985    After jump chain elimination the 'ja L1' instructions is converted to
2986    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
2987    up being implemented in their long form (if L2 is far away) using:
2988    
2989            jbe     L3      % jump if below or equal to L3
2990            jmp     L2
2991         L3:
2992            ...
2993    
2994    
2995    For large compilation units L2  may be far away.
2996    
2997    
2998    ----------------------------------------------------------------------
2999    Name: Matthias Blume
3000    Date: 2002/03/11 13:30:00 EST
3001    Tag: blume-20020311-mltreeeval
3002    Description:
3003    
3004    A functor parameter was missing.
3005    
3006    ----------------------------------------------------------------------
3007    Name: Allen Leung
3008    Date: 2002/03/11 10:30:00 EST
3009    Tag: leunga-20020311-runtime-string0
3010    Description:
3011    
3012       The representation of the empty string now points to a
3013    legal null terminated C string instead of unit.  It is now possible
3014    to convert an ML string into C string with InlineT.CharVector.getData.
3015    This compiles into one single machine instruction.
3016    
3017    ----------------------------------------------------------------------
3018    Name: Allen Leung
3019    Date: 2002/03/10 23:55:00 EST
3020    Tag: leunga-20020310-x86-call
3021    Description:
3022    
3023       Added machine generation for CALL instruction (relative displacement mode)
3024    
3025    ----------------------------------------------------------------------
3026    Name: Matthias Blume
3027    Date: 2002/03/08 16:05:00
3028    Tag: blume-20020308-entrypoints
3029    Description:
3030    
3031    Version number bumped to 110.39.1.  NEW BOOTFILES!
3032    
3033    Entrypoints: non-zero offset into a code object where execution should begin.
3034    
3035    - Added the notion of an entrypoint to CodeObj.
3036    - Added reading/writing of entrypoint info to Binfile.
3037    - Made runtime system bootloader aware of entrypoints.
3038    - Use the address of the label of the first function given to mlriscGen
3039      as the entrypoint.  This address is currently always 0, but it will
3040      not be 0 once we turn on block placement.
3041    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3042      with entry points) from mlriscGen.
3043    
3044    ----------------------------------------------------------------------
3045    Name: Allen Leung
3046    Date: 2002/03/07 20:45:00 EST
3047    Tag: leunga-20020307-x86-cmov
3048    Description:
3049    
3050       Bug fixes for CMOVcc on x86.
3051    
3052       1. Added machine code generation for CMOVcc
3053       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3054       3. CMOVcc cannot have an immediate operand as argument.
3055    
3056    ----------------------------------------------------------------------
3057    Name: Matthias Blume
3058    Date: 2002/03/07 16:15:00 EST
3059    Tag: blume-20020307-controls
3060    Description:
3061    
3062    This is a very large but mostly boring patch which makes (almost)
3063    every tuneable compiler knob (i.e., pretty much everything under
3064    Control.* plus a few other things) configurable via both the command
3065    line and environment variables in the style CM did its configuration
3066    until now.
3067    
3068    Try starting sml with '-h' (or, if you are brave, '-H')
3069    
3070    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3071    implements the underlying generic mechanism.
3072    
3073    The interface to some of the existing such facilities has changed somewhat.
3074    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3075    (The getFoo interface is still there for backward-compatibility, but its
3076    use is deprecated.)
3077    
3078    The ml-build script passes -Cxxx=yyy command-line arguments through so
3079    that one can now twiddle the compiler settings when using this "batch"
3080    compiler.
3081    
3082    TODO items:
3083    
3084    We should go through and throw out all controls that are no longer
3085    connected to anything.  Moreover, we should go through and provide
3086    meaningful (and correct!) documentation strings for those controls
3087    that still are connected.
3088    
3089    Currently, multiple calls to Controls.new are accepted (only the first
3090    has any effect).  Eventually we should make sure that every control
3091    is being made (via Controls.new) exactly once.  Future access can then
3092    be done using Controls.acc.
3093    
3094    Finally, it would probably be a good idea to use the getter-setter
3095    interface to controls rather than ref cells.  For the time being, both
3096    styles are provided by the Controls module, but getter-setter pairs are
3097    better if thread-safety is of any concern because they can be wrapped.
3098    
3099    *****************************************
3100    
3101    One bug fix: The function blockPlacement in three of the MLRISC
3102    backpatch files used to be hard-wired to one of two possibilities at
3103    link time (according to the value of the placementFlag).  But (I
3104    think) it should rather sense the flag every time.
3105    
3106    *****************************************
3107    
3108    Other assorted changes (by other people who did not supply a HISTORY entry):
3109    
3110    1. the cross-module inliner now works much better (Monnier)
3111    2. representation of weights, frequencies, and probabilities in MLRISC
3112       changed in preparation of using those for weighted block placement
3113       (Reppy, George)
3114    
3115    ----------------------------------------------------------------------
3116    Name: Lal George
3117    Date: 2002/03/07 14:44:24 EST 2002
3118    Tag: george-20020307-weighted-block-placement
3119    
3120    Tested the weighted block placement optimization on all architectures
3121    (except the hppa) using AMPL to generate the block and edge frequencies.
3122    Changes were required in the machine properties to correctly
3123    categorize trap instructions. There is an MLRISC flag
3124    "weighted-block-placement" that can be used to enable weighted block
3125    placement, but this will be ineffective without block/edge
3126    frequencies (coming soon).
3127    
3128    
3129    ----------------------------------------------------------------------
3130    Name: Lal George
3131    Date: 2002/03/05 17:24:48 EST
3132    Tag: george-20020305-linkage-cluster
3133    
3134    In order to support the block placement optimization, a new cluster
3135    is generated as the very first cluster (called the linkage cluster).
3136    It contains a single jump to the 'real' entry point for the compilation
3137    unit. Block placement has no effect on the linkage cluster itself, but
3138    all the other clusters  have full freedom in the manner in which they
3139    reorder blocks or functions.
3140    
3141    On the x86 the typical linkage code that is generated is:
3142       ----------------------
3143            .align 2
3144       L0:
3145            addl    $L1-L0, 72(%esp)
3146            jmp     L1
3147    
3148    
3149            .align  2
3150       L1:
3151       ----------------------
3152    
3153    72(%esp) is the memory location for the stdlink register. This
3154    must contain the address of the CPS function being called. In the
3155    above example, it contains the address of  L0; before
3156    calling L1 (the real entry point for the compilation unit), it
3157    must contain the address for L1, and hence
3158    
3159            addl $L1-L0, 72(%esp)
3160    
3161    I have tested this on all architectures except the hppa.The increase
3162    in code size is of course negligible
3163    
3164    ----------------------------------------------------------------------
3165    Name: Allen Leung
3166    Date: 2002/03/03 13:20:00 EST
3167    Tag: leunga-20020303-mlrisc-tools
3168    
3169      Added #[ ... ] expressions to mlrisc tools
3170    
3171    ----------------------------------------------------------------------
3172    Name: Matthias Blume
3173    Date: 2002/02/27 12:29:00 EST
3174    Tag: blume-20020227-cdebug
3175    Description:
3176    
3177    - made types in structure C and C_Debug to be equal
3178    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3179    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3180    
3181    ----------------------------------------------------------------------
3182    Name: Matthias Blume
3183    Date: 2002/02/26 12:00:00 EST
3184    Tag: blume-20020226-ffi
3185    Description:
3186    
3187    1. Fixed a minor bug in CM's "noweb" tool:
3188       If numbering is turned off, then truly don't number (i.e., do not
3189       supply the -L option to noweb).  The previous behavior was to supply
3190       -L'' -- which caused noweb to use the "default" line numbering scheme.
3191       Thanks to Chris Richards for pointing this out (and supplying the fix).
3192    
3193    2. Once again, I reworked some aspects of the FFI:
3194    
3195       A. The incomplete/complete type business:
3196    
3197       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3198         gone!
3199       - ML types representing an incomplete type are now *equal* to
3200         ML types representing their corresponding complete types (just like
3201         in C).  This is still safe because ml-nlffigen will not generate
3202         RTTI for incomplete types, nor will it generate functions that
3203         require access to such RTTI.   But when ML code generated from both
3204         incomplete and complete versions of the C type meet, the ML types
3205         are trivially interoperable.
3206    
3207         NOTE:  These changes restore the full generality of the translation
3208         (which was previously lost when I eliminated functorization)!
3209    
3210       B. Enum types:
3211    
3212       - Structure C now has a type constructor "enum" that is similar to
3213         how the "su" constructor works.  However, "enum" is not a phantom
3214         type because each "T enum" has values (and is isomorphic to
3215         MLRep.Signed.int).
3216       - There are generic access operations for enum objects (using
3217         MLRep.Signed.int).
3218       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3219         * The structure contains the definition of type "mlrep" (the ML-side
3220         representation type of the enum).  Normally, mlrep is the same
3221         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3222         then mlrep will be defined as a datatype -- thus facilitating
3223         pattern matching on mlrep values.
3224         ("-ec" will be suppressed if there are duplicate values in an
3225          enumeration.)
3226         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3227         will be generated for each C enum constant xxx.
3228         * Conversion functions m2i and i2m convert between mlrep and
3229         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3230         * Coversion functions c and ml convert between mlrep and "tag enum".
3231         * Access functions (get/set) fetch and store mlrep values.
3232       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3233         enumerations are merged into one single enumeration represented by
3234         structure E_'.
3235    
3236    ----------------------------------------------------------------------
3237    Name: Allen Leung
3238    Date: 2002/02/25 04:45:00 EST
3239    Tag: leunga-20020225-cps-spill
3240    
3241    This is a new implementation of the CPS spill phase.
3242    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3243    In case of problems, replace it with the old file spill.sml
3244    
3245    The current compiler runs into some serious performance problems when
3246    constructing a large record.  This can happen when we try to compile a
3247    structure with many items.  Even a very simple structure like the following
3248    makes the compiler slow down.
3249    
3250        structure Foo = struct
3251           val x_1 = 0w1 : Word32.int
3252           val x_2 = 0w2 : Word32.int
3253           val x_3 = 0w3 : Word32.int
3254           ...
3255           val x_N = 0wN : Word32.int
3256        end
3257    
3258    The following table shows the compile time, from N=1000 to N=4000,
3259    with the old compiler:
3260    
3261    N
3262    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3263           MLRISC ra                               0.06u  0.00s  0.05g
3264              (spills = 0 reloads = 0)
3265           TOTAL                                   0.63u  0.07s  0.21g
3266    
3267    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3268           MLRISC ra                               5.68u  0.59s  3.93g
3269              (spills = 0 reloads = 0)
3270           TOTAL                                   14.71u  0.99s  4.81g
3271    
3272    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3273           MLRISC ra                               5.54u  0.65s  3.91g
3274              (spills = 543 reloads = 1082)
3275           TOTAL                                   65.40u  3.13s  6.00g
3276    
3277    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3278           MLRISC ra                               0.80u  0.10s  0.55g
3279              (spills = 42 reloads = 84)
3280           TOTAL                                   129.42u  5.10s  4.13g
3281    
3282    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3283           MLRISC ra                               2.69u  0.27s  1.38g
3284              (spills = 62 reloads = 124)
3285           TOTAL                                   682.48u  19.61s  13.99g
3286    
3287    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3288           MLRISC ra                               4.96u  0.27s  2.72g
3289              (spills = 85 reloads = 170)
3290           TOTAL                                   2375.26u  57.21s  48.00g
3291    
3292    As you can see the old cps spill module suffers from some serious
3293    performance problem.  But since I cannot decipher the old code fully,
3294    instead of patching the problems up, I'm reimplementing it
3295    with a different algorithm.  The new code is more modular,
3296    smaller when compiled, and substantially faster
3297    (O(n log n) time and O(n) space).  Timing of the new spill module:
3298    
3299    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3300          MLRISC ra                               0.25u  0.02s  0.15g
3301             (spills=1 reloads=3)
3302          TOTAL                                   7.74u  0.34s  1.62g
3303    
3304    Implementation details:
3305    
3306    As far as I can tell, the purpose of the CPS spill module is to make sure the
3307    number of live variables at any program point (the bandwidth)
3308    does not exceed a certain limit, which is determined by the
3309    size of the spill area.
3310    
3311    When the bandwidth is too large, we decrease the register pressure by
3312    packing live variables into spill records.  How we achieve this is
3313    completely different than what we did in the old code.
3314    
3315    First, there is something about the MLRiscGen code generator
3316    that we should be aware of:
3317    
3318    o MLRiscGen performs code motion!
3319    
3320       In particular, it will move floating point computations and
3321       address computations involving only the heap pointer to
3322       their use sites (if there is only a single use).
3323       What this means is that if we have a CPS record construction
3324       statement
3325    
3326           RECORD(k,vl,w,e)
3327    
3328       we should never count the new record address w as live if w
3329       has only one use (which is often the case).
3330    
3331       We should do something similar to floating point, but the transformation
3332       there is much more complex, so I won't deal with that.
3333    
3334    Secondly, there are now two new cps primops at our disposal:
3335    
3336     1. rawrecord of record_kind option
3337        This pure operator allocates some uninitialized storage from the heap.
3338        There are two forms:
3339    
3340         rawrecord NONE [INT n]  allocates a tagless record of length n
3341         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3342                                     and initializes the tag.
3343    
3344     2. rawupdate of cty
3345          rawupdate cty (v,i,x)
3346          Assigns to x to the ith component of record v.
3347          The storelist is not updated.
3348    
3349    We use these new primops for both spilling and increment record construction.
3350    
3351     1. Spilling.
3352    
3353        This is implemented with a linear scan algorithm (but generalized
3354        to trees).  The algorithm will create a single spill record at the
3355        beginning of the cps function and use rawupdate to spill to it,
3356        and SELECT or SELp to reload from it.  So both spills and reloads
3357        are fine-grain operations.  In contrast, in the old algorithm
3358        "spills" have to be bundled together in records.
3359    
3360        Ideally, we should sink the spill record construction to where
3361        it is needed.  We can even split the spill record into multiple ones
3362        at the places where they are needed.  But CPS is not a good
3363        representation for global code motion, so I'll keep it simple and
3364        am not attempting this.
3365    
3366     2. Incremental record construction (aka record splitting).
3367    
3368        Long records with many component values which are simulatenously live
3369        (recall that single use record addresses are not considered to
3370         be live) are constructed with rawrecord and rawupdate.
3371        We allocate space on the heap with rawrecord first, then gradually
3372        fill it in with rawupdate.  This is the technique suggested to me
3373        by Matthias.
3374    
3375        Some restrictions on when this is applicable:
3376        1. It is not a VECTOR record.  The code generator currently does not handle
3377           this case. VECTOR record uses double indirection like arrays.
3378        2. All the record component values are defined in the same "basic block"
3379           as the record constructor.  This is to prevent speculative
3380           record construction.
3381    
3382    ----------------------------------------------------------------------
3383    Name: Allen Leung
3384    Date: 2002/02/22 01:02:00 EST
3385    Tag: leunga-20020222-mlrisc-tools
3386    
3387    Minor bug fixes in the parser and rewriter
3388    
3389    ----------------------------------------------------------------------
3390    Name: Allen Leung
3391    Date: 2002/02/21 20:20:00 EST
3392    Tag: leunga-20020221-peephole
3393    
3394    Regenerated the peephole files.  Some contained typos in the specification
3395    and some didn't compile because of pretty printing bugs in the old version
3396    of 'nowhere'.
3397    
3398    ----------------------------------------------------------------------
3399    Name: Allen Leung
3400    Date: 2002/02/19 20:20:00 EST
3401    Tag: leunga-20020219-mlrisc-tools
3402    Description:
3403    
3404       Minor bug fixes to the mlrisc-tools library:
3405    
3406       1.  Fixed up parsing colon suffixed keywords
3407       2.  Added the ability to shut the error messages up
3408       3.  Reimplemented the pretty printer and fixed up/improved
3409           the pretty printing of handle and -> types.
3410       4.  Fixed up generation of literal symbols in the nowhere tool.
3411       5.  Added some SML keywords to to sml.sty
3412    
3413    ----------------------------------------------------------------------
3414    Name: Matthias Blume
3415    Date: 2002/02/19 16:20:00 EST
3416    Tag: blume-20020219-cmffi
3417    Description:
3418    
3419    A wild mix of changes, some minor, some major:
3420    
3421    * All C FFI-related libraries are now anchored under $c:
3422        $/c.cm      --> $c/c.cm
3423        $/c-int.cm  --> $c/internals/c-int.cm
3424        $/memory.cm --> $c/memory/memory.cm
3425    
3426    * "make" tool (in CM) now treats its argument pathname slightly
3427      differently:
3428        1. If the native expansion is an absolute name, then before invoking
3429           the "make" command on it, CM will apply OS.Path.mkRelative
3430           (with relativeTo = OS.FileSys.getDir()) to it.
3431        2. The argument will be passed through to subsequent phases of CM
3432           processing without "going native".  In particular, if the argument
3433           was an anchored path, then "make" will not lose track of that anchor.
3434    
3435    * Compiler backends now "know" their respective C calling conventions
3436      instead of having to be told about it by ml-nlffigen.  This relieves
3437      ml-nlffigen from one of its burdens.
3438    
3439    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3440    
3441    * Export C_DEBUG and C_Debug from $c/c.cm.
3442    
3443    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3444      subtyping relationship between incomplete pointers and their complete
3445      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3446      with the convention of instantiating 'o with ('t, 'c) obj whenever
3447      the pointer target type is complete.  In the incomplete case, 'o
3448      will be instantiated with some "'c iobj" -- a type obtained by
3449      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3450    
3451      Operations that work on both incomplete and complete pointer types are
3452      typed as taking an 'o ptr while operations that require the target to
3453      be known are typed as taking some ('t, 'c) obj ptr.
3454    
3455      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3456      where void is an eqtype without any values.  This makes it possible
3457      to work on voidptr values using functions meant to operate on light
3458      incomplete pointers.
3459    
3460    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3461      been vastly simplified.
3462    
3463    ----------------------------------------------------------------------
3464    Name: Matthias Blume
3465    Date: 2002/02/19 10:48:00 EST
3466    Tag: blume-20020219-pqfix
3467    Description:
3468    
3469    Applied Chris Okasaki's bug fix for priority queues.
3470    
3471    ----------------------------------------------------------------------
3472    Name: Matthias Blume
3473    Date: 2002/02/15 17:05:00
3474    Tag: Release_110_39
3475    Description:
3476    
3477    Last-minute retagging is becoming a tradition... :-(
3478    
3479    This is the working release 110.39.
3480    
3481    ----------------------------------------------------------------------
3482    Name: Matthias Blume
3483    Date: 2002/02/15 16:00:00 EST
3484    Tag: Release_110_39-orig
3485    Description:
3486    
3487    Working release 110.39.  New bootfiles.
3488    
3489    (Update: There was a small bug in the installer so it wouldn't work
3490    with all shells.  So I retagged. -Matthias)
3491    
3492    ----------------------------------------------------------------------
3493    Name: Matthias Blume
3494    Date: 2002/02/15 14:17:00 EST
3495    Tag: blume-20020215-showbindings
3496    Description:
3497    
3498    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3499    the latter can be useful for exploring what bindings are available at
3500    the interactive prompt.  (The first function returns only the list
3501    of symbols that are really bound, the second prints those but also the
3502    ones that CM's autoloading mechanism knows about.)
3503    
3504    ----------------------------------------------------------------------
3505    Name: Matthias Blume
3506    Date: 2002/02/15 12:08:00 EST
3507    Tag: blume-20020215-iptrs
3508    Description:
3509    
3510    Two improvements to ml-nlffigen:
3511    
3512      1. Write files only if they do not exist or if their current contents
3513         do not coincide with what's being written.  (That is, avoid messing
3514         with the time stamps unless absolutely necessary.)
3515    
3516      2. Implement a "repository" mechanism for generated files related
3517         to "incomplete pointer types".   See the README file for details.
3518    
3519    ----------------------------------------------------------------------
3520    Name: Matthias Blume
3521    Date: 2002/02/14 11:50:00 EST
3522    Tag: blume-20020214-quote
3523    Description:
3524    
3525    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3526    because of the new and improved tag generation scheme.  (Thanks to Allen
3527    Leung for pointing it out.)
3528    
3529    ----------------------------------------------------------------------
3530    Name: Lal George
3531    Date: 2002/02/14 09:55:27 EST 2002
3532    Tag: george-20020214-isabelle-bug
3533    Description:
3534    
3535    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3536    of Isabelle on the x86.
3537    
3538    From Allen:
3539    -----------
3540     I've found the problem:
3541    
3542         in ra-core.sml, I use the counter "blocked" to keep track of the
3543         true number of elements in the freeze queue.  When the counter goes
3544         to zero, I skip examining the queue.  But I've messed up the
3545         bookkeeping in combine():
3546    
3547             else ();
3548             case !ucol of
3549               PSEUDO => (if !cntv > 0 then
3550                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3551                                        ^^^^^^^^^^^^^^^^^^^^^^^
3552                      moveu := mergeMoveList(!movev, !moveu)
3553                     )
3554                  else ();
3555    
3556         combine() is called to coalesce two nodes u and v.
3557         I think I was thinking that if the move counts of u and v are both
3558         greater than zero then after they are coalesced then one node is
3559         removed from the freeze queue.  Apparently I was thinking that
3560         both u and v are of low degree, but that's clearly not necessarily true.
3561    
3562    
3563    02/12/2002:
3564        Here's the patch.  HOL now compiles.
3565    
3566        I don't know how this impact on performance (compile
3567        time or runtime).  This bug caused the RA (especially on the x86)
3568        to go thru the potential spill phase when there are still nodes on the
3569        freeze queue.
3570    
3571    
3572    
3573    
3574    ----------------------------------------------------------------------
3575    Name: Matthias Blume
3576    Date: 2002/02/13 22:40:00 EST
3577    Tag: blume-20020213-fptr-rtti
3578    Description:
3579    
3580    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3581    updates.
3582    
3583    ----------------------------------------------------------------------
3584    Name: Matthias Blume
3585    Date: 2002/02/13 16:41:00 EST
3586    Tag: blume-20020213-cmlpq
3587    Description:
3588    
3589    Added new priority queue export symbols (which have just been added to
3590    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3591    not compile and the installer would choke.)
3592    
3593    ----------------------------------------------------------------------
3594    Name: Matthias Blume
3595    Date: 2002/02/13 16:15:00 EST
3596    Tag: blume-20020213-various
3597    Description:
3598    
3599    1. More tweaks to ml-nlffigen:
3600    
3601       - better internal datastructures (resulting in slight speedup)
3602       - "-match" option requires exact match
3603       - "localized" gensym counters (untagged structs/unions nested within
3604         other structs/unions or within typedefs get a fresh counter; their
3605         tag will be prefixed by a concatenation of their parents' tags)
3606       - bug fixes (related to calculation of transitive closure of types
3607         to be included in the output)
3608    
3609    2. Minor Basis updates:
3610    
3611       - added implementations for List.collate and Option.app
3612    
3613    ----------------------------------------------------------------------
3614    Name: Matthias Blume
3615    Date: 2002/02/11 15:55:00 EST
3616    Tag: blume-20020211-gensym
3617    Description:
3618    
3619    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3620    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3621    names (ML structure names that correspond to unnamed C structs, unions,
3622    and enums), so that separate runs of ml-nlffigen do not clash.
3623    
3624    ----------------------------------------------------------------------
3625    Name: Matthias Blume
3626    Date: 2002/02/11 12:05:00 EST
3627    Tag: blume-20020211-gensml
3628    Description:
3629    
3630    A quick fix for a problem with GenSML (in the pgraph-util library):
3631    Make generation of toplevel "local" optional.  (Strictly speaking,
3632    signature definitions within "local" are not legal SML.)
3633    
3634    Other than that: updates to INSTALL and cm/TODO.
3635    
3636    ----------------------------------------------------------------------
3637    Name: Matthias Blume
3638    Date: 2002/02/08 15:00:00 EST
3639    Tag: blume-20020208-uniquepid
3640    Description:
3641    
3642    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3643    
3644    1. The installer (config/install.sh) has gotten smarter:
3645    
3646         - Configuration options are a bit easier to specify now
3647           (in config/targets).
3648         - Bug in recognizing .tar.bz2 files fixed.
3649         - Installer automatically resolves dependencies between
3650           configuration options (e.g., if you ask for eXene, you will
3651           also get cml -- regardless whether you asked for it or not).
3652         - Installer can run in "quieter mode" by setting the environment
3653           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3654           "completely silent", though.
3655         - Build HashCons library as part of smlnj-lib.
3656    
3657    2. A new scheme for assigning persistent identifiers to compilation
3658       units (and, by extension, to types etc.) has been put into place.
3659       This fixes a long-standing bug where types and even dynamic values
3660       can get internally confused, thereby compromising type safety
3661       (abstraction) and dynamic correctness.  See
3662    
3663         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3664    
3665       for an example of how things could go wrong until now.
3666    
3667       The downside of the new scheme is that pids are not quite as
3668       persistent as they used to be: CM will generate a fresh pid
3669       for every compilation unit that it thinks it sees for the first
3670       time.  That means that if you compile starting from a clean, fresh
3671       source tree at two different times, you end up with different
3672       binaries.
3673    
3674       Cutoff recompilation, however, has not been compromised because
3675       CM keeps pid information in special caches between runs.
3676    
3677    ----------------------------------------------------------------------
3678    Name: Lal George
3679    Date: 2002/02/07 15:34:13 EST 2002
3680    Tag: <none>
3681    Description:
3682    
3683    Compilers that generate assembly code may produce  global labels
3684    whose value is resolved at link time. The various peephole optimization
3685    modules did not take this in account.
3686    
3687    TODO. The Labels.addrOf function should really return an option
3688    type so that clients are forced to deal with this issue, rather
3689    than an exception being raised.
3690    
3691    ----------------------------------------------------------------------
3692    Name: Lal George
3693    Date: 2002/02/06 13:55:02 EST
3694    Tag: george-20020206-ra-breakup
3695    Description:
3696    
3697    1. A bug fix from Allen.
3698    
3699        A typo causes extra fstp %st(0)'s to be generated at compensation
3700        edges, which might cause stack underflow traps at runtime.  This
3701        occurs in fft where there are extraneous fstps right before the 'into'
3702        trap instruction (in this case they are harmless since none of the
3703        integers overflow.)
3704    
3705    2. Pulled out various utility modules that were embedded in the modules
3706       of the register allocator. I need these modules for other purposes, but
3707       they are not complete enough to put into a library (just yet).
3708    ----------------------------------------------------------------------
3709    Name: Matthias Blume
3710    Date: 2002/01/31 16:05:00 EST
3711    Tag: blume-20020131-sparc-ccalls
3712    Description:
3713    
3714    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3715       of extra stack space by mistake.  Fixed.
3716    
3717    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3718    
3719    ----------------------------------------------------------------------
3720    Name: Allen Leung
3721    Date: 2002/01/30
3722    Tag: leunga-20020130-nowhere-bug-fix
3723    Description:
3724    
3725       MLRISC bug fixes:
3726       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3727       2. MachineInt.fromString was negating its value.
3728    
3729    ----------------------------------------------------------------------
3730    Name: Matthias Blume
3731    Date: 2002/01/29
3732    Tag: blume-20020129-INSTALL
3733    Description:
3734    
3735    - Added somewhat detailed installation instructions (file INSTALL).
3736    - Fixed curl-detection bug in config/install.sh.
3737    - It is now possible to select the URL getter using the URLGETTER
3738      environment variable:
3739    
3740          not set / "unknown"      --> automatic detection (script tries wget,
3741                                       curl, and lynx)
3742          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3743                                       how to properly invoke them)
3744          other                    --> use $URLGETTER directly, it must take
3745                                       precisely two command-line arguments
3746                                       (source URL and destination file name)
3747    
3748    ----------------------------------------------------------------------
3749    Name: Matthias Blume
3750    Date: 2002/01/28
3751    Tag: blume-20020128-sparc-ccalls
3752    Description:
3753    
3754    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3755    - Make use of the allocParam argument in sparc-c-calls.
3756    
3757    ----------------------------------------------------------------------
3758    Name: Matthias Blume
3759    Date: 2002/01/28
3760    Tag: blume-20020128-allocParam
3761    Description:
3762    
3763    John Reppy:  Changes c-calls API to accept client-callback for
3764    allocating extra stack space.
3765    me: Corresponding changes to mlriscGen (using a dummy argument that
3766        does not change the current behavior).
3767    
3768    ----------------------------------------------------------------------
3769    Name: Matthias Blume
3770    Date: 2002/01/28 12:00:00
3771    Tag: Release_110_38
3772    Description:
3773    
3774    This time for real!!!
3775    
3776    ----------------------------------------------------------------------
3777    Name: Matthias Blume
3778    Date: 2002/01/28 10:56:00 EST
3779    Tag: blume-20020128-retraction
3780    Description:
3781    
3782    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
3783       with blume-Release_110_38-retracted.)
3784    
3785    1. Fixed a problem with incorrect rounding modes in real64.sml.
3786       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
3787    
3788    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
3789       and enums fixed.  The naming of corresponding ML identifiers should
3790       now be consistent again.
3791    
3792    ----------------------------------------------------------------------
3793    Name: Allen Leung
3794    Date: 2002/01/27
3795    Tag: leunga-20020127-nowhere
3796    Description:
3797    
3798       Added a target called nowhere in the configuration scripts.
3799       Enabling this will build the MLRISC 'nowhere' tool (for translating
3800       programs with where-clauses into legal SML code) during installation.
3801    
3802    ----------------------------------------------------------------------
3803    Name: Matthias Blume
3804    Date: 2002/01/25 21:27:00 EST
3805    Tag: blume-Release_110_38-retracted
3806    Description:
3807    
3808    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
3809    
3810    README will be added later.
3811    
3812    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
3813    (Release_110_38) removed.  Reason: Last-minute bug fixes.
3814    
3815    ----------------------------------------------------------------------
3816    Name: Matthias Blume
3817    Date: 2002/01/25
3818    Tag: blume-20020125-ffi
3819    Description:
3820    
3821    A large number of tweaks and improvements to ml-nlffi-lib and
3822    ml-nlffigen:
3823    
3824       - ML represenation types have been streamlined
3825       - getter and setter functions work with concrete values, not abstract
3826         ones where possible
3827       - ml-nlffigen command line more flexible (see README file there)
3828       - some bugs have been fixed (hopefully)
3829    
3830    ----------------------------------------------------------------------
3831    Name: Lal George
3832    Date: 2002/01/24
3833    Tag: george-20020124-risc-ra-interface
3834    Description:
3835    
3836       There is a dramatic simplification in the interface to the
3837       register allocator for RISC architectures as a result of making
3838       parallel copy instructions explicit.
3839    
3840    ----------------------------------------------------------------------
3841    Name: Matthias Blume
3842    Date: 2002/01/22
3843    Tag: blume-20020122-x86-ccalls
3844    Description:
3845    
3846    Bug fix for c-calls on x86 (having to do with how char- and
3847    short-arguments are being handled).
3848    
3849    ----------------------------------------------------------------------
3850    Name: Matthias Blume
3851    Date: 2002/01/21
3852    Tag: blume-20020121-ff
3853    Description:
3854    
3855