Home My Page Projects Code Snippets Project Openings SML/NJ
Summary Activity Forums Tracker Lists Tasks Docs Surveys News SCM Files

SCM Repository

[smlnj] Diff of /sml/trunk/HISTORY
ViewVC logotype

Diff of /sml/trunk/HISTORY

Parent Directory Parent Directory | Revision Log Revision Log | View Patch Patch

revision 972, Sun Oct 28 00:37:02 2001 UTC revision 1785, Wed May 4 15:51:20 2005 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15  ----------------------------------------------------------------------  ----------------------------------------------------------------------
16    Name: Allen Leung (leunga (at) reservoir (dot) com)
17    Date: 2005/05/04 11:50:00 EST
18    Tag: leunga-20050504-checkgc
19    Description:
20    
21    Added a gc protocol checking phase.  This phase is enabled with
22    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
23    
24    ----------------------------------------------------------------------
25    Name: Matthias Blume (blume (at) tti - c (dot) org)
26    Date: 2005/05/04 10:45:00 CDT
27    Tag: blume-20050504-intinf
28    Description:
29    
30    Fixed a bug in the implementation of div and mod for IntInf.
31    Thanks to Neophytos Michael for reporting the problem.
32    
33    ----------------------------------------------------------------------
34    Name: Matthias Blume (blume (at) tti - c (dot) org)
35    Date: 2005/05/04 10:35:00 CDT
36    Tag: blume-20050504-join
37    Description:
38    
39    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
40    
41    ----------------------------------------------------------------------
42    Name: Matthias Blume (blume (at) tti - c (dot) org)
43    Date: 2005/02/28 23:40:00 CST
44    Tag: blume-20050228-mVar
45    Description:
46    
47    Fixed serious bug (brown paper bag variety) in new implementation of
48    structure Atom in CML.  (I had accidentally used a mailbox instead of
49    an mvar, leaving the door open for races.)
50    
51    ----------------------------------------------------------------------
52    Name: Matthias Blume (blume (at) tti - c (dot) org)
53    Date: 2005/02/25 15:00:00 CST
54    Tag: Release_110_53
55    Description:
56    
57    New working version (110.53).  NEW BOOTFILES!
58    
59    ----------------------------------------------------------------------
60    Name: Matthias Blume (blume (at) tti - c (dot) org)
61    Date: 2005/02/25 14:50:00 CST
62    Tag: blume-20050225-susp
63    Description:
64    
65    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
66    implemented in Core -- which means that it is the same as the one used
67    by the lazy extension.
68    
69    ----------------------------------------------------------------------
70    Name: Matthias Blume (blume (at) tti - c (dot) org)
71    Date: 2005/02/24 16:50:00 CST
72    Tag: blume-20050224-cml-atom
73    Description:
74    
75    Simpler and at the same time more general implementation of structure
76    Atom in CML.
77    
78    ----------------------------------------------------------------------
79    Name: Matthias Blume (blume (at) tti - c (dot) org)
80    Date: 2005/02/15 17:35:00 CST
81    Tag: blume-20050215-tools
82    Description:
83    
84    Created new "tools" directory under "src" and moved "TraceDebugProf"
85    there.
86    
87    ----------------------------------------------------------------------
88    Name: Matthias Blume (blume (at) tti - c (dot) org)
89    Date: 2005/02/10 17:55:00 CST
90    Tag: blume-20050210-longlong
91    Description:
92    
93    Implemented "long long" arguments and results for NLFFI.  (Only the
94    PPC/MacOS implementation is complete, the other backends still need to
95    be updated.)
96    
97    ----------------------------------------------------------------------
98    Name: Matthias Blume (blume (at) tti - c (dot) org)
99    Date: 2005/01/24 17:40:00 CST
100    Tag: blume-20050124-mlyacc
101    Description:
102    
103    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
104    problem with certain "as" patterns which previously got rendered
105    using incorrect syntax.
106    
107    ----------------------------------------------------------------------
108    Name: Matthias Blume (blume (at) tti - c (dot) org)
109    Date: 2005/01/18 12:00:00 CST
110    Tag: blume-20050118-profile
111    Description:
112    
113    Made time profiling code (interrupt handler) in runtime system aware
114    of new array representation.
115    
116    ----------------------------------------------------------------------
117    Name: Matthias Blume (blume (at) tti - c (dot) org)
118    Date: 2005/01/14 18:00:00 CST
119    Tag: blume-20050114-heap2exec
120    Description:
121    
122    Implemented new (but still experimental) heap2exec facility.  This is
123    tested under Mac OS X and should work under Linux (will test shortly).
124    It will probably also work on the Sparc (will test some time later).
125    
126      - removed old "HACKED_STANDALONE" hack from runtime
127    
128    To be able to test this, uncomment the request for "heap2asm" in
129    config/targets prior to installation.  (Notice that this is different
130    from "heap2exec" mentioned below.  Not a typo.)
131    
132    To perform an actual test, run the command
133    
134       $ bin/heap2exec heapfile execfile
135    
136    (You can put heap2exec on your shell's path.)
137    
138    For example, run
139    
140      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
141    
142    This will create a standalone executable called "mly" which you can
143    then invoke directly as a command.
144    
145    ----------------------------------------------------------------------
146    Name: Matthias Blume (blume (at) tti - c (dot) org)
147    Date: 2005/01/07 11:44:00 CST
148    Tag: blume-20050107-mlstring
149    Description:
150    
151    fixed off-by-one error in ML_STRING macro (globals.c)
152    
153    ----------------------------------------------------------------------
154    Name: Matthias Blume (blume (at) tti - c (dot) org)
155    Date: 2004/12/23 18:00:00 CST
156    Tag: blume-20041223-santa
157    Description:
158    
159    Made ml-build script "smarter" (but only very little).
160    
161    ----------------------------------------------------------------------
162    Name: Matthias Blume (blume (at) tti - c (dot) org)
163    Date: 2004/12/21 15:05:00 CST
164    Tag: blume-20041221-longlong
165    Description:
166    
167    * Implemented access to signed and unsigned long long data in NLFFI.
168    (The parameter-passing part of the picture has not complete. But
169    data structure access seems to work.)
170    
171    * Fixed CM's incorrect assumption that the PPC is little-endian.
172      (On the Mac, it is big-endian.  And that's currently our only
173      PPC platform.)
174    
175    ----------------------------------------------------------------------
176    Name: Matthias Blume (blume (at) tti - c (dot) org)
177    Date: 2004/12/21 12:50:00 CST
178    Tag: blume-20041221-memory
179    Description:
180    
181    Some cleanup in the $c/memory.cm library: separated some concerns by
182    moving allocation code and memory access code each into their own
183    files.
184    
185    ----------------------------------------------------------------------
186    Name: Allen Leung (leunga (at) reservoir (dot) com)
187    Date: 2004/12/17 16:12:00 EST
188    Tag: leunga-20041217-cygwin-smlnj-home
189    Description:
190    
191        The Unix I/O library of SML/NJ on cygwin does not understand
192    Windows style pathname, so problems arise when SMLNJ_HOME is set
193    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
194    to a POSIX pathname on cygwin.
195    
196    ----------------------------------------------------------------------
197    Name: Matthias Blume (blume (at) tti - c (dot) org)
198    Date: 2004/12/16 13:35:00 CST
199    Tag: Release_110_52
200    Description:
201    
202    Last-minute changes incorporated into 110.52.  Release tag moved.
203    
204    The changes:
205       - HashString.hashString' -> HashString.hashSubstring
206       - bug fix in UnivariateStats
207    
208    ----------------------------------------------------------------------
209    Name: Matthias Blume (blume (at) tti - c (dot) org)
210    Date: 2004/12/15 23:40:00 CST
211    Tag: blume-20041215-hashSubstring
212    Description:
213    
214    - HashString.hashString' -> HashString.hashSubstring
215    - corresponding changes in atom.sml
216    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
217      added some comments
218    
219    ----------------------------------------------------------------------
220    Name: Matthias Blume (blume (at) tti - c (dot) org)
221    Date: 2004/12/15 15:30:00 CST
222    Tag: (Release_110_52)
223    Description:
224    
225    New working version (110.52).  NEW BOOTFILES!
226    
227    ----------------------------------------------------------------------
228    Name: Matthias Blume (blume (at) tti - c (dot) org)
229    Date: 2004/12/15 12:45:00 CST
230    Tag: blume-20041215-spaces
231    Description:
232    
233    More on the space problem (this time for Win32).
234    
235    ----------------------------------------------------------------------
236    Name: Matthias Blume (blume (at) tti - c (dot) org)
237    Date: 2004/12/14 17:30:00 CST
238    Tag: blume-20041214-spaces
239    Description:
240    
241    Hacked some of the scripts (in particular: the installer) to cope with
242    spaces in filenames a bit better.  But beware: the current "solution"
243    is likely still full of bugs and inherently incomplete.  (We need to
244    do away with those shell scripts for a comprehensive solution.)
245    
246    ----------------------------------------------------------------------
247    Name: Matthias Blume (blume (at) tti - c (dot) org)
248    Date: 2004/12/13 14:45:00 CST
249    Tag: blume-20041213-ml-makedepend
250    Description:
251    
252    Fixed bug in code for ml-makedepend.
253    
254    ----------------------------------------------------------------------
255    Name: Matthias Blume (blume (at) tti - c (dot) org)
256    Date: 2004/12/09 16:30:00 CST
257    Tag: blume-20041209-statistics
258    Description:
259    
260    Added two simple but potentially useful statistics modules to SML/NJ Library.
261    (See CHANGES file there.)
262    
263    ----------------------------------------------------------------------
264    Name: Matthias Blume (blume (at) tti - c (dot) org)
265    Date: 2004/12/01 16:50:00 CST
266    Tag: blume-20041201-atom
267    Description:
268    
269    smlnj-lib:
270    
271      Added function HashString.hashString' for substrings.
272      Hand-inlined CharVector.foldl into HashString (for speed).
273      Modified implementation of structure Atom to avoid extracting
274      strings from substrings unless necessary.
275      (Also see CHANGES file for smlnj-lib.)
276    
277    ----------------------------------------------------------------------
278    Name: Matthias Blume (blume (at) tti - c (dot) org)
279    Date: 2004/11/24 22:15:00 CST
280    Tag: blume-20041124-cml
281    Description:
282    
283    Made sure CML compiles when Position = Int64.
284    
285    ----------------------------------------------------------------------
286    Name: Matthias Blume (blume (at) tti - c (dot) org)
287    Date: 2004/11/24 14:30:00 CST
288    Tag: blume-20041124-position
289    Description:
290    
291    The compiler can now be compiled in a mode that makes structure Position
292    equal to Int64.  The default, however, is unchanged (Position = Int31)
293    for the time being.
294    
295    To enable 64-bit positions, use the following procedure:
296    
297    1. Start sml
298    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
299    3. Type
300          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
301    4. Run CMB.make() as usual.
302    
303    This is barely tested.  The only test so far was a little SML program
304    counting the number of characters in an 8-gigabyte file by
305    reading it character-by-character.  That test was successful.
306    
307    In support of 64-bit positions, a number of new functions have been
308    added to the runtime system.
309    
310    ----------------------------------------------------------------------
311    Name: Matthias Blume (blume (at) tti - c (dot) org)
312    Date: 2004/11/23 14:45:00 CST
313    Tag: blume-20041123-useFile
314    Description:
315    
316    Fixed a problem with unhelpful error messages related to problems with
317    .cm- or .sml files that appear as part of the sml command line.
318    
319    ----------------------------------------------------------------------
320    Name: Matthias Blume (blume (at) tti - c (dot) org)
321    Date: 2004/11/18 15:40:00 CST
322    Tag: Release_110_51
323    Description:
324    
325    New working version (110.51).  NEW BOOTFILES!
326    
327    ----------------------------------------------------------------------
328    Name: Matthias Blume (blume (at) tti - c (dot) org)
329    Date: 2004/11/18 15:35:00 CST
330    Tag: <none>
331    Description:
332    
333    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
334    Joemann).
335    
336    ----------------------------------------------------------------------
337    Name: Allen Leung (leunga (at) reservoir (dot) com)
338    Date: 2004/11/17 16:05:21 EST 2004
339    Tag: leunga-20041117-mlrisc-live-kill
340    Description:
341    
342       Added support for MLTree constructs LIVE and KILL
343    to all the architectures.
344    
345    ----------------------------------------------------------------------
346    Name: Matthias Blume (blume (at) tti - c (dot) org)
347    Date: 2004/11/13 00:20:00 CST
348    Tag: blume-20041113-versiontool
349    Description:
350    
351    - Stripped down the versiontool: It now only handles the version number.
352      The date string is generated at bootstrap time (during makeml).
353    
354    - In a previous commit, fixed a minor issue with how polyequal is being
355      translated.  In particular, the code now "looks through" abstractions.
356      This results in slightly fewer polyEqual warnings and hopefully slightly
357      more efficient code.  Important examples for where this matters are
358      the new int64 and word64 types.
359    
360    ----------------------------------------------------------------------
361    Name: Matthias Blume (blume (at) tti - c (dot) org)
362    Date: 2004/11/12 00:30:00 CST
363    Tag: blume-20041112-int64
364    Description:
365    
366    Structure Int64 fully hooked in.  (The implementation is not very
367    efficient, though.)
368    
369    ----------------------------------------------------------------------
370    Name: Matthias Blume (blume (at) tti - c (dot) org)
371    Date: 2004/11/11 17:30:00 CST
372    Tag: blume-20041111-more64
373    Description:
374    
375    All the pieces of Word64 are now there, with the exception of the
376    conversions from and to LargeWord. (Eventually these need to be identities,
377    but for the time being they don't even make sense because LargeWord is
378    32-bit wide.)
379    
380    Also started to add similar support for Int64, but major pieces of that
381    are still missing.
382    
383    ----------------------------------------------------------------------
384    Name: Matthias Blume (blume (at) tti - c (dot) org)
385    Date: 2004/11/11 00:15:00 CST
386    Tag: blume-20041111-word64
387    Description:
388    
389    Structure Word64 is now (almost) complete, word literals and patterns
390    seem to work.  There are a few odd pieces missing.  In particular,
391    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
392    at the moment.
393    
394    Making Word64 official would mean that LargeWord becomes Word64.  But
395    this requires extreme care because most word-word conversions have to
396    go through LargeWord, so making a mistake means loss of efficiency or
397    worse.  Eventually there will be a solution similar to (but actually
398    simpler than) what I did with IntInf.
399    
400    ----------------------------------------------------------------------
401    Name: Matthias Blume (blume (at) tti - c (dot) org)
402    Date: 2004/11/10 18:12:00 CST
403    Tag: blume-20041110-64bit
404    Description:
405    
406    More 64-bit hacking (but still not even half-way there yet).
407    Also, some assorted improvements to the handling of 8-bit words.
408    
409    ----------------------------------------------------------------------
410    Name: Matthias Blume (blume (at) tti - c (dot) org)
411    Date: 2004/11/09 17:50:00 CST
412    Tag: <none>
413    Description:
414    
415    Started adding some infrastructure for supporting 64-bit int- and
416    word-types.  (Still in its very early stages.)
417    
418    ----------------------------------------------------------------------
419    Name: Matthias Blume (blume (at) tti - c (dot) org)
420    Date: 2004/10/28 10:45:00 CDT
421    Tag: Release_110_50
422    Description:
423    
424    New working version (110.50).  NEW BOOTFILES!
425    
426                 =====================
427    
428    Also:
429    
430      - Changend config/srcarchiveurl from a file just
431        containing the URL string into a file containing
432        shell script code.  The code has access to the $VERSION variable.
433      - Made corresponding changes to config/install.sh and config/unpack.
434      - Default contents of config/srcarchiveurl uses $VERSION and
435        normally does not have to be edited to reflect a version change.
436    
437      (As a result, a version change can be done by just editing
438       config/version, the rest is now automatic.)
439    
440    ----------------------------------------------------------------------
441    Name: Matthias Blume (blume (at) tti - c (dot) org)
442    Date: 2004/10/27 17:50:00 CDT
443    Tag: blume-20041027-btrace-msg
444    Description:
445    
446    BackTrace.monitor now also reports the source of the exception that
447    triggered the trace.
448    
449    ----------------------------------------------------------------------
450    Name: Matthias Blume (blume (at) tti - c (dot) org)
451    Date: 2004/10/27 17:20:00 CDT
452    Tag: blume-20041027-x86-c-calls
453    Description:
454    
455    This is the HISTORY entry for two earlier commits, both concerning
456    the x86 c-calls code in MLRISC:
457    
458      - added a missing LOAD in the code that deals with struct arguments
459      - made sure the caller does not add the wrong number of bytes to the
460        stack pointer after a call of a function returning a struct
461        (the callee already pops the implicit argument which points to
462         the space reserved for the result)
463    
464    ----------------------------------------------------------------------
465    Name: Allen Leung (leunga (at) reservoir (dot) com)
466    Date: 2004/10/24 14:00:00 EST
467    Tag: leunga-20041024-x86-gas-fucomip
468    Description:
469    
470       John discovered a bug in the syntax of fucomip.
471       The opcodes FU?COMIP? have been changed to
472    
473           fu?comip? %st(i), %st
474    
475    ----------------------------------------------------------------------
476    Name: Matthias Blume (blume (at) tti - c (dot) org)
477    Date: 2004/10/20 15:06:00 CDT
478    Tag: blume-20041020-standalone-backtrace
479    Description:
480    
481    Added a mechanism for getting back-trace information from standalone
482    programs.  Here is how it works:
483    
484       1. The part of the program from which you want to get backtrace
485          information (usually the whole program) should be wrapped with
486          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
487          main program could be modified from something like
488    
489             fun main (pgm, args) = ...
490    
491         to
492    
493            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
494    
495       2. To be able to access BackTrace.monitor, you have to add
496          library $smlnj-tdp/plugins.cm to the .cm file that contains your
497          main function.
498    
499       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
500          CM might have created in the past for your project).
501    
502       4. Build the system using this command line:
503    
504             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
505                      myprog.cm MyProg.main myprog
506    
507          instead of the usual
508    
509             ml-build myprog.cm MyProg.main myprog
510    
511    I changed a library name:
512    
513       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
514    
515    New libraries:
516    
517       $smlnj-tdp/back-trace.cm
518          -- when loaded causes the back-trace plugin to be installed
519       $smlnj-tdp/coverage.cm
520          -- when loaded causes the coverage plugin to be installed
521    
522    ----------------------------------------------------------------------
523    Name: Matthias Blume (blume (at) tti - c (dot) org)
524    Date: 2004/10/18 16:45:00 CDT
525    Tag: blume-20041018-groupowner
526    Description:
527    
528    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
529    
530    Eliminated group owner specs from .cm files throughout the source tree.
531    
532    ----------------------------------------------------------------------
533    Name: Matthias Blume (blume (at) tti - c (dot) org)
534    Date: 2004/10/15 15:45:00 CDT
535    Tag: blume-20041015-coverage
536    Description:
537    
538    * Test coverage tool added!
539    * Further reorganization of tracing-, debugging-, and profiling support:
540    
541        - moved original BTImp -- now called BackTrace -- into a separate
542          library called $/trace-debug-profile.cm
543        - eliminated all mentions of BTrace from SMLofNJ.Internals
544        - only the instrumentation mechanism is now left in the compiler proper
545        - BackTrace module is a plugin which is NOT plugged in by default
546        - Coverage module is another such plugin
547    
548    To get the benefits of any of these plugin modules, the code in
549    question must be compiled with tdp instrumentation turned on.  This
550    can be done by setting SMLofNJ.Internals.TDP.mode to true.
551    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
552    
553    Plugins are selected at link time.  (Pre-compiled instrumented code
554    can be re-loaded with different plugins in effect.)  When an
555    instrumented module is linked, whatever plugins are at that time
556    enabled will come into effect for that module.
557    
558    To enable the back-trace plugin, load library $/trace-debug-profile.cm
559    and invoke BackTrace.install() (e.g., from the interactive prompt).
560    To enable the coverage plugin, load the same library and invoke
561    Coverage.install().
562    
563    Back-traces are generated automatically on uncaught exceptions and
564    when the code in question explicitly invokes BackTrace.trigger().
565    
566    Coverage (and execution frequency-) information must be queried
567    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
568    
569    ----------------------------------------------------------------------
570    Name: Matthias Blume (blume (at) tti - c (dot) org)
571    Date: 2004/10/14 17:40:00 CDT
572    Tag: blume-20041014-tdp-core
573    Description:
574    
575    Snapshot of a significant overhaul of how the trace/debug/profile support
576    is hooked into the system (specifically: Core and SMLofNJ.Internals).
577    
578    ----------------------------------------------------------------------
579    Name: Matthias Blume (blume (at) tti - c (dot) org)
580    Date: 2004/10/13 16:34:00 CDT
581    Tag: blume-20041013-tdp
582    Description:
583    
584    Some rationalization of names:
585    
586       structure BTrace -> structure TDPInstrument
587       etc.
588    
589    This is is preparation of using the original back-trace
590    instrumentation for other purposes.  "TDP" stands for
591    Trace/Debug/Profile.
592    
593    The control flag controlling whether instrumentation is on or off is now
594    registered under a different name, so instead of running sml as
595    
596     sml -Cinstrument.btrace-mode=true
597    
598    one has to say
599    
600     sml -Ctdp.instrument=true
601    
602    ----------------------------------------------------------------------
603    Name: Matthias Blume (blume (at) tti - c (dot) org)
604    Date: 2004/10/11 16:37:00 CDT
605    Tag: blume-20041011-regions
606    Description:
607    
608    Made some minor modifications to elabcore.sml to have source regions
609    be propagated more tightly -- resulting in better (i.e., smaller)
610    regions being reported in error- and debug messages.
611    
612    ----------------------------------------------------------------------
613    Name: Matthias Blume (blume (at) tti - c (dot) org)
614    Date: 2004/10/08 22:50:00 CDT
615    Tag: blume-20041008-cmkw
616    Description:
617    
618    Fixed handling of keywords in .cm files: After seeing "is" the lexer
619    treats subsequent occurrences of "group", "library", "source", "is",
620    "*", and "-" as ordinary identifiers rather than keywords.
621    
622    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
623    is supposed to accept a tool argument called "source", but this did
624    not work because of the clash with the keyword.
625    
626    ----------------------------------------------------------------------
627    Name: Matthias Blume (blume (at) tti - c (dot) org)
628    Date: 2004/10/07 16:00:00 CDT
629    Tag: blume-20041007-cleanup
630    Description:
631    
632    Assorted cleanup work:
633    
634      - got rid of intstrmap in favor of using the library's
635        hash table implementation
636      - threw out most of the pathnames stuff, as it was not used anyway
637      - simplified tokentable implementation
638      - fixed some minor spelling errors
639    
640    ----------------------------------------------------------------------
641    Name: Matthias Blume (blume (at) tti - c (dot) org)
642    Date: 2004/10/06 15:15:15 CDT
643    Tag: blume-20041006-handler
644    Description:
645    
646    Cleaned up the absyn to reflect the invariant that HANDLE always
647    carries a FNexp as part of the type definition.  This eliminates some
648    superfluous sanity checks at runtime down the road.
649    
650    Some minor cleanup of the btrace code.
651    
652    ----------------------------------------------------------------------
653    Name: Matthias Blume (blume (at) tti - c (dot) org)
654    Date: 2004/10/01 10:20:30 CDT
655    Tag: blume-20041001-slave
656    Description:
657    
658    Added hack to make slave mode work in the presence of the version
659    tool.  (Still, since the master does two passes over the code for
660    CMB.make, the release number gets bumped twice when slaves are
661    attached.  I don't know if this is worth fixing...)
662    
663    ----------------------------------------------------------------------
664    Name: Matthias Blume (blume (at) tti - c (dot) org)
665    Date: 2004/09/30 10:55:00 CDT
666    Tag: blume-20040930-version
667    Description:
668    
669    * Moved the "version" magic into its own little library under
670      src/system/smlnj/internal.  This avoids expensive reconstruction of
671      a stable src/compiler/core.cm.
672    
673    * At the same time, structure CompilerVersion is now known as
674      structure SMLNJVersion.
675    
676    * Arranged for the version tool to NOT kick in when rebuilding the system
677      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
678      Also, loading the versiontool does not work when rebuilding the system
679      because CM is not properly initialized at that time.
680    
681    ----------------------------------------------------------------------
682    Name: Matthias Blume (blume (at) tti - c (dot) org)
683    Date: 2004/09/29 14:00:00 CDT
684    Tag: blume-20040929-autoversion
685    Description:
686    
687    Implemented some CM magic to have
688        file src/compiler/TopLevel/main/version.sml
689    generated automagically.
690    The version is taken from two files: config/version and config/release.
691    The first is expected to contain a two-part version number such as 110.49.
692    The second should contain a single number, but it may be missing.
693    
694    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
695    time the version tool is loaded (which is the first time you say CMB.make),
696    then the tool will increment the value stored in config/release every
697    time CMB.make is invoked.
698    
699    The binfile format is now insensitive to anything beyond the first
700    two components of a version number, so bumping the release does not render
701    binfiles incompatible.  Auto-bumping can be used to keep track of versions
702    during development without invalidating existing binfiles.
703    
704    In any case, every CMB.make updates the date information in version.sml.
705    (This is the date that is printed in the banner.)
706    
707    ----------------------------------------------------------------------
708    Name: Matthias Blume (blume (at) tti - c (dot) org)
709    Date: 2004/09/28 10:53:00 CDT
710    Tag: blume-20040928-controls
711    Description:
712    
713    Some cleanup of the controls code.
714    
715    ----------------------------------------------------------------------
716    Name: Matthias Blume (blume (at) tti - c (dot) org)
717    Date: 2004/09/27 22:08:00 CDT
718    Tag: blume-20040927-controls
719    Description:
720    
721    Added two pieces of functionality to the Controls interface:
722    
723    1.   val save'restore: 'a control -> unit -> unit
724    
725       grabs the current value of the control in stage 1 and restores it
726       in stage 2.
727    
728    2.   val set' : 'a control * 'a -> unit -> unit
729    
730       stores the given value into the control in stage 2 (i.e., delayed)
731       but does all error checking in stage 1.
732       (This is for string controls that need to do parse their argument --
733        something that might fail.  In some cases, notably in CM, one
734        already knows the intended argument but wants to delay the actual
735        assignment until a time when error recovery would be more difficult.)
736    
737    Changed the handling of controls in tool arguments to classes "sml" and
738    "lazysml":
739       - use Controls.save'restore as a more robust way of restoring the
740         old value (in particular: without having to re-parse the string)
741       - use controls to handle the "overload" keyword in the init group
742         (I believe this change actually fixes a long-standing obscure bug.)
743    
744    ----------------------------------------------------------------------
745    Name: Matthias Blume (blume (at) tti - c (dot) org)
746    Date: 2004/09/27 17:00:00 CDT
747    Tag: blume-20040927-lazysml
748    Description:
749    
750    Added a new tool class called "lazysml" to CM's tool chest.  The only
751    difference to "sml" is that compilation is done with Control.lazysml
752    set to true.  A source of class "lazysml" is automatically recognized
753    by a file name suffix of ".lml".
754    
755    In addition to the above feature, the original class "sml" now also
756    supports a tool argument "lazy" which has the same effect.  As a
757    result, the following three lines are equivalent:
758    
759        foo.sml : lazysml
760        foo.sml : sml (lazy)
761        foo.sml (lazy)
762    
763    The setting goes into effect both during parsing and during
764    compilation.  The original setting is restored right after parsing and
765    after compilation, respectively.
766    
767    In addition to all the above, there is also a general mechanism to set
768    ANY of the "controls" that are available at the command line via
769    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
770    apply as well. (In fact, "lazy" is implemented as a special case of
771    the general mechanism.)
772    
773    The .cm file syntax uses a new keyword tool argument called "with".
774    There are several ways of indicating the desired settings:
775    
776        foo.sml (with:parser.quotations=true)
777        foo.sml (with:(name:parser.quotations value:true))
778        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
779        foo.sml (with:(name1=value1 name2=value2 ...))
780        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
781    
782        etc.
783    
784    Another possible abbreviation is to leave out the =v or value:v part
785    if the name refers to a boolean control (in which case the value is
786    taken to be true).  Thus, one could get lazy sml also by saying:
787    
788        foo.sml (with:parser.lazy-keyword=true)
789        foo.sml (with:parser.lazy-keyword)
790        foo.sml (with:(name:parser.lazy-keyword value:true))
791        foo.sml (with:(name:parser.lazy-keyword))
792    
793    ----------------------------------------------------------------------
794    Name: Matthias Blume (blume (at) tti - c (dot) org)
795    Date: 2004/09/24 16:22:00 CDT
796    Tag: blume-20040924-ppc-long-branch
797    Description:
798    
799    Turned message about "emiting long form of branch" off by default.
800    Added a control flag to turn it back on when desired.
801    
802    ----------------------------------------------------------------------
803    Name: Matthias Blume (blume (at) tti - c (dot) org)
804    Date: 2004/09/24 16:05:00 CDT
805    Tag: blume-20040924-rounding
806    Description:
807    
808    Applied patch for setting rounding modes under Mac OS X.  Thanks to
809    Melissa O'Neill for providing the code!
810    
811    ----------------------------------------------------------------------
812    Name: Matthias Blume (blume (at) tti - c (dot) org)
813    Date: 2004/09/23 17:30:00 CDT
814    Tag: blume-20040923-envvars
815    Description:
816    
817    1. Changed definition of type ControlRegistry.registry_tree to
818       include control_info (i.e., the name of the controlling environment
819       variable).
820    
821    2. Added command-line flags -e and -E to print the names of environment
822       variables that can be used to control internal settings.
823       (This uses the new API mentioned in 1.)
824    
825    ----------------------------------------------------------------------
826    Name: Matthias Blume (blume (at) tti - c (dot) org)
827    Date: 2004/09/13 16:50:00 CDT
828    Tag: Release_110_49
829    Description:
830    
831    New working version (110.49).  NEW BOOTFILES!
832    
833    ----------------------------------------------------------------------
834    Name: Matthias Blume
835    Date: 2004/09/13 16:20:00 CDT
836    Tag: blume-20040913-config-mlrisc
837    Description:
838    
839    Put target "mlrisc" back into the default list.
840    (There is no harm in having it, and some users have expressed their
841    wish to have "mlrisc" included by default.)
842    
843    ----------------------------------------------------------------------
844    Name: John Reppy
845    Date: 2004/09/13
846    Tag: jhr-20040913-signals
847    Description:
848    
849    Fixed the signal masking code to properly nest mask/unmask operations
850    on a per-signal basis.
851    
852    ----------------------------------------------------------------------
853    Name: Matthias Blume (blume (at) tti - c (dot) org)
854    Date: 2004/09/08 13:20:00 CDT
855    Tag: blume-20040908-heap-magic
856    Description:
857    
858    Bumped the heap macig number to 0x09082004 to account for the changed
859    layout of the ML frame under MacOS X.
860    
861    ----------------------------------------------------------------------
862    Name: Allen Leung (leunga (at) reservoir (dot) com)
863    Date: 2004/09/03 11:26:00 EST
864    Tag: leunga-20040903-cygwin-install
865    Description:
866    
867       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
868    The Cygwin runtime is turned on by setting the environment
869    variable SMLNJ_CYGWIN_RUNTIME to 1.
870    
871    ----------------------------------------------------------------------
872    Name: Matthias Blume (blume (at) tti - c (dot) org)
873    Date: 2004/08/31 17:15:00 CDT
874    Tag: blume-20040831-core
875    Description:
876    
877    Added some exports to src/compiler/core.cm upon request by J. Joemann.
878    
879    ----------------------------------------------------------------------
880    Name: Matthias Blume (blume (at) tti - c (dot) org)
881    Date: 2004/08/30 17:55:00 CDT
882    Tag: blume-20040830-installer
883    Description:
884    
885    Upon request by Johannes Joemann:
886    
887    - improved ML code of installer to fall back to coping when renaming
888      fails (i.e., when source and target are on different file systems);
889      the code compiles but has yet to be tested in anger
890    - removed mlrisc from list of default targets (config/targets)
891    
892    ----------------------------------------------------------------------
893    Name: Matthias Blume (blume (at) tti - c (dot) org)
894    Date: 2004/08/27 17:20:00 CDT
895    Tag: blume-20040827-ptreql
896    Description:
897    
898    Added ptreql primop to structure InlineT (upon request from Larry
899    Paulson).
900    
901    ----------------------------------------------------------------------
902    Name: Allen Leung (leunga (at) reservoir (dot) com)
903    Date: 2004/08/15 21:21:00 EST
904    Tag: leunga-110_48-udgraph
905    Description:
906    
907    Another bug fix from Carl Hauser:
908    
909    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
910    > 48c48
911    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
912    > e::L)
913    > ---
914    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
915    > Without this, any deletion of an edge in an undirected graph does severe
916    > violence to the graph.
917    
918    ----------------------------------------------------------------------
919    Name: Allen Leung (leunga (at) reservoir (dot) com)
920    Date: 2004/08/10 23:35:00 EST
921    Tag: leunga-110_48-ppc
922    Description:
923    
924       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
925    
926    ----------------------------------------------------------------------
927    Name: Matthias Blume (blume (at) tti - c (dot) org)
928    Date: 2004/08/10 12:00:00 CDT
929    Tag: Release_110_48
930    Description:
931    
932    New working version (110.48).  NEW BOOTFILES!
933    
934    ----------------------------------------------------------------------
935    Name: Allen Leung (leunga (at) reservoir (dot) com)
936    Date: 2004/08/09 12:21:00 EST
937    Tag: leunga-110_47-dijsktra
938    Description:
939    
940    Bug fix from Carl Hauser:
941    
942    single_source_shortest_paths in dijkstra.sml was observed to get wrong
943    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
944    
945    The problem is that following the expression A.update(dist,s,Num.zero)
946    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
947    
948    ----------------------------------------------------------------------
949    Name: Matthias Blume (blume (at) tti - c (dot) org)
950    Date: 2004/08/06 18:10:00 CDT
951    Tag: blume-20040806-cmdline
952    Description:
953    
954    Fiddled with handling of command-line options:
955    
956      * sml now quits after processing the command line
957        if -H, -S, -h<n>, or -s<n> appears as the last
958        command-line argument
959      * a new option -q terminates the session when encountered on
960        the command line; subsequent arguments will be ignored
961      * bug fixes: short (erroneous) arguments are no longer ignored
962        completely
963    
964    ----------------------------------------------------------------------
965    Name: Allen Leung (leunga (at) reservoir (dot) com)
966    Date: 2004/08/04 18:17:00 EST
967    Tag: leunga-110_47-ppc-ibm-asm
968    Description:
969    
970      - Added minimal IBM assembly syntax support for PowerPC.
971    
972      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
973        been altered in the runtime.  We need an automatic way to keep the file
974        in sync.
975    
976    ----------------------------------------------------------------------
977    Name: Matthias Blume (blume (at) tti - c (dot) org)
978    Date: 2004/08/04 14:00:00 CDT
979    Tag: Release_110_47
980    Description:
981    
982    New working version (110.47).  NEW BOOTFILES!
983    
984    ----------------------------------------------------------------------
985    Name: Matthias Blume (blume (at) tti - c (dot) org)
986    Date: 2004/08/03 14:25:00 CDT
987    Tag: blume-20040803-callingconv
988    Description:
989    
990    Added low-level support for choosing C calling conventions by
991    twiddling the type of rawccall.  (See
992    src/compiler/Semant/types/cproto.sml for details.)
993    
994    ----------------------------------------------------------------------
995    Name: Matthias Blume (blume (at) tti - c (dot) org)
996    Date: 2004/08/02 15:55:00 CDT
997    Tag: blume-20040802-backout
998    Description:
999    
1000    Backed out of change to win32-filesys.c.  The earlier patch to
1001    get_file_time caused CM to produce files with the wrong time stamp.
1002    
1003    ----------------------------------------------------------------------
1004    Name: Matthias Blume (blume (at) tti - c (dot) org)
1005    Date: 2004/08/02 14:45:00 CDT
1006    Tag: blume-20040802-nlffi-win32
1007    Description:
1008    
1009    Added NLFFI support for Win32, adapted from a patch provided by David
1010    Hansel.  This is currently completely untested.  Also, the issue
1011    concerning stdcall vs. ccall is still unresolved.
1012    
1013    ----------------------------------------------------------------------
1014    Name: Matthias Blume (blume (at) tti - c (dot) org)
1015    Date: 2004/07/30 17:55:00 CDT
1016    Tag: blume-20040730-various
1017    Description:
1018    
1019    Gearing up towards 110.47...
1020    
1021    - various minor bugfixes to ml-nlffigen
1022    - a beginning of a manual for nlffi
1023    
1024    - eliminated 'export name=value' in config/install.sh as this does
1025      not work with certain versions of /bin/sh
1026      (Thanks to David King at Motorola for catching this.)
1027    
1028    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1029      - added a test for tm==NULL to gmtime.c and localtime.c
1030      - applied patch for incorrect GetFileTime under win32
1031      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1032    
1033    ----------------------------------------------------------------------
1034    Name: Matthias Blume (blume (at) tti - c (dot) org)
1035    Date: 2004/07/21 18:20:00 CDT
1036    Tag: blume-20040721-nlffigen
1037    Description:
1038    
1039    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1040      for a typedef to an incomplete type, but leave out the "typ" member.
1041      (This is just for consistency.)
1042    - Started to produce what is supposed to become better (i.e., comprehensive)
1043      documentation of what ml-nlffigen does and produces.
1044    
1045    ----------------------------------------------------------------------
1046    Name: Matthias Blume (blume (at) tti - c (dot) org)
1047    Date: 2004/07/14 16:25:00 CDT
1048    Tag: blume-20040714-union
1049    Description:
1050    
1051    Added C_UNION to c-calls/c-types.sml and updated the machinery
1052    (ml-nlffigen, cproto.sml) that conveys C function interface
1053    information to the code generator.
1054    
1055    However, the actual architecture-specific implementation of function
1056    arguments and results that are C unions is still not implemented.
1057    
1058    ----------------------------------------------------------------------
1059    Name: Allen Leung (leunga (at) reservoir (dot) com)
1060    Date: 2004/07/14 14:38:00 EST
1061    Tag: leunga-110_46_1-ppc-lwzu
1062    Description:
1063    
1064        Added these instructions to the PowerPC architecture:
1065        LBZU(X), LHZU(X), LWZU(X),
1066        STWU(X), STFDU, STFSU
1067    
1068        etc...
1069    
1070        Note: I haven't added their instruction encoding into the description.
1071    
1072    ----------------------------------------------------------------------
1073    Name: Allen Leung (leunga (at) reservoir (dot) com)
1074    Date: 2004/07/13 15:04:00 EST
1075    Tag: leunga-110_46_1-ppc-lwarx
1076    Description:
1077    
1078       Added the two instructions LWARX and STWCX to the PowerPC
1079    instruction set.
1080    
1081       A (untested) rewrite of loop-structure.sml.  The old version
1082    is completely broken.
1083    
1084    ----------------------------------------------------------------------
1085    Name: Matthias Blume (blume (at) tti - c (dot) org)
1086    Date: 2004/07/13 13:50:00 CDT
1087    Tag: blume-20040713-nlffi
1088    Description:
1089    
1090    - use paramAlloc to report c-calls with too many arguments
1091      (for PPC version where parameter area is pre-allocated)
1092    - added ccall_maxargspace to machspec (to implement the above)
1093    - made "make" commend in CM's "make" tool configurable
1094    - added option (default: on) for passing the name of the SML/NJ's "bin"
1095      directory to "make"; the call looks like this:
1096    
1097         make <options> SMLNJ_BINDIR=<dir> <target>
1098    
1099      This can be used by the Makefile to, e.g., pick the "right" version
1100      of ml-nlffigen.
1101    - minor code tweaks
1102    
1103    ----------------------------------------------------------------------
1104    Name: Matthias Blume (blume (at) tti - c (dot) org)
1105    Date: 2004/07/12 22:50:00 CDT
1106    Tag: blume-110_46_1-macosx-nlffi
1107    Description:
1108    
1109    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1110    though.
1111    
1112    Note:  1. You have to make a new, clean build of the runtime system.
1113           2. There are new BOOTFILES, you have to use them!
1114              (Doing the bootstrap process yourself would be *very* painful!
1115               If you absolutely have to do it, build the system under
1116               a different architecture and then cross-compile.)
1117    
1118    Version bumped to 110.46.1 to account for runtime data format changes.
1119    
1120    ----------------------------------------------------------------------
1121    Name: Matthias Blume (blume (at) tti - c (dot) org)
1122    Date: 2004/06/18 14:30:00 CDT
1123    Tag: blume-20040618-unix
1124    Description:
1125    
1126    Changed the implementation of structure Unix so that the same stream
1127    is returned every time one of the {text,bin}{In,Out}streamOf functions
1128    is invoked on the same proc.  This is not what the spec currently
1129    says -- although IMO it arguably should.  (See discussion below.)
1130    
1131    ----------------------------------------------------------------------
1132    Name: Matthias Blume (blume (at) tti - c (dot) org)
1133    Date: 2004/06/17 18:15:00 CDT
1134    Tag: Release_110_46
1135    Description:
1136    
1137    New working version (110.46).  NEW BOOTFILES!
1138    
1139    ----------------------------------------------------------------------
1140    Name: Matthias Blume (blume (at) tti - c (dot) org)
1141    Date: 2004/06/17 17:20:00 CDT
1142    Tag: blume-20040617-timer-unix
1143    Description:
1144    
1145    Changed the interface of structures Timer and Unix to match the most
1146    recent Basis spec.
1147    
1148    In the case of Unix there still seems to be an open/weird issue:
1149    
1150       The {text,bin}{In,Out}streamOf functions are supposed to create
1151       fresh streams whenever they are called -- as opposed to have them
1152       return the same stream every time.  This design is supposed to
1153       prevent space leaks caused by proc values hanging on to streams.
1154    
1155       The reap function, on the other hand, is supposed to close the
1156       streams.  This cannot be done without having a handle on the
1157       stream in proc after all...
1158    
1159    I took the liberty to implement the following stopgap solution:
1160    
1161      The proc value hangs on to the most recently created stream(s).
1162      Reap closes those.  If either or both of the two streams hadn't
1163      been created at all yet, then reap will close the corresponding
1164      file descriptors directly.
1165    
1166    PS: I don't understand the original space leak argument anymore.  If
1167    a proc hangs on to the imperative stream, then I/O operations on those
1168    will advance the state of the cached stream and avoid the space leak.
1169    
1170    ----------------------------------------------------------------------
1171    Name: Matthias Blume (blume (at) tti - c (dot) org)
1172    Date: 2004/05/28 16:45:00 CDT
1173    Tag: blume-20040528-basis
1174    Description:
1175    
1176    Added signature PACK_REAL and exported functor PrimIO.
1177    
1178    ----------------------------------------------------------------------
1179    Name: Matthias Blume (blume (at) tti - c (dot) org)
1180    Date: 2004/05/25 16:00:00 CDT
1181    Tag: blume-20040525-group-owner
1182    Description:
1183    
1184    CM now ignores (but still accepts) the "owner" information in group
1185    descriptions.  The owner of a group is its next enclosing
1186    library. Each group must have a unique owner.  (There is a virtual
1187    "toplevel" library that owns groups which are not nested within a real
1188    library.)  Previously, each group had to explicitly declare its owner,
1189    and CM would check that such a declaration is correct.  The new scheme
1190    is to have CM check that for each group there is precisely one owning
1191    library.
1192    
1193    The advantage of the new scheme is that the programmer no longer needs
1194    to maintain the somewhat annoying owner information.  The downside is
1195    that CM cannot enforce the ownership rule across multiple runs of
1196    CM.make.  Fortunately, enclosing the same group in two different
1197    libraries A and B which are not part of the same program does not
1198    cause real problems.
1199    
1200    ----------------------------------------------------------------------
1201    Name: Matthias Blume (blume (at) tti - c (dot) org)
1202    Date: 2004/05/20 16:00:00 CDT
1203    Tag: blume-20040520-win32
1204    Description:
1205    
1206    Made the win32 version work again.  (Strangely, a misplaced comma had
1207    slipped into win32-process.c which prevented the runtime from being
1208    compiled correctly.)
1209    
1210    Also, included a minor addition to ml-build.bat analogous to what was
1211    done in blume-20040519-ml-build.
1212    
1213    ----------------------------------------------------------------------
1214    Name: Matthias Blume (blume (at) tti - c (dot) org)
1215    Date: 2004/05/19 22:10:00 CDT
1216    Tag: blume-20040519-ml-build
1217    Description:
1218    
1219    Arranged for ml-build to clean up after itself a little bit better.
1220    The script generates a temporary SML source file and compiles it using
1221    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1222    gets rid of those at the end, so they don't accumulate under .cm.
1223    
1224    This required a minor change to install.sh because the name of the
1225    metadata directory (default: .cm) is actually configurable at
1226    installation time.
1227    
1228    ----------------------------------------------------------------------
1229    Name: Matthias Blume (blume (at) tti - c (dot) org)
1230    Date: 2004/05/18 15:50:00 CDT
1231    Tag: blume-20040518-mkreader
1232    Description:
1233    
1234    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1235    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1236    
1237    ----------------------------------------------------------------------
1238    Name: Matthias Blume (blume (at) tti - c (dot) org)
1239    Date: 2004/05/11 14:35:00 CDT
1240    Tag: blume-20040511-win32sock
1241    Description:
1242    
1243    Added previously missing support for many socket-related functions
1244    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1245    for the voluminous patch!
1246    
1247    (I have not tested this patch under win32 yet.)
1248    
1249    Here is David's e-mail:
1250    
1251    Hi,
1252    
1253    Attached to this email you find a diff against sml/nj 110.45
1254    that will enable socket support under Windows.
1255    
1256    To apply the patch (using unix or cygwin)
1257    1) gunzip runtime.diff.gz
1258    2) "cd" into "src/runtime" in the source tree of a fresh
1259       110.45 installation.
1260    3) patch -p 1 < [your/path/to]runtime.diff
1261    
1262    The code compiles fine but has NOT yet been extensively tested.
1263    I only ran a few tests for basic socket client functionality
1264    (which worked fine).  Especially the functions that use ioctl
1265    are not tested at all and might not work (see below).
1266    
1267    I implemented this since we want to move to a newer version of sml/nj
1268    but need socket support in order to use it.  This is the first time I
1269    even had a look at the sml/nj source,  so please review my changes
1270    before making this part of the distribution!  Here are a few issues
1271    that I think might be better for someone to solve who is more
1272    familiar with the sml/nj source (and socket programming):
1273    
1274    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1275      exception since I could not figure out what the windows equivalent
1276      of these functions is
1277    
1278    - In sockets-osdep.h there are a some #include statements that are
1279      only used in a few files that include sockets-osdep.h
1280    
1281    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1282      does not process its return value since I don't know how to
1283      report an error upwards.
1284    
1285    - It would probably be good to have a call to WSACleanup() when
1286      the library is unloaded (if there is such a possibility).
1287      Otherwise I think Windows will take care of this automatically
1288      when the process finishes.
1289    
1290    - I used ioctlsocket() as a replacement for ioctl() but I have
1291      no idea if that is actually the proper replacement on Windows.
1292    
1293    - All these issues are marked in the code by "FIXME" comments.
1294    
1295    We use sml/nj extensively in our products and are quite happy
1296    with it.  I hope this contribution will help you.
1297    
1298    Keep up the good work!
1299    
1300    David
1301    
1302    ----------------------------------------------------------------------
1303    Name: Matthias Blume (blume (at) tti - c (dot) org)
1304    Date: 2004/05/11 14:20:00 CDT
1305    Tag: blume-20040511-installml
1306    Description:
1307    
1308    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1309    the patch.)
1310    
1311    ----------------------------------------------------------------------
1312    Name: Matthias Blume (blume (at) tti - c (dot) org)
1313    Date: 2004/05/11 14:05:00 CDT
1314    Tag: blume-20040511-nlffi-netbsd
1315    Description:
1316    
1317    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1318    the patch.)
1319    
1320    ----------------------------------------------------------------------
1321    Name: Matthias Blume (blume (at) tti - c (dot) org)
1322    Date: 2004/05/11 12:05:00 CDT
1323    Tag: blume-20040511-exports
1324    Description:
1325    
1326    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1327    export lists in compiler-related .cm-files.
1328    
1329    ----------------------------------------------------------------------
1330    Name: Matthias Blume (blume (at) tti - c (dot) org)
1331    Date: 2004/05/11 11:35:00 CDT
1332    Tag: blume-20040511-allsource
1333    Description:
1334    
1335    The installer now honors the "src-smlnj" target again, although its meaning
1336    has changed from "all sources required for the compiler" to "all sources
1337    the installer knows about".  In other words, if you enable "src-smlnj"
1338    in the "targets" file, then the installer will pull in sources for
1339    everything.  (Notice that this refers to source code only.  Compiled
1340    code is still only installed for modules that were requested explicitly
1341    or which are required for other modules that were requested explicitly.)
1342    
1343    ----------------------------------------------------------------------
1344    Name: Matthias Blume (blume (at) tti - c (dot) org)
1345    Date: 2004/04/23 17:40:00 CDT
1346    Tag: blume-20040423-ieee-scan
1347    Description:
1348    
1349    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1350    in the exponent calculation we get INF or ZERO (depending on the mantissa
1351    and the sign of the exponent).
1352    
1353    ----------------------------------------------------------------------
1354    Name: Matthias Blume (blume (at) tti - c (dot) org)
1355    Date: 2004/04/23 10:40:00 CDT
1356    Tag: blume-20040423-ml-build
1357    Description:
1358    
1359    The ml-build script now terminates with a non-0 status when something
1360    goes wrong.
1361    
1362    ----------------------------------------------------------------------
1363    Name: Matthias Blume (blume (at) tti - c (dot) org)
1364    Date: 2004/04/22 16:35:00 CDT
1365    Tag: blume-20040422-Option
1366    Description:
1367    
1368    Made exception Option to be the same as exception Option.Option
1369    (as it should be).
1370    
1371    ----------------------------------------------------------------------
1372    Name: Allen Leung (leunga (at) reservoir (dot) com)
1373    Date: 2004/03/19 14:40:00 EST
1374    Tag: leunga-20040319-cygwin-nlffi
1375    Description:
1376    
1377    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1378    of SML/NJ.  The problem is that
1379    
1380       lib = dlopen(NULL, ...)
1381       f   = dlsym(lib, "malloc");
1382    
1383    does not work on Windows unless we explicitly export symbols
1384    such as 'malloc' during linking.  We fixed this by explicitly
1385    exporting the required symbols with the magic gcc incantation:
1386    
1387        -Wl,--export-all cygwin.def
1388    
1389    where cygwin.def is a file containing all the symbols that
1390    we wish to export.
1391    
1392    I suspect this is a Windows problem and we'll have to
1393    do the same (somehow with windows compilers) when
1394    we build the native win32 version with the system
1395    calls LoadLibrary/GetProcAddress.
1396    
1397    ----------------------------------------------------------------------
1398    Name: Matthias Blume (blume (at) tti - c (dot) org)
1399    Date: 2004/03/04 16:35:00 CST
1400    Tag: blume-20040304-intinf-fmt
1401    Description:
1402    
1403    Fixed problem with IntInf.fmt (sign would show up on the right instead
1404    of on the left for BIN, OCT, and HEX).
1405    
1406    ----------------------------------------------------------------------
1407    Name: Matthias Blume (blume (at) tti - c (dot) org)
1408    Date: 2004/03/04 11:25:00 CST
1409    Tag: blume-20040304-symlinks
1410    Description:
1411    
1412    Fixed problem with installer script (unix only) where bin/ml-yacc and
1413    friends pointed (via symlinks) to absolute locations instead of just
1414    .run-sml.  This was reported by Vesa A Norrman.
1415    
1416    ----------------------------------------------------------------------
1417    Name: Matthias Blume (blume (at) tti - c (dot) org)
1418    Date: 2004/02/13 14:50:00 CST
1419    Tag: Release_110_45
1420    Description:
1421    
1422    New working version (110.45). New bootfiles.
1423    
1424    ----------------------------------------------------------------------
1425    Name: Matthias Blume (blume (at) tti - c (dot) org)
1426    Date: 2004/01/26 15:15:15 CST
1427    Tag: blume-20040126-toplevel
1428    Description:
1429    
1430    Improved handling of exceptions at the interactive toplevel.
1431    
1432    ----------------------------------------------------------------------
1433    Name: Matthias Blume (blume (at) tti - c (dot) org)
1434    Date: 2004/01/26 11:25:00
1435    Tag: blume-20040126-app
1436    Description:
1437    
1438    Type of top-level "app" corrected.
1439    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1440    
1441    ----------------------------------------------------------------------
1442    Name: Matthias Blume (blume (at) tti - c (dot) org)
1443    Date: 2003/11/18 17:10 CST
1444    Tag: blume-20031118-basis-fiddle
1445    Description:
1446    
1447    - changed Timer interface to what might become the spec
1448    - POSIX_FLAGS -> BIT_FLAGS according to spec
1449    - some other minor discrepancies wrt. spec eliminated
1450    
1451    ----------------------------------------------------------------------
1452    Name: Matthias Blume (blume (at) tti - c (dot) org)
1453    Date: 2003/11/06 12:00:00 CST
1454    Tag: Release_110_44
1455    Description:
1456    
1457    New working version (110.44). New bootfiles.
1458    
1459    ----------------------------------------------------------------------
1460    Name: Matthias Blume (blume (at) tti - c (dot) org)
1461    Date: 2003/11/04 11:50:00 CST
1462    Tag: blume-20031104-move-libraries
1463    Description:
1464    
1465    Eliminated the "dont_move_libraries" directive in config/targets.
1466    (The mechanism was broken and could not be fixed easily.  Moreover,
1467    there does not seem to be any reason not to move all libraries into
1468    lib during installation.  I originally implemented this directive as a
1469    backward-compatibility feature when I first introduced the new CM.
1470    Now that things have been stable for a long time and going back to the
1471    old CM is not an option, there is no reason to keep it around.)
1472    
1473    ----------------------------------------------------------------------
1474    Name: Matthias Blume (blume (at) tti - c (dot) org)
1475    Date: 2003/11/03 16:00:00 CST
1476    Tag: blume-20031103-installdir
1477    Description:
1478    
1479    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1480    Richards for pointing out the problem and providing the solution.)
1481    
1482    ----------------------------------------------------------------------
1483    Name: Matthias Blume (blume (at) tti - c (dot) org)
1484    Date: 2003/10/01 17:05:00 CDT
1485    Tag: blume-20031001-lal-mlrisc
1486    Description:
1487    
1488    MLRISC bug fix from Lal.
1489    
1490    ----------------------------------------------------------------------
1491    Name: Matthias Blume (blume (at) tti - c (dot) org)
1492    Date: 2003/09/30 16:10:00 CDT
1493    Tag: blume-20030930-primio-bat
1494    Description:
1495    
1496    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1497    2. Improved .bat files (for Win32 port) to make things work under Win95.
1498       (thanks to Aaron S. Hawley for this one)
1499    
1500    ----------------------------------------------------------------------
1501    Name: Matthias Blume (blume (at) tti - c (dot) org)
1502    Date: 2003/09/26 16:05:00 CDT
1503    Tag: blume-20030926-wrappriv
1504    Description:
1505    
1506    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1507    
1508    ----------------------------------------------------------------------
1509    Name: Matthias Blume (blume (at) tti - c (dot) org)
1510    Date: 2003/09/26 15:00:00 CDT
1511    Tag: blume-20030926-110_43_3
1512    Description:
1513    
1514    - additional cleanup
1515    - version number bump, NEW BOOTFILES
1516    
1517    ----------------------------------------------------------------------
1518    Name: Matthias Blume (blume (at) tti - c (dot) org)
1519    Date: 2003/09/26 12:00:00 CDT
1520    Tag: blume-20030926-ppautoload
1521    Description:
1522    
1523    I modified the read-eval-print loop so that the autoloader gets
1524    invoked whenever the prettyprinter tries to look up a symbol that
1525    is not currently defined in the toplevel environment but which
1526    appears in CM's autoload registry.  As a result, we see far fewer of
1527    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1528    
1529    In addition to this I tried to clean up some pieces of the Basis
1530    implementation (e.g., Socket, Word8Array) in order to prevent other
1531    instances of these ?.Foo.Bar.xxx names from being printed.
1532    
1533    The mechanism that picks names for types still needs some work, though.
1534    (Right now it seems that if there is a type A.t which is defined to
1535    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1536    "?.B.u" although the perhaps more sensible solution would be to use
1537    "A.t" in this case.  In other words, the prettyprinter should follow
1538    a chain of DEFtycs not farther than there are corresponding toplevel
1539    names in the current environment.)
1540    
1541    ----------------------------------------------------------------------
1542    Name: Matthias Blume (blume (at) tti - c (dot) org)
1543    Date: 2003/09/24 16:31:00 CDT
1544    Tag: blume-20030924-installer
1545    Description:
1546    
1547    Another installer tweak:  All the ML code for the installer is now
1548    compiled during CMB.make and put into a little library called
1549    $smlnj/installer.cm.  The installation then simply invokes
1550    
1551       sml -m $smlnj/installer.cm
1552    
1553    and everything happens automagically.
1554    
1555    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1556    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1557            and BINDIR.
1558    
1559    The new scheme guarantees that the ML code responsible for the installation
1560    is in sync with the APIs of the main system.  Also, the installer is
1561    somewhat faster because the installer script is precompiled.
1562    
1563    ----------------------------------------------------------------------
1564    Name: Matthias Blume (blume (at) tti - c (dot) org)
1565    Date: 2003/09/24 15:35:00 CDT
1566    Tag: blume-20030924-synsock
1567    Description:
1568    
1569    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1570    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1571    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1572    implementations of non-blocking operations from CML's Socket
1573    structure.
1574    
1575    ----------------------------------------------------------------------
1576    Name: Matthias Blume (blume (at) tti - c (dot) org)
1577    Date: 2003/09/24 15:10:05 CDT
1578    Tag: blume-20030924-sockets
1579    Description:
1580    
1581    1. Fixed SOCKET API and implementation to match Basis spec.
1582       This required changing the internal representation of sockets to one
1583       that remembers (for each socket file descriptor) whether it is currently
1584       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1585       call is made only if the state actually needs to change).
1586    
1587    2. OS-specific details of sockets were moved into separate files, thus
1588       making it possible to unify the bulk of the socket implementations
1589       between Unix and Win32.
1590    
1591    3. CML's socket API changed accordingly.
1592       (Note that we need to remove non-blocking functions from this API
1593       since they are redundant in the case of CML!)
1594    
1595    4. CML's socket implementation now makes use of non-blocking functions
1596       provided by Basis, thus removing all OS-dependent code from this part
1597       of CML.
1598    
1599    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1600    
1601    ----------------------------------------------------------------------
1602    Name: Matthias Blume (blume (at) tti - c (dot) org)
1603    Date: 2003/09/22 12:10:00 CDT
1604    Tag: blume-20030922-110_43_2
1605    Description:
1606    
1607    Made a new interim version and bootfiles for developer's bootstrapping
1608    convenience.
1609    
1610       110.43.2 -- NEW BOOTFILES
1611    
1612    ----------------------------------------------------------------------
1613    Name: Matthias Blume (blume (at) tti - c (dot) org)
1614    Date: 2003/09/19 15:55:00 CDT
1615    Tag: blume-20030919-cmdir
1616    Description:
1617    
1618    1. new-install.sh -> install.sh
1619    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1620    3. tweaked installer so that another name instead of .cm can be chosen
1621       at install time (by setting the CM_DIR_ARC environment variable
1622       during installation); once installation is complete, the name is
1623       fixed
1624    
1625    ----------------------------------------------------------------------
1626    Name: Matthias Blume (blume (at) tti - c (dot) org)
1627    Date: 2003/09/18 16:00:00 CDT
1628    Tag: blume-20030918-110_43_1
1629    Description:
1630    
1631    Made a new interim version and bootfiles for developer's bootstrapping
1632    convenience.
1633    
1634       110.43.1 -- NEW BOOTFILES
1635    
1636    ----------------------------------------------------------------------
1637    Name: Matthias Blume (blume (at) tti - c (dot) org)
1638    Date: 2003/09/18 15:20:00 CDT
1639    Tag: blume-20030918-misc
1640    Description:
1641    
1642    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1643       this seems to be controversial at the moment) and used those in
1644       Posix.ProcEnv.times.
1645    
1646    2. Added Time.{from,to}Nanoseconds to Time.
1647    
1648    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1649       For example, fromLargeInt never needs to look at more than 3 "big
1650       digits" to get its 53 bits of precision.
1651    
1652    ----------------------------------------------------------------------
1653    Name: Matthias Blume (blume (at) tti - c (dot) org)
1654    Date: 2003/09/17 16:30:00 CDT
1655    Tag: blume-20030917-real32-slices
1656    Description:
1657    
1658    Added an entry to the primitive environment
1659    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1660    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1661    
1662    Removed some of the "magic" constants in real64.sml and replaced them
1663    with code that generates these values from their corresponding
1664    integer counterparts.
1665    
1666    Made all(?) the slice-related changes to the Basis and made everything
1667    compile again...
1668    
1669    ----------------------------------------------------------------------
1670    Name: Matthias Blume (blume (at) tti - c (dot) org)
1671    Date: 2003/09/15 17:45:00 CDT
1672    Tag: blume-20030915-rbase
1673    Description:
1674    
1675    Fixed bug in Real.fromLargeInt.
1676    
1677    ----------------------------------------------------------------------
1678    Name: Matthias Blume (blume (at) tti - c (dot) org)
1679    Date: 2003/09/13 18:11:00 CDT
1680    Tag: blume-20030913-libinstall
1681    Description:
1682    
1683    Minor bugfix in config/libinstall (set anchor with path to
1684    standalone tool after installing it, otherwise libraries that
1685    need ml-lex or ml-yacc won't compile the first time the installer
1686    runs).
1687    
1688    ----------------------------------------------------------------------
1689    Name: Matthias Blume (blume (at) tti - c (dot) org)
1690    Date: 2003/09/12 11:45:00 CDT
1691    Tag: blume-20030912-various
1692    Description:
1693    
1694     - fixed bug in Real.toLargeInt
1695     - fixed bug in Posix.ProcEnv.times
1696     - changed inputLine functions to return an option
1697     - minor installer improvements / bugfixes
1698     - changed default @SMLalloc parameter for x86/celeron to 64k
1699    
1700    ----------------------------------------------------------------------
1701    Name: Matthias Blume (blume (at) tti - c (dot) org)
1702    Date: 2003/09/09 22:00:00 CDT
1703    Tag: Release_110_43
1704    Description:
1705    
1706        New working release 110.43.  New bootfiles.
1707    
1708    ----------------------------------------------------------------------
1709    Name: Matthias Blume (blume (at) tti - c (dot) org)
1710    Date: 2003/09/09 19:20:00 CDT
1711    Tag: blume-20030909-installer
1712    Description:
1713    
1714    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1715    Modified config/install.bat to take advantage of it.  Also modified
1716    config/install.sh (and called it config/new-install.sh) to take advantage
1717    of it on Unix systems.  (The SML code is (supposed to be) platform-
1718    independent.)
1719    
1720    The installer can now install everything under Win32
1721    as well as under *nix as long as it compiles.
1722    
1723    Other changes:
1724    
1725      - made CML compile again under Win32
1726      - made eXene compile under Win32 (by providing a fake structure UnixSock
1727        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1728      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1729        same as type int;  under Win32 it isn't
1730      - fixed some slice-related problems in the win32-specific parts of CML
1731      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1732        (under Win32, the volume name is case-insensitive, and the
1733        OS.Path code compares volume names for equality)
1734    
1735    ----------------------------------------------------------------------
1736    Name: Matthias Blume (blume (at) tti - c (dot) org)
1737    Date: 2003/09/08 11:55:00 CDT
1738    Tag: blume-20030908-fullpath
1739    Description:
1740    
1741    Made Win32 version of OS.FileSys.fullPath return current directory
1742    when given an empty string.  This is what the spec says, and incidentally,
1743    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1744    cases when presented with the name of a non-existing .cm file.)
1745    
1746    ----------------------------------------------------------------------
1747    Name: Matthias Blume (blume (at) tti - c (dot) org)
1748    Date: 2003/09/04 16:30:00 CDT
1749    Tag: blume-20030905-slices-etc
1750    Description:
1751    
1752      1. Changed interface to vectors and arrays in Basis to match
1753         (draft) Basis spec.
1754      2. Added signatures and implementations of slices according to
1755         Basis spec.
1756      3. Edited source code throughout the system to make it compile again
1757         under 1. and 2.  (In some cases code had to be added to have it
1758         match the new signatures.)
1759      4. MLRISC should be backward-compatible: the copies of the originals
1760         of files that needed to change under 3. were retained, the .cm files
1761         check the compiler version number and use old versions when
1762         appropriate.
1763      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1764         dirstream -> string option (in accordance with Basis spec).
1765      6. When generating code that counts lines, ml-lex used function
1766         CharVector.foldli, taking advantage of its old interface.
1767         This has been replaced with the corresponding code from
1768         CharVectorSlice.  (html-lex must be re-lexed!)
1769      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1770         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1771      8. Removed temporary additions (fromInternal, toInternal) from the
1772         (now obsolete) IntInf in smlnj-lib/Util.
1773      9. Cleaned up structure Byte.
1774     10. Added localOffset, scan, and fromString to Date (according to spec).
1775         Cleaned/corrected implementation of Date.
1776         (Still need to check for correctness; implement better canonicalizeDate.)
1777     11. Added "scan" to signature IEEE_REAL.
1778     12. Some improvements to IntInf [in particular: efficiency-hack for
1779         mod and rem when second operand is 2 (for parity checks).]
1780     13. Changed representation of type Time.time, using a single IntInf.int
1781         value counting microseconds.  This considerably simplified the
1782         implementation of structure Time.  We now support negative time
1783         values; scan and fromString handle signs.
1784     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1785         ArraySlice).
1786    
1787    ----------------------------------------------------------------------
1788    Name: Matthias Blume (blume (at) tti - c (dot) org)
1789    Date: 2003/08/28 17:00:00 CDT
1790    Tag: blume-20030828-intinf
1791    Description:
1792    
1793    This is a major update which comes with a version number bump
1794    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1795    and an implementation of IntInf in the Basis.
1796    
1797    There are a fairly large number of related changes and updates throughout
1798    the system:
1799    
1800    Basis:
1801            - Implemented IntInf.
1802            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1803            - Added some missing Real64 operations, most notably Real.toLargeInt.
1804            - Added FixedInt as a synonym for Int32.
1805    
1806    compiler:
1807            * Added support for a built-in intinf type.
1808              - literals
1809              - pattern matching
1810              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
1811              - overloading on literals and operations
1812    
1813            This required adding a primitive type intinf, some additional
1814            primops, and implementations for several non-trivial intinf
1815            operations in Core.  (The intinf type is completely abstract
1816            to the compiler; all operations get delegated back to the Core.)
1817    
1818            * Intinf equality is handled by polyequal.  However, the compiler
1819            does not print its usual warning in this case (since polyequal
1820            is the right thing to do there).
1821    
1822            * Improved the organization of structure InlineT.
1823    
1824            * A word about conversion primops:
1825            If conversions involving intinf do not cancel out during
1826            CPS contract, then the compiler must insert calls to Core functions.
1827            Since all core access must be resolved already during the FLINT
1828            translate phase, it would be too late a the time of CPS contract
1829            to add new Core calls.  For this reason, conversion primops
1830            for intinf carry two arguments:  1. the numeric argument that
1831            they are supposed to convert, and 2. the Core function that
1832            can help with this conversion if necessary.  If CPS contract
1833            eliminates a primop, then the associated Core function becomes
1834            dead and goes away.  Intinf conversion primops that do not get
1835            eliminated by CPS contract get rewritten into calls of their
1836            core functions by a separate, new phase.
1837    
1838    interactive system:
1839            - Control.Print.intinfDepth controls max length of intinf constants
1840            being printed. (Analogous to Control.Print.stringDepth.)
1841            - Cleanup in printutil and pputil: got rid of unused stuff and
1842            duplicates; replaced some of the code with code that makes better
1843            use of library functionality.
1844    
1845    CM:
1846            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
1847            are no longer silent.
1848    
1849    CKIT:
1850            Fixed mismatched uses of Int32 and LargeInt.  I always decided
1851            in favor of LargeInt -- which is now the same as IntInf.
1852            CKIT-knowledgable people should check whether this is what's
1853            intended and otherwise change things back to using Int32 or
1854            FixedInt.
1855    
1856    Throughout the code:
1857            Started using IntInf.int literals and built-in operations
1858            (e.g., comparison with 0) where this seems appropriate.
1859    
1860    
1861    ----------------------------------------------------------------------
1862    Name: Dave MacQueen (dbm@cs.uchicago.edu)
1863    Date: 2003/08/13 11:36:00 CDT
1864    Tag: dbm-20030813-mcz-merge1
1865    Description:
1866    
1867    Merging changes from the mcz-branch development branch into trunk.
1868    These changes involve replacement of the emulated old prettyprinter
1869    interface with direct use of the SML/NJ Lib PP library, and fixing
1870    of a couple of bugs (895, 1186) relating to error messages.  A new
1871    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
1872    has been added.
1873    
1874    ----------------------------------------------------------------------
1875    Name: Matthias Blume (blume (at) tti - c (dot) org)
1876    Date: 2003/08/11 15:45:00 CDT
1877    Tag: blume-20030811-windows
1878    Description:
1879    
1880    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
1881    
1882        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
1883    
1884    This patch restores SML/NJ's ability to run under win32.  There are a
1885    number of changes, including fixes for several bugs that had gone
1886    unnoticed until now:
1887    
1888      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
1889        trying to run the win32 version from within cygwin.)
1890    
1891      - There are a number of simple .bat scripts that substitute for their
1892        corresponding Unix shell-scripts.  (See below.)
1893    
1894      - The internals of ml-build have been modified slightly.  The main
1895        difference is that instead of calling ".link-sml" (or link-sml.bat)
1896        using OS.Process.system, the ML process delegates this task back
1897        to the script.  Otherwise problems arise in mixed environments such
1898        as Cygwin where scripts look and work like Unix scripts, but
1899        where OS.Process.system cannot run them.
1900    
1901      - In CM, the srcpath pickler used native pathname syntax -- which
1902        is incorrect in the case of cross-compilation.  The new pickle format
1903        is independent of platform-specific naming conventions.
1904    
1905      - Path configuration files (such as lib/pathconfig) can now choose
1906        between native and standard syntax.  Placing a line of the form
1907    
1908            standard!
1909    
1910        into the file causes all subsequent paths to be interpreted using
1911        CM standard pathname syntax (= Unix conventions); a line
1912    
1913            native!
1914    
1915        switches back to native style.  This was needed so that
1916        path config files can be written portably, see src/system/pathconfig.
1917    
1918      - Runtime system:
1919    
1920          - win32-filesys.c:  get_file_time and set_file_time now
1921            access modification time, not creation time.
1922    
1923          - I/O code made aware of new array representation.
1924    
1925          - Bug fixes in X86.prim.masm.
1926    
1927       - src/system/makeml made aware of win32. (For use under cygwin
1928            and other Unix-environments for windows.)
1929    
1930       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
1931         which caused BinIO.inputAll to fail consistently.
1932    
1933    .bat scripts:
1934    
1935       Windows .bat scripts assume that SMLNJ_HOME is defined.
1936    
1937       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
1938         applications (sml, ml-yacc, ml-lex).
1939       - ml-build.bat: analogous to ml-build.
1940       - config\install.bat:  Analogous to config/install.sh.  This requires
1941         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
1942         (nmake etc. must be on the path, and vcvars32 must have been run.)
1943         Moreover, sources for ml-lex and ml-yacc need to exist under src,
1944         and the bootfile hierarchy must have been unpacked under
1945         sml.boot.x86-win32.
1946         The script is very primitive and does a poor job at error checking.
1947         It only installs the base system, ml-lex, and ml-yacc.  No other
1948         libraries are being installed (i.e., you get only those that
1949         are part of the compiler.)
1950       - link-sml.bat:  analogous to .link-sml, but not currently used
1951    
1952    Unrelated bug fixes:
1953    
1954      - ml-nlffigen now exports structures ST_* corresponding to incomplete
1955        types.
1956      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
1957        otherwise.)
1958    
1959    ----------------------------------------------------------------------
1960    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
1961    Date: 2003/06/17
1962    Tag: macqueen-20030617-bug895
1963    Description:
1964    
1965    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
1966    Tag will be used for new development branch (mcz-branch) for
1967    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
1968    
1969    ----------------------------------------------------------------------
1970    Name: Matthias Blume (blume (at) tti - c (dot) org)
1971    Date: 2003/05/27 16:55:00 CDT
1972    Tag: blume-20030527-polyeq
1973    Description:
1974    
1975    Tried to eliminated most cases of polymorphic equality.
1976    
1977    ----------------------------------------------------------------------
1978    Name: Matthias Blume (blume (at) tti - c (dot) org)
1979    Date: 2003/05/21 17:45:00 CDT
1980    Tag: blume-20030517-complete
1981    Description:
1982    
1983    Two changes:
1984    
1985      1. Added a flag for controlling whether non-exhaustive bindings will
1986         be treated as errors (default is false).
1987      2. Cleaned up the *entire* source tree so that CMB.make goes through
1988         without a single non-exhaustive match- or bind warning.
1989    
1990    ----------------------------------------------------------------------
1991    Name: Matthias Blume (blume (at) tti - c (dot) org)
1992    Date: 2003/05/17 10:20:00 CDT
1993    Tag: blume-20030517-absyn
1994    Description:
1995    
1996    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
1997    
1998       This mainly affects the quality of error messages.  However, some
1999       of the code is now more straightforward than before. (Treatment of
2000       the above four constructs in translate.sml is much simpler than
2001       the "macro-expansion" that was going on before.  Plus, the mach-
2002       compiler no longer gets invoked just to be able to compile an
2003       if-expression.)
2004    
2005    2. The ErrorMsg.Error exception is now caught and absorbed by the
2006       interactive loop.
2007    
2008    ----------------------------------------------------------------------
2009    Name: Allen Leung
2010    Date: 2003/05/16 13:05:00 CDT
2011    Tag: leunga-20030516-cygwin-runtime
2012    Description:
2013    
2014        Ported the runtime system to cygwin, which uses the unix
2015    x86-unix bin files.  Missing/buggy features:
2016    
2017       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2018         the Cygwin library.
2019       o Ctrl-C handling may be flaky.
2020       o Windows system calls and Windows I/O are not supported.
2021    
2022       A new set of binfiles is located at:
2023    
2024         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2025    
2026       This is only needed for bootstrapping the cygwin version of smlnj.
2027       Other x86 versions can use the existing binfiles.
2028    
2029    ----------------------------------------------------------------------
2030    Name: Matthias Blume
2031    Date: 2003/04/08 15:42:00 CDT
2032    Tag: blume-20030408-listpair
2033    Description:
2034    
2035    1. Added a target 'mlrisc' to installer.
2036    
2037    2. Added missing elements to structure ListPair.
2038    
2039    ----------------------------------------------------------------------
2040    Name: Allen Leung
2041    Date: 2003/01/07 10:40:00 EST
2042    Tag: leunga-20030107-int-rem
2043    Description:
2044    
2045    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2046    The arguments to the SUBL instruction were swapped.
2047    
2048    ----------------------------------------------------------------------
2049    Name: Matthias Blume
2050    Date: 2002/12/12 16:25:00 EST
2051    Tag: blume-20021212-risc-ra
2052    Description:
2053    
2054    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2055    sent the RA into an infinite loop when floating point registers get
2056    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2057    110.37 and 110.38.)
2058    There was another set of potential problems related to the handling of
2059    MLRISC annotations (but those did not yet cause real problems, apparently).
2060    
2061    ----------------------------------------------------------------------
2062    Name: Matthias Blume
2063    Date: 2002/12/06 22:40:00 EST
2064    Tag: blume-20021206-cm-fileid
2065    Description:
2066    
2067    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2068    This fixes the problem of CM getting confused by files that suddenly
2069    change their identity (e.g., by getting unlinked and recreated by some
2070    text editor such as vi).  There might be a better/cheaper/cleaner way
2071    of doing this, but for now this will have to do.
2072    
2073    ----------------------------------------------------------------------
2074    Name: Matthias Blume
2075    Date: 2002/10/28 09:50:00 EST
2076    Tag: blume-20021028-typecheck
2077    Description:
2078    
2079    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2080    
2081    ----------------------------------------------------------------------
2082    Name: Matthias Blume
2083    Date: 2002/10/17 09:10:00 EDT
2084    Tag: Release_110_42
2085    Description:
2086    
2087    In good old tradition, there has been a slight hiccup so that we have
2088    to patch 110.42 after the fact.   The old release tag has been replaced
2089    (see below).
2090    
2091    The change solves a problem with two competing approaches the
2092    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2093    each other's way.
2094    
2095    This change only affects the runtime system code and the installer script.
2096    (No new bootfiles.)
2097    
2098    ----------------------------------------------------------------------
2099    Name: Matthias Blume
2100    Date: 2002/10/16 12:00:00 EDT
2101    Tag: Release_110_42_removed
2102    Description:
2103    
2104    New working release.  New bootfiles.
2105    
2106    ----------------------------------------------------------------------
2107    Name: Matthias Blume
2108    Date: 2002/10/10 13:10:00 EDT
2109    Tag: blume-20021010-ppc-divs
2110    Description:
2111    
2112    The mltree operator DIVS must be implemented with an overflow check on
2113    the PPC because the hardware indicates divide-by-zero using "overflow" as
2114    well.
2115    
2116    ----------------------------------------------------------------------
2117    Name: Matthias Blume
2118    Date: 2002/07/23 11:45:00
2119    Tag: blume-20020723-smlnj-home
2120    Description:
2121    
2122    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2123    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2124    CM_PATHCONFIG is also set), the path configuration file is assumed
2125    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2126    the entire tree to some other place and everything will "just work".
2127    
2128    (Companion commands such as ml-build and ml-makedepend also sense this
2129    variable.)
2130    
2131    ----------------------------------------------------------------------
2132    Name: Matthias Blume
2133    Date: 2002/07/12 21:19:00 EDT
2134    Tag: blume-20020712-liveness
2135    Description:
2136    
2137    Exported two useful "step" functions from liveness module (MLRISC).
2138    
2139    ----------------------------------------------------------------------
2140    Name: Matthias Blume
2141    Date: 2002/07/05 16:00 EDT
2142    Tag: Release_110_41
2143    Description:
2144    
2145    New working release.  New bootfiles.
2146    
2147    ----------------------------------------------------------------------
2148    Name: Matthias Blume
2149    Date: 2002/07/05 10:25:00 EDT
2150    Tag: blume-20020705-btimp
2151    Description:
2152    
2153    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2154    other clients can set up backtracing support.
2155    
2156    ----------------------------------------------------------------------
2157    Name: Matthias Blume
2158    Date: 2002/06/25 17:23:00 EDT
2159    Tag: blume-20020625-fpmax
2160    Description:
2161    
2162    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2163    case.  (The sense of the isNaN test was reversed -- which made min and
2164    max always return their first argument.)
2165    
2166    ----------------------------------------------------------------------
2167    Name: Matthias Blume
2168    Date: 2002/06/11
2169    Tag: blume-20020611-unixpath
2170    Description:
2171    
2172    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2173    
2174    ----------------------------------------------------------------------
2175    Name: Matthias Blume
2176    Date: 2002/06/10 16:35:00 EDT
2177    Tag: blume-20020610-ieeereal
2178    Description:
2179    
2180    I back-ported my implementation of IEEEReal.fromString from the
2181    idlbasis-devel branch so that we can test it.
2182    
2183    Another small change is that ppDec tries to give more information
2184    than just "<sig>" in the case of functors.  However, this code is
2185    broken in some mysterious way if the functor's body's signature
2186    has not been declared by ascription but gets inferred from the
2187    implementation.  This needs fixing...
2188    
2189    ----------------------------------------------------------------------
2190    Name: Matthias Blume
2191    Date: 2002/05/31
2192    Tag: blume-20020531-btrace-mode
2193    Description:
2194    
2195    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2196    the wayside when I switched over to using Controls everywhere.)
2197    
2198    ----------------------------------------------------------------------
2199    Name: Lal George
2200    Date: 2002/05/23 12:21:40 EDT
2201    Tag: george-20020523-visual-labels
2202    Description:
2203    
2204    Labels are now displayed in the graphical output to make
2205    the fall-through and target blocks obvious.
2206    
2207    ----------------------------------------------------------------------
2208    Name: Matthias Blume
2209    Date: 2002/05/22 11:03:00 EDT
2210    Tag: blume-20020522-shrink
2211    Description:
2212    
2213    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2214    situation that comes up when allocating huge arrays.
2215    
2216    ----------------------------------------------------------------------
2217    Name: Matthias Blume
2218    Date: 2002/05/21 16:00:00 EDT
2219    Tag: Release_110_40
2220    Description:
2221    
2222    New working release (110.40).  New bootfiles.
2223    
2224    [Also: John Reppy fixed GC bug 1131.]
2225    
2226    ----------------------------------------------------------------------
2227    Name: Matthias Blume
2228    Date: 2002/05/21 12:35:00 EDT
2229    Tag: blume-20020521-cmdoc
2230    Description:
2231    
2232    CM documentation update.
2233    
2234    ----------------------------------------------------------------------
2235    Name: Matthias Blume
2236    Date: 2002/05/21 10:55:00 EDT
2237    Tag: blume-20020521-misc
2238    Description:
2239    
2240    - John tweaked runtime to be silent on heap export (except when
2241      GC messages are on).
2242    - I added a few more things (cross-compiling versions of CMB) to
2243      config/preloads (as suggestions).
2244    
2245    ----------------------------------------------------------------------
2246    Name: Matthias Blume
2247    Date: 2002/05/20 22:25:00 EDT
2248    Tag: blume-20020520-controls
2249    Description:
2250    
2251    - Added ControlUtil structure to control-lib.cm.
2252    - Use it throughout.
2253    - Used Controls facility to define MLRISC controls (as opposed to
2254      registering MLRISC control ref cells with Controls after the fact)
2255    - Fixed messed-up controls priorities.
2256    
2257    * Removed again all the stuff from config/preloads that one wouldn't
2258      be able to preload at the time the initial heap image is built.
2259      (Many libraries, e.g., CML, do not exist yet at this time.  The
2260       only libraries that can be preloaded via config/preloads are those
2261       that come bundled with the bootfiles.)
2262    
2263    ----------------------------------------------------------------------
2264    Name: Matthias Blume
2265    Date: 2002/05/20 10:59:00 EDT
2266    Tag: blume-20020520-preloads
2267    Description:
2268    
2269    Added a lot of commented-out suggestions for things to be included
2270    in config/preloads.
2271    
2272    ----------------------------------------------------------------------
2273    Name: Allen Leung
2274    Date: 2002/05/18 14:20:00 EDT
2275    Tag: leunga-20020518-mdl
2276    Description:
2277    
2278    o Made the mdl tool stuff compile and run again.
2279    o I've disabled all the stuff that depends on RTL specifications; they
2280      are all badly broken anyway.
2281    
2282    ----------------------------------------------------------------------
2283    Name: Matthias Blume
2284    Date: 2002/05/17 16:49:00 EDT
2285    Tag: blume-20020517-controls
2286    Description:
2287    
2288    0. John Reppy made several modifications to the SML/NJ library.
2289       In particular, there is a shiny new controls-lib.cm.
2290    
2291    1. Pushed new controls interface through compiler so that everything
2292       compiles again.
2293    
2294    2. Added FormatComb and FORMAT_COMB to the CML version of the
2295       SML/NJ library (so that CML compiles again).
2296    
2297    3. Modified init scripts because XXX_DEFAULT environment variables
2298       are no longer with us.  (Boot-time initialization is now done
2299       using the same environment variables that are also used for
2300       startup-time initialization of controls.)
2301    
2302    ----------------------------------------------------------------------
2303    Name: Lal George
2304    Date: 2002/05/15 09:20:10 EDT
2305    Tag: george-20020515-pseudo-op-decls
2306    Description:
2307    
2308            All pseudo-ops emitted before the first segment declaration
2309    such as TEXT, DATA, and BSS directives are assumed to be global
2310    declarations and are emitted first in the assembly file. This is
2311    useful in a number of situations where one has pseudo-ops that are not
2312    specific to any segment, and also works around the constraint that one
2313    cannot have client pseudo-ops in the TEXT segment.
2314    
2315            Because no segment is associated with these declarations it is
2316    an error to allocate any space or objects before the first segment
2317    directive and an exception will be raised. However, we cannot make
2318    this check for client pseudo-ops.
2319    
2320            These top level declarations are a field in the CFG graph_info.
2321    In theory you can continue to add to this field after the CFG has been
2322    built -- provided you know what you are doing;-)
2323    
2324    ----------------------------------------------------------------------
2325    Name: Matthias Blume
2326    Date: 2002/05/13 16:40:00 EDT
2327    Tag: blume-20020513-pp-etc
2328    Description:
2329    
2330    A few minor bugfixes:
2331    
2332      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2333        (Bogus printouts for redefined bindings still occur.  Compiler
2334        bug should no longer occur now.  We need to redo the prettyprinter
2335        from scratch.)
2336    
2337      - CM pathname printer now also adds escape sequences for ( and )
2338    
2339      - commend and docu fixes for ml-nlffi
2340    
2341    ----------------------------------------------------------------------
2342    Name: Matthias Blume
2343    Date: 2002/05/10 16:40:00 EDT
2344    Tag: blume-20020510-erg-textio
2345    Description:
2346    
2347    Applied the following bugfix provided by Emden Gansner:
2348    
2349        Output is corrupted when outputSubstr is used rather than output.
2350    
2351        The problem occurs when a substring
2352    
2353            ss = (s, dataStart, dataLen)
2354    
2355        where dataStart > 0, fills a stream buffer with avail bytes left.
2356        avail bytes of s, starting at index dataStart, are copied into the
2357        buffer, the buffer is flushed, and then the remaining dataLen-avail
2358        bytes of ss are copied into the beginning of the buffer. Instead of
2359        starting this copy at index dataStart+avail in s, the current code
2360        starts the copy at index avail.
2361    
2362        Fix:
2363        In text-io-fn.sml, change line 695 from
2364             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2365        to
2366             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2367    
2368    ----------------------------------------------------------------------
2369    Name: Matthias Blume
2370    Date: 2002/04/12 13:55:00 EDT
2371    Tag: blume-20020412-assyntax
2372    Description:
2373    
2374    1. Grabbed newer assyntax.h from the XFree86 project.
2375    2. Fiddled with how to compile X86.prim.asm without warnings.
2376    3. (Very) Minor cleanup in CM.
2377    
2378    ----------------------------------------------------------------------
2379    Name: Matthias Blume
2380    Date: 2002/04/01 (no joke!) 17:07:00 EST
2381    Tag: blume-20020401-x86div
2382    Description:
2383    
2384    Added full support for div/mod/rem/quot on the x86, using the machine
2385    instruction's two results (without clumsily recomputing the remainder)
2386    directly where appropriate.
2387    
2388    Some more extensive power-of-two support was added to the x86 instruction
2389    selector (avoiding expensive divs, mods, and muls where they can be
2390    replaced with cheaper shifts and masks).  However, this sort of thing
2391    ought to be done earlier, e.g., within the CPS optimizer so that
2392    all architectures benefit from it.
2393    
2394    The compiler compiles to a fixed point, but changes might be somewhat
2395    fragile nevertheless.  Please, report any strange things that you might
2396    see wrt. div/mod/quot/rem...
2397    
2398    ----------------------------------------------------------------------
2399    Name: Matthias Blume
2400    Date: 2002/03/29 17:22:00
2401    Tag: blume-20020329-div
2402    Description:
2403    
2404    Fixed my broken div/mod logic.  Unfortunately, this means that the
2405    inline code for div/mod now has one more comparison than before.
2406    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2407    The problem was with quotient = 0, because that alone does not tell
2408    us which way the rounding went.  One then has to look at whether
2409    remainder and divisor have the same sign...  :(
2410    
2411    Anyway, I replaced the bootfiles with fresh ones...
2412    
2413    ----------------------------------------------------------------------
2414    Name: Matthias Blume
2415    Date: 2002/03/29 14:10:00 EST
2416    Tag: blume-20020329-inlprims
2417    Description:
2418    
2419    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2420    
2421    Primops have changed. This means that the bin/boot-file formats have
2422    changed as well.
2423    
2424    To make sure that there is no confusion, I made a new version.
2425    
2426    
2427    CHANGES:
2428    
2429    * removed REMT from mltree (remainder should never overflow).
2430    
2431    * added primops to deal with divisions of all flavors to the frontend
2432    
2433    * handled these primops all the way through so they map to their respective
2434      MLRISC support
2435    
2436    * used these primops in the implementation of Int, Int32, Word, Word32
2437    
2438    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2439    
2440    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2441    
2442    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2443      floating point
2444    
2445    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2446      and Real (but Real.abs maps to a separate floating-point-only primop)
2447    
2448    
2449    TODO items:
2450    
2451    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2452      instructions because the machine instruction encoder cannot handle
2453      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2454      REM in the first place, and REM is supported by the x86 machine coder.)
2455    
2456    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2457      selection phase.  (The two can be streamlined because the hardware
2458      delivers both quotient and remainder at the same time anyway.)
2459    
2460    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2461      (Currently the behavior is inconsistent both across architectures and
2462      wrt. the draft Basis spec.)
2463    
2464    * Word8 should eventually be handled natively, too.
2465    
2466    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2467      as if there currently is no execution path that could trigger it in
2468      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2469      hold for things like multiplication and division.)
2470    
2471    ----------------------------------------------------------------------
2472    Name: Matthias Blume
2473    Date: 2002/03/27 16:27:00 EST
2474    Tag: blume-20020327-mlrisc-divisions
2475    Description:
2476    
2477    Added support for all four division operations (ML's div, mod, quot,
2478    and rem) to MLRISC.  In the course of doing so, I also rationalized
2479    the naming (no more annoying switch-around of DIV and QUOT), by
2480    parameterizing the operation by div_rounding_mode (which can be either
2481    DIV_TO_ZERO or DIV_TO_NEGINF).
2482    
2483    The generic MLTreeGen functor takes care of compiling all four
2484    operations down to only round-to-zero div.
2485    
2486    Missing pieces:
2487    
2488      * Doing something smarter than relying on MLTreeGen on architectures
2489        like, e.g., the x86 where hardware division delivers both quotient and
2490        remainder at the same time.  With this, the implementation of the
2491        round-to-neginf operations could be further streamlined.
2492    
2493      * Remove inlining support for div/mod/rem from the frontend and replace it
2494        with primops that get carried through to the backend.  Do this for all
2495        int and word types.
2496    
2497    ----------------------------------------------------------------------
2498    Name: Matthias Blume
2499    Date: 2002/03/25 17:25:00 EST
2500    Tag: blume-20020325-divmod
2501    Description:
2502    
2503    I improved (hopefully without breaking them) the implementation of Int.div,
2504    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2505    advantage of the following observations:
2506    
2507      Let  q = x quot y      r = x rem y
2508           d = x div  y      m = x mod y
2509    
2510    where "quot" is the round-to-zero version of integer division that
2511    hardware usually provides.  Then we have:
2512    
2513         r = x - q * y        where neither the * nor the - will overflow
2514         d = if q >= 0 orelse x = q * y then q else q - 1
2515                              where neither the * nor the - will overflow
2516         m = if q >= 0 orelse r = 0 then r else r + y
2517                              where the + will not overflow
2518    
2519    This results in substantial simplification of the generated code.
2520    The following table shows the number of CFG nodes and edges generated
2521    for
2522            fun f (x, y) = x OPER y
2523            (* with OPER \in div, mod, quot, rem *)
2524    
2525    
2526        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2527        --------------------------------------------------------
2528         div |         24 |         39 |         12 |         16
2529         mod |         41 |         71 |         12 |         16
2530        quot |          8 |         10 |          8 |         10
2531         rem |         10 |         14 |          8 |         10
2532    
2533    
2534    ----------------------------------------------------------------------
2535    Name: Matthias Blume
2536    Date: 2002/03/25 22:06:00 EST
2537    Tag: blume-20020325-cprotobug
2538    Description:
2539    
2540    Fixed a bug in cproto (c prototype decoder).
2541    
2542    ----------------------------------------------------------------------
2543    Name: Matthias Blume
2544    Date: 2002/03/25 16:00:00 EST
2545    Tag: blume-20020325-raw-primops
2546    Description:
2547    
2548    I did some cleanup to Allen's new primop code and
2549    replaced yesterday's bootfiles with new ones.
2550    (But they are stored in the same place.)
2551    
2552    ----------------------------------------------------------------------
2553    Name: Matthias Blume
2554    Date: 2002/03/24 22:40:00 EST
2555    Tag: blume-20020324-bootfiles
2556    Description:
2557    
2558    Made the bootfiles that Allen asked for.
2559    
2560    ----------------------------------------------------------------------
2561    Name: Allen Leung
2562    Date: 2002/03/23 15:50:00 EST
2563    Tag: leunga-20020323-flint-cps-rcc-primops
2564    Description:
2565    
2566      1. Changes to FLINT primops:
2567    
2568        (* make a call to a C-function;
2569         * The primop carries C function prototype information and specifies
2570         * which of its (ML-) arguments are floating point. C prototype
2571         * information is for use by the backend, ML information is for
2572         * use by the CPS converter. *)
2573      | RAW_CCALL of { c_proto: CTypes.c_proto,
2574                       ml_args: ccall_type list,
2575                       ml_res_opt: ccall_type option,
2576                       reentrant : bool
2577                     } option
2578       (* Allocate uninitialized storage on the heap.
2579        * The record is meant to hold short-lived C objects, i.e., they
2580        * are not ML pointers.  With the tag, the representation is
2581        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2582        *)
2583      | RAW_RECORD of {tag:bool,sz:int}
2584      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2585    
2586      2.  These CPS primops are now overloaded:
2587    
2588           rawload of {kind:numkind}
2589           rawstore of {kind:numkind}
2590    
2591          The one argument form is:
2592    
2593             rawload {kind} address
2594    
2595          The two argument form is:
2596    
2597             rawload {kind} [ml object, byte-offset]
2598    
2599      3. RAW_CCALL/RCC now takes two extra arguments:
2600    
2601         a. The first is whether the C call is reentrant, i.e., whether
2602            ML state should be saved and restored.
2603         b. The second argument is a string argument specifying the name of
2604            library and the C function.
2605    
2606         These things are currently not handled in the code generator, yet.
2607    
2608      4. In CProto,
2609    
2610         An encoding type of "bool" means "ml object" and is mapped into
2611         C prototype of PTR.  Note that "bool" is different than "string",
2612         even though "string" is also mapped into PTR, because "bool"
2613         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2614    
2615      5. Pickler/unpicker
2616    
2617         Changed to handle RAW_RECORD and newest RAW_CCALL
2618    
2619      6. MLRiscGen,
2620    
2621         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2622         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2623            in the file CodeGen/cpscompile/cps-c-calls.sml
2624    
2625      7. Added the conditional move operator
2626    
2627             condmove of branch
2628    
2629         to cps.  Generation of this is still buggy so it is currently
2630         disabled.
2631    
2632    ----------------------------------------------------------------------
2633    Name: Lal George
2634    Date: 2002/03/22 14:18:25 EST
2635    Tag: george-20020322-cps-branch-prob
2636    Description:
2637    
2638    Implemented the Ball-Larus branch prediction-heuristics, and
2639    incorporated graphical viewers for control flow graphs.
2640    
2641    Ball-Larus Heuristics:
2642    ---------------------
2643    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2644    
2645    By design it uses the Dempster-Shafer theory for combining
2646    probabilities.  For example, in the function:
2647    
2648        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2649    
2650    the ball-larus heuristics predicts that the n=0 is unlikely
2651    (OH-heuristic), and the 'then' branch is unlikely because of the
2652    RH-heuristic -- giving the 'then' branch an even lower combined
2653    probability using the Dempster-Shafer theory.
2654    
2655    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2656    probability of the 'then' branch because of the loop in the else
2657    branch.
2658    
2659    
2660    Graphical Viewing:
2661    ------------------
2662    I merely plugged in Allen's graphical viewers into the compiler. The
2663    additional code is not much. At the top level, saying:
2664    
2665            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2666    
2667    will display the graphical view of the control flow graph just before
2668    back-patching.  daVinci must be in your path for this to work. If
2669    daVinci is not available, then the default viewer can be changed
2670    using:
2671    
2672            Control.MLRISC.getString "viewer"
2673    
2674    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2675    course, these viewers must be in your path.
2676    
2677    The above will display the compilation unit at the level of clusters,
2678    many of which are small, boring, and un-interesting. Also setting:
2679    
2680            Control.MLRISC.getInt "cfg-graphical-view_size"
2681    
2682    will display clusters that are larger than the value set by the above.
2683    
2684    
2685    ----------------------------------------------------------------------
2686    Name: Matthias Blume
2687    Date: 2002/03/21 22:20:00 EST
2688    Tag: blume-20020321-kmp-bugfix
2689    Description:
2690    
2691    Changed the interface to the KMP routine in PreString and fixed
2692    a minor bug in one place where it was used.
2693    
2694    ----------------------------------------------------------------------
2695    Name: Allen Leung
2696    Date: 2002/03/21 20:30:00 EST
2697    Tag: leunga-20020321-cfg
2698    Description:
2699    
2700      Fixed a potential problem in cfg edge splitting.
2701    
2702    ----------------------------------------------------------------------
2703    Name: Allen Leung
2704    Date: 2002/03/21 17:15:00 EST
2705    Tag: leunga-20020321-x86-fp-cfg
2706    Description:
2707    
2708      1. Recoded the buggy parts of x86-fp.
2709    
2710         a. All the block reordering code has been removed.
2711            We now depend on the block placement phases to do this work.
2712    
2713         b. Critical edge splitting code has been simplified and moved into the
2714            CFG modules, as where they belong.
2715    
2716         Both of these were quite buggy and complex.  The code is now much, much
2717         simpler.
2718    
2719      2. X86 backend.
2720    
2721         a. Added instructions for 64-bit support.  Instruction selection for
2722            64-bit has not been committed, however, since that
2723            requires changes to MLTREE which haven't been approved by
2724            Lal and John.
2725    
2726         b. Added support for FUCOMI and FUCOMIP when generating code for
2727            PentiumPro and above.  We only generate these instructions in
2728            the fast-fp mode.
2729    
2730         c. Added cases for JP and JNP in X86FreqProps.
2731    
2732      3. CFG
2733    
2734         CFG now has a bunch of methods for edge splitting and merging.
2735    
2736      4. Machine description.
2737    
2738         John's simplification of MLTREE_BASIS.fcond broke a few machine
2739         description things:
2740    
2741         rtl-build.{sig,sml} and hppa.mdl fixed.
2742    
2743         NOTE: the machine description stuff in the repository is still broken.
2744               Again, I can't put my fixes in because that involves
2745               changes to MLTREE.
2746    
2747    ----------------------------------------------------------------------
2748    Name: Matthias Blume
2749    Date: 2002/03/20 15:55:00 EST
2750    Tag: blume-20020320-kmp
2751    Description:
2752    
2753    Implemented Knuth-Morris-Pratt string matching in PreString and used
2754    it for String.isSubstring, Substring.isSubstring, and
2755    Substring.position.
2756    
2757    (Might need some stress-testing.  Simple examples worked fine.)
2758    
2759    ----------------------------------------------------------------------
2760    Name: Matthias Blume
2761    Date: 2002/03/19 16:37:00 EST
2762    Tag: blume-20020319-witnesses
2763    Description:
2764    
2765    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2766    
2767    This implements a generic mechanism for changing constness qualifiers
2768    anywhere within big C types without resorting to outright "casts".
2769    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2770    modify the constness at the outermost level.)
2771    The implementation of "convert" is based on the idea of "witness"
2772    values -- values that are not used by the operation but whose types
2773    "testify" to their applicability.  On the implementation side, "convert"
2774    is simply a projection (returning its second curried argument).  With
2775    cross-module inlining, it should not result in any machine code being
2776    generated.
2777    
2778    ----------------------------------------------------------------------
2779    Name: Matthias Blume
2780    Date: 2002/03/15 16:40:00 EST
2781    Tag: blume-20020315-basis
2782    Description:
2783    
2784    Provided (preliminary?) implementations for
2785    
2786      {String,Substring}.{concatWith,isSuffix,isSubstring}
2787    
2788    and
2789    
2790      Substring.full
2791    
2792    Those are in the Basis spec but they were missing in SML/NJ.
2793    
2794    ----------------------------------------------------------------------
2795    Name: Matthias Blume
2796    Date: 2002/03/14 21:30:00 EST
2797    Tag: blume-20020314-controls
2798    Description:
2799    
2800    Controls:
2801    ---------
2802    
2803    1. Factored out the recently-added Controls : CONTROLS stuff and put
2804       it into its own library $/controls-lib.cm.  The source tree for
2805       this is under src/smlnj-lib/Controls.
2806    
2807    2. Changed the names of types and functions in this interface, so they
2808       make a bit more "sense":
2809    
2810          module -> registry
2811          'a registry -> 'a group
2812    
2813    3. The interface now deals in ref cells only.  The getter/setter interface
2814       is (mostly) gone.
2815    
2816    4. Added a function that lets one register an already-existing ref cell.
2817    
2818    5. Made the corresponding modifications to the rest of the code so that
2819       everything compiles again.
2820    
2821    6. Changed the implementation of Controls.MLRISC back to something closer
2822       to the original.  In particular, this module (and therefore MLRISC)
2823       does not depend on Controls.  There now is some link-time code in
2824       int-sys.sml that registers the MLRISC controls with the Controls
2825       module.
2826    
2827    CM:
2828    ---
2829    
2830      * One can now specify the lambda-split aggressiveness in init.cmi.
2831    
2832    ----------------------------------------------------------------------
2833    Name: Allen Leung
2834    Date: 2002/03/13 17:30:00 EST
2835    Tag: leunga-20020313-x86-fp-unary
2836    Description:
2837    
2838    Bug fix for:
2839    
2840    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
2841    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
2842    > - fun f(x,(y,z)) = Real.~ y;
2843    > [autoloading]
2844    > [autoloading done]
2845    >       fchsl   (%eax), 184(%esp)
2846    > Error: MLRisc bug: X86MCEmitter.emitInstr
2847    >
2848    > uncaught exception Error
2849    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
2850    
2851    The problem was that the code generator did not generate any fp registers
2852    in this case, and the ra didn't know that it needed to run the X86FP phase to
2853    translate the pseudo fp instruction.   This only happened with unary fp
2854    operators in certain situations.
2855    
2856    ----------------------------------------------------------------------
2857    Name: Matthias Blume
2858    Date: 2002/03/13 14:00:00 EST
2859    Tag: blume-20020313-overload-etc
2860    Description:
2861    
2862    1. Added _overload as a synonym for overload for backward compatibility.
2863       (Control.overloadKW must be true for either version to be accepted.)
2864    
2865    2. Fixed bug in install script that caused more things to be installed
2866       than what was requested in config/targets.
2867    
2868    3. Made CM aware of the (_)overload construct so that autoloading
2869       works.
2870    
2871    ----------------------------------------------------------------------
2872    Name: Matthias Blume
2873    Date: 2002/03/12 22:03:00 EST
2874    Tag: blume-20020312-url
2875    Description:
2876    
2877    Forgot to update BOOT and srcarchiveurl.
2878    
2879    ----------------------------------------------------------------------
2880    Name: Matthias Blume
2881    Date: 2002/03/12 17:30:00 EST
2882    Tag: blume-20020312-version110392
2883    Description:
2884    
2885    Yet another version number bump (because of small changes to the
2886    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
2887    
2888    Changes:
2889    
2890      The new pid generation scheme described a few weeks ago was overly
2891      complicated.  I implemented a new mechanism that is simpler and
2892      provides a bit more "stability":  Once CM has seen a compilation
2893      unit, it keeps its identity constant (as long as you do not delete
2894      those crucial CM/GUID/* files).  This means that when you change
2895      an interface, compile, then go back to the old interface, and
2896      compile again, you arrive at the original pid.
2897    
2898      There now also is a mechanism that instructs CM to use the plain
2899      environment hash as a module's pid (effectively making its GUID
2900      the empty string).  For this, "noguid" must be specified as an
2901      option to the .sml file in question within its .cm file.
2902      This is most useful for code that is being generated by tools such
2903      as ml-nlffigen (because during development programmers tend to
2904      erase the tool's entire output directory tree including CM's cached
2905      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
2906      revert to the old, broken behavior of SML/NJ, but in specific cases
2907      where there is no danger of interface confusion, its use is ok
2908      (I think).
2909    
2910      ml-nlffigen by default generates "noguid" annotations.  They can be
2911      turned off by specifying -guid in its command line.
2912    
2913    ----------------------------------------------------------------------
2914    Name: Lal George
2915    Date: 2002/03/12 12 14:42:36 EST
2916    Tag: george-20020312-frequency-computation
2917    Description:
2918    
2919    Integrated jump chaining and static block frequency into the
2920    compiler. More details and numbers later.
2921    
2922    ----------------------------------------------------------------------
2923    Name: Lal George
2924    Date: 2002/03/11 11 22:38:53 EST
2925    Tag: george-20020311-jump-chain-elim
2926    Description:
2927    
2928    Tested the jump chain elimination on all architectures (except the
2929    hppa).  This is on by default right now and is profitable for the
2930    alpha and x86, however, it may not be profitable for the sparc and ppc
2931    when compiling the compiler.
2932    
2933    The gc test will typically jump to a label at the end of the cluster,
2934    where there is another jump to an external cluster containing the actual
2935    code to invoke gc. This is to allow factoring of common gc invocation
2936    sequences. That is to say, we generate:
2937    
2938            f:
2939               testgc
2940               ja   L1      % jump if above to L1
2941    
2942            L1:
2943               jmp L2
2944    
2945    
2946    After jump chain elimination the 'ja L1' instructions is converted to
2947    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
2948    up being implemented in their long form (if L2 is far away) using:
2949    
2950            jbe     L3      % jump if below or equal to L3
2951            jmp     L2
2952         L3:
2953            ...
2954    
2955    
2956    For large compilation units L2  may be far away.
2957    
2958    
2959    ----------------------------------------------------------------------
2960    Name: Matthias Blume
2961    Date: 2002/03/11 13:30:00 EST
2962    Tag: blume-20020311-mltreeeval
2963    Description:
2964    
2965    A functor parameter was missing.
2966    
2967    ----------------------------------------------------------------------
2968    Name: Allen Leung
2969    Date: 2002/03/11 10:30:00 EST
2970    Tag: leunga-20020311-runtime-string0
2971    Description:
2972    
2973       The representation of the empty string now points to a
2974    legal null terminated C string instead of unit.  It is now possible
2975    to convert an ML string into C string with InlineT.CharVector.getData.
2976    This compiles into one single machine instruction.
2977    
2978    ----------------------------------------------------------------------
2979    Name: Allen Leung
2980    Date: 2002/03/10 23:55:00 EST
2981    Tag: leunga-20020310-x86-call
2982    Description:
2983    
2984       Added machine generation for CALL instruction (relative displacement mode)
2985    
2986    ----------------------------------------------------------------------
2987    Name: Matthias Blume
2988    Date: 2002/03/08 16:05:00
2989    Tag: blume-20020308-entrypoints
2990    Description:
2991    
2992    Version number bumped to 110.39.1.  NEW BOOTFILES!
2993    
2994    Entrypoints: non-zero offset into a code object where execution should begin.
2995    
2996    - Added the notion of an entrypoint to CodeObj.
2997    - Added reading/writing of entrypoint info to Binfile.
2998    - Made runtime system bootloader aware of entrypoints.
2999    - Use the address of the label of the first function given to mlriscGen
3000      as the entrypoint.  This address is currently always 0, but it will
3001      not be 0 once we turn on block placement.
3002    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3003      with entry points) from mlriscGen.
3004    
3005    ----------------------------------------------------------------------
3006    Name: Allen Leung
3007    Date: 2002/03/07 20:45:00 EST
3008    Tag: leunga-20020307-x86-cmov
3009    Description:
3010    
3011       Bug fixes for CMOVcc on x86.
3012    
3013       1. Added machine code generation for CMOVcc
3014       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3015       3. CMOVcc cannot have an immediate operand as argument.
3016    
3017    ----------------------------------------------------------------------
3018    Name: Matthias Blume
3019    Date: 2002/03/07 16:15:00 EST
3020    Tag: blume-20020307-controls
3021    Description:
3022    
3023    This is a very large but mostly boring patch which makes (almost)
3024    every tuneable compiler knob (i.e., pretty much everything under
3025    Control.* plus a few other things) configurable via both the command
3026    line and environment variables in the style CM did its configuration
3027    until now.
3028    
3029    Try starting sml with '-h' (or, if you are brave, '-H')
3030    
3031    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3032    implements the underlying generic mechanism.
3033    
3034    The interface to some of the existing such facilities has changed somewhat.
3035    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3036    (The getFoo interface is still there for backward-compatibility, but its
3037    use is deprecated.)
3038    
3039    The ml-build script passes -Cxxx=yyy command-line arguments through so
3040    that one can now twiddle the compiler settings when using this "batch"
3041    compiler.
3042    
3043    TODO items:
3044    
3045    We should go through and throw out all controls that are no longer
3046    connected to anything.  Moreover, we should go through and provide
3047    meaningful (and correct!) documentation strings for those controls
3048    that still are connected.
3049    
3050    Currently, multiple calls to Controls.new are accepted (only the first
3051    has any effect).  Eventually we should make sure that every control
3052    is being made (via Controls.new) exactly once.  Future access can then
3053    be done using Controls.acc.
3054    
3055    Finally, it would probably be a good idea to use the getter-setter
3056    interface to controls rather than ref cells.  For the time being, both
3057    styles are provided by the Controls module, but getter-setter pairs are
3058    better if thread-safety is of any concern because they can be wrapped.
3059    
3060    *****************************************
3061    
3062    One bug fix: The function blockPlacement in three of the MLRISC
3063    backpatch files used to be hard-wired to one of two possibilities at
3064    link time (according to the value of the placementFlag).  But (I
3065    think) it should rather sense the flag every time.
3066    
3067    *****************************************
3068    
3069    Other assorted changes (by other people who did not supply a HISTORY entry):
3070    
3071    1. the cross-module inliner now works much better (Monnier)
3072    2. representation of weights, frequencies, and probabilities in MLRISC
3073       changed in preparation of using those for weighted block placement
3074       (Reppy, George)
3075    
3076    ----------------------------------------------------------------------
3077    Name: Lal George
3078    Date: 2002/03/07 14:44:24 EST 2002
3079    Tag: george-20020307-weighted-block-placement
3080    
3081    Tested the weighted block placement optimization on all architectures
3082    (except the hppa) using AMPL to generate the block and edge frequencies.
3083    Changes were required in the machine properties to correctly
3084    categorize trap instructions. There is an MLRISC flag
3085    "weighted-block-placement" that can be used to enable weighted block
3086    placement, but this will be ineffective without block/edge
3087    frequencies (coming soon).
3088    
3089    
3090    ----------------------------------------------------------------------
3091    Name: Lal George
3092    Date: 2002/03/05 17:24:48 EST
3093    Tag: george-20020305-linkage-cluster
3094    
3095    In order to support the block placement optimization, a new cluster
3096    is generated as the very first cluster (called the linkage cluster).
3097    It contains a single jump to the 'real' entry point for the compilation
3098    unit. Block placement has no effect on the linkage cluster itself, but
3099    all the other clusters  have full freedom in the manner in which they
3100    reorder blocks or functions.
3101    
3102    On the x86 the typical linkage code that is generated is:
3103       ----------------------
3104            .align 2
3105       L0:
3106            addl    $L1-L0, 72(%esp)
3107            jmp     L1
3108    
3109    
3110            .align  2
3111       L1:
3112       ----------------------
3113    
3114    72(%esp) is the memory location for the stdlink register. This
3115    must contain the address of the CPS function being called. In the
3116    above example, it contains the address of  L0; before
3117    calling L1 (the real entry point for the compilation unit), it
3118    must contain the address for L1, and hence
3119    
3120            addl $L1-L0, 72(%esp)
3121    
3122    I have tested this on all architectures except the hppa.The increase
3123    in code size is of course negligible
3124    
3125    ----------------------------------------------------------------------
3126    Name: Allen Leung
3127    Date: 2002/03/03 13:20:00 EST
3128    Tag: leunga-20020303-mlrisc-tools
3129    
3130      Added #[ ... ] expressions to mlrisc tools
3131    
3132    ----------------------------------------------------------------------
3133    Name: Matthias Blume
3134    Date: 2002/02/27 12:29:00 EST
3135    Tag: blume-20020227-cdebug
3136    Description:
3137    
3138    - made types in structure C and C_Debug to be equal
3139    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3140    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3141    
3142    ----------------------------------------------------------------------
3143    Name: Matthias Blume
3144    Date: 2002/02/26 12:00:00 EST
3145    Tag: blume-20020226-ffi
3146    Description:
3147    
3148    1. Fixed a minor bug in CM's "noweb" tool:
3149       If numbering is turned off, then truly don't number (i.e., do not
3150       supply the -L option to noweb).  The previous behavior was to supply
3151       -L'' -- which caused noweb to use the "default" line numbering scheme.
3152       Thanks to Chris Richards for pointing this out (and supplying the fix).
3153    
3154    2. Once again, I reworked some aspects of the FFI:
3155    
3156       A. The incomplete/complete type business:
3157    
3158       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3159         gone!
3160       - ML types representing an incomplete type are now *equal* to
3161         ML types representing their corresponding complete types (just like
3162         in C).  This is still safe because ml-nlffigen will not generate
3163         RTTI for incomplete types, nor will it generate functions that
3164         require access to such RTTI.   But when ML code generated from both
3165         incomplete and complete versions of the C type meet, the ML types
3166         are trivially interoperable.
3167    
3168         NOTE:  These changes restore the full generality of the translation
3169         (which was previously lost when I eliminated functorization)!
3170    
3171       B. Enum types:
3172    
3173       - Structure C now has a type constructor "enum" that is similar to
3174         how the "su" constructor works.  However, "enum" is not a phantom
3175         type because each "T enum" has values (and is isomorphic to
3176         MLRep.Signed.int).
3177       - There are generic access operations for enum objects (using
3178         MLRep.Signed.int).
3179       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3180         * The structure contains the definition of type "mlrep" (the ML-side
3181         representation type of the enum).  Normally, mlrep is the same
3182         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3183         then mlrep will be defined as a datatype -- thus facilitating
3184         pattern matching on mlrep values.
3185         ("-ec" will be suppressed if there are duplicate values in an
3186          enumeration.)
3187         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3188         will be generated for each C enum constant xxx.
3189         * Conversion functions m2i and i2m convert between mlrep and
3190         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3191         * Coversion functions c and ml convert between mlrep and "tag enum".
3192         * Access functions (get/set) fetch and store mlrep values.
3193       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3194         enumerations are merged into one single enumeration represented by
3195         structure E_'.
3196    
3197    ----------------------------------------------------------------------
3198    Name: Allen Leung
3199    Date: 2002/02/25 04:45:00 EST
3200    Tag: leunga-20020225-cps-spill
3201    
3202    This is a new implementation of the CPS spill phase.
3203    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3204    In case of problems, replace it with the old file spill.sml
3205    
3206    The current compiler runs into some serious performance problems when
3207    constructing a large record.  This can happen when we try to compile a
3208    structure with many items.  Even a very simple structure like the following
3209    makes the compiler slow down.
3210    
3211        structure Foo = struct
3212           val x_1 = 0w1 : Word32.int
3213           val x_2 = 0w2 : Word32.int
3214           val x_3 = 0w3 : Word32.int
3215           ...
3216           val x_N = 0wN : Word32.int
3217        end
3218    
3219    The following table shows the compile time, from N=1000 to N=4000,
3220    with the old compiler:
3221    
3222    N
3223    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3224           MLRISC ra                               0.06u  0.00s  0.05g
3225              (spills = 0 reloads = 0)
3226           TOTAL                                   0.63u  0.07s  0.21g
3227    
3228    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3229           MLRISC ra                               5.68u  0.59s  3.93g
3230              (spills = 0 reloads = 0)
3231           TOTAL                                   14.71u  0.99s  4.81g
3232    
3233    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3234           MLRISC ra                               5.54u  0.65s  3.91g
3235              (spills = 543 reloads = 1082)
3236           TOTAL                                   65.40u  3.13s  6.00g
3237    
3238    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3239           MLRISC ra                               0.80u  0.10s  0.55g
3240              (spills = 42 reloads = 84)
3241           TOTAL                                   129.42u  5.10s  4.13g
3242    
3243    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3244           MLRISC ra                               2.69u  0.27s  1.38g
3245              (spills = 62 reloads = 124)
3246           TOTAL                                   682.48u  19.61s  13.99g
3247    
3248    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3249           MLRISC ra                               4.96u  0.27s  2.72g
3250              (spills = 85 reloads = 170)
3251           TOTAL                                   2375.26u  57.21s  48.00g
3252    
3253    As you can see the old cps spill module suffers from some serious
3254    performance problem.  But since I cannot decipher the old code fully,
3255    instead of patching the problems up, I'm reimplementing it
3256    with a different algorithm.  The new code is more modular,
3257    smaller when compiled, and substantially faster
3258    (O(n log n) time and O(n) space).  Timing of the new spill module:
3259    
3260    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3261          MLRISC ra                               0.25u  0.02s  0.15g
3262             (spills=1 reloads=3)
3263          TOTAL                                   7.74u  0.34s  1.62g
3264    
3265    Implementation details:
3266    
3267    As far as I can tell, the purpose of the CPS spill module is to make sure the
3268    number of live variables at any program point (the bandwidth)
3269    does not exceed a certain limit, which is determined by the
3270    size of the spill area.
3271    
3272    When the bandwidth is too large, we decrease the register pressure by
3273    packing live variables into spill records.  How we achieve this is
3274    completely different than what we did in the old code.
3275    
3276    First, there is something about the MLRiscGen code generator
3277    that we should be aware of:
3278    
3279    o MLRiscGen performs code motion!
3280    
3281       In particular, it will move floating point computations and
3282       address computations involving only the heap pointer to
3283       their use sites (if there is only a single use).
3284       What this means is that if we have a CPS record construction
3285       statement
3286    
3287           RECORD(k,vl,w,e)
3288    
3289       we should never count the new record address w as live if w
3290       has only one use (which is often the case).
3291    
3292       We should do something similar to floating point, but the transformation
3293       there is much more complex, so I won't deal with that.
3294    
3295    Secondly, there are now two new cps primops at our disposal:
3296    
3297     1. rawrecord of record_kind option
3298        This pure operator allocates some uninitialized storage from the heap.
3299        There are two forms:
3300    
3301         rawrecord NONE [INT n]  allocates a tagless record of length n
3302         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3303                                     and initializes the tag.
3304    
3305     2. rawupdate of cty
3306          rawupdate cty (v,i,x)
3307          Assigns to x to the ith component of record v.
3308          The storelist is not updated.
3309    
3310    We use these new primops for both spilling and increment record construction.
3311    
3312     1. Spilling.
3313    
3314        This is implemented with a linear scan algorithm (but generalized
3315        to trees).  The algorithm will create a single spill record at the
3316        beginning of the cps function and use rawupdate to spill to it,
3317        and SELECT or SELp to reload from it.  So both spills and reloads
3318        are fine-grain operations.  In contrast, in the old algorithm
3319        "spills" have to be bundled together in records.
3320    
3321        Ideally, we should sink the spill record construction to where
3322        it is needed.  We can even split the spill record into multiple ones
3323        at the places where they are needed.  But CPS is not a good
3324        representation for global code motion, so I'll keep it simple and
3325        am not attempting this.
3326    
3327     2. Incremental record construction (aka record splitting).
3328    
3329        Long records with many component values which are simulatenously live
3330        (recall that single use record addresses are not considered to
3331         be live) are constructed with rawrecord and rawupdate.
3332        We allocate space on the heap with rawrecord first, then gradually
3333        fill it in with rawupdate.  This is the technique suggested to me
3334        by Matthias.
3335    
3336        Some restrictions on when this is applicable:
3337        1. It is not a VECTOR record.  The code generator currently does not handle
3338           this case. VECTOR record uses double indirection like arrays.
3339        2. All the record component values are defined in the same "basic block"
3340           as the record constructor.  This is to prevent speculative
3341           record construction.
3342    
3343    ----------------------------------------------------------------------
3344    Name: Allen Leung
3345    Date: 2002/02/22 01:02:00 EST
3346    Tag: leunga-20020222-mlrisc-tools
3347    
3348    Minor bug fixes in the parser and rewriter
3349    
3350    ----------------------------------------------------------------------
3351    Name: Allen Leung
3352    Date: 2002/02/21 20:20:00 EST
3353    Tag: leunga-20020221-peephole
3354    
3355    Regenerated the peephole files.  Some contained typos in the specification
3356    and some didn't compile because of pretty printing bugs in the old version
3357    of 'nowhere'.
3358    
3359    ----------------------------------------------------------------------
3360    Name: Allen Leung
3361    Date: 2002/02/19 20:20:00 EST
3362    Tag: leunga-20020219-mlrisc-tools
3363    Description:
3364    
3365       Minor bug fixes to the mlrisc-tools library:
3366    
3367       1.  Fixed up parsing colon suffixed keywords
3368       2.  Added the ability to shut the error messages up
3369       3.  Reimplemented the pretty printer and fixed up/improved
3370           the pretty printing of handle and -> types.
3371       4.  Fixed up generation of literal symbols in the nowhere tool.
3372       5.  Added some SML keywords to to sml.sty
3373    
3374    ----------------------------------------------------------------------
3375    Name: Matthias Blume
3376    Date: 2002/02/19 16:20:00 EST
3377    Tag: blume-20020219-cmffi
3378    Description:
3379    
3380    A wild mix of changes, some minor, some major:
3381    
3382    * All C FFI-related libraries are now anchored under $c:
3383        $/c.cm      --> $c/c.cm
3384        $/c-int.cm  --> $c/internals/c-int.cm
3385        $/memory.cm --> $c/memory/memory.cm
3386    
3387    * "make" tool (in CM) now treats its argument pathname slightly
3388      differently:
3389        1. If the native expansion is an absolute name, then before invoking
3390           the "make" command on it, CM will apply OS.Path.mkRelative
3391           (with relativeTo = OS.FileSys.getDir()) to it.
3392        2. The argument will be passed through to subsequent phases of CM
3393           processing without "going native".  In particular, if the argument
3394           was an anchored path, then "make" will not lose track of that anchor.
3395    
3396    * Compiler backends now "know" their respective C calling conventions
3397      instead of having to be told about it by ml-nlffigen.  This relieves
3398      ml-nlffigen from one of its burdens.
3399    
3400    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3401    
3402    * Export C_DEBUG and C_Debug from $c/c.cm.
3403    
3404    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3405      subtyping relationship between incomplete pointers and their complete
3406      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3407      with the convention of instantiating 'o with ('t, 'c) obj whenever
3408      the pointer target type is complete.  In the incomplete case, 'o
3409      will be instantiated with some "'c iobj" -- a type obtained by
3410      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3411    
3412      Operations that work on both incomplete and complete pointer types are
3413      typed as taking an 'o ptr while operations that require the target to
3414      be known are typed as taking some ('t, 'c) obj ptr.
3415    
3416      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3417      where void is an eqtype without any values.  This makes it possible
3418      to work on voidptr values using functions meant to operate on light
3419      incomplete pointers.
3420    
3421    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3422      been vastly simplified.
3423    
3424    ----------------------------------------------------------------------
3425    Name: Matthias Blume
3426    Date: 2002/02/19 10:48:00 EST
3427    Tag: blume-20020219-pqfix
3428    Description:
3429    
3430    Applied Chris Okasaki's bug fix for priority queues.
3431    
3432    ----------------------------------------------------------------------
3433    Name: Matthias Blume
3434    Date: 2002/02/15 17:05:00
3435    Tag: Release_110_39
3436    Description:
3437    
3438    Last-minute retagging is becoming a tradition... :-(
3439    
3440    This is the working release 110.39.
3441    
3442    ----------------------------------------------------------------------
3443    Name: Matthias Blume
3444    Date: 2002/02/15 16:00:00 EST
3445    Tag: Release_110_39-orig
3446    Description:
3447    
3448    Working release 110.39.  New bootfiles.
3449    
3450    (Update: There was a small bug in the installer so it wouldn't work
3451    with all shells.  So I retagged. -Matthias)
3452    
3453    ----------------------------------------------------------------------
3454    Name: Matthias Blume
3455    Date: 2002/02/15 14:17:00 EST
3456    Tag: blume-20020215-showbindings
3457    Description:
3458    
3459    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3460    the latter can be useful for exploring what bindings are available at
3461    the interactive prompt.  (The first function returns only the list
3462    of symbols that are really bound, the second prints those but also the
3463    ones that CM's autoloading mechanism knows about.)
3464    
3465    ----------------------------------------------------------------------
3466    Name: Matthias Blume
3467    Date: 2002/02/15 12:08:00 EST
3468    Tag: blume-20020215-iptrs
3469    Description:
3470    
3471    Two improvements to ml-nlffigen:
3472    
3473      1. Write files only if they do not exist or if their current contents
3474         do not coincide with what's being written.  (That is, avoid messing
3475         with the time stamps unless absolutely necessary.)
3476    
3477      2. Implement a "repository" mechanism for generated files related
3478         to "incomplete pointer types".   See the README file for details.
3479    
3480    ----------------------------------------------------------------------
3481    Name: Matthias Blume
3482    Date: 2002/02/14 11:50:00 EST
3483    Tag: blume-20020214-quote
3484    Description:
3485    
3486    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3487    because of the new and improved tag generation scheme.  (Thanks to Allen
3488    Leung for pointing it out.)
3489    
3490    ----------------------------------------------------------------------
3491    Name: Lal George
3492    Date: 2002/02/14 09:55:27 EST 2002
3493    Tag: george-20020214-isabelle-bug
3494    Description:
3495    
3496    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3497    of Isabelle on the x86.
3498    
3499    From Allen:
3500    -----------
3501     I've found the problem:
3502    
3503         in ra-core.sml, I use the counter "blocked" to keep track of the
3504         true number of elements in the freeze queue.  When the counter goes
3505         to zero, I skip examining the queue.  But I've messed up the
3506         bookkeeping in combine():
3507    
3508             else ();
3509             case !ucol of
3510               PSEUDO => (if !cntv > 0 then
3511                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3512                                        ^^^^^^^^^^^^^^^^^^^^^^^
3513                      moveu := mergeMoveList(!movev, !moveu)
3514                     )
3515                  else ();
3516    
3517         combine() is called to coalesce two nodes u and v.
3518         I think I was thinking that if the move counts of u and v are both
3519         greater than zero then after they are coalesced then one node is
3520         removed from the freeze queue.  Apparently I was thinking that
3521         both u and v are of low degree, but that's clearly not necessarily true.
3522    
3523    
3524    02/12/2002:
3525        Here's the patch.  HOL now compiles.
3526    
3527        I don't know how this impact on performance (compile
3528        time or runtime).  This bug caused the RA (especially on the x86)
3529        to go thru the potential spill phase when there are still nodes on the
3530        freeze queue.
3531    
3532    
3533    
3534    
3535    ----------------------------------------------------------------------
3536    Name: Matthias Blume
3537    Date: 2002/02/13 22:40:00 EST
3538    Tag: blume-20020213-fptr-rtti
3539    Description:
3540    
3541    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3542    updates.
3543    
3544    ----------------------------------------------------------------------
3545    Name: Matthias Blume
3546    Date: 2002/02/13 16:41:00 EST
3547    Tag: blume-20020213-cmlpq
3548    Description:
3549    
3550    Added new priority queue export symbols (which have just been added to
3551    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3552    not compile and the installer would choke.)
3553    
3554    ----------------------------------------------------------------------
3555    Name: Matthias Blume
3556    Date: 2002/02/13 16:15:00 EST
3557    Tag: blume-20020213-various
3558    Description:
3559    
3560    1. More tweaks to ml-nlffigen:
3561    
3562       - better internal datastructures (resulting in slight speedup)
3563       - "-match" option requires exact match
3564       - "localized" gensym counters (untagged structs/unions nested within
3565         other structs/unions or within typedefs get a fresh counter; their
3566         tag will be prefixed by a concatenation of their parents' tags)
3567       - bug fixes (related to calculation of transitive closure of types
3568         to be included in the output)
3569    
3570    2. Minor Basis updates:
3571    
3572       - added implementations for List.collate and Option.app
3573    
3574    ----------------------------------------------------------------------
3575    Name: Matthias Blume
3576    Date: 2002/02/11 15:55:00 EST
3577    Tag: blume-20020211-gensym
3578    Description:
3579    
3580    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3581    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3582    names (ML structure names that correspond to unnamed C structs, unions,
3583    and enums), so that separate runs of ml-nlffigen do not clash.
3584    
3585    ----------------------------------------------------------------------
3586    Name: Matthias Blume
3587    Date: 2002/02/11 12:05:00 EST
3588    Tag: blume-20020211-gensml
3589    Description:
3590    
3591    A quick fix for a problem with GenSML (in the pgraph-util library):
3592    Make generation of toplevel "local" optional.  (Strictly speaking,
3593    signature definitions within "local" are not legal SML.)
3594    
3595    Other than that: updates to INSTALL and cm/TODO.
3596    
3597    ----------------------------------------------------------------------
3598    Name: Matthias Blume
3599    Date: 2002/02/08 15:00:00 EST
3600    Tag: blume-20020208-uniquepid
3601    Description:
3602    
3603    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3604    
3605    1. The installer (config/install.sh) has gotten smarter:
3606    
3607         - Configuration options are a bit easier to specify now
3608           (in config/targets).
3609         - Bug in recognizing .tar.bz2 files fixed.
3610         - Installer automatically resolves dependencies between
3611           configuration options (e.g., if you ask for eXene, you will
3612           also get cml -- regardless whether you asked for it or not).
3613         - Installer can run in "quieter mode" by setting the environment
3614           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3615           "completely silent", though.
3616         - Build HashCons library as part of smlnj-lib.
3617    
3618    2. A new scheme for assigning persistent identifiers to compilation
3619       units (and, by extension, to types etc.) has been put into place.
3620       This fixes a long-standing bug where types and even dynamic values
3621       can get internally confused, thereby compromising type safety
3622       (abstraction) and dynamic correctness.  See
3623    
3624         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3625    
3626       for an example of how things could go wrong until now.
3627    
3628       The downside of the new scheme is that pids are not quite as
3629       persistent as they used to be: CM will generate a fresh pid
3630       for every compilation unit that it thinks it sees for the first
3631       time.  That means that if you compile starting from a clean, fresh
3632       source tree at two different times, you end up with different
3633       binaries.
3634    
3635       Cutoff recompilation, however, has not been compromised because
3636       CM keeps pid information in special caches between runs.
3637    
3638    ----------------------------------------------------------------------
3639    Name: Lal George
3640    Date: 2002/02/07 15:34:13 EST 2002
3641    Tag: <none>
3642    Description:
3643    
3644    Compilers that generate assembly code may produce  global labels
3645    whose value is resolved at link time. The various peephole optimization
3646    modules did not take this in account.
3647    
3648    TODO. The Labels.addrOf function should really return an option
3649    type so that clients are forced to deal with this issue, rather
3650    than an exception being raised.
3651    
3652    ----------------------------------------------------------------------
3653    Name: Lal George
3654    Date: 2002/02/06 13:55:02 EST
3655    Tag: george-20020206-ra-breakup
3656    Description:
3657    
3658    1. A bug fix from Allen.
3659    
3660        A typo causes extra fstp %st(0)'s to be generated at compensation
3661        edges, which might cause stack underflow traps at runtime.  This
3662        occurs in fft where there are extraneous fstps right before the 'into'
3663        trap instruction (in this case they are harmless since none of the
3664        integers overflow.)
3665    
3666    2. Pulled out various utility modules that were embedded in the modules
3667       of the register allocator. I need these modules for other purposes, but
3668       they are not complete enough to put into a library (just yet).
3669    ----------------------------------------------------------------------
3670    Name: Matthias Blume
3671    Date: 2002/01/31 16:05:00 EST
3672    Tag: blume-20020131-sparc-ccalls
3673    Description:
3674    
3675    1. C-calls on Sparc needlessly allocated a huge chunk (96 bytes)
3676       of extra stack space by mistake.  Fixed.
3677    
3678    2. Bug in logic of handling of command-line options in ml-nlffigen fixed.
3679    
3680    ----------------------------------------------------------------------
3681    Name: Allen Leung
3682    Date: 2002/01/30
3683    Tag: leunga-20020130-nowhere-bug-fix
3684    Description:
3685    
3686       MLRISC bug fixes:
3687       1. Fixed a bindings computation bug in the 'nowhere' program generator tool.
3688       2. MachineInt.fromString was negating its value.
3689    
3690    ----------------------------------------------------------------------
3691    Name: Matthias Blume
3692    Date: 2002/01/29
3693    Tag: blume-20020129-INSTALL
3694    Description:
3695    
3696    - Added somewhat detailed installation instructions (file INSTALL).
3697    - Fixed curl-detection bug in config/install.sh.
3698    - It is now possible to select the URL getter using the URLGETTER
3699      environment variable:
3700    
3701          not set / "unknown"      --> automatic detection (script tries wget,
3702                                       curl, and lynx)
3703          "wget" / "curl" / "lynx" --> use the specified program (script "knows"
3704                                       how to properly invoke them)
3705          other                    --> use $URLGETTER directly, it must take
3706                                       precisely two command-line arguments
3707                                       (source URL and destination file name)
3708    
3709    ----------------------------------------------------------------------
3710    Name: Matthias Blume
3711    Date: 2002/01/28
3712    Tag: blume-20020128-sparc-ccalls
3713    Description:
3714    
3715    - Fixed problem with calculation of "used" registers in sparc-c-calls.
3716    - Make use of the allocParam argument in sparc-c-calls.
3717    
3718    ----------------------------------------------------------------------
3719    Name: Matthias Blume
3720    Date: 2002/01/28
3721    Tag: blume-20020128-allocParam
3722    Description:
3723    
3724    John Reppy:  Changes c-calls API to accept client-callback for
3725    allocating extra stack space.
3726    me: Corresponding changes to mlriscGen (using a dummy argument that
3727        does not change the current behavior).
3728    
3729    ----------------------------------------------------------------------
3730    Name: Matthias Blume
3731    Date: 2002/01/28 12:00:00
3732    Tag: Release_110_38
3733    Description:
3734    
3735    This time for real!!!
3736    
3737    ----------------------------------------------------------------------
3738    Name: Matthias Blume
3739    Date: 2002/01/28 10:56:00 EST
3740    Tag: blume-20020128-retraction
3741    Description:
3742    
3743    0. Retracted earlier 110.38.  (The Release_110_38 tag has been replaced
3744       with blume-Release_110_38-retracted.)
3745    
3746    1. Fixed a problem with incorrect rounding modes in real64.sml.
3747       (Thanks to Andrew Mccreight <andrew.mccreight@yale.edu>.)
3748    
3749    2. A bug in ml-nlffigen related to the handling of unnamed structs, unions,
3750       and enums fixed.  The naming of corresponding ML identifiers should
3751       now be consistent again.
3752    
3753    ----------------------------------------------------------------------
3754    Name: Allen Leung
3755    Date: 2002/01/27
3756    Tag: leunga-20020127-nowhere
3757    Description:
3758    
3759       Added a target called nowhere in the configuration scripts.
3760       Enabling this will build the MLRISC 'nowhere' tool (for translating
3761       programs with where-clauses into legal SML code) during installation.
3762    
3763    ----------------------------------------------------------------------
3764    Name: Matthias Blume
3765    Date: 2002/01/25 21:27:00 EST
3766    Tag: blume-Release_110_38-retracted
3767    Description:
3768    
3769    Call it a (working) release!  Version is 110.38. Bootfiles are ready.
3770    
3771    README will be added later.
3772    
3773    !!! NOTE:  Re-tagged as blume-Release_110_38-retracted. Original tag
3774    (Release_110_38) removed.  Reason: Last-minute bug fixes.
3775    
3776    ----------------------------------------------------------------------
3777    Name: Matthias Blume
3778    Date: 2002/01/25
3779    Tag: blume-20020125-ffi
3780    Description:
3781    
3782    A large number of tweaks and improvements to ml-nlffi-lib and
3783    ml-nlffigen:
3784    
3785       - ML represenation types have been streamlined
3786       - getter and setter functions work with concrete values, not abstract
3787         ones where possible
3788       - ml-nlffigen command line more flexible (see README file there)
3789       - some bugs have been fixed (hopefully)
3790    
3791    ----------------------------------------------------------------------
3792    Name: Lal George
3793    Date: 2002/01/24
3794    Tag: george-20020124-risc-ra-interface
3795    Description:
3796    
3797       There is a dramatic simplification in the interface to the
3798       register allocator for RISC architectures as a result of making
3799       parallel copy instructions explicit.
3800    
3801    ----------------------------------------------------------------------
3802    Name: Matthias Blume
3803    Date: 2002/01/22
3804    Tag: blume-20020122-x86-ccalls
3805    Description:
3806    
3807    Bug fix for c-calls on x86 (having to do with how char- and
3808    short-arguments are being handled).
3809    
3810    ----------------------------------------------------------------------
3811    Name: Matthias Blume
3812    Date: 2002/01/21
3813    Tag: blume-20020121-ff
3814    Description:
3815    
3816    Another day of fiddling with the FFI...
3817    
3818    1. Bug fix/workaround:  CKIT does not complain about negative array
3819       dimensions, so ml-nlffigen has to guard itself against this possibility.
3820       (Otherwise a negative dimension would send it into an infinite loop.)
3821    
3822    2. Some of the abstract types (light objects, light pointers, most "base"
3823       types) in structure C are now eqtypes.
3824    
3825    3. Added constructors and test functions for NULL function pointers.
3826    
3827    ----------------------------------------------------------------------
3828    Name: Matthias Blume
3829    Date: 2002/01/18
3830    Tag: blume-20020118-ready-for-new-release
3831    Description:
3832    
3833    Made config/srcarchiveurl point to a new place.  (Will provide boot
3834    files shortly.)
3835    
3836    Maybe we christen this to be 110.38?
3837    
3838    ----------------------------------------------------------------------
3839    Name: Matthias Blume
3840    Date: 2002/01/18
3841    Tag: blume-20020118-more-ffifiddle
3842    Description:
3843    
3844    Today's FFI fiddling:
3845    
3846      - Provided a structure CGetSet with "convenient" versions of C.Get.* and
3847        C.Set.* that use concrete (MLRep.*) arguments and results instead
3848        of abstract ones.
3849    
3850      - Provided word-style bit operations etc. for "int" representation
3851        types in MLRep.S<Foo>Bitops where <Foo> ranges over Char, Int, Short,
3852        and Long.
3853    
3854    ----------------------------------------------------------------------
3855 &nb