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/NOTES/HISTORY
ViewVC logotype

Diff of /sml/trunk/NOTES/HISTORY

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

revision 594, Tue Apr 4 23:41:47 2000 UTC revision 1864, Wed Nov 16 03:37:03 2005 UTC
# Line 8  Line 8 
8  The form of an entry should be:  The form of an entry should be:
9    
10  Name:  Name:
11  Date:  Date: yyyy/mm/dd
12  Tag: <post-commit CVS tag>  Tag: <post-commit CVS tag>
13  Description:  Description:
14    
15    ----------------------------------------------------------------------
16    Name: Matthias Blume (http://www.tti-c.org/blume/)
17    Date: 2005/11/15 21:35:00 CST
18    Tag: blume-22051115-versiontool
19    Description:
20    
21    Reversed change to src/system/smlnj/internals/versiontool.cm.  This
22    file gets loaded as a tool -- by the equivalent of CM.make during the
23    run of CMB.make.  Thus, CMB's path configuration is meaningless for
24    it.  Instead, it has the status of "user code", so it should use
25    $/basis.cm to refer to the Basis library.  (At least that's true for
26    the purpose of bootstrapping the previous change.  In the future it
27    might make sense to have versiontool.cm refer to
28    $smlnj/basis/basis.cm, i.e., the version of the Basis that the
29    compiler itself uses.)
30    
31    Also patched src/system/testml to have it activate those extra anchor
32    bindings in config/extrapathconfig.
33    
34    ----------------------------------------------------------------------
35    Name: Matthias Blume (http://www.tti-c.org/blume/)
36    Date: 2005/11/15 16:35:00 CST
37    Tag: blume-22051115-libraries
38    Description:
39    
40    This change affects the way the following libraries are tied into
41    the system:
42    
43       $/basis.cm
44       $/smlnj-lib.cm
45       $/pp-lib.cm
46       $/controls-lib.cm
47       $/html-lib.cm
48       $/ml-yacc-lib.cm
49    
50    These libraries are now internally (as seen from the source code
51    of the implementation itself) known by the following names:
52    
53       $smlnj/basis/basis.cm
54       $smlnj/smlnj-lib/smlnj-lib.cm
55       $smlnj/smlnj-lib/pp-lib.cm
56       $smlnj/smlnj-lib/controls-lib.cm
57       $smlnj/smlnj-lib/html-lib.cm
58       $smlnj/ml-yacc/ml-yacc-lib.cm
59    
60    This makes it possible to work with code that requires different
61    versions of these libraries, and which refers to these libraries using
62    their "default" names (i.e., the first set of names above).  In other
63    words, one can un-define or re-define those default names without
64    compromising the proper functioning of the compiler itself.
65    
66    A similar procedure had already been performed for several of the
67    MLRISC libraries that are linked into the compiler.  I did some
68    cleanup on this code.
69    
70    A new file in the config directory (named extrapathconfig) is
71    responsible for setting up path anchors that the compiler itself does
72    not need, but that are typically required by user code.
73    
74    ----------------------------------------------------------------------
75    Name: Matthias Blume
76    Date: 2005/11/07 17:05:00 CST
77    Tag: blume-20051107-slice
78    Description:
79    
80    Fixed erroneous out-of-bounds test in the "update" function of
81    various *ArraySlice modules.  (Thanks to Vesa A. Norrman for pointing
82    out the problem.)
83    
84    Pushed some Basis changes through ML-Lex, CML, and eXene.
85    
86    ----------------------------------------------------------------------
87    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
88    Date: 2005/11/07
89    Description:
90    Fixed a Basis incompatibility: The depreciated function Substring.all
91    was removed (use Substring.full instead).
92    
93    ----------------------------------------------------------------------
94    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
95    Date: 2005/11/05
96    Description:
97    Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
98    
99    ----------------------------------------------------------------------
100    Name: Matthias Blume (http://www.tti-c.org/blume/)
101    Date: 2005/11/02 16:25:00 CST
102    Tag: blume-20051102-hex
103    Description:
104    
105    * Runtime system bootstrap code now accepts hex digits in BOOTLIST
106      in either upper case or lower case format.
107    * Pushed changes to names of Pack<N>{Big,Little} structures through
108      CML and eXene.
109    
110    ----------------------------------------------------------------------
111    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
112    Date: 2005/11/02
113    Description:
114    Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
115    be called PackWord<N>{Big,Little}.
116    
117    ----------------------------------------------------------------------
118    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
119    Date: 2005/10/28
120    Description:
121    Fixed a minor Basis incompatibility: hex digits should be upper case.
122    
123    ----------------------------------------------------------------------
124    Name: Matthias Blume (blume (at) tti - c (dot) org)
125    Date: 2005/10/25 16:00:00 CDT
126    Tag: Release_110_56
127    Description:
128    
129    New working version (110.56).  NEW BOOTFILES!
130    
131    ----------------------------------------------------------------------
132    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
133    Date: 2005/10/25
134    Description:
135    Added interval sets to utility library (signatures INTERVAL_DOMAIN
136    and INTERVAL_SET, and functor IntervalSetFn).
137    
138    ----------------------------------------------------------------------
139    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
140    Date: 2005/10/14
141    Description:
142    Add Zhong Shao's fix for datatype equality functions.
143    
144    ----------------------------------------------------------------------
145    Name: Allen Leung (leunga (at) reservoir (dot) com)
146    Date: 2005/10/14 12:30:00 EST
147    Tag: leunga-20051014-x86-spill
148    Description:
149    
150    Bug fix a bug found by Carl Hauser.
151    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
152    
153    ----------------------------------------------------------------------
154    Name: Allen Leung (leunga (at) reservoir (dot) com)
155    Date: 2005/10/14 11:15:00 EST
156    Tag: leunga-20051014-x86-asm
157    Description:
158    
159    Removed some debugging code in file x86Asm.sml.
160    The function emit_operand was printing out debugging output.
161    
162    ----------------------------------------------------------------------
163    Name: John Reppy
164    Date: 2005/07/27
165    Tag:
166    Description:
167    
168    Fixed ml-lex to recognize "\r" as representing carriage return.
169    
170    ----------------------------------------------------------------------
171    Name: John Reppy
172    Date: 2005/07/27
173    Tag:
174    Description:
175    
176    Fixed ml-yacc to work on files with non-native end-of-line
177    encodings (e.g., Windows text file on a Unix system).
178    
179    ----------------------------------------------------------------------
180    Name: John Reppy
181    Date: 2005/07/20
182    Tag:
183    Description:
184    
185    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
186    to support HPUX 11.
187    
188    ----------------------------------------------------------------------
189    Name: John Reppy
190    Date: 2005/07/06
191    Tag:
192    Description:
193    
194    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
195    
196    ----------------------------------------------------------------------
197    Name: Matthias Blume (blume (at) tti - c (dot) org)
198    Date: 2005/07/06 00:45:00 CDT
199    Tag: blume-20050706-slice-copy
200    Description:
201    
202    Fixed reversed logic for deciding whether to "copy up" or "copy down"
203    in *-array-slice.sml.
204    
205    ----------------------------------------------------------------------
206    Name: Allen Leung (leunga (at) reservoir (dot) com)
207    Date: 2005/05/31 17:00:00 EST
208    Tag: leunga-20050531-cygwin-fault-2
209    Description:
210    
211    A typo in the cygwin code fixed.
212    
213    ----------------------------------------------------------------------
214    Name: Allen Leung (leunga (at) reservoir (dot) com)
215    Date: 2005/05/31 16:47:00 EST
216    Tag: leunga-20050531-cygwin-fault
217    Description:
218    
219    Updated Cygwin's fault/signal handling to match the Windows version.
220    Updated the export list.
221    
222    ----------------------------------------------------------------------
223    Name: Matthias Blume (blume (at) tti - c (dot) org)
224    Date: 2005/05/18 13:35:00 CDT
225    Tag: Release_110_54
226    Description:
227    
228    New working version (110.54).  NEW BOOTFILES!
229    
230    ----------------------------------------------------------------------
231    Name: Matthias Blume (blume (at) tti - c (dot) org)
232    Date: 2005/05/18 11:58:00 CDT
233    Tag: blume-20050518-installer
234    Description:
235    
236    Added support scripts for Mac OS X PackageMaker and modified
237    config/install.sh so that it supports re-dumping a heap image
238    after customization.
239    
240    ----------------------------------------------------------------------
241    Name: Matthias Blume (blume (at) tti - c (dot) org)
242    Date: 2005/05/18 10:55:00 CDT
243    Tag: blume-20050518-realdiv-noovld
244    Description:
245    
246    Un-overloaded / to work around bug in overloading resolution code.
247    
248    ----------------------------------------------------------------------
249    Name: Matthias Blume (blume (at) tti - c (dot) org)
250    Date: 2005/05/16 23:50:00 CDT
251    Tag: blume-20050516-redump-heap
252    Description:
253    
254    Added mechanism for re-creating a heap file for the interactive system
255    after configuration variables have been changed.
256    
257       CM.redump_heap : string -> unit
258    
259    This is much like SMLofNJ.exportML, but starting from the resulting
260    heap does not return to the caller of CM.redump_heap but
261    restarts the interactive system from scratch.  The original call of
262    CM.redump_heap does not return but ends the interactive session.
263    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
264    
265    Internally, redump_heap winds the dynamic execution context back to
266    the point where the original heap image was created and re-executes
267    the heap image generation code in the boot code.
268    
269    ----------------------------------------------------------------------
270    Name: Matthias Blume (blume (at) tti - c (dot) org)
271    Date: 2005/05/09 21:55:00 CDT
272    Tag: blume-20050509-word64
273    Description:
274    
275    Added a hack to the existing hack known as Word64 to make fromString
276    behave correctly.  I am still not sure whether Word64.scan will work
277    as specified with respect to the interaction of radix and prefix.
278    
279    ----------------------------------------------------------------------
280    Name: Allen Leung (leunga (at) reservoir (dot) com)
281    Date: 2005/05/04 11:50:00 EST
282    Tag: leunga-20050504-checkgc
283    Description:
284    
285    Added a gc protocol checking phase.  This phase is enabled with
286    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
287    
288    ----------------------------------------------------------------------
289    Name: Matthias Blume (blume (at) tti - c (dot) org)
290    Date: 2005/05/04 10:45:00 CDT
291    Tag: blume-20050504-intinf
292    Description:
293    
294    Fixed a bug in the implementation of div and mod for IntInf.
295    Thanks to Neophytos Michael for reporting the problem.
296    
297    ----------------------------------------------------------------------
298    Name: Matthias Blume (blume (at) tti - c (dot) org)
299    Date: 2005/05/04 10:35:00 CDT
300    Tag: blume-20050504-join
301    Description:
302    
303    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
304    
305    ----------------------------------------------------------------------
306    Name: Matthias Blume (blume (at) tti - c (dot) org)
307    Date: 2005/02/28 23:40:00 CST
308    Tag: blume-20050228-mVar
309    Description:
310    
311    Fixed serious bug (brown paper bag variety) in new implementation of
312    structure Atom in CML.  (I had accidentally used a mailbox instead of
313    an mvar, leaving the door open for races.)
314    
315    ----------------------------------------------------------------------
316    Name: Matthias Blume (blume (at) tti - c (dot) org)
317    Date: 2005/02/25 15:00:00 CST
318    Tag: Release_110_53
319    Description:
320    
321    New working version (110.53).  NEW BOOTFILES!
322    
323    ----------------------------------------------------------------------
324    Name: Matthias Blume (blume (at) tti - c (dot) org)
325    Date: 2005/02/25 14:50:00 CST
326    Tag: blume-20050225-susp
327    Description:
328    
329    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
330    implemented in Core -- which means that it is the same as the one used
331    by the lazy extension.
332    
333    ----------------------------------------------------------------------
334    Name: Matthias Blume (blume (at) tti - c (dot) org)
335    Date: 2005/02/24 16:50:00 CST
336    Tag: blume-20050224-cml-atom
337    Description:
338    
339    Simpler and at the same time more general implementation of structure
340    Atom in CML.
341    
342    ----------------------------------------------------------------------
343    Name: Matthias Blume (blume (at) tti - c (dot) org)
344    Date: 2005/02/15 17:35:00 CST
345    Tag: blume-20050215-tools
346    Description:
347    
348    Created new "tools" directory under "src" and moved "TraceDebugProf"
349    there.
350    
351    ----------------------------------------------------------------------
352    Name: Matthias Blume (blume (at) tti - c (dot) org)
353    Date: 2005/02/10 17:55:00 CST
354    Tag: blume-20050210-longlong
355    Description:
356    
357    Implemented "long long" arguments and results for NLFFI.  (Only the
358    PPC/MacOS implementation is complete, the other backends still need to
359    be updated.)
360    
361    ----------------------------------------------------------------------
362    Name: Matthias Blume (blume (at) tti - c (dot) org)
363    Date: 2005/01/24 17:40:00 CST
364    Tag: blume-20050124-mlyacc
365    Description:
366    
367    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
368    problem with certain "as" patterns which previously got rendered
369    using incorrect syntax.
370    
371    ----------------------------------------------------------------------
372    Name: Matthias Blume (blume (at) tti - c (dot) org)
373    Date: 2005/01/18 12:00:00 CST
374    Tag: blume-20050118-profile
375    Description:
376    
377    Made time profiling code (interrupt handler) in runtime system aware
378    of new array representation.
379    
380    ----------------------------------------------------------------------
381    Name: Matthias Blume (blume (at) tti - c (dot) org)
382    Date: 2005/01/14 18:00:00 CST
383    Tag: blume-20050114-heap2exec
384    Description:
385    
386    Implemented new (but still experimental) heap2exec facility.  This is
387    tested under Mac OS X and should work under Linux (will test shortly).
388    It will probably also work on the Sparc (will test some time later).
389    
390      - removed old "HACKED_STANDALONE" hack from runtime
391    
392    To be able to test this, uncomment the request for "heap2asm" in
393    config/targets prior to installation.  (Notice that this is different
394    from "heap2exec" mentioned below.  Not a typo.)
395    
396    To perform an actual test, run the command
397    
398       $ bin/heap2exec heapfile execfile
399    
400    (You can put heap2exec on your shell's path.)
401    
402    For example, run
403    
404      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
405    
406    This will create a standalone executable called "mly" which you can
407    then invoke directly as a command.
408    
409    ----------------------------------------------------------------------
410    Name: Matthias Blume (blume (at) tti - c (dot) org)
411    Date: 2005/01/07 11:44:00 CST
412    Tag: blume-20050107-mlstring
413    Description:
414    
415    fixed off-by-one error in ML_STRING macro (globals.c)
416    
417    ----------------------------------------------------------------------
418    Name: Matthias Blume (blume (at) tti - c (dot) org)
419    Date: 2004/12/23 18:00:00 CST
420    Tag: blume-20041223-santa
421    Description:
422    
423    Made ml-build script "smarter" (but only very little).
424    
425    ----------------------------------------------------------------------
426    Name: Matthias Blume (blume (at) tti - c (dot) org)
427    Date: 2004/12/21 15:05:00 CST
428    Tag: blume-20041221-longlong
429    Description:
430    
431    * Implemented access to signed and unsigned long long data in NLFFI.
432    (The parameter-passing part of the picture has not complete. But
433    data structure access seems to work.)
434    
435    * Fixed CM's incorrect assumption that the PPC is little-endian.
436      (On the Mac, it is big-endian.  And that's currently our only
437      PPC platform.)
438    
439    ----------------------------------------------------------------------
440    Name: Matthias Blume (blume (at) tti - c (dot) org)
441    Date: 2004/12/21 12:50:00 CST
442    Tag: blume-20041221-memory
443    Description:
444    
445    Some cleanup in the $c/memory.cm library: separated some concerns by
446    moving allocation code and memory access code each into their own
447    files.
448    
449    ----------------------------------------------------------------------
450    Name: Allen Leung (leunga (at) reservoir (dot) com)
451    Date: 2004/12/17 16:12:00 EST
452    Tag: leunga-20041217-cygwin-smlnj-home
453    Description:
454    
455        The Unix I/O library of SML/NJ on cygwin does not understand
456    Windows style pathname, so problems arise when SMLNJ_HOME is set
457    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
458    to a POSIX pathname on cygwin.
459    
460    ----------------------------------------------------------------------
461    Name: Matthias Blume (blume (at) tti - c (dot) org)
462    Date: 2004/12/16 13:35:00 CST
463    Tag: Release_110_52
464    Description:
465    
466    Last-minute changes incorporated into 110.52.  Release tag moved.
467    
468    The changes:
469       - HashString.hashString' -> HashString.hashSubstring
470       - bug fix in UnivariateStats
471    
472    ----------------------------------------------------------------------
473    Name: Matthias Blume (blume (at) tti - c (dot) org)
474    Date: 2004/12/15 23:40:00 CST
475    Tag: blume-20041215-hashSubstring
476    Description:
477    
478    - HashString.hashString' -> HashString.hashSubstring
479    - corresponding changes in atom.sml
480    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
481      added some comments
482    
483    ----------------------------------------------------------------------
484    Name: Matthias Blume (blume (at) tti - c (dot) org)
485    Date: 2004/12/15 15:30:00 CST
486    Tag: (Release_110_52)
487    Description:
488    
489    New working version (110.52).  NEW BOOTFILES!
490    
491    ----------------------------------------------------------------------
492    Name: Matthias Blume (blume (at) tti - c (dot) org)
493    Date: 2004/12/15 12:45:00 CST
494    Tag: blume-20041215-spaces
495    Description:
496    
497    More on the space problem (this time for Win32).
498    
499    ----------------------------------------------------------------------
500    Name: Matthias Blume (blume (at) tti - c (dot) org)
501    Date: 2004/12/14 17:30:00 CST
502    Tag: blume-20041214-spaces
503    Description:
504    
505    Hacked some of the scripts (in particular: the installer) to cope with
506    spaces in filenames a bit better.  But beware: the current "solution"
507    is likely still full of bugs and inherently incomplete.  (We need to
508    do away with those shell scripts for a comprehensive solution.)
509    
510    ----------------------------------------------------------------------
511    Name: Matthias Blume (blume (at) tti - c (dot) org)
512    Date: 2004/12/13 14:45:00 CST
513    Tag: blume-20041213-ml-makedepend
514    Description:
515    
516    Fixed bug in code for ml-makedepend.
517    
518    ----------------------------------------------------------------------
519    Name: Matthias Blume (blume (at) tti - c (dot) org)
520    Date: 2004/12/09 16:30:00 CST
521    Tag: blume-20041209-statistics
522    Description:
523    
524    Added two simple but potentially useful statistics modules to SML/NJ Library.
525    (See CHANGES file there.)
526    
527    ----------------------------------------------------------------------
528    Name: Matthias Blume (blume (at) tti - c (dot) org)
529    Date: 2004/12/01 16:50:00 CST
530    Tag: blume-20041201-atom
531    Description:
532    
533    smlnj-lib:
534    
535      Added function HashString.hashString' for substrings.
536      Hand-inlined CharVector.foldl into HashString (for speed).
537      Modified implementation of structure Atom to avoid extracting
538      strings from substrings unless necessary.
539      (Also see CHANGES file for smlnj-lib.)
540    
541    ----------------------------------------------------------------------
542    Name: Matthias Blume (blume (at) tti - c (dot) org)
543    Date: 2004/11/24 22:15:00 CST
544    Tag: blume-20041124-cml
545    Description:
546    
547    Made sure CML compiles when Position = Int64.
548    
549    ----------------------------------------------------------------------
550    Name: Matthias Blume (blume (at) tti - c (dot) org)
551    Date: 2004/11/24 14:30:00 CST
552    Tag: blume-20041124-position
553    Description:
554    
555    The compiler can now be compiled in a mode that makes structure Position
556    equal to Int64.  The default, however, is unchanged (Position = Int31)
557    for the time being.
558    
559    To enable 64-bit positions, use the following procedure:
560    
561    1. Start sml
562    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
563    3. Type
564          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
565    4. Run CMB.make() as usual.
566    
567    This is barely tested.  The only test so far was a little SML program
568    counting the number of characters in an 8-gigabyte file by
569    reading it character-by-character.  That test was successful.
570    
571    In support of 64-bit positions, a number of new functions have been
572    added to the runtime system.
573    
574    ----------------------------------------------------------------------
575    Name: Matthias Blume (blume (at) tti - c (dot) org)
576    Date: 2004/11/23 14:45:00 CST
577    Tag: blume-20041123-useFile
578    Description:
579    
580    Fixed a problem with unhelpful error messages related to problems with
581    .cm- or .sml files that appear as part of the sml command line.
582    
583    ----------------------------------------------------------------------
584    Name: Matthias Blume (blume (at) tti - c (dot) org)
585    Date: 2004/11/18 15:40:00 CST
586    Tag: Release_110_51
587    Description:
588    
589    New working version (110.51).  NEW BOOTFILES!
590    
591    ----------------------------------------------------------------------
592    Name: Matthias Blume (blume (at) tti - c (dot) org)
593    Date: 2004/11/18 15:35:00 CST
594    Tag: <none>
595    Description:
596    
597    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
598    Joemann).
599    
600    ----------------------------------------------------------------------
601    Name: Allen Leung (leunga (at) reservoir (dot) com)
602    Date: 2004/11/17 16:05:21 EST 2004
603    Tag: leunga-20041117-mlrisc-live-kill
604    Description:
605    
606       Added support for MLTree constructs LIVE and KILL
607    to all the architectures.
608    
609    ----------------------------------------------------------------------
610    Name: Matthias Blume (blume (at) tti - c (dot) org)
611    Date: 2004/11/13 00:20:00 CST
612    Tag: blume-20041113-versiontool
613    Description:
614    
615    - Stripped down the versiontool: It now only handles the version number.
616      The date string is generated at bootstrap time (during makeml).
617    
618    - In a previous commit, fixed a minor issue with how polyequal is being
619      translated.  In particular, the code now "looks through" abstractions.
620      This results in slightly fewer polyEqual warnings and hopefully slightly
621      more efficient code.  Important examples for where this matters are
622      the new int64 and word64 types.
623    
624    ----------------------------------------------------------------------
625    Name: Matthias Blume (blume (at) tti - c (dot) org)
626    Date: 2004/11/12 00:30:00 CST
627    Tag: blume-20041112-int64
628    Description:
629    
630    Structure Int64 fully hooked in.  (The implementation is not very
631    efficient, though.)
632    
633    ----------------------------------------------------------------------
634    Name: Matthias Blume (blume (at) tti - c (dot) org)
635    Date: 2004/11/11 17:30:00 CST
636    Tag: blume-20041111-more64
637    Description:
638    
639    All the pieces of Word64 are now there, with the exception of the
640    conversions from and to LargeWord. (Eventually these need to be identities,
641    but for the time being they don't even make sense because LargeWord is
642    32-bit wide.)
643    
644    Also started to add similar support for Int64, but major pieces of that
645    are still missing.
646    
647    ----------------------------------------------------------------------
648    Name: Matthias Blume (blume (at) tti - c (dot) org)
649    Date: 2004/11/11 00:15:00 CST
650    Tag: blume-20041111-word64
651    Description:
652    
653    Structure Word64 is now (almost) complete, word literals and patterns
654    seem to work.  There are a few odd pieces missing.  In particular,
655    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
656    at the moment.
657    
658    Making Word64 official would mean that LargeWord becomes Word64.  But
659    this requires extreme care because most word-word conversions have to
660    go through LargeWord, so making a mistake means loss of efficiency or
661    worse.  Eventually there will be a solution similar to (but actually
662    simpler than) what I did with IntInf.
663    
664    ----------------------------------------------------------------------
665    Name: Matthias Blume (blume (at) tti - c (dot) org)
666    Date: 2004/11/10 18:12:00 CST
667    Tag: blume-20041110-64bit
668    Description:
669    
670    More 64-bit hacking (but still not even half-way there yet).
671    Also, some assorted improvements to the handling of 8-bit words.
672    
673    ----------------------------------------------------------------------
674    Name: Matthias Blume (blume (at) tti - c (dot) org)
675    Date: 2004/11/09 17:50:00 CST
676    Tag: <none>
677    Description:
678    
679    Started adding some infrastructure for supporting 64-bit int- and
680    word-types.  (Still in its very early stages.)
681    
682    ----------------------------------------------------------------------
683    Name: Matthias Blume (blume (at) tti - c (dot) org)
684    Date: 2004/10/28 10:45:00 CDT
685    Tag: Release_110_50
686    Description:
687    
688    New working version (110.50).  NEW BOOTFILES!
689    
690                 =====================
691    
692    Also:
693    
694      - Changend config/srcarchiveurl from a file just
695        containing the URL string into a file containing
696        shell script code.  The code has access to the $VERSION variable.
697      - Made corresponding changes to config/install.sh and config/unpack.
698      - Default contents of config/srcarchiveurl uses $VERSION and
699        normally does not have to be edited to reflect a version change.
700    
701      (As a result, a version change can be done by just editing
702       config/version, the rest is now automatic.)
703    
704    ----------------------------------------------------------------------
705    Name: Matthias Blume (blume (at) tti - c (dot) org)
706    Date: 2004/10/27 17:50:00 CDT
707    Tag: blume-20041027-btrace-msg
708    Description:
709    
710    BackTrace.monitor now also reports the source of the exception that
711    triggered the trace.
712    
713    ----------------------------------------------------------------------
714    Name: Matthias Blume (blume (at) tti - c (dot) org)
715    Date: 2004/10/27 17:20:00 CDT
716    Tag: blume-20041027-x86-c-calls
717    Description:
718    
719    This is the HISTORY entry for two earlier commits, both concerning
720    the x86 c-calls code in MLRISC:
721    
722      - added a missing LOAD in the code that deals with struct arguments
723      - made sure the caller does not add the wrong number of bytes to the
724        stack pointer after a call of a function returning a struct
725        (the callee already pops the implicit argument which points to
726         the space reserved for the result)
727    
728    ----------------------------------------------------------------------
729    Name: Allen Leung (leunga (at) reservoir (dot) com)
730    Date: 2004/10/24 14:00:00 EST
731    Tag: leunga-20041024-x86-gas-fucomip
732    Description:
733    
734       John discovered a bug in the syntax of fucomip.
735       The opcodes FU?COMIP? have been changed to
736    
737           fu?comip? %st(i), %st
738    
739    ----------------------------------------------------------------------
740    Name: Matthias Blume (blume (at) tti - c (dot) org)
741    Date: 2004/10/20 15:06:00 CDT
742    Tag: blume-20041020-standalone-backtrace
743    Description:
744    
745    Added a mechanism for getting back-trace information from standalone
746    programs.  Here is how it works:
747    
748       1. The part of the program from which you want to get backtrace
749          information (usually the whole program) should be wrapped with
750          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
751          main program could be modified from something like
752    
753             fun main (pgm, args) = ...
754    
755         to
756    
757            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
758    
759       2. To be able to access BackTrace.monitor, you have to add
760          library $smlnj-tdp/plugins.cm to the .cm file that contains your
761          main function.
762    
763       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
764          CM might have created in the past for your project).
765    
766       4. Build the system using this command line:
767    
768             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
769                      myprog.cm MyProg.main myprog
770    
771          instead of the usual
772    
773             ml-build myprog.cm MyProg.main myprog
774    
775    I changed a library name:
776    
777       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
778    
779    New libraries:
780    
781       $smlnj-tdp/back-trace.cm
782          -- when loaded causes the back-trace plugin to be installed
783       $smlnj-tdp/coverage.cm
784          -- when loaded causes the coverage plugin to be installed
785    
786    ----------------------------------------------------------------------
787    Name: Matthias Blume (blume (at) tti - c (dot) org)
788    Date: 2004/10/18 16:45:00 CDT
789    Tag: blume-20041018-groupowner
790    Description:
791    
792    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
793    
794    Eliminated group owner specs from .cm files throughout the source tree.
795    
796    ----------------------------------------------------------------------
797    Name: Matthias Blume (blume (at) tti - c (dot) org)
798    Date: 2004/10/15 15:45:00 CDT
799    Tag: blume-20041015-coverage
800    Description:
801    
802    * Test coverage tool added!
803    * Further reorganization of tracing-, debugging-, and profiling support:
804    
805        - moved original BTImp -- now called BackTrace -- into a separate
806          library called $/trace-debug-profile.cm
807        - eliminated all mentions of BTrace from SMLofNJ.Internals
808        - only the instrumentation mechanism is now left in the compiler proper
809        - BackTrace module is a plugin which is NOT plugged in by default
810        - Coverage module is another such plugin
811    
812    To get the benefits of any of these plugin modules, the code in
813    question must be compiled with tdp instrumentation turned on.  This
814    can be done by setting SMLofNJ.Internals.TDP.mode to true.
815    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
816    
817    Plugins are selected at link time.  (Pre-compiled instrumented code
818    can be re-loaded with different plugins in effect.)  When an
819    instrumented module is linked, whatever plugins are at that time
820    enabled will come into effect for that module.
821    
822    To enable the back-trace plugin, load library $/trace-debug-profile.cm
823    and invoke BackTrace.install() (e.g., from the interactive prompt).
824    To enable the coverage plugin, load the same library and invoke
825    Coverage.install().
826    
827    Back-traces are generated automatically on uncaught exceptions and
828    when the code in question explicitly invokes BackTrace.trigger().
829    
830    Coverage (and execution frequency-) information must be queried
831    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
832    
833    ----------------------------------------------------------------------
834    Name: Matthias Blume (blume (at) tti - c (dot) org)
835    Date: 2004/10/14 17:40:00 CDT
836    Tag: blume-20041014-tdp-core
837    Description:
838    
839    Snapshot of a significant overhaul of how the trace/debug/profile support
840    is hooked into the system (specifically: Core and SMLofNJ.Internals).
841    
842    ----------------------------------------------------------------------
843    Name: Matthias Blume (blume (at) tti - c (dot) org)
844    Date: 2004/10/13 16:34:00 CDT
845    Tag: blume-20041013-tdp
846    Description:
847    
848    Some rationalization of names:
849    
850       structure BTrace -> structure TDPInstrument
851       etc.
852    
853    This is is preparation of using the original back-trace
854    instrumentation for other purposes.  "TDP" stands for
855    Trace/Debug/Profile.
856    
857    The control flag controlling whether instrumentation is on or off is now
858    registered under a different name, so instead of running sml as
859    
860     sml -Cinstrument.btrace-mode=true
861    
862    one has to say
863    
864     sml -Ctdp.instrument=true
865    
866    ----------------------------------------------------------------------
867    Name: Matthias Blume (blume (at) tti - c (dot) org)
868    Date: 2004/10/11 16:37:00 CDT
869    Tag: blume-20041011-regions
870    Description:
871    
872    Made some minor modifications to elabcore.sml to have source regions
873    be propagated more tightly -- resulting in better (i.e., smaller)
874    regions being reported in error- and debug messages.
875    
876    ----------------------------------------------------------------------
877    Name: Matthias Blume (blume (at) tti - c (dot) org)
878    Date: 2004/10/08 22:50:00 CDT
879    Tag: blume-20041008-cmkw
880    Description:
881    
882    Fixed handling of keywords in .cm files: After seeing "is" the lexer
883    treats subsequent occurrences of "group", "library", "source", "is",
884    "*", and "-" as ordinary identifiers rather than keywords.
885    
886    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
887    is supposed to accept a tool argument called "source", but this did
888    not work because of the clash with the keyword.
889    
890    ----------------------------------------------------------------------
891    Name: Matthias Blume (blume (at) tti - c (dot) org)
892    Date: 2004/10/07 16:00:00 CDT
893    Tag: blume-20041007-cleanup
894    Description:
895    
896    Assorted cleanup work:
897    
898      - got rid of intstrmap in favor of using the library's
899        hash table implementation
900      - threw out most of the pathnames stuff, as it was not used anyway
901      - simplified tokentable implementation
902      - fixed some minor spelling errors
903    
904    ----------------------------------------------------------------------
905    Name: Matthias Blume (blume (at) tti - c (dot) org)
906    Date: 2004/10/06 15:15:15 CDT
907    Tag: blume-20041006-handler
908    Description:
909    
910    Cleaned up the absyn to reflect the invariant that HANDLE always
911    carries a FNexp as part of the type definition.  This eliminates some
912    superfluous sanity checks at runtime down the road.
913    
914    Some minor cleanup of the btrace code.
915    
916    ----------------------------------------------------------------------
917    Name: Matthias Blume (blume (at) tti - c (dot) org)
918    Date: 2004/10/01 10:20:30 CDT
919    Tag: blume-20041001-slave
920    Description:
921    
922    Added hack to make slave mode work in the presence of the version
923    tool.  (Still, since the master does two passes over the code for
924    CMB.make, the release number gets bumped twice when slaves are
925    attached.  I don't know if this is worth fixing...)
926    
927    ----------------------------------------------------------------------
928    Name: Matthias Blume (blume (at) tti - c (dot) org)
929    Date: 2004/09/30 10:55:00 CDT
930    Tag: blume-20040930-version
931    Description:
932    
933    * Moved the "version" magic into its own little library under
934      src/system/smlnj/internal.  This avoids expensive reconstruction of
935      a stable src/compiler/core.cm.
936    
937    * At the same time, structure CompilerVersion is now known as
938      structure SMLNJVersion.
939    
940    * Arranged for the version tool to NOT kick in when rebuilding the system
941      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
942      Also, loading the versiontool does not work when rebuilding the system
943      because CM is not properly initialized at that time.
944    
945    ----------------------------------------------------------------------
946    Name: Matthias Blume (blume (at) tti - c (dot) org)
947    Date: 2004/09/29 14:00:00 CDT
948    Tag: blume-20040929-autoversion
949    Description:
950    
951    Implemented some CM magic to have
952        file src/compiler/TopLevel/main/version.sml
953    generated automagically.
954    The version is taken from two files: config/version and config/release.
955    The first is expected to contain a two-part version number such as 110.49.
956    The second should contain a single number, but it may be missing.
957    
958    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
959    time the version tool is loaded (which is the first time you say CMB.make),
960    then the tool will increment the value stored in config/release every
961    time CMB.make is invoked.
962    
963    The binfile format is now insensitive to anything beyond the first
964    two components of a version number, so bumping the release does not render
965    binfiles incompatible.  Auto-bumping can be used to keep track of versions
966    during development without invalidating existing binfiles.
967    
968    In any case, every CMB.make updates the date information in version.sml.
969    (This is the date that is printed in the banner.)
970    
971    ----------------------------------------------------------------------
972    Name: Matthias Blume (blume (at) tti - c (dot) org)
973    Date: 2004/09/28 10:53:00 CDT
974    Tag: blume-20040928-controls
975    Description:
976    
977    Some cleanup of the controls code.
978    
979    ----------------------------------------------------------------------
980    Name: Matthias Blume (blume (at) tti - c (dot) org)
981    Date: 2004/09/27 22:08:00 CDT
982    Tag: blume-20040927-controls
983    Description:
984    
985    Added two pieces of functionality to the Controls interface:
986    
987    1.   val save'restore: 'a control -> unit -> unit
988    
989       grabs the current value of the control in stage 1 and restores it
990       in stage 2.
991    
992    2.   val set' : 'a control * 'a -> unit -> unit
993    
994       stores the given value into the control in stage 2 (i.e., delayed)
995       but does all error checking in stage 1.
996       (This is for string controls that need to do parse their argument --
997        something that might fail.  In some cases, notably in CM, one
998        already knows the intended argument but wants to delay the actual
999        assignment until a time when error recovery would be more difficult.)
1000    
1001    Changed the handling of controls in tool arguments to classes "sml" and
1002    "lazysml":
1003       - use Controls.save'restore as a more robust way of restoring the
1004         old value (in particular: without having to re-parse the string)
1005       - use controls to handle the "overload" keyword in the init group
1006         (I believe this change actually fixes a long-standing obscure bug.)
1007    
1008    ----------------------------------------------------------------------
1009    Name: Matthias Blume (blume (at) tti - c (dot) org)
1010    Date: 2004/09/27 17:00:00 CDT
1011    Tag: blume-20040927-lazysml
1012    Description:
1013    
1014    Added a new tool class called "lazysml" to CM's tool chest.  The only
1015    difference to "sml" is that compilation is done with Control.lazysml
1016    set to true.  A source of class "lazysml" is automatically recognized
1017    by a file name suffix of ".lml".
1018    
1019    In addition to the above feature, the original class "sml" now also
1020    supports a tool argument "lazy" which has the same effect.  As a
1021    result, the following three lines are equivalent:
1022    
1023        foo.sml : lazysml
1024        foo.sml : sml (lazy)
1025        foo.sml (lazy)
1026    
1027    The setting goes into effect both during parsing and during
1028    compilation.  The original setting is restored right after parsing and
1029    after compilation, respectively.
1030    
1031    In addition to all the above, there is also a general mechanism to set
1032    ANY of the "controls" that are available at the command line via
1033    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
1034    apply as well. (In fact, "lazy" is implemented as a special case of
1035    the general mechanism.)
1036    
1037    The .cm file syntax uses a new keyword tool argument called "with".
1038    There are several ways of indicating the desired settings:
1039    
1040        foo.sml (with:parser.quotations=true)
1041        foo.sml (with:(name:parser.quotations value:true))
1042        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
1043        foo.sml (with:(name1=value1 name2=value2 ...))
1044        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
1045    
1046        etc.
1047    
1048    Another possible abbreviation is to leave out the =v or value:v part
1049    if the name refers to a boolean control (in which case the value is
1050    taken to be true).  Thus, one could get lazy sml also by saying:
1051    
1052        foo.sml (with:parser.lazy-keyword=true)
1053        foo.sml (with:parser.lazy-keyword)
1054        foo.sml (with:(name:parser.lazy-keyword value:true))
1055        foo.sml (with:(name:parser.lazy-keyword))
1056    
1057    ----------------------------------------------------------------------
1058    Name: Matthias Blume (blume (at) tti - c (dot) org)
1059    Date: 2004/09/24 16:22:00 CDT
1060    Tag: blume-20040924-ppc-long-branch
1061    Description:
1062    
1063    Turned message about "emiting long form of branch" off by default.
1064    Added a control flag to turn it back on when desired.
1065    
1066    ----------------------------------------------------------------------
1067    Name: Matthias Blume (blume (at) tti - c (dot) org)
1068    Date: 2004/09/24 16:05:00 CDT
1069    Tag: blume-20040924-rounding
1070    Description:
1071    
1072    Applied patch for setting rounding modes under Mac OS X.  Thanks to
1073    Melissa O'Neill for providing the code!
1074    
1075    ----------------------------------------------------------------------
1076    Name: Matthias Blume (blume (at) tti - c (dot) org)
1077    Date: 2004/09/23 17:30:00 CDT
1078    Tag: blume-20040923-envvars
1079    Description:
1080    
1081    1. Changed definition of type ControlRegistry.registry_tree to
1082       include control_info (i.e., the name of the controlling environment
1083       variable).
1084    
1085    2. Added command-line flags -e and -E to print the names of environment
1086       variables that can be used to control internal settings.
1087       (This uses the new API mentioned in 1.)
1088    
1089    ----------------------------------------------------------------------
1090    Name: Matthias Blume (blume (at) tti - c (dot) org)
1091    Date: 2004/09/13 16:50:00 CDT
1092    Tag: Release_110_49
1093    Description:
1094    
1095    New working version (110.49).  NEW BOOTFILES!
1096    
1097    ----------------------------------------------------------------------
1098    Name: Matthias Blume
1099    Date: 2004/09/13 16:20:00 CDT
1100    Tag: blume-20040913-config-mlrisc
1101    Description:
1102    
1103    Put target "mlrisc" back into the default list.
1104    (There is no harm in having it, and some users have expressed their
1105    wish to have "mlrisc" included by default.)
1106    
1107    ----------------------------------------------------------------------
1108    Name: John Reppy
1109    Date: 2004/09/13
1110    Tag: jhr-20040913-signals
1111    Description:
1112    
1113    Fixed the signal masking code to properly nest mask/unmask operations
1114    on a per-signal basis.
1115    
1116    ----------------------------------------------------------------------
1117    Name: Matthias Blume (blume (at) tti - c (dot) org)
1118    Date: 2004/09/08 13:20:00 CDT
1119    Tag: blume-20040908-heap-magic
1120    Description:
1121    
1122    Bumped the heap macig number to 0x09082004 to account for the changed
1123    layout of the ML frame under MacOS X.
1124    
1125    ----------------------------------------------------------------------
1126    Name: Allen Leung (leunga (at) reservoir (dot) com)
1127    Date: 2004/09/03 11:26:00 EST
1128    Tag: leunga-20040903-cygwin-install
1129    Description:
1130    
1131       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1132    The Cygwin runtime is turned on by setting the environment
1133    variable SMLNJ_CYGWIN_RUNTIME to 1.
1134    
1135    ----------------------------------------------------------------------
1136    Name: Matthias Blume (blume (at) tti - c (dot) org)
1137    Date: 2004/08/31 17:15:00 CDT
1138    Tag: blume-20040831-core
1139    Description:
1140    
1141    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1142    
1143    ----------------------------------------------------------------------
1144    Name: Matthias Blume (blume (at) tti - c (dot) org)
1145    Date: 2004/08/30 17:55:00 CDT
1146    Tag: blume-20040830-installer
1147    Description:
1148    
1149    Upon request by Johannes Joemann:
1150    
1151    - improved ML code of installer to fall back to coping when renaming
1152      fails (i.e., when source and target are on different file systems);
1153      the code compiles but has yet to be tested in anger
1154    - removed mlrisc from list of default targets (config/targets)
1155    
1156    ----------------------------------------------------------------------
1157    Name: Matthias Blume (blume (at) tti - c (dot) org)
1158    Date: 2004/08/27 17:20:00 CDT
1159    Tag: blume-20040827-ptreql
1160    Description:
1161    
1162    Added ptreql primop to structure InlineT (upon request from Larry
1163    Paulson).
1164    
1165    ----------------------------------------------------------------------
1166    Name: Allen Leung (leunga (at) reservoir (dot) com)
1167    Date: 2004/08/15 21:21:00 EST
1168    Tag: leunga-110_48-udgraph
1169    Description:
1170    
1171    Another bug fix from Carl Hauser:
1172    
1173    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1174    > 48c48
1175    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1176    > e::L)
1177    > ---
1178    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1179    > Without this, any deletion of an edge in an undirected graph does severe
1180    > violence to the graph.
1181    
1182    ----------------------------------------------------------------------
1183    Name: Allen Leung (leunga (at) reservoir (dot) com)
1184    Date: 2004/08/10 23:35:00 EST
1185    Tag: leunga-110_48-ppc
1186    Description:
1187    
1188       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1189    
1190    ----------------------------------------------------------------------
1191    Name: Matthias Blume (blume (at) tti - c (dot) org)
1192    Date: 2004/08/10 12:00:00 CDT
1193    Tag: Release_110_48
1194    Description:
1195    
1196    New working version (110.48).  NEW BOOTFILES!
1197    
1198    ----------------------------------------------------------------------
1199    Name: Allen Leung (leunga (at) reservoir (dot) com)
1200    Date: 2004/08/09 12:21:00 EST
1201    Tag: leunga-110_47-dijsktra
1202    Description:
1203    
1204    Bug fix from Carl Hauser:
1205    
1206    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1207    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1208    
1209    The problem is that following the expression A.update(dist,s,Num.zero)
1210    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1211    
1212    ----------------------------------------------------------------------
1213    Name: Matthias Blume (blume (at) tti - c (dot) org)
1214    Date: 2004/08/06 18:10:00 CDT
1215    Tag: blume-20040806-cmdline
1216    Description:
1217    
1218    Fiddled with handling of command-line options:
1219    
1220      * sml now quits after processing the command line
1221        if -H, -S, -h<n>, or -s<n> appears as the last
1222        command-line argument
1223      * a new option -q terminates the session when encountered on
1224        the command line; subsequent arguments will be ignored
1225      * bug fixes: short (erroneous) arguments are no longer ignored
1226        completely
1227    
1228    ----------------------------------------------------------------------
1229    Name: Allen Leung (leunga (at) reservoir (dot) com)
1230    Date: 2004/08/04 18:17:00 EST
1231    Tag: leunga-110_47-ppc-ibm-asm
1232    Description:
1233    
1234      - Added minimal IBM assembly syntax support for PowerPC.
1235    
1236      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1237        been altered in the runtime.  We need an automatic way to keep the file
1238        in sync.
1239    
1240    ----------------------------------------------------------------------
1241    Name: Matthias Blume (blume (at) tti - c (dot) org)
1242    Date: 2004/08/04 14:00:00 CDT
1243    Tag: Release_110_47
1244    Description:
1245    
1246    New working version (110.47).  NEW BOOTFILES!
1247    
1248    ----------------------------------------------------------------------
1249    Name: Matthias Blume (blume (at) tti - c (dot) org)
1250    Date: 2004/08/03 14:25:00 CDT
1251    Tag: blume-20040803-callingconv
1252    Description:
1253    
1254    Added low-level support for choosing C calling conventions by
1255    twiddling the type of rawccall.  (See
1256    src/compiler/Semant/types/cproto.sml for details.)
1257    
1258    ----------------------------------------------------------------------
1259    Name: Matthias Blume (blume (at) tti - c (dot) org)
1260    Date: 2004/08/02 15:55:00 CDT
1261    Tag: blume-20040802-backout
1262    Description:
1263    
1264    Backed out of change to win32-filesys.c.  The earlier patch to
1265    get_file_time caused CM to produce files with the wrong time stamp.
1266    
1267    ----------------------------------------------------------------------
1268    Name: Matthias Blume (blume (at) tti - c (dot) org)
1269    Date: 2004/08/02 14:45:00 CDT
1270    Tag: blume-20040802-nlffi-win32
1271    Description:
1272    
1273    Added NLFFI support for Win32, adapted from a patch provided by David
1274    Hansel.  This is currently completely untested.  Also, the issue
1275    concerning stdcall vs. ccall is still unresolved.
1276    
1277    ----------------------------------------------------------------------
1278    Name: Matthias Blume (blume (at) tti - c (dot) org)
1279    Date: 2004/07/30 17:55:00 CDT
1280    Tag: blume-20040730-various
1281    Description:
1282    
1283    Gearing up towards 110.47...
1284    
1285    - various minor bugfixes to ml-nlffigen
1286    - a beginning of a manual for nlffi
1287    
1288    - eliminated 'export name=value' in config/install.sh as this does
1289      not work with certain versions of /bin/sh
1290      (Thanks to David King at Motorola for catching this.)
1291    
1292    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1293      - added a test for tm==NULL to gmtime.c and localtime.c
1294      - applied patch for incorrect GetFileTime under win32
1295      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1296    
1297    ----------------------------------------------------------------------
1298    Name: Matthias Blume (blume (at) tti - c (dot) org)
1299    Date: 2004/07/21 18:20:00 CDT
1300    Tag: blume-20040721-nlffigen
1301    Description:
1302    
1303    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1304      for a typedef to an incomplete type, but leave out the "typ" member.
1305      (This is just for consistency.)
1306    - Started to produce what is supposed to become better (i.e., comprehensive)
1307      documentation of what ml-nlffigen does and produces.
1308    
1309    ----------------------------------------------------------------------
1310    Name: Matthias Blume (blume (at) tti - c (dot) org)
1311    Date: 2004/07/14 16:25:00 CDT
1312    Tag: blume-20040714-union
1313    Description:
1314    
1315    Added C_UNION to c-calls/c-types.sml and updated the machinery
1316    (ml-nlffigen, cproto.sml) that conveys C function interface
1317    information to the code generator.
1318    
1319    However, the actual architecture-specific implementation of function
1320    arguments and results that are C unions is still not implemented.
1321    
1322    ----------------------------------------------------------------------
1323    Name: Allen Leung (leunga (at) reservoir (dot) com)
1324    Date: 2004/07/14 14:38:00 EST
1325    Tag: leunga-110_46_1-ppc-lwzu
1326    Description:
1327    
1328        Added these instructions to the PowerPC architecture:
1329        LBZU(X), LHZU(X), LWZU(X),
1330        STWU(X), STFDU, STFSU
1331    
1332        etc...
1333    
1334        Note: I haven't added their instruction encoding into the description.
1335    
1336    ----------------------------------------------------------------------
1337    Name: Allen Leung (leunga (at) reservoir (dot) com)
1338    Date: 2004/07/13 15:04:00 EST
1339    Tag: leunga-110_46_1-ppc-lwarx
1340    Description:
1341    
1342       Added the two instructions LWARX and STWCX to the PowerPC
1343    instruction set.
1344    
1345       A (untested) rewrite of loop-structure.sml.  The old version
1346    is completely broken.
1347    
1348    ----------------------------------------------------------------------
1349    Name: Matthias Blume (blume (at) tti - c (dot) org)
1350    Date: 2004/07/13 13:50:00 CDT
1351    Tag: blume-20040713-nlffi
1352    Description:
1353    
1354    - use paramAlloc to report c-calls with too many arguments
1355      (for PPC version where parameter area is pre-allocated)
1356    - added ccall_maxargspace to machspec (to implement the above)
1357    - made "make" commend in CM's "make" tool configurable
1358    - added option (default: on) for passing the name of the SML/NJ's "bin"
1359      directory to "make"; the call looks like this:
1360    
1361         make <options> SMLNJ_BINDIR=<dir> <target>
1362    
1363      This can be used by the Makefile to, e.g., pick the "right" version
1364      of ml-nlffigen.
1365    - minor code tweaks
1366    
1367    ----------------------------------------------------------------------
1368    Name: Matthias Blume (blume (at) tti - c (dot) org)
1369    Date: 2004/07/12 22:50:00 CDT
1370    Tag: blume-110_46_1-macosx-nlffi
1371    Description:
1372    
1373    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1374    though.
1375    
1376    Note:  1. You have to make a new, clean build of the runtime system.
1377           2. There are new BOOTFILES, you have to use them!
1378              (Doing the bootstrap process yourself would be *very* painful!
1379               If you absolutely have to do it, build the system under
1380               a different architecture and then cross-compile.)
1381    
1382    Version bumped to 110.46.1 to account for runtime data format changes.
1383    
1384    ----------------------------------------------------------------------
1385    Name: Matthias Blume (blume (at) tti - c (dot) org)
1386    Date: 2004/06/18 14:30:00 CDT
1387    Tag: blume-20040618-unix
1388    Description:
1389    
1390    Changed the implementation of structure Unix so that the same stream
1391    is returned every time one of the {text,bin}{In,Out}streamOf functions
1392    is invoked on the same proc.  This is not what the spec currently
1393    says -- although IMO it arguably should.  (See discussion below.)
1394    
1395    ----------------------------------------------------------------------
1396    Name: Matthias Blume (blume (at) tti - c (dot) org)
1397    Date: 2004/06/17 18:15:00 CDT
1398    Tag: Release_110_46
1399    Description:
1400    
1401    New working version (110.46).  NEW BOOTFILES!
1402    
1403    ----------------------------------------------------------------------
1404    Name: Matthias Blume (blume (at) tti - c (dot) org)
1405    Date: 2004/06/17 17:20:00 CDT
1406    Tag: blume-20040617-timer-unix
1407    Description:
1408    
1409    Changed the interface of structures Timer and Unix to match the most
1410    recent Basis spec.
1411    
1412    In the case of Unix there still seems to be an open/weird issue:
1413    
1414       The {text,bin}{In,Out}streamOf functions are supposed to create
1415       fresh streams whenever they are called -- as opposed to have them
1416       return the same stream every time.  This design is supposed to
1417       prevent space leaks caused by proc values hanging on to streams.
1418    
1419       The reap function, on the other hand, is supposed to close the
1420       streams.  This cannot be done without having a handle on the
1421       stream in proc after all...
1422    
1423    I took the liberty to implement the following stopgap solution:
1424    
1425      The proc value hangs on to the most recently created stream(s).
1426      Reap closes those.  If either or both of the two streams hadn't
1427      been created at all yet, then reap will close the corresponding
1428      file descriptors directly.
1429    
1430    PS: I don't understand the original space leak argument anymore.  If
1431    a proc hangs on to the imperative stream, then I/O operations on those
1432    will advance the state of the cached stream and avoid the space leak.
1433    
1434    ----------------------------------------------------------------------
1435    Name: Matthias Blume (blume (at) tti - c (dot) org)
1436    Date: 2004/05/28 16:45:00 CDT
1437    Tag: blume-20040528-basis
1438    Description:
1439    
1440    Added signature PACK_REAL and exported functor PrimIO.
1441    
1442    ----------------------------------------------------------------------
1443    Name: Matthias Blume (blume (at) tti - c (dot) org)
1444    Date: 2004/05/25 16:00:00 CDT
1445    Tag: blume-20040525-group-owner
1446    Description:
1447    
1448    CM now ignores (but still accepts) the "owner" information in group
1449    descriptions.  The owner of a group is its next enclosing
1450    library. Each group must have a unique owner.  (There is a virtual
1451    "toplevel" library that owns groups which are not nested within a real
1452    library.)  Previously, each group had to explicitly declare its owner,
1453    and CM would check that such a declaration is correct.  The new scheme
1454    is to have CM check that for each group there is precisely one owning
1455    library.
1456    
1457    The advantage of the new scheme is that the programmer no longer needs
1458    to maintain the somewhat annoying owner information.  The downside is
1459    that CM cannot enforce the ownership rule across multiple runs of
1460    CM.make.  Fortunately, enclosing the same group in two different
1461    libraries A and B which are not part of the same program does not
1462    cause real problems.
1463    
1464    ----------------------------------------------------------------------
1465    Name: Matthias Blume (blume (at) tti - c (dot) org)
1466    Date: 2004/05/20 16:00:00 CDT
1467    Tag: blume-20040520-win32
1468    Description:
1469    
1470    Made the win32 version work again.  (Strangely, a misplaced comma had
1471    slipped into win32-process.c which prevented the runtime from being
1472    compiled correctly.)
1473    
1474    Also, included a minor addition to ml-build.bat analogous to what was
1475    done in blume-20040519-ml-build.
1476    
1477    ----------------------------------------------------------------------
1478    Name: Matthias Blume (blume (at) tti - c (dot) org)
1479    Date: 2004/05/19 22:10:00 CDT
1480    Tag: blume-20040519-ml-build
1481    Description:
1482    
1483    Arranged for ml-build to clean up after itself a little bit better.
1484    The script generates a temporary SML source file and compiles it using
1485    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1486    gets rid of those at the end, so they don't accumulate under .cm.
1487    
1488    This required a minor change to install.sh because the name of the
1489    metadata directory (default: .cm) is actually configurable at
1490    installation time.
1491    
1492    ----------------------------------------------------------------------
1493    Name: Matthias Blume (blume (at) tti - c (dot) org)
1494    Date: 2004/05/18 15:50:00 CDT
1495    Tag: blume-20040518-mkreader
1496    Description:
1497    
1498    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1499    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1500    
1501    ----------------------------------------------------------------------
1502    Name: Matthias Blume (blume (at) tti - c (dot) org)
1503    Date: 2004/05/11 14:35:00 CDT
1504    Tag: blume-20040511-win32sock
1505    Description:
1506    
1507    Added previously missing support for many socket-related functions
1508    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1509    for the voluminous patch!
1510    
1511    (I have not tested this patch under win32 yet.)
1512    
1513    Here is David's e-mail:
1514    
1515    Hi,
1516    
1517    Attached to this email you find a diff against sml/nj 110.45
1518    that will enable socket support under Windows.
1519    
1520    To apply the patch (using unix or cygwin)
1521    1) gunzip runtime.diff.gz
1522    2) "cd" into "src/runtime" in the source tree of a fresh
1523       110.45 installation.
1524    3) patch -p 1 < [your/path/to]runtime.diff
1525    
1526    The code compiles fine but has NOT yet been extensively tested.
1527    I only ran a few tests for basic socket client functionality
1528    (which worked fine).  Especially the functions that use ioctl
1529    are not tested at all and might not work (see below).
1530    
1531    I implemented this since we want to move to a newer version of sml/nj
1532    but need socket support in order to use it.  This is the first time I
1533    even had a look at the sml/nj source,  so please review my changes
1534    before making this part of the distribution!  Here are a few issues
1535    that I think might be better for someone to solve who is more
1536    familiar with the sml/nj source (and socket programming):
1537    
1538    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1539      exception since I could not figure out what the windows equivalent
1540      of these functions is
1541    
1542    - In sockets-osdep.h there are a some #include statements that are
1543      only used in a few files that include sockets-osdep.h
1544    
1545    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1546      does not process its return value since I don't know how to
1547      report an error upwards.
1548    
1549    - It would probably be good to have a call to WSACleanup() when
1550      the library is unloaded (if there is such a possibility).
1551      Otherwise I think Windows will take care of this automatically
1552      when the process finishes.
1553    
1554    - I used ioctlsocket() as a replacement for ioctl() but I have
1555      no idea if that is actually the proper replacement on Windows.
1556    
1557    - All these issues are marked in the code by "FIXME" comments.
1558    
1559    We use sml/nj extensively in our products and are quite happy
1560    with it.  I hope this contribution will help you.
1561    
1562    Keep up the good work!
1563    
1564    David
1565    
1566    ----------------------------------------------------------------------
1567    Name: Matthias Blume (blume (at) tti - c (dot) org)
1568    Date: 2004/05/11 14:20:00 CDT
1569    Tag: blume-20040511-installml
1570    Description:
1571    
1572    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1573    the patch.)
1574    
1575    ----------------------------------------------------------------------
1576    Name: Matthias Blume (blume (at) tti - c (dot) org)
1577    Date: 2004/05/11 14:05:00 CDT
1578    Tag: blume-20040511-nlffi-netbsd
1579    Description:
1580    
1581    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1582    the patch.)
1583    
1584    ----------------------------------------------------------------------
1585    Name: Matthias Blume (blume (at) tti - c (dot) org)
1586    Date: 2004/05/11 12:05:00 CDT
1587    Tag: blume-20040511-exports
1588    Description:
1589    
1590    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1591    export lists in compiler-related .cm-files.
1592    
1593    ----------------------------------------------------------------------
1594    Name: Matthias Blume (blume (at) tti - c (dot) org)
1595    Date: 2004/05/11 11:35:00 CDT
1596    Tag: blume-20040511-allsource
1597    Description:
1598    
1599    The installer now honors the "src-smlnj" target again, although its meaning
1600    has changed from "all sources required for the compiler" to "all sources
1601    the installer knows about".  In other words, if you enable "src-smlnj"
1602    in the "targets" file, then the installer will pull in sources for
1603    everything.  (Notice that this refers to source code only.  Compiled
1604    code is still only installed for modules that were requested explicitly
1605    or which are required for other modules that were requested explicitly.)
1606    
1607    ----------------------------------------------------------------------
1608    Name: Matthias Blume (blume (at) tti - c (dot) org)
1609    Date: 2004/04/23 17:40:00 CDT
1610    Tag: blume-20040423-ieee-scan
1611    Description:
1612    
1613    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1614    in the exponent calculation we get INF or ZERO (depending on the mantissa
1615    and the sign of the exponent).
1616    
1617    ----------------------------------------------------------------------
1618    Name: Matthias Blume (blume (at) tti - c (dot) org)
1619    Date: 2004/04/23 10:40:00 CDT
1620    Tag: blume-20040423-ml-build
1621    Description:
1622    
1623    The ml-build script now terminates with a non-0 status when something
1624    goes wrong.
1625    
1626    ----------------------------------------------------------------------
1627    Name: Matthias Blume (blume (at) tti - c (dot) org)
1628    Date: 2004/04/22 16:35:00 CDT
1629    Tag: blume-20040422-Option
1630    Description:
1631    
1632    Made exception Option to be the same as exception Option.Option
1633    (as it should be).
1634    
1635    ----------------------------------------------------------------------
1636    Name: Allen Leung (leunga (at) reservoir (dot) com)
1637    Date: 2004/03/19 14:40:00 EST
1638    Tag: leunga-20040319-cygwin-nlffi
1639    Description:
1640    
1641    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1642    of SML/NJ.  The problem is that
1643    
1644       lib = dlopen(NULL, ...)
1645       f   = dlsym(lib, "malloc");
1646    
1647    does not work on Windows unless we explicitly export symbols
1648    such as 'malloc' during linking.  We fixed this by explicitly
1649    exporting the required symbols with the magic gcc incantation:
1650    
1651        -Wl,--export-all cygwin.def
1652    
1653    where cygwin.def is a file containing all the symbols that
1654    we wish to export.
1655    
1656    I suspect this is a Windows problem and we'll have to
1657    do the same (somehow with windows compilers) when
1658    we build the native win32 version with the system
1659    calls LoadLibrary/GetProcAddress.
1660    
1661    ----------------------------------------------------------------------
1662    Name: Matthias Blume (blume (at) tti - c (dot) org)
1663    Date: 2004/03/04 16:35:00 CST
1664    Tag: blume-20040304-intinf-fmt
1665    Description:
1666    
1667    Fixed problem with IntInf.fmt (sign would show up on the right instead
1668    of on the left for BIN, OCT, and HEX).
1669    
1670    ----------------------------------------------------------------------
1671    Name: Matthias Blume (blume (at) tti - c (dot) org)
1672    Date: 2004/03/04 11:25:00 CST
1673    Tag: blume-20040304-symlinks
1674    Description:
1675    
1676    Fixed problem with installer script (unix only) where bin/ml-yacc and
1677    friends pointed (via symlinks) to absolute locations instead of just
1678    .run-sml.  This was reported by Vesa A Norrman.
1679    
1680    ----------------------------------------------------------------------
1681    Name: Matthias Blume (blume (at) tti - c (dot) org)
1682    Date: 2004/02/13 14:50:00 CST
1683    Tag: Release_110_45
1684    Description:
1685    
1686    New working version (110.45). New bootfiles.
1687    
1688    ----------------------------------------------------------------------
1689    Name: Matthias Blume (blume (at) tti - c (dot) org)
1690    Date: 2004/01/26 15:15:15 CST
1691    Tag: blume-20040126-toplevel
1692    Description:
1693    
1694    Improved handling of exceptions at the interactive toplevel.
1695    
1696    ----------------------------------------------------------------------
1697    Name: Matthias Blume (blume (at) tti - c (dot) org)
1698    Date: 2004/01/26 11:25:00
1699    Tag: blume-20040126-app
1700    Description:
1701    
1702    Type of top-level "app" corrected.
1703    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1704    
1705    ----------------------------------------------------------------------
1706    Name: Matthias Blume (blume (at) tti - c (dot) org)
1707    Date: 2003/11/18 17:10 CST
1708    Tag: blume-20031118-basis-fiddle
1709    Description:
1710    
1711    - changed Timer interface to what might become the spec
1712    - POSIX_FLAGS -> BIT_FLAGS according to spec
1713    - some other minor discrepancies wrt. spec eliminated
1714    
1715    ----------------------------------------------------------------------
1716    Name: Matthias Blume (blume (at) tti - c (dot) org)
1717    Date: 2003/11/06 12:00:00 CST
1718    Tag: Release_110_44
1719    Description:
1720    
1721    New working version (110.44). New bootfiles.
1722    
1723    ----------------------------------------------------------------------
1724    Name: Matthias Blume (blume (at) tti - c (dot) org)
1725    Date: 2003/11/04 11:50:00 CST
1726    Tag: blume-20031104-move-libraries
1727    Description:
1728    
1729    Eliminated the "dont_move_libraries" directive in config/targets.
1730    (The mechanism was broken and could not be fixed easily.  Moreover,
1731    there does not seem to be any reason not to move all libraries into
1732    lib during installation.  I originally implemented this directive as a
1733    backward-compatibility feature when I first introduced the new CM.
1734    Now that things have been stable for a long time and going back to the
1735    old CM is not an option, there is no reason to keep it around.)
1736    
1737    ----------------------------------------------------------------------
1738    Name: Matthias Blume (blume (at) tti - c (dot) org)
1739    Date: 2003/11/03 16:00:00 CST
1740    Tag: blume-20031103-installdir
1741    Description:
1742    
1743    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1744    Richards for pointing out the problem and providing the solution.)
1745    
1746    ----------------------------------------------------------------------
1747    Name: Matthias Blume (blume (at) tti - c (dot) org)
1748    Date: 2003/10/01 17:05:00 CDT
1749    Tag: blume-20031001-lal-mlrisc
1750    Description:
1751    
1752    MLRISC bug fix from Lal.
1753    
1754    ----------------------------------------------------------------------
1755    Name: Matthias Blume (blume (at) tti - c (dot) org)
1756    Date: 2003/09/30 16:10:00 CDT
1757    Tag: blume-20030930-primio-bat
1758    Description:
1759    
1760    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1761    2. Improved .bat files (for Win32 port) to make things work under Win95.
1762       (thanks to Aaron S. Hawley for this one)
1763    
1764    ----------------------------------------------------------------------
1765    Name: Matthias Blume (blume (at) tti - c (dot) org)
1766    Date: 2003/09/26 16:05:00 CDT
1767    Tag: blume-20030926-wrappriv
1768    Description:
1769    
1770    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1771    
1772    ----------------------------------------------------------------------
1773    Name: Matthias Blume (blume (at) tti - c (dot) org)
1774    Date: 2003/09/26 15:00:00 CDT
1775    Tag: blume-20030926-110_43_3
1776    Description:
1777    
1778    - additional cleanup
1779    - version number bump, NEW BOOTFILES
1780    
1781    ----------------------------------------------------------------------
1782    Name: Matthias Blume (blume (at) tti - c (dot) org)
1783    Date: 2003/09/26 12:00:00 CDT
1784    Tag: blume-20030926-ppautoload
1785    Description:
1786    
1787    I modified the read-eval-print loop so that the autoloader gets
1788    invoked whenever the prettyprinter tries to look up a symbol that
1789    is not currently defined in the toplevel environment but which
1790    appears in CM's autoload registry.  As a result, we see far fewer of
1791    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1792    
1793    In addition to this I tried to clean up some pieces of the Basis
1794    implementation (e.g., Socket, Word8Array) in order to prevent other
1795    instances of these ?.Foo.Bar.xxx names from being printed.
1796    
1797    The mechanism that picks names for types still needs some work, though.
1798    (Right now it seems that if there is a type A.t which is defined to
1799    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1800    "?.B.u" although the perhaps more sensible solution would be to use
1801    "A.t" in this case.  In other words, the prettyprinter should follow
1802    a chain of DEFtycs not farther than there are corresponding toplevel
1803    names in the current environment.)
1804    
1805    ----------------------------------------------------------------------
1806    Name: Matthias Blume (blume (at) tti - c (dot) org)
1807    Date: 2003/09/24 16:31:00 CDT
1808    Tag: blume-20030924-installer
1809    Description:
1810    
1811    Another installer tweak:  All the ML code for the installer is now
1812    compiled during CMB.make and put into a little library called
1813    $smlnj/installer.cm.  The installation then simply invokes
1814    
1815       sml -m $smlnj/installer.cm
1816    
1817    and everything happens automagically.
1818    
1819    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1820    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1821            and BINDIR.
1822    
1823    The new scheme guarantees that the ML code responsible for the installation
1824    is in sync with the APIs of the main system.  Also, the installer is
1825    somewhat faster because the installer script is precompiled.
1826    
1827    ----------------------------------------------------------------------
1828    Name: Matthias Blume (blume (at) tti - c (dot) org)
1829    Date: 2003/09/24 15:35:00 CDT
1830    Tag: blume-20030924-synsock
1831    Description:
1832    
1833    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1834    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1835    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1836    implementations of non-blocking operations from CML's Socket
1837    structure.
1838    
1839    ----------------------------------------------------------------------
1840    Name: Matthias Blume (blume (at) tti - c (dot) org)
1841    Date: 2003/09/24 15:10:05 CDT
1842    Tag: blume-20030924-sockets
1843    Description:
1844    
1845    1. Fixed SOCKET API and implementation to match Basis spec.
1846       This required changing the internal representation of sockets to one
1847       that remembers (for each socket file descriptor) whether it is currently
1848       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1849       call is made only if the state actually needs to change).
1850    
1851    2. OS-specific details of sockets were moved into separate files, thus
1852       making it possible to unify the bulk of the socket implementations
1853       between Unix and Win32.
1854    
1855    3. CML's socket API changed accordingly.
1856       (Note that we need to remove non-blocking functions from this API
1857       since they are redundant in the case of CML!)
1858    
1859    4. CML's socket implementation now makes use of non-blocking functions
1860       provided by Basis, thus removing all OS-dependent code from this part
1861       of CML.
1862    
1863    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1864    
1865    ----------------------------------------------------------------------
1866    Name: Matthias Blume (blume (at) tti - c (dot) org)
1867    Date: 2003/09/22 12:10:00 CDT
1868    Tag: blume-20030922-110_43_2
1869    Description:
1870    
1871    Made a new interim version and bootfiles for developer's bootstrapping
1872    convenience.
1873    
1874       110.43.2 -- NEW BOOTFILES
1875    
1876    ----------------------------------------------------------------------
1877    Name: Matthias Blume (blume (at) tti - c (dot) org)
1878    Date: 2003/09/19 15:55:00 CDT
1879    Tag: blume-20030919-cmdir
1880    Description:
1881    
1882    1. new-install.sh -> install.sh
1883    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1884    3. tweaked installer so that another name instead of .cm can be chosen
1885       at install time (by setting the CM_DIR_ARC environment variable
1886       during installation); once installation is complete, the name is
1887       fixed
1888    
1889    ----------------------------------------------------------------------
1890    Name: Matthias Blume (blume (at) tti - c (dot) org)
1891    Date: 2003/09/18 16:00:00 CDT
1892    Tag: blume-20030918-110_43_1
1893    Description:
1894    
1895    Made a new interim version and bootfiles for developer's bootstrapping
1896    convenience.
1897    
1898       110.43.1 -- NEW BOOTFILES
1899    
1900    ----------------------------------------------------------------------
1901    Name: Matthias Blume (blume (at) tti - c (dot) org)
1902    Date: 2003/09/18 15:20:00 CDT
1903    Tag: blume-20030918-misc
1904    Description:
1905    
1906    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1907       this seems to be controversial at the moment) and used those in
1908       Posix.ProcEnv.times.
1909    
1910    2. Added Time.{from,to}Nanoseconds to Time.
1911    
1912    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1913       For example, fromLargeInt never needs to look at more than 3 "big
1914       digits" to get its 53 bits of precision.
1915    
1916    ----------------------------------------------------------------------
1917    Name: Matthias Blume (blume (at) tti - c (dot) org)
1918    Date: 2003/09/17 16:30:00 CDT
1919    Tag: blume-20030917-real32-slices
1920    Description:
1921    
1922    Added an entry to the primitive environment
1923    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1924    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1925    
1926    Removed some of the "magic" constants in real64.sml and replaced them
1927    with code that generates these values from their corresponding
1928    integer counterparts.
1929    
1930    Made all(?) the slice-related changes to the Basis and made everything
1931    compile again...
1932    
1933    ----------------------------------------------------------------------
1934    Name: Matthias Blume (blume (at) tti - c (dot) org)
1935    Date: 2003/09/15 17:45:00 CDT
1936    Tag: blume-20030915-rbase
1937    Description:
1938    
1939    Fixed bug in Real.fromLargeInt.
1940    
1941    ----------------------------------------------------------------------
1942    Name: Matthias Blume (blume (at) tti - c (dot) org)
1943    Date: 2003/09/13 18:11:00 CDT
1944    Tag: blume-20030913-libinstall
1945    Description:
1946    
1947    Minor bugfix in config/libinstall (set anchor with path to
1948    standalone tool after installing it, otherwise libraries that
1949    need ml-lex or ml-yacc won't compile the first time the installer
1950    runs).
1951    
1952    ----------------------------------------------------------------------
1953    Name: Matthias Blume (blume (at) tti - c (dot) org)
1954    Date: 2003/09/12 11:45:00 CDT
1955    Tag: blume-20030912-various
1956    Description:
1957    
1958     - fixed bug in Real.toLargeInt
1959     - fixed bug in Posix.ProcEnv.times
1960     - changed inputLine functions to return an option
1961     - minor installer improvements / bugfixes
1962     - changed default @SMLalloc parameter for x86/celeron to 64k
1963    
1964    ----------------------------------------------------------------------
1965    Name: Matthias Blume (blume (at) tti - c (dot) org)
1966    Date: 2003/09/09 22:00:00 CDT
1967    Tag: Release_110_43
1968    Description:
1969    
1970        New working release 110.43.  New bootfiles.
1971    
1972    ----------------------------------------------------------------------
1973    Name: Matthias Blume (blume (at) tti - c (dot) org)
1974    Date: 2003/09/09 19:20:00 CDT
1975    Tag: blume-20030909-installer
1976    Description:
1977    
1978    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1979    Modified config/install.bat to take advantage of it.  Also modified
1980    config/install.sh (and called it config/new-install.sh) to take advantage
1981    of it on Unix systems.  (The SML code is (supposed to be) platform-
1982    independent.)
1983    
1984    The installer can now install everything under Win32
1985    as well as under *nix as long as it compiles.
1986    
1987    Other changes:
1988    
1989      - made CML compile again under Win32
1990      - made eXene compile under Win32 (by providing a fake structure UnixSock
1991        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1992      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1993        same as type int;  under Win32 it isn't
1994      - fixed some slice-related problems in the win32-specific parts of CML
1995      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1996        (under Win32, the volume name is case-insensitive, and the
1997        OS.Path code compares volume names for equality)
1998    
1999    ----------------------------------------------------------------------
2000    Name: Matthias Blume (blume (at) tti - c (dot) org)
2001    Date: 2003/09/08 11:55:00 CDT
2002    Tag: blume-20030908-fullpath
2003    Description:
2004    
2005    Made Win32 version of OS.FileSys.fullPath return current directory
2006    when given an empty string.  This is what the spec says, and incidentally,
2007    CM depends on it.  (CM otherwise goes into an infinite loop in certain
2008    cases when presented with the name of a non-existing .cm file.)
2009    
2010    ----------------------------------------------------------------------
2011    Name: Matthias Blume (blume (at) tti - c (dot) org)
2012    Date: 2003/09/04 16:30:00 CDT
2013    Tag: blume-20030905-slices-etc
2014    Description:
2015    
2016      1. Changed interface to vectors and arrays in Basis to match
2017         (draft) Basis spec.
2018      2. Added signatures and implementations of slices according to
2019         Basis spec.
2020      3. Edited source code throughout the system to make it compile again
2021         under 1. and 2.  (In some cases code had to be added to have it
2022         match the new signatures.)
2023      4. MLRISC should be backward-compatible: the copies of the originals
2024         of files that needed to change under 3. were retained, the .cm files
2025         check the compiler version number and use old versions when
2026         appropriate.
2027      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
2028         dirstream -> string option (in accordance with Basis spec).
2029      6. When generating code that counts lines, ml-lex used function
2030         CharVector.foldli, taking advantage of its old interface.
2031         This has been replaced with the corresponding code from
2032         CharVectorSlice.  (html-lex must be re-lexed!)
2033      7. BitArray in smlnj-lib/Util has been extended/modified to match the
2034         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
2035      8. Removed temporary additions (fromInternal, toInternal) from the
2036         (now obsolete) IntInf in smlnj-lib/Util.
2037      9. Cleaned up structure Byte.
2038     10. Added localOffset, scan, and fromString to Date (according to spec).
2039         Cleaned/corrected implementation of Date.
2040         (Still need to check for correctness; implement better canonicalizeDate.)
2041     11. Added "scan" to signature IEEE_REAL.
2042     12. Some improvements to IntInf [in particular: efficiency-hack for
2043         mod and rem when second operand is 2 (for parity checks).]
2044     13. Changed representation of type Time.time, using a single IntInf.int
2045         value counting microseconds.  This considerably simplified the
2046         implementation of structure Time.  We now support negative time
2047         values; scan and fromString handle signs.
2048     14. Functor PrimIO now takes two additional arguments (VectorSlice and
2049         ArraySlice).
2050    
2051    ----------------------------------------------------------------------
2052    Name: Matthias Blume (blume (at) tti - c (dot) org)
2053    Date: 2003/08/28 17:00:00 CDT
2054    Tag: blume-20030828-intinf
2055    Description:
2056    
2057    This is a major update which comes with a version number bump
2058    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
2059    and an implementation of IntInf in the Basis.
2060    
2061    There are a fairly large number of related changes and updates throughout
2062    the system:
2063    
2064    Basis:
2065            - Implemented IntInf.
2066            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
2067            - Added some missing Real64 operations, most notably Real.toLargeInt.
2068            - Added FixedInt as a synonym for Int32.
2069    
2070    compiler:
2071            * Added support for a built-in intinf type.
2072              - literals
2073              - pattern matching
2074              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2075              - overloading on literals and operations
2076    
2077            This required adding a primitive type intinf, some additional
2078            primops, and implementations for several non-trivial intinf
2079            operations in Core.  (The intinf type is completely abstract
2080            to the compiler; all operations get delegated back to the Core.)
2081    
2082            * Intinf equality is handled by polyequal.  However, the compiler
2083            does not print its usual warning in this case (since polyequal
2084            is the right thing to do there).
2085    
2086            * Improved the organization of structure InlineT.
2087    
2088            * A word about conversion primops:
2089            If conversions involving intinf do not cancel out during
2090            CPS contract, then the compiler must insert calls to Core functions.
2091            Since all core access must be resolved already during the FLINT
2092            translate phase, it would be too late a the time of CPS contract
2093            to add new Core calls.  For this reason, conversion primops
2094            for intinf carry two arguments:  1. the numeric argument that
2095            they are supposed to convert, and 2. the Core function that
2096            can help with this conversion if necessary.  If CPS contract
2097            eliminates a primop, then the associated Core function becomes
2098            dead and goes away.  Intinf conversion primops that do not get
2099            eliminated by CPS contract get rewritten into calls of their
2100            core functions by a separate, new phase.
2101    
2102    interactive system:
2103            - Control.Print.intinfDepth controls max length of intinf constants
2104            being printed. (Analogous to Control.Print.stringDepth.)
2105            - Cleanup in printutil and pputil: got rid of unused stuff and
2106            duplicates; replaced some of the code with code that makes better
2107            use of library functionality.
2108    
2109    CM:
2110            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2111            are no longer silent.
2112    
2113    CKIT:
2114            Fixed mismatched uses of Int32 and LargeInt.  I always decided
2115            in favor of LargeInt -- which is now the same as IntInf.
2116            CKIT-knowledgable people should check whether this is what's
2117            intended and otherwise change things back to using Int32 or
2118            FixedInt.
2119    
2120    Throughout the code:
2121            Started using IntInf.int literals and built-in operations
2122            (e.g., comparison with 0) where this seems appropriate.
2123    
2124    
2125    ----------------------------------------------------------------------
2126    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2127    Date: 2003/08/13 11:36:00 CDT
2128    Tag: dbm-20030813-mcz-merge1
2129    Description:
2130    
2131    Merging changes from the mcz-branch development branch into trunk.
2132    These changes involve replacement of the emulated old prettyprinter
2133    interface with direct use of the SML/NJ Lib PP library, and fixing
2134    of a couple of bugs (895, 1186) relating to error messages.  A new
2135    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2136    has been added.
2137    
2138    ----------------------------------------------------------------------
2139    Name: Matthias Blume (blume (at) tti - c (dot) org)
2140    Date: 2003/08/11 15:45:00 CDT
2141    Tag: blume-20030811-windows
2142    Description:
2143    
2144    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2145    
2146        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2147    
2148    This patch restores SML/NJ's ability to run under win32.  There are a
2149    number of changes, including fixes for several bugs that had gone
2150    unnoticed until now:
2151    
2152      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2153        trying to run the win32 version from within cygwin.)
2154    
2155      - There are a number of simple .bat scripts that substitute for their
2156        corresponding Unix shell-scripts.  (See below.)
2157    
2158      - The internals of ml-build have been modified slightly.  The main
2159        difference is that instead of calling ".link-sml" (or link-sml.bat)
2160        using OS.Process.system, the ML process delegates this task back
2161        to the script.  Otherwise problems arise in mixed environments such
2162        as Cygwin where scripts look and work like Unix scripts, but
2163        where OS.Process.system cannot run them.
2164    
2165      - In CM, the srcpath pickler used native pathname syntax -- which
2166        is incorrect in the case of cross-compilation.  The new pickle format
2167        is independent of platform-specific naming conventions.
2168    
2169      - Path configuration files (such as lib/pathconfig) can now choose
2170        between native and standard syntax.  Placing a line of the form
2171    
2172            standard!
2173    
2174        into the file causes all subsequent paths to be interpreted using
2175        CM standard pathname syntax (= Unix conventions); a line
2176    
2177            native!
2178    
2179        switches back to native style.  This was needed so that
2180        path config files can be written portably, see src/system/pathconfig.
2181    
2182      - Runtime system:
2183    
2184          - win32-filesys.c:  get_file_time and set_file_time now
2185            access modification time, not creation time.
2186    
2187          - I/O code made aware of new array representation.
2188    
2189          - Bug fixes in X86.prim.masm.
2190    
2191       - src/system/makeml made aware of win32. (For use under cygwin
2192            and other Unix-environments for windows.)
2193    
2194       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2195         which caused BinIO.inputAll to fail consistently.
2196    
2197    .bat scripts:
2198    
2199       Windows .bat scripts assume that SMLNJ_HOME is defined.
2200    
2201       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2202         applications (sml, ml-yacc, ml-lex).
2203       - ml-build.bat: analogous to ml-build.
2204       - config\install.bat:  Analogous to config/install.sh.  This requires
2205         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2206         (nmake etc. must be on the path, and vcvars32 must have been run.)
2207         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2208         and the bootfile hierarchy must have been unpacked under
2209         sml.boot.x86-win32.
2210         The script is very primitive and does a poor job at error checking.
2211         It only installs the base system, ml-lex, and ml-yacc.  No other
2212         libraries are being installed (i.e., you get only those that
2213         are part of the compiler.)
2214       - link-sml.bat:  analogous to .link-sml, but not currently used
2215    
2216    Unrelated bug fixes:
2217    
2218      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2219        types.
2220      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2221        otherwise.)
2222    
2223    ----------------------------------------------------------------------
2224    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2225    Date: 2003/06/17
2226    Tag: macqueen-20030617-bug895
2227    Description:
2228    
2229    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2230    Tag will be used for new development branch (mcz-branch) for
2231    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2232    
2233    ----------------------------------------------------------------------
2234    Name: Matthias Blume (blume (at) tti - c (dot) org)
2235    Date: 2003/05/27 16:55:00 CDT
2236    Tag: blume-20030527-polyeq
2237    Description:
2238    
2239    Tried to eliminated most cases of polymorphic equality.
2240    
2241    ----------------------------------------------------------------------
2242    Name: Matthias Blume (blume (at) tti - c (dot) org)
2243    Date: 2003/05/21 17:45:00 CDT
2244    Tag: blume-20030517-complete
2245    Description:
2246    
2247    Two changes:
2248    
2249      1. Added a flag for controlling whether non-exhaustive bindings will
2250         be treated as errors (default is false).
2251      2. Cleaned up the *entire* source tree so that CMB.make goes through
2252         without a single non-exhaustive match- or bind warning.
2253    
2254    ----------------------------------------------------------------------
2255    Name: Matthias Blume (blume (at) tti - c (dot) org)
2256    Date: 2003/05/17 10:20:00 CDT
2257    Tag: blume-20030517-absyn
2258    Description:
2259    
2260    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2261    
2262       This mainly affects the quality of error messages.  However, some
2263       of the code is now more straightforward than before. (Treatment of
2264       the above four constructs in translate.sml is much simpler than
2265       the "macro-expansion" that was going on before.  Plus, the mach-
2266       compiler no longer gets invoked just to be able to compile an
2267       if-expression.)
2268    
2269    2. The ErrorMsg.Error exception is now caught and absorbed by the
2270       interactive loop.
2271    
2272    ----------------------------------------------------------------------
2273    Name: Allen Leung
2274    Date: 2003/05/16 13:05:00 CDT
2275    Tag: leunga-20030516-cygwin-runtime
2276    Description:
2277    
2278        Ported the runtime system to cygwin, which uses the unix
2279    x86-unix bin files.  Missing/buggy features:
2280    
2281       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2282         the Cygwin library.
2283       o Ctrl-C handling may be flaky.
2284       o Windows system calls and Windows I/O are not supported.
2285    
2286       A new set of binfiles is located at:
2287    
2288         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2289    
2290       This is only needed for bootstrapping the cygwin version of smlnj.
2291       Other x86 versions can use the existing binfiles.
2292    
2293    ----------------------------------------------------------------------
2294    Name: Matthias Blume
2295    Date: 2003/04/08 15:42:00 CDT
2296    Tag: blume-20030408-listpair
2297    Description:
2298    
2299    1. Added a target 'mlrisc' to installer.
2300    
2301    2. Added missing elements to structure ListPair.
2302    
2303    ----------------------------------------------------------------------
2304    Name: Allen Leung
2305    Date: 2003/01/07 10:40:00 EST
2306    Tag: leunga-20030107-int-rem
2307    Description:
2308    
2309    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2310    The arguments to the SUBL instruction were swapped.
2311    
2312    ----------------------------------------------------------------------
2313    Name: Matthias Blume
2314    Date: 2002/12/12 16:25:00 EST
2315    Tag: blume-20021212-risc-ra
2316    Description:
2317    
2318    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2319    sent the RA into an infinite loop when floating point registers get
2320    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2321    110.37 and 110.38.)
2322    There was another set of potential problems related to the handling of
2323    MLRISC annotations (but those did not yet cause real problems, apparently).
2324    
2325    ----------------------------------------------------------------------
2326    Name: Matthias Blume
2327    Date: 2002/12/06 22:40:00 EST
2328    Tag: blume-20021206-cm-fileid
2329    Description:
2330    
2331    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2332    This fixes the problem of CM getting confused by files that suddenly
2333    change their identity (e.g., by getting unlinked and recreated by some
2334    text editor such as vi).  There might be a better/cheaper/cleaner way
2335    of doing this, but for now this will have to do.
2336    
2337    ----------------------------------------------------------------------
2338    Name: Matthias Blume
2339    Date: 2002/10/28 09:50:00 EST
2340    Tag: blume-20021028-typecheck
2341    Description:
2342    
2343    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2344    
2345    ----------------------------------------------------------------------
2346    Name: Matthias Blume
2347    Date: 2002/10/17 09:10:00 EDT
2348    Tag: Release_110_42
2349    Description:
2350    
2351    In good old tradition, there has been a slight hiccup so that we have
2352    to patch 110.42 after the fact.   The old release tag has been replaced
2353    (see below).
2354    
2355    The change solves a problem with two competing approaches the
2356    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2357    each other's way.
2358    
2359    This change only affects the runtime system code and the installer script.
2360    (No new bootfiles.)
2361    
2362    ----------------------------------------------------------------------
2363    Name: Matthias Blume
2364    Date: 2002/10/16 12:00:00 EDT
2365    Tag: Release_110_42_removed
2366    Description:
2367    
2368    New working release.  New bootfiles.
2369    
2370    ----------------------------------------------------------------------
2371    Name: Matthias Blume
2372    Date: 2002/10/10 13:10:00 EDT
2373    Tag: blume-20021010-ppc-divs
2374    Description:
2375    
2376    The mltree operator DIVS must be implemented with an overflow check on
2377    the PPC because the hardware indicates divide-by-zero using "overflow" as
2378    well.
2379    
2380    ----------------------------------------------------------------------
2381    Name: Matthias Blume
2382    Date: 2002/07/23 11:45:00
2383    Tag: blume-20020723-smlnj-home
2384    Description:
2385    
2386    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2387    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2388    CM_PATHCONFIG is also set), the path configuration file is assumed
2389    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2390    the entire tree to some other place and everything will "just work".
2391    
2392    (Companion commands such as ml-build and ml-makedepend also sense this
2393    variable.)
2394    
2395    ----------------------------------------------------------------------
2396    Name: Matthias Blume
2397    Date: 2002/07/12 21:19:00 EDT
2398    Tag: blume-20020712-liveness
2399    Description:
2400    
2401    Exported two useful "step" functions from liveness module (MLRISC).
2402    
2403    ----------------------------------------------------------------------
2404    Name: Matthias Blume
2405    Date: 2002/07/05 16:00 EDT
2406    Tag: Release_110_41
2407    Description:
2408    
2409    New working release.  New bootfiles.
2410    
2411    ----------------------------------------------------------------------
2412    Name: Matthias Blume
2413    Date: 2002/07/05 10:25:00 EDT
2414    Tag: blume-20020705-btimp
2415    Description:
2416    
2417    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2418    other clients can set up backtracing support.
2419    
2420    ----------------------------------------------------------------------
2421    Name: Matthias Blume
2422    Date: 2002/06/25 17:23:00 EDT
2423    Tag: blume-20020625-fpmax
2424    Description:
2425    
2426    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2427    case.  (The sense of the isNaN test was reversed -- which made min and
2428    max always return their first argument.)
2429    
2430    ----------------------------------------------------------------------
2431    Name: Matthias Blume
2432    Date: 2002/06/11
2433    Tag: blume-20020611-unixpath
2434    Description:
2435    
2436    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2437    
2438    ----------------------------------------------------------------------
2439    Name: Matthias Blume
2440    Date: 2002/06/10 16:35:00 EDT
2441    Tag: blume-20020610-ieeereal
2442    Description:
2443    
2444    I back-ported my implementation of IEEEReal.fromString from the
2445    idlbasis-devel branch so that we can test it.
2446    
2447    Another small change is that ppDec tries to give more information
2448    than just "<sig>" in the case of functors.  However, this code is
2449    broken in some mysterious way if the functor's body's signature
2450    has not been declared by ascription but gets inferred from the
2451    implementation.  This needs fixing...
2452    
2453    ----------------------------------------------------------------------
2454    Name: Matthias Blume
2455    Date: 2002/05/31
2456    Tag: blume-20020531-btrace-mode
2457    Description:
2458    
2459    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2460    the wayside when I switched over to using Controls everywhere.)
2461    
2462    ----------------------------------------------------------------------
2463    Name: Lal George
2464    Date: 2002/05/23 12:21:40 EDT
2465    Tag: george-20020523-visual-labels
2466    Description:
2467    
2468    Labels are now displayed in the graphical output to make
2469    the fall-through and target blocks obvious.
2470    
2471    ----------------------------------------------------------------------
2472    Name: Matthias Blume
2473    Date: 2002/05/22 11:03:00 EDT
2474    Tag: blume-20020522-shrink
2475    Description:
2476    
2477    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2478    situation that comes up when allocating huge arrays.
2479    
2480    ----------------------------------------------------------------------
2481    Name: Matthias Blume
2482    Date: 2002/05/21 16:00:00 EDT
2483    Tag: Release_110_40
2484    Description:
2485    
2486    New working release (110.40).  New bootfiles.
2487    
2488    [Also: John Reppy fixed GC bug 1131.]
2489    
2490    ----------------------------------------------------------------------
2491    Name: Matthias Blume
2492    Date: 2002/05/21 12:35:00 EDT
2493    Tag: blume-20020521-cmdoc
2494    Description:
2495    
2496    CM documentation update.
2497    
2498    ----------------------------------------------------------------------
2499    Name: Matthias Blume
2500    Date: 2002/05/21 10:55:00 EDT
2501    Tag: blume-20020521-misc
2502    Description:
2503    
2504    - John tweaked runtime to be silent on heap export (except when
2505      GC messages are on).
2506    - I added a few more things (cross-compiling versions of CMB) to
2507      config/preloads (as suggestions).
2508    
2509    ----------------------------------------------------------------------
2510    Name: Matthias Blume
2511    Date: 2002/05/20 22:25:00 EDT
2512    Tag: blume-20020520-controls
2513    Description:
2514    
2515    - Added ControlUtil structure to control-lib.cm.
2516    - Use it throughout.
2517    - Used Controls facility to define MLRISC controls (as opposed to
2518      registering MLRISC control ref cells with Controls after the fact)
2519    - Fixed messed-up controls priorities.
2520    
2521    * Removed again all the stuff from config/preloads that one wouldn't
2522      be able to preload at the time the initial heap image is built.
2523      (Many libraries, e.g., CML, do not exist yet at this time.  The
2524       only libraries that can be preloaded via config/preloads are those
2525       that come bundled with the bootfiles.)
2526    
2527    ----------------------------------------------------------------------
2528    Name: Matthias Blume
2529    Date: 2002/05/20 10:59:00 EDT
2530    Tag: blume-20020520-preloads
2531    Description:
2532    
2533    Added a lot of commented-out suggestions for things to be included
2534    in config/preloads.
2535    
2536    ----------------------------------------------------------------------
2537    Name: Allen Leung
2538    Date: 2002/05/18 14:20:00 EDT
2539    Tag: leunga-20020518-mdl
2540    Description:
2541    
2542    o Made the mdl tool stuff compile and run again.
2543    o I've disabled all the stuff that depends on RTL specifications; they
2544      are all badly broken anyway.
2545    
2546    ----------------------------------------------------------------------
2547    Name: Matthias Blume
2548    Date: 2002/05/17 16:49:00 EDT
2549    Tag: blume-20020517-controls
2550    Description:
2551    
2552    0. John Reppy made several modifications to the SML/NJ library.
2553       In particular, there is a shiny new controls-lib.cm.
2554    
2555    1. Pushed new controls interface through compiler so that everything
2556       compiles again.
2557    
2558    2. Added FormatComb and FORMAT_COMB to the CML version of the
2559       SML/NJ library (so that CML compiles again).
2560    
2561    3. Modified init scripts because XXX_DEFAULT environment variables
2562       are no longer with us.  (Boot-time initialization is now done
2563       using the same environment variables that are also used for
2564       startup-time initialization of controls.)
2565    
2566    ----------------------------------------------------------------------
2567    Name: Lal George
2568    Date: 2002/05/15 09:20:10 EDT
2569    Tag: george-20020515-pseudo-op-decls
2570    Description:
2571    
2572            All pseudo-ops emitted before the first segment declaration
2573    such as TEXT, DATA, and BSS directives are assumed to be global
2574    declarations and are emitted first in the assembly file. This is
2575    useful in a number of situations where one has pseudo-ops that are not
2576    specific to any segment, and also works around the constraint that one
2577    cannot have client pseudo-ops in the TEXT segment.
2578    
2579            Because no segment is associated with these declarations it is
2580    an error to allocate any space or objects before the first segment
2581    directive and an exception will be raised. However, we cannot make
2582    this check for client pseudo-ops.
2583    
2584            These top level declarations are a field in the CFG graph_info.
2585    In theory you can continue to add to this field after the CFG has been
2586    built -- provided you know what you are doing;-)
2587    
2588    ----------------------------------------------------------------------
2589    Name: Matthias Blume
2590    Date: 2002/05/13 16:40:00 EDT
2591    Tag: blume-20020513-pp-etc
2592    Description:
2593    
2594    A few minor bugfixes:
2595    
2596      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2597        (Bogus printouts for redefined bindings still occur.  Compiler
2598        bug should no longer occur now.  We need to redo the prettyprinter
2599        from scratch.)
2600    
2601      - CM pathname printer now also adds escape sequences for ( and )
2602    
2603      - commend and docu fixes for ml-nlffi
2604    
2605    ----------------------------------------------------------------------
2606    Name: Matthias Blume
2607    Date: 2002/05/10 16:40:00 EDT
2608    Tag: blume-20020510-erg-textio
2609    Description:
2610    
2611    Applied the following bugfix provided by Emden Gansner:
2612    
2613        Output is corrupted when outputSubstr is used rather than output.
2614    
2615        The problem occurs when a substring
2616    
2617            ss = (s, dataStart, dataLen)
2618    
2619        where dataStart > 0, fills a stream buffer with avail bytes left.
2620        avail bytes of s, starting at index dataStart, are copied into the
2621        buffer, the buffer is flushed, and then the remaining dataLen-avail
2622        bytes of ss are copied into the beginning of the buffer. Instead of
2623        starting this copy at index dataStart+avail in s, the current code
2624        starts the copy at index avail.
2625    
2626        Fix:
2627        In text-io-fn.sml, change line 695 from
2628             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2629        to
2630             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2631    
2632    ----------------------------------------------------------------------
2633    Name: Matthias Blume
2634    Date: 2002/04/12 13:55:00 EDT
2635    Tag: blume-20020412-assyntax
2636    Description:
2637    
2638    1. Grabbed newer assyntax.h from the XFree86 project.
2639    2. Fiddled with how to compile X86.prim.asm without warnings.
2640    3. (Very) Minor cleanup in CM.
2641    
2642    ----------------------------------------------------------------------
2643    Name: Matthias Blume
2644    Date: 2002/04/01 (no joke!) 17:07:00 EST
2645    Tag: blume-20020401-x86div
2646    Description:
2647    
2648    Added full support for div/mod/rem/quot on the x86, using the machine
2649    instruction's two results (without clumsily recomputing the remainder)
2650    directly where appropriate.
2651    
2652    Some more extensive power-of-two support was added to the x86 instruction
2653    selector (avoiding expensive divs, mods, and muls where they can be
2654    replaced with cheaper shifts and masks).  However, this sort of thing
2655    ought to be done earlier, e.g., within the CPS optimizer so that
2656    all architectures benefit from it.
2657    
2658    The compiler compiles to a fixed point, but changes might be somewhat
2659    fragile nevertheless.  Please, report any strange things that you might
2660    see wrt. div/mod/quot/rem...
2661    
2662    ----------------------------------------------------------------------
2663    Name: Matthias Blume
2664    Date: 2002/03/29 17:22:00
2665    Tag: blume-20020329-div
2666    Description:
2667    
2668    Fixed my broken div/mod logic.  Unfortunately, this means that the
2669    inline code for div/mod now has one more comparison than before.
2670    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2671    The problem was with quotient = 0, because that alone does not tell
2672    us which way the rounding went.  One then has to look at whether
2673    remainder and divisor have the same sign...  :(
2674    
2675    Anyway, I replaced the bootfiles with fresh ones...
2676    
2677    ----------------------------------------------------------------------
2678    Name: Matthias Blume
2679    Date: 2002/03/29 14:10:00 EST
2680    Tag: blume-20020329-inlprims
2681    Description:
2682    
2683    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2684    
2685    Primops have changed. This means that the bin/boot-file formats have
2686    changed as well.
2687    
2688    To make sure that there is no confusion, I made a new version.
2689    
2690    
2691    CHANGES:
2692    
2693    * removed REMT from mltree (remainder should never overflow).
2694    
2695    * added primops to deal with divisions of all flavors to the frontend
2696    
2697    * handled these primops all the way through so they map to their respective
2698      MLRISC support
2699    
2700    * used these primops in the implementation of Int, Int32, Word, Word32
2701    
2702    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2703    
2704    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2705    
2706    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2707      floating point
2708    
2709    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2710      and Real (but Real.abs maps to a separate floating-point-only primop)
2711    
2712    
2713    TODO items:
2714    
2715    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2716      instructions because the machine instruction encoder cannot handle
2717      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2718      REM in the first place, and REM is supported by the x86 machine coder.)
2719    
2720    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2721      selection phase.  (The two can be streamlined because the hardware
2722      delivers both quotient and remainder at the same time anyway.)
2723    
2724    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2725      (Currently the behavior is inconsistent both across architectures and
2726      wrt. the draft Basis spec.)
2727    
2728    * Word8 should eventually be handled natively, too.
2729    
2730    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2731      as if there currently is no execution path that could trigger it in
2732      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2733      hold for things like multiplication and division.)
2734    
2735    ----------------------------------------------------------------------
2736    Name: Matthias Blume
2737    Date: 2002/03/27 16:27:00 EST
2738    Tag: blume-20020327-mlrisc-divisions
2739    Description:
2740    
2741    Added support for all four division operations (ML's div, mod, quot,
2742    and rem) to MLRISC.  In the course of doing so, I also rationalized
2743    the naming (no more annoying switch-around of DIV and QUOT), by
2744    parameterizing the operation by div_rounding_mode (which can be either
2745    DIV_TO_ZERO or DIV_TO_NEGINF).
2746    
2747    The generic MLTreeGen functor takes care of compiling all four
2748    operations down to only round-to-zero div.
2749    
2750    Missing pieces:
2751    
2752      * Doing something smarter than relying on MLTreeGen on architectures
2753        like, e.g., the x86 where hardware division delivers both quotient and
2754        remainder at the same time.  With this, the implementation of the
2755        round-to-neginf operations could be further streamlined.
2756    
2757      * Remove inlining support for div/mod/rem from the frontend and replace it
2758        with primops that get carried through to the backend.  Do this for all
2759        int and word types.
2760    
2761    ----------------------------------------------------------------------
2762    Name: Matthias Blume
2763    Date: 2002/03/25 17:25:00 EST
2764    Tag: blume-20020325-divmod
2765    Description:
2766    
2767    I improved (hopefully without breaking them) the implementation of Int.div,
2768    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2769    advantage of the following observations:
2770    
2771      Let  q = x quot y      r = x rem y
2772           d = x div  y      m = x mod y
2773    
2774    where "quot" is the round-to-zero version of integer division that
2775    hardware usually provides.  Then we have:
2776    
2777         r = x - q * y        where neither the * nor the - will overflow
2778         d = if q >= 0 orelse x = q * y then q else q - 1
2779                              where neither the * nor the - will overflow
2780         m = if q >= 0 orelse r = 0 then r else r + y
2781                              where the + will not overflow
2782    
2783    This results in substantial simplification of the generated code.
2784    The following table shows the number of CFG nodes and edges generated
2785    for
2786            fun f (x, y) = x OPER y
2787            (* with OPER \in div, mod, quot, rem *)
2788    
2789    
2790        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2791        --------------------------------------------------------
2792         div |         24 |         39 |         12 |         16
2793         mod |         41 |         71 |         12 |         16
2794        quot |          8 |         10 |          8 |         10
2795         rem |         10 |         14 |          8 |         10
2796    
2797    
2798    ----------------------------------------------------------------------
2799    Name: Matthias Blume
2800    Date: 2002/03/25 22:06:00 EST
2801    Tag: blume-20020325-cprotobug
2802    Description:
2803    
2804    Fixed a bug in cproto (c prototype decoder).
2805    
2806    ----------------------------------------------------------------------
2807    Name: Matthias Blume
2808    Date: 2002/03/25 16:00:00 EST
2809    Tag: blume-20020325-raw-primops
2810    Description:
2811    
2812    I did some cleanup to Allen's new primop code and
2813    replaced yesterday's bootfiles with new ones.
2814    (But they are stored in the same place.)
2815    
2816    ----------------------------------------------------------------------
2817    Name: Matthias Blume
2818    Date: 2002/03/24 22:40:00 EST
2819    Tag: blume-20020324-bootfiles
2820    Description:
2821    
2822    Made the bootfiles that Allen asked for.
2823    
2824    ----------------------------------------------------------------------
2825    Name: Allen Leung
2826    Date: 2002/03/23 15:50:00 EST
2827    Tag: leunga-20020323-flint-cps-rcc-primops
2828    Description:
2829    
2830      1. Changes to FLINT primops:
2831    
2832        (* make a call to a C-function;
2833         * The primop carries C function prototype information and specifies
2834         * which of its (ML-) arguments are floating point. C prototype
2835         * information is for use by the backend, ML information is for
2836         * use by the CPS converter. *)
2837      | RAW_CCALL of { c_proto: CTypes.c_proto,
2838                       ml_args: ccall_type list,
2839                       ml_res_opt: ccall_type option,
2840                       reentrant : bool
2841                     } option
2842       (* Allocate uninitialized storage on the heap.
2843        * The record is meant to hold short-lived C objects, i.e., they
2844        * are not ML pointers.  With the tag, the representation is
2845        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2846        *)
2847      | RAW_RECORD of {tag:bool,sz:int}
2848      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2849    
2850      2.  These CPS primops are now overloaded:
2851    
2852           rawload of {kind:numkind}
2853           rawstore of {kind:numkind}
2854    
2855          The one argument form is:
2856    
2857             rawload {kind} address
2858    
2859          The two argument form is:
2860    
2861             rawload {kind} [ml object, byte-offset]
2862    
2863      3. RAW_CCALL/RCC now takes two extra arguments:
2864    
2865         a. The first is whether the C call is reentrant, i.e., whether
2866            ML state should be saved and restored.
2867         b. The second argument is a string argument specifying the name of
2868            library and the C function.
2869    
2870         These things are currently not handled in the code generator, yet.
2871    
2872      4. In CProto,
2873    
2874         An encoding type of "bool" means "ml object" and is mapped into
2875         C prototype of PTR.  Note that "bool" is different than "string",
2876         even though "string" is also mapped into PTR, because "bool"
2877         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2878    
2879      5. Pickler/unpicker
2880    
2881         Changed to handle RAW_RECORD and newest RAW_CCALL
2882    
2883      6. MLRiscGen,
2884    
2885         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2886         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2887            in the file CodeGen/cpscompile/cps-c-calls.sml
2888    
2889      7. Added the conditional move operator
2890    
2891             condmove of branch
2892    
2893         to cps.  Generation of this is still buggy so it is currently
2894         disabled.
2895    
2896    ----------------------------------------------------------------------
2897    Name: Lal George
2898    Date: 2002/03/22 14:18:25 EST
2899    Tag: george-20020322-cps-branch-prob
2900    Description:
2901    
2902    Implemented the Ball-Larus branch prediction-heuristics, and
2903    incorporated graphical viewers for control flow graphs.
2904    
2905    Ball-Larus Heuristics:
2906    ---------------------
2907    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2908    
2909    By design it uses the Dempster-Shafer theory for combining
2910    probabilities.  For example, in the function:
2911    
2912        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2913    
2914    the ball-larus heuristics predicts that the n=0 is unlikely
2915    (OH-heuristic), and the 'then' branch is unlikely because of the
2916    RH-heuristic -- giving the 'then' branch an even lower combined
2917    probability using the Dempster-Shafer theory.
2918    
2919    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2920    probability of the 'then' branch because of the loop in the else
2921    branch.
2922    
2923    
2924    Graphical Viewing:
2925    ------------------
2926    I merely plugged in Allen's graphical viewers into the compiler. The
2927    additional code is not much. At the top level, saying:
2928    
2929            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2930    
2931    will display the graphical view of the control flow graph just before
2932    back-patching.  daVinci must be in your path for this to work. If
2933    daVinci is not available, then the default viewer can be changed
2934    using:
2935    
2936            Control.MLRISC.getString "viewer"
2937    
2938    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2939    course, these viewers must be in your path.
2940    
2941    The above will display the compilation unit at the level of clusters,
2942    many of which are small, boring, and un-interesting. Also setting:
2943    
2944            Control.MLRISC.getInt "cfg-graphical-view_size"
2945    
2946    will display clusters that are larger than the value set by the above.
2947    
2948    
2949    ----------------------------------------------------------------------
2950    Name: Matthias Blume
2951    Date: 2002/03/21 22:20:00 EST
2952    Tag: blume-20020321-kmp-bugfix
2953    Description:
2954    
2955    Changed the interface to the KMP routine in PreString and fixed
2956    a minor bug in one place where it was used.
2957    
2958    ----------------------------------------------------------------------
2959    Name: Allen Leung
2960    Date: 2002/03/21 20:30:00 EST
2961    Tag: leunga-20020321-cfg
2962    Description:
2963    
2964      Fixed a potential problem in cfg edge splitting.
2965    
2966    ----------------------------------------------------------------------
2967    Name: Allen Leung
2968    Date: 2002/03/21 17:15:00 EST
2969    Tag: leunga-20020321-x86-fp-cfg
2970    Description:
2971    
2972      1. Recoded the buggy parts of x86-fp.
2973    
2974         a. All the block reordering code has been removed.
2975            We now depend on the block placement phases to do this work.
2976    
2977         b. Critical edge splitting code has been simplified and moved into the
2978            CFG modules, as where they belong.
2979    
2980         Both of these were quite buggy and complex.  The code is now much, much
2981         simpler.
2982    
2983      2. X86 backend.
2984    
2985         a. Added instructions for 64-bit support.  Instruction selection for
2986            64-bit has not been committed, however, since that
2987            requires changes to MLTREE which haven't been approved by
2988            Lal and John.
2989    
2990         b. Added support for FUCOMI and FUCOMIP when generating code for
2991            PentiumPro and above.  We only generate these instructions in
2992            the fast-fp mode.
2993    
2994         c. Added cases for JP and JNP in X86FreqProps.
2995    
2996      3. CFG
2997    
2998         CFG now has a bunch of methods for edge splitting and merging.
2999    
3000      4. Machine description.
3001    
3002         John's simplification of MLTREE_BASIS.fcond broke a few machine
3003         description things:
3004    
3005         rtl-build.{sig,sml} and hppa.mdl fixed.
3006    
3007         NOTE: the machine description stuff in the repository is still broken.
3008               Again, I can't put my fixes in because that involves
3009               changes to MLTREE.
3010    
3011    ----------------------------------------------------------------------
3012    Name: Matthias Blume
3013    Date: 2002/03/20 15:55:00 EST
3014    Tag: blume-20020320-kmp
3015    Description:
3016    
3017    Implemented Knuth-Morris-Pratt string matching in PreString and used
3018    it for String.isSubstring, Substring.isSubstring, and
3019    Substring.position.
3020    
3021    (Might need some stress-testing.  Simple examples worked fine.)
3022    
3023    ----------------------------------------------------------------------
3024    Name: Matthias Blume
3025    Date: 2002/03/19 16:37:00 EST
3026    Tag: blume-20020319-witnesses
3027    Description:
3028    
3029    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
3030    
3031    This implements a generic mechanism for changing constness qualifiers
3032    anywhere within big C types without resorting to outright "casts".
3033    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
3034    modify the constness at the outermost level.)
3035    The implementation of "convert" is based on the idea of "witness"
3036    values -- values that are not used by the operation but whose types
3037    "testify" to their applicability.  On the implementation side, "convert"
3038    is simply a projection (returning its second curried argument).  With
3039    cross-module inlining, it should not result in any machine code being
3040    generated.
3041    
3042    ----------------------------------------------------------------------
3043    Name: Matthias Blume
3044    Date: 2002/03/15 16:40:00 EST
3045    Tag: blume-20020315-basis
3046    Description:
3047    
3048    Provided (preliminary?) implementations for
3049    
3050      {String,Substring}.{concatWith,isSuffix,isSubstring}
3051    
3052    and
3053    
3054      Substring.full
3055    
3056    Those are in the Basis spec but they were missing in SML/NJ.
3057    
3058    ----------------------------------------------------------------------
3059    Name: Matthias Blume
3060    Date: 2002/03/14 21:30:00 EST
3061    Tag: blume-20020314-controls
3062    Description:
3063    
3064    Controls:
3065    ---------
3066    
3067    1. Factored out the recently-added Controls : CONTROLS stuff and put
3068       it into its own library $/controls-lib.cm.  The source tree for
3069       this is under src/smlnj-lib/Controls.
3070    
3071    2. Changed the names of types and functions in this interface, so they
3072       make a bit more "sense":
3073    
3074          module -> registry
3075          'a registry -> 'a group
3076    
3077    3. The interface now deals in ref cells only.  The getter/setter interface
3078       is (mostly) gone.
3079    
3080    4. Added a function that lets one register an already-existing ref cell.
3081    
3082    5. Made the corresponding modifications to the rest of the code so that
3083       everything compiles again.
3084    
3085    6. Changed the implementation of Controls.MLRISC back to something closer
3086       to the original.  In particular, this module (and therefore MLRISC)
3087       does not depend on Controls.  There now is some link-time code in
3088       int-sys.sml that registers the MLRISC controls with the Controls
3089       module.
3090    
3091    CM:
3092    ---
3093    
3094      * One can now specify the lambda-split aggressiveness in init.cmi.
3095    
3096    ----------------------------------------------------------------------
3097    Name: Allen Leung
3098    Date: 2002/03/13 17:30:00 EST
3099    Tag: leunga-20020313-x86-fp-unary
3100    Description:
3101    
3102    Bug fix for:
3103    
3104    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3105    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3106    > - fun f(x,(y,z)) = Real.~ y;
3107    > [autoloading]
3108    > [autoloading done]
3109    >       fchsl   (%eax), 184(%esp)
3110    > Error: MLRisc bug: X86MCEmitter.emitInstr
3111    >
3112    > uncaught exception Error
3113    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3114    
3115    The problem was that the code generator did not generate any fp registers
3116    in this case, and the ra didn't know that it needed to run the X86FP phase to
3117    translate the pseudo fp instruction.   This only happened with unary fp
3118    operators in certain situations.
3119    
3120    ----------------------------------------------------------------------
3121    Name: Matthias Blume
3122    Date: 2002/03/13 14:00:00 EST
3123    Tag: blume-20020313-overload-etc
3124    Description:
3125    
3126    1. Added _overload as a synonym for overload for backward compatibility.
3127       (Control.overloadKW must be true for either version to be accepted.)
3128    
3129    2. Fixed bug in install script that caused more things to be installed
3130       than what was requested in config/targets.
3131    
3132    3. Made CM aware of the (_)overload construct so that autoloading
3133       works.
3134    
3135    ----------------------------------------------------------------------
3136    Name: Matthias Blume
3137    Date: 2002/03/12 22:03:00 EST
3138    Tag: blume-20020312-url
3139    Description:
3140    
3141    Forgot to update BOOT and srcarchiveurl.
3142    
3143    ----------------------------------------------------------------------
3144    Name: Matthias Blume
3145    Date: 2002/03/12 17:30:00 EST
3146    Tag: blume-20020312-version110392
3147    Description:
3148    
3149    Yet another version number bump (because of small changes to the
3150    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3151    
3152    Changes:
3153    
3154      The new pid generation scheme described a few weeks ago was overly
3155      complicated.  I implemented a new mechanism that is simpler and
3156      provides a bit more "stability":  Once CM has seen a compilation
3157      unit, it keeps its identity constant (as long as you do not delete
3158      those crucial CM/GUID/* files).  This means that when you change
3159      an interface, compile, then go back to the old interface, and
3160      compile again, you arrive at the original pid.
3161    
3162      There now also is a mechanism that instructs CM to use the plain
3163      environment hash as a module's pid (effectively making its GUID
3164      the empty string).  For this, "noguid" must be specified as an
3165      option to the .sml file in question within its .cm file.
3166      This is most useful for code that is being generated by tools such
3167      as ml-nlffigen (because during development programmers tend to
3168      erase the tool's entire output directory tree including CM's cached
3169      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3170      revert to the old, broken behavior of SML/NJ, but in specific cases
3171      where there is no danger of interface confusion, its use is ok
3172      (I think).
3173    
3174      ml-nlffigen by default generates "noguid" annotations.  They can be
3175      turned off by specifying -guid in its command line.
3176    
3177    ----------------------------------------------------------------------
3178    Name: Lal George
3179    Date: 2002/03/12 12 14:42:36 EST
3180    Tag: george-20020312-frequency-computation
3181    Description:
3182    
3183    Integrated jump chaining and static block frequency into the
3184    compiler. More details and numbers later.
3185    
3186    ----------------------------------------------------------------------
3187    Name: Lal George
3188    Date: 2002/03/11 11 22:38:53 EST
3189    Tag: george-20020311-jump-chain-elim
3190    Description:
3191    
3192    Tested the jump chain elimination on all architectures (except the
3193    hppa).  This is on by default right now and is profitable for the
3194    alpha and x86, however, it may not be profitable for the sparc and ppc
3195    when compiling the compiler.
3196    
3197    The gc test will typically jump to a label at the end of the cluster,
3198    where there is another jump to an external cluster containing the actual
3199    code to invoke gc. This is to allow factoring of common gc invocation
3200    sequences. That is to say, we generate:
3201    
3202            f:
3203               testgc
3204               ja   L1      % jump if above to L1
3205    
3206            L1:
3207               jmp L2
3208    
3209    
3210    After jump chain elimination the 'ja L1' instructions is converted to
3211    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3212    up being implemented in their long form (if L2 is far away) using:
3213    
3214            jbe     L3      % jump if below or equal to L3
3215            jmp     L2
3216         L3:
3217            ...
3218    
3219    
3220    For large compilation units L2  may be far away.
3221    
3222    
3223    ----------------------------------------------------------------------
3224    Name: Matthias Blume
3225    Date: 2002/03/11 13:30:00 EST
3226    Tag: blume-20020311-mltreeeval
3227    Description:
3228    
3229    A functor parameter was missing.
3230    
3231    ----------------------------------------------------------------------
3232    Name: Allen Leung
3233    Date: 2002/03/11 10:30:00 EST
3234    Tag: leunga-20020311-runtime-string0
3235    Description:
3236    
3237       The representation of the empty string now points to a
3238    legal null terminated C string instead of unit.  It is now possible
3239    to convert an ML string into C string with InlineT.CharVector.getData.
3240    This compiles into one single machine instruction.
3241    
3242    ----------------------------------------------------------------------
3243    Name: Allen Leung
3244    Date: 2002/03/10 23:55:00 EST
3245    Tag: leunga-20020310-x86-call
3246    Description:
3247    
3248       Added machine generation for CALL instruction (relative displacement mode)
3249    
3250    ----------------------------------------------------------------------
3251    Name: Matthias Blume
3252    Date: 2002/03/08 16:05:00
3253    Tag: blume-20020308-entrypoints
3254    Description:
3255    
3256    Version number bumped to 110.39.1.  NEW BOOTFILES!
3257    
3258    Entrypoints: non-zero offset into a code object where execution should begin.
3259    
3260    - Added the notion of an entrypoint to CodeObj.
3261    - Added reading/writing of entrypoint info to Binfile.
3262    - Made runtime system bootloader aware of entrypoints.
3263    - Use the address of the label of the first function given to mlriscGen
3264      as the entrypoint.  This address is currently always 0, but it will
3265      not be 0 once we turn on block placement.
3266    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3267      with entry points) from mlriscGen.
3268    
3269    ----------------------------------------------------------------------
3270    Name: Allen Leung
3271    Date: 2002/03/07 20:45:00 EST
3272    Tag: leunga-20020307-x86-cmov
3273    Description:
3274    
3275       Bug fixes for CMOVcc on x86.
3276    
3277       1. Added machine code generation for CMOVcc
3278       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3279       3. CMOVcc cannot have an immediate operand as argument.
3280    
3281    ----------------------------------------------------------------------
3282    Name: Matthias Blume
3283    Date: 2002/03/07 16:15:00 EST
3284    Tag: blume-20020307-controls
3285    Description:
3286    
3287    This is a very large but mostly boring patch which makes (almost)
3288    every tuneable compiler knob (i.e., pretty much everything under
3289    Control.* plus a few other things) configurable via both the command
3290    line and environment variables in the style CM did its configuration
3291    until now.
3292    
3293    Try starting sml with '-h' (or, if you are brave, '-H')
3294    
3295    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3296    implements the underlying generic mechanism.
3297    
3298    The interface to some of the existing such facilities has changed somewhat.
3299    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3300    (The getFoo interface is still there for backward-compatibility, but its
3301    use is deprecated.)
3302    
3303    The ml-build script passes -Cxxx=yyy command-line arguments through so
3304    that one can now twiddle the compiler settings when using this "batch"
3305    compiler.
3306    
3307    TODO items:
3308    
3309    We should go through and throw out all controls that are no longer
3310    connected to anything.  Moreover, we should go through and provide
3311    meaningful (and correct!) documentation strings for those controls
3312    that still are connected.
3313    
3314    Currently, multiple calls to Controls.new are accepted (only the first
3315    has any effect).  Eventually we should make sure that every control
3316    is being made (via Controls.new) exactly once.  Future access can then
3317    be done using Controls.acc.
3318    
3319    Finally, it would probably be a good idea to use the getter-setter
3320    interface to controls rather than ref cells.  For the time being, both
3321    styles are provided by the Controls module, but getter-setter pairs are
3322    better if thread-safety is of any concern because they can be wrapped.
3323    
3324    *****************************************
3325    
3326    One bug fix: The function blockPlacement in three of the MLRISC
3327    backpatch files used to be hard-wired to one of two possibilities at
3328    link time (according to the value of the placementFlag).  But (I
3329    think) it should rather sense the flag every time.
3330    
3331    *****************************************
3332    
3333    Other assorted changes (by other people who did not supply a HISTORY entry):
3334    
3335    1. the cross-module inliner now works much better (Monnier)
3336    2. representation of weights, frequencies, and probabilities in MLRISC
3337       changed in preparation of using those for weighted block placement
3338       (Reppy, George)
3339    
3340    ----------------------------------------------------------------------
3341    Name: Lal George
3342    Date: 2002/03/07 14:44:24 EST 2002
3343    Tag: george-20020307-weighted-block-placement
3344    
3345    Tested the weighted block placement optimization on all architectures
3346    (except the hppa) using AMPL to generate the block and edge frequencies.
3347    Changes were required in the machine properties to correctly
3348    categorize trap instructions. There is an MLRISC flag
3349    "weighted-block-placement" that can be used to enable weighted block
3350    placement, but this will be ineffective without block/edge
3351    frequencies (coming soon).
3352    
3353    
3354    ----------------------------------------------------------------------
3355    Name: Lal George
3356    Date: 2002/03/05 17:24:48 EST
3357    Tag: george-20020305-linkage-cluster
3358    
3359    In order to support the block placement optimization, a new cluster
3360    is generated as the very first cluster (called the linkage cluster).
3361    It contains a single jump to the 'real' entry point for the compilation
3362    unit. Block placement has no effect on the linkage cluster itself, but
3363    all the other clusters  have full freedom in the manner in which they
3364    reorder blocks or functions.
3365    
3366    On the x86 the typical linkage code that is generated is:
3367       ----------------------
3368            .align 2
3369       L0:
3370            addl    $L1-L0, 72(%esp)
3371            jmp     L1
3372    
3373    
3374            .align  2
3375       L1:
3376       ----------------------
3377    
3378    72(%esp) is the memory location for the stdlink register. This
3379    must contain the address of the CPS function being called. In the
3380    above example, it contains the address of  L0; before
3381    calling L1 (the real entry point for the compilation unit), it
3382    must contain the address for L1, and hence
3383    
3384            addl $L1-L0, 72(%esp)
3385    
3386    I have tested this on all architectures except the hppa.The increase
3387    in code size is of course negligible
3388    
3389    ----------------------------------------------------------------------
3390    Name: Allen Leung
3391    Date: 2002/03/03 13:20:00 EST
3392    Tag: leunga-20020303-mlrisc-tools
3393    
3394      Added #[ ... ] expressions to mlrisc tools
3395    
3396    ----------------------------------------------------------------------
3397    Name: Matthias Blume
3398    Date: 2002/02/27 12:29:00 EST
3399    Tag: blume-20020227-cdebug
3400    Description:
3401    
3402    - made types in structure C and C_Debug to be equal
3403    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3404    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3405    
3406    ----------------------------------------------------------------------
3407    Name: Matthias Blume
3408    Date: 2002/02/26 12:00:00 EST
3409    Tag: blume-20020226-ffi
3410    Description:
3411    
3412    1. Fixed a minor bug in CM's "noweb" tool:
3413       If numbering is turned off, then truly don't number (i.e., do not
3414       supply the -L option to noweb).  The previous behavior was to supply
3415       -L'' -- which caused noweb to use the "default" line numbering scheme.
3416       Thanks to Chris Richards for pointing this out (and supplying the fix).
3417    
3418    2. Once again, I reworked some aspects of the FFI:
3419    
3420       A. The incomplete/complete type business:
3421    
3422       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3423         gone!
3424       - ML types representing an incomplete type are now *equal* to
3425         ML types representing their corresponding complete types (just like
3426         in C).  This is still safe because ml-nlffigen will not generate
3427         RTTI for incomplete types, nor will it generate functions that
3428         require access to such RTTI.   But when ML code generated from both
3429         incomplete and complete versions of the C type meet, the ML types
3430         are trivially interoperable.
3431    
3432         NOTE:  These changes restore the full generality of the translation
3433         (which was previously lost when I eliminated functorization)!
3434    
3435       B. Enum types:
3436    
3437       - Structure C now has a type constructor "enum" that is similar to
3438         how the "su" constructor works.  However, "enum" is not a phantom
3439         type because each "T enum" has values (and is isomorphic to
3440         MLRep.Signed.int).
3441       - There are generic access operations for enum objects (using
3442         MLRep.Signed.int).
3443       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3444         * The structure contains the definition of type "mlrep" (the ML-side
3445         representation type of the enum).  Normally, mlrep is the same
3446         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3447         then mlrep will be defined as a datatype -- thus facilitating
3448         pattern matching on mlrep values.
3449         ("-ec" will be suppressed if there are duplicate values in an
3450          enumeration.)
3451         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3452         will be generated for each C enum constant xxx.
3453         * Conversion functions m2i and i2m convert between mlrep and
3454         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3455         * Coversion functions c and ml convert between mlrep and "tag enum".
3456         * Access functions (get/set) fetch and store mlrep values.
3457       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3458         enumerations are merged into one single enumeration represented by
3459         structure E_'.
3460    
3461    ----------------------------------------------------------------------
3462    Name: Allen Leung
3463    Date: 2002/02/25 04:45:00 EST
3464    Tag: leunga-20020225-cps-spill
3465    
3466    This is a new implementation of the CPS spill phase.
3467    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3468    In case of problems, replace it with the old file spill.sml
3469    
3470    The current compiler runs into some serious performance problems when
3471    constructing a large record.  This can happen when we try to compile a
3472    structure with many items.  Even a very simple structure like the following
3473    makes the compiler slow down.
3474    
3475        structure Foo = struct
3476           val x_1 = 0w1 : Word32.int
3477           val x_2 = 0w2 : Word32.int
3478           val x_3 = 0w3 : Word32.int
3479           ...
3480           val x_N = 0wN : Word32.int
3481        end
3482    
3483    The following table shows the compile time, from N=1000 to N=4000,
3484    with the old compiler:
3485    
3486    N
3487    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3488           MLRISC ra                               0.06u  0.00s  0.05g
3489              (spills = 0 reloads = 0)
3490           TOTAL                                   0.63u  0.07s  0.21g
3491    
3492    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3493           MLRISC ra                               5.68u  0.59s  3.93g
3494              (spills = 0 reloads = 0)
3495           TOTAL                                   14.71u  0.99s  4.81g
3496    
3497    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3498           MLRISC ra                               5.54u  0.65s  3.91g
3499              (spills = 543 reloads = 1082)
3500           TOTAL                                   65.40u  3.13s  6.00g
3501    
3502    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3503           MLRISC ra                               0.80u  0.10s  0.55g
3504              (spills = 42 reloads = 84)
3505           TOTAL                                   129.42u  5.10s  4.13g
3506    
3507    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3508           MLRISC ra                               2.69u  0.27s  1.38g
3509              (spills = 62 reloads = 124)
3510           TOTAL                                   682.48u  19.61s  13.99g
3511    
3512    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3513           MLRISC ra                               4.96u  0.27s  2.72g
3514              (spills = 85 reloads = 170)
3515           TOTAL                                   2375.26u  57.21s  48.00g
3516    
3517    As you can see the old cps spill module suffers from some serious
3518    performance problem.  But since I cannot decipher the old code fully,
3519    instead of patching the problems up, I'm reimplementing it
3520    with a different algorithm.  The new code is more modular,
3521    smaller when compiled, and substantially faster
3522    (O(n log n) time and O(n) space).  Timing of the new spill module:
3523    
3524    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3525          MLRISC ra                               0.25u  0.02s  0.15g
3526             (spills=1 reloads=3)
3527          TOTAL                                   7.74u  0.34s  1.62g
3528    
3529    Implementation details:
3530    
3531    As far as I can tell, the purpose of the CPS spill module is to make sure the
3532    number of live variables at any program point (the bandwidth)
3533    does not exceed a certain limit, which is determined by the
3534    size of the spill area.
3535    
3536    When the bandwidth is too large, we decrease the register pressure by
3537    packing live variables into spill records.  How we achieve this is
3538    completely different than what we did in the old code.
3539    
3540    First, there is something about the MLRiscGen code generator
3541    that we should be aware of:
3542    
3543    o MLRiscGen performs code motion!
3544    
3545       In particular, it will move floating point computations and
3546       address computations involving only the heap pointer to
3547       their use sites (if there is only a single use).
3548       What this means is that if we have a CPS record construction
3549       statement
3550    
3551           RECORD(k,vl,w,e)
3552    
3553       we should never count the new record address w as live if w
3554       has only one use (which is often the case).
3555    
3556       We should do something similar to floating point, but the transformation
3557       there is much more complex, so I won't deal with that.
3558    
3559    Secondly, there are now two new cps primops at our disposal:
3560    
3561     1. rawrecord of record_kind option
3562        This pure operator allocates some uninitialized storage from the heap.
3563        There are two forms:
3564    
3565         rawrecord NONE [INT n]  allocates a tagless record of length n
3566         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3567                                     and initializes the tag.
3568    
3569     2. rawupdate of cty
3570          rawupdate cty (v,i,x)
3571          Assigns to x to the ith component of record v.
3572          The storelist is not updated.
3573    
3574    We use these new primops for both spilling and increment record construction.
3575    
3576     1. Spilling.
3577    
3578        This is implemented with a linear scan algorithm (but generalized
3579        to trees).  The algorithm will create a single spill record at the
3580        beginning of the cps function and use rawupdate to spill to it,
3581        and SELECT or SELp to reload from it.  So both spills and reloads
3582        are fine-grain operations.  In contrast, in the old algorithm
3583        "spills" have to be bundled together in records.
3584    
3585        Ideally, we should sink the spill record construction to where
3586        it is needed.  We can even split the spill record into multiple ones
3587        at the places where they are needed.  But CPS is not a good
3588        representation for global code motion, so I'll keep it simple and
3589        am not attempting this.
3590    
3591     2. Incremental record construction (aka record splitting).
3592    
3593        Long records with many component values which are simulatenously live
3594        (recall that single use record addresses are not considered to
3595         be live) are constructed with rawrecord and rawupdate.
3596        We allocate space on the heap with rawrecord first, then gradually
3597        fill it in with rawupdate.  This is the technique suggested to me
3598        by Matthias.
3599    
3600        Some restrictions on when this is applicable:
3601        1. It is not a VECTOR record.  The code generator currently does not handle
3602           this case. VECTOR record uses double indirection like arrays.
3603        2. All the record component values are defined in the same "basic block"
3604           as the record constructor.  This is to prevent speculative
3605           record construction.
3606    
3607    ----------------------------------------------------------------------
3608    Name: Allen Leung
3609    Date: 2002/02/22 01:02:00 EST
3610    Tag: leunga-20020222-mlrisc-tools
3611    
3612    Minor bug fixes in the parser and rewriter
3613    
3614    ----------------------------------------------------------------------
3615    Name: Allen Leung
3616    Date: 2002/02/21 20:20:00 EST
3617    Tag: leunga-20020221-peephole
3618    
3619    Regenerated the peephole files.  Some contained typos in the specification
3620    and some didn't compile because of pretty printing bugs in the old version
3621    of 'nowhere'.
3622    
3623    ----------------------------------------------------------------------
3624    Name: Allen Leung
3625    Date: 2002/02/19 20:20:00 EST
3626    Tag: leunga-20020219-mlrisc-tools
3627    Description:
3628    
3629       Minor bug fixes to the mlrisc-tools library: