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

SCM Repository

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

Diff of /sml/trunk/HISTORY

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

revision 993, Fri Nov 23 02:37:10 2001 UTC revision 1861, Mon Nov 7 18:17:09 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: John Reppy (http://www.cs.uchicago.edu/~jhr)
17    Date: 2005/11/07
18    Description:
19    Fixed a Basis incompatibility: The depreciated function Substring.all
20    was removed (use Substring.full instead).
21    
22    ----------------------------------------------------------------------
23    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
24    Date: 2005/11/05
25    Description:
26    Tweaked interval set API in SML/NJ library; see the CHANGES file for details.
27    
28    ----------------------------------------------------------------------
29    Name: Matthias Blume (http://www.tti-c.org/blume/)
30    Date: 2005/11/02 16:25:00 CST
31    Tag: blume-20051102-hex
32    Description:
33    
34    * Runtime system bootstrap code now accepts hex digits in BOOTLIST
35      in either upper case or lower case format.
36    * Pushed changes to names of Pack<N>{Big,Little} structures through
37      CML and eXene.
38    
39    ----------------------------------------------------------------------
40    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
41    Date: 2005/11/02
42    Description:
43    Fixed a Basis incompatibility: Pack<N>{Big,Little} structures should
44    be called PackWord<N>{Big,Little}.
45    
46    ----------------------------------------------------------------------
47    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
48    Date: 2005/10/28
49    Description:
50    Fixed a minor Basis incompatibility: hex digits should be upper case.
51    
52    ----------------------------------------------------------------------
53    Name: Matthias Blume (blume (at) tti - c (dot) org)
54    Date: 2005/10/25 16:00:00 CDT
55    Tag: Release_110_56
56    Description:
57    
58    New working version (110.56).  NEW BOOTFILES!
59    
60    ----------------------------------------------------------------------
61    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
62    Date: 2005/10/25
63    Description:
64    Added interval sets to utility library (signatures INTERVAL_DOMAIN
65    and INTERVAL_SET, and functor IntervalSetFn).
66    
67    ----------------------------------------------------------------------
68    Name: John Reppy (http://www.cs.uchicago.edu/~jhr)
69    Date: 2005/10/14
70    Description:
71    Add Zhong Shao's fix for datatype equality functions.
72    
73    ----------------------------------------------------------------------
74    Name: Allen Leung (leunga (at) reservoir (dot) com)
75    Date: 2005/10/14 12:30:00 EST
76    Tag: leunga-20051014-x86-spill
77    Description:
78    
79    Bug fix a bug found by Carl Hauser.
80    There was a typo in the reload code for FCMP in x86SpillInstr.sml.
81    
82    ----------------------------------------------------------------------
83    Name: Allen Leung (leunga (at) reservoir (dot) com)
84    Date: 2005/10/14 11:15:00 EST
85    Tag: leunga-20051014-x86-asm
86    Description:
87    
88    Removed some debugging code in file x86Asm.sml.
89    The function emit_operand was printing out debugging output.
90    
91    ----------------------------------------------------------------------
92    Name: John Reppy
93    Date: 2005/07/27
94    Tag:
95    Description:
96    
97    Fixed ml-lex to recognize "\r" as representing carriage return.
98    
99    ----------------------------------------------------------------------
100    Name: John Reppy
101    Date: 2005/07/27
102    Tag:
103    Description:
104    
105    Fixed ml-yacc to work on files with non-native end-of-line
106    encodings (e.g., Windows text file on a Unix system).
107    
108    ----------------------------------------------------------------------
109    Name: John Reppy
110    Date: 2005/07/20
111    Tag:
112    Description:
113    
114    Added changes from Dominic Evans (oldmanuk (at) gmail (dot) com)
115    to support HPUX 11.
116    
117    ----------------------------------------------------------------------
118    Name: John Reppy
119    Date: 2005/07/06
120    Tag:
121    Description:
122    
123    Changes to the SML/NJ library.  See smlnj-lib/CHANGES for details.
124    
125    ----------------------------------------------------------------------
126    Name: Matthias Blume (blume (at) tti - c (dot) org)
127    Date: 2005/07/06 00:45:00 CDT
128    Tag: blume-20050706-slice-copy
129    Description:
130    
131    Fixed reversed logic for deciding whether to "copy up" or "copy down"
132    in *-array-slice.sml.
133    
134    ----------------------------------------------------------------------
135    Name: Allen Leung (leunga (at) reservoir (dot) com)
136    Date: 2005/05/31 17:00:00 EST
137    Tag: leunga-20050531-cygwin-fault-2
138    Description:
139    
140    A typo in the cygwin code fixed.
141    
142    ----------------------------------------------------------------------
143    Name: Allen Leung (leunga (at) reservoir (dot) com)
144    Date: 2005/05/31 16:47:00 EST
145    Tag: leunga-20050531-cygwin-fault
146    Description:
147    
148    Updated Cygwin's fault/signal handling to match the Windows version.
149    Updated the export list.
150    
151    ----------------------------------------------------------------------
152    Name: Matthias Blume (blume (at) tti - c (dot) org)
153    Date: 2005/05/18 13:35:00 CDT
154    Tag: Release_110_54
155    Description:
156    
157    New working version (110.54).  NEW BOOTFILES!
158    
159    ----------------------------------------------------------------------
160    Name: Matthias Blume (blume (at) tti - c (dot) org)
161    Date: 2005/05/18 11:58:00 CDT
162    Tag: blume-20050518-installer
163    Description:
164    
165    Added support scripts for Mac OS X PackageMaker and modified
166    config/install.sh so that it supports re-dumping a heap image
167    after customization.
168    
169    ----------------------------------------------------------------------
170    Name: Matthias Blume (blume (at) tti - c (dot) org)
171    Date: 2005/05/18 10:55:00 CDT
172    Tag: blume-20050518-realdiv-noovld
173    Description:
174    
175    Un-overloaded / to work around bug in overloading resolution code.
176    
177    ----------------------------------------------------------------------
178    Name: Matthias Blume (blume (at) tti - c (dot) org)
179    Date: 2005/05/16 23:50:00 CDT
180    Tag: blume-20050516-redump-heap
181    Description:
182    
183    Added mechanism for re-creating a heap file for the interactive system
184    after configuration variables have been changed.
185    
186       CM.redump_heap : string -> unit
187    
188    This is much like SMLofNJ.exportML, but starting from the resulting
189    heap does not return to the caller of CM.redump_heap but
190    restarts the interactive system from scratch.  The original call of
191    CM.redump_heap does not return but ends the interactive session.
192    Thus, CM.redump_heap is a lot like SMLofNJ.exportFn.
193    
194    Internally, redump_heap winds the dynamic execution context back to
195    the point where the original heap image was created and re-executes
196    the heap image generation code in the boot code.
197    
198    ----------------------------------------------------------------------
199    Name: Matthias Blume (blume (at) tti - c (dot) org)
200    Date: 2005/05/09 21:55:00 CDT
201    Tag: blume-20050509-word64
202    Description:
203    
204    Added a hack to the existing hack known as Word64 to make fromString
205    behave correctly.  I am still not sure whether Word64.scan will work
206    as specified with respect to the interaction of radix and prefix.
207    
208    ----------------------------------------------------------------------
209    Name: Allen Leung (leunga (at) reservoir (dot) com)
210    Date: 2005/05/04 11:50:00 EST
211    Tag: leunga-20050504-checkgc
212    Description:
213    
214    Added a gc protocol checking phase.  This phase is enabled with
215    the flag "check-gc".  "debug-check-gc" turns on the verbose mode.
216    
217    ----------------------------------------------------------------------
218    Name: Matthias Blume (blume (at) tti - c (dot) org)
219    Date: 2005/05/04 10:45:00 CDT
220    Tag: blume-20050504-intinf
221    Description:
222    
223    Fixed a bug in the implementation of div and mod for IntInf.
224    Thanks to Neophytos Michael for reporting the problem.
225    
226    ----------------------------------------------------------------------
227    Name: Matthias Blume (blume (at) tti - c (dot) org)
228    Date: 2005/05/04 10:35:00 CDT
229    Tag: blume-20050504-join
230    Description:
231    
232    Added a "join" combinator to the ParserComb module in smlnj-lib.cm.
233    
234    ----------------------------------------------------------------------
235    Name: Matthias Blume (blume (at) tti - c (dot) org)
236    Date: 2005/02/28 23:40:00 CST
237    Tag: blume-20050228-mVar
238    Description:
239    
240    Fixed serious bug (brown paper bag variety) in new implementation of
241    structure Atom in CML.  (I had accidentally used a mailbox instead of
242    an mvar, leaving the door open for races.)
243    
244    ----------------------------------------------------------------------
245    Name: Matthias Blume (blume (at) tti - c (dot) org)
246    Date: 2005/02/25 15:00:00 CST
247    Tag: Release_110_53
248    Description:
249    
250    New working version (110.53).  NEW BOOTFILES!
251    
252    ----------------------------------------------------------------------
253    Name: Matthias Blume (blume (at) tti - c (dot) org)
254    Date: 2005/02/25 14:50:00 CST
255    Tag: blume-20050225-susp
256    Description:
257    
258    Brought back SMLofNJ.Susp.  The underlying suspension type is the one
259    implemented in Core -- which means that it is the same as the one used
260    by the lazy extension.
261    
262    ----------------------------------------------------------------------
263    Name: Matthias Blume (blume (at) tti - c (dot) org)
264    Date: 2005/02/24 16:50:00 CST
265    Tag: blume-20050224-cml-atom
266    Description:
267    
268    Simpler and at the same time more general implementation of structure
269    Atom in CML.
270    
271    ----------------------------------------------------------------------
272    Name: Matthias Blume (blume (at) tti - c (dot) org)
273    Date: 2005/02/15 17:35:00 CST
274    Tag: blume-20050215-tools
275    Description:
276    
277    Created new "tools" directory under "src" and moved "TraceDebugProf"
278    there.
279    
280    ----------------------------------------------------------------------
281    Name: Matthias Blume (blume (at) tti - c (dot) org)
282    Date: 2005/02/10 17:55:00 CST
283    Tag: blume-20050210-longlong
284    Description:
285    
286    Implemented "long long" arguments and results for NLFFI.  (Only the
287    PPC/MacOS implementation is complete, the other backends still need to
288    be updated.)
289    
290    ----------------------------------------------------------------------
291    Name: Matthias Blume (blume (at) tti - c (dot) org)
292    Date: 2005/01/24 17:40:00 CST
293    Tag: blume-20050124-mlyacc
294    Description:
295    
296    Minor cleanup in ML-Yacc rule printing mechanism.  This should fix a
297    problem with certain "as" patterns which previously got rendered
298    using incorrect syntax.
299    
300    ----------------------------------------------------------------------
301    Name: Matthias Blume (blume (at) tti - c (dot) org)
302    Date: 2005/01/18 12:00:00 CST
303    Tag: blume-20050118-profile
304    Description:
305    
306    Made time profiling code (interrupt handler) in runtime system aware
307    of new array representation.
308    
309    ----------------------------------------------------------------------
310    Name: Matthias Blume (blume (at) tti - c (dot) org)
311    Date: 2005/01/14 18:00:00 CST
312    Tag: blume-20050114-heap2exec
313    Description:
314    
315    Implemented new (but still experimental) heap2exec facility.  This is
316    tested under Mac OS X and should work under Linux (will test shortly).
317    It will probably also work on the Sparc (will test some time later).
318    
319      - removed old "HACKED_STANDALONE" hack from runtime
320    
321    To be able to test this, uncomment the request for "heap2asm" in
322    config/targets prior to installation.  (Notice that this is different
323    from "heap2exec" mentioned below.  Not a typo.)
324    
325    To perform an actual test, run the command
326    
327       $ bin/heap2exec heapfile execfile
328    
329    (You can put heap2exec on your shell's path.)
330    
331    For example, run
332    
333      $ bin/heap2exec bin/.heap/ml-yacc.ppc-darwin mly
334    
335    This will create a standalone executable called "mly" which you can
336    then invoke directly as a command.
337    
338    ----------------------------------------------------------------------
339    Name: Matthias Blume (blume (at) tti - c (dot) org)
340    Date: 2005/01/07 11:44:00 CST
341    Tag: blume-20050107-mlstring
342    Description:
343    
344    fixed off-by-one error in ML_STRING macro (globals.c)
345    
346    ----------------------------------------------------------------------
347    Name: Matthias Blume (blume (at) tti - c (dot) org)
348    Date: 2004/12/23 18:00:00 CST
349    Tag: blume-20041223-santa
350    Description:
351    
352    Made ml-build script "smarter" (but only very little).
353    
354    ----------------------------------------------------------------------
355    Name: Matthias Blume (blume (at) tti - c (dot) org)
356    Date: 2004/12/21 15:05:00 CST
357    Tag: blume-20041221-longlong
358    Description:
359    
360    * Implemented access to signed and unsigned long long data in NLFFI.
361    (The parameter-passing part of the picture has not complete. But
362    data structure access seems to work.)
363    
364    * Fixed CM's incorrect assumption that the PPC is little-endian.
365      (On the Mac, it is big-endian.  And that's currently our only
366      PPC platform.)
367    
368    ----------------------------------------------------------------------
369    Name: Matthias Blume (blume (at) tti - c (dot) org)
370    Date: 2004/12/21 12:50:00 CST
371    Tag: blume-20041221-memory
372    Description:
373    
374    Some cleanup in the $c/memory.cm library: separated some concerns by
375    moving allocation code and memory access code each into their own
376    files.
377    
378    ----------------------------------------------------------------------
379    Name: Allen Leung (leunga (at) reservoir (dot) com)
380    Date: 2004/12/17 16:12:00 EST
381    Tag: leunga-20041217-cygwin-smlnj-home
382    Description:
383    
384        The Unix I/O library of SML/NJ on cygwin does not understand
385    Windows style pathname, so problems arise when SMLNJ_HOME is set
386    to a Windows style pathname.  _run-sml now converts SMLNJ_HOME
387    to a POSIX pathname on cygwin.
388    
389    ----------------------------------------------------------------------
390    Name: Matthias Blume (blume (at) tti - c (dot) org)
391    Date: 2004/12/16 13:35:00 CST
392    Tag: Release_110_52
393    Description:
394    
395    Last-minute changes incorporated into 110.52.  Release tag moved.
396    
397    The changes:
398       - HashString.hashString' -> HashString.hashSubstring
399       - bug fix in UnivariateStats
400    
401    ----------------------------------------------------------------------
402    Name: Matthias Blume (blume (at) tti - c (dot) org)
403    Date: 2004/12/15 23:40:00 CST
404    Tag: blume-20041215-hashSubstring
405    Description:
406    
407    - HashString.hashString' -> HashString.hashSubstring
408    - corresponding changes in atom.sml
409    - "de-compressed" (aka. un-obfuscated) code for UnivariateStats and
410      added some comments
411    
412    ----------------------------------------------------------------------
413    Name: Matthias Blume (blume (at) tti - c (dot) org)
414    Date: 2004/12/15 15:30:00 CST
415    Tag: (Release_110_52)
416    Description:
417    
418    New working version (110.52).  NEW BOOTFILES!
419    
420    ----------------------------------------------------------------------
421    Name: Matthias Blume (blume (at) tti - c (dot) org)
422    Date: 2004/12/15 12:45:00 CST
423    Tag: blume-20041215-spaces
424    Description:
425    
426    More on the space problem (this time for Win32).
427    
428    ----------------------------------------------------------------------
429    Name: Matthias Blume (blume (at) tti - c (dot) org)
430    Date: 2004/12/14 17:30:00 CST
431    Tag: blume-20041214-spaces
432    Description:
433    
434    Hacked some of the scripts (in particular: the installer) to cope with
435    spaces in filenames a bit better.  But beware: the current "solution"
436    is likely still full of bugs and inherently incomplete.  (We need to
437    do away with those shell scripts for a comprehensive solution.)
438    
439    ----------------------------------------------------------------------
440    Name: Matthias Blume (blume (at) tti - c (dot) org)
441    Date: 2004/12/13 14:45:00 CST
442    Tag: blume-20041213-ml-makedepend
443    Description:
444    
445    Fixed bug in code for ml-makedepend.
446    
447    ----------------------------------------------------------------------
448    Name: Matthias Blume (blume (at) tti - c (dot) org)
449    Date: 2004/12/09 16:30:00 CST
450    Tag: blume-20041209-statistics
451    Description:
452    
453    Added two simple but potentially useful statistics modules to SML/NJ Library.
454    (See CHANGES file there.)
455    
456    ----------------------------------------------------------------------
457    Name: Matthias Blume (blume (at) tti - c (dot) org)
458    Date: 2004/12/01 16:50:00 CST
459    Tag: blume-20041201-atom
460    Description:
461    
462    smlnj-lib:
463    
464      Added function HashString.hashString' for substrings.
465      Hand-inlined CharVector.foldl into HashString (for speed).
466      Modified implementation of structure Atom to avoid extracting
467      strings from substrings unless necessary.
468      (Also see CHANGES file for smlnj-lib.)
469    
470    ----------------------------------------------------------------------
471    Name: Matthias Blume (blume (at) tti - c (dot) org)
472    Date: 2004/11/24 22:15:00 CST
473    Tag: blume-20041124-cml
474    Description:
475    
476    Made sure CML compiles when Position = Int64.
477    
478    ----------------------------------------------------------------------
479    Name: Matthias Blume (blume (at) tti - c (dot) org)
480    Date: 2004/11/24 14:30:00 CST
481    Tag: blume-20041124-position
482    Description:
483    
484    The compiler can now be compiled in a mode that makes structure Position
485    equal to Int64.  The default, however, is unchanged (Position = Int31)
486    for the time being.
487    
488    To enable 64-bit positions, use the following procedure:
489    
490    1. Start sml
491    2. Autoload $smlnj/cmb.cm (if not already autoloaded)
492    3. Type
493          #set (CMB.symval "USE_64_BIT_POSITIONS") (SOME 1);
494    4. Run CMB.make() as usual.
495    
496    This is barely tested.  The only test so far was a little SML program
497    counting the number of characters in an 8-gigabyte file by
498    reading it character-by-character.  That test was successful.
499    
500    In support of 64-bit positions, a number of new functions have been
501    added to the runtime system.
502    
503    ----------------------------------------------------------------------
504    Name: Matthias Blume (blume (at) tti - c (dot) org)
505    Date: 2004/11/23 14:45:00 CST
506    Tag: blume-20041123-useFile
507    Description:
508    
509    Fixed a problem with unhelpful error messages related to problems with
510    .cm- or .sml files that appear as part of the sml command line.
511    
512    ----------------------------------------------------------------------
513    Name: Matthias Blume (blume (at) tti - c (dot) org)
514    Date: 2004/11/18 15:40:00 CST
515    Tag: Release_110_51
516    Description:
517    
518    New working version (110.51).  NEW BOOTFILES!
519    
520    ----------------------------------------------------------------------
521    Name: Matthias Blume (blume (at) tti - c (dot) org)
522    Date: 2004/11/18 15:35:00 CST
523    Tag: <none>
524    Description:
525    
526    Enabled dlopen and friends for FreeBSD (as recommended by Johannes 5
527    Joemann).
528    
529    ----------------------------------------------------------------------
530    Name: Allen Leung (leunga (at) reservoir (dot) com)
531    Date: 2004/11/17 16:05:21 EST 2004
532    Tag: leunga-20041117-mlrisc-live-kill
533    Description:
534    
535       Added support for MLTree constructs LIVE and KILL
536    to all the architectures.
537    
538    ----------------------------------------------------------------------
539    Name: Matthias Blume (blume (at) tti - c (dot) org)
540    Date: 2004/11/13 00:20:00 CST
541    Tag: blume-20041113-versiontool
542    Description:
543    
544    - Stripped down the versiontool: It now only handles the version number.
545      The date string is generated at bootstrap time (during makeml).
546    
547    - In a previous commit, fixed a minor issue with how polyequal is being
548      translated.  In particular, the code now "looks through" abstractions.
549      This results in slightly fewer polyEqual warnings and hopefully slightly
550      more efficient code.  Important examples for where this matters are
551      the new int64 and word64 types.
552    
553    ----------------------------------------------------------------------
554    Name: Matthias Blume (blume (at) tti - c (dot) org)
555    Date: 2004/11/12 00:30:00 CST
556    Tag: blume-20041112-int64
557    Description:
558    
559    Structure Int64 fully hooked in.  (The implementation is not very
560    efficient, though.)
561    
562    ----------------------------------------------------------------------
563    Name: Matthias Blume (blume (at) tti - c (dot) org)
564    Date: 2004/11/11 17:30:00 CST
565    Tag: blume-20041111-more64
566    Description:
567    
568    All the pieces of Word64 are now there, with the exception of the
569    conversions from and to LargeWord. (Eventually these need to be identities,
570    but for the time being they don't even make sense because LargeWord is
571    32-bit wide.)
572    
573    Also started to add similar support for Int64, but major pieces of that
574    are still missing.
575    
576    ----------------------------------------------------------------------
577    Name: Matthias Blume (blume (at) tti - c (dot) org)
578    Date: 2004/11/11 00:15:00 CST
579    Tag: blume-20041111-word64
580    Description:
581    
582    Structure Word64 is now (almost) complete, word literals and patterns
583    seem to work.  There are a few odd pieces missing.  In particular,
584    I didn't do the {from,to}LargeWord parts because LargeWord is still Word32
585    at the moment.
586    
587    Making Word64 official would mean that LargeWord becomes Word64.  But
588    this requires extreme care because most word-word conversions have to
589    go through LargeWord, so making a mistake means loss of efficiency or
590    worse.  Eventually there will be a solution similar to (but actually
591    simpler than) what I did with IntInf.
592    
593    ----------------------------------------------------------------------
594    Name: Matthias Blume (blume (at) tti - c (dot) org)
595    Date: 2004/11/10 18:12:00 CST
596    Tag: blume-20041110-64bit
597    Description:
598    
599    More 64-bit hacking (but still not even half-way there yet).
600    Also, some assorted improvements to the handling of 8-bit words.
601    
602    ----------------------------------------------------------------------
603    Name: Matthias Blume (blume (at) tti - c (dot) org)
604    Date: 2004/11/09 17:50:00 CST
605    Tag: <none>
606    Description:
607    
608    Started adding some infrastructure for supporting 64-bit int- and
609    word-types.  (Still in its very early stages.)
610    
611    ----------------------------------------------------------------------
612    Name: Matthias Blume (blume (at) tti - c (dot) org)
613    Date: 2004/10/28 10:45:00 CDT
614    Tag: Release_110_50
615    Description:
616    
617    New working version (110.50).  NEW BOOTFILES!
618    
619                 =====================
620    
621    Also:
622    
623      - Changend config/srcarchiveurl from a file just
624        containing the URL string into a file containing
625        shell script code.  The code has access to the $VERSION variable.
626      - Made corresponding changes to config/install.sh and config/unpack.
627      - Default contents of config/srcarchiveurl uses $VERSION and
628        normally does not have to be edited to reflect a version change.
629    
630      (As a result, a version change can be done by just editing
631       config/version, the rest is now automatic.)
632    
633    ----------------------------------------------------------------------
634    Name: Matthias Blume (blume (at) tti - c (dot) org)
635    Date: 2004/10/27 17:50:00 CDT
636    Tag: blume-20041027-btrace-msg
637    Description:
638    
639    BackTrace.monitor now also reports the source of the exception that
640    triggered the trace.
641    
642    ----------------------------------------------------------------------
643    Name: Matthias Blume (blume (at) tti - c (dot) org)
644    Date: 2004/10/27 17:20:00 CDT
645    Tag: blume-20041027-x86-c-calls
646    Description:
647    
648    This is the HISTORY entry for two earlier commits, both concerning
649    the x86 c-calls code in MLRISC:
650    
651      - added a missing LOAD in the code that deals with struct arguments
652      - made sure the caller does not add the wrong number of bytes to the
653        stack pointer after a call of a function returning a struct
654        (the callee already pops the implicit argument which points to
655         the space reserved for the result)
656    
657    ----------------------------------------------------------------------
658    Name: Allen Leung (leunga (at) reservoir (dot) com)
659    Date: 2004/10/24 14:00:00 EST
660    Tag: leunga-20041024-x86-gas-fucomip
661    Description:
662    
663       John discovered a bug in the syntax of fucomip.
664       The opcodes FU?COMIP? have been changed to
665    
666           fu?comip? %st(i), %st
667    
668    ----------------------------------------------------------------------
669    Name: Matthias Blume (blume (at) tti - c (dot) org)
670    Date: 2004/10/20 15:06:00 CDT
671    Tag: blume-20041020-standalone-backtrace
672    Description:
673    
674    Added a mechanism for getting back-trace information from standalone
675    programs.  Here is how it works:
676    
677       1. The part of the program from which you want to get backtrace
678          information (usually the whole program) should be wrapped with
679          BackTrace.monitor.  This is a (unit->'a)->'a function, and your
680          main program could be modified from something like
681    
682             fun main (pgm, args) = ...
683    
684         to
685    
686            fun main (pgm, args) = BackTrace.monitor (fn () => ...)
687    
688       2. To be able to access BackTrace.monitor, you have to add
689          library $smlnj-tdp/plugins.cm to the .cm file that contains your
690          main function.
691    
692       3. Remove all compiled code (i.e., all the .cm/ subdirectories that
693          CM might have created in the past for your project).
694    
695       4. Build the system using this command line:
696    
697             ml-build -Ctdp.instrument=true \$smlnj-tdp/back-trace.cm \
698                      myprog.cm MyProg.main myprog
699    
700          instead of the usual
701    
702             ml-build myprog.cm MyProg.main myprog
703    
704    I changed a library name:
705    
706       $/trace-debug-profile.cm --> $smlnj-tdp/plugins.cm
707    
708    New libraries:
709    
710       $smlnj-tdp/back-trace.cm
711          -- when loaded causes the back-trace plugin to be installed
712       $smlnj-tdp/coverage.cm
713          -- when loaded causes the coverage plugin to be installed
714    
715    ----------------------------------------------------------------------
716    Name: Matthias Blume (blume (at) tti - c (dot) org)
717    Date: 2004/10/18 16:45:00 CDT
718    Tag: blume-20041018-groupowner
719    Description:
720    
721    Added an "obsolete" warning for the "group owner" syntax to CM's parser.
722    
723    Eliminated group owner specs from .cm files throughout the source tree.
724    
725    ----------------------------------------------------------------------
726    Name: Matthias Blume (blume (at) tti - c (dot) org)
727    Date: 2004/10/15 15:45:00 CDT
728    Tag: blume-20041015-coverage
729    Description:
730    
731    * Test coverage tool added!
732    * Further reorganization of tracing-, debugging-, and profiling support:
733    
734        - moved original BTImp -- now called BackTrace -- into a separate
735          library called $/trace-debug-profile.cm
736        - eliminated all mentions of BTrace from SMLofNJ.Internals
737        - only the instrumentation mechanism is now left in the compiler proper
738        - BackTrace module is a plugin which is NOT plugged in by default
739        - Coverage module is another such plugin
740    
741    To get the benefits of any of these plugin modules, the code in
742    question must be compiled with tdp instrumentation turned on.  This
743    can be done by setting SMLofNJ.Internals.TDP.mode to true.
744    (The ref cell is also controlled via the -Ctdp.instrument=... switch.)
745    
746    Plugins are selected at link time.  (Pre-compiled instrumented code
747    can be re-loaded with different plugins in effect.)  When an
748    instrumented module is linked, whatever plugins are at that time
749    enabled will come into effect for that module.
750    
751    To enable the back-trace plugin, load library $/trace-debug-profile.cm
752    and invoke BackTrace.install() (e.g., from the interactive prompt).
753    To enable the coverage plugin, load the same library and invoke
754    Coverage.install().
755    
756    Back-traces are generated automatically on uncaught exceptions and
757    when the code in question explicitly invokes BackTrace.trigger().
758    
759    Coverage (and execution frequency-) information must be queried
760    explicitly by calling Coverage.not_covered and Coverage.hot_spots.
761    
762    ----------------------------------------------------------------------
763    Name: Matthias Blume (blume (at) tti - c (dot) org)
764    Date: 2004/10/14 17:40:00 CDT
765    Tag: blume-20041014-tdp-core
766    Description:
767    
768    Snapshot of a significant overhaul of how the trace/debug/profile support
769    is hooked into the system (specifically: Core and SMLofNJ.Internals).
770    
771    ----------------------------------------------------------------------
772    Name: Matthias Blume (blume (at) tti - c (dot) org)
773    Date: 2004/10/13 16:34:00 CDT
774    Tag: blume-20041013-tdp
775    Description:
776    
777    Some rationalization of names:
778    
779       structure BTrace -> structure TDPInstrument
780       etc.
781    
782    This is is preparation of using the original back-trace
783    instrumentation for other purposes.  "TDP" stands for
784    Trace/Debug/Profile.
785    
786    The control flag controlling whether instrumentation is on or off is now
787    registered under a different name, so instead of running sml as
788    
789     sml -Cinstrument.btrace-mode=true
790    
791    one has to say
792    
793     sml -Ctdp.instrument=true
794    
795    ----------------------------------------------------------------------
796    Name: Matthias Blume (blume (at) tti - c (dot) org)
797    Date: 2004/10/11 16:37:00 CDT
798    Tag: blume-20041011-regions
799    Description:
800    
801    Made some minor modifications to elabcore.sml to have source regions
802    be propagated more tightly -- resulting in better (i.e., smaller)
803    regions being reported in error- and debug messages.
804    
805    ----------------------------------------------------------------------
806    Name: Matthias Blume (blume (at) tti - c (dot) org)
807    Date: 2004/10/08 22:50:00 CDT
808    Tag: blume-20041008-cmkw
809    Description:
810    
811    Fixed handling of keywords in .cm files: After seeing "is" the lexer
812    treats subsequent occurrences of "group", "library", "source", "is",
813    "*", and "-" as ordinary identifiers rather than keywords.
814    
815    Most seriously, this fixes a problem with CM's "shell" tool.  The tool
816    is supposed to accept a tool argument called "source", but this did
817    not work because of the clash with the keyword.
818    
819    ----------------------------------------------------------------------
820    Name: Matthias Blume (blume (at) tti - c (dot) org)
821    Date: 2004/10/07 16:00:00 CDT
822    Tag: blume-20041007-cleanup
823    Description:
824    
825    Assorted cleanup work:
826    
827      - got rid of intstrmap in favor of using the library's
828        hash table implementation
829      - threw out most of the pathnames stuff, as it was not used anyway
830      - simplified tokentable implementation
831      - fixed some minor spelling errors
832    
833    ----------------------------------------------------------------------
834    Name: Matthias Blume (blume (at) tti - c (dot) org)
835    Date: 2004/10/06 15:15:15 CDT
836    Tag: blume-20041006-handler
837    Description:
838    
839    Cleaned up the absyn to reflect the invariant that HANDLE always
840    carries a FNexp as part of the type definition.  This eliminates some
841    superfluous sanity checks at runtime down the road.
842    
843    Some minor cleanup of the btrace code.
844    
845    ----------------------------------------------------------------------
846    Name: Matthias Blume (blume (at) tti - c (dot) org)
847    Date: 2004/10/01 10:20:30 CDT
848    Tag: blume-20041001-slave
849    Description:
850    
851    Added hack to make slave mode work in the presence of the version
852    tool.  (Still, since the master does two passes over the code for
853    CMB.make, the release number gets bumped twice when slaves are
854    attached.  I don't know if this is worth fixing...)
855    
856    ----------------------------------------------------------------------
857    Name: Matthias Blume (blume (at) tti - c (dot) org)
858    Date: 2004/09/30 10:55:00 CDT
859    Tag: blume-20040930-version
860    Description:
861    
862    * Moved the "version" magic into its own little library under
863      src/system/smlnj/internal.  This avoids expensive reconstruction of
864      a stable src/compiler/core.cm.
865    
866    * At the same time, structure CompilerVersion is now known as
867      structure SMLNJVersion.
868    
869    * Arranged for the version tool to NOT kick in when rebuilding the system
870      (makeml -rebuild, fixpt).  Otherwise one would never reach a fixpoint.
871      Also, loading the versiontool does not work when rebuilding the system
872      because CM is not properly initialized at that time.
873    
874    ----------------------------------------------------------------------
875    Name: Matthias Blume (blume (at) tti - c (dot) org)
876    Date: 2004/09/29 14:00:00 CDT
877    Tag: blume-20040929-autoversion
878    Description:
879    
880    Implemented some CM magic to have
881        file src/compiler/TopLevel/main/version.sml
882    generated automagically.
883    The version is taken from two files: config/version and config/release.
884    The first is expected to contain a two-part version number such as 110.49.
885    The second should contain a single number, but it may be missing.
886    
887    If the environment variable VERSIONTOOL_BUMP_RELEASE is defined at the
888    time the version tool is loaded (which is the first time you say CMB.make),
889    then the tool will increment the value stored in config/release every
890    time CMB.make is invoked.
891    
892    The binfile format is now insensitive to anything beyond the first
893    two components of a version number, so bumping the release does not render
894    binfiles incompatible.  Auto-bumping can be used to keep track of versions
895    during development without invalidating existing binfiles.
896    
897    In any case, every CMB.make updates the date information in version.sml.
898    (This is the date that is printed in the banner.)
899    
900    ----------------------------------------------------------------------
901    Name: Matthias Blume (blume (at) tti - c (dot) org)
902    Date: 2004/09/28 10:53:00 CDT
903    Tag: blume-20040928-controls
904    Description:
905    
906    Some cleanup of the controls code.
907    
908    ----------------------------------------------------------------------
909    Name: Matthias Blume (blume (at) tti - c (dot) org)
910    Date: 2004/09/27 22:08:00 CDT
911    Tag: blume-20040927-controls
912    Description:
913    
914    Added two pieces of functionality to the Controls interface:
915    
916    1.   val save'restore: 'a control -> unit -> unit
917    
918       grabs the current value of the control in stage 1 and restores it
919       in stage 2.
920    
921    2.   val set' : 'a control * 'a -> unit -> unit
922    
923       stores the given value into the control in stage 2 (i.e., delayed)
924       but does all error checking in stage 1.
925       (This is for string controls that need to do parse their argument --
926        something that might fail.  In some cases, notably in CM, one
927        already knows the intended argument but wants to delay the actual
928        assignment until a time when error recovery would be more difficult.)
929    
930    Changed the handling of controls in tool arguments to classes "sml" and
931    "lazysml":
932       - use Controls.save'restore as a more robust way of restoring the
933         old value (in particular: without having to re-parse the string)
934       - use controls to handle the "overload" keyword in the init group
935         (I believe this change actually fixes a long-standing obscure bug.)
936    
937    ----------------------------------------------------------------------
938    Name: Matthias Blume (blume (at) tti - c (dot) org)
939    Date: 2004/09/27 17:00:00 CDT
940    Tag: blume-20040927-lazysml
941    Description:
942    
943    Added a new tool class called "lazysml" to CM's tool chest.  The only
944    difference to "sml" is that compilation is done with Control.lazysml
945    set to true.  A source of class "lazysml" is automatically recognized
946    by a file name suffix of ".lml".
947    
948    In addition to the above feature, the original class "sml" now also
949    supports a tool argument "lazy" which has the same effect.  As a
950    result, the following three lines are equivalent:
951    
952        foo.sml : lazysml
953        foo.sml : sml (lazy)
954        foo.sml (lazy)
955    
956    The setting goes into effect both during parsing and during
957    compilation.  The original setting is restored right after parsing and
958    after compilation, respectively.
959    
960    In addition to all the above, there is also a general mechanism to set
961    ANY of the "controls" that are available at the command line via
962    "-C..." on a per-sml-file basis.  The same rules that apply for "lazy"
963    apply as well. (In fact, "lazy" is implemented as a special case of
964    the general mechanism.)
965    
966    The .cm file syntax uses a new keyword tool argument called "with".
967    There are several ways of indicating the desired settings:
968    
969        foo.sml (with:parser.quotations=true)
970        foo.sml (with:(name:parser.quotations value:true))
971        foo.sml (with:(name:name1 value:value1 name:name2 value:value2 ...))
972        foo.sml (with:(name1=value1 name2=value2 ...))
973        foo.sml (with:(name1=value1 name:name2 value:value2 name3=value3 ...))
974    
975        etc.
976    
977    Another possible abbreviation is to leave out the =v or value:v part
978    if the name refers to a boolean control (in which case the value is
979    taken to be true).  Thus, one could get lazy sml also by saying:
980    
981        foo.sml (with:parser.lazy-keyword=true)
982        foo.sml (with:parser.lazy-keyword)
983        foo.sml (with:(name:parser.lazy-keyword value:true))
984        foo.sml (with:(name:parser.lazy-keyword))
985    
986    ----------------------------------------------------------------------
987    Name: Matthias Blume (blume (at) tti - c (dot) org)
988    Date: 2004/09/24 16:22:00 CDT
989    Tag: blume-20040924-ppc-long-branch
990    Description:
991    
992    Turned message about "emiting long form of branch" off by default.
993    Added a control flag to turn it back on when desired.
994    
995    ----------------------------------------------------------------------
996    Name: Matthias Blume (blume (at) tti - c (dot) org)
997    Date: 2004/09/24 16:05:00 CDT
998    Tag: blume-20040924-rounding
999    Description:
1000    
1001    Applied patch for setting rounding modes under Mac OS X.  Thanks to
1002    Melissa O'Neill for providing the code!
1003    
1004    ----------------------------------------------------------------------
1005    Name: Matthias Blume (blume (at) tti - c (dot) org)
1006    Date: 2004/09/23 17:30:00 CDT
1007    Tag: blume-20040923-envvars
1008    Description:
1009    
1010    1. Changed definition of type ControlRegistry.registry_tree to
1011       include control_info (i.e., the name of the controlling environment
1012       variable).
1013    
1014    2. Added command-line flags -e and -E to print the names of environment
1015       variables that can be used to control internal settings.
1016       (This uses the new API mentioned in 1.)
1017    
1018    ----------------------------------------------------------------------
1019    Name: Matthias Blume (blume (at) tti - c (dot) org)
1020    Date: 2004/09/13 16:50:00 CDT
1021    Tag: Release_110_49
1022    Description:
1023    
1024    New working version (110.49).  NEW BOOTFILES!
1025    
1026    ----------------------------------------------------------------------
1027    Name: Matthias Blume
1028    Date: 2004/09/13 16:20:00 CDT
1029    Tag: blume-20040913-config-mlrisc
1030    Description:
1031    
1032    Put target "mlrisc" back into the default list.
1033    (There is no harm in having it, and some users have expressed their
1034    wish to have "mlrisc" included by default.)
1035    
1036    ----------------------------------------------------------------------
1037    Name: John Reppy
1038    Date: 2004/09/13
1039    Tag: jhr-20040913-signals
1040    Description:
1041    
1042    Fixed the signal masking code to properly nest mask/unmask operations
1043    on a per-signal basis.
1044    
1045    ----------------------------------------------------------------------
1046    Name: Matthias Blume (blume (at) tti - c (dot) org)
1047    Date: 2004/09/08 13:20:00 CDT
1048    Tag: blume-20040908-heap-magic
1049    Description:
1050    
1051    Bumped the heap macig number to 0x09082004 to account for the changed
1052    layout of the ML frame under MacOS X.
1053    
1054    ----------------------------------------------------------------------
1055    Name: Allen Leung (leunga (at) reservoir (dot) com)
1056    Date: 2004/09/03 11:26:00 EST
1057    Tag: leunga-20040903-cygwin-install
1058    Description:
1059    
1060       Added a patch to _arch-n-opsys to enable the Cygwin runtime.
1061    The Cygwin runtime is turned on by setting the environment
1062    variable SMLNJ_CYGWIN_RUNTIME to 1.
1063    
1064    ----------------------------------------------------------------------
1065    Name: Matthias Blume (blume (at) tti - c (dot) org)
1066    Date: 2004/08/31 17:15:00 CDT
1067    Tag: blume-20040831-core
1068    Description:
1069    
1070    Added some exports to src/compiler/core.cm upon request by J. Joemann.
1071    
1072    ----------------------------------------------------------------------
1073    Name: Matthias Blume (blume (at) tti - c (dot) org)
1074    Date: 2004/08/30 17:55:00 CDT
1075    Tag: blume-20040830-installer
1076    Description:
1077    
1078    Upon request by Johannes Joemann:
1079    
1080    - improved ML code of installer to fall back to coping when renaming
1081      fails (i.e., when source and target are on different file systems);
1082      the code compiles but has yet to be tested in anger
1083    - removed mlrisc from list of default targets (config/targets)
1084    
1085    ----------------------------------------------------------------------
1086    Name: Matthias Blume (blume (at) tti - c (dot) org)
1087    Date: 2004/08/27 17:20:00 CDT
1088    Tag: blume-20040827-ptreql
1089    Description:
1090    
1091    Added ptreql primop to structure InlineT (upon request from Larry
1092    Paulson).
1093    
1094    ----------------------------------------------------------------------
1095    Name: Allen Leung (leunga (at) reservoir (dot) com)
1096    Date: 2004/08/15 21:21:00 EST
1097    Tag: leunga-110_48-udgraph
1098    Description:
1099    
1100    Another bug fix from Carl Hauser:
1101    
1102    diff /net/niflab/smlnj48/src/MLRISC/graphs/udgraph.sml udgraph.sml
1103    > 48c48
1104    > <              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then es else
1105    > e::L)
1106    > ---
1107    > >              | rmv((e as (k,_))::es,L) = rmv(es,if k = i then L else e::L)
1108    > Without this, any deletion of an edge in an undirected graph does severe
1109    > violence to the graph.
1110    
1111    ----------------------------------------------------------------------
1112    Name: Allen Leung (leunga (at) reservoir (dot) com)
1113    Date: 2004/08/10 23:35:00 EST
1114    Tag: leunga-110_48-ppc
1115    Description:
1116    
1117       The IBM/MacOS syntax switch on PPC was incorrectly swapped.
1118    
1119    ----------------------------------------------------------------------
1120    Name: Matthias Blume (blume (at) tti - c (dot) org)
1121    Date: 2004/08/10 12:00:00 CDT
1122    Tag: Release_110_48
1123    Description:
1124    
1125    New working version (110.48).  NEW BOOTFILES!
1126    
1127    ----------------------------------------------------------------------
1128    Name: Allen Leung (leunga (at) reservoir (dot) com)
1129    Date: 2004/08/09 12:21:00 EST
1130    Tag: leunga-110_47-dijsktra
1131    Description:
1132    
1133    Bug fix from Carl Hauser:
1134    
1135    single_source_shortest_paths in dijkstra.sml was observed to get wrong
1136    answers (by comparing to single_source_shortest_paths in bellman-ford.sml).
1137    
1138    The problem is that following the expression A.update(dist,s,Num.zero)
1139    it is necessary to update the priority queue using Q.decreaseWeight(Q,s).
1140    
1141    ----------------------------------------------------------------------
1142    Name: Matthias Blume (blume (at) tti - c (dot) org)
1143    Date: 2004/08/06 18:10:00 CDT
1144    Tag: blume-20040806-cmdline
1145    Description:
1146    
1147    Fiddled with handling of command-line options:
1148    
1149      * sml now quits after processing the command line
1150        if -H, -S, -h<n>, or -s<n> appears as the last
1151        command-line argument
1152      * a new option -q terminates the session when encountered on
1153        the command line; subsequent arguments will be ignored
1154      * bug fixes: short (erroneous) arguments are no longer ignored
1155        completely
1156    
1157    ----------------------------------------------------------------------
1158    Name: Allen Leung (leunga (at) reservoir (dot) com)
1159    Date: 2004/08/04 18:17:00 EST
1160    Tag: leunga-110_47-ppc-ibm-asm
1161    Description:
1162    
1163      - Added minimal IBM assembly syntax support for PowerPC.
1164    
1165      - Cygwin: manually changed the file cygwin.def.  Some exported symbols have
1166        been altered in the runtime.  We need an automatic way to keep the file
1167        in sync.
1168    
1169    ----------------------------------------------------------------------
1170    Name: Matthias Blume (blume (at) tti - c (dot) org)
1171    Date: 2004/08/04 14:00:00 CDT
1172    Tag: Release_110_47
1173    Description:
1174    
1175    New working version (110.47).  NEW BOOTFILES!
1176    
1177    ----------------------------------------------------------------------
1178    Name: Matthias Blume (blume (at) tti - c (dot) org)
1179    Date: 2004/08/03 14:25:00 CDT
1180    Tag: blume-20040803-callingconv
1181    Description:
1182    
1183    Added low-level support for choosing C calling conventions by
1184    twiddling the type of rawccall.  (See
1185    src/compiler/Semant/types/cproto.sml for details.)
1186    
1187    ----------------------------------------------------------------------
1188    Name: Matthias Blume (blume (at) tti - c (dot) org)
1189    Date: 2004/08/02 15:55:00 CDT
1190    Tag: blume-20040802-backout
1191    Description:
1192    
1193    Backed out of change to win32-filesys.c.  The earlier patch to
1194    get_file_time caused CM to produce files with the wrong time stamp.
1195    
1196    ----------------------------------------------------------------------
1197    Name: Matthias Blume (blume (at) tti - c (dot) org)
1198    Date: 2004/08/02 14:45:00 CDT
1199    Tag: blume-20040802-nlffi-win32
1200    Description:
1201    
1202    Added NLFFI support for Win32, adapted from a patch provided by David
1203    Hansel.  This is currently completely untested.  Also, the issue
1204    concerning stdcall vs. ccall is still unresolved.
1205    
1206    ----------------------------------------------------------------------
1207    Name: Matthias Blume (blume (at) tti - c (dot) org)
1208    Date: 2004/07/30 17:55:00 CDT
1209    Tag: blume-20040730-various
1210    Description:
1211    
1212    Gearing up towards 110.47...
1213    
1214    - various minor bugfixes to ml-nlffigen
1215    - a beginning of a manual for nlffi
1216    
1217    - eliminated 'export name=value' in config/install.sh as this does
1218      not work with certain versions of /bin/sh
1219      (Thanks to David King at Motorola for catching this.)
1220    
1221    - several bugfixes provided or suggested by David Hansel at Reactive Systems:
1222      - added a test for tm==NULL to gmtime.c and localtime.c
1223      - applied patch for incorrect GetFileTime under win32
1224      - toSeconds -> toMilliseconds in Win32/win32-process.sml
1225    
1226    ----------------------------------------------------------------------
1227    Name: Matthias Blume (blume (at) tti - c (dot) org)
1228    Date: 2004/07/21 18:20:00 CDT
1229    Tag: blume-20040721-nlffigen
1230    Description:
1231    
1232    - Fixed minor issue in ml-nlffigen: Now generate structure T_foo
1233      for a typedef to an incomplete type, but leave out the "typ" member.
1234      (This is just for consistency.)
1235    - Started to produce what is supposed to become better (i.e., comprehensive)
1236      documentation of what ml-nlffigen does and produces.
1237    
1238    ----------------------------------------------------------------------
1239    Name: Matthias Blume (blume (at) tti - c (dot) org)
1240    Date: 2004/07/14 16:25:00 CDT
1241    Tag: blume-20040714-union
1242    Description:
1243    
1244    Added C_UNION to c-calls/c-types.sml and updated the machinery
1245    (ml-nlffigen, cproto.sml) that conveys C function interface
1246    information to the code generator.
1247    
1248    However, the actual architecture-specific implementation of function
1249    arguments and results that are C unions is still not implemented.
1250    
1251    ----------------------------------------------------------------------
1252    Name: Allen Leung (leunga (at) reservoir (dot) com)
1253    Date: 2004/07/14 14:38:00 EST
1254    Tag: leunga-110_46_1-ppc-lwzu
1255    Description:
1256    
1257        Added these instructions to the PowerPC architecture:
1258        LBZU(X), LHZU(X), LWZU(X),
1259        STWU(X), STFDU, STFSU
1260    
1261        etc...
1262    
1263        Note: I haven't added their instruction encoding into the description.
1264    
1265    ----------------------------------------------------------------------
1266    Name: Allen Leung (leunga (at) reservoir (dot) com)
1267    Date: 2004/07/13 15:04:00 EST
1268    Tag: leunga-110_46_1-ppc-lwarx
1269    Description:
1270    
1271       Added the two instructions LWARX and STWCX to the PowerPC
1272    instruction set.
1273    
1274       A (untested) rewrite of loop-structure.sml.  The old version
1275    is completely broken.
1276    
1277    ----------------------------------------------------------------------
1278    Name: Matthias Blume (blume (at) tti - c (dot) org)
1279    Date: 2004/07/13 13:50:00 CDT
1280    Tag: blume-20040713-nlffi
1281    Description:
1282    
1283    - use paramAlloc to report c-calls with too many arguments
1284      (for PPC version where parameter area is pre-allocated)
1285    - added ccall_maxargspace to machspec (to implement the above)
1286    - made "make" commend in CM's "make" tool configurable
1287    - added option (default: on) for passing the name of the SML/NJ's "bin"
1288      directory to "make"; the call looks like this:
1289    
1290         make <options> SMLNJ_BINDIR=<dir> <target>
1291    
1292      This can be used by the Makefile to, e.g., pick the "right" version
1293      of ml-nlffigen.
1294    - minor code tweaks
1295    
1296    ----------------------------------------------------------------------
1297    Name: Matthias Blume (blume (at) tti - c (dot) org)
1298    Date: 2004/07/12 22:50:00 CDT
1299    Tag: blume-110_46_1-macosx-nlffi
1300    Description:
1301    
1302    NLFFI under Mac OS X now working (sort of).  This is largely untested,
1303    though.
1304    
1305    Note:  1. You have to make a new, clean build of the runtime system.
1306           2. There are new BOOTFILES, you have to use them!
1307              (Doing the bootstrap process yourself would be *very* painful!
1308               If you absolutely have to do it, build the system under
1309               a different architecture and then cross-compile.)
1310    
1311    Version bumped to 110.46.1 to account for runtime data format changes.
1312    
1313    ----------------------------------------------------------------------
1314    Name: Matthias Blume (blume (at) tti - c (dot) org)
1315    Date: 2004/06/18 14:30:00 CDT
1316    Tag: blume-20040618-unix
1317    Description:
1318    
1319    Changed the implementation of structure Unix so that the same stream
1320    is returned every time one of the {text,bin}{In,Out}streamOf functions
1321    is invoked on the same proc.  This is not what the spec currently
1322    says -- although IMO it arguably should.  (See discussion below.)
1323    
1324    ----------------------------------------------------------------------
1325    Name: Matthias Blume (blume (at) tti - c (dot) org)
1326    Date: 2004/06/17 18:15:00 CDT
1327    Tag: Release_110_46
1328    Description:
1329    
1330    New working version (110.46).  NEW BOOTFILES!
1331    
1332    ----------------------------------------------------------------------
1333    Name: Matthias Blume (blume (at) tti - c (dot) org)
1334    Date: 2004/06/17 17:20:00 CDT
1335    Tag: blume-20040617-timer-unix
1336    Description:
1337    
1338    Changed the interface of structures Timer and Unix to match the most
1339    recent Basis spec.
1340    
1341    In the case of Unix there still seems to be an open/weird issue:
1342    
1343       The {text,bin}{In,Out}streamOf functions are supposed to create
1344       fresh streams whenever they are called -- as opposed to have them
1345       return the same stream every time.  This design is supposed to
1346       prevent space leaks caused by proc values hanging on to streams.
1347    
1348       The reap function, on the other hand, is supposed to close the
1349       streams.  This cannot be done without having a handle on the
1350       stream in proc after all...
1351    
1352    I took the liberty to implement the following stopgap solution:
1353    
1354      The proc value hangs on to the most recently created stream(s).
1355      Reap closes those.  If either or both of the two streams hadn't
1356      been created at all yet, then reap will close the corresponding
1357      file descriptors directly.
1358    
1359    PS: I don't understand the original space leak argument anymore.  If
1360    a proc hangs on to the imperative stream, then I/O operations on those
1361    will advance the state of the cached stream and avoid the space leak.
1362    
1363    ----------------------------------------------------------------------
1364    Name: Matthias Blume (blume (at) tti - c (dot) org)
1365    Date: 2004/05/28 16:45:00 CDT
1366    Tag: blume-20040528-basis
1367    Description:
1368    
1369    Added signature PACK_REAL and exported functor PrimIO.
1370    
1371    ----------------------------------------------------------------------
1372    Name: Matthias Blume (blume (at) tti - c (dot) org)
1373    Date: 2004/05/25 16:00:00 CDT
1374    Tag: blume-20040525-group-owner
1375    Description:
1376    
1377    CM now ignores (but still accepts) the "owner" information in group
1378    descriptions.  The owner of a group is its next enclosing
1379    library. Each group must have a unique owner.  (There is a virtual
1380    "toplevel" library that owns groups which are not nested within a real
1381    library.)  Previously, each group had to explicitly declare its owner,
1382    and CM would check that such a declaration is correct.  The new scheme
1383    is to have CM check that for each group there is precisely one owning
1384    library.
1385    
1386    The advantage of the new scheme is that the programmer no longer needs
1387    to maintain the somewhat annoying owner information.  The downside is
1388    that CM cannot enforce the ownership rule across multiple runs of
1389    CM.make.  Fortunately, enclosing the same group in two different
1390    libraries A and B which are not part of the same program does not
1391    cause real problems.
1392    
1393    ----------------------------------------------------------------------
1394    Name: Matthias Blume (blume (at) tti - c (dot) org)
1395    Date: 2004/05/20 16:00:00 CDT
1396    Tag: blume-20040520-win32
1397    Description:
1398    
1399    Made the win32 version work again.  (Strangely, a misplaced comma had
1400    slipped into win32-process.c which prevented the runtime from being
1401    compiled correctly.)
1402    
1403    Also, included a minor addition to ml-build.bat analogous to what was
1404    done in blume-20040519-ml-build.
1405    
1406    ----------------------------------------------------------------------
1407    Name: Matthias Blume (blume (at) tti - c (dot) org)
1408    Date: 2004/05/19 22:10:00 CDT
1409    Tag: blume-20040519-ml-build
1410    Description:
1411    
1412    Arranged for ml-build to clean up after itself a little bit better.
1413    The script generates a temporary SML source file and compiles it using
1414    CM, so CM generates metadata (GUID, SKEL, objectfile) for it.  It now
1415    gets rid of those at the end, so they don't accumulate under .cm.
1416    
1417    This required a minor change to install.sh because the name of the
1418    metadata directory (default: .cm) is actually configurable at
1419    installation time.
1420    
1421    ----------------------------------------------------------------------
1422    Name: Matthias Blume (blume (at) tti - c (dot) org)
1423    Date: 2004/05/18 15:50:00 CDT
1424    Tag: blume-20040518-mkreader
1425    Description:
1426    
1427    Added Posix.IO.mk{Bin,Text}{Reader,Writer} by lifting their respective
1428    implementations from internal modules PosixBinPrimIO and PosixTextPrimIO.
1429    
1430    ----------------------------------------------------------------------
1431    Name: Matthias Blume (blume (at) tti - c (dot) org)
1432    Date: 2004/05/11 14:35:00 CDT
1433    Tag: blume-20040511-win32sock
1434    Description:
1435    
1436    Added previously missing support for many socket-related functions
1437    under win32.  Thanks to David Hansel <hansel@reactive-systems.com>
1438    for the voluminous patch!
1439    
1440    (I have not tested this patch under win32 yet.)
1441    
1442    Here is David's e-mail:
1443    
1444    Hi,
1445    
1446    Attached to this email you find a diff against sml/nj 110.45
1447    that will enable socket support under Windows.
1448    
1449    To apply the patch (using unix or cygwin)
1450    1) gunzip runtime.diff.gz
1451    2) "cd" into "src/runtime" in the source tree of a fresh
1452       110.45 installation.
1453    3) patch -p 1 < [your/path/to]runtime.diff
1454    
1455    The code compiles fine but has NOT yet been extensively tested.
1456    I only ran a few tests for basic socket client functionality
1457    (which worked fine).  Especially the functions that use ioctl
1458    are not tested at all and might not work (see below).
1459    
1460    I implemented this since we want to move to a newer version of sml/nj
1461    but need socket support in order to use it.  This is the first time I
1462    even had a look at the sml/nj source,  so please review my changes
1463    before making this part of the distribution!  Here are a few issues
1464    that I think might be better for someone to solve who is more
1465    familiar with the sml/nj source (and socket programming):
1466    
1467    - getnetbyaddr.c and getnetbyname.c will raise a "not implemented"
1468      exception since I could not figure out what the windows equivalent
1469      of these functions is
1470    
1471    - In sockets-osdep.h there are a some #include statements that are
1472      only used in a few files that include sockets-osdep.h
1473    
1474    - In smlnj-sock-lib.c, function init_fn() calls WSAStartup() but
1475      does not process its return value since I don't know how to
1476      report an error upwards.
1477    
1478    - It would probably be good to have a call to WSACleanup() when
1479      the library is unloaded (if there is such a possibility).
1480      Otherwise I think Windows will take care of this automatically
1481      when the process finishes.
1482    
1483    - I used ioctlsocket() as a replacement for ioctl() but I have
1484      no idea if that is actually the proper replacement on Windows.
1485    
1486    - All these issues are marked in the code by "FIXME" comments.
1487    
1488    We use sml/nj extensively in our products and are quite happy
1489    with it.  I hope this contribution will help you.
1490    
1491    Keep up the good work!
1492    
1493    David
1494    
1495    ----------------------------------------------------------------------
1496    Name: Matthias Blume (blume (at) tti - c (dot) org)
1497    Date: 2004/05/11 14:20:00 CDT
1498    Tag: blume-20040511-installml
1499    Description:
1500    
1501    Fixed two bugs in installml script.  (Thanks to Vesa A. Norrman for
1502    the patch.)
1503    
1504    ----------------------------------------------------------------------
1505    Name: Matthias Blume (blume (at) tti - c (dot) org)
1506    Date: 2004/05/11 14:05:00 CDT
1507    Tag: blume-20040511-nlffi-netbsd
1508    Description:
1509    
1510    Added support for nlffi under netbsd.  (Thanks to Vesa A. Norrman for
1511    the patch.)
1512    
1513    ----------------------------------------------------------------------
1514    Name: Matthias Blume (blume (at) tti - c (dot) org)
1515    Date: 2004/05/11 12:05:00 CDT
1516    Tag: blume-20040511-exports
1517    Description:
1518    
1519    As per request by Adam Chlipala <adam@hcoop.net>, extended various
1520    export lists in compiler-related .cm-files.
1521    
1522    ----------------------------------------------------------------------
1523    Name: Matthias Blume (blume (at) tti - c (dot) org)
1524    Date: 2004/05/11 11:35:00 CDT
1525    Tag: blume-20040511-allsource
1526    Description:
1527    
1528    The installer now honors the "src-smlnj" target again, although its meaning
1529    has changed from "all sources required for the compiler" to "all sources
1530    the installer knows about".  In other words, if you enable "src-smlnj"
1531    in the "targets" file, then the installer will pull in sources for
1532    everything.  (Notice that this refers to source code only.  Compiled
1533    code is still only installed for modules that were requested explicitly
1534    or which are required for other modules that were requested explicitly.)
1535    
1536    ----------------------------------------------------------------------
1537    Name: Matthias Blume (blume (at) tti - c (dot) org)
1538    Date: 2004/04/23 17:40:00 CDT
1539    Tag: blume-20040423-ieee-scan
1540    Description:
1541    
1542    Fixed IEEEReal.scan (and .fromString) so that if there is an overflow
1543    in the exponent calculation we get INF or ZERO (depending on the mantissa
1544    and the sign of the exponent).
1545    
1546    ----------------------------------------------------------------------
1547    Name: Matthias Blume (blume (at) tti - c (dot) org)
1548    Date: 2004/04/23 10:40:00 CDT
1549    Tag: blume-20040423-ml-build
1550    Description:
1551    
1552    The ml-build script now terminates with a non-0 status when something
1553    goes wrong.
1554    
1555    ----------------------------------------------------------------------
1556    Name: Matthias Blume (blume (at) tti - c (dot) org)
1557    Date: 2004/04/22 16:35:00 CDT
1558    Tag: blume-20040422-Option
1559    Description:
1560    
1561    Made exception Option to be the same as exception Option.Option
1562    (as it should be).
1563    
1564    ----------------------------------------------------------------------
1565    Name: Allen Leung (leunga (at) reservoir (dot) com)
1566    Date: 2004/03/19 14:40:00 EST
1567    Tag: leunga-20040319-cygwin-nlffi
1568    Description:
1569    
1570    Fixed the runtime so that ml-nlffi-lib runs on the cygwin version
1571    of SML/NJ.  The problem is that
1572    
1573       lib = dlopen(NULL, ...)
1574       f   = dlsym(lib, "malloc");
1575    
1576    does not work on Windows unless we explicitly export symbols
1577    such as 'malloc' during linking.  We fixed this by explicitly
1578    exporting the required symbols with the magic gcc incantation:
1579    
1580        -Wl,--export-all cygwin.def
1581    
1582    where cygwin.def is a file containing all the symbols that
1583    we wish to export.
1584    
1585    I suspect this is a Windows problem and we'll have to
1586    do the same (somehow with windows compilers) when
1587    we build the native win32 version with the system
1588    calls LoadLibrary/GetProcAddress.
1589    
1590    ----------------------------------------------------------------------
1591    Name: Matthias Blume (blume (at) tti - c (dot) org)
1592    Date: 2004/03/04 16:35:00 CST
1593    Tag: blume-20040304-intinf-fmt
1594    Description:
1595    
1596    Fixed problem with IntInf.fmt (sign would show up on the right instead
1597    of on the left for BIN, OCT, and HEX).
1598    
1599    ----------------------------------------------------------------------
1600    Name: Matthias Blume (blume (at) tti - c (dot) org)
1601    Date: 2004/03/04 11:25:00 CST
1602    Tag: blume-20040304-symlinks
1603    Description:
1604    
1605    Fixed problem with installer script (unix only) where bin/ml-yacc and
1606    friends pointed (via symlinks) to absolute locations instead of just
1607    .run-sml.  This was reported by Vesa A Norrman.
1608    
1609    ----------------------------------------------------------------------
1610    Name: Matthias Blume (blume (at) tti - c (dot) org)
1611    Date: 2004/02/13 14:50:00 CST
1612    Tag: Release_110_45
1613    Description:
1614    
1615    New working version (110.45). New bootfiles.
1616    
1617    ----------------------------------------------------------------------
1618    Name: Matthias Blume (blume (at) tti - c (dot) org)
1619    Date: 2004/01/26 15:15:15 CST
1620    Tag: blume-20040126-toplevel
1621    Description:
1622    
1623    Improved handling of exceptions at the interactive toplevel.
1624    
1625    ----------------------------------------------------------------------
1626    Name: Matthias Blume (blume (at) tti - c (dot) org)
1627    Date: 2004/01/26 11:25:00
1628    Tag: blume-20040126-app
1629    Description:
1630    
1631    Type of top-level "app" corrected.
1632    Added code for setting vp_limitPtrMask to Win32-specific runtime.
1633    
1634    ----------------------------------------------------------------------
1635    Name: Matthias Blume (blume (at) tti - c (dot) org)
1636    Date: 2003/11/18 17:10 CST
1637    Tag: blume-20031118-basis-fiddle
1638    Description:
1639    
1640    - changed Timer interface to what might become the spec
1641    - POSIX_FLAGS -> BIT_FLAGS according to spec
1642    - some other minor discrepancies wrt. spec eliminated
1643    
1644    ----------------------------------------------------------------------
1645    Name: Matthias Blume (blume (at) tti - c (dot) org)
1646    Date: 2003/11/06 12:00:00 CST
1647    Tag: Release_110_44
1648    Description:
1649    
1650    New working version (110.44). New bootfiles.
1651    
1652    ----------------------------------------------------------------------
1653    Name: Matthias Blume (blume (at) tti - c (dot) org)
1654    Date: 2003/11/04 11:50:00 CST
1655    Tag: blume-20031104-move-libraries
1656    Description:
1657    
1658    Eliminated the "dont_move_libraries" directive in config/targets.
1659    (The mechanism was broken and could not be fixed easily.  Moreover,
1660    there does not seem to be any reason not to move all libraries into
1661    lib during installation.  I originally implemented this directive as a
1662    backward-compatibility feature when I first introduced the new CM.
1663    Now that things have been stable for a long time and going back to the
1664    old CM is not an option, there is no reason to keep it around.)
1665    
1666    ----------------------------------------------------------------------
1667    Name: Matthias Blume (blume (at) tti - c (dot) org)
1668    Date: 2003/11/03 16:00:00 CST
1669    Tag: blume-20031103-installdir
1670    Description:
1671    
1672    Made installer honor INSTALLDIR variable again.  (Thanks to Chris
1673    Richards for pointing out the problem and providing the solution.)
1674    
1675    ----------------------------------------------------------------------
1676    Name: Matthias Blume (blume (at) tti - c (dot) org)
1677    Date: 2003/10/01 17:05:00 CDT
1678    Tag: blume-20031001-lal-mlrisc
1679    Description:
1680    
1681    MLRISC bug fix from Lal.
1682    
1683    ----------------------------------------------------------------------
1684    Name: Matthias Blume (blume (at) tti - c (dot) org)
1685    Date: 2003/09/30 16:10:00 CDT
1686    Tag: blume-20030930-primio-bat
1687    Description:
1688    
1689    1. Added openVector, nullRd, and nullWr to PRIM_IO.
1690    2. Improved .bat files (for Win32 port) to make things work under Win95.
1691       (thanks to Aaron S. Hawley for this one)
1692    
1693    ----------------------------------------------------------------------
1694    Name: Matthias Blume (blume (at) tti - c (dot) org)
1695    Date: 2003/09/26 16:05:00 CDT
1696    Tag: blume-20030926-wrappriv
1697    Description:
1698    
1699    Added missing wrapper for privilege "primitive" in $smlnj/viscomp/core.cm.
1700    
1701    ----------------------------------------------------------------------
1702    Name: Matthias Blume (blume (at) tti - c (dot) org)
1703    Date: 2003/09/26 15:00:00 CDT
1704    Tag: blume-20030926-110_43_3
1705    Description:
1706    
1707    - additional cleanup
1708    - version number bump, NEW BOOTFILES
1709    
1710    ----------------------------------------------------------------------
1711    Name: Matthias Blume (blume (at) tti - c (dot) org)
1712    Date: 2003/09/26 12:00:00 CDT
1713    Tag: blume-20030926-ppautoload
1714    Description:
1715    
1716    I modified the read-eval-print loop so that the autoloader gets
1717    invoked whenever the prettyprinter tries to look up a symbol that
1718    is not currently defined in the toplevel environment but which
1719    appears in CM's autoload registry.  As a result, we see far fewer of
1720    those ?.Foo.Bar.xxx names in the prettyprinter's output.
1721    
1722    In addition to this I tried to clean up some pieces of the Basis
1723    implementation (e.g., Socket, Word8Array) in order to prevent other
1724    instances of these ?.Foo.Bar.xxx names from being printed.
1725    
1726    The mechanism that picks names for types still needs some work, though.
1727    (Right now it seems that if there is a type A.t which is defined to
1728    be B.u, but B is unavailable at toplevel, then A.t gets printed as
1729    "?.B.u" although the perhaps more sensible solution would be to use
1730    "A.t" in this case.  In other words, the prettyprinter should follow
1731    a chain of DEFtycs not farther than there are corresponding toplevel
1732    names in the current environment.)
1733    
1734    ----------------------------------------------------------------------
1735    Name: Matthias Blume (blume (at) tti - c (dot) org)
1736    Date: 2003/09/24 16:31:00 CDT
1737    Tag: blume-20030924-installer
1738    Description:
1739    
1740    Another installer tweak:  All the ML code for the installer is now
1741    compiled during CMB.make and put into a little library called
1742    $smlnj/installer.cm.  The installation then simply invokes
1743    
1744       sml -m $smlnj/installer.cm
1745    
1746    and everything happens automagically.
1747    
1748    Win32:  ML code senses value of environment variable SMLNJ_HOME.
1749    Unix:   ML code senses values of environment variables ROOT, CONFIGDIR,
1750            and BINDIR.
1751    
1752    The new scheme guarantees that the ML code responsible for the installation
1753    is in sync with the APIs of the main system.  Also, the installer is
1754    somewhat faster because the installer script is precompiled.
1755    
1756    ----------------------------------------------------------------------
1757    Name: Matthias Blume (blume (at) tti - c (dot) org)
1758    Date: 2003/09/24 15:35:00 CDT
1759    Tag: blume-20030924-synsock
1760    Description:
1761    
1762    Added a signature SYNCHRONOUS_SOCKET to basis.cm.  This is like SOCKET
1763    but excludes all non-blocking operations.  Defined SOCKET (in Basis)
1764    and CML_SOCKET in terms of SYNCHRONOUS_SOCKET.  Removed superfluous
1765    implementations of non-blocking operations from CML's Socket
1766    structure.
1767    
1768    ----------------------------------------------------------------------
1769    Name: Matthias Blume (blume (at) tti - c (dot) org)
1770    Date: 2003/09/24 15:10:05 CDT
1771    Tag: blume-20030924-sockets
1772    Description:
1773    
1774    1. Fixed SOCKET API and implementation to match Basis spec.
1775       This required changing the internal representation of sockets to one
1776       that remembers (for each socket file descriptor) whether it is currently
1777       blocking or non-blocking.  This state is maintained lazily (i.e., a system
1778       call is made only if the state actually needs to change).
1779    
1780    2. OS-specific details of sockets were moved into separate files, thus
1781       making it possible to unify the bulk of the socket implementations
1782       between Unix and Win32.
1783    
1784    3. CML's socket API changed accordingly.
1785       (Note that we need to remove non-blocking functions from this API
1786       since they are redundant in the case of CML!)
1787    
1788    4. CML's socket implementation now makes use of non-blocking functions
1789       provided by Basis, thus removing all OS-dependent code from this part
1790       of CML.
1791    
1792    5. Changed Real64.precision from 52 to 53.  Minor cleanup in Real64 code.
1793    
1794    ----------------------------------------------------------------------
1795    Name: Matthias Blume (blume (at) tti - c (dot) org)
1796    Date: 2003/09/22 12:10:00 CDT
1797    Tag: blume-20030922-110_43_2
1798    Description:
1799    
1800    Made a new interim version and bootfiles for developer's bootstrapping
1801    convenience.
1802    
1803       110.43.2 -- NEW BOOTFILES
1804    
1805    ----------------------------------------------------------------------
1806    Name: Matthias Blume (blume (at) tti - c (dot) org)
1807    Date: 2003/09/19 15:55:00 CDT
1808    Tag: blume-20030919-cmdir
1809    Description:
1810    
1811    1. new-install.sh -> install.sh
1812    2. changed default CM "metadata" directory name to ".cm" (instead of "CM")
1813    3. tweaked installer so that another name instead of .cm can be chosen
1814       at install time (by setting the CM_DIR_ARC environment variable
1815       during installation); once installation is complete, the name is
1816       fixed
1817    
1818    ----------------------------------------------------------------------
1819    Name: Matthias Blume (blume (at) tti - c (dot) org)
1820    Date: 2003/09/18 16:00:00 CDT
1821    Tag: blume-20030918-110_43_1
1822    Description:
1823    
1824    Made a new interim version and bootfiles for developer's bootstrapping
1825    convenience.
1826    
1827       110.43.1 -- NEW BOOTFILES
1828    
1829    ----------------------------------------------------------------------
1830    Name: Matthias Blume (blume (at) tti - c (dot) org)
1831    Date: 2003/09/18 15:20:00 CDT
1832    Tag: blume-20030918-misc
1833    Description:
1834    
1835    1. Exported fractionsPerSecond etc. from TimeImp (but not from Time as
1836       this seems to be controversial at the moment) and used those in
1837       Posix.ProcEnv.times.
1838    
1839    2. Added Time.{from,to}Nanoseconds to Time.
1840    
1841    3. Improved Real.{from,to}LargeInt by avoiding needless calculations.
1842       For example, fromLargeInt never needs to look at more than 3 "big
1843       digits" to get its 53 bits of precision.
1844    
1845    ----------------------------------------------------------------------
1846    Name: Matthias Blume (blume (at) tti - c (dot) org)
1847    Date: 2003/09/17 16:30:00 CDT
1848    Tag: blume-20030917-real32-slices
1849    Description:
1850    
1851    Added an entry to the primitive environment
1852    (compiler/Semant/statenv/prim.sml) for int32->real64 conversion and
1853    added code to compiler/CodeGen/main/mlriscGen.sml to implement it.
1854    
1855    Removed some of the "magic" constants in real64.sml and replaced them
1856    with code that generates these values from their corresponding
1857    integer counterparts.
1858    
1859    Made all(?) the slice-related changes to the Basis and made everything
1860    compile again...
1861    
1862    ----------------------------------------------------------------------
1863    Name: Matthias Blume (blume (at) tti - c (dot) org)
1864    Date: 2003/09/15 17:45:00 CDT
1865    Tag: blume-20030915-rbase
1866    Description:
1867    
1868    Fixed bug in Real.fromLargeInt.
1869    
1870    ----------------------------------------------------------------------
1871    Name: Matthias Blume (blume (at) tti - c (dot) org)
1872    Date: 2003/09/13 18:11:00 CDT
1873    Tag: blume-20030913-libinstall
1874    Description:
1875    
1876    Minor bugfix in config/libinstall (set anchor with path to
1877    standalone tool after installing it, otherwise libraries that
1878    need ml-lex or ml-yacc won't compile the first time the installer
1879    runs).
1880    
1881    ----------------------------------------------------------------------
1882    Name: Matthias Blume (blume (at) tti - c (dot) org)
1883    Date: 2003/09/12 11:45:00 CDT
1884    Tag: blume-20030912-various
1885    Description:
1886    
1887     - fixed bug in Real.toLargeInt
1888     - fixed bug in Posix.ProcEnv.times
1889     - changed inputLine functions to return an option
1890     - minor installer improvements / bugfixes
1891     - changed default @SMLalloc parameter for x86/celeron to 64k
1892    
1893    ----------------------------------------------------------------------
1894    Name: Matthias Blume (blume (at) tti - c (dot) org)
1895    Date: 2003/09/09 22:00:00 CDT
1896    Tag: Release_110_43
1897    Description:
1898    
1899        New working release 110.43.  New bootfiles.
1900    
1901    ----------------------------------------------------------------------
1902    Name: Matthias Blume (blume (at) tti - c (dot) org)
1903    Date: 2003/09/09 19:20:00 CDT
1904    Tag: blume-20030909-installer
1905    Description:
1906    
1907    Rewrote large parts of config/install.sh in SML (config/libinstall.sml).
1908    Modified config/install.bat to take advantage of it.  Also modified
1909    config/install.sh (and called it config/new-install.sh) to take advantage
1910    of it on Unix systems.  (The SML code is (supposed to be) platform-
1911    independent.)
1912    
1913    The installer can now install everything under Win32
1914    as well as under *nix as long as it compiles.
1915    
1916    Other changes:
1917    
1918      - made CML compile again under Win32
1919      - made eXene compile under Win32 (by providing a fake structure UnixSock
1920        and by using OS.Process.getEnv instead of Posix.ProcEnv.getenv)
1921      - fixed a bug in nowhere: it assumed that type OS.Process.status is the
1922        same as type int;  under Win32 it isn't
1923      - fixed some slice-related problems in the win32-specific parts of CML
1924      - added a functor argument "sameVol" to os-path-fn.sml in the Basis
1925        (under Win32, the volume name is case-insensitive, and the
1926        OS.Path code compares volume names for equality)
1927    
1928    ----------------------------------------------------------------------
1929    Name: Matthias Blume (blume (at) tti - c (dot) org)
1930    Date: 2003/09/08 11:55:00 CDT
1931    Tag: blume-20030908-fullpath
1932    Description:
1933    
1934    Made Win32 version of OS.FileSys.fullPath return current directory
1935    when given an empty string.  This is what the spec says, and incidentally,
1936    CM depends on it.  (CM otherwise goes into an infinite loop in certain
1937    cases when presented with the name of a non-existing .cm file.)
1938    
1939    ----------------------------------------------------------------------
1940    Name: Matthias Blume (blume (at) tti - c (dot) org)
1941    Date: 2003/09/04 16:30:00 CDT
1942    Tag: blume-20030905-slices-etc
1943    Description:
1944    
1945      1. Changed interface to vectors and arrays in Basis to match
1946         (draft) Basis spec.
1947      2. Added signatures and implementations of slices according to
1948         Basis spec.
1949      3. Edited source code throughout the system to make it compile again
1950         under 1. and 2.  (In some cases code had to be added to have it
1951         match the new signatures.)
1952      4. MLRISC should be backward-compatible: the copies of the originals
1953         of files that needed to change under 3. were retained, the .cm files
1954         check the compiler version number and use old versions when
1955         appropriate.
1956      5. Changed type of OS.FileSys.readDir and Posix.FileSys.readdir to
1957         dirstream -> string option (in accordance with Basis spec).
1958      6. When generating code that counts lines, ml-lex used function
1959         CharVector.foldli, taking advantage of its old interface.
1960         This has been replaced with the corresponding code from
1961         CharVectorSlice.  (html-lex must be re-lexed!)
1962      7. BitArray in smlnj-lib/Util has been extended/modified to match the
1963         new MONO_ARRAY signature.  (Do we need BitArraySlice?)
1964      8. Removed temporary additions (fromInternal, toInternal) from the
1965         (now obsolete) IntInf in smlnj-lib/Util.
1966      9. Cleaned up structure Byte.
1967     10. Added localOffset, scan, and fromString to Date (according to spec).
1968         Cleaned/corrected implementation of Date.
1969         (Still need to check for correctness; implement better canonicalizeDate.)
1970     11. Added "scan" to signature IEEE_REAL.
1971     12. Some improvements to IntInf [in particular: efficiency-hack for
1972         mod and rem when second operand is 2 (for parity checks).]
1973     13. Changed representation of type Time.time, using a single IntInf.int
1974         value counting microseconds.  This considerably simplified the
1975         implementation of structure Time.  We now support negative time
1976         values; scan and fromString handle signs.
1977     14. Functor PrimIO now takes two additional arguments (VectorSlice and
1978         ArraySlice).
1979    
1980    ----------------------------------------------------------------------
1981    Name: Matthias Blume (blume (at) tti - c (dot) org)
1982    Date: 2003/08/28 17:00:00 CDT
1983    Tag: blume-20030828-intinf
1984    Description:
1985    
1986    This is a major update which comes with a version number bump
1987    (110.42.99 -- yes, we are really close to 110.43 :-), NEW BOOTFILES,
1988    and an implementation of IntInf in the Basis.
1989    
1990    There are a fairly large number of related changes and updates throughout
1991    the system:
1992    
1993    Basis:
1994            - Implemented IntInf.
1995            - Made LargeInt a projection of IntInf (by filtering through INTEGER).
1996            - Added some missing Real64 operations, most notably Real.toLargeInt.
1997            - Added FixedInt as a synonym for Int32.
1998    
1999    compiler:
2000            * Added support for a built-in intinf type.
2001              - literals
2002              - pattern matching
2003              - conversion shortcuts (Int32.fromLarge o Int.toLarge etc.)
2004              - overloading on literals and operations
2005    
2006            This required adding a primitive type intinf, some additional
2007            primops, and implementations for several non-trivial intinf
2008            operations in Core.  (The intinf type is completely abstract
2009            to the compiler; all operations get delegated back to the Core.)
2010    
2011            * Intinf equality is handled by polyequal.  However, the compiler
2012            does not print its usual warning in this case (since polyequal
2013            is the right thing to do there).
2014    
2015            * Improved the organization of structure InlineT.
2016    
2017            * A word about conversion primops:
2018            If conversions involving intinf do not cancel out during
2019            CPS contract, then the compiler must insert calls to Core functions.
2020            Since all core access must be resolved already during the FLINT
2021            translate phase, it would be too late a the time of CPS contract
2022            to add new Core calls.  For this reason, conversion primops
2023            for intinf carry two arguments:  1. the numeric argument that
2024            they are supposed to convert, and 2. the Core function that
2025            can help with this conversion if necessary.  If CPS contract
2026            eliminates a primop, then the associated Core function becomes
2027            dead and goes away.  Intinf conversion primops that do not get
2028            eliminated by CPS contract get rewritten into calls of their
2029            core functions by a separate, new phase.
2030    
2031    interactive system:
2032            - Control.Print.intinfDepth controls max length of intinf constants
2033            being printed. (Analogous to Control.Print.stringDepth.)
2034            - Cleanup in printutil and pputil: got rid of unused stuff and
2035            duplicates; replaced some of the code with code that makes better
2036            use of library functionality.
2037    
2038    CM:
2039            Bugfix: parse-errors in init group (system/smlnj/init/init.cmi)
2040            are no longer silent.
2041    
2042    CKIT:
2043            Fixed mismatched uses of Int32 and LargeInt.  I always decided
2044            in favor of LargeInt -- which is now the same as IntInf.
2045            CKIT-knowledgable people should check whether this is what's
2046            intended and otherwise change things back to using Int32 or
2047            FixedInt.
2048    
2049    Throughout the code:
2050            Started using IntInf.int literals and built-in operations
2051            (e.g., comparison with 0) where this seems appropriate.
2052    
2053    
2054    ----------------------------------------------------------------------
2055    Name: Dave MacQueen (dbm@cs.uchicago.edu)
2056    Date: 2003/08/13 11:36:00 CDT
2057    Tag: dbm-20030813-mcz-merge1
2058    Description:
2059    
2060    Merging changes from the mcz-branch development branch into trunk.
2061    These changes involve replacement of the emulated old prettyprinter
2062    interface with direct use of the SML/NJ Lib PP library, and fixing
2063    of a couple of bugs (895, 1186) relating to error messages.  A new
2064    prettyprinter for ast datatypes (Elaborator/print/ppast.{sig,sml})
2065    has been added.
2066    
2067    ----------------------------------------------------------------------
2068    Name: Matthias Blume (blume (at) tti - c (dot) org)
2069    Date: 2003/08/11 15:45:00 CDT
2070    Tag: blume-20030811-windows
2071    Description:
2072    
2073    Version number bumped to 110.42.9.  NEW BOOTFILES!!!
2074    
2075        http://smlnj.cs.uchicago.edu/dist/working/110.42.9/
2076    
2077    This patch restores SML/NJ's ability to run under win32.  There are a
2078    number of changes, including fixes for several bugs that had gone
2079    unnoticed until now:
2080    
2081      - uname "CYGWIN_NT*" is recognized as win32 (This is relevant only when
2082        trying to run the win32 version from within cygwin.)
2083    
2084      - There are a number of simple .bat scripts that substitute for their
2085        corresponding Unix shell-scripts.  (See below.)
2086    
2087      - The internals of ml-build have been modified slightly.  The main
2088        difference is that instead of calling ".link-sml" (or link-sml.bat)
2089        using OS.Process.system, the ML process delegates this task back
2090        to the script.  Otherwise problems arise in mixed environments such
2091        as Cygwin where scripts look and work like Unix scripts, but
2092        where OS.Process.system cannot run them.
2093    
2094      - In CM, the srcpath pickler used native pathname syntax -- which
2095        is incorrect in the case of cross-compilation.  The new pickle format
2096        is independent of platform-specific naming conventions.
2097    
2098      - Path configuration files (such as lib/pathconfig) can now choose
2099        between native and standard syntax.  Placing a line of the form
2100    
2101            standard!
2102    
2103        into the file causes all subsequent paths to be interpreted using
2104        CM standard pathname syntax (= Unix conventions); a line
2105    
2106            native!
2107    
2108        switches back to native style.  This was needed so that
2109        path config files can be written portably, see src/system/pathconfig.
2110    
2111      - Runtime system:
2112    
2113          - win32-filesys.c:  get_file_time and set_file_time now
2114            access modification time, not creation time.
2115    
2116          - I/O code made aware of new array representation.
2117    
2118          - Bug fixes in X86.prim.masm.
2119    
2120       - src/system/makeml made aware of win32. (For use under cygwin
2121            and other Unix-environments for windows.)
2122    
2123       - In Basis, fixed off-by-one error in win32-io.sml (function vecF)
2124         which caused BinIO.inputAll to fail consistently.
2125    
2126    .bat scripts:
2127    
2128       Windows .bat scripts assume that SMLNJ_HOME is defined.
2129    
2130       - sml.bat, ml-yacc.bat, ml-lex.bat:  Driver scripts for standalone
2131         applications (sml, ml-yacc, ml-lex).
2132       - ml-build.bat: analogous to ml-build.
2133       - config\install.bat:  Analogous to config/install.sh.  This requires
2134         that SMLNJ_HOME is set and that Microsoft Visual C is ready to use.
2135         (nmake etc. must be on the path, and vcvars32 must have been run.)
2136         Moreover, sources for ml-lex and ml-yacc need to exist under src,
2137         and the bootfile hierarchy must have been unpacked under
2138         sml.boot.x86-win32.
2139         The script is very primitive and does a poor job at error checking.
2140         It only installs the base system, ml-lex, and ml-yacc.  No other
2141         libraries are being installed (i.e., you get only those that
2142         are part of the compiler.)
2143       - link-sml.bat:  analogous to .link-sml, but not currently used
2144    
2145    Unrelated bug fixes:
2146    
2147      - ml-nlffigen now exports structures ST_* corresponding to incomplete
2148        types.
2149      - Added getDevice to PP/src/pp-debug-fn.sml.  (Would not compile
2150        otherwise.)
2151    
2152    ----------------------------------------------------------------------
2153    Name: Dave MacQueen (macqueen@cs.uchicago.edu)
2154    Date: 2003/06/17
2155    Tag: macqueen-20030617-bug895
2156    Description:
2157    
2158    Modified compiler/Elaborator/print/pptype.sml to fix bug 895.
2159    Tag will be used for new development branch (mcz-branch) for
2160    use by MacQueen, (Lucasz) Zairek, and (George) Cao at uchicago.
2161    
2162    ----------------------------------------------------------------------
2163    Name: Matthias Blume (blume (at) tti - c (dot) org)
2164    Date: 2003/05/27 16:55:00 CDT
2165    Tag: blume-20030527-polyeq
2166    Description:
2167    
2168    Tried to eliminated most cases of polymorphic equality.
2169    
2170    ----------------------------------------------------------------------
2171    Name: Matthias Blume (blume (at) tti - c (dot) org)
2172    Date: 2003/05/21 17:45:00 CDT
2173    Tag: blume-20030517-complete
2174    Description:
2175    
2176    Two changes:
2177    
2178      1. Added a flag for controlling whether non-exhaustive bindings will
2179         be treated as errors (default is false).
2180      2. Cleaned up the *entire* source tree so that CMB.make goes through
2181         without a single non-exhaustive match- or bind warning.
2182    
2183    ----------------------------------------------------------------------
2184    Name: Matthias Blume (blume (at) tti - c (dot) org)
2185    Date: 2003/05/17 10:20:00 CDT
2186    Tag: blume-20030517-absyn
2187    Description:
2188    
2189    1. Added cases for IF, WHILE, ANDALSO, and ORELSE to Absyn.
2190    
2191       This mainly affects the quality of error messages.  However, some
2192       of the code is now more straightforward than before. (Treatment of
2193       the above four constructs in translate.sml is much simpler than
2194       the "macro-expansion" that was going on before.  Plus, the mach-
2195       compiler no longer gets invoked just to be able to compile an
2196       if-expression.)
2197    
2198    2. The ErrorMsg.Error exception is now caught and absorbed by the
2199       interactive loop.
2200    
2201    ----------------------------------------------------------------------
2202    Name: Allen Leung
2203    Date: 2003/05/16 13:05:00 CDT
2204    Tag: leunga-20030516-cygwin-runtime
2205    Description:
2206    
2207        Ported the runtime system to cygwin, which uses the unix
2208    x86-unix bin files.  Missing/buggy features:
2209    
2210       o getnetbyname, getnetbyaddr: these functions seem to be missing in
2211         the Cygwin library.
2212       o Ctrl-C handling may be flaky.
2213       o Windows system calls and Windows I/O are not supported.
2214    
2215       A new set of binfiles is located at:
2216    
2217         http://www.dorsai.org/~leunga/boot.x86-unix.tgz
2218    
2219       This is only needed for bootstrapping the cygwin version of smlnj.
2220       Other x86 versions can use the existing binfiles.
2221    
2222    ----------------------------------------------------------------------
2223    Name: Matthias Blume
2224    Date: 2003/04/08 15:42:00 CDT
2225    Tag: blume-20030408-listpair
2226    Description:
2227    
2228    1. Added a target 'mlrisc' to installer.
2229    
2230    2. Added missing elements to structure ListPair.
2231    
2232    ----------------------------------------------------------------------
2233    Name: Allen Leung
2234    Date: 2003/01/07 10:40:00 EST
2235    Tag: leunga-20030107-int-rem
2236    Description:
2237    
2238    Fixed a bug in Int.rem(x,y) where y is a power of 2 on x86.
2239    The arguments to the SUBL instruction were swapped.
2240    
2241    ----------------------------------------------------------------------
2242    Name: Matthias Blume
2243    Date: 2002/12/12 16:25:00 EST
2244    Tag: blume-20021212-risc-ra
2245    Description:
2246    
2247    Fixed a serious bug in the rewrite code for FP spilling/reloading that
2248    sent the RA into an infinite loop when floating point registers get
2249    spilled.  (Because of this bug, e.g., nucleic stopped compiling between
2250    110.37 and 110.38.)
2251    There was another set of potential problems related to the handling of
2252    MLRISC annotations (but those did not yet cause real problems, apparently).
2253    
2254    ----------------------------------------------------------------------
2255    Name: Matthias Blume
2256    Date: 2002/12/06 22:40:00 EST
2257    Tag: blume-20021206-cm-fileid
2258    Description:
2259    
2260    Added a call of SrcPath.sync at the beginning of Parse.parse (in CM).
2261    This fixes the problem of CM getting confused by files that suddenly
2262    change their identity (e.g., by getting unlinked and recreated by some
2263    text editor such as vi).  There might be a better/cheaper/cleaner way
2264    of doing this, but for now this will have to do.
2265    
2266    ----------------------------------------------------------------------
2267    Name: Matthias Blume
2268    Date: 2002/10/28 09:50:00 EST
2269    Tag: blume-20021028-typecheck
2270    Description:
2271    
2272    Exported structure Typecheck from $smlnj/viscomp/core.cm.
2273    
2274    ----------------------------------------------------------------------
2275    Name: Matthias Blume
2276    Date: 2002/10/17 09:10:00 EDT
2277    Tag: Release_110_42
2278    Description:
2279    
2280    In good old tradition, there has been a slight hiccup so that we have
2281    to patch 110.42 after the fact.   The old release tag has been replaced
2282    (see below).
2283    
2284    The change solves a problem with two competing approaches the
2285    configuration problem regarding MacOS 10.1 vs. MacOS 10.2 which got in
2286    each other's way.
2287    
2288    This change only affects the runtime system code and the installer script.
2289    (No new bootfiles.)
2290    
2291    ----------------------------------------------------------------------
2292    Name: Matthias Blume
2293    Date: 2002/10/16 12:00:00 EDT
2294    Tag: Release_110_42_removed
2295    Description:
2296    
2297    New working release.  New bootfiles.
2298    
2299    ----------------------------------------------------------------------
2300    Name: Matthias Blume
2301    Date: 2002/10/10 13:10:00 EDT
2302    Tag: blume-20021010-ppc-divs
2303    Description:
2304    
2305    The mltree operator DIVS must be implemented with an overflow check on
2306    the PPC because the hardware indicates divide-by-zero using "overflow" as
2307    well.
2308    
2309    ----------------------------------------------------------------------
2310    Name: Matthias Blume
2311    Date: 2002/07/23 11:45:00
2312    Tag: blume-20020723-smlnj-home
2313    Description:
2314    
2315    Sml now senses the SMLNJ_HOME environment variable.  If this is set,
2316    then the bin dir is assumed to be in $SMLNJ_HOME/bin and (unless
2317    CM_PATHCONFIG is also set), the path configuration file is assumed
2318    to be in $SMLNJ_HOME/lib/pathconfig.  This way one can easily move
2319    the entire tree to some other place and everything will "just work".
2320    
2321    (Companion commands such as ml-build and ml-makedepend also sense this
2322    variable.)
2323    
2324    ----------------------------------------------------------------------
2325    Name: Matthias Blume
2326    Date: 2002/07/12 21:19:00 EDT
2327    Tag: blume-20020712-liveness
2328    Description:
2329    
2330    Exported two useful "step" functions from liveness module (MLRISC).
2331    
2332    ----------------------------------------------------------------------
2333    Name: Matthias Blume
2334    Date: 2002/07/05 16:00 EDT
2335    Tag: Release_110_41
2336    Description:
2337    
2338    New working release.  New bootfiles.
2339    
2340    ----------------------------------------------------------------------
2341    Name: Matthias Blume
2342    Date: 2002/07/05 10:25:00 EDT
2343    Tag: blume-20020705-btimp
2344    Description:
2345    
2346    Exported structure BTImp from $smlnj/viscomp/debugprof.cm so that
2347    other clients can set up backtracing support.
2348    
2349    ----------------------------------------------------------------------
2350    Name: Matthias Blume
2351    Date: 2002/06/25 17:23:00 EDT
2352    Tag: blume-20020625-fpmax
2353    Description:
2354    
2355    Fixed a bug in translation of INLMAX (and INLMIN) for the floating-point
2356    case.  (The sense of the isNaN test was reversed -- which made min and
2357    max always return their first argument.)
2358    
2359    ----------------------------------------------------------------------
2360    Name: Matthias Blume
2361    Date: 2002/06/11
2362    Tag: blume-20020611-unixpath
2363    Description:
2364    
2365    Back-ported OS.Path.{from,to}UnixPath from idlbasis-devel branch.
2366    
2367    ----------------------------------------------------------------------
2368    Name: Matthias Blume
2369    Date: 2002/06/10 16:35:00 EDT
2370    Tag: blume-20020610-ieeereal
2371    Description:
2372    
2373    I back-ported my implementation of IEEEReal.fromString from the
2374    idlbasis-devel branch so that we can test it.
2375    
2376    Another small change is that ppDec tries to give more information
2377    than just "<sig>" in the case of functors.  However, this code is
2378    broken in some mysterious way if the functor's body's signature
2379    has not been declared by ascription but gets inferred from the
2380    implementation.  This needs fixing...
2381    
2382    ----------------------------------------------------------------------
2383    Name: Matthias Blume
2384    Date: 2002/05/31
2385    Tag: blume-20020531-btrace-mode
2386    Description:
2387    
2388    Resurrected SMLofNJ.Internals.BTrace.mode.  (It accidentally fell by
2389    the wayside when I switched over to using Controls everywhere.)
2390    
2391    ----------------------------------------------------------------------
2392    Name: Lal George
2393    Date: 2002/05/23 12:21:40 EDT
2394    Tag: george-20020523-visual-labels
2395    Description:
2396    
2397    Labels are now displayed in the graphical output to make
2398    the fall-through and target blocks obvious.
2399    
2400    ----------------------------------------------------------------------
2401    Name: Matthias Blume
2402    Date: 2002/05/22 11:03:00 EDT
2403    Tag: blume-20020522-shrink
2404    Description:
2405    
2406    John tweaked yesterday's fix for 1131 to handle an out-of-memory
2407    situation that comes up when allocating huge arrays.
2408    
2409    ----------------------------------------------------------------------
2410    Name: Matthias Blume
2411    Date: 2002/05/21 16:00:00 EDT
2412    Tag: Release_110_40
2413    Description:
2414    
2415    New working release (110.40).  New bootfiles.
2416    
2417    [Also: John Reppy fixed GC bug 1131.]
2418    
2419    ----------------------------------------------------------------------
2420    Name: Matthias Blume
2421    Date: 2002/05/21 12:35:00 EDT
2422    Tag: blume-20020521-cmdoc
2423    Description:
2424    
2425    CM documentation update.
2426    
2427    ----------------------------------------------------------------------
2428    Name: Matthias Blume
2429    Date: 2002/05/21 10:55:00 EDT
2430    Tag: blume-20020521-misc
2431    Description:
2432    
2433    - John tweaked runtime to be silent on heap export (except when
2434      GC messages are on).
2435    - I added a few more things (cross-compiling versions of CMB) to
2436      config/preloads (as suggestions).
2437    
2438    ----------------------------------------------------------------------
2439    Name: Matthias Blume
2440    Date: 2002/05/20 22:25:00 EDT
2441    Tag: blume-20020520-controls
2442    Description:
2443    
2444    - Added ControlUtil structure to control-lib.cm.
2445    - Use it throughout.
2446    - Used Controls facility to define MLRISC controls (as opposed to
2447      registering MLRISC control ref cells with Controls after the fact)
2448    - Fixed messed-up controls priorities.
2449    
2450    * Removed again all the stuff from config/preloads that one wouldn't
2451      be able to preload at the time the initial heap image is built.
2452      (Many libraries, e.g., CML, do not exist yet at this time.  The
2453       only libraries that can be preloaded via config/preloads are those
2454       that come bundled with the bootfiles.)
2455    
2456    ----------------------------------------------------------------------
2457    Name: Matthias Blume
2458    Date: 2002/05/20 10:59:00 EDT
2459    Tag: blume-20020520-preloads
2460    Description:
2461    
2462    Added a lot of commented-out suggestions for things to be included
2463    in config/preloads.
2464    
2465    ----------------------------------------------------------------------
2466    Name: Allen Leung
2467    Date: 2002/05/18 14:20:00 EDT
2468    Tag: leunga-20020518-mdl
2469    Description:
2470    
2471    o Made the mdl tool stuff compile and run again.
2472    o I've disabled all the stuff that depends on RTL specifications; they
2473      are all badly broken anyway.
2474    
2475    ----------------------------------------------------------------------
2476    Name: Matthias Blume
2477    Date: 2002/05/17 16:49:00 EDT
2478    Tag: blume-20020517-controls
2479    Description:
2480    
2481    0. John Reppy made several modifications to the SML/NJ library.
2482       In particular, there is a shiny new controls-lib.cm.
2483    
2484    1. Pushed new controls interface through compiler so that everything
2485       compiles again.
2486    
2487    2. Added FormatComb and FORMAT_COMB to the CML version of the
2488       SML/NJ library (so that CML compiles again).
2489    
2490    3. Modified init scripts because XXX_DEFAULT environment variables
2491       are no longer with us.  (Boot-time initialization is now done
2492       using the same environment variables that are also used for
2493       startup-time initialization of controls.)
2494    
2495    ----------------------------------------------------------------------
2496    Name: Lal George
2497    Date: 2002/05/15 09:20:10 EDT
2498    Tag: george-20020515-pseudo-op-decls
2499    Description:
2500    
2501            All pseudo-ops emitted before the first segment declaration
2502    such as TEXT, DATA, and BSS directives are assumed to be global
2503    declarations and are emitted first in the assembly file. This is
2504    useful in a number of situations where one has pseudo-ops that are not
2505    specific to any segment, and also works around the constraint that one
2506    cannot have client pseudo-ops in the TEXT segment.
2507    
2508            Because no segment is associated with these declarations it is
2509    an error to allocate any space or objects before the first segment
2510    directive and an exception will be raised. However, we cannot make
2511    this check for client pseudo-ops.
2512    
2513            These top level declarations are a field in the CFG graph_info.
2514    In theory you can continue to add to this field after the CFG has been
2515    built -- provided you know what you are doing;-)
2516    
2517    ----------------------------------------------------------------------
2518    Name: Matthias Blume
2519    Date: 2002/05/13 16:40:00 EDT
2520    Tag: blume-20020513-pp-etc
2521    Description:
2522    
2523    A few minor bugfixes:
2524    
2525      - Stopgap measure for bug recently reported by Elsa Gunter (ppDec).
2526        (Bogus printouts for redefined bindings still occur.  Compiler
2527        bug should no longer occur now.  We need to redo the prettyprinter
2528        from scratch.)
2529    
2530      - CM pathname printer now also adds escape sequences for ( and )
2531    
2532      - commend and docu fixes for ml-nlffi
2533    
2534    ----------------------------------------------------------------------
2535    Name: Matthias Blume
2536    Date: 2002/05/10 16:40:00 EDT
2537    Tag: blume-20020510-erg-textio
2538    Description:
2539    
2540    Applied the following bugfix provided by Emden Gansner:
2541    
2542        Output is corrupted when outputSubstr is used rather than output.
2543    
2544        The problem occurs when a substring
2545    
2546            ss = (s, dataStart, dataLen)
2547    
2548        where dataStart > 0, fills a stream buffer with avail bytes left.
2549        avail bytes of s, starting at index dataStart, are copied into the
2550        buffer, the buffer is flushed, and then the remaining dataLen-avail
2551        bytes of ss are copied into the beginning of the buffer. Instead of
2552        starting this copy at index dataStart+avail in s, the current code
2553        starts the copy at index avail.
2554    
2555        Fix:
2556        In text-io-fn.sml, change line 695 from
2557             val needsFlush = copyVec(v, avail, dataLen-avail, buf, 0)
2558        to
2559             val needsFlush = copyVec(v, dataStart+avail, dataLen-avail, buf, 0)
2560    
2561    ----------------------------------------------------------------------
2562    Name: Matthias Blume
2563    Date: 2002/04/12 13:55:00 EDT
2564    Tag: blume-20020412-assyntax
2565    Description:
2566    
2567    1. Grabbed newer assyntax.h from the XFree86 project.
2568    2. Fiddled with how to compile X86.prim.asm without warnings.
2569    3. (Very) Minor cleanup in CM.
2570    
2571    ----------------------------------------------------------------------
2572    Name: Matthias Blume
2573    Date: 2002/04/01 (no joke!) 17:07:00 EST
2574    Tag: blume-20020401-x86div
2575    Description:
2576    
2577    Added full support for div/mod/rem/quot on the x86, using the machine
2578    instruction's two results (without clumsily recomputing the remainder)
2579    directly where appropriate.
2580    
2581    Some more extensive power-of-two support was added to the x86 instruction
2582    selector (avoiding expensive divs, mods, and muls where they can be
2583    replaced with cheaper shifts and masks).  However, this sort of thing
2584    ought to be done earlier, e.g., within the CPS optimizer so that
2585    all architectures benefit from it.
2586    
2587    The compiler compiles to a fixed point, but changes might be somewhat
2588    fragile nevertheless.  Please, report any strange things that you might
2589    see wrt. div/mod/quot/rem...
2590    
2591    ----------------------------------------------------------------------
2592    Name: Matthias Blume
2593    Date: 2002/03/29 17:22:00
2594    Tag: blume-20020329-div
2595    Description:
2596    
2597    Fixed my broken div/mod logic.  Unfortunately, this means that the
2598    inline code for div/mod now has one more comparison than before.
2599    Fast paths (quotient > 0 or remainder = 0) are not affected, though.
2600    The problem was with quotient = 0, because that alone does not tell
2601    us which way the rounding went.  One then has to look at whether
2602    remainder and divisor have the same sign...  :(
2603    
2604    Anyway, I replaced the bootfiles with fresh ones...
2605    
2606    ----------------------------------------------------------------------
2607    Name: Matthias Blume
2608    Date: 2002/03/29 14:10:00 EST
2609    Tag: blume-20020329-inlprims
2610    Description:
2611    
2612    NEW BOOTFILES!!!    Version number bumped to 110.39.3.
2613    
2614    Primops have changed. This means that the bin/boot-file formats have
2615    changed as well.
2616    
2617    To make sure that there is no confusion, I made a new version.
2618    
2619    
2620    CHANGES:
2621    
2622    * removed REMT from mltree (remainder should never overflow).
2623    
2624    * added primops to deal with divisions of all flavors to the frontend
2625    
2626    * handled these primops all the way through so they map to their respective
2627      MLRISC support
2628    
2629    * used these primops in the implementation of Int, Int32, Word, Word32
2630    
2631    * removed INLDIV, INLMOD, and INLREM as they are no longer necessary
2632    
2633    * parameterized INLMIN, INLMAX, and INLABS by a numkind
2634    
2635    * translate.sml now deals with all flavors of INL{MIN,MAX,ABS}, including
2636      floating point
2637    
2638    * used INL{MIN,MAX,ABS} in the implementation of Int, Int32, Word, Word32,
2639      and Real (but Real.abs maps to a separate floating-point-only primop)
2640    
2641    
2642    TODO items:
2643    
2644    * Hacked Alpha32 instruction selection, disabling the selection of REMx
2645      instructions because the machine instruction encoder cannot handle
2646      them.  (Hppa, PPC, and Sparc instruction selection did not handle
2647      REM in the first place, and REM is supported by the x86 machine coder.)
2648    
2649    * Handle DIV and MOD with DIV_TO_NEGINF directly in the x86 instruction
2650      selection phase.  (The two can be streamlined because the hardware
2651      delivers both quotient and remainder at the same time anyway.)
2652    
2653    * Think about what to do with "valOf(Int32.minInt) div ~1" and friends.
2654      (Currently the behavior is inconsistent both across architectures and
2655      wrt. the draft Basis spec.)
2656    
2657    * Word8 should eventually be handled natively, too.
2658    
2659    * There seems to be one serious bug in mltree-gen.sml.  It appears, though,
2660      as if there currently is no execution path that could trigger it in
2661      SML/NJ.  (The assumptions underlying functions arith and promotable do not
2662      hold for things like multiplication and division.)
2663    
2664    ----------------------------------------------------------------------
2665    Name: Matthias Blume
2666    Date: 2002/03/27 16:27:00 EST
2667    Tag: blume-20020327-mlrisc-divisions
2668    Description:
2669    
2670    Added support for all four division operations (ML's div, mod, quot,
2671    and rem) to MLRISC.  In the course of doing so, I also rationalized
2672    the naming (no more annoying switch-around of DIV and QUOT), by
2673    parameterizing the operation by div_rounding_mode (which can be either
2674    DIV_TO_ZERO or DIV_TO_NEGINF).
2675    
2676    The generic MLTreeGen functor takes care of compiling all four
2677    operations down to only round-to-zero div.
2678    
2679    Missing pieces:
2680    
2681      * Doing something smarter than relying on MLTreeGen on architectures
2682        like, e.g., the x86 where hardware division delivers both quotient and
2683        remainder at the same time.  With this, the implementation of the
2684        round-to-neginf operations could be further streamlined.
2685    
2686      * Remove inlining support for div/mod/rem from the frontend and replace it
2687        with primops that get carried through to the backend.  Do this for all
2688        int and word types.
2689    
2690    ----------------------------------------------------------------------
2691    Name: Matthias Blume
2692    Date: 2002/03/25 17:25:00 EST
2693    Tag: blume-20020325-divmod
2694    Description:
2695    
2696    I improved (hopefully without breaking them) the implementation of Int.div,
2697    Int.mod, and Int.rem.   For this, the code in translate.sml now takes
2698    advantage of the following observations:
2699    
2700      Let  q = x quot y      r = x rem y
2701           d = x div  y      m = x mod y
2702    
2703    where "quot" is the round-to-zero version of integer division that
2704    hardware usually provides.  Then we have:
2705    
2706         r = x - q * y        where neither the * nor the - will overflow
2707         d = if q >= 0 orelse x = q * y then q else q - 1
2708                              where neither the * nor the - will overflow
2709         m = if q >= 0 orelse r = 0 then r else r + y
2710                              where the + will not overflow
2711    
2712    This results in substantial simplification of the generated code.
2713    The following table shows the number of CFG nodes and edges generated
2714    for
2715            fun f (x, y) = x OPER y
2716            (* with OPER \in div, mod, quot, rem *)
2717    
2718    
2719        OPER | nodes(old) | edges(old) | nodes(new) | edges(new)
2720        --------------------------------------------------------
2721         div |         24 |         39 |         12 |         16
2722         mod |         41 |         71 |         12 |         16
2723        quot |          8 |         10 |          8 |         10
2724         rem |         10 |         14 |          8 |         10
2725    
2726    
2727    ----------------------------------------------------------------------
2728    Name: Matthias Blume
2729    Date: 2002/03/25 22:06:00 EST
2730    Tag: blume-20020325-cprotobug
2731    Description:
2732    
2733    Fixed a bug in cproto (c prototype decoder).
2734    
2735    ----------------------------------------------------------------------
2736    Name: Matthias Blume
2737    Date: 2002/03/25 16:00:00 EST
2738    Tag: blume-20020325-raw-primops
2739    Description:
2740    
2741    I did some cleanup to Allen's new primop code and
2742    replaced yesterday's bootfiles with new ones.
2743    (But they are stored in the same place.)
2744    
2745    ----------------------------------------------------------------------
2746    Name: Matthias Blume
2747    Date: 2002/03/24 22:40:00 EST
2748    Tag: blume-20020324-bootfiles
2749    Description:
2750    
2751    Made the bootfiles that Allen asked for.
2752    
2753    ----------------------------------------------------------------------
2754    Name: Allen Leung
2755    Date: 2002/03/23 15:50:00 EST
2756    Tag: leunga-20020323-flint-cps-rcc-primops
2757    Description:
2758    
2759      1. Changes to FLINT primops:
2760    
2761        (* make a call to a C-function;
2762         * The primop carries C function prototype information and specifies
2763         * which of its (ML-) arguments are floating point. C prototype
2764         * information is for use by the backend, ML information is for
2765         * use by the CPS converter. *)
2766      | RAW_CCALL of { c_proto: CTypes.c_proto,
2767                       ml_args: ccall_type list,
2768                       ml_res_opt: ccall_type option,
2769                       reentrant : bool
2770                     } option
2771       (* Allocate uninitialized storage on the heap.
2772        * The record is meant to hold short-lived C objects, i.e., they
2773        * are not ML pointers.  With the tag, the representation is
2774        * the same as RECORD with tag tag_raw32 (sz=4), or tag_fblock (sz=8)
2775        *)
2776      | RAW_RECORD of {tag:bool,sz:int}
2777      and ccall_type = CCALL_INT32 | CCALL_REAL64 | CCALL_ML_PTR
2778    
2779      2.  These CPS primops are now overloaded:
2780    
2781           rawload of {kind:numkind}
2782           rawstore of {kind:numkind}
2783    
2784          The one argument form is:
2785    
2786             rawload {kind} address
2787    
2788          The two argument form is:
2789    
2790             rawload {kind} [ml object, byte-offset]
2791    
2792      3. RAW_CCALL/RCC now takes two extra arguments:
2793    
2794         a. The first is whether the C call is reentrant, i.e., whether
2795            ML state should be saved and restored.
2796         b. The second argument is a string argument specifying the name of
2797            library and the C function.
2798    
2799         These things are currently not handled in the code generator, yet.
2800    
2801      4. In CProto,
2802    
2803         An encoding type of "bool" means "ml object" and is mapped into
2804         C prototype of PTR.  Note that "bool" is different than "string",
2805         even though "string" is also mapped into PTR, because "bool"
2806         is assigned an CPS type of BOGt, while "string" is assigned INT32t.
2807    
2808      5. Pickler/unpicker
2809    
2810         Changed to handle RAW_RECORD and newest RAW_CCALL
2811    
2812      6. MLRiscGen,
2813    
2814         1. Changed to handle the new rawload/rawstore/rawrecord operators.
2815         2. Code for handling C Calls has been moved to a new module CPSCCalls,
2816            in the file CodeGen/cpscompile/cps-c-calls.sml
2817    
2818      7. Added the conditional move operator
2819    
2820             condmove of branch
2821    
2822         to cps.  Generation of this is still buggy so it is currently
2823         disabled.
2824    
2825    ----------------------------------------------------------------------
2826    Name: Lal George
2827    Date: 2002/03/22 14:18:25 EST
2828    Tag: george-20020322-cps-branch-prob
2829    Description:
2830    
2831    Implemented the Ball-Larus branch prediction-heuristics, and
2832    incorporated graphical viewers for control flow graphs.
2833    
2834    Ball-Larus Heuristics:
2835    ---------------------
2836    See the file compiler/CodeGen/cpscompile/cpsBranchProb.sml.
2837    
2838    By design it uses the Dempster-Shafer theory for combining
2839    probabilities.  For example, in the function:
2840    
2841        fun f(n,acc) = if n = 0 then acc else f(n-1, n*acc)
2842    
2843    the ball-larus heuristics predicts that the n=0 is unlikely
2844    (OH-heuristic), and the 'then' branch is unlikely because of the
2845    RH-heuristic -- giving the 'then' branch an even lower combined
2846    probability using the Dempster-Shafer theory.
2847    
2848    Finally, John Reppy's loop analysis in MLRISC, further lowers the
2849    probability of the 'then' branch because of the loop in the else
2850    branch.
2851    
2852    
2853    Graphical Viewing:
2854    ------------------
2855    I merely plugged in Allen's graphical viewers into the compiler. The
2856    additional code is not much. At the top level, saying:
2857    
2858            Control.MLRISC.getFlag "cfg-graphical-view" := true;
2859    
2860    will display the graphical view of the control flow graph just before
2861    back-patching.  daVinci must be in your path for this to work. If
2862    daVinci is not available, then the default viewer can be changed
2863    using:
2864    
2865            Control.MLRISC.getString "viewer"
2866    
2867    which can be set to "dot" or "vcg" for the corresponding viewers. Of
2868    course, these viewers must be in your path.
2869    
2870    The above will display the compilation unit at the level of clusters,
2871    many of which are small, boring, and un-interesting. Also setting:
2872    
2873            Control.MLRISC.getInt "cfg-graphical-view_size"
2874    
2875    will display clusters that are larger than the value set by the above.
2876    
2877    
2878    ----------------------------------------------------------------------
2879    Name: Matthias Blume
2880    Date: 2002/03/21 22:20:00 EST
2881    Tag: blume-20020321-kmp-bugfix
2882    Description:
2883    
2884    Changed the interface to the KMP routine in PreString and fixed
2885    a minor bug in one place where it was used.
2886    
2887    ----------------------------------------------------------------------
2888    Name: Allen Leung
2889    Date: 2002/03/21 20:30:00 EST
2890    Tag: leunga-20020321-cfg
2891    Description:
2892    
2893      Fixed a potential problem in cfg edge splitting.
2894    
2895    ----------------------------------------------------------------------
2896    Name: Allen Leung
2897    Date: 2002/03/21 17:15:00 EST
2898    Tag: leunga-20020321-x86-fp-cfg
2899    Description:
2900    
2901      1. Recoded the buggy parts of x86-fp.
2902    
2903         a. All the block reordering code has been removed.
2904            We now depend on the block placement phases to do this work.
2905    
2906         b. Critical edge splitting code has been simplified and moved into the
2907            CFG modules, as where they belong.
2908    
2909         Both of these were quite buggy and complex.  The code is now much, much
2910         simpler.
2911    
2912      2. X86 backend.
2913    
2914         a. Added instructions for 64-bit support.  Instruction selection for
2915            64-bit has not been committed, however, since that
2916            requires changes to MLTREE which haven't been approved by
2917            Lal and John.
2918    
2919         b. Added support for FUCOMI and FUCOMIP when generating code for
2920            PentiumPro and above.  We only generate these instructions in
2921            the fast-fp mode.
2922    
2923         c. Added cases for JP and JNP in X86FreqProps.
2924    
2925      3. CFG
2926    
2927         CFG now has a bunch of methods for edge splitting and merging.
2928    
2929      4. Machine description.
2930    
2931         John's simplification of MLTREE_BASIS.fcond broke a few machine
2932         description things:
2933    
2934         rtl-build.{sig,sml} and hppa.mdl fixed.
2935    
2936         NOTE: the machine description stuff in the repository is still broken.
2937               Again, I can't put my fixes in because that involves
2938               changes to MLTREE.
2939    
2940    ----------------------------------------------------------------------
2941    Name: Matthias Blume
2942    Date: 2002/03/20 15:55:00 EST
2943    Tag: blume-20020320-kmp
2944    Description:
2945    
2946    Implemented Knuth-Morris-Pratt string matching in PreString and used
2947    it for String.isSubstring, Substring.isSubstring, and
2948    Substring.position.
2949    
2950    (Might need some stress-testing.  Simple examples worked fine.)
2951    
2952    ----------------------------------------------------------------------
2953    Name: Matthias Blume
2954    Date: 2002/03/19 16:37:00 EST
2955    Tag: blume-20020319-witnesses
2956    Description:
2957    
2958    Added a structure C.W and functions convert/Ptr.convert to ml-nlffi-lib.
2959    
2960    This implements a generic mechanism for changing constness qualifiers
2961    anywhere within big C types without resorting to outright "casts".
2962    (So far, functions such as C.rw/C.ro or C.Ptr.rw/C.Ptr.ro only let you
2963    modify the constness at the outermost level.)
2964    The implementation of "convert" is based on the idea of "witness"
2965    values -- values that are not used by the operation but whose types
2966    "testify" to their applicability.  On the implementation side, "convert"
2967    is simply a projection (returning its second curried argument).  With
2968    cross-module inlining, it should not result in any machine code being
2969    generated.
2970    
2971    ----------------------------------------------------------------------
2972    Name: Matthias Blume
2973    Date: 2002/03/15 16:40:00 EST
2974    Tag: blume-20020315-basis
2975    Description:
2976    
2977    Provided (preliminary?) implementations for
2978    
2979      {String,Substring}.{concatWith,isSuffix,isSubstring}
2980    
2981    and
2982    
2983      Substring.full
2984    
2985    Those are in the Basis spec but they were missing in SML/NJ.
2986    
2987    ----------------------------------------------------------------------
2988    Name: Matthias Blume
2989    Date: 2002/03/14 21:30:00 EST
2990    Tag: blume-20020314-controls
2991    Description:
2992    
2993    Controls:
2994    ---------
2995    
2996    1. Factored out the recently-added Controls : CONTROLS stuff and put
2997       it into its own library $/controls-lib.cm.  The source tree for
2998       this is under src/smlnj-lib/Controls.
2999    
3000    2. Changed the names of types and functions in this interface, so they
3001       make a bit more "sense":
3002    
3003          module -> registry
3004          'a registry -> 'a group
3005    
3006    3. The interface now deals in ref cells only.  The getter/setter interface
3007       is (mostly) gone.
3008    
3009    4. Added a function that lets one register an already-existing ref cell.
3010    
3011    5. Made the corresponding modifications to the rest of the code so that
3012       everything compiles again.
3013    
3014    6. Changed the implementation of Controls.MLRISC back to something closer
3015       to the original.  In particular, this module (and therefore MLRISC)
3016       does not depend on Controls.  There now is some link-time code in
3017       int-sys.sml that registers the MLRISC controls with the Controls
3018       module.
3019    
3020    CM:
3021    ---
3022    
3023      * One can now specify the lambda-split aggressiveness in init.cmi.
3024    
3025    ----------------------------------------------------------------------
3026    Name: Allen Leung
3027    Date: 2002/03/13 17:30:00 EST
3028    Tag: leunga-20020313-x86-fp-unary
3029    Description:
3030    
3031    Bug fix for:
3032    
3033    > leunga@weaselbane:~/Yale/tmp/sml-dist{21} bin/sml
3034    > Standard ML of New Jersey v110.39.1 [FLINT v1.5], March 08, 2002
3035    > - fun f(x,(y,z)) = Real.~ y;
3036    > [autoloading]
3037    > [autoloading done]
3038    >       fchsl   (%eax), 184(%esp)
3039    > Error: MLRisc bug: X86MCEmitter.emitInstr
3040    >
3041    > uncaught exception Error
3042    >   raised at: ../MLRISC/control/mlriscErrormsg.sml:16.14-16.19
3043    
3044    The problem was that the code generator did not generate any fp registers
3045    in this case, and the ra didn't know that it needed to run the X86FP phase to
3046    translate the pseudo fp instruction.   This only happened with unary fp
3047    operators in certain situations.
3048    
3049    ----------------------------------------------------------------------
3050    Name: Matthias Blume
3051    Date: 2002/03/13 14:00:00 EST
3052    Tag: blume-20020313-overload-etc
3053    Description:
3054    
3055    1. Added _overload as a synonym for overload for backward compatibility.
3056       (Control.overloadKW must be true for either version to be accepted.)
3057    
3058    2. Fixed bug in install script that caused more things to be installed
3059       than what was requested in config/targets.
3060    
3061    3. Made CM aware of the (_)overload construct so that autoloading
3062       works.
3063    
3064    ----------------------------------------------------------------------
3065    Name: Matthias Blume
3066    Date: 2002/03/12 22:03:00 EST
3067    Tag: blume-20020312-url
3068    Description:
3069    
3070    Forgot to update BOOT and srcarchiveurl.
3071    
3072    ----------------------------------------------------------------------
3073    Name: Matthias Blume
3074    Date: 2002/03/12 17:30:00 EST
3075    Tag: blume-20020312-version110392
3076    Description:
3077    
3078    Yet another version number bump (because of small changes to the
3079    binfile format).  Version number is now 110.39.2.  NEW BOOTFILES!
3080    
3081    Changes:
3082    
3083      The new pid generation scheme described a few weeks ago was overly
3084      complicated.  I implemented a new mechanism that is simpler and
3085      provides a bit more "stability":  Once CM has seen a compilation
3086      unit, it keeps its identity constant (as long as you do not delete
3087      those crucial CM/GUID/* files).  This means that when you change
3088      an interface, compile, then go back to the old interface, and
3089      compile again, you arrive at the original pid.
3090    
3091      There now also is a mechanism that instructs CM to use the plain
3092      environment hash as a module's pid (effectively making its GUID
3093      the empty string).  For this, "noguid" must be specified as an
3094      option to the .sml file in question within its .cm file.
3095      This is most useful for code that is being generated by tools such
3096      as ml-nlffigen (because during development programmers tend to
3097      erase the tool's entire output directory tree including CM's cached
3098      GUIDs).  "noguid" is somewhat dangerous (since it can be used to locally
3099      revert to the old, broken behavior of SML/NJ, but in specific cases
3100      where there is no danger of interface confusion, its use is ok
3101      (I think).
3102    
3103      ml-nlffigen by default generates "noguid" annotations.  They can be
3104      turned off by specifying -guid in its command line.
3105    
3106    ----------------------------------------------------------------------
3107    Name: Lal George
3108    Date: 2002/03/12 12 14:42:36 EST
3109    Tag: george-20020312-frequency-computation
3110    Description:
3111    
3112    Integrated jump chaining and static block frequency into the
3113    compiler. More details and numbers later.
3114    
3115    ----------------------------------------------------------------------
3116    Name: Lal George
3117    Date: 2002/03/11 11 22:38:53 EST
3118    Tag: george-20020311-jump-chain-elim
3119    Description:
3120    
3121    Tested the jump chain elimination on all architectures (except the
3122    hppa).  This is on by default right now and is profitable for the
3123    alpha and x86, however, it may not be profitable for the sparc and ppc
3124    when compiling the compiler.
3125    
3126    The gc test will typically jump to a label at the end of the cluster,
3127    where there is another jump to an external cluster containing the actual
3128    code to invoke gc. This is to allow factoring of common gc invocation
3129    sequences. That is to say, we generate:
3130    
3131            f:
3132               testgc
3133               ja   L1      % jump if above to L1
3134    
3135            L1:
3136               jmp L2
3137    
3138    
3139    After jump chain elimination the 'ja L1' instructions is converted to
3140    'ja L2'. On the sparc and ppc, many of the 'ja L2' instructions may end
3141    up being implemented in their long form (if L2 is far away) using:
3142    
3143            jbe     L3      % jump if below or equal to L3
3144            jmp     L2
3145         L3:
3146            ...
3147    
3148    
3149    For large compilation units L2  may be far away.
3150    
3151    
3152    ----------------------------------------------------------------------
3153    Name: Matthias Blume
3154    Date: 2002/03/11 13:30:00 EST
3155    Tag: blume-20020311-mltreeeval
3156    Description:
3157    
3158    A functor parameter was missing.
3159    
3160    ----------------------------------------------------------------------
3161    Name: Allen Leung
3162    Date: 2002/03/11 10:30:00 EST
3163    Tag: leunga-20020311-runtime-string0
3164    Description:
3165    
3166       The representation of the empty string now points to a
3167    legal null terminated C string instead of unit.  It is now possible
3168    to convert an ML string into C string with InlineT.CharVector.getData.
3169    This compiles into one single machine instruction.
3170    
3171    ----------------------------------------------------------------------
3172    Name: Allen Leung
3173    Date: 2002/03/10 23:55:00 EST
3174    Tag: leunga-20020310-x86-call
3175    Description:
3176    
3177       Added machine generation for CALL instruction (relative displacement mode)
3178    
3179    ----------------------------------------------------------------------
3180    Name: Matthias Blume
3181    Date: 2002/03/08 16:05:00
3182    Tag: blume-20020308-entrypoints
3183    Description:
3184    
3185    Version number bumped to 110.39.1.  NEW BOOTFILES!
3186    
3187    Entrypoints: non-zero offset into a code object where execution should begin.
3188    
3189    - Added the notion of an entrypoint to CodeObj.
3190    - Added reading/writing of entrypoint info to Binfile.
3191    - Made runtime system bootloader aware of entrypoints.
3192    - Use the address of the label of the first function given to mlriscGen
3193      as the entrypoint.  This address is currently always 0, but it will
3194      not be 0 once we turn on block placement.
3195    - Removed the linkage cluster code (which was The Other Way(tm) of dealing
3196      with entry points) from mlriscGen.
3197    
3198    ----------------------------------------------------------------------
3199    Name: Allen Leung
3200    Date: 2002/03/07 20:45:00 EST
3201    Tag: leunga-20020307-x86-cmov
3202    Description:
3203    
3204       Bug fixes for CMOVcc on x86.
3205    
3206       1. Added machine code generation for CMOVcc
3207       2. CMOVcc is now generated in preference over SETcc on PentiumPro or above.
3208       3. CMOVcc cannot have an immediate operand as argument.
3209    
3210    ----------------------------------------------------------------------
3211    Name: Matthias Blume
3212    Date: 2002/03/07 16:15:00 EST
3213    Tag: blume-20020307-controls
3214    Description:
3215    
3216    This is a very large but mostly boring patch which makes (almost)
3217    every tuneable compiler knob (i.e., pretty much everything under
3218    Control.* plus a few other things) configurable via both the command
3219    line and environment variables in the style CM did its configuration
3220    until now.
3221    
3222    Try starting sml with '-h' (or, if you are brave, '-H')
3223    
3224    To this end, I added a structure Controls : CONTROLS to smlnj-lib.cm which
3225    implements the underlying generic mechanism.
3226    
3227    The interface to some of the existing such facilities has changed somewhat.
3228    For example, the MLRiscControl module now provides mkFoo instead of getFoo.
3229    (The getFoo interface is still there for backward-compatibility, but its
3230    use is deprecated.)
3231    
3232    The ml-build script passes -Cxxx=yyy command-line arguments through so
3233    that one can now twiddle the compiler settings when using this "batch"
3234    compiler.
3235    
3236    TODO items:
3237    
3238    We should go through and throw out all controls that are no longer
3239    connected to anything.  Moreover, we should go through and provide
3240    meaningful (and correct!) documentation strings for those controls
3241    that still are connected.
3242    
3243    Currently, multiple calls to Controls.new are accepted (only the first
3244    has any effect).  Eventually we should make sure that every control
3245    is being made (via Controls.new) exactly once.  Future access can then
3246    be done using Controls.acc.
3247    
3248    Finally, it would probably be a good idea to use the getter-setter
3249    interface to controls rather than ref cells.  For the time being, both
3250    styles are provided by the Controls module, but getter-setter pairs are
3251    better if thread-safety is of any concern because they can be wrapped.
3252    
3253    *****************************************
3254    
3255    One bug fix: The function blockPlacement in three of the MLRISC
3256    backpatch files used to be hard-wired to one of two possibilities at
3257    link time (according to the value of the placementFlag).  But (I
3258    think) it should rather sense the flag every time.
3259    
3260    *****************************************
3261    
3262    Other assorted changes (by other people who did not supply a HISTORY entry):
3263    
3264    1. the cross-module inliner now works much better (Monnier)
3265    2. representation of weights, frequencies, and probabilities in MLRISC
3266       changed in preparation of using those for weighted block placement
3267       (Reppy, George)
3268    
3269    ----------------------------------------------------------------------
3270    Name: Lal George
3271    Date: 2002/03/07 14:44:24 EST 2002
3272    Tag: george-20020307-weighted-block-placement
3273    
3274    Tested the weighted block placement optimization on all architectures
3275    (except the hppa) using AMPL to generate the block and edge frequencies.
3276    Changes were required in the machine properties to correctly
3277    categorize trap instructions. There is an MLRISC flag
3278    "weighted-block-placement" that can be used to enable weighted block
3279    placement, but this will be ineffective without block/edge
3280    frequencies (coming soon).
3281    
3282    
3283    ----------------------------------------------------------------------
3284    Name: Lal George
3285    Date: 2002/03/05 17:24:48 EST
3286    Tag: george-20020305-linkage-cluster
3287    
3288    In order to support the block placement optimization, a new cluster
3289    is generated as the very first cluster (called the linkage cluster).
3290    It contains a single jump to the 'real' entry point for the compilation
3291    unit. Block placement has no effect on the linkage cluster itself, but
3292    all the other clusters  have full freedom in the manner in which they
3293    reorder blocks or functions.
3294    
3295    On the x86 the typical linkage code that is generated is:
3296       ----------------------
3297            .align 2
3298       L0:
3299            addl    $L1-L0, 72(%esp)
3300            jmp     L1
3301    
3302    
3303            .align  2
3304       L1:
3305       ----------------------
3306    
3307    72(%esp) is the memory location for the stdlink register. This
3308    must contain the address of the CPS function being called. In the
3309    above example, it contains the address of  L0; before
3310    calling L1 (the real entry point for the compilation unit), it
3311    must contain the address for L1, and hence
3312    
3313            addl $L1-L0, 72(%esp)
3314    
3315    I have tested this on all architectures except the hppa.The increase
3316    in code size is of course negligible
3317    
3318    ----------------------------------------------------------------------
3319    Name: Allen Leung
3320    Date: 2002/03/03 13:20:00 EST
3321    Tag: leunga-20020303-mlrisc-tools
3322    
3323      Added #[ ... ] expressions to mlrisc tools
3324    
3325    ----------------------------------------------------------------------
3326    Name: Matthias Blume
3327    Date: 2002/02/27 12:29:00 EST
3328    Tag: blume-20020227-cdebug
3329    Description:
3330    
3331    - made types in structure C and C_Debug to be equal
3332    - got rid of code duplication (c-int.sml vs. c-int-debug.sml)
3333    - there no longer is a C_Int_Debug (C_Debug is directly derived from C)
3334    
3335    ----------------------------------------------------------------------
3336    Name: Matthias Blume
3337    Date: 2002/02/26 12:00:00 EST
3338    Tag: blume-20020226-ffi
3339    Description:
3340    
3341    1. Fixed a minor bug in CM's "noweb" tool:
3342       If numbering is turned off, then truly don't number (i.e., do not
3343       supply the -L option to noweb).  The previous behavior was to supply
3344       -L'' -- which caused noweb to use the "default" line numbering scheme.
3345       Thanks to Chris Richards for pointing this out (and supplying the fix).
3346    
3347    2. Once again, I reworked some aspects of the FFI:
3348    
3349       A. The incomplete/complete type business:
3350    
3351       - Signatures POINTER_TO_INCOMPLETE_TYPE and accompanying functors are
3352         gone!
3353       - ML types representing an incomplete type are now *equal* to
3354         ML types representing their corresponding complete types (just like
3355         in C).  This is still safe because ml-nlffigen will not generate
3356         RTTI for incomplete types, nor will it generate functions that
3357         require access to such RTTI.   But when ML code generated from both
3358         incomplete and complete versions of the C type meet, the ML types
3359         are trivially interoperable.
3360    
3361         NOTE:  These changes restore the full generality of the translation
3362         (which was previously lost when I eliminated functorization)!
3363    
3364       B. Enum types:
3365    
3366       - Structure C now has a type constructor "enum" that is similar to
3367         how the "su" constructor works.  However, "enum" is not a phantom
3368         type because each "T enum" has values (and is isomorphic to
3369         MLRep.Signed.int).
3370       - There are generic access operations for enum objects (using
3371         MLRep.Signed.int).
3372       - ml-nlffigen will generate a structure E_foo for each "enum foo".
3373         * The structure contains the definition of type "mlrep" (the ML-side
3374         representation type of the enum).  Normally, mlrep is the same
3375         as "MLRep.Signed.int", but if ml-nlffigen was invoked with "-ec",
3376         then mlrep will be defined as a datatype -- thus facilitating
3377         pattern matching on mlrep values.
3378         ("-ec" will be suppressed if there are duplicate values in an
3379          enumeration.)
3380         * Constructors ("-ec") or values (no "-ec") e_xxx of type mlrep
3381         will be generated for each C enum constant xxx.
3382         * Conversion functions m2i and i2m convert between mlrep and
3383         MLRep.Signed.int.  (Without "-ec", these functions are identities.)
3384         * Coversion functions c and ml convert between mlrep and "tag enum".
3385         * Access functions (get/set) fetch and store mlrep values.
3386       - By default (unless ml-nlffigen was invoked with "-nocollect"), unnamed
3387         enumerations are merged into one single enumeration represented by
3388         structure E_'.
3389    
3390    ----------------------------------------------------------------------
3391    Name: Allen Leung
3392    Date: 2002/02/25 04:45:00 EST
3393    Tag: leunga-20020225-cps-spill
3394    
3395    This is a new implementation of the CPS spill phase.
3396    The new phase is in the new file compiler/CodeGen/cpscompile/spill-new.sml
3397    In case of problems, replace it with the old file spill.sml
3398    
3399    The current compiler runs into some serious performance problems when
3400    constructing a large record.  This can happen when we try to compile a
3401    structure with many items.  Even a very simple structure like the following
3402    makes the compiler slow down.
3403    
3404        structure Foo = struct
3405           val x_1 = 0w1 : Word32.int
3406           val x_2 = 0w2 : Word32.int
3407           val x_3 = 0w3 : Word32.int
3408           ...
3409           val x_N = 0wN : Word32.int
3410        end
3411    
3412    The following table shows the compile time, from N=1000 to N=4000,
3413    with the old compiler:
3414    
3415    N
3416    1000   CPS 100 spill                           0.04u  0.00s  0.00g
3417           MLRISC ra                               0.06u  0.00s  0.05g
3418              (spills = 0 reloads = 0)
3419           TOTAL                                   0.63u  0.07s  0.21g
3420    
3421    1100   CPS 100 spill                           8.25u  0.32s  0.64g
3422           MLRISC ra                               5.68u  0.59s  3.93g
3423              (spills = 0 reloads = 0)
3424           TOTAL                                   14.71u  0.99s  4.81g
3425    
3426    1500   CPS 100 spill                           58.55u  2.34s  1.74g
3427           MLRISC ra                               5.54u  0.65s  3.91g
3428              (spills = 543 reloads = 1082)
3429           TOTAL                                   65.40u  3.13s  6.00g
3430    
3431    2000   CPS 100 spill                           126.69u  4.84s  3.08g
3432           MLRISC ra                               0.80u  0.10s  0.55g
3433              (spills = 42 reloads = 84)
3434           TOTAL                                   129.42u  5.10s  4.13g
3435    
3436    3000   CPS 100 spill                           675.59u  19.03s  11.64g
3437           MLRISC ra                               2.69u  0.27s  1.38g
3438              (spills = 62 reloads = 124)
3439           TOTAL                                   682.48u  19.61s  13.99g
3440    
3441    4000   CPS 100 spill                           2362.82u  56.28s  43.60g
3442           MLRISC ra                               4.96u  0.27s  2.72g
3443              (spills = 85 reloads = 170)
3444           TOTAL                                   2375.26u  57.21s  48.00g
3445    
3446    As you can see the old cps spill module suffers from some serious
3447    performance problem.  But since I cannot decipher the old code fully,
3448    instead of patching the problems up, I'm reimplementing it
3449    with a different algorithm.  The new code is more modular,
3450    smaller when compiled, and substantially faster
3451    (O(n log n) time and O(n) space).  Timing of the new spill module:
3452    
3453    4000  CPS 100 spill                           0.02u  0.00s  0.00g
3454          MLRISC ra                               0.25u  0.02s  0.15g
3455             (spills=1 reloads=3)
3456          TOTAL                                   7.74u  0.34s  1.62g
3457    
3458    Implementation details:
3459    
3460    As far as I can tell, the purpose of the CPS spill module is to make sure the
3461    number of live variables at any program point (the bandwidth)
3462    does not exceed a certain limit, which is determined by the
3463    size of the spill area.
3464    
3465    When the bandwidth is too large, we decrease the register pressure by
3466    packing live variables into spill records.  How we achieve this is
3467    completely different than what we did in the old code.
3468    
3469    First, there is something about the MLRiscGen code generator
3470    that we should be aware of:
3471    
3472    o MLRiscGen performs code motion!
3473    
3474       In particular, it will move floating point computations and
3475       address computations involving only the heap pointer to
3476       their use sites (if there is only a single use).
3477       What this means is that if we have a CPS record construction
3478       statement
3479    
3480           RECORD(k,vl,w,e)
3481    
3482       we should never count the new record address w as live if w
3483       has only one use (which is often the case).
3484    
3485       We should do something similar to floating point, but the transformation
3486       there is much more complex, so I won't deal with that.
3487    
3488    Secondly, there are now two new cps primops at our disposal:
3489    
3490     1. rawrecord of record_kind option
3491        This pure operator allocates some uninitialized storage from the heap.
3492        There are two forms:
3493    
3494         rawrecord NONE [INT n]  allocates a tagless record of length n
3495         rawrecord (SOME rk) [INT n] allocates a tagged record of length n
3496                                     and initializes the tag.
3497    
3498     2. rawupdate of cty
3499          rawupdate cty (v,i,x)
3500          Assigns to x to the ith component of record v.
3501          The storelist is not updated.
3502    
3503    We use these new primops for both spilling and increment record construction.
3504    
3505     1. Spilling.
3506    
3507        This is implemented with a linear scan algorithm (but generalized
3508        to trees).  The algorithm will create a single spill record at the
3509        beginning of the cps function and use rawupdate to spill to it,
3510        and SELECT or SELp to reload from it.  So both spills and reloads
3511        are fine-grain operations.  In contrast, in the old algorithm
3512        "spills" have to be bundled together in records.
3513    
3514        Ideally, we should sink the spill record construction to where
3515        it is needed.  We can even split the spill record into multiple ones
3516        at the places where they are needed.  But CPS is not a good
3517        representation for global code motion, so I'll keep it simple and
3518        am not attempting this.
3519    
3520     2. Incremental record construction (aka record splitting).
3521    
3522        Long records with many component values which are simulatenously live
3523        (recall that single use record addresses are not considered to
3524         be live) are constructed with rawrecord and rawupdate.
3525        We allocate space on the heap with rawrecord first, then gradually
3526        fill it in with rawupdate.  This is the technique suggested to me
3527        by Matthias.
3528    
3529        Some restrictions on when this is applicable:
3530        1. It is not a VECTOR record.  The code generator currently does not handle
3531           this case. VECTOR record uses double indirection like arrays.
3532        2. All the record component values are defined in the same "basic block"
3533           as the record constructor.  This is to prevent speculative
3534           record construction.
3535    
3536    ----------------------------------------------------------------------
3537    Name: Allen Leung
3538    Date: 2002/02/22 01:02:00 EST
3539    Tag: leunga-20020222-mlrisc-tools
3540    
3541    Minor bug fixes in the parser and rewriter
3542    
3543    ----------------------------------------------------------------------
3544    Name: Allen Leung
3545    Date: 2002/02/21 20:20:00 EST
3546    Tag: leunga-20020221-peephole
3547    
3548    Regenerated the peephole files.  Some contained typos in the specification
3549    and some didn't compile because of pretty printing bugs in the old version
3550    of 'nowhere'.
3551    
3552    ----------------------------------------------------------------------
3553    Name: Allen Leung
3554    Date: 2002/02/19 20:20:00 EST
3555    Tag: leunga-20020219-mlrisc-tools
3556    Description:
3557    
3558       Minor bug fixes to the mlrisc-tools library:
3559    
3560       1.  Fixed up parsing colon suffixed keywords
3561       2.  Added the ability to shut the error messages up
3562       3.  Reimplemented the pretty printer and fixed up/improved
3563           the pretty printing of handle and -> types.
3564       4.  Fixed up generation of literal symbols in the nowhere tool.
3565       5.  Added some SML keywords to to sml.sty
3566    
3567    ----------------------------------------------------------------------
3568    Name: Matthias Blume
3569    Date: 2002/02/19 16:20:00 EST
3570    Tag: blume-20020219-cmffi
3571    Description:
3572    
3573    A wild mix of changes, some minor, some major:
3574    
3575    * All C FFI-related libraries are now anchored under $c:
3576        $/c.cm      --> $c/c.cm
3577        $/c-int.cm  --> $c/internals/c-int.cm
3578        $/memory.cm --> $c/memory/memory.cm
3579    
3580    * "make" tool (in CM) now treats its argument pathname slightly
3581      differently:
3582        1. If the native expansion is an absolute name, then before invoking
3583           the "make" command on it, CM will apply OS.Path.mkRelative
3584           (with relativeTo = OS.FileSys.getDir()) to it.
3585        2. The argument will be passed through to subsequent phases of CM
3586           processing without "going native".  In particular, if the argument
3587           was an anchored path, then "make" will not lose track of that anchor.
3588    
3589    * Compiler backends now "know" their respective C calling conventions
3590      instead of having to be told about it by ml-nlffigen.  This relieves
3591      ml-nlffigen from one of its burdens.
3592    
3593    * The X86Backend has been split into X86CCallBackend and X86StdCallBackend.
3594    
3595    * Export C_DEBUG and C_Debug from $c/c.cm.
3596    
3597    * C type encoding in ml-nlffi-lib has been improved to model the conceptual
3598      subtyping relationship between incomplete pointers and their complete
3599      counterparts.  For this, ('t, 'c) ptr has been changed to 'o ptr --
3600      with the convention of instantiating 'o with ('t, 'c) obj whenever
3601      the pointer target type is complete.  In the incomplete case, 'o
3602      will be instantiated with some "'c iobj" -- a type obtained by
3603      using one of the functors PointerToIncompleteType or PointerToCompleteType.
3604    
3605      Operations that work on both incomplete and complete pointer types are
3606      typed as taking an 'o ptr while operations that require the target to
3607      be known are typed as taking some ('t, 'c) obj ptr.
3608    
3609      voidptr is now a bit "more concrete", namely "type voidptr = void ptr'"
3610      where void is an eqtype without any values.  This makes it possible
3611      to work on voidptr values using functions meant to operate on light
3612      incomplete pointers.
3613    
3614    * As a result of the above, signature POINTER_TO_INCOMPLETE_TYPE has
3615      been vastly simplified.
3616    
3617    ----------------------------------------------------------------------
3618    Name: Matthias Blume
3619    Date: 2002/02/19 10:48:00 EST
3620    Tag: blume-20020219-pqfix
3621    Description:
3622    
3623    Applied Chris Okasaki's bug fix for priority queues.
3624    
3625    ----------------------------------------------------------------------
3626    Name: Matthias Blume
3627    Date: 2002/02/15 17:05:00
3628    Tag: Release_110_39
3629    Description:
3630    
3631    Last-minute retagging is becoming a tradition... :-(
3632    
3633    This is the working release 110.39.
3634    
3635    ----------------------------------------------------------------------
3636    Name: Matthias Blume
3637    Date: 2002/02/15 16:00:00 EST
3638    Tag: Release_110_39-orig
3639    Description:
3640    
3641    Working release 110.39.  New bootfiles.
3642    
3643    (Update: There was a small bug in the installer so it wouldn't work
3644    with all shells.  So I retagged. -Matthias)
3645    
3646    ----------------------------------------------------------------------
3647    Name: Matthias Blume
3648    Date: 2002/02/15 14:17:00 EST
3649    Tag: blume-20020215-showbindings
3650    Description:
3651    
3652    Added EnvRef.listBoundSymbols and CM.State.showBindings.  Especially
3653    the latter can be useful for exploring what bindings are available at
3654    the interactive prompt.  (The first function returns only the list
3655    of symbols that are really bound, the second prints those but also the
3656    ones that CM's autoloading mechanism knows about.)
3657    
3658    ----------------------------------------------------------------------
3659    Name: Matthias Blume
3660    Date: 2002/02/15 12:08:00 EST
3661    Tag: blume-20020215-iptrs
3662    Description:
3663    
3664    Two improvements to ml-nlffigen:
3665    
3666      1. Write files only if they do not exist or if their current contents
3667         do not coincide with what's being written.  (That is, avoid messing
3668         with the time stamps unless absolutely necessary.)
3669    
3670      2. Implement a "repository" mechanism for generated files related
3671         to "incomplete pointer types".   See the README file for details.
3672    
3673    ----------------------------------------------------------------------
3674    Name: Matthias Blume
3675    Date: 2002/02/14 11:50:00 EST
3676    Tag: blume-20020214-quote
3677    Description:
3678    
3679    Added a type 't t_' to tag.sml (in ml-nlffi-lib.cm).  This is required
3680    because of the new and improved tag generation scheme.  (Thanks to Allen
3681    Leung for pointing it out.)
3682    
3683    ----------------------------------------------------------------------
3684    Name: Lal George
3685    Date: 2002/02/14 09:55:27 EST 2002
3686    Tag: george-20020214-isabelle-bug
3687    Description:
3688    
3689    Fixed the MLRISC bug sent by Markus Wenzel regarding the compilation
3690    of Isabelle on the x86.
3691    
3692    From Allen:
3693    -----------
3694     I've found the problem:
3695    
3696         in ra-core.sml, I use the counter "blocked" to keep track of the
3697         true number of elements in the freeze queue.  When the counter goes
3698         to zero, I skip examining the queue.  But I've messed up the
3699         bookkeeping in combine():
3700    
3701             else ();
3702             case !ucol of
3703               PSEUDO => (if !cntv > 0 then
3704                     (if !cntu > 0 then blocked := !blocked - 1 else ();
3705                                        ^^^^^^^^^^^^^^^^^^^^^^^
3706                      moveu := mergeMoveList(!movev, !moveu)
3707                     )
3708                  else ();
3709    
3710         combine() is called to coalesce two nodes u and v.
3711         I think I was thinking that if the move counts of u and v are both
3712         greater than zero then after they are coalesced then one node is
3713         removed from the freeze queue.  Apparently I was thinking that
3714         both u and v are of low degree, but that's clearly not necessarily true.
3715    
3716    
3717    02/12/2002:
3718        Here's the patch.  HOL now compiles.
3719    
3720        I don't know how this impact on performance (compile
3721        time or runtime).  This bug caused the RA (especially on the x86)
3722        to go thru the potential spill phase when there are still nodes on the
3723        freeze queue.
3724    
3725    
3726    
3727    
3728    ----------------------------------------------------------------------
3729    Name: Matthias Blume
3730    Date: 2002/02/13 22:40:00 EST
3731    Tag: blume-20020213-fptr-rtti
3732    Description:
3733    
3734    Fixed a bug in ml-nlffigen that was introduced with one of the previous
3735    updates.
3736    
3737    ----------------------------------------------------------------------
3738    Name: Matthias Blume
3739    Date: 2002/02/13 16:41:00 EST
3740    Tag: blume-20020213-cmlpq
3741    Description:
3742    
3743    Added new priority queue export symbols (which have just been added to
3744    smlnj-lib.cm) to CML's version of smlnj-lib.cm.  (Otherwise CML would
3745    not compile and the installer would choke.)
3746    
3747    ----------------------------------------------------------------------
3748    Name: Matthias Blume
3749    Date: 2002/02/13 16:15:00 EST
3750    Tag: blume-20020213-various
3751    Description:
3752    
3753    1. More tweaks to ml-nlffigen:
3754    
3755       - better internal datastructures (resulting in slight speedup)
3756       - "-match" option requires exact match
3757       - "localized" gensym counters (untagged structs/unions nested within
3758         other structs/unions or within typedefs get a fresh counter; their
3759         tag will be prefixed by a concatenation of their parents' tags)
3760       - bug fixes (related to calculation of transitive closure of types
3761         to be included in the output)
3762    
3763    2. Minor Basis updates:
3764    
3765       - added implementations for List.collate and Option.app
3766    
3767    ----------------------------------------------------------------------
3768    Name: Matthias Blume
3769    Date: 2002/02/11 15:55:00 EST
3770    Tag: blume-20020211-gensym
3771    Description:
3772    
3773    Added a "-gensym" option to command line of ml-nlffigen.  This can be
3774    used to specify a "stem" -- a string that is inserted in all "gensym'd"
3775    names (ML structure names that correspond to unnamed C structs, unions,
3776    and enums), so that separate runs of ml-nlffigen do not clash.
3777    
3778    ----------------------------------------------------------------------
3779    Name: Matthias Blume
3780    Date: 2002/02/11 12:05:00 EST
3781    Tag: blume-20020211-gensml
3782    Description:
3783    
3784    A quick fix for a problem with GenSML (in the pgraph-util library):
3785    Make generation of toplevel "local" optional.  (Strictly speaking,
3786    signature definitions within "local" are not legal SML.)
3787    
3788    Other than that: updates to INSTALL and cm/TODO.
3789    
3790    ----------------------------------------------------------------------
3791    Name: Matthias Blume
3792    Date: 2002/02/08 15:00:00 EST
3793    Tag: blume-20020208-uniquepid
3794    Description:
3795    
3796    0. Version number has been bumped to 110.38.1.  NEW BOOTFILES!!!
3797    
3798    1. The installer (config/install.sh) has gotten smarter:
3799    
3800         - Configuration options are a bit easier to specify now
3801           (in config/targets).
3802         - Bug in recognizing .tar.bz2 files fixed.
3803         - Installer automatically resolves dependencies between
3804           configuration options (e.g., if you ask for eXene, you will
3805           also get cml -- regardless whether you asked for it or not).
3806         - Installer can run in "quieter mode" by setting the environment
3807           variable INSTALL_QUIETLY to "true".  "Quieter" does not mean
3808           "completely silent", though.
3809         - Build HashCons library as part of smlnj-lib.
3810    
3811    2. A new scheme for assigning persistent identifiers to compilation
3812       units (and, by extension, to types etc.) has been put into place.
3813       This fixes a long-standing bug where types and even dynamic values
3814       can get internally confused, thereby compromising type safety
3815       (abstraction) and dynamic correctness.  See
3816    
3817         http://cm.bell-labs.com/cm/cs/who/blume/pid-confusion.tgz
3818    
3819       for an example of how things could go wrong until now.
3820    
3821       The downside of the new scheme is that pids are not quite as
3822       persistent as they used to be: CM will generate a fresh pid
3823       for every compilation unit that it thinks it sees for the first
3824       time.  That means that if you compile starting from a clean, fresh
3825       source tree at two different times, you end up with different
3826       binaries.
3827    
3828       Cutoff recompilation, however, has not been compromised because
3829       CM keeps pid information in special caches between runs.
3830    
3831    ----------------------------------------------------------------------
3832    Name: Lal George
3833    Date: 2002/02/07 15:34:13 EST 2002
3834    Tag: <none>
3835    Description:
3836    
3837    Compilers that generate assembly code may produce  global labels
3838    whose value is resolved at link time. The various peephole optimization
3839    modules did not take this in account.
3840    
3841    TODO. The Labels.addrOf function should really return an option
3842    type so that clients are forced to deal with this issue, rather
3843    than an exception being raised.
3844    
3845    ----------------------------------------------------------------------
3846    Name: Lal George
3847    Date: 2002/02/06 13:55:02 EST
3848    Tag: george-20020206-ra-breakup
3849    Description:
3850    
3851    1. A bug fix from Allen.
3852    
3853        A typo causes extra fstp %st(0)'s to be generated at compensation
3854        edges, which might cause stack underflow traps at runtime.  This
3855        occurs in fft where there are extraneous fstps right before the 'into'
3856        trap instruction (in this case they are harmless since none of the